Isolation and Characterization of a Novel Pituitary ... - Semantic Scholar

0888-8809/04/$15.00/0 Printed in U.S.A.

Molecular Endocrinology 18(7):1827–1839 Copyright © 2004 by The Endocrine Society doi: 10.1210/me.2004-0087

Isolation and Characterization of a Novel Pituitary Tumor Apoptosis Gene ADIL BAHAR, DAVID J. SIMPSON, STEVE J. CUTTY, JOHN E. BICKNELL, PAUL R. HOBAN, SARAH HOLLEY, MIRNA MOURTADA-MAARABOUNI, GWYN T. WILLIAMS, RICHARD N. CLAYTON, AND WILLIAM E. FARRELL Institute for Science and Technology in Medicine (A.B., D.J.S., S.J.C., J.E.B., P.R.H., S.H., R.N.C., W.E.F.) Medical Research Unit, Keele University, North Staffordshire Hospital, Stoke-on-Trent ST4 7QB, United Kingdom; and School of Life Sciences (M.M.-M., G.T.W.), Keele University, Keele ST5 5BG, United Kingdom To determine mechanisms for pituitary neoplasia we used methylation-sensitive arbitrarily primedPCR to isolate novel genes that are differentially methylated relative to normal pituitary. We report the isolation of a novel differentially methylated chromosome 22 CpG island-associated gene (C22orf3). Sodium bisulfite sequencing of pooled tumor cohorts, used in the isolation of this gene, showed that only a proportion of the adenomas within the pools were methylated; however, expression analysis by quantitative RT-PCR of individual adenoma irrespective of subtype showed the majority (30 of 38; 79%) failed to express this gene relative to normal pituitary. Sodium bisulfite sequencing of individual adenomas showed that 6 of 30 (20%) that failed to express pituitary tumor apoptosis gene (PTAG) were methylated; however, genetic change as determined by loss of heterozygosity and sequence analysis was not apparent in the remaining tumors that failed to express this gene. In those cases where the CpG island of these genes was methylated it was invariably associated with loss of transcript expression. Enforced expression of C22orf3 in AtT20 cells had no measur-

able effects on cell proliferation or viability; however, in response to bromocriptine challenge (10–40 ␮M) cells expressing this gene showed a significantly augmented apoptotic response as determined by both acridine orange staining and TUNEL labeling. The apoptotic response to bromocriptine challenge was inhibited in coincubation experiments with the general caspase inhibitor zVAD-fmk. In addition, in time course experiments, direct measurement of active caspases by fluorochrome-labeled inhibition of caspases, showed an augmented increase (⬃2.4 fold) in active caspases in response to bromocriptine challenge in cells expressing C22orf3 relative to those harboring an empty vector control. The pituitary tumor derivation and its role in apoptosis of this gene led us to assign the acronym PTAG to this gene and its protein product. The ability of cells, showing reduced expression of PTAG, to evade or show a blunted apoptotic response may underlie oncogenic transformation in both the pituitary and other tumor types. (Molecular Endocrinology 18: 1827–1839, 2004)

P

display and cDNA representation difference analysis, the pituitary tumor transforming gene, GADD45␥, and MEG3, have been identified as being inappropriately expressed in pituitary tumors (10–12). Overexpression of pituitary tumor transforming gene and loss of expression of the growth-regulatory genes, GADD45␥ and MEG3, are found in the majority of pituitary tumors; however, mechanisms directly responsible for their dysregulation are yet to be defined (11–14). Recent studies have highlighted epigenetic mechanisms associated with gene silencing. Inappropriate methylation of CpG islands of key cell cycle control and growth-regulatory genes has been demonstrated for pituitary tumors that include CDKN2A/ p16 (15–19), RB1 (20), DAP kinase (death associated protein kinase) (21) and GADD45␥ (22), which is associated with gene silencing. Furthermore, the association of methylation with gene silencing provides a detectable epigenetic mark for the isolation

ITUITARY ADENOMAS ARE mostly monoclonal benign neoplasms accounting for 10–15% of all diagnosed adult intracranial tumors (1–3). One third of pituitary tumors will show invasive growth characteristics, and an even smaller proportion (⬍1%) harbor the potential to develop into malignant carcinomas and metastasize beyond the central nervous system (4). With the exception of the gsp oncogene, mutations of known protooncogenes and tumor suppressor genes occur infrequently (5–9). Employing differential Abbreviations: FLICA, Fluorochrome-labeled inhibition of caspases; IPTG, isopropyl-␤-D-thiogalactopyranoside; LOH, loss of heterozygosity; MsAP-PCR, methylation-sensitive arbitrarily primed PCR; PBGD, porphobilinogen deaminase; PTAG, pituitary tumor apoptosis gene; qRT-PCR, quantitative RT-PCR. Molecular Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.

1827

1828 Mol Endocrinol, July 2004, 18(7):1827–1839

and characterization of novel genes implicated in tumorigenesis (23). Several techniques have been described that characterize DNA methylation status including methylation-sensitive PCR (24), sodium bisulfite sequencing (25), and methylation-sensitive Southern blotting (26). A limitation of these techniques is that they require some prior knowledge of the DNA sequence being studied. However, other techniques have been described that identify novel CpG islands that are differentially methylated in cancer; these techniques include restriction landmark genomic scanning (27), methylation-sensitive representational difference analysis (28), methylation-sensitive arbitrarily primed PCR (MsAPPCR) (29–31), CpG island microarrays (32, 33), and methyl-CpG binding domain chromatography (34). In this study we used MsAP-PCR to isolate and identify CpG islands that are differentially methylated in pituitary tumors relative to normal pituitary tissue. Because gene-associated CpG islands frequently encompass or extend into transcribed regions, subsequent sequence analysis allowed us to identify and characterize their role in this tumor type. In this study, we report the isolation and characterization of a novel chromosome 22 CpG island-associated gene (C22orf3) that is not expressed in a significant proportion of pituitary tumors. Functional characterization of this gene, through induced expression in the pituitary tumor cell line AtT20, shows it to be a novel pituitary-derived proapoptotic gene, which we have termed pituitary tumor apoptosis gene (PTAG).

