Abstract
It has been demonstrated that transforming growth factor-β (TGFβ) and other members of TGFβ superfamily play an important role in thyroid proliferative diseases. The deficiencies of SMAD4 are responsible to accelerate the malignant progression of neoplastic lesions in several types of tumors. Therefore, the objective of the present study was to determine the functional role of reduced expression of SMAD4 in human papillary thyroid carcinogenesis. For this purpose, we examined the TGFβ response in two cell lines, TPC-1 and BCPAP. Our data demonstrated for the first time that these cells showed a strong reduction in the level of SMAD4 protein, which was responsible for an alteration of TGFβ signaling and for some of the TGFβ-mediated biological effects. The overexpression of SMAD4, restoring TGFβ transduction, determined a significant increase of antiproliferative response to TGFβ, and reduced the invasive behavior of these cells. Therefore, our data indicated that reduction of SMAD4 may play a significant role in thyroid carcinogenesis.
Introduction
Transforming growth factor-β (TGFβ) belongs to a large superfamily of growth factors and controls a plethora of cellular responses, including cell proliferation, differentiation, motility, adhesion, extracellular matrix protein expression, apoptosis, and specification of developmental fate, during embryogenesis as well as in mature tissues, in species ranging from flies and worms to mammals (Patterson & Padgett 2000, Shi & Massagué 2003). Hence, in normal cells, TGFβ acts as a tumor suppressor by inhibiting cell growth or by promoting cellular differentiation or apoptosis.
At some time during the stepwise transition towards malignancy, human cancer cells acquire several abilities that most normal cells do not have, including the ability to become, at least partially, resistant to growth inhibition, to proliferate without dependence on growth factors, to replicate without limit, to invade, and to metastasize. TGFβ is one of the complex signal transduction pathways that control normal cellular homeostasis regulating these processes. TGFβ signaling is propagated via cell surface Ser/Thr kinases, TGFβ type I receptor (TβRI, listed in the HGNC database as TGFBR1), and TGFβ type II receptor (TβRII, TGFBR2), followed by an intracellular cascade of events involving receptor-regulated SMADs (R-SMADs), SMAD2 and SMAD3, and their interacting partner, common-mediator SMAD, SMAD4. Once formed the complexes, the SMADs translocate in the nucleus where they regulate target genes (Attisano & Wrana 2000). In contrast, the inhibitory SMAD (I-SMAD), SMAD7, binds to TβRI and prevents the phosphorylation of R-SMADs, resulting in the inhibition of TGFβ signaling (Imamura et al. 1997, Nakao et al. 1997). A SMAD ubiquitin regulatory factor 1 (SMURF1), being a HECT-type E3 ubiquitin ligase, interacts with I-SMAD7 and enhances the turnover of TβRI (Suzuki et al. 2002). TGFβ elicits context-dependent and cell-specific effects that often appear conflicting. It is puzzling how a seemingly direct transduction scheme, dependent on shuttling of SMAD proteins from the cytoplasm to the nucleus, can mediate such diverse array of responses.
TGFβ is normally expressed and secreted in epithelial follicular thyroid cells. In thyroid cell lines, the differentiated phenotype is controlled by TGFβ, which inhibits iodide trapping (Taton et al. 1993, Pekary & Hershman 1998) and thyroglobulin synthesis (Colletta et al. 1989); some of these effects are exerted through SMAD signaling (Nicolussi et al. 2003, Costamagna et al. 2004). TGFβ is also the negative regulator of thyrocyte proliferation: it antagonizes the mitogenic effects of the main growth factors in thyroid cells of rat (Morris et al. 1988, Colletta et al. 1989, Coppa et al. 1995, Carneiro et al. 1998), of porcine (Franzén et al. 1999), and of humans (Wyllie et al. 1991), and delays progression during the mid-late G1 phase (Carneiro et al. 1998, Depoortere et al. 2000).
Several studies on surgical samples and on thyroid cell cultures have been performed showing the role of TGFβ and other members of TGFβ superfamily in thyroid proliferative diseases: benign cases (Morosini et al. 1994) and malignant ones (Heldin et al. 1999, Kimura et al. 1999, Cerutti et al. 2003). These data support the hypothesis that this growth factor is important in thyroid cell physiology, modulating thyroid functions and proliferation, and justify the need to study also the proteins involved in signaling from the cytoplasm to the nucleus.
The objective of this study was to determine the functional role of TGFβ pathway in human papillary thyroid carcinogenesis. For this purpose, we examined the TGFβ response in two cell lines, TPC-1 and BCPAP, and in a group of 23 cases of human papillary thyroid carcinomas (PTCs). Our data demonstrated for the first time that these cells showed a strong reduction in the level of SMAD4 protein, which was responsible for an alteration of TGFβ signaling and for some of the TGFβ-mediated biological effects. The overexpression of SMAD4, restoring TGFβ transduction, determined a significant increase of antiproliferative response to TGFβ, and reduced the invasive behavior of these cells. Therefore, these data indicate that reduction of SMAD4 may play a significant role in thyroid carcinogenesis.
