Immunohistochemical detection of transforming growth factor β receptor type I in papillary carcinoma (×250). A, Intense cytoplasmic expression is observed in nonneoplastic thyroid parenchyma (upper left) and papillary carcinoma (lower right). B, Strong receptor expression is seen in lymph node metastasis (right).
Immunohistochemical detection of transforming growth factor β receptor type II in papillary carcinoma (×250). A, Receptor expressed in nonneoplastic thyroid parenchyma (upper) and markedly reduced in adjacent papillary carcinoma (lower). No expression was observed in central areas of the tumor (B) and lymph node metastasis (C).
Immunohistochemical detection of p27kip in papillary carcinoma (×100). A, Nuclear expression is observed in nonneoplastic thyroid parenchyma (left). No expression is observed in papillary carcinoma (right). B, No expression is observed in the lymph node metastasis (lower half).
Muro-Cacho CA, Muñoz-Antonia T, Livingston S, Klotch D. Transforming Growth Factor β Receptors and p27kip in Thyroid Carcinoma. Arch Otolaryngol Head Neck Surg. 1999;125(1):76-81. doi:10.1001/archotol.125.1.76
To investigate the role of cell cycle regulators in the pathogenesis of papillary carcinoma of the thyroid.
Resistance to transforming growth factor β–mediated inhibition is a well-known pathogenic mechanism in epithelial neoplasias. In a retrospective study, the expression of transforming growth factor β receptors types I and II, cyclin D1, and the cyclin-dependent inhibitor p27kip, was analyzed by immunohistochemistry. Results were interpreted in the context of clinicopathological data. Patient follow-up ranged from 1 to 18 years, with a mean of 4 years.
Twenty conventional primary papillary carcinomas and their metastases were selected according to current pathologic criteria. Nonconventional papillary carcinomas (eg, tall-cell, columnar) were excluded from the analysis.
Cyclin D1 was expressed more intensely in the tumor than in adjacent nonneoplastic parenchyma. Within a given tumor, however, there was significant heterogeneity in expression intensity and percentage of positive cells, particularly in metastases. Type I receptors were strongly expressed in 90% of tumors, while 80% of the tumors revealed low to no expression of type II receptors. In 10% of tumors, type I receptors were absent and type II receptors expressed. Simultaneous absence of both receptors was not observed. While p27kip was strongly expressed in nonneoplastic thyroid, it was not detected in any of the primary tumors or their metastases.
The results strongly suggest that functional abnormalities in type II receptors result in increased levels of cyclin D1 and down-regulation of p27kip. This would maintain cells in a proliferative state and would promote tumor progression.
IN RECENT YEARS, a large body of evidence has accumulated with regard to the central role played by cell cycle regulatory mechanisms in many cancer types.1 The G1 phase of the cell cycle has proved to be a rate-limiting step beyond which cells are committed to division.2 Progression through this G1 checkpoint depends on the availability of specific growth factors and can be blocked by negative regulators such as transforming growth factor β (TGFβ). Current evidence indicates that the TGFβ receptors type I (TβR-I) and type II (TβR-II) are directly involved in TGFβ-mediated signal transduction. Signaling requires the formation of a heterotetramer containing 2 molecules each of TβR-I and TβR-II. A model for the mechanism of action of this complex has been proposed by Wrana and Attisano.3 According to this model, TβR-II is constitutively autophosphorylated, and ligand binding results in recruitment and phosphorylation of TβR-I. Kinase activity of TβR-I would then be responsible for the phosphorylation of unknown downstream substrates and the generation of biological responses.
Many tumors become insensitive to the growth-inhibitory actions of TGFβ by down-regulating TβR-II.4,5 This decrease in TβR-II expression correlates with tumorigenicity. For instance, the breast carcinoma cell line MCF7(-) does not express TβR-II, is resistant to TGFβ-mediated antiproliferative effects, and is tumorigenic. Transfection of this line with TβR-II complementary DNA increases the number of surface TβR-II, restores TGFβ sensitivity, and decreases tumorigenicity.6 Furthermore, TGFβ treatment of nontransformed epithelial cells blocks DNA synthesis, suppresses cyclin D1 messenger RNA (mRNA) synthesis and protein expression, and increases the expression of certain cyclin-dependent kinase inhibitors (p27kip1, p15ink4), resulting in G1 arrest.7
We have previously reported that cyclin D1 protein is overexpressed in thyroid papillary carcinomas and that the levels of cyclin D1 correlate with disease severity.8 In this article, we examine a mechanism for this cyclin D1 overexpression in thyroid papillary carcinomas, that is, the dysregulation of its expression by lack of TGFβ-mediated control. We report the absence of TβR-II and p27kip in primary papillary carcinomas of the thyroid and their metastases. These findings suggest a link between absence of TGFβ-mediated cell growth inhibition, abnormal G1 phase regulation, and thyroid oncogenesis.
Twenty conventional papillary carcinomas of the thyroid were selected from the tissue archives of the Pathology Department at the H. Lee Moffitt Cancer Center and Research Institute, Tampa, Fla. The diagnosis of papillary carcinoma was made, in hematoxylin-eosin–stained tissue sections, following standard criteria that included the presence of papillae, psammoma bodies, nuclear clearing, nuclear grooves, intranuclear pseudoinclusions, and fibrosis.9 Clinicopathological staging was done according to the recommendations of the American Joint Committee on Cancer10 as follows: T0, no evidence of primary tumor; T1, tumor measuring 1 cm or less; T2, tumor measuring more than 1 cm and less than 4 cm; T3, tumor measuring more than 4 cm; T4, tumor of any size extending beyond the thyroid capsule; N0, no regional lymph node metastasis; N1a, metastasis to ipsilateral cervical lymph nodes; N1b, metastasis to bilateral, midline, or contralateral lymph nodes; M0: no distant metastasis; M1, distant metastases; stage I, any T, any N, M0 in patients younger than 45 years, and T1 N0 M0 in patients older than 45 years; stage II, any T, any N, M1 in patients younger than 45 years, and T2 N0 M0 or T3 N0 M0 in patients older than 45 years; stage III, T4 N0 M0 and any T, N1, M0 in patients older than 45 years; and stage IV, any T, any N, M1 in patients older than 45 years. Tissue blocks containing both tumor and adjacent nonneoplastic thyroid parenchyma and blocks from regional lymph nodes containing metastatic disease were selected for immunohistochemical analysis in each of the 20 cases. The study followed institutional review board guidelines, without imposing added risk or morbidity to patients, and preserving patient confidentiality. All studies were performed at the Pathology Research Core Facility of the University of South Florida College of Medicine and H. Lee Moffitt Cancer Center and Research Institute.
Sections (5 µm) from formalin-fixed, paraffin-embedded tissues were cut and placed in poly-L-lysine–coated slides. The slides were subjected to deparaffination in xylene and hydration, through a series of decreasing alcohol concentrations, following standard procedures. To allow penetration of reagents, tissue slides were digested with 50-mg/mL proteinase K in sodium chloride sodium phosphate buffer (pH 7.4) at 37°C. After 30 minutes, the slides were heated and extensively washed in sodium phosphate buffer solution (PBS) (pH 7.4) to inactivate the proteinase. Endogenous peroxidase was quenched with a 3% solution of hydrogen peroxide for 20 minutes at 37°C. The slides were washed in deionized water for 5 minutes. Antigen retrieval was performed by placing the slides in a clear plastic container with vented top containing citrate buffer (0.1-mol/L citric acid, 4.5 mL; 0.1-mol/L sodium citrate, 21.5 mL; deionized water, 225 mL) in a microwave oven set on "high" for 2 for 5 minutes. The slides were allowed to cool for 10 to 20 minutes, rinsed extensively in deionized water, placed in PBS for 5 minutes, and drained. Blocking serum was applied, and the slides were incubated in a humid chamber for 20 minutes at room temperature. After blotting, the following primary antibodies were applied at room temperature at a 1:100 dilution: anti-cyclin D1 (C-terminal) polyclonal antibody (Immunotech Inc, Westbrook, Me); anti–TβR-II (C-terminal domain), anti–TβR-I, and anti-p27kip (Santa Cruz Biotechnology Inc, Santa Cruz, Calif). After 1 hour, the slides were rinsed in PBS for 5 minutes. For detection, the avidin-biotin complex method (Vectastain ABC Kit, Rabbit IgG, Elite series; Vector Laboratories Inc, Burlingame, Calif) was used following manufacturer's specifications. The biotinylated secondary antibody was applied for 20 minutes at room temperature in a humid chamber. At the end of this incubation, the slides were rinsed in PBS for 5 minutes, followed by the addition of the avidin-biotin complex. Slides were then incubated in a humid chamber for 30 minutes at room temperature and rinsed and placed in PBS for 5 minutes. Diaminobenzidine, prepared following manufacturer's specifications, was applied and color development monitored. When the desired intensity was reached (2-5 minutes), the slides were rinsed in water and counterstained with modified Mayer hematoxylin for 30 seconds. Slides were then washed in running water for 10 minutes, dehydrated, cleared, and mounted with resinous mounting medium. Antigen expression was classified as 0 (no expression), 1+ (<25% of cells), 2+ (25%-50% of cells), 3+ (50%-75% of cells), and 4+ (>75% of cells).
A summary of the data for all 20 papillary carcinomas of the thyroid examined is presented in Table 1. Cases are listed according to disease stage and increasing patient age at the time of initial clinical presentation. The intensity of immunohistochemical expression is listed independently, in each case, for normal thyroid, tumor, and metastasis.
To examine the relationship between overexpression of cyclin D1 and the status of TβR-II in thyroid papillary carcinomas, 20 primary tumors, their metastases, and corresponding normal tissue counterparts were subjected to immunohistochemical analysis. In agreement with our previous results in the thyroid, and those of others in other tumors,11,12 cyclin D1 expression was observed both in the nucleus and in the cytoplasm (data not shown). As reported previously,8 intensity of expression of cyclin D1 was higher in tumors than in adjacent normal thyroid. Within a given tumor, however, there was significant heterogeneity in expression intensity and percentage of positive cells, particularly in the metastasis. Southern blot analysis of thyroid papillary carcinomas overexpressing cyclin D1 protein did not reveal an increased number of copies or other alterations in the cyclin D1 gene (data not shown). This is in agreement with the results of Lazzereschi et al,13 who used the cyclin D1 gene to normalize their Southern blots, and did not observe genetic alterations.
Since increase in cyclin D1 expression could be due to a defect in signaling through TGFβ receptors,14 expression of TβR-I and TβR-II was evaluated in the same 20 thyroid papillary carcinomas. While both receptors were strongly expressed in normal thyroid tissue (Figure 1, A, and Figure 2, A) and TβR-I was strongly expressed in the vast majority of tumors (Figure 1, A) and their metastases (Figure 1, B), 80% of the primary tumors (Figures 2, A and B) and the vast majority of the metastases (Figure 2, C) revealed low to no expression of TβR-II. Of the 4 cases (20%) expressing TβR-II, 3 also showed receptor expression in the metastases. However, in 1 case, the primary tumor expressed TβR-II but the metastasis did not. In only 2 tumors (10%), TβR-I was absent and TβR-II expressed. In no case was the simultaneous absence of both receptors detected, indicating that redundancy is not necessary to escape growth inhibition control. These results are in agreement with those previously reported by Lazzereschi et al13 who reported decreased levels of TβR-II in nonpapillary thyroid carcinomas, and suggest that escape from TGFβ antiproliferative control might be a generalized mechanism in thyroid cancer.
In proliferating cells, p27kip is bound to active cyclin-cdk4 complexes. Stimulation of TGFβ induces a redistribution of p27kip from cyclin D-cdk4 to cyclin E-cdk2, resulting in inhibition of cdk-2.7,15 To investigate the role played by p27kip in papillary carcinoma, we have analyzed its expression by immunohistochemistry. As it has been reported in other tumor types,16 the location of p27kip, in thyroid papillary carcinoma, is predominantly nuclear. While p27kip is strongly expressed in normal thyroid, p27kip was not detected in any of the primary tumors or their metastases (Figure 3). These results are in agreement with previous reports in other neoplasias where p27kip expression was decreased or absent.16
Cell cycle regulatory mechanisms play a fundamental role in oncogenesis. Transition through the different phases of the cycle is the result of the sequential activation and deactivation of a family of cyclin-dependent kinases (CDKs). Enzymatic activity of CDK is regulated at 3 different levels: (1) binding to cyclin proteins, (2) activation of the cyclin-CDK complex by subunit phosphorylation, and (3) CDK inhibition.17 During the G1 phase, a critical checkpoint of the cell cycle, D-type cyclins (D1, D2, and D3) bind to cdk4 and/or cdk6, and phosphorylate the Rb protein, resulting in release of the E2F family of transcriptional activators and entry into the S-phase.18 Cells that overexpress cyclin D1 have a shortened G1 phase, are less dependent on growth factors, and are able to traverse the G0/G1 and/or the G1/S transition under conditions that limit the growth of normal cells.19 Overexpression of cyclin D1 through a variety of mechanisms has been reported in a variety of neoplasias.20 We have previously shown that cyclin D1 is overexpressed in papillary carcinoma of the thyroid, and that the levels of cyclin D1 correlate with disease severity.8 Our preliminary Southern blot and Northern blot analyses, and those of others,13 have failed to reveal gene amplifications, translocations, or increased mRNA transcription (data not shown). Here we have explored posttranslational mechanisms that may explain the overexpression of cyclin D1 in thyroid papillary carcinomas and its role in thyroid oncogenesis.
Transforming growth factor β is an important modulator of cellular proliferation and a potent growth inhibitor of many epithelial cells. It exerts its action at the G1 phase via specific TGFβ receptors (TβR). Despite normal levels of TGFβ, many tumors become insensitive to TGFβ-mediated, growth-inhibitory effects through abnormal regulation of TβR.4,14,21 Loss of this negative regulation has been implicated as a major contributor to the development of epithelial tumors. Transforming growth factor β blocks G1 progression by specifically suppressing cyclin D1 mRNA and protein expression, without affecting the levels of other D-type cyclins (D2 and D3).20 Furthermore, in the absence of TGFβ function, cyclin D-cdk4 complexes serve both as a p27kip reservoir and as down-regulators of its activity, allowing G1 progression.4,7 The activities of the cyclin-CDK complexes are negatively regulated by CDK inhibitors that, in response to a variety of growth-modulating signals, physically associate with and inhibit the activity of CDKs.17,22 The role played in cell cycle regulation by p27kip, as G1 inhibitor, has been highlighted in recent years. The p27kip protein, a CDK inhibitor expressed in most cells, is present in maximal amounts during the quiescent (G0) and prereplicative (G1) phases, and decreases as cells are induced to enter the cell cycle.7,23 Mutations in the p27kip gene and alterations of the p27kip locus, in chromosome 12p13, have been found in a small number of tumors.14 Expression of p27kip has been shown to have antiproliferative activity, in vivo, in a variety of tissues while absence or decreased expression of p27kip leads to hyperplasia and has been observed in both benign and malignant neoplasias.16
The marked decrease in expression of TβR-II and the absence of p27kip found in the present study indicate that the mechanisms responsible for regulation of the G1 phase of the cell cycle are severely impaired in papillary carcinoma of the thyroid. Similar findings have been described in intestinal adenomas, where lack of TβR-II has been observed concomitantly to increased expression of cyclin D1.20 Transforming growth factor receptor type I, the other component of the receptor heterodimer, is, however, readily detected both in normal thyroid cells and in papillary carcinoma. Our preliminary studies (unpublished results) have revealed point mutations in the TβR-II promoter in a small group of these papillary carcinomas. These mutations are accompanied by lack of TβR-II mRNA transcription. This suggests that, in the majority of papillary carcinomas, a selective abnormality in the function of the TβR-II5,24 may be primarily responsible for the lack of inhibitory control mediated by TGFβ, which is expressed and secreted by normal thyroid cells.25 We postulate that, in papillary carcinoma, abnormal expression of TβR-II results in lack of TGFβ-mediated inhibitory function, increased expression of cyclin D1, and down-regulation of p27kip. This would maintain cells in a proliferative state, promoting tumor progression.
Accepted for publication July 27, 1998.
This work was sponsored in part by the American Cancer Society, Atlanta, Ga (Institutional Research Grant 202) (project 6115-107 LO).
Presented as an abstract at the 40th Annual Meeting of the American Society for Head and Neck Surgery, Palm Beach, Fla, May 14, 1998.
Reprints: Carlos A. Muro-Cacho, MD, PhD, Pathology Department, H. Lee Moffitt Cancer Center and Research Institute, 12902 Magnolia Dr, Tampa, FL 33612 (e-mail: firstname.lastname@example.org).