RESULTS MsAP-PCR Identifies CpG Islands We used MsAP-PCR to isolate and identify novel CpG islands that are differentially methylated in pituitary tumors relative to normal pituitaries. Figure 1 shows representative examples of differential PCR amplicons generated by MsAP-PCR in separate pooled samples of pituitary adenomas relative to postmortem normal pituitaries. In some cases, differential amplicons were generated that were specific to a particular pituitary tumor subtype (Fig. 1A), whereas others were common to both nonfunctional adenomas and somatotrophinomas (Fig. 1B). A total of 42 differential PCR products were isolated and subject to DNA sequence analysis, which showed that 30 conformed to the minimal criteria that define CpG islands (ⱖ200 bp, GC content ⬎50%, CpG/GpC ⬎0.5) (35). DNA sequence homology analysis of the CpG islands using the BLAST program revealed that nine were localized to the 5⬘-regions of known or novel genes and extended into their coding region. The size of each PCR amplicon, percentage of GC nucleotides, CpG/GpC ratio, and sequence homology results are summarized in Table 1. Of the nine transcript-associated CpG islands, four were isolated from the pool of

Bahar et al. • Isolation and Characterization of PTAG

Fig. 1. Representative Examples of MsAP-PCRs Performed with Undigested Normal Pituitary DNA (NP), Restricted Normal Pituitary DNA (NP⫹), or Pooled Restricted DNA from Somatotrophinomas (SOM⫹) and Nonfunctional Adenomas (NF⫹) PCRs were performed with either single arbitrary primers or a combination of two different primers, shown above each gel. Putative differentially methylated sequences were defined as amplicons present in both digested pooled tumor samples and undigested normal pituitary but absent in digested normal pituitary DNA (indicated with arrows). Differential products were identified as specific to only one subtype (A) or as common amplicons between tumor subtypes (B).

nonfunctional tumors, three were isolated from the somatotrophinoma pool, and two were derived from both tumor subtypes. Identification of Differentially Methylated CpG Islands Bisulfite sequencing was performed to determine the methylation status of the CpG islands isolated by MsAP-PCR. For each of the nine CpG islands, five molecules were subject to bisulfite sequencing from each of the two separate pools of tumors and normal pituitaries. Five CpG islands (1, 2, 4, 5, and 7; Table 1) were unmethylated in both pools of tumors and normal pituitaries, whereas four CpG islands (3, 6, 8 and 9; Table 1) showed tumor-specific methylation relative to normal pituitaries. To favor detection of frequent epigenetic change, sodium bisulfite sequencing was restricted to five molecules across each of the islands analyzed; however, we are aware that this does not represent a saturated analysis of individual islands. Thus, sequencing an increase number of molecules from each adenoma pool may well identify islands methylated as an infrequent event. Table 1 shows the number of methylated molecules (of five sequenced) in the somatotrophinoma and nonfunctional pools of tumors for each of the nine CpG islands. The CpG island


Mol Endocrinol, July 2004, 18(7):1827–1839

1829

Table 1. Summary of Nine CpG Islands Isolated by MsAP-PCR Sequence

Sizea

Tumor Typeb

% GCc

CpG/ GpCc

1 2

905 523

NF NF

72.63 71.86

0.85 0.88

3 4 5

382 764 419

NF Som/NF Som/NF

68.1 68.69 61.97

0.69 0.66 0.73

6

287

Som

73.6

0.72

7

273

Som

68.3

0.69

8 9

322 716

Som NF

71.5 69.16

0.7 0.93

Blast Homologyd

Transcription factor RELB Protein-tyrosine phosphatase MU precursor Vesicle transport-related protein Novel transcript Proline and glutamic acid rich nuclear protein Inter-␣-trypsin inhibitor heavy chain H2 Peptide deformylase mitochondrial precursor Novel transcript Protein C22 ORF3

Chromosome Mapd

Ensembl Accession No.d

Methylated Moleculese Som

NF

19q13.32 18p11.23

ENSG00000104856 ENSG00000168519

0 0

0 0

14q12 7q34 17p13.3

ENSG00000092108 ENSG00000006459 ENSG00000141456

1 0 0

0 0 0

10p14

ENSG00000151655

0

1

16q22.1

ENSG00000157312

0

0

11q12.3 22q12.2

ENSG00000149481 ENSG00000100263

1 2

1 2

For clarity, information relating to sequence 9 is shown in bold. a Size of the PCR amplicons in base pairs. b Subtype of pituitary tumor from which MsAP-PCR products were isolated: either from a single subtype or common to both (Som, somatotrophinomas; NF, nonfunctionals). c CpG islands were identified based on criteria described previously (38); minimum length, 200 bp; GC content, greater than 50%; CpG/CpG more than 0.5. d BLAST homology analysis was performed for each sequence using the Ensembl database. e The number of methylated molecules in the somatotrophinomas and nonfunctional pool of tumors, respectively.

associated with the novel gene C22orf3 (see Fig. 2) showed dense methylation; however, this was confined to two of five molecules in each pool of tumors. These findings, derived from pooled adenoma specimens, suggested that not all of the individual adenomas harbored this epigenetic change. However, because the CpG island associated with C22orf3 (PTAG) demonstrated the highest density and frequency of methylation relative to the other CpG islands isolated, we examined the role of this gene and its protein product in greater detail. Loss of Expression of PTAG in Pituitary Adenomas To assess whether PTAG represented a novel gene that is silenced in association with methylation of its CpG island, we used quantitative RT-PCR (qRT-PCR) to examine the expression of PTAG in pituitary tumors. qRT-PCR was performed on 38 individual pituitary adenomas that make up the major pituitary tumor subtypes. Expression of PTAG was significantly reduced relative to normal pituitary in 30 of 38 (⬃79%) pituitary tumors. Subdivision of tumors based on subtype showed that all seven corticotrophinomas and six prolactinomas showed significantly reduced PTAG expression relative to normal pituitary. For the somatotrophinomas and nonfunctional tumors, qRT-PCR showed reduced levels of PTAG expression in eight of 11 (⬃73%) and nine of 14 (⬃64%) tumors, respectively. Figure 3 shows the expression of PTAG in tumors from each subtype, relative to postmortem normal pituitary.

Association between Expression of PTAG and Methylation of its CpG Island To determine associations between loss of PTAG expression and methylation of its CpG island in pituitary adenomas, we used bisulfite sequencing of individual tumors. Of 30 tumors that failed to express PTAG, six (20%) were methylated and showed dense but heterogeneous methylation of CpG dinucleotides (Fig. 4). In normal pituitaries and those tumors that expressed PTAG, infrequent methylation was confined to single CpG dinucleotides. Figure 4 shows representative examples of molecules sequenced from tumors and normal pituitary tissue. The low frequency of methylationassociated loss of PTAG expression suggested that aberrations other than or in addition to this epigenetic mechanism may account for the loss of PTAG transcript in pituitary adenomas. Loss of Heterozygosity (LOH) and Sequence Analysis of PTAG To identify aberrations other than CpG methylation that might account for reduced PTAG expression, we employed LOH and direct sequence analysis of PTAG in 10 tumors that failed to express this gene. Using a polymorphic microsatellite marker (D22S448) at the cytogenetic location of PTAG (22q12.2) we found no LOH in 10 tumors studied. Furthermore, sequence analysis of all five coding exons of PTAG did not reveal changes relative to postmortem pituitaries or to the published DNA sequence (data not shown).



Fig. 2. Bisulfite Sequencing of a Portion of the PTAG Genes CpG Island A, The genomic structure of PTAG, comprising seven exons. The five coding exons (3–7) are indicated (solid squares). A CpG island extends from nucleotides ⫺989 to ⫹295 relative to the first nucleotide of exon 1 (bent arrow). The structure of the CpG island is shown in greater detail including individual CpG dinucleotides as thin vertical bars. Groups of two or more CpG dinucleotides are shown as thick vertical bars. The annealing sites of the bisulfite sequencing primers used in primary and secondary reactions are arrowed. F, Forward primer; R, reverse primer. B, Individual molecules were sequenced post-TA cloning from the two pools of tumors used for MsAP-PCR. The circles represent individual CpG dinucleotides: solid circles, methylated; open circles, unmethylated. Each row represents a single clone. Across the 21 CpG dinucleotides examined there was dense but heterogeneous methylation in two of five molecules sequenced from each of the two tumor pools. Postmortem normal pituitaries showed infrequent methylation in individual molecules.