Materials and methods
Reagents and constructs
DMEM, Coon's modified Ham's F-12 medium, PBS, FBS, trypsin–EDTA, l-glutamine 100× (200 mM), and the six-hormone mixture (6H) containing TSH (10 mU/ml), insulin (10 μg/ml), hydrocortisone (10−8 M), transferrin (5 μg/ml), glycyl-l-histidyl-l-lysine acetate (10 ng/ml), and somatostatin (10 μg/ml) were purchased from Sigma–Aldrich, Co.
The human recombinant TGFβ1 isoform was purchased by Peprotech (Rocky Hill, NJ, USA). MG132 was provided by Calbiochem (La Jolla, CA, USA).
Phenylmethylsulfonyl fluoride (PMSF), protease inhibitor cocktail containing 1 mg/ml leupeptin, 2.5 mg/ml aprotinin, and 1 mg/ml benzamide hydrochloride, and phosphatase inhibitor cocktail were purchased from Sigma–Aldrich, Co. Antibodies to TβRI and TβRII, green fluorescent protein (GFP), SMURF1 (H60), SMAD7 (H79), SMAD4 (B8), SMAD4 (H552), N-cadherin (H63), β-actin (C4), and Lamin B (C20) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Antibodies to the phosphorylated form of SMAD2 (SMAD2-P) and SMAD3 (SMAD3-P) were obtained from Cell Signaling Technology (Beverly, MA, USA), to SMAD2-P from Millipore (Billerica, MA, USA), to E-cadherin from Immunological Sciences (Rome, Italy), to SMAD3 from Zymed Laboratories (San Francisco, CA, USA), and to SMAD2/ SMAD3 from BD Biosciences PharMingen (San Diego, CA, USA). Texas Red-conjugated goat anti-mouse IgG was obtained from Jackson Laboratories (Bar Harbor, ME, USA); FITC-conjugated goat anti-mouse and goat anti-rabbit IgG was obtained from Cappel (Organon Teknica Corp., West Chester, PA, USA).
Transfections were performed using lipofectin kit provided by LipofectAMINE PLUS (Gibco BRL, Life Technologies).
GFP-tagged SMAD4 wt was constructed subcloning the human SMAD4 gene in CLONTECH pEGFPC3 vector, which allows in-frame fusion to the C terminus of GFP. pCMV5 3TP-Lux, pCMV5-βGal, and pCMV5 FLAG-SMAD2 wt have been gently provided by Prof. Liliana Attisano.
Cell cultures and transient transfection analysis
TPC-1 and BCPAP (human PTC cell lines), and FRO (human anaplastic thyroid carcinoma cell lines) were kindly provided by Prof. Alfredo Fusco and were maintained in culture as previously described (Iuliano et al. 2003). The FRTL-5 cells were kindly provided by Dr L D Kohn. This cell line, diploid between their 5th and 25th passage, maintains the functional characteristics of iodide uptake, thyroglobulin synthesis, and cyclic nucleotide metabolism over prolonged periods of culture and doubling time of ∼36 h (Ambesi-Impiombato et al. 1980). These cells were grown as previously described (Nicolussi et al. 2003) at 37 °C in a humidified atmosphere of 5% CO2 air in w/o supplemented with 5% calf serum and 6H mixture. FTC133, human follicular thyroid tumor cell line (Goretzki et al. 1990), kindly provided by Prof. Salvatore Ulisse, MCF7 and MDA MB468 (Schutte et al. 1996), breast cancer cell lines, HepG2, human hepatoma cell line, and HEK293, derived from human embryonic kidney cells, purchased from American Type-Culture Collection (ATCC, Rockville, MD, USA), were grown at 37 °C in a humidified atmosphere of 5% CO2 air in DMEM supplemented with 10% calf serum.
TPC-1 and BCPAP were transiently cotransfected with empty vectors or pEGFPC3-SMAD4 wt and/or pCMV5 FLAG-SMAD2 wt using lipofectin technique following the manufacturer's instructions. After 24 h of incubation, transfected cells were used in immunofluorescences, luciferase assays, proliferation assays, wound-healing assays, and western blot assays.
Proliferation assays
The cell proliferation was assessed by cell counting. Briefly, 5×104 cells were seeded into 35 mm plates and allowed overnight at 37 °C in a humidified incubator with 5% CO2. Cell numbers were determined using a hemacytometer both at time zero (T0), to determine the number of cells in the plates before the TGFβ1 treatment, and after 24 and 48 h with or without 10 ng/ml of TGFβ1 in serum-free medium. Cells were counted three times by two independent investigators. Inter-observer variation was below 5%. Values represent the mean of triplicate determination ±s.d. of three experiments, and the results were also presented as a percentage of growth inhibition with respect to relative untreated control for each time point.
Luciferase assays
These assays were carried out as described previously (Nicolussi et al. 2003, 2006). After transfection, cells were treated or untreated with 10 ng/ml of TGFβ1 in serum-free medium. Luciferase activities were normalized for transfection efficiency using the β-galactosidase-expressing control vector. The experiments were repeated four times in triplicate.
