Carcinogenesis vol.28 no.5 pp.932–939, 2007 doi:10.1093/carcin/bgl231 Advance Access publication November 24, 2006

The pituitary tumor-transforming gene promotes angiogenesis in a mouse model of follicular thyroid cancer Caroline S.Kim, Hao Ying, Mark C.Willingham1 and Sheue-yann Cheng
Laboratory of Molecular Biology, Center for Cancer Research, National Cancer Institute, 37 Convent Drive, Room 5128, Bethesda, MD 20892-4264, USA and 1Department of Pathology, Wake Forest University, Winston-Salem, NC 27157-3001, USA
To whom correspondence should be addressed. Tel: þ301 496 4280; Fax: þ301 402 8262; Email: [email protected]

Overexpression of the pituitary tumor-transforming gene (PTTG) has been associated with tumorigenesis. In a mouse model that spontaneously develops follicular thyroid cancer (FTC) with distant metastasis (TRbPV mouse), PTTG is overexpressed, similar to human thyroid cancer. To evaluate the role of PTTG in thyroid carcinogenesis, we studied the offspring of TRbPV mice with mice lacking PTTG (PTTG/ mice). The thyroids of TRbPV/PV PTTG/ mice were significantly smaller than TRbPV/PV mice. Ki-67 staining showed a decrease in thyroid proliferation in TRbPV/PV PTTG/ mice. Our evaluation of the Rb–E2F pathway, a central mediator of cell growth, found that TRbPV/PV PTTG/ mice exhibited a decrease in protein levels of phosphorylated Rb along with an elevation of the cdk inhibitor p21. Histological examination documented no difference in FTC occurrence between TRbPV/PV and TRbPV/PV PTTG/ mice, which indicates that PTTG removal does not prevent the initiation of FTC. However, TRbPV/PV PTTG/ mice had a significant decrease in vascular invasion and less development of lung metastasis as they progressively aged. CD31 staining also showed a decrease in vessel density in TRbPV/PV PTTG/ versus TRbPV/PV thyroids. Given the decreased vascular invasion in the PTTG knockout mice, we studied genes involved in angiogenesis. Real-time reverse transcription–polymerase chain reaction showed a consistent decrease in pro-angiogenic factors, fibroblast growth factor (FGF2), its receptor FGFR1 and vascular endothelial growth factor. Our results highlight the dual roles of PTTG as a regulator of thyroid growth and contributor to tumor progression. The separation of the pathways regulating cell proliferation, tumor initiation and tumor progression should direct future therapeutic options.

Introduction The pituitary tumor-transforming gene (PTTG) is a human securin protein and has a central role in maintaining sister chromatid cohesiveness during mitosis. The transition from metaphase to anaphase in mitosis is characterized by sister chromatid separation. The protein cohesin binds to the sister chromatids to prevent premature separation in metaphase. During anaphase, the protein separase degrades cohesin, which facilitates the progression from metaphase to anaphase. PTTG binds to separase to inhibit its function. Degradation of PTTG by the anaphase-promoting complex leads to an active separase and ultimately to the separation of sister chromatids. Alteration of normal levels of PTTG may disrupt this orderly process of chromatid segregation with subsequent aneuploidy in the daughter cells (1). PTTG’s known role in maintenance of chromosomal stability along with its roles as a transcription factor and in DNA repair suggest that it could also contribute to tumorigenesis. One group found that PTTG expresAbbreviations: FGF, fibroblast growth factor; FTC, follicular thyroid cancer; PTTG, pituitary tumor-transforming gene; PBS, phosphate-buffered saline; RTH, resistance to thyroid hormone; TSH, thyroid-stimulating hormone; VEGF, vascular endothelial growth factor. Published by Oxford University Press 2006.