Functional Characterization of PTAG

PTAG Is a Novel Proapoptotic Gene

Sequence homology analysis, using DNA and protein databases, identified PTAG as a novel chromosome 22 gene the protein product of which has no obvious functional domains except for a putative ubiquitinassociated domain and a c-myb DNA binding domain. To determine functional characteristics, stable transfectants of AtT20 cells were generated using an inducible expression vector harboring the complete cDNA sequence for human PTAG. Before transfection we determined the expression status of the murine homolog of PTAG (GenBank accession no. NM177370) in this cell line by RT-PCR analysis. AtT20 cells failed to express this gene relative to normal mouse pituitary, thus providing a suitable model system for functional analysis (Fig. 5A). Induced expression of PTAG was confirmed by RT-PCR analysis in response to a range of isopropyl-␤-D-thiogalactopyranoside (IPTG) concentrations (Fig. 5B). Growth curve analysis over 12 d in the presence of inducing agent (IPTG) showed no discernable difference in doubling time or cell viability as assessed by trypan blue exclusion relative to control cells harboring the empty expression vector (Fig. 5C).

Although induced expression of PTAG had no discernable affect on cell proliferation in AtT20 cells, this did not preclude the possibility of other novel cell regulatory functions of PTAG, including apoptosis. To test this hypothesis the dopamine D2 analog bromocriptine was used to induce apoptosis in these cells. AtT20 cells were treated with a range of drug doses for 72 h, before acridine orange staining and assessment of morphological changes characteristic of apoptosis (Fig. 6A). There was a dose-dependent increase in the number of cells showing apoptotic morphology, irrespective of PTAG expression status. However, PTAG-expressing cells showed an enhanced apoptotic response to bromocriptine challenge. At the highest drug dose (40 ␮M) there was a 3-fold increase (P ⫽ 0.001) in the number of apoptotic cells relative to control cells harboring an empty vector. The apoptotic response of PTAG-expressing cells to bromocriptine was confirmed with the general caspase inhibitor z-VAD-fmk (Fig. 6B). Irrespective of PTAG expression status there was an approximately 4-fold decrease in the percentage of bromocriptineinduced apoptotic cells pretreated with z-VAD-fmk, relative to cells incubated in the presence of bro-



1831

Fig. 3. Expression of PTAG in Pituitary Tumors as Determined by qRT-PCR Analysis The expression of PTAG in tumors relative to normal pituitary was determined using the relative standard curve method, normalized using 18S RNA as an internal control. Each value is the SEM from three separate experiments. Tumor numbers are shown below each sample. A, Somatotrophinomas; B, nonfunctional adenomas; C, corticotrophinomas; and D, prolactinomas.

increase in the levels of active caspases in AtT20 cells treated with 40 ␮M bromocriptine, irrespective of PTAG expression status. However, at all time points PTAG-expressing cells showed an enhanced response. At the 24-h time point PTAG-expressing cells showed a 2.4-fold (P ⫽ 0.001) increase in caspase activation relative to control cells. PTAG Enhances Apoptosis-Mediated Cell Death Fig. 4. Bisulfite Sequencing of a Portion of the PTAG Genes CpG Island in Individual Adenomas as Described in Fig. 2 Representative examples of tumors that either expressed (493) or did not express (484, 440) PTAG. Infrequent methylation was observed in normal pituitary and the adenoma (493) that expressed PTAG. In the representative adenomas that did not express this gene, one (484) showed dense methylation across the majority of CpG dinulceotides whereas the other (440) did not.

mocriptine alone. These results suggest that the enhanced proapoptotic response observed in bromocriptine-challenged cells expressing PTAG is most likely mediated through caspase activation.

Further evidence that PTAG augmented caspase-mediated cell death was derived from a DNA fragmentation labeling assay (TUNEL) to measure the latter stages of apoptosis. AtT20 cells were challenged with 40 ␮M bromocriptine in a time course experiment (Fig. 8). A time-dependent increase in the percentage of TUNEL-positive cells was observed for both PTAGexpressing and control cells. At all time points, PTAGexpressing cells showed higher levels of TUNEL-positive cells, and at 72 h showed a 2.5-fold increase (P ⫽ 0.002) relative to control cells.

DISCUSSION

PTAG-Associated Increase of Caspase Activation To directly determine PTAG-associated caspase activation in the apoptotic response, a fluorochrome-labeled general caspase inhibitor (fam-VAD-fmk) was used to detect relative levels of active caspases over a 24-h time course (Fig. 7). There was a time-dependent

CpG island methylation is a common aberration frequently associated with gene silencing in multiple tumor types including those of pituitary origin (reviewed in Ref. 36). To identify novel targets of methylation in human pituitary adenomas we used MsAP-PCR to isolate and characterize targets of this epigenetic


Fig. 5. Characterization of PTAG in the Mouse Pituitary Cell Line AtT20 A, The expression of the murine homolog of PTAG (mPTAG) was assessed in AtT20 cells by RT-PCR analysis. No PCR product was observed in AtT20 cells relative to mouse normal pituitary. PBGD was used as an internal control. NP, Normal mouse pituitary; ⫺VE, negative control. B, Inducible expression of PTAG was assessed in AtT20 cells by RT-PCR analysis over a range of IPTG doses relative to cells harboring an empty vector [multiple cloning site (MCS)]. Transcript identities are indicated with arrows, and doses of IPTG (1–10 mM) are shown above the gel. Maximal expression was observed at an IPTG dose of 5 mM. No expression was observed in control cells (MCS) harboring empty vector. C, Growth curve analysis of AtT20 cells expressing PTAG relative to control cells in the presence of 5 mM IPTG. Solid squares represent cells expressing PTAG; solid diamonds indicate empty vector control. Data represent the mean and SEM of three separate experiments. No discernable difference was observed in the growth profile.

change. We identified several novel sequences on the basis of their differential methylation in tumors vs. normal pituitary. For one of these sequences, encoding a chromosome 22-derived transcript, functional studies defined a role in apoptosis and on this basis was assigned the acronym PTAG. To increase the likelihood of identifying gene-specific differential methylation in tumors vs. normal pituitary, pooled tumor cohorts were used, and subsequent methylation status was determined within the pool by bisulfite sequencing. Within each of these pools and in contrast to normal pituitary, bisulfite sequencing showed tumor-specific differential methylation of PTAG. However, in each adenoma pool


(somatotrophinoma and nonfunctional), although methylation of the PTAG-associated CpG island was dense, it was confined, in each case, to two of five molecules, suggesting that not all of the individual adenomas within the pools were methylated. Several previous studies have employed MsAPPCR to isolate DNA sequences that are differentially methylated in a range of cancer types including colon, breast, and lung (29–31). In these studies, confirmation of methylation was determined by methylationsensitive restriction digestion and Southern blotting. We have extended this approach by using bisulfite sequencing as a rigorous method of determining the methylation status of differentially isolated products. Whereas previous studies have successfully employed MsAP-PCR to identify novel altered methylation profiles associated with oncogenic transformation, correlations between this change and cognate gene expression status have not been described. To determine associations between methylation and expression of PTAG, we used qRT-PCR of cDNA derived from individual pituitary adenomas that made up the major subtypes. Relative to normal pituitary, these studies showed that a significant proportion (⬃79%) of adenomas, irrespective of subtype, failed to express this gene, suggesting that loss of PTAG transcript is a common underlying aberration in pituitary tumors. Sodium bisulfite sequencing of these individual adenomas showed that loss of expression was infrequently associated with methylation of this gene’s CpG island. However, in those adenomas that were methylated, we did not detect expression of PTAG. Although these findings would suggest mechanisms in addition to methylation to be responsible for loss of PTAG, our analysis did not reveal genetic change as determined by either LOH or sequence analysis as an alternate mechanism. In this context, recent studies have described loss or substantially reduced expression of both GADD45␥ and MEG3 transcripts in a significant proportion of pituitary tumors (11, 12). Although our own studies have now described a significant correlation between methylation of the GADD45␥ gene’s CpG island and loss of transcript expression (22), mechanisms responsible for loss of MEG3 await further experimental study. The isolation of PTAG was performed on the basis of discriminating sequences differentially methylated in pituitary tumor DNA relative to normal tissue. The use of pooled tumor DNA for MsAP-PCR greatly increased the probability of identifying sequences that are methylated. Indeed, within the pooled tumors used to isolate this novel gene, our sequencing of individual clones (molecules) confirmed this. In these cases, although methylation was dense, it was confined to two of five molecules. However, the high frequency of loss across individual adenomas, as determined by qRTPCR, together with infrequent methylation and the absence of mutations or genetic loss, suggests either more subtle changes outside of the regions we have



1833

Fig. 6. Effect of Various Doses of Bromocriptine on the Percentage of AtT20 Cells Exhibiting Morphological Changes Indicative of Apoptosis Apoptotic cells were determined by staining with acridine orange after 72 h of treatment with bromocriptine. Solid bars, AtT20 cells expressing PTAG; open bars, empty vector controls. A, PTAG expressing cells were more sensitive to bromocriptine-induced apoptosis than controls [multiple cloning site (MCS)]. Cells showed a dose-dependent increase in apoptosis, and at the highest dose (40 ␮M), PTAG-expressing cells were 3-fold more sensitive to drug than empty vector controls. B, Inhibition of apoptosis by the general caspase inhibitor z-VAD-fmk. PTAG cells and control cells (MCS) were coincubated with 40 ␮M bromocriptine and 50 ␮g ml⫺1 of z-VAD-fmk for 72 h. The inhibitor significantly reduced the number of apoptotic cells relative to cells not treated with z-VAD-fmk. The data presented are the mean from three separate experiments. The mean ⫾ SD is shown for each sample.

examined or dysfunction in trans-acting factors that regulate expression of this gene. Because database searches revealed no obvious functional roles for the PTAG gene, we initially determined the consequences of enforced PTAG expression on proliferation in the mouse corticotroph cell line AtT20. In the absence of a suitable human pituitary cell line we considered AtT20 cells a suitable model system because they do not express the murine homolog of PTAG as determined by RT-PCR analysis, and our previous studies had shown that enforced expression of p16 in this cell line resulted in a G1-mediated growth arrest (37). However, enforced expression of PTAG in these stable transfected cells did not result in either growth arrest or decrease in cell viability. In contradistinction to these finding, enforced expression of PTAG in AtT20 cells significantly augmented the apoptotic response to bromocriptine challenge, suggesting that the function of this gene is linked to programmed cell death as a proapoptotic mediator.

We initially assessed bromocriptine-induced apoptosis by acridine orange staining, which showed that a greater proportion of PTAG-expressing cells displayed apoptotic morphology relative to cells harboring an empty-vector control. In addition, the general caspase inhibitor z-VAD-fmk significantly reduced the number of apoptotic cells, suggesting that apoptosis was through a caspase-mediated pathway. Increased levels of active caspases in PTAG-expressing cells were more directly confirmed in time course experiments using a fluorescent labeled caspase inhibitor that binds irreversibly to activated caspases. To further define the role of PTAG as a proapoptotic mediator, the latter stages of this process, DNA fragmentation, were measured by TUNEL labeling. In response to bromocriptine, PTAG-expressing AtT20 cells were significantly more sensitive to challenge that those harboring an empty vector. Previous studies have shown that the pituitary cell lines AtT20 and GH3 show decreased proliferation and

1834



Fig. 7. Relative Levels of Caspase Activity in Cells Treated with 40 ␮M Bromocriptine PTAG expressing AtT20 cells and controls [multiple cloning site (MCS)] were harvested at the indicated time points and subject to FLICA labeling. Fluorescence was measured using a fluorescence 96-well plate reader at 485-nm excitation and 535-nm emission wavelengths. A, There was a time-dependent increase in intrinsic caspase activity irrespective of PTAG expression status; however, PTAG-expressing cells showed a 2.4-fold greater level of intrinsic caspase activity relative to control cells at the 24-h time point. Data represent the mean and SD of three separate experiments. B, Representative examples of FLICA staining in PTAG-expressing and control cells in the absence (⫺VE) and presence (⫹VE) of 40 ␮M bromocriptine for 24 h. Left panels show phase contrast images, and right panels show fluorescence microscopy images.

increased apoptosis associated with increased p53 expression and reduced bcl-2 levels in response to bromocriptine challenge (38–40); however, it is not clear whether this response is mediated through D2 receptors (41–43). Our data also show, at least in AtT20 cells, that the absence of the murine homolog of PTAG did not confer complete protection from bromocriptine-induced apoptosis. However, the augmented apoptotic response, seen in these cells through enforced expression of PTAG, would support its role as a mediator rather than an obligate effector in this pathway. The intracellular cascades activated in both receptor-mediated extrinsic and stress-induced intrinsic apoptotic pathways are complex with activation of proximal and subsequent effector caspases (44, 45). In addition, cooperation between these mediators and amplification of the extrinsic pathway through cross-talk with the intrinsic pathway are also apparent (46). Thus, singular loss of a proapoptotic mediator within these pathways may result in a blunted apoptotic response. In this context, in mouse embryo fibroblasts derived from either bad- (47) or bax-deficient (48) mice, singular loss of these characterized proapoptotic genes shows that they do not account exclusively for physiological cell death.

The frequent loss of PTAG expression in pituitary tumors suggests that this is an early change in pituitary tumorigenesis leading to a blunted apoptotic response. Indeed, several studies have suggested a central role for a compromise apoptotic response in tumor evolution and progression (45, 49, 50). In these cases, resistance to apoptosis will allow damaged cells to proliferate, thus accruing additional genetic damage and tumor outgrowth. Whereas the role of this novel proapoptotic gene in other tumor types awaits investigation, its reactivation in tumors of pituitary origin may sensitize these cells to known and perhaps previously untested apoptosis-inducing agents.

MATERIALS AND METHODS Tumor Material Pituitary tumors, along with matched blood samples, were collected from patients who had undergone hypophysectomy and graded following standard histological assessment as previously described (51, 52). In addition, histologically normal postmortem pituitaries were obtained within 12 h of death. All samples were stored at ⫺70 C.



1835

Fig. 8. DNA Fragmentation Labeling Assay Using the TUNEL Reaction Cells were challenged with 40 ␮M bromocriptine over 72 h before TUNEL labeling. TUNEL-positive cells were counted using a fluorescent microscope from three separate experiments. A, A time-dependent increase in the percentage of TUNEL-positive cells was observed; however, PTAG-expressing cells showed increased sensitivity to drug with a 2.5-fold greater proportion of TUNEL-positive cells at 72 h relative to control cells. The mean ⫾ SD are shown. B, Representative examples of TUNEL staining in PTAG-expressing cells relative to control cells in the absence (⫺VE) and in the presence (⫹VE) of 40 ␮M bromocriptine at the 72-h time point. Left panels show phase contrast images, and right panels show fluorescence microscopy images.

DNA and RNA Preparation DNA was prepared from frozen pituitary tissue using the Nucleon DNA extraction kit (Anachem, Bedfordshire, UK) as described by the manufacturer. Total RNA was prepared from frozen pituitary material as previously described (53). Restriction Enzyme Digestion of Genomic DNA To determine differences in the methylation pattern between tumor and normal pituitary tissue, pooled DNA samples were first subject to cleavage with a panel of methylation-sensitive restriction enzymes. Separate DNA pools comprised a cohort of 10 somatotrophinomas, 10 nonfunctional tumors, and five postmortem normal pituitaries, respectively. Specimens were selected from male individuals to avoid confounding effects of X-chromosome inactivation. A total of 20 ␮g of DNA for each pooled sample were first treated with a panel of methylation-sensitive restriction enzymes. Briefly, restriction digest reactions were performed using SmaI (100 U) at 25 C for 6 h, SacII (100 U) and HpaII (50 U) at 37 C for a further 16 h, and finally BstUI (100 U) at 60 C for 6 h, after which DNA was precipitated with ethanol and resuspended in 70 ␮l H2O. Samples were then digested using 50 U of the methylation-insensitive enzyme MseI by incubating at 37 C for 6 h, purified by phenol/

chloroform extraction, and stored at ⫺20 C. All restriction enzymes were purchased from New England Biolabs (Hertfordshire, UK). MsAP-PCR Differentially methylated sequences within tumor samples were identified by MsAP-PCR using arbitrary primers; LI (5⬘CGTTCGTATCGACGGCGCGA-3⬘), BS5 (5⬘-CTCCCACGCG3⬘), and BS13 (5⬘-CGGGGCGCGA-3⬘) as previously described (29). PCRs were performed using 0.5 ␮g DNA from the digested pooled tumors, normal pituitaries, and undigested normal pituitary, in a total of 50 ␮l containing 10 mM Tris-HCl (pH 9.0), 50 mM KCl, 0.1% Triton-X, 1 U Taq DNA polymerase, 200 ␮M each of dATP, dCTP, dGTP, dTTP, 0.5 M Betaine, 2% (vol/vol) dimethylsulfoxide, and 40 pmol of arbitrary primer. Because the primers used are arbitrary in nature, but biased toward CpG-rich regions, the same primer may be used in both the forward and reverse reaction. Therefore, PCRs were performed using either a single primer or a combination of two primers. In all cases, PCRs were performed using the following conditions: 1 cycle at 96 C for 5 min, 35 cycles at 42 C for 1 min, 72 C for 1 min, 94 C for 0.5 min, and a final extension at 72 C for 5 min. PCR products were resolved on 8% polyacrylamide gels and visualized by silver staining, as previously described (21).

1836



Isolation and Characterization of DNA Fragments

Sequence and LOH Analysis of PTAG

Following PAGE, PCR products appearing to be differentially methylated by MsAP-PCR in pooled tumors vs. normal pituitaries were excised and eluted into sterile H2O. The eluate (2–5 ␮l) was reamplified using the same primer sets used in the original MsAP-PCR, under the same conditions as described above. Amplified products were resolved on 1% (wt/vol) agarose gels stained with ethidium bromide, and PCR products were excised and purified using the GeneClean II kit (Anachem, Bedfordshire, UK) followed by cloning into the PGem TA vector (Promega, Southampton, UK). Individual colonies were sequenced using a Big Dye Terminator cycle sequencing protocol (PE Applied Biosystems, Cheshire, UK), on an ABI Prism 310 Genetic analyzer, and data were analyzed with Sequencing Analysis V3.0 software (ABI Prism, PE Applied Biosystems). Sequence data were used to determine genomic information including homology to characterized or novel genes and cytogenetic map positions employing the Ensembl database (http://www.ensembl.org). Furthermore, each sequence was analyzed for CpG content and CpG/GpC ratio using the GrailEXP (http://compbio.ornl.gov/grailexp) program to predict conformity to criteria defining CpG islands (35).

The coding sequence of PTAG was sequenced using intronic primers designed to encompass each of the five coding exons (Table 2). Tumor and normal pituitary DNA (200 ng) was subject to PCR amplification as described above, and PCR amplicons were subject to cycle sequencing reactions (as described above) using both sense and antisense PCR primers. For LOH analysis, oligonucleotide primers to a polymorphic microsatellite marker (D22S448) centromeric to PTAG (22q12.2) were used to PCR amplify 200 ng DNA from matching blood and tumor samples using 25–28 cycles, an annealing temperature of 55 C, and 20 pmol of each forward and reverse primer, as described above (Table 2). Products were resolved on 8% nondenaturing polyacrylamide gels and visualized by silver staining as described (21). LOH was defined in tumor DNA by a reduction in allele intensity of greater than 80% (as assessed by scanning densitometry) or by the absence of one of the PCR-amplified alleles from tumor DNA.

Sodium Bisulfite Modification of DNA Sodium bisulfite conversion before sequencing was carried out as previously described (21). Briefly, 5 ␮g DNA were denatured by incubating at 37 C for 15 min in the presence of 0.4 M NaOH; 370 ␮l of freshly prepared sodium bisulfite solution (2.2 M sodium bisulfite, 8 M urea, 10 mM hydroquinone, pH 5.0) was added to the DNA and incubated at 55 C for 4 h. Modified DNA was recovered using the Gene Clean II kit (Anachem), followed by desulfonation of the modified DNA by incubating samples with 0.3 M NaOH at 37 C for 15 min. DNA was precipitated with ethanol and resuspended in 20 ␮l of TE buffer (pH 8.0). Bisulfite Sequencing of CpG Islands We employed bisulfite sequencing to determine the methylation status of the CpG islands isolated by MsAP-PCR, using primers designed according to criteria previously described (25). Primer sequences are available on request. PCRs were performed with 1–2 ␮l of modified DNA in 50 ␮l containing 10 mM Tris-HCl (pH 9.0), 50 mM KCl, 0.1% Triton-X, 1 U Taq DNA polymerase, 200 ␮M each of dATP, dCTP, dGTP, dTTP, and 20 pmol of each primer. PCR products were resolved by agarose gel electrophoresis, cloned into the PGem TA vector (Promega), and individual molecules were sequenced as described above. Quantitative RT-PCR Analysis of C22orf3 (PTAG) We employed qRT-PCR to determine expression status of PTAG in pituitary tumors relative to normal pituitary. Total RNA (5 ␮g) was used for cDNA synthesis using Superscript II RT according to the manufacturer’s instructions (Invitrogen, Paisley, Scotland, UK). Quantitative PCR was performed using 1 ␮l cDNA, corresponding to 250 ng initial RNA, and TaqMan MGB probes specific to PTAG (Assay ID: Hs00202161_m1) and 18S rRNA (Assay ID: Hs99999901_s1) as an endogenous control, according to manufacturer’s instructions (Applied Biosystems, Warrington, UK). Quantitation of PTAG gene expression in pituitary tumors relative to normal pituitary was determined using the relative standard curve method, with normal pituitary as the calibrator. Real time fluorescence measurement of PCR samples was performed using an ABI Prism 7000 sequence detection system, and data were analyzed using ABI Prism 7000 SDS software.