Protein extraction and western blotting analysis
Protein extracts were obtained as previously described (Nicolussi et al. 2006). Subconfluent cells, transfected or not, were treated or untreated with 10 ng/ml of TGFβ1 for 30, 60, or 120 min or with MG132 (30 μM) for 24 h.
To obtain the cytosolic and nuclear fractions, cells were lysed in 0.4 ml hypotonic buffer A (10 mM Hepes, pH 7.4, 42 mM KCl, and 5 mM MgCl2) supplemented with protease inhibitor cocktail. After centrifugation of lysate at 1200 g for 10 min, the supernatant containing the cytosolic fraction was clarified by centrifugation at 15 000 g for 1 h at 4 °C. The pellet containing nuclear fraction was repeatedly washed in a specific buffer B (10 mM Hepes, pH 7.4, 10 mM KCl, 1.5 mM MgCl2, and 0.1 mM EDTA) and resuspended in extraction buffer C (20 mM Hepes, pH 7.4, 25% glycerol, 0.4 M NaCl, 1.5 mM MgCl2, and 0.2 mM EDTA), supplemented with protease inhibitor cocktail, PMSF, and dithiothreitol. Protein extracts were subjected to immunoblotting as previously described (Nicolussi et al. 2006) using primary antibodies to TβRI, TβRII (1:600), SMAD2-P (1:1000), SMAD3-P (1:1000), SMAD3 (1:1000), GFP (1:500), E-cadherin (1:1000), N-cadherin (1:500), Lamin B (1:1000), SMAD2/SMAD3 (1:500), SMAD4 (1:500), and β-actin (1:5000). SMAD4, E-cadherin, SMAD2-P, and SMAD3-P bands were analyzed using Bio-Rad Laboratories software. Data, obtained from three different protein extracts, were collected in terms of average intensity of bands of each proteins per average intensity of bands of β-actin or SMAD2.
RNA isolation and analysis
RNA was extracted using Tri Reagent (Sigma–Aldrich, Co.), following the manufacturer's instructions. Using 1 μg RNA, cDNAs were synthesized using MuLV reverse transcriptase (Applied Biosystems, Hammonton, NJ, USA) and random primers (Roche) according to the manufacturer's instructions. The primers used in the amplification of human SMAD4 (SMAD4-A forward 5′-CCTTGCAACGTTAGCTGTTG-3′, reverse 5′-CTTCAGTGGACAACGATG-3′; SMAD4-B forward 5′-CACTGCAGAGTAATGCTCC-3′, reverse 5′-GTCCACGTATCCATCAAC-3′; SMAD4-C forward 5′-GATGTTCAGGTAGGAGAG-3′, reverse 5′-CCAACGGTAAAAGACCTCAG-3′) and GAPDH genes as internal control (GAPDH forward 5′-ACCACAGTCCATGCCATCAC-3′, reverse GAPDH 5′-TCCACCACCCTGTTGCTGTA-3′) were designed to cross intron–exon junctions. The amplified cDNAs were purified with a gel cleanup kit, and subjected to sequencing reactions (MWG Biotech AG, Martinsried, Germany). All the obtained sequences corresponded to the expected ones (data not shown).
Real-time reverse transcription (RT)-PCR was performed three times in six replicates on a 7500 RT-PCR System (Applied Biosystems) using the SYBR Green detection system. mRNA levels were compared with reference curves and normalized to GAPDH mRNA. Each experiment was repeated three times using different total RNA extracts.
Immunofluorescence experiments
Immunofluorescences were carried out as described previously (Nicolussi et al. 2006). Cells were incubated with or without 10 ng/ml of TGFβ1 in serum-free medium for 1 h and then fixed with 4% paraformaldehyde/PBS for 30 min, followed by treatment with 0.1 M glycine/PBS for 20 min and permeabilized with 0.1% Triton X-100 for additional 5 min. For detection of SMAD2/SMAD3 and SMAD2-P, cells were incubated with anti-SMAD2/3 antibody (1:50) and SMAD2-P antibody (1:50) for 1 h respectively, followed by Texas Red-conjugated goat anti-mouse IgG (1:50). For detection of SMAD4, cells were incubated with anti-SMAD4 antibody (1:100), followed by FITC-conjugated goat anti-mouse IgG (1:10). Nuclei were stained with 10 μg/ml 4′,6′-diamidino-2-phenylindole (Sigma–Aldrich, Co.).
Immunofluorescences in cells transiently transfected with pEGFPC3-SMAD4 wt or pEGFPC3 empty vector were carried out as described above. The anti-GFP antibody (1:100) was used to detect GFP-tagged SMAD4 protein. Fluorescence was observed with an Olympus BX51 microscope and photographed at 40× optical magnification. Each experiment was repeated three times.
Wound-healing assay
TPC-1, BCPAP, and FRTL-5 cells, native or transiently transfected, were plated in 35 mm dish and cultured to confluence of 90%. Cell layers were then scratched with a pipette tip, and cellular debris has been removed washing gently. Some plates were fixed and photographed immediately after scratching (T0 control). The cells that were allowed to incubate with serum-free medium for 24 h were fixed with 4% paraformaldehyde for 30 min, and dishes were plated under a phase contrast microscope Axio Observer (Carl Zeiss, Inc., Oberkochen, