Expression of Phosphorylated p27Kip1 Protein and Jun Activation

0013-7227/02/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 87(6):2635–2643 Copyright © 2002 by The Endocrine Society

Expression of Phosphorylated p27Kip1 Protein and Jun Activation Domain-Binding Protein 1 in Human Pituitary Tumors ´ RTA KORBONITS*, HARVINDER S. CHAHAL*, GREGORY KALTSAS, SUZANNE JORDAN, MA YULDUZ URMANOVA†, ZAMIRA KHALIMOVA†, PHILIP E. HARRIS, WILLIAM E. FARRELL, FRANCOIS-XAVIER CLARET, AND ASHLEY B. GROSSMAN Department of Endocrinology (M.K., H.S.C., G.K., S.J., Y.U., Z.K., A.B.G.), St. Bartholomew’s Hospital, London EC1A 7BE, United Kingdom; King’s College Hospital (P.E.H.), London SE15 5QN, United Kingdom; Center for Cell and Molecular Medicine (W.E.F.), University of Keele School of Postgraduate Medicine, Stoke-on Trent ST4 7QB, United Kingdom; and Department of Molecular Therapeutics (F.-X.C.), The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030 The cyclin-dependent kinase inhibitor p27Kip1 (p27) plays a pivotal role in controlling cell proliferation during development and tumorigenesis. p27 has been implicated in pituitary tumorigenesis in studies of knockout mice and in analyses of human pituitary tumor samples. In this study, we further explored the role of p27 in human pituitary tumors by measuring levels of phosphorylated p27 (P-p27), and also Jun activation domain-binding protein 1 (Jab1), which is thought to facilitate the phosphorylation and degradation of p27, in normal pituitary tissue (n ⴝ 21), pituitary adenomas (n ⴝ 75), and pituitary carcinomas (n ⴝ 10). The amount of p27 protein in corticotroph adenomas and pituitary carcinomas was much lower than that in normal pituitary tissue or other types of pituitary adenoma. Nuclear P-p27 protein levels were significantly decreased in the adenomas, compared with the normals, and were much lower in the carcinomas, compared with either normal pituitary tissue or pituitary adenomas. However, Pp27 levels in corticotroph adenomas were similar to normal pituitary tissue, thus demonstrating a greatly increased ratio of P-p27 to p27 specifically in corticotroph tumors. No dif-

T

HE CYCLIN-DEPENDENT KINASE inhibitor p27Kip1 (p27) is an essential participant in the regulation of cell cycle progression (1). Reduced expression of p27 protein has been frequently observed in a variety of human malignancies, and a significant correlation between low p27 protein expression and high tumor grade has been described (for review, see Ref. 2). In humans, pituitary adenomas are common neoplasms occurring in 10 –27% of unselected autopsy series. The pathogenesis of these primarily slow-growing, benign tumors is unknown, but it has been suggested that p27 could play an important role in pituitary tumorigenesis, because p27-deficient or haploinsufficient mice show a specific propensity for the development of pituitary tumors arising from the intermediate lobe and secreting ACTH (3– 6). p27 is rarely mutated in human tumors (7), including pituitary tumors (8 –10), but its gene was shown to be siAbbreviations: AP, Activator protein; CDK, cyclin-dependent kinase; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; Jab1, Jun activation domain-binding protein 1; NFPA, nonfunctioning pituitary adenomas; p27, cyclin-dependent kinase inhibitor p27Kip1; P-p27, phosphorylated p27; Thr187, threonine residue 187.

ference was found in Jab1 protein levels in either corticotroph tumors or other pituitary adenomas, compared with normal tissue, but there was a small but significant increase in Jab1 levels in carcinomas. Corticotroph and metastatic tumors both showed a significantly higher Ki-67 labeling index than normal pituitary or other types of pituitary adenomas, and in general the Ki-67 labeling index was negatively correlated with p27 nuclear staining. The amount of p27 and Jab1 mRNA was positively correlated in all pituitary samples studied but did not correlate with the changes in immunostaining. Our findings suggest that in corticotroph tumors there is an accentuated phosphorylation of p27 into P-p27, possibly related to increased cyclin E expression, whereas both p27 and P-p27 are subject to increased degradation in pituitary carcinomas. Such variations in phosphorylation may play a role in pituitary tumorigenesis, but modulation of Jab1 is unlikely to be important in the pathogenesis of pituitary adenomas. (J Clin Endocrinol Metab 87: 2635–2643, 2002)

lenced by methylation in a pituitary cell line (11). In addition, trisomy of chromosome 12, the location of the gene of p27, has been described in a subset of pituitary adenomas (12). Although regulation of p27 mRNA expression has been shown to play an important role in certain situations (2, 11, 13, 14), most of the studies suggest that p27 is primarily regulated at the protein level, and factors influencing p27 degradation may be important regulators of cell cycle progression. p27 undergoes phosphorylation before transportation to the cytoplasm for ubiquitin-mediated degradation (15, 16). Recently, a new protein has been implicated in the transport of p27 into the cytoplasm: Jun activation domainbinding protein 1 (Jab1) (17, 18). We and others have previously shown that p27 protein expression was significantly less in human pituitary adenomas compared with the normal pituitary, and virtually absent in corticotroph or malignant pituitary tumors (19, 20). Our goals in this study were to measure the expression of p27, phosphorylated p27 (P-p27), and Jab1 protein and mRNA in a variety of pituitary tumors, and to compare these to the Ki-67 labeling index, a marker of cell proliferation.