sion was 1 of 17 genes that could be used to profile the metastatic potential of solid primary tumors (2). PTTG was first isolated from a rat pituitary tumor cell line (3) and its overexpression has been described in a variety of human tumors, including pituitary, thyroid, lung and colon cancer (4–6). Moreover, PTTG overexpression is able to transform NIH-3T3 cells to form tumors in nude mice (3). Some of PTTG’s oncogenic properties are via induction of angiogenic genes fibroblast growth factor (FGF-2) and vascular endothelial growth factor (VEGF), respectively (4,7), along with its transcriptional properties on p53 (8,9). Other groups report that up-regulation of PTTG results in an increase of the oncogene b-catenin (10). Although PTTG is considered an oncogene, there was also a report that overexpression of PTTG in the human cancer cell lines HeLa and A549 results in a decrease in cell growth along with up-regulation of the cdk inhibitor p21 (11). In addition, PTTG has been found by some to induce apoptosis through both p53-dependent and -independent pathways (12,13). Elucidation of PTTG’s role in tumorigenesis has been advanced through study of PTTG knockout mice. There exists at least three human isoforms of PTTG (PTTG 1–3) with various levels of abundance (14). PTTG1 is most often studied and associated with tumorigenesis. PTTG1 is normally found in testis and thymus, PTTG2 is considered to have a broader tissue distribution in organs such as the heart, pituitary and liver and PTTG3 is found in the kidneys and prostate (15). PTTG does not have significant homology to other classes of proteins, and given its role in mitosis, it was surprising to find that PTTG1 knockout mice (PTTG1/) are viable and fertile (15). PTTG1/ mice had smaller testes and spleens than wild-type mice, whereas mouse embryo fibroblasts from these mice had increased chromosomal instability with premature division of the centromeres (15). Both the under- and overexpression of PTTG may adversely affect chromosomal stability; however, PTTG1 alone is not necessary for maintenance of sister chromatid binding. Interestingly, when these knockout mice were crossed with the Rbþ/ mouse that develops pituitary tumors, the Rbþ/PTTG1/ mice had decreased incidence of pituitary tumor development compared with Rbþ/ mice (16). Further observations on these mice showed that deficiency of PTTG1 resulted in an increase of the cdk inhibitor p21 and that there was a reduction in cell proliferation. Donangelo et al. (17) have also recently studied transgenic mice that overexpress PTTG1 in the pituitary (a-GSU.PTTG) and found that these mice can develop pituitary microadenomas. In addition, crossing the a-GSU.PTTG mice with Rbþ/ mice resulted in the double-mutant mice developing greater pituitary tumors than Rbþ/ mice. These studies suggest that PTTG1 is a proto-oncogene in the pituitary and deletion of PTTG1 may delay or prevent tumor development. PTTG has also been associated with thyroid cancer. Thyroid cancer is the most common endocrine malignancy and the two most common types are papillary and follicular thyroid cancer (FTC). Although derived from the same epithelial follicular cell, these two types of thyroid cancer are distinct from each other in appearance, biological behavior and genetic causes. However, both papillary and FTC have been found to have elevated levels of PTTG at the mRNA level (18). One mechanism for PTTG’s role in thyroid tumorigenesis may be through promotion of genetic instability (5). In addition, microarray analysis comparing overexpression of PTTG in thyroid cell lines compared with the original line found that PTTG results in an increase in the angiogenic factors Id3 and a decrease in the antiangiogenic marker TSP-1 (19). Given the association of PTTG overexpression and thyroid cancer along with studies showing a reduction in tumor incidence when PTTG1 is ablated in mice, we focused on whether PTTG1 deletion would prevent thyroid tumorigenesis in a mouse model that develops thyroid

932

PTTG promotes angiogenesis in mouse model of FTC

cancer similar to human FTC (TRbPV mouse). Hereafter, the remainder of the paper will refer to PTTG1 as PTTG. The TRbPV mouse model expresses a dominant-negative mutation in the thyroid hormone receptor b. Human mutations in TRb are known to cause the syndrome resistance to thyroid hormone (RTH) where there is decreased tissue responsiveness to thyroid hormone. Interestingly, homozygous TRbPV mice spontaneously develop FTC with progression from capsular and vascular invasion to distant metastasis and anaplastic transformation (20). The similarities in FTC progression between this model and humans provide unique opportunities to evaluate the changes required for metastasis. Previous microarray analysis comparing the thyroids of TRbPV/PV versus wild-type mice identified genes differentially expressed in the mutant mice and found that PTTG was overexpressed over 5-fold (21). This study examines the in vivo effects of PTTG deletion in TRbPV mice. We found that the TRbPV/PV PTTG/ mice still developed FTC at the same rate as TRbPV/PV mice; however, there was less aggressive disease as seen by a reduction in vascular invasion and enhanced survival rate. Histological examination showed a reduction in thyroid size when PTTG was ablated and immunohistochemistry revealed that there was less cell proliferation in the thyroids of TRbPV/PV PTTG/ mice. We found that an up-regulation of p21 in the thyroids of TRbPV/PV PTTG/ mice may also contribute to a reduction in thyroid size along with a decrease in invasion. There was also less neovascularization in TRbPV/PV mice lacking PTTG that may be attributed to a decrease in the angiogenic factors FGF2, FGFR1 and VEGF. Given that PTTG deletion did not prevent FTC development, PTTG may be considered a contributor to thyroid tumorigenesis and tumor progression versus an initiating factor.

paraffin. For immunohistochemistry, after antigen retrieval using antigen unmasking solution (Vector Laboratories, Burlingame, CA) and incubation at 95°C for 20 min, 5 lm thick sections were blocked with 1% bovine serum albumin in phosphate-buffered saline (PBS) for 30 min followed by sequential incubations in rabbit anti-Ki-67 (NeoMarkers, Fremont, CA) or anti-CD31 (PharMingen/BD Biosciences). This was followed by indirect labeling with horseradish peroxidase-labeled goat anti-IgG (Jackson, ImmunoResearch, West Grove, PA) according to standard methods with counterstain using light green (Biomeda, Foster City, CA).

Materials and methods

Primary cell culture Primary thyroid cell lines were grown at 37°C, 5% CO2 atmosphere in primary cell media as described previously (23). Cells were then used for either western blot analysis or cell motility assay (see below).

Mouse strains All aspects regarding the care and handling of the animals used in this study were approved by the National Cancer Institute Animal Care and Use Committee. TRbPV mice were generated via homologous recombination, as described previously (22). Genotyping was performed using the primers and method as described previously (22). PTTG knockout mice were graciously provided by Dr Shlomo Melmed and the origin of these mice is described previously (15). Quantitative reverse transcription–polymerase chain reaction Total RNA was isolated from thyroid tissue as described previously using TRIzol (Invitrogen, Carlsbad, CA) as per manufacturer’s instructions. The LightCycler RNA amplification kit SYBR Green I (Roche, Indianopolis, IN) was used as per manufacturer’s instructions. Results were normalized to glyceraldehyde-3-phosphate dehydrogenase and analysis carried out as described previously (21). Primer sequences are as follows: FGF2—F 170– 190: CACACGTCAAACTACAACTCC and R 390–372: CCCAGTTCGTTTCAGTGCC; FGFR1—F: GTAGCTCCCTACTGGACATCC and R: GAGGCTACAAGGTTCGCTATGC; VEGF—F 909–928: AGGCGAGGCAGCTTGAGTTA and R 1245–1224: CTTGGCGATTTAGCAGCAGAT. Western blot Immunoblotting was performed as described previously with adjustments as described below (23). Thyroid lysate was prepared from age- and gendermatched mice using a final concentration of 50 mM Tris, 100 mM NaCl, 0.1% Triton X-100 and protease inhibitors (1 mM phenylmethylsulphonylfluoride, 0.3 lM aprotinin and 0.4 mM leupeptin 5 Complete Mini tab). A total of 50 lg of protein was loaded for each sample. Experiments were repeated three times using different mice. Primary antibodies used included Rb (sc-50, Santa Cruz Biotechnology, Santa Cruz, CA) and p21 (sc-6246, Santa Cruz Biotechnology). Valosin-containing protein (p97) antibody was a generous gift from Dr Lawrence Samelson (National Cancer Institute, Bethesda, MD). Secondary antibodies used included horseradish peroxidase-conjugated goat anti-mouse or anti-rabbit IgG (Amersham Biosciences, Piscataway, NJ) and visualized with the Western Lightning chemiluminescence reagent plus system (PerkinElmer Life Sciences, Boston, MA). The blots were stripped with ReBlot Plus (Chemicon, Temacula, CA) and normalized to a-Tubulin (1:6000 dilution; Sigma–Aldrich, St. Louis, MO). Immunohistochemistry Ki-67 staining was done as described previously (24). Briefly, thyroids were dissected and fixed in 10% neutral buffered formalin, and then embedded in

Flow cytometry analysis of thyroid tissue Preparation of cells from frozen thyroid tissue for cell cycle analysis was adapted from a protocol described previously (25). Briefly, thyroids were dissected from mice and frozen in liquid nitrogen. About 30 mg of thyroid tissue was thawed on ice. Afterwards, the thyroid sample was placed in a 40 lm cell strainer on top of a 50 ml tube (BD Falcon, Bedford, MA) and gently pushed against the strainer in 0.5 ml of PBS. Afterwards, the strainer was rinsed with 4 ml of additional PBS. The suspension was counted using a cell counter (Beckman Coulter, Fullerton, CA). Afterwards, the sample was centrifuged at 400g and re-suspended to a concentration of 1  107 to 2  107 cell/ml in PBS. Approximately 1  106 cells were aliquoted into a 12  75 mm Falcon tube (BD Falcon) and incubated in an ice bath for 15 min. Cells were fixed with 70% ethanol and kept at 4°C overnight. The next day, cells were centrifuged and the ethanol was decanted. The samples were incubated for 20 min at room temperature with 100 U DNAse-inactivated RNAse (Sigma–Aldrich). Propidium iodide was added at a concentration of 40 lg/ml (Molecular Probes, Eugene, OR) and incubated at 4°C for 30 min. Cell cycle analysis was performed using a FACSCalibur system (BD Biosciences). Data from 10 000 to 20 000 single-cell events per sample were collected. Cell cycle histograms were analyzed using the ModFit LT program (Verity Software House, Topsham, ME). All samples had low coefficient variation of the G0/G1 peak (coefficient variation between 3 and 6%).

Cell motility assay As described previously with the following adaptation (23), cells were incubated at 37°C for 8 h. Experiments were performed in duplicate. Data expressed are from combined experiments. Fold of increase or decrease in motility was calculated by dividing experimental % motility by control % motility. Hormone assays Determination of serum total T4 and total T3 (TT4 and TT3, respectively) were performed as described previously (26) using a Gamma Coat T4 and T3 Radioimmunoassay kit (Dia-Sorin, Stillwater, MN) according to the manufacturer’s instructions. Serum thyroid-stimulating hormone (TSH) levels were obtained using a method described previously (24). Statistical analysis Kaplan–Meier cumulative survival analysis was determined for TRbPV/PV and TRbPV/PV PTTG/ mice using GraphPad PRISM 4.0a (GraphPad Software, San Diego, CA). Thyroid function tests were determined using analysis of variance followed by Tukey’s multiple comparison test using GraphPad PRISM 4.0a software. Cell cycle differences were determined using analysis of variance and cell motility results were analyzed using the Student’s t-test where P , 0.05 was considered significant. One-sided analysis of variance was used to analyze the histological progression and thyroid weights.

Results Lack of PTTG does not significantly alter the pituitary–thyroid dysregulation in TRbPV/PV mice TRbPV/PV mice model the human syndrome RTH, which comprised decreased tissue responsiveness to thyroid hormone. The typical biochemical abnormalities in RTH patients include elevation of the thyroid hormones (T3 and T4) with a non-suppressed TSH. TRbPV/PV mice recapitulate the thyroid function test abnormalities of RTH patients, but the overall magnitude of difference compared with wildtype mice is magnified given the presence of two mutant TRb alleles (22,20). With the creation of a new double-mutant mouse (i.e. TRbPV/PV PTTG/ mouse), we proceeded to evaluate whether PTTG plays a significant role in the modulation of the pituitary–thyroid axis.

933

C.S.Kim et al.

Thyroid function tests were obtained for the different genotypes of mice at various ages (Figure 1). In Figure 1A, there is no appreciable difference in serum total T4 levels between wild-type and PTTG/ mice. Similar to prior studies, TRbPV/PV mice have extremely high levels of T4 compared with mice with wild-type TRb; this finding is not influenced by the presence or absence of PTTG (Figure 1A). The levels of the biologically active thyroid hormone, T3, are determined by both the direct production in the thyroid and conversion of T4 to T3 (27). Similar to T4, the total T3 levels between wild-type mice and PTTG/ mice, in either age group, were not different from each other (Figure 1B). However, in the older age group, the TRbPV/PV mice had higher levels of T3 than TRbPV/PV PTTG/ mice by 2-fold (Figure 1B). Despite the elevated T3 levels in the older age group, the TSH levels were not different between the TRbPV/PV and TRbPV/PV PTTG/ mice (Figure 1C). As similar serum levels of TSH exist in mice with or without PTTG, this suggests that the thyrotrope population in the pituitaries of PTTG-null mice is not significantly altered. Figure 2A-a supports this hypothesis, as there is no significant reduction in pituitary size in the pituitaries of TRbPV/PV PTTG/ mice compared with TRbPV/PV mice. These results suggest that deletion of PTTG does not significantly influence the pituitary–thyroid axis. However, consistent with that reported for PTTG/ mice (15), we found that in TRb wildtype mice, pituitary size of female wild-type mice was significantly reduced by the lack of PTTG (Figure 2A-b). Taken together, these results suggest that the mutation of the TRb gene could affect the action of PTTG in pituitary growth. Deficiency of PTTG affects thyroid growth As TSH is a major stimulus for thyrocyte growth and could confound the interpretation of FTC biology in TRbPV/PV mice, we next determined the thyroid weights in mice with or without PTTG. Given that no differences in TSH levels and pituitary size were found in our mice, we might also expect there to be no difference in thyroid size. Indeed, there was no difference in thyroid weights between wild-type TRb mice that express or do not express PTTG (TRbþ/þ PTTGþ/þ versus TRbþ/þ PTTG/, Figure 2B). We further evaluated the thyroid weights at different ages of TRbPV/PV and TRbPV/PV PTTG/ mice. Consistent with prior published studies, compared with the thyroid weights of wild-type mice (lanes 2 and 4), we see in Figure 2B that TRbPV/PV mice develop very large goiters that increase as the mice age (26,28; lanes 6, 8 and 10, Figure 2B). However, Figure 2B shows that across all ages, TRbPV/PV mice exhibit significantly larger thyroids compared with TRbPV/PV mice lacking PTTG (lanes 5, 7, 9 compared with lanes 6, 8 and 10, respectively). Since there were no significant differences in TSH levels between TRbPV/PV PTTGþ/þ and TRbPV/PV PTTG/ mice, these findings suggest that PTTG may be critical in times of active thyroid growth, for example, in response to increased levels of TSH, whereas PTTG may not be crucial for maintenance of physiologic thyrocyte growth. PTTG ablation decreases thyrocyte proliferation To understand the mechanism for the reduced thyroid size in TRbPV/PV PTTG/ versus TRbPV/PV mice when TSH levels are comparable, we evaluated thyrocyte proliferation status using Ki-67 staining. Figure 3A shows representative slides of thyroid tissues from the two groups. Panel A shows the negative control and panel B represents the amount of Ki-67 staining in normal thyroid tissue. As can be seen in panels C and D, there is an increase in Ki-67 in TRbPV/PV thyroids compared with wild type as represented by the arrows. However, there is a decrease in Ki-67 staining in TRbPV/PV PTTG/ thyroids compared with TRbPV/PV thyroids, which indicates a lower rate of proliferation in the absence of PTTG (panels E and F). Reduction in the rate of thyrocyte proliferation supports the findings of the decrease in gross thyroid weight. TRbPV/PV PTTG/ thyroids exhibit altered cell cycle parameters with up-regulation of p21 To assess whether there are any differences in the cell cycle of TRbPV/PV versus TRbPV/PV PTTG/ thyroids, we performed flow cytometry of

934

Fig. 1. Thyroid function tests between TRbPV/PV and TRbPV/PV PTTG/ mice show no significant differences in total T4, T3 and TSH levels (A–C) except for total T3 levels in the older mice. The bar represents the mean level. Between 5 and 10 mice in each genotype were determined at the ages indicated. Hormone level determination was performed as described in Materials and methods.

PTTG promotes angiogenesis in mouse model of FTC

Fig. 2. (A) Pituitary weights are compared between TRbPV/PV and TRbPV/PV PTTG/ mice (A-a) or between wild-type TRbPV/PV mice with or without PTTG (numbers of mice analyzed in each group are indicated). (B) Thyroid weights comparing TRbPV/PV and TRbPV/PV PTTG/ mice reveal a significant increase in the thyroid size of TRbPV/PV mice across all ages. The mean ages and the number of mice analyzed in each group were indicated. Statistical significance was determined at P, 0.05 using the Student’s t-test.

frozen thyroid tissue samples. Cell cycle analysis using frozen tissue specimens is advantageous in that it may provide a more representative picture of the thyroid in vivo. Figure 3B shows that there is a decrease of TRbPV/PV PTTG/ cells in G0/G1 with an increased percentage of thyrocytes in G2/M. The increase of PTTG-null cells in G2/M is consistent with other cell cycle studies of PTTG-null cell lines (15) and may correspond with the smaller thyroid weights seen in Figure 2B. We next examined known cell cycle regulators that may contribute to alteration of the cell cycle. PTTG has been shown to modulate p21 expression in the pituitary (16). Our western blot analysis of thyroid lysate from TRbPV/PV PTTG/ and TRbPV/PV PTTGþ/þ mice also showed an up-regulation of p21 in PTTG-deficient mice (lanes 4–6 compared with lanes 1–3; Figure 3C-a). One consequence of elevated p21 is inhibition of cdk (i.e. cdk2) and decreased phosphorylation of Rb, ultimately leading to impairment of transition from the G0/G1 phase to the S phase of the cell cycle. Indeed, we found a reduction of phosphorylated Rb in the thyroids of TRbPV/PV PTTG/ mice (lanes 4–6, Figure 3C-b) compared with TRbPV/PV thyroid (lanes 1–3, Figure 3C-b). A slight reduction in the unphosphorylated Rb was also observed in TRbPV/PV PTTG/ mice. The reduced phosphorylation of Rb may lead to a decrease in thyrocyte proliferation. Thus, the reduction in thyroid growth may be mediated in part by the up-regulation of p21.

Ablation of PTTG does not reduce FTC occurrence but alters FTC aggressiveness Although PTTG deletion negatively affects thyroid proliferation in TRbPV/PV mice, these mice still develop goiters. With prior reports describing a protective effect of PTTG ablation on pituitary tumor development, we next focused on FTC tumorigenesis in TRbPV/PV PTTG/ mice. The criteria for a diagnosis of FTC includes capsular or vascular invasion present in the thyroid. The histopathological analysis of the thyroids and of the heart and lungs from mice that were killed due to moribund status, beginning at 4–5 months of age in both groups, show that both TRbPV/PV and TRbPV/PV PTTG/ mice develop hyperplasia at a young age (Figure 4A). Hyperplasia is consistent with the increased thyroid weights and elevated TSH levels, but this is not a criterion for FTC. Figure 4A shows that there is no difference in the occurrence of capsular invasion, which signifies that both groups of mice develop FTC. These findings suggest that PTTG ablation is not protective against the development of FTC which is in contrast to prior reports of PTTG deletion preventing pituitary tumor development (16). Interestingly, there is a significant decrease in vascular invasion seen in TRbPV/PV PTTG/ mice. As FTC spreads hematogenously, the reduced vascular invasion may lead to decreased metastasis. Figure 4A shows that there is a trend toward decreased lung metastasis. Additional analysis of the survival rates comparing

935

C.S.Kim et al.

these groups support the premise that PTTG may influence the metastatic potential of FTC. That thyroid tumor progression is delayed is further supported by the survival curve in which a significant reduction in mortality in TRbPV/PV PTTG/ mice (median age 367 versus 267 days, P , 0.05) was observed (Figure 4B) at the age when thyroid growth was significantly rapid (9 months of age; Figure 2B) as compared with that of TRbPV/PV PTTGþ/þ mice. PTTG deficiency decreases neovascularization in the thyroid The decrease in vascular invasion and reduction in lung metastasis in TRbPV/PV PTTG/ mice suggests that PTTG may alter the balance of angiogenic factors in the thyroid. The thyroid is a very vascular organ and to evaluate whether any differences in the vascular pattern

Fig. 3. (A) Immunohistochemical staining of Ki-67 in representative thyroids from wild-type, TRbPV/PV and TRbPV/PV PTTG/ mice. Thyroids were dissected from gender- and age-matched mice and fixed and embedded in paraffin. Ki-67 detection was done using horseradish peroxidase as described in Materials and methods. Panel A is a control, whereas panel B represents wild-type thyroid. Panels C and D show Ki-67 staining in TRbPV/PV, seen in brown, highlighted by the arrows. Panels E and F are representative sections from TRbPV/PV PTTG/ thyroids. Thyroid sections from TRbPV/PV mice have increased Ki-67 staining compared with TRbPV/PV PTTG/ mice. (B) Flow cytometry from frozen thyroid tissue shows a delay in G2/M phase in TRbPV/PV PTTG/ mice represented by the white bars, compared with TRbPV/PV (black bars). Four age- and gender-matched mouse thyroids from each genotype were analyzed. Statistical significance was determined at P , 0.05 using analysis of variance. (C) Western blot analysis showing an increase in p21 (3C–a) and decrease in Rb and phosphorylated Rb in TRbPV/PV PTTG/ compared with TRbPV/PV thyroids (3C-b) For p21

936

Fig. 4. Histopathological comparison of thyroids and the heart and lungs from TRbPV/PV and TRbPV/PV PTTG/ mice (black and white bars, respectively); N 5 19–26 mice for thyroid and 15–19 heart and lungs examined from each genotype. Tissue sections from moribund mice were collected, fixed in formalin, embedded in paraffin and stained by hematoxylin and eosin for examination. Both groups develop FTC as defined by either the presence of capsular or vascular invasion. TRbPV/PV mice have an increased incidence of vascular invasion while there is a trend towards increased lung metastasis. Data are expressed as the percentage of occurrence of the total number of mutant mice examined in each genotype. (B) Kaplan–Meier survival curves for TRbPV/PV and TRbPV/PV PTTG/ mice (black triangles and open squares, respectively). TRbPV/PV PTTG/ mice had a statistically significant increase in survival (P , 0.05) as determined by analysis of variance with Tukey’s post hoc analysis. In all, 32 TRbPV/PV PTTG/ mice and 26 TRbPV/PV mice were followed.

analysis, three thyroids from each genotype of age- and gender-matched mice were evaluated. P21 is increased in the thyroids of mice lacking PTTG1. a-Tubulin was used as a loading control. Evaluation of three mice in each genotype for Rb and its phosphorylated form reveals a decrease in TRbPV/PV PTTG/ mice.

PTTG promotes angiogenesis in mouse model of FTC

Fig. 5. (A) Immunohistochemical analysis of vessel density in the thyroids of TRbPV/PV PTTG/ and TRbPV/PV mice. Thyroid specimens were prepared as described in Materials and methods. Antibody to CD31 (PECAM1) was used to visualize endothelial cells as shown in brown. TRbPV/PV thyroids (panel A) show an increase in CD31 staining compared with TRbPV/PV PTTG/ thyroids (panel B). Shown are representative samples of an age- and gender-matched pair. A total of three pairs were evaluated. (B) Angiogenic factors increased in the thyroids of TRbPV/PV mice. Using quantitative reverse transcription–polymerase chain reaction, PTTG downstream target gene, FGF-2, its receptor FGFR1, and VEGF are decreased in mouse thyroids lacking PTTG. (C) Western blot analysis of thyroid for the protein p97, a marker for vascular invasion in human FTC, shows a decrease in p97 levels in TRbPV/PV PTTG/ mice. A total of three mice in each genotype were used for RNA and protein analysis.

exist in PTTG-deficient thyroids, we performed immunohistochemistry using the endothelial marker CD31. Figure 5A shows that TRbPV/PV PTTG/ have less staining for CD31, compared with TRbPV/PV PTTGþ/þ thyroids. These results may help to explain the observed decrease in vascular invasion. To understand which angiogenic factors may be influential in the thyroid, we performed real-time reverse transcription–polymerase chain reaction analysis on downstream targets of PTTG, FGF2, its receptor FGFR1 and VEGF that are known to contribute to angiogenesis. Figure 5B shows that in the thyroids of the mice lacking PTTG have reduced levels of FGF2 (Figure 5B-a), FGFR1 (Figure 5B-b) and VEGF (Figure 5B-c). Another marker for vascular invasion in human FTC is the valosin-containing protein, which is known as p97 (29). It is also decreased in the thyroids of TRbPV/PV PTTG/ mice compared with TRbPV/PV thyroids (Figure 5C) and corroborates the hypothesis that deletion of PTTG in FTC tumors may reduce its invasive ability. PTTG may alter cell motility The significant reduction in vascular invasion suggests that PTTG could be involved in cell migration. Our histology showed that there

was a trend toward decreased lung metastasis (Figure 4A). The downstream effectors of PTTG are being investigated, but given PTTG’s multifunctional properties, it may also regulate cell motility. To investigate whether PTTG ablation could be a disadvantage for cell migration, we performed in vitro motility assays using cultured primary thyroid tumor cells derived from both TRbPV/PV PTTG/ and TRbPV/PV mice. Figure 6 shows that the cell line from TRbPV/PV PTTG/ mice moved at a slower rate than those derived from TRbPV/PV mice, indicating that PTTG has effects on cell motility independent of angiogenesis and cell proliferation. Discussion PTTG is considered a proto-oncogene and its overexpression is associated with many tumors, including human thyroid cancer. There are many mechanisms through which PTTG contributes as an oncogene, including the areas of cell proliferation and angiogenesis. Previous reports found that in pituitary tumorigenesis in the mouse, PTTG is a critical factor as the ablation of PTTG was protective against pituitary tumor development (16). In addition, PTTG overexpression was

937

C.S.Kim et al.

Fig. 6. TRbPV/PV mouse and TRbPV/PV PTTG/ mouse thyroids were removed and prepared into a single-cell suspension as described in Materials and methods. Cell proliferation rate for cell lines was determined by seeding cells in a 12-well plate and counting for 4 days. Cell motility experiments were performed by placing the cells in 8 lm pore transwells in 0.1% serumcontaining media at the top and primary cell media on the bottom (10% serum). The cells were incubated for 5 h; two independent experiments using cell lines with a similar growth rate were performed in triplicate. Data presented as the mean ± SEM (n 5 6),
P, 0.05

sufficient to promote pituitary microadenomas in mice, suggesting an initiating role for PTTG in pituitary tumor development (30). To understand the molecular basis of thyroid cancer, we have utilized the TRbPV/PV mouse model that spontaneously develops FTC similar to humans. The genetic changes responsible for FTC development and behavior are not well known and mutations such as the fusion Pax8–PPARc rearrangement are not entirely specific for FTC (31). Elevation of PTTG in human thyroid cancers has been associated with an increase in pro-angiogenic genes and chromosomal instability (19). As these studies suggest that PTTG may have an important role in FTC tumorigenesis, we proceeded to determine the effects of PTTG ablation in the TRbPV/PV mouse model. Our results show that TRbPV/PV PTTG/ mice are viable. As it is not known whether PTTG has a direct effect on TRb signaling pathways, we evaluated the thyroid function tests of TRbPV/PV and TRbPV/ PV PTTG/ mice to determine if there is any significant alteration in the regulation of the pituitary–thyroid axis. We found that both groups of mice had similarly elevated thyroid hormone levels along with a non-suppressed TSH. There was no statistically significant difference in serum total T4 or TSH levels at two different time points suggesting that PTTG may not directly contribute to the negative regulation of TSH by TRb. There are no prior reports on the thyroid function tests of PTTG/ mice, although they were found to have smaller pituitaries compared with wild-type mice (16). Recent creation of mice expressing human PTTG under the control of the asubunit glycoprotein hormone (a-GSU) promoter resulted in PTTG overexpression in the pituitary (a-GSU.PTTG mouse). The aGSU.PTTG mice developed pituitary hyperplasia and microadenomas (30). Interestingly, some of these transgenic mice overexpressed PTTG in pituitary cells that co-stained for TSH, but there was no elevation of serum TSH or total T4 levels (30). These findings suggest that PTTG may not have a prominent role in TSH secretion. Similar levels of TSH in TRbPV/PV and TRbPV/PV PTTG/ mice also suggest that TSH is not sufficient to account for any observed differences in the thyroid, but that PTTG may have more direct effects on the thyroid. There is currently no data suggesting that PTTG may act as a coregulatory protein for nuclear receptors. Our studies do reveal a difference in serum T3 levels in the older TRbPV/PV versus TRbPV/PV PTTG/ mice. It is unclear as to the cause as TSH levels are not significantly different in TRbPV/PV with or without PTTG. This may be due to the larger thyroid size in the TRbPV/PV mice that may coincide with increased T3 production or increased conversion of T4 to

938

T3 by tumor. One group found increased T4 to T3 conversion in human FTC patients due to overexpression of type 2 deiodinase (32). Future investigation into whether PTTG affects deiodinase activity may clarify this issue, but the important point for our studies is that TSH is not significantly different between the two groups. The decrease in thyroid size in the TRbPV/PV PTTG/ mice was seen at an early age and is unlikely to be attributed to the serum TSH levels. PTTG’s effects on cell cycle regulation are more likely to impact thyroid growth. We found on evaluation of cell cycle regulators that thyroids without PTTG have an increase in p21 levels with concomitant decrease in phosphorylated Rb. The present study therefore identified PTTG as a novel regulator for thyroid growth. It is important to note that although the thyroids of TRbPV/PV PTTG/ mice grew slower, the reduced proliferation did not prevent the development of FTC. These observations suggest that PTTG-mediated proliferation may not directly contribute to the development of thyroid carcinogenesis in TRbPV/PV mice. PTTG removal did result in a decrease in vascular invasion and eventually to reduced lung metastasis. Recent studies showed that prevention of goiter development could be accomplished by inhibiting the VEGF, angiopoietin, and fibroblast growth signaling pathways (31). There is also growing evidence that PTTG may activate proangiogenic factors such as FGF2 and VEGF (19). Consistent with these observations, our TRbPV/PV PTTG/ mice also showed a significant reduction of FGF2, its receptor FGFR1 and VEGF at the mRNA level. The protein p97 is being evaluated as a marker for FTC recurrence as it is another angiogenic marker that is decreased in the TRbPV/PV PTTG/ mice. P97 is well known to regulate endoplasmic reticulum-associated degradation of proteins by removing substrates from the endoplasmic reticulum and transferring them to the cytosol for proteasomal degradation. Although p97 can interact with multiple proteins, there is no current data showing an interaction between p97 and PTTG. Future study on how PTTG may alter p97 expression with resultant alteration of the vascular micro-environment may aid in understanding FTC pathogenesis. The phenotype of a decrease in vessel density seen by immunohistochemistry along with a reduction in vascular invasion and metastasis in PTTG-null mice suggests that PTTG removal may impede the ability for FTC to escape outside the thyroid. The multifunctional properties of PTTG in contributing to tumorigenesis may also include its ability to affect cell motility. Our cell motility experiments show that the ablation of PTTG impaired motility. Identification of the downstream targets of PTTG that may directly influence cell motility may also help our understanding of PTTG’s role in modulating the metastatic potential of FTC. Our in vivo study of PTTG ablation in the thyroids of TRbPV/PV mice, which are predisposed to develop thyroid cancer, found a decrease in thyrocyte growth and increase in p21 and alteration of the Rb–E2F pathway. These findings are similar to another group that studied the effects of PTTG removal in the Rbþ/ mouse that develops pituitary tumors (16). However, there are also tissue-specific effects of PTTG ablation in the thyroid with the most striking being that deletion of PTTG was not protective against FTC tumorigenesis. Our findings are in contrast to the prior study showing that PTTG reduces the occurrence of pituitary tumors. Part of these findings may be attributed to the other mutation present in each respective strain (i.e. Rb heterozygous versus dominant-negative TRb). Our results suggest that deletion of PTTG in TRbPV/PV mouse thyroids decreases cell proliferation, up-regulates p21, reduces angiogenic factors such as FGF2 and leads to an overall improvement in survival. Our findings suggest that PTTG may be a valid target for retarding both thyrocyte growth and FTC progression. The TRbPV mouse provides a valuable preclinical model to investigate the initiators and contributors to FTC tumorigenesis.

Acknowledgements We thank Dr Shlomo Melmed for the PTTG/ mice and Dr Lawrence Samelson for his p97 antibody. This research was supported, in part, by the

PTTG promotes angiogenesis in mouse model of FTC

Intramural Research Program of the National Institutes of Health, National Cancer Institute, Center for Cancer Research.

References 1. Zou,H. et al. (1999) Identification of a vertebrate sister-chromatid separation inhibitor involved in transformation and tumorigenesis. Science, 285, 418–422. 2. Ramaswamy,S. et al. (2003) A molecular signature of metastasis in primary solid tumors. Nat. Genet., 33, 49–54. 3. Pei,L. et al. (1997) Isolation and characterization of a pituitary tumortransforming gene (PTTG). Mol. Endocrinol., 11, 433–441. 4. Zhang,X. et al. (1999) Structure, expression and function of human pituitary tumor transforming gene (PTTG). Mol. Endocrinol., 13, 156–166. 5. Kim,D. et al. (2005) Pituitary tumor transforming gene (PTTG) induces genetic instability in thyroid cells. Oncogene, 24, 4861–4866. 6. Heaney,A.P. et al. (2000) Expression of pituitary-tumour transforming gene in colorectal tumours. Lancet, 355, 716–719. 7. Ishikawa,H. et al. (2001) Human pituitary tumor-transforming gene induces angiogenesis. J. Clin. Endocrinol. Metab., 86, 867–874. 8. Jallepalli,P.V. et al. (2001) Securin is required for chromosomal stability in human cells. Cell, 105, 445–457. 9. Hamid,T. et al. (2004) PTTG/securin activates expression of p53 and modulates its function. Mol. Cancer, 3, 18. 10. Zhou,C. et al. (2005) Overexpression of human pituitary tumor transforming gene (hPTTG) is regulated by beta-catenin/TCF pathway in human esophageal squamous cell carcinoma. Int. J. Cancer, 113, 891–898. 11. Mu,Y.M. et al. (2003) Human pituitary tumor transforming gene (hPTTG) inhibits human lung cancer A549 cell growth through activation of p21 (WAF1/CIO1). Endocr. J., 50, 771–781. 12. Bernal,J.A. et al. (2002) Human securin interacts with p53 and modulates p53-mediated transcriptional activity and apoptosis. Nat. Genet., 32, 306–311. 13. Yu,R. et al. (2000) Pituitary tumor transforming gene (PTTG) regulates placental JEG-3 cell division and survival: evidence from live cell imaging. Mol. Endocrinol., 14, 1137–1146. 14. Chen,L. et al. (2000) Identification of the human pituitary tumor transforming gene (hPTTG) family: molecular structure, expression and chromosomal localization. Gene, 248, 41–50. 15. Wang,Z. et al. (2001) Mice lacking pituitary tumor transforming gene show testicular and splenic hypoplasia, thymic hyperplasia, thrombocytopenia, aberrant cell cycle progression, and premature centromere division. Mol. Endocrinol., 15, 1870–1879. 16. Chesnokova,V. et al. (2005) Pituitary hypoplasia in PTTG/ mice is protective for Rbþ/ pituitary tumorigenesis. Mol. Endocrinol., 19, 2371–2379. 17. Donangelo,I. et al. (2006) PTTG over-expression facilitates pituitary tumor development. Endocrinology, 147, 4781–4791.

18. Heaney,A.P. et al. (2001) Transforming events in thyroid tumorigenesis and their association with follicular lesions. J. Clin. Endocrinol. Metab., 86, 5025–5032. 19. Kim,D.S. et al. (2006) Pituitary tumor transforming gene (PTTG) regulates multiple downstream angiogenic genes in thyroid cancer. J. Clin. Endocrinol. Metab., 91, 1119–1128. 20. Suzuki,H. et al. (2003) Compensatory role of thyroid hormone receptor (TR)a1 in resistance to thyroid hormone: study in mice with a targeted mutation in the TRb gene and deficient in TRa1. Mol. Endocrinol., 17, 1647–1655. 21. Ying,H. et al. (2003) Alterations in genomic profiles during tumor progression in a mouse model of follicular thyroid carcinoma. Carcinogenesis, 24, 1467–1479. 22. Kaneshige,M. et al. (2000) Mice with a targeted mutation in the thyroid hormone b receptor gene exhibit impaired growth and resistance to thyroid hormone. Proc. Natl Acad. Sci. USA, 97, 13209–13214. 23. Kim,C.S. et al. (2005) AKT activation promotes metastasis in a mouse model of follicular thyroid carcinoma. Endocrinology, 146, 4456–4464. 24. Furumoto,H. et al. (2005) An unliganded thyroid hormone b receptor activates the cyclin D1/cyclin-dependent kinase/retinoblastoma/E2F pathway and induces pituitary tumorigenesis. Mol. Cell. Biol., 25, 124–135. 25. Zitzmann,S. et al. (2002) Arginine-glycine-aspartic acid (RGD)-peptide binds to both tumor and tumor-endothelial cells in vivo. Cancer Res., 62, 5139–5143. 26. Ying,H. et al. (2005) Dual functions of the steroid hormone receptor coactivator 3 in modulating resistance to thyroid hormone. Mol. Cell. Biol., 25, 7687–7695. 27. Zavacki,A.M. et al. (2005) Type 1 iodothyronine deiodinase is a sensitive marker of peripheral thyroid status in the mouse. Endocrinology, 146, 1568–1575. 28. Suzuki,H. et al. (2002) Mice with a mutation in the thyroid hormone receptor b gene spontaneously develop thyroid carcinoma: a mouse model of thyroid carcinogenesis. Thyroid, 12, 963–969. 29. Yamamoto,S. et al. (2005) Increased expression of valosin-containing protein (p97) is correlated with disease recurrence in follicular thyroid cancer. Ann. Surg. Oncol., 12, 925–934. 30. Abbud,R.A. et al. (2005) Early multipotential pituitary focal hyperplasia in the alpha-subunit of glycoprotein hormone-driven pituitary tumor-transforming gene transgenic mice. Mol. Endocrinol., 19, 1383–1391. 31. Ramsden,J.D. et al. (2005) Complete inhibition of goiter in mice requires combined gene therapy modification of angiopoietin, vascular endothelial growth factor, and fibroblast growth factor signaling. Endocrinology, 146, 2895–2902. 32. Kim,B.W. et al. (2003) Overexpression of type 2 iodothyronine deiodinase in follicular carcinoma as a cause of low circulating free thyroxine levels. J. Clin. Endocrinol. Metab., 88, 594–598. Received August 24, 2006; revised November 06, 2006; accepted November 17, 2006

939