Generation of Stable Transfected Cell Lines The complete coding sequence of human PTAG (GenBank accession no. NM012265), obtained in a clone from the Mammalian Gene Collection (MGC: 3580, Cambridge, UK), was subcloned into a modified Gateway-compatible inducible vector, pOPRSVI (Stratagene, Cambridge UK). Recombination-mediated transfer was performed using the Gateway cloning system according to the manufacturer’s instructions (Invitrogen), and orientation and sequence integrity of the insert were confirmed by sequence analysis. The mouse pituitary cell line AtT20 was characterized as a suitable expression model, because these cells do not express endogenous PTAG as assessed by RT-PCR analysis using primers specific to the murine homolog of PTAG (GenBank accession no. NM177370) and porphobilinogen deaminase (PBGD) (GenBank accession no. NM013551) as an internal control (Table 2). The inducible vector containing the complete coding sequence of PTAG (pOP-PTAG) was transfected into AtT20 cells, harboring the constitutive expression vector (pCMVlacI) for the lac repressor (lacI), and stable transfectants generated as previously described (37). Control transfectants were also generated harboring the empty expression vector (pOP-MCS). Cells were cultured as previously described (37), and selection was performed in the presence of 500 ␮g ml⫺1 Geneticin and Hygromycin B (Invitrogen) followed by soft agar cloning to isolate single clones. Maximal expression of PTAG was determined in IPTG dose-response experiments and RT-PCR analysis (see Table 2 for primers and PCR conditions) as previously described (37). Growth Curve Analysis The effect of induced expression of PTAG was assessed in stable transfected clones (AtT20-PTAG) relative to a clone stably transfected with empty vector (AtT20-MCS). Cells were seeded at a density of 2.5 ⫻ 105 per 75 cm2 flask in 5 ml of medium and induced with 5 mM IPTG or PBS alone, and media were replenished at 2-d intervals, as previously described (37). Triplicate individual flasks were sacrificed at each time point (2 d) and cells were counted and their viability determined by trypan blue exclusion. The experiment was repeated three times. Detection of Apoptosis by Acridine Orange Staining AtT20-PTAG and AtT20-MCS control cells (2 ⫻ 105) were seeded into individual wells of a six-well cell culture plate in 2 ml of medium and induced with 5 mM IPTG for 48 h before



1837

Table 2. Primer Sequences Used for the Analysis of PTAG Sense Primer

Bisulfite sequencinga PCR Primary Secondary RT-PCRb Gene hPTAG mPTAG PBGD DNA sequencing PTAG exonc 1 2 3 4 5 Microsatellite analysis D22S448 Marker

Antisense Primer

Tm

PCR Product (bp)

5⬘-GGTTTTTTAGATGTTATTTTGGTT-3⬘ 5⬘-TAGAGAGGATTGGGATAAGAG-3⬘

5⬘-TTGGTTGGGAAGATTGAGTGGAGTTGT-3⬘ 5⬘-GGTGAGTATGGTGGAATT-3⬘

60 50

352 222

5⬘-ACTGGGAAGACTCAGCCCTG-3⬘ 5⬘-ACTGGCCTAGCCTATGCTGCT-3⬘ 5⬘-CTTCTGCAGACACCAGGGGA-3⬘

5⬘-AGCTGCTGCAGCCGCAGAG-3⬘ 5⬘-TCTTCCATGTCAGCTGTCCAGG-3⬘ 5⬘-GCTCTTCTCTCCAATCTTAGAG-3⬘

60 60 55

250 275 281

5⬘-TCTCCTGCTTCCTCCAAGGTC-3⬘ 5⬘-ACGGGCCCTGGGCTCAGG-3⬘ 5⬘-GCCACAGTGCCCACCACCTGGA-3⬘ 5⬘-AGAAGTGGGGAGTGTCAGGGCCA-3⬘ 5⬘-TGATGCCAACACAGAAGAGTC-3⬘

5⬘-TTATCCCCAGCACACAGCCC-3⬘ 5⬘-CACTAACCTGGCACATAAGC-3⬘ 5⬘-ACGCTATTCCATGCTGGTCCACA-3⬘ 5⬘-TCCAGGTGGGCACTGTGGC-3⬘ 5⬘-GGCCCTGACACTCCCCAC-3⬘

60 55 65 59 58

239 448 353 419 458

5⬘-TCAACCTTGTCCTCTCAGGG-3⬘

5⬘-GATTAGGGGAGGGCATGG-3⬘

55 186–208

Tm, Annealing temperature. a Primers were designed to the bisulfite-converted DNA sequence. PCRs were performed using a primary and secondary (nested) reaction. b Oligonucleotide primer sequences specific for human PTAG (GenBank accession no: NM_012265 (hPTAG), the murine homolog of PTAG (GenBank accession no. NM_177370 (mPTAG), and the murine housekeeping gene PBGD (GenBank accession no. NM_013551). c Oligonucleotide primer sequences specific for each of the five coding exons of PTAG.

bromocriptine challenge. Cells were assayed for apoptotic response as described (54). Briefly, adherent and nonadherent cells were collected and washed in PBS before being pelleted; 10 ␮l of cell suspension were then mixed with 10 ␮l acridine orange (50 ␮g ml⫺1 in PBS) and analyzed on a wet mount slide using a Leica DMR fluorescent microscope (Leica Corp., Deerfield, IL). The percentage of cells with apoptotic morphology (nuclear and cytoplasmic condensation, nuclear fragmentation, membrane blebbing, and apoptotic body formation) was determined from 200 cells in each of three separate fields, and the mean from three independent experiments was determined. Where indicated, bromocriptine challenge was also performed in the presence of 50 ␮g ml⫺1 of the general caspase inhibitor z-VAD-fmk (Flowgen, Leicestershire, UK) before staining. Measurement of Caspase Activation in Situ by Fluorochrome-Labeled Inhibition of Caspases (FLICA) Caspase activity in AtT20 cells was measured using the fluorochrome (FAM)-labeled broad spectrum inhibitor of caspases (fam-VAD-fmk). This reporter molecule irreversibly binds to active caspases allowing measurement of caspase activation in situ (55). AtT20-PTAG and AtT20-MCS control cells were seeded and induced with IPTG as described above and were then challenged with 40 ␮M bromocriptine for 72 h before labeling according to the manufacturer’s instructions (Flowgen). Briefly, a 300-␮l aliquot of cells at a density of 1 x 106 ml⫺1 was labeled using 10 ␮l of a freshly prepared 30⫻ working dilution of fam-VAD-fmk by incubation at 37 C and 5% CO2 for 15 min, followed by two washes with 2 ml of a 1⫻ working dilution wash buffer. Cells were resuspended in 320 ␮l PBS and placed on ice before analysis. Fluorescence was measured at an excitation wavelength of 485 nm and emission at 535 nm using an automated fluorescence plate reader (Wallac 1420 Victor, Milton Keynes, UK).