2635

2636

J Clin Endocrinol Metab, June 2002, 87(6):2635–2643

Materials and Methods Samples Human pituitary samples (n ⫽ 106) removed at transsphenoidal surgery were classified histologically using hematoxylin and eosin, reticulin and immunohistochemical staining, into 21 normal pituitaries and 85 pituitary tumors. Abnormal pituitary tissue was classified as benign adenoma, aggressive adenoma, or metastatic tumor. Adenomas were defined as samples removed at pituitary surgery with a disrupted reticulin pattern. The 65 benign adenomas were categorized as GHsecreting tumors (n ⫽ 19), ACTH-secreting tumors (n ⫽ 13), prolactinomas (n ⫽ 11), nonfunctioning pituitary adenomas (NFPAs) (n ⫽ 16), TSH-omas (n ⫽ 2), and FSH-omas (n ⫽ 4). For the NFPAs, immunostaining showed no staining with antisera for any of the pituitary hormones (ACTH, GH, PRL, FSH, LH, TSH, ␣-subunit, or the ␤-subunit of human chorionic gonadotropin) in 7 of the 16 NFPA samples, whereas the rest showed variable PRL, LH, ␣-subunit, and the ␤-subunit of human chorionic gonadotropin staining. The aggressive tumors included NFPAs (n ⫽ 5), prolactinomas (n ⫽ 3), and somatotroph adenomas (n ⫽ 2). The definition of aggressiveness was taken to include those showing cavernous sinus invasion at operation or recurrence after surgery and radiotherapy. Metastatic carcinomas were defined as pituitary tumors with histologically verified distinct extrapituitary metastases. The metastatic group (n ⫽ 10) included ACTH-secreting (n ⫽ 4; metastatic sites, liver, cervical vertebra, lung, and thoracic spine), PRL-secreting (n ⫽ 4; metastatic sites, internal jugular vein, base of skull, vault of skull, and brain stem), and nonfunctioning tumors (n ⫽ 2; metastatic sites, ribs, lumbar spine, and intradural cord). In each case, the primary tumor was analyzed, except one of the ACTH-secreting carcinomas when the cervical vertebra sample was studied. All patients had a 100-mg hydrocortisone im injection before surgery, whereas patients with the clinical diagnosis of Cushing’s disease routinely received 6– 8 wk of medical therapy with metyrapone, ketoconazole, or both to normalize cortisol levels before surgery. To avoid inconsistency in the treatment of tissue samples that is unavoidable with autopsy controls, the control normal pituitaries used in the immunohistochemical studies were part of resection specimens removed at transsphenoidal surgery for presumptive tumors that proved on staining to consist of normal pituitary tissue and architecture. The normal pituitaries included tissue from patients with the clinical diagnosis of Cushing’s disease (n ⫽ 14), prolactinoma (n ⫽ 3), acromegaly (n ⫽ 2), NFPA (n ⫽ 1), and an arachnoid cyst (n ⫽ 1). Patient samples studied by RT-PCR (n ⫽ 47) included 11 somatotroph, 5 corticotroph, 4 gonadotroph, 5 lactotroph adenomas, and 22 NFPAs, and their data were compared with 4 normal pituitaries removed at autopsy from patients with nonendocrine disease. The institutional Ethics Committee approved all studies.

Immunostaining for p27, phospho-p27, Jab1, and Ki-67 Tissue samples were collected at transsphenoidal surgery and prepared for pathological examination by formalin fixation and paraffin embedding. Sections underwent heat-mediated antigen retrieval treatment (21) before immunohistochemical analysis with the standard avidin-biotin complex immunoperoxidase system (Vectastain Elite, Vector Laboratories, Inc., Peterborough, UK) (21). Human anti-p27 antibody raised against the full-length p27 protein was used at 1:50 as described previously (22), and part of the data on p27 staining was used from our own previously published study (19). For P-p27 staining, we used an epitope affinity-purified rabbit antihuman polyclonal antibody (1:250) (Zymed Laboratories, Inc., San Francisco, CA), raised against the 22 amino acid p27 peptide fragment containing the phosphorylated threonine residue 187 (Thr187) in the C terminus. For Jab1 staining, an affinity-purified antihuman monoclonal Jab1 antibody (GeneTex, Wiltshire, UK), raised against a peptide mapping at the amino terminus of Jab1 (17), was used at a dilution of 1:500. For Ki-67 staining, a rabbit antihuman polyclonal Ki-67 antibody was used at a dilution of 1:200 (DAKO Corp., Cambridgeshire, UK). The Ki-67 proliferative index was determined as the percentage of positive cells of all counted cells and expressed as the labeling index. Human tonsil tissue was used as a positive control because this contains lymphoid tissue with variable proliferative activity. In the mantle (peripheral) zone of the follicle, the cells are mainly quiescent, whereas cells in the germinal centers are highly proliferative. For as-

Korbonits et al. • Phospho-p27 and Jab1 in Pituitary Tumors

sessment of p27, P-p27, and Jab1 immunostaining in pituitary samples, approximately 500 cells were counted and assessed for the intensity of staining. Cells with strong or moderate staining were counted as positive, cells with no staining were counted as negative, while cells with weak staining were scored separately. This type of quantitative analysis of immunostaining has been used previously in a number of publications (23–26). Sections were chosen on each slide from a randomized grid array and were counted blind to the diagnosis. Cytoplasmic staining was assessed as focal or diffuse, and as negative, weak, moderate, or strong. Every case had a negative control in which the primary antibody was omitted and replaced by 1% BSA. Staining specificity was assessed by pretreating the slides with the appropriate antigen used in the antibody production (blocking peptide): full-length p27, the 22 amino acid fragment of P-p27, and the N-terminal fragment of the Jab1 protein. Crossreactivity of p27 and P-p27 antibodies was assessed using p27 blocking peptide and P-p27 antibody, and P-p27 blocking peptide and p27 antibody, in tonsil tissue. Data showed no cross-reaction between the two antibodies. Figure 1 illustrates staining with p27 antibody with a) no pretreatment (positive staining), b) in the presence of added p27 peptide (no positive staining), and c) in the presence of P-p27 peptide (positive staining); it also shows staining with P-p27 antibody and d) no pretreatment (positive staining), e) in the presence of added p27 peptide (positive staining), and f) P-p27 protein (no staining). Jab1 staining with or without pretreatment with Jab1 blocking peptide is shown in Fig. 1, g and h. Photographs of slides were taken using a Leica Corp. DMR microscope and a Leica Corp. DC200 digital camera. They were printed by an HP Inkjet printer on HP PremiumPhoto paper.

Semiquantitative RT-PCR Relative amounts of mRNA for p27 and Jab1 were compared by RT-PCR. Total RNA was obtained and reverse-transcribed into cDNA by a standardized technique, as previously described (27). RT-PCRs with omission of reverse transcriptase, and with water replacing the template, were used as negative controls. The PCR was performed using primers spanning one or more introns of the genes being studied to allow genomic DNA contamination to be excluded. Primers used for p27 (358-bp product) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (198-bp product) were described elsewhere (8, 28); a new primer was used for Jab1 (539-bp product; GenBank accession no. U65928; sense, 5⬘ GTGATGGGTCTGATGCTAGGAA 3⬘; antisense, 5⬘ AGCAAGCTAGAAGAACTCAACG 3⬘). The sequence of the Jab1 PCR product was confirmed by direct sequencing. Gene expression was determined by multiplex PCR with the p27, Jab1, and GAPDH primers. The PCRs were performed during the linear phase of the synthesis curve for each PCR product (data not shown). cDNA (2.5 ␮l; 250 ng RNA equivalent) template was incubated with 0.5 ␮l of 20 ␮m deoxynucleotides (Promega Corp., Southampton, UK), 0.4 ␮mol Jab1 and p27 primers, and 0.15 ␮mol GAPDH primers, 0.125 U HotStart Taq DNA polymerase enzyme (QIAGEN, Crawley, UK), 5 ␮l Q solution, and 2.5 ␮l QiaBuffer containing 1.5 mmol/liter MgCl2, according to the manufacturer’s instructions in a 25 ␮l PCR. Twenty-six cycles were performed at 94 C for 1 min, 58 C for 1 min, and 72 C for 1 min after a denaturing cycle of 95 C for 15 min. A final extension cycle of 10 min at 72 C was used. The PCR products were visualized on ethidium bromide-stained 2% agarose gels. The absorbance values were measured for each band by densitometry (Model DS670 image densitometer, Bio-Rad Laboratories, Inc. Hemel-Hempstead, Hertfordshire, UK), using the Molecular Analyst PC software for Bio-Rad Laboratories Image Analysis systems, and expressed as OD units. A ratio between p27 or Jab1 and GAPDH was obtained for each individual sample.

Immunoblot analysis Tissue from a nonfunctioning adenoma was ground with a few strokes of a Teflon pestle and lysed in lysis buffer [25 mm HEPES (pH 7.7), 400 mm NaCl, 0.5% Triton X-100, 1.5 mm MgCl2, 2 mm EDTA, 2 mm dithiothreithol, 0.1 mm phenylmethylsulfonyl fluoride; the protease inhibitors leupeptin 10 ␮g/ml, peptstatin 2 ␮g/ml, antipain 50 ␮g/ml, aprotinin 2 ␮g/ml, chymostatin 20 ␮g/ml, and benzamidine 2 ␮g/ml; and the phosphatase inhibitors 2 mm NaF, 1 mm Na3VO4, and 20 mm ␤-glycerophosphate] for 20 min at 4 C. The cell pellet was spun at 13,000 rpm for 15 min at 4 C. Proteins of crude tissue extract (40 ␮g protein/


J Clin Endocrinol Metab, June 2002, 87(6):2635–2643 2637

FIG. 1. Staining in human tonsil. p27 antibody: a, no pretreatment (positive staining); b, in the presence of added p27 peptide (no positive staining); and c, in the presence of P-p27 peptide (positive staining). P-p27 antibody: d, no pretreatment (positive staining); e, in the presence of added p27 peptide (positive staining); and f, P-p27 protein (no staining). Jab1 antibody: g, no pretreatment (positive staining); h, in the presence of added Jab1 peptide (no staining). Ki-67 antibody: i, positive staining. Magnification, ⫻200.

lane) were resolved on 10% SDS-PAGE and were transferred to polyvinylidene difluoride membrane. Membranes were blocked with PBSTween 20 containing 5% milk and probed with primary antibodies as specified. Reactions were visualized with suitable secondary antibody conjugated with horseradish peroxidase (Bio-Rad Laboratories, Inc., Hercules, CA) using enhanced chemiluminescence reagents (Amersham Pharmacia Biotech, Piscataway, NJ).

staining cells were compared between normal pituitary tissue and abnormal pituitary tissue (i.e. all tumors) with the Mann-Whitney U test. Comparisons between normal tissue and each type of tumor were carried out with the Kruskal-Wallis test. Spearman’s rank order correlation tests were performed to see whether there was any correlation between p27, P-p27, Jab1, and Ki-67 staining. Data on figures and in text are shown as mean ⫾ se (unless otherwise stated). Significance was taken at P less than 0.05.

Statistical analysis The statistical analysis was performed with Arcus Quickstat version 1.2 (I. Buchan; Addison Wesley Longman Ltd., Cambridge, UK). After the Shapiro-Wilk test showed the data to be nonnormally distributed, the data were analyzed with nonparametric tests as follows. For each antibody, the percentages of the combined strongly and moderately

Results p27 and P-p27 protein levels

p27 staining in human tonsil was shown primarily in the resting cells of the mantle zone, with little or no staining in

2638


the center of the follicle (Fig. 1A). This is an opposite pattern to the one observed with P-p27, in which the positive staining is seen in the rapidly dividing cells in the center of the follicle, whereas cells in the mantle zone showed very little P-p27 staining (Fig. 1D). There was no cross-reaction between P-p27 protein and p27 antibody, and p27 protein and phospho-p27 antibody (Fig. 1, A–F). The specific immunodetection of p27, phospho-p27, and Jab1 antibodies was tested against whole tissue extract from pituitary adenomas by immunoblotting. As shown in Fig. 2, when total protein extract from pituitary adenomas were separated, blotted, and then probed with Jab1, P-p27, and p27 antibodies, only a single band was recognized for each of the antibodies (38.5 kDa, Jab1; 27 kDa, p27). p27 staining was observed in the nuclei of the pituitary cells. In general, normal pituitary expressed high levels of p27, pituitary adenomas showed less p27 staining, and malignant transformation led to nearly complete loss of p27 immunoreactivity (Figs. 3A and 4). The exceptions to this pattern were the corticotroph adenomas, which expressed very low levels of p27 similar to pituitary carcinomas (19). Significantly more P-p27 was found in normal pituitary tissue than in abnormal tissue (P ⬍ 0.0001). Comparing individual tumor types with the normal pituitaries, adenomas showed significantly less P-p27 staining in all benign tumors except corticotroph tumors, which were similar to normals (70 ⫾ 5 vs. 60 ⫾ 5%, NS; Fig. 3B). The metastatic tumors showed a very low P-p27 staining index (11.6 ⫾ 7.3%), notably different to the corticotroph tumors (Fig. 3A). With regard to cytoplasmic vs. nuclear distribution, some variable cytoplasmic staining was observed, but did not differ among tissue types. An example of P-p27 staining is demonstrated in Fig. 4; the amount of nuclear P-p27 in the corticotroph tumor was high, but nuclear P-p27 staining in the metastatic pituitary carcinoma was relatively low. It is notable that the very low levels of p27 staining in corticotroph tumors is associated with levels of P-p27 similar to normal tissue, whereas the malignant tumors are characterized by low levels of both p27 and P-p27. In fact, the ratio between P-p27 and p27 staining showed a significant difference between the


TABLE 1. Ratio of P-p27 to p27 staining in normal and tumorous pituitary tissue P-p27/p27 staining (median)

Normal GH ACTH NFPA PRL FSH TSH Aggressive Metastatic

1.4 0.8 89.8 1.3 1.3 2.2 0.2 0.2 0.3

different tumor types (P ⫽ 0.002; Table 1). Specifically, corticotroph tumors had a significantly higher P-p27/p27 ratio (89.8) compared with metastatic tumors (0.3), aggressive tumors (0.2), or TSH-omas (0.2; this latter group contained only two samples), whereas metastatic tumors were not significantly different from normal tissue or GH-, ACTH-, FSH-, or PRL-secreting or nonfunctioning adenomas. Jab1 staining

In human tonsil, Jab1 staining was observed in the central, proliferating zone of the follicle (Fig. 1G), and this was blocked by pretreatment with Jab1 peptide (Fig. 1H). The distribution was similar to P-p27 and opposite to p27 (Fig. 1, D and A, respectively). The antibody was tested on pituitary tissue by Western blotting and confirmed the antigen to be the predicted size (see above and Fig. 2). In the pituitary, both nuclear and cytoplasmic staining with the Jab1 antibody varied widely in intensity. No significant differences were seen between normal and abnormal tissues as a whole. Jab1 staining was fairly homogeneous among the adenomas, but seemed to be less in prolactinomas (Fig. 3C). Pituitary carcinoma samples demonstrated slightly but significantly stronger nuclear Jab1 staining (P ⫽ 0.05). Representative images of Jab1 staining in a corticotroph adenoma and a metastatic carcinoma are shown in Fig. 4. There is more apparent cytoplasmic staining in this corticotroph tumor than in the metastatic tumor, but the differences in cytoplasmic staining among the different groups of samples (normal vs. abnormal or individual comparisons) were not statistically significant. Ki-67 labeling index

FIG. 2. Detection of Jab1 (lane 1), p27 (lane 2), and P-p27 (lane 3) in whole-tissue extract from a pituitary adenoma (38.5 kDa, Jab1; 27 kDa, p27). Protein lysates were separated and blotted onto a polyvinylidene difluoride membrane; immunoblotting (IB) was performed with the indicated antibodies. MW, Molecular weight.

In the control human tonsils, Ki-67 staining was observed primarily in the central zone of the follicle, with occasional positive cells in the mantle zone (Fig. 1I). In pituitary tissue, the Ki-67 labeling index was significantly higher in tumor samples compared with the normal tissue (P ⫽ 0.0089). Individual comparisons showed that the metastatic (2.6 ⫾ 0.6), aggressive (2.0 ⫾ 0.6), and corticotropinoma samples (2.4 ⫾ 1.0) had significantly higher labeling indices compared with normal pituitary tissue (0.64 ⫾ 0.1) (P ⫽ 0.0003, 0.03, and 0.009, respectively; Fig. 3D). The high levels of Ki-67 staining in a corticotroph adenoma and in a metastatic tumor are illustrated in Fig. 4. p27 staining of tumors showed a negative correlation with the Ki-67 labeling index (Spearman’s correlation coefficient, ⫺0.3; P ⫽ 0.004). No other correlation



FIG. 3. A, Nuclear p27 staining in pituitary tumors. Samples from patients with Cushing’s disease or metastatic pituitary tumors had significantly less nuclear p27 staining than did normal tumors or other tumor types (updated from Ref. 19). B, P-p27 staining in pituitary tumors. All samples showed significantly less P-p27 staining than normal pituitaries except corticotroph tumors. *, Significant difference (P ⬍ 0.05) compared with normal samples. C, Jab1 staining in pituitary tumors. PRL-secreting adenomas showed less Jab1 staining, and metastatic samples showed higher Jab1 protein expression. ⫹, P ⫽ 0.05 compared with normal samples. D, Ki-67 staining shows significantly higher labeling index in corticotroph tumors and metastatic tumors than normal or other adenomatous samples.

was observed between the different types of staining, grouping the samples either as normals and tumors or according to the individual hormone-secreting types. Within the normal samples, no differences were found in p27, P-p27, Jab1, or Ki-67 staining between tissue removed from patients with the clinical diagnosis of Cushing’s disease or other pituitary abnormalities (adenomas or arachnoid cysts). Similarly, no differences were observed in p27, P-p27, Jab1, and Ki-67 staining within the metastatic group according to the hormone-secreting status of the tumors (ACTH- vs. PRL-secreting or nonfunctioning tumors). Because clinically aggressive adenomas showed similar staining to nonaggressive adenomas, we repeated the calculations with the clinically aggressive adenomas (but not the carcinomas) regrouped into their relevant hormone-secreting group. The results were similar to the calculations shown above. p27 and Jab1 mRNA expression

p27 and Jab1 mRNA were detectable in every sample studied (Fig. 5, A and B). p27 and Jab1 mRNA expression was positively correlated in the overall pituitary tumor group (Fig. 5A; Spearman’s correlation coefficient, ⫹0.7; P ⬍ 0.0001). Within individual tumor subgroups, a significant positive correlation was found analyzing NFPAs (Spearman’s correlation coefficient, ⫹0.7; P ⬍ 0.0001) and PRLsecreting adenomas alone (Spearman’s correlation coeffi-

cient, ⫹0.9; P ⫽ 0.017). Similar trends were apparent in the other tumor types, but none attained statistical significance. Discussion

We found that both p27 and P-p27 were expressed in normal and abnormal tissue, generally at lower levels in adenomatous samples than in normal tissue. Moreover, the ratio of p27 to P-p27 was preserved in most cases, including the pituitary carcinomas, in which extremely low expression of both p27 and P-p27 was observed. However, a notable exception to this finding was seen in corticotroph tumors, in which P-p27 levels similar to normal tissue accompanied the low expression of p27. P-p27 was observed both in the cytoplasm and in the nucleus, but only nuclear staining showed differences between the different tumor types. Jab1 staining was also seen in both the nucleus and the cytoplasm. Although generally normal and abnormal pituitary samples showed similar Jab1 levels, metastatic pituitary carcinomas showed higher Jab1 staining compared with normal samples. Ki-67 staining was higher in pituitary carcinomas, aggressive tumors, and corticotroph adenomas, and correlated inversely with nuclear p27 immunostaining. p27 and Jab1 mRNA levels showed a positive correlation in both normal and abnormal samples. Oncogenic processes are likely to involve regulators of cell cycle progression through the G1 phase (1). One of the im-

2640



FIG. 4. Row 1, p27 staining in a corticotroph adenoma (left) and a metastatic tumor (right), both showing low level of nuclear p27 staining. Row 2, Strong phospho-p27 staining in a corticotroph adenoma (left); a metastatic tumor with little P-p27 staining (right). Row 3, A corticotroph adenoma with moderate amount of nuclear and cytoplasmic Jab1 staining (left); a metastatic tumor with strong nuclear and little cytoplasmic Jab1 staining (right). Row 4, Ki-67 staining in a corticotroph adenoma (left) and metastatic tumor (right) both with several Ki-67 positive cells. Magnification, ⫻1000.

portant functions of cyclin E/cyclin-dependent kinase (CDK) 2 is that it inactivates p27 via phosphorylation (15, 16). Phosphorylation at the Thr187 residue renders p27 available for degradation via the ubiquitin/proteasome pathway (16, 29). p27 thus can act as a substrate as well as an inhibitor for the cyclin E/CDK2 complex, depending on the relative concentrations of p27 and cyclin E/CDK2 (16, 30). The activation of cyclin E/CDK2 therefore initiates a positive feedback loop; once this checkpoint is passed, the cell cycle can progress without external stimulants. Although the transcriptional regulation of p27 is important in some situations (2, 11, 13, 14), the expression of p27 mRNA remains stable throughout the cell cycle; by contrast, its protein expression varies widely during the cell cycle, with high levels in G0 phase, declining levels in G1, and lowest levels in S, G2, and M, suggesting that p27 is primarily regulated post-transcriptionally. Several studies suggest that after phosphorylation, p27, like other cell cycle proteins, is conjugated to multiple ubiquitin mol-

ecules and degraded by the proteasome pathway (15, 23, 31). Recently, it has been suggested that early in the G1 phase a Thr187-independent pathway also exists that directs p27 to ubiquitination and degradation by the proteasome (32). Because ubiquitination and degradation probably occur in the cytoplasm, the P-p27 must be transported there to be degraded. One potential transporter is Jab1, an activator protein (AP)-1 complex coactivator protein that may accelerate p27 degradation by facilitating its phosphorylation and by bringing it to the degradation machinery in the cytoplasm (18, 33). However, other data suggest that cyclin E/CDK2 forms a trimeric complex with p27; in a concentrationdependent manner, this facilitates its ubiquitination and degradation by the proteasome within the nucleus without the need to be transported into the cytoplasm (29, 34). Jab1 is a 38 kDa protein and was first described as a coactivator of c-Jun and Jun D (17). The mechanism proposed for Jab-induced coactivation is the stabilization of c-Jun com-



FIG. 5. A, Correlation between p27 and Jab1 expression in pituitary tumors. Spearman’s correlation coefficient, 0.69; P ⬍ 0.0001. Normal samples are shown with open diamonds, NFPAs with filled diamonds, somatrotropinomas with crosses, lactotroph tumors with open circles, corticotropinoma samples with filled triangles, and gonadotroph tumors with open triangles. B, PCR with primers for p27, Jab1, and the housekeeping gene GAPDH show mRNA expression in an NFPA. M, Size markers.

plexes with AP-1 sites. In addition, Jab1 was found to be present in a novel protein complex, the Jab1 signalosome, that was identified through its association with the 26S proteasome (35). The Jab1 signalosome possesses kinase activity in vitro and can phosphorylate c-Jun and other proteins as well (IK␤␣ and p105) (35). Jab1 is present in both the cytoplasm and the nucleus and interacts with a number of factors. Activation of cell membrane receptor LFA-1, an integrin adhesion molecule, is followed by an increase in the nuclear pool of Jab1 paralleled by enhanced transactivation of an AP-1-dependent promoter (36). Using two-hybrid screens, a nuclear pore-associated protein, mNPAP60, has been identified that interacts with Arg90 on the P-p27 molecule and supports transport across the nuclear membrane (37). Mutant p27 at the mNPAP60 binding site is resistant to ubiquitin-mediated degradation (37). Thus, Jab1 and mNPAP60 might act in combination or in parallel to transport p27 out of the nucleus. Although the amount of Jab1 is tightly controlled in the cell (38), recent data suggest that the Jab1 level increases during tumor progression both in childhood neuroblastomas (39) and in breast cancer (Claret, F.-X., unpublished data), and also shows a tight correlation with cell cycle stages (Claret, F.-X., personal communication). These suggest that the higher Jab1 levels in our limited number of pituitary carcinoma cases are similar to other more malignant tumor types. Exactly how Jab1 interacts with p27 is unclear. Also, the precise mechanism by which Jab1 enhances p27 degradation (e.g. by facilitating phosphorylation of Thr187 or by bringing it to the degradation machinery in the cytoplasm) is uncertain; perhaps Jab1 simply shuttles p27 from the nucleus to the cytoplasm. Alternatively, because Jab1 is present in mam-

malian COP9 complexes (which seem to contain enzymes that add phosphate groups), it may also be involved in the phosphorylation of p27; Jab1 may itself also physically interact with components of the ubiquitin/proteasome system. Interestingly, ectopic expression of p27 inhibited the entry of BALB/c-3T3 cells into S phase. Among regulators of cell cycle progression, those specific for G1 phase are often found to be altered in human tumors (1), indicating that G1 regulation is closely connected with tumor suppression. The precise interplay between this Jab1 regulatory pathway and the relative role of Jab1 in tumors will require identification of the cellular proteins that associate with Jab1. p27-deficient mice develop pituitary adenomas arising from the intermediate lobe that stain positive for ACTH (3–5, 40), whereas a double knockout of p27 and p18 develops this abnormality earlier. Our own earlier study found no p27 sequence abnormalities apart from known polymorphisms in pituitary tumors; no loss of heterozygosity was observed, and no difference in p27 mRNA expression was seen in corticotropin-secreting pituitary tumors (8). However, we also showed that at the protein level significantly decreased p27 expression was observed generally in pituitary tumors compared with normals, and especially low levels in corticotroph adenomas and in metastatic pituitary tumors (19). This is clearly post-translational, because it is not mirrored by changes in p27 mRNA. These data correspond with those of others (9, 10, 20, 41– 43). We also showed that normal pituitary corticotrophs have a relatively lower p27 expression compared with other normal pituitary cell types (19). The p27 gene can also be silenced via methylation in pituitary cell lines, but this does not appear to be a characteristic of human pituitary tumors (11). The present studies also con-

2642 J Clin Endocrinol Metab, June 2002, 87(6):2635–2643

firm that any changes in p27 regulation in pituitary tumors are not secondary to alterations in mRNA. In the present study, we studied the expression of P-p27 and correlated it with p27 expression and Ki-67 labeling index. Our data suggest that pituitary tumors as a group show lower p27 and P-p27 and higher Ki-67 staining compared with the normal pituitary. Analyzing individual tumor types, metastatic tumors, as expected, showed low levels of p27 and higher levels of Ki-67. The P-p27 staining was similarly low compared with that of p27, lower than in other tumor types or in normal tissue. Aggressive adenomas showed higher Ki-67 labeling but did not show any differences in terms of p27 and P-p27 compared with their more benign counterparts. By contrast, corticotroph adenomas behaved considerably different from other benign pituitary adenomas. Corticotroph tumors have relatively high levels of P-p27, compared with other pituitary tumors, and indeed were comparable to normal pituitary. This is in contrast with the fact that Cushing’s disease samples showed significantly less p27 staining; the relatively high P-p27 staining suggests that increased phosphorylation is characteristic of corticotroph tumors. The high P-p27/p27 ratio observed in corticotropinomas might suggest that the balance between P-p27 and p27 is altered in corticotroph tumors, leading to increased inactivation of p27. Thus, the high labeling index indicative of increased proliferative activity is associated with low p27 expression in both corticotroph and malignant tumors, but the mechanism of p27 inactivation appears to differ in the two tumor types. In malignant tumors, little or no p27 (phosphorylated or not) appears to remain in the nucleus, being either degraded in situ or exported out of the nucleus before ubiquitination. By contrast, in corticotroph tumors there is extensive phosphorylation of p27, to a much greater extent than that seen in normal pituitary or in other pituitary adenomas. Ki-67 labeling index values are somewhat different between different laboratories. Our values in normal pituitary are higher than in some other publications, whereas the metastatic samples show lower indices (44 – 46). These differences could be caused by different sample fixation procedures, different specific antibodies, and/or differences in the method of counting. Cyclin E is known to phosphorylate p27, and we have shown slight but significant overexpression of cyclin E in corticotroph tumors (47). Our present data and that of others (45) on Ki-67 suggest that corticotroph adenomas have a higher proliferative index than other pituitary adenomas; this could be explained by the relative overexpression of cyclin E and consequent high levels of P-p27 and low p27 expression compared with their levels in other pituitary tumors. This may be a characteristic of corticotroph adenomas per se, although we cannot rule out the possibility that local hormonal factors may play a role in the decline of p27 because normal corticotrophs contain less p27 than other types of hormone-secreting normal pituitary cells (19). Such factors could include glucocorticoids; corticotroph cells contain an especially high concentration of glucocorticoid receptors (48), but their number is, if anything, increased in corticotroph tumors (49). However, we have recently demonstrated overexpression of the enzyme 11␤-hydroxy-dehydrogenase


type 2 in pituitary adenomas, which will decrease effective feedback and may thus increase proliferative activity (50). Thus, we have shown a number of different markers suggesting that corticotroph tumors are more proliferative than other pituitary tumors; they have a higher Ki-67 labeling index, lower p27 expression, and increased cyclin E expression (47), and this is supported by earlier studies in which increased proliferative behavior has been observed in corticotroph tumors (51, 52). On the basis of the p27, cyclin E, and Ki-67 data, together with earlier in vitro studies, we suggest that corticotroph adenomas are more proliferative than other types of benign pituitary tumors. This might seem in sharp contrast with the clinical finding of the majority of corticotroph adenomas being microadenomas. However, because high levels of ACTH result in high cortisol levels, the clinical symptoms are more rapidly recognized than other types of pituitary tumors, including GH- or PRL-secreting adenomas, and especially nonfunctioning adenomas. It seems likely that the exuberant clinical signs lead to earlier diagnosis and smaller tumors at imaging and at operation. In summary, our data suggest that low p27 protein expression characteristic of corticotroph and malignant pituitary tumors is associated with a marker of high proliferative activity (Ki-67), but the mechanism of p27 inactivation is different in the two tumor types. Corticotroph tumors show a high ratio of phosphorylated to unphosphorylated p27, possibly related to increased cyclin E expression. By contrast, malignant tumors show only very low levels of both p27 and P-p27 expression. In the corticotroph adenomas, the low p27 expression is unlikely to be related to excessive expression of the putative oncogene Jab1, whereas in malignant tumors this might be the case, and further studies are necessary to dissect the role of Jab1 in decreasing p27 levels. Acknowledgments We thank Lin Tiang (University of Texas M.D Anderson Cancer Center, Houston, TX) for technical assistance. We are grateful to Xin Lu (Ludwig Institute for Cancer Research, St. Mary’s Hospital, London, UK) for the p27 antibody and p27 blocking peptide and to Barbara Christy (Department of Cellular and Structural Biology, University of Texas Health Science Center, San Antonio, TX) for the Jab1 peptide for the blocking studies. Received October 9, 2001. Accepted January 22, 2002. Address all correspondence and requests for reprints to: Prof. A. B. Grossman, Department of Endocrinology, St. Bartholomew’s Hospital, West Smithfield, London EC1A 7BE, United Kingdom. E-mail: [email protected]. *These authors contributed equally to this work. †Present address: Institute of Endocrinology, Tashkent, Uzbekistan. This study was supported by the Medical Research Council (to M.K.), the Cancer Research Committee (to M.K. and A.B.G.), and the Joint Research Board of St. Bartholomew’s Hospital (to H.S.C.); NIH Grants 5P50CA83639 and CA90853-01A1; institutional research grants (IRG, IRS-BRS and PRS) (to F.-X.C.); and the Royal Society (to Y.U. and Z.K.).

References 1. Sherr CJ 1996 Cancer cell cycles. Science 274:1672–1677 2. Sgambato A, Cittadini A, Faraglia B, Weinstein IB 2000 Multiple functions of p27(Kip1) and its alterations in tumor cells: a review. J Cell Physiol 183:18 –27 3. Nakayama K, Ishida N, Shirane M, Inomata A, Inoue T, Shishido N, Horii I, Loh DY, Nakayama K 1996 Mice lacking p27(Kip1) display increased body


4.

5.

6. 7.

8.

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

20. 21. 22.

23.

24.

25. 26.

27.

28.

size, multiple organ hyperplasia, retinal dysplasia, and pituitary tumors. Cell 85:707–720 Kiyokawa H, Kineman RD, Manova-Todorova KO, Soares VC, Hoffman ES, Ono M, Khanam D, Hayday AC, Frohman LA, Koff A 1996 Enhanced growth of mice lacking the cyclin-dependent kinase inhibitor function of p27(Kip1). Cell 85:721–732 Fero ML, Rivkin M, Tasch M, Porter P, Carow CE, Firpo E, Polyak K, Tsai LH, Broudy V, Perlmutter RM, Kaushansky K, Roberts JM 1996 A syndrome of multiorgan hyperplasia with features of gigantism, tumorigenesis, and female sterility in p27(Kip1)-deficient mice. Cell 85:733–744 Fero ML, Randel E, Gurley KE, Roberts JM, Kemp CJ 1998 The murine gene p27Kip1 is haplo-insufficient for tumour suppression. Nature 396:177–180 Ponce-Castaneda MV, Lee MH, Latres E, Polyak K, Lacombe L, Montgomery K, Mathew S, Krauter K, Sheinfeld J, Massague J, Cordoncardo L 1995 p27Kip1: chromosomal mapping to 12p12–12p13.1 and absence of mutations in human tumors. Cancer Res 55:1211–1214 Dahia PLM, Aguiar RC, Honegger J, Fahlbush R, Jordan S, Lowe DG, Lu X, Clayton RN, Besser GM, Grossman AB 1998 Mutation and expression analysis of the p27/kip1 gene in corticotrophin-secreting tumours. Oncogene 16: 69 –76 Takeuchi S, Koeffler HP, Hinton DR, Miyoshi I, Melmed S, Shimon I 1998 Mutation and expression analysis of the cyclin-dependent kinase inhibitor gene p27/Kip1 in pituitary tumors. J Endocrinol 157:337–341 Ikeda H, Yoshimoto T, Shida N 1997 Molecular analysis of p21 and p27 genes in human pituitary adenomas. Br J Cancer 76:1119 –1123 Qian X, Jin L, Kulig E, Lloyd RV 1998 DNA methylation regulates p27kip1 expression in rodent pituitary cell lines. Am J Pathol 153:1475–1482 Tanaka C, Yoshimoto K, Yang P, Kimura T, Yamada S, Moritani M, Sano T, Itakura M 1997 Infrequent mutations of p27Kip1 gene and trisomy 12 in a subset of pituitary adenomas. J Clin Endocrinol Metab 82:3141–3147 Medema RH, Kops GJPL, Bos JL, Burgering BMT 2000 AFX-like Forkhead transcriptional factors mediate cell-cycle regulation by Ras and PKB through p27kip1. Nature 404:782–787 Vairo G, Soos TJ, Upton TM, Zalvide J, DeCaprio JA, Ewen ME, Koff A, Adams JM 2000 Bcl-2 retards cell cycle entry through p27(Kip1), pRB relative p130, and altered E2F regulation. Mol Cell Biol 20:4745– 4753 Vlach J, Hennecke S, Amati B 1997 Phosphorylation-dependent degradation of the cyclin-dependent kinase inhibitor p27Kip1. EMBO J 16:5334 –5344 Sheaff RJ, Groudine M, Gordon M, Roberts JM, Clurman BE 1997 Cyclin E-CDK2 is a regulator of p27Kip1. Genes Dev 11:1464 –1478 Claret FX, Hibi M, Dhut S, Toda T, Karin M 1996 A new group of conserved coactivators that increase the specificity of AP-1 transcription factors. Nature 383:453– 457 Tomoda K, Kubota Y, Kato JY 1999 Degradation of the cyclin-dependentkinase inhibitor p27Kip1 is instigated by Jab1. Nature 398:160 –165 Lidhar K, Korbonits M, Jordan S, Khalimova Z, Kaltsas G, Lu X, Clayton RN, Jenkins PJ, Monson JP, Besser GM, Lowe DG, Grossman AB 1999 Low expression of the cell cycle inhibitor p27 Kip1 in normal corticotroph cells, corticotroph tumors, and malignant pituitary tumors. J Clin Endocrinol Metab 84:3823–3830 Lloyd RV, Jin L, Qian X, Kulig E 1997 Aberrant p27kip1 expression in endocrine and other tumors. Am J Pathol 150:401– 407 Norton AJ, Jordan S, Yeomans P 1994 Brief, high-temperature heat denaturation (pressure cooking): a simple and effective method of antigen retrieval for routinely processed tissues. J Pathol 73:371–379 Fredersdorf S, Burns J, Milne AM, Packham G, Fallis L, Gillett CE, Royds JA, Peston D, Hall PA, Hanby AM, Barnes DM, Shousha S, O’Hare MJ, Lu X 1997 High level expression of p27(kip1) and cyclin D1 in some human breast cancer cells: inverse correlation between the expression of p27(kip1) and degree of malignancy in human breast and colorectal cancers. Proc Natl Acad Sci USA 94:6380 – 6385 Loda M, Cukor B, Tam SW, Lavin P, Fiorentino M, Draetta GF, Jessup JM, Pagano M 1997 Increased proteasome-dependent degradation of the cyclindependent kinase inhibitor p27 in aggressive colorectal carcinomas. Nat Med 3:231–234 Porter PL, Malone KE, Heagerty PJ, Alexander GM, Gatti LA, Firpo EJ, Daling JR, Roberts JM 1997 Expression of cell-cycle regulators p27Kip1 and cyclin E, alone and in combination, correlate with survival in young breast cancer patients. Nat Med 3:222–225 Mori M, Mimori K, Shiraishi T, Tanaka S, Ueo H, Sugimachi K, Akiyoshi T 1997 p27 expression and gastric carcinoma. Nat Med 3:593 Catzavelos C, Bhattacharya N, Ung YC, Wilson JA, Roncari L, Sandhu C, Shaw P, Yeger H, Morava-Protzner I, Kapusta L, Franssen E, Pritchard KI, Slingerland JM 1997 Decreased levels of the cell cycle inhibitor p27/kip1 protein: prognostic implications in primary breast cancer. Nat Med 3:227–230 Korbonits M, Chitnis MM, Norman D, Norman D, Rosenfelder N, Suliman M, Jones TH, Fabbri KN, Besser GM, Burrin JM, Grossman AB 2001 The release of leptin and its effect on hormone release from human pituitary adenomas. Clin Endocrinol (Oxf) 54:781–789 Korbonits M, Jacobs RA, Aylwin SJB, Burrin JM, Dahia PL, Monson JP, Honegger J, Fahlbush R, Trainer PJ, Chew SL, Besser GM, Grossman AB


29. 30. 31.

32. 33. 34. 35.

36. 37. 38. 39.

40.

41. 42. 43.

44. 45.

46. 47. 48.

49.

50.

51. 52.

1998 Expression of the growth hormone secretagogue receptor in pituitary adenomas and other neuroendocrine tumors. J Clin Endocrinol Metab 83: 3624 –3630 Montagnoli A, Fiore F, Eytan E, Carrano AC, Draetta GF, Hershko A, Pagano M 1999 Ubiquitination of p27 is regulated by Cdk-dependent phosphorylation and trimeric complex formation. Genes Dev 13:1181–1189 Nguyen H, Gitig DM, Koff A 1999 Cell-free degradation of p27(kip1), a G1 cyclin-dependent kinase inhibitor, is dependent on CDK2 activity and the proteasome. Mol Cell Biol 19:1190 –1201 Pagano M, Tam SW, Theodoras AM, Beer-Romero P, Del Sal G, Chau V, Yew PR, Draetta GF, Rolfe M 1995 Role of the ubiquitin-proteasome pathway in regulating abundance of the cyclin-dependent kinase inhibitor p27. Science 269:682– 685 Malek NP, Sundberg H, McGrew S, Nakayama K, Kyriakidis TR, Roberts JM 2001 A mouse knock-in model exposes sequential proteolytic pathways that regulate p27Kip1 in G1 and S phase. Nature 413:323–327 Scheffner M 1999 Moving protein heads for breakdown. Nature 398:103–104 Swanson C, Ross J, Jackson PK 2000 Nuclear accumulation of cyclin E/Cdk2 triggers a concentration-dependent switch for the destruction of p27Xic1. Proc Natl Acad Sci USA 97:7796 –7801 Seeger M, Kraft R, Ferrell K, Bech-Otschir D, Dumdey R, Schade R, Gordon C, Naumann M, Dubiel W 1998 A novel protein complex involved in the signal transduction processing similar to 26S proteasome subunits. FASEB J 12:469 – 478 Bianchi E, Denti S, Granata A, Bossi G, Geginat J, Villa A, Rogge L, Pardi R 2000 Integrin LFA-1 interacts with the transcriptional co-activator JAB1 to modulate AP-1 activity. Nature 404:617– 621 Muller D, Thieke K, Burgin A, Dickmanns A, Eilers M 2000 Cyclin Emediated elimination of p27 requires its interaction with the nuclear poreassociated protein mNPAP60. EMBO J 19:2168 –2180 Bounpheng MA, Melnikova IN, Dodds SG, Chen H, Copeland NG, Gilbert DJ, Jenkins NA, Christy BA 2000 Characterization of the mouse JAB1 cDNA and protein. Gene 242:41–50 Shen L, Tsuchida R, Miyauchi J, Saeki M, Honna T, Tsunematsu Y, Kato J, Mizutani S 2000 Differentiation-associated expression and intracellular localization of cyclin-dependent kinase inhibitor p27(KIP1) and c- Jun co-activator JAB1 in neuroblastoma. Int J Oncol 17:749 –754 Franklin DS, Godfrey VL, Lee H, Kovalev GI, Schoonhoven R, Chen-Kiang S, Su L, Xiong Y 1998 CDK inhibitors p18(INK4c) and p27(Kip1) mediate two separate pathways to collaboratively suppress pituitary tumorigenesis. Genes Dev 12:2899 –2911 Bamberger CM, Fehn M, Bamberger AM, Ludecke DK, Beil FU, Saeger W, Schulte HM 1999 Reduced expression levels of the cell-cycle inhibitor p27Kip1 in human pituitary adenomas. Eur J Endocrinol 140:250 –255 Qian X, Jin L, Grande JP, Lloyd RV 1996 Transforming growth factor-␤ and p27 expression in pituitary cells. Endocrinology 137:3051–3060 Jin L, Qian X, Kulig E, Sanno N, Scheithauer BW, Kovacs K, Young Jr WF, Lloyd RV 1997 Transforming growth factor-␤, transforming growth factor-␤ receptor II, and p27Kip1 expression in nontumorous and neoplastic human pituitaries. Am J Pathol 151:509 –519 Thapar K, Yamada Y, Scheithauer B, Kovacs K, Yamada S, Stefaneanu L 1996 Assessment of mitotic activity in pituitary adenomas and carcinomas. Endocrine Pathol 7:215–221 Mastronardi L, Guiducci A, Spera C, Puzilli F, Liberati F, Maira G 1999 Ki-67 labelling index and invasiveness among anterior pituitary adenomas: analysis of 103 cases using the MIB-1 monoclonal antibody. Am J Pathol 52:107–111 Turner HE, Wass JA 1999 Are markers of proliferation valuable in the histological assessment of pituitary tumours? Pituitary 1:147–151 Jordan S, Lidhar K, Korbonits M, Lowe DG, Grossman AB 2000 Cyclin D and cyclin E expression in normal and adenomatous pituitary. Eur J Endocrinol 143:R1–R6 Kovacs K, Rotondo F, Stefaneanu L, Kovais R, Rohoudo F, Stefenesnu L, Fereidooni F, Horvath E, Lloyd RV, Scheithauer BW 2001 Glucocorticoid receptor expression in nontumorous human pituitaries and pituitary adenomas. Endocrine Pathol, 11:267–275 Dahia PLM, Honegger J, Reincke M, Jacobs RA, Mirtella A, Fahlbusch R, Besser GM, Chew SL, Grossman AB 1997 Expression of glucocorticoid receptor gene isoforms in corticotropin-secreting tumors. J Clin Endocrinol Metab 82:1088 –1093 Korbonits M, Bujalska I, Shimojo M, Nobes J, Jordan S, Grossman AB, Stewart PM 2001 Expression of 11␤-hydroxysteroid dehydrogenase type 1 and 2 in the pituitary gland: induction of type 2 isozyme expression in corticotrophinomas. J Clin Endocrinol Metab 86:2728 –2733 Gulya´s M, Acs Z, Rappay G, Makara GB 1993 Corticotroph, somatotroph and mammotroph cell kinetics in the postnatal infant female rat. Histochemistry 100:503–507 Buchfelder M, Fahlbush R, Adams EF, Kiesewetter F, Thierauf P 1996 Proliferation parameters for pituitary adenomas. Acta Neurochir 65 (Suppl):18 –21