DNA Fragmentation Labeling Assay (TUNEL) Apoptosis-induced DNA fragmentation was measured using the APO-BRDU kit (BD Biosciences, Oxford, UK). Cells were seeded, induced with IPTG, and challenged with 40 ␮M bromocriptine as described above, followed by fixation in 1% paraformaldehyde in PBS at 4 C for 30–60 min, postfixed in 70% ethanol and stored at ⫺20 C for 12–18 h before staining. Briefly, fixed cells (1 ⫻ 106) were washed twice with PBS before treatment with a TUNEL reaction mix (bromolated deoxyuridine triphosphate, terminal deoxynucleotidyl transferase, and reaction buffer) for 60 min at 37 C, followed by washing and incubation with a fluorescein isothiocyanatelabeled bromodeoxyuridine monoclonal antibody for 30 min at room temperature. TUNEL-positive cells were counted from 200 cells in each of three separate fields for each measurement, using a Leica DMR fluorescent microscope. The mean from three separate experiments was determined.

Acknowledgments Received March 2, 2004. Accepted April 16, 2004. Address all correspondence and requests for reprints to: Dr. W. E. Farrell, Institute for Science and Technology in Medicine, School of Postgraduate Medicine, Keele University, North Staffordshire Hospital, Stoke-on-Trent ST4 7QB, United Kingdom. E-mail: [email protected]. This work was supported by a Medical Research Council project grant award (to W.E.F.).

REFERENCES 1. Asa SL, Ezzat S 2002 The pathogenesis of pituitary tumors. Nat Rev Cancer 2:836–848


2. Lloyd RV 2001 Molecular pathology of pituitary adenomas. J Neurooncol 54:111–119 3. Alexander JM, Biller BMK, Bikkal H, Zervas NT, Arnold A, Klibanski A 1990 Clinically non-functioning pituitary tumors are monoclonal in origin. J Clin Invest 86:336–340 4. Pernicone PJ, Scheithauer BW, Sebo TJ, Kovacs KT, Horvath E, Young WF, Lloyd RV, Davis DH, Guthrie BH, Schoene WC 1997 Pituitary carcinoma: a clinicopathological study of 15 cases. Cancer 79:804–812 5. Landis CA, Mastyers SB, Spada A, Pace AM, Boune HR, Valler A 1989 GTPase inhibiting mutations activate the ␣ chain of Gs and stimulate adenyl cyclase in human pituitary adenomas. Nature 340:692–696 6. Clementi E, Malgaretti N, Meldoteri J, Taramelli R 1990 A new constitutively activating mutation in Gs protein ␣ subunit-gsp oncogene is found in human pituitary tumors. Oncogene 5:1059–1061 7. Woloschak M, Roberts JL, Post K 1994 c-myc, c-fos, and c-myb Gene expression in human pituitary adenomas. J Clin Endocrinol Metab 79:914–257 8. Levy A, Hall S, Yeudall MA, Lightman SL 1994 p53 Gene mutations in pituitary adenoma: rare events. Clin Endocrinol (Oxf) 41:809–814 9. Tanaka C, Yoshimoto K, Yang P, Kimura T, Yamada S, Moritani M, Samo T, Itakura M 1997 Infrequent mutations of p27kip1 gene and trisomy 12 in a subset of human pituitary adenomas. J Clin Endocrinol Metab 82: 3141–3147 10. Pei L, Melmed S 1997 Isolation and characterisation of a pituitary tumor-transforming gene (PTTG). Mol Endocrinol 11:433–411 11. Zhang X, Sun H, Danila DC, Johnson SR, Zhou Y, Swearingen B, Klibanski A 2002 Loss of expression of GADD45␥, a growth inhibitory gene, in human pituitary adenomas: implications for tumorigenesis. J Clin Endocrinol Metab 87:1262–1267 12. 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. J Clin Endocrinol Metab 88:5119–5126 13. Zhang X, Horvitz GA, Heaney AP, Nakashima M, Prezant TR, Bronstein MD, Melmed S 1999 Pituitary tumor transforming gene (PTTG) expression in pituitary adenomas. J Clin Endocrinol Metab 84:761–767 14. Zhang X, Horvitz GA, Prezant TR, Valentini A, Nakashima M, Bronstein MD, Melmed S 1999 Structure, expression and function of human pituitary tumor transforming gene (PTTG). Mol Endocrinol 13:156–166 15. Woloschak M, Yu A, Post KD 1997 Frequent inactivation of the p16 gene in human pituitary tumors by gene methylation. Mol Carcinog 19:221–224 16. Simpson DJ, Bicknell JE, McNicol AM, Clayton RN, Farrell WE 1999 Hypermethylation of the p16/CDKN2A/ MTS1 gene and loss of protein expression is associated with nonfunctional pituitary adenomas but not somatotrophinomas. Genes Chromosomes Cancer 24: 328–336 17. Jaffrain-Rea ML, Ferretti E, Toniato E, Carmita K, Santono A, Di Stefano D, Ricevuto E, Masoder M, Tamburrano G, Cantore G, Gulino A, Martinotti S 1999 p16 (INK4a, MTS-1) gene polymorphism and methylation status in human pituitary tumors. Clin Endocrinol (Oxf) 51: 317–325 18. Seeman N, Kuhn D, Wrocklage C, Keyvani K, Hackl W, Buchfelder M, Fahlbusch R, Paulus W 2001 CDKN2A/ p16 inactivation is related to pituitary adenoma type and size. J Pathol 193:491–497 19. Ruebel KH, Jin L, Zhang S, Scheithauer BW, Lloyd RV 2001 Inactivation of the p16 gene in human pituitary nonfunctioning tumors by hypermethylation is more common in null cell adenomas. Endocr Pathol 12: 281–289


20. 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 Res 60:1211–1216 21. Simspon DJ, Clayton RN, Farrell WE 2002 Preferential loss of death associated protein kinase expression in invasive pituitary tumors is associated with either CpG island methylation or homozygous deletion. Oncogene 21:1217–1224 22. Bahar A, Bicknell JE, Simpson DJ, Clayton RN, Farrell WE 2003 Loss of expression of the growth inhibitory gene GADD45␥, in human pituitary adenomas, is associated with CpG island methylation. Oncogene 23: 934–944 23. Baylin SB, Herman JG, Graff JR, Vertino PM, Issa J-P 1997 Alterations in DNA methylation: a fundamental aspect of neoplasia. Adv Cancer Res 72:141–196 24. Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB 1996 Methylation-specific PCR. A novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci USA 93:9821–9826 25. Clark SJ, Harrison J, Paul CL, Frommer M 1994 High sensitivity mapping of methylated cytosines. Nucleic Acid Res 22:2990–2997 26. Kamb A, Gruis NA, Weaver-Feldhaus J, Liu Q, Harshman K, Tavtigian SV, Stockert E, Day III RS, Johnson BE, Skolnick MH 1994 A cell cycle regulator potentially involved in genesis of many tumor types. Science 264: 436–440 27. Kawai J, Hirotsune S, Hirose K, Fushiki S, Watanabe S, Hayashizaki Y 1993 Methylation profiles of genomic DNA of mouse developmental brain detected by restriction landmark genomic scanning (RLGS) method. Nucleic Acids Res 21:5604–5608 28. Toyata M, Ho C, Ahuja N, Jair KW, Li Q, Toyata MO, Baylin SB, Issa J-P 1999 Identification of differentially methylated sequences in colorectal cancer by methylated CpG island amplification. Cancer Res 59: 2307–2312 29. Gonzalgo ML, Liang G, Spruck CH, Zingg JM, Rideout WH, Jones PA 1997 Identification and characterisation of differentially methylated regions of genomic DNA by methylation-sensitive arbitrarily primed PCR. Cancer Res 57:594–599 30. Huang THM, Laurx DE, Hamlin BC, Tran P, Tran H, Lubahn DM 1997 Identification of DNA methylation markers for human breast carcinomas using the methylation-sensitive restriction fingerprinting technique. Cancer Res 57:1030–1034 31. Kohno T, Kawarrishi M, Inazana J, Yokota J 1998 Identification of CpG islands hypermethylated in human lung cancer by the arbitrarily primed-PCR method. Hum Genet 102:258–264 32. Yan PS, Chen CM, Shi H, Rahmatpanah F, Wei SH, Caldwell CW, Huang TH 2001 Dissecting complex epigenetic alterations in breast cancer using CpG island microarrays. Cancer Res 61:8375–8380 33. Hatada I, Kato A, Morita S, Obata Y, Nagoaka A, Sakurada A, Sato M, Horii A, Tsujimoto A, Matsubara K 2002 A microarray-based method for detecting methylated loci. J Hum Genet 47:448–451 34. Cross SH, Charlton JA, Nan X, Bird AP 1994 Purification of CpG islands using a methylated DNA binding column. Nat Genet 6:236–244 35. Gardiner-Garden M, Frommer M 1987 CpG islands in vertebrate genomes. J Mol Biol 196:261–282 36. Farrell WE, Clayton RN 2003 Epigenetic change in pituitary tumorigenesis. Endocr Relat Cancer 10:225–236 37. 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


38. 39.

40. 41.

42.

43.

44.

45.

46.

AtT20 pituitary tumor cell line. Mol Endocrinol 13: 1801–1810 Yin D, Kondo S, Takeuchi J, Morimura T 1993 Induction of apoptosis in murine ACTH-secreting pituitary adenoma cells by bromocriptine. FEBS Lett 339:73–75 Yin D, Tamaki N, Kokunai T, Yasuo K, Yonezawa K 1999 Bromocriptine-induced apoptosis in pituitary adenoma cells: relationship to p53 and bcl-2 expression. J Clin Neurosci 6:326–331 Yin D, Kondo S, Takeuchi J, Morimura T 1993 Induction of apoptosis in rat somatotrophin-secreting pituitary adenoma cells by bromocriptine. Oncol Res 5:383–387 Kanasaki H, Fukunaga K, Takahashi K, Miyazaki K, Miyamoto E 2000 Involvement of p38 mitogen-activated protein kinase activation in bromocriptine-induced apoptosis in rat pituitary GH3 cells. Biol Reprod 62: 1486–1494 Johnson JM, Wood DF, Read S, Johnston DG 1993 Dopamine regulates D2 receptor gene expression in normal but not in tumorous rat pituitary cells. Mol Cell Biol 92:63–68 Missale C, Castelletti L, Boroni F, Memo M, Spano P 1991 Epidermal growth factor induces the functional expression of dopamine receptors in the GH3 cell line. Endocrinology 128:13–20 Peart MJ, Tainton KM, Ruefli AA, Dear AE, Sedelies KA, O’Reilley LA, Waterhouse NJ, Trapani JA, Johnstone RW 2003 Novel mechanisms of apoptosis induced by histone deacetylase inhibitors. Cancer Res 63:4460–4471 Zindy F, Eischen CM, Randle DH, Kamijo T, Cleveland JL, Sherr CJ, Roussel MF 1998 Myc signaling via the ARF tumor suppressor regulates p53-dependent apoptosis and immortalization. Genes Dev 12:2424–2433 Igney FH, Krammer PH 2002 Death and anti-death: tumor resistance to apoptosis. Nat Rev Cancer 2:277–288


1839

47. Ranger AM, Zha J, Harada H, Datta SR, Danial NN, Gilmore AP, Kutok JL, Le Beau MM, Greenberg ME, Korsmeyer SJ 2003 Bad-deficient mice develop diffuse large B cell lymphoma. Proc Natl Acad Sci USA 100: 9324–9329 48. McCurrach ME, Conner TME, Knudson CM, Korsmeyer SJ, Lowe SW 1997 bax-Deficiency promotes drug resistance and oncogenic transformation by attenuating p53dependent apoptosis. Proc Natl Acad Sci USA 94: 2345–2349 49. Ferri KF, Kroemer G 2001 Organelle-specific initiation of cell death pathways. Nat Cell Biol 3:E255–E263 50. Jiang M, Milner J 2003 Bcl-2 constitutively suppresses p53-dependent apoptosis in colorectal cancer cells. Genes Dev 17:832–837 51. Hibberts NA, Simpson DJ, Bicknell JE, Broome JC, Hoban PR, Clayton RN, Farrell WE 1999 Analysis of cyclin D1 (CCND1) allelic imbalance and overexpression in sporadic pituitary tumors. Clin Cancer Res 5:2133–2139 52. Bates AS, Farrell WE, Bicknell JE, Talbot AJ, Broome JC, Perrett CW, Thakker RV, Clayton RN 1997 Allelic deletion in pituitary adenomas reflects aggressive biological activity and has potential value as a prognostic marker. J Clin Endocrinol Metab 82:818–824 53. Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acid guanidinium thiocynate-phenolchloroform extraction. Anal Biochem 162:156–159 54. Schonhoff CM, Gaston B, Mannick JB 2003 Nitrosylation of cytochrome c during apoptosis. J Biol Chem 278: 18265–18270 55. Bedner E, Smolewski P, Amstad P, Darzynkiewicz Z 2000 Activation of caspases measured in situ by binding of fluorochrome-labelled inhibitors of caspases (FLICA): correlation with DNA fragmentation. Exp Cell Res 259:308–313

Molecular Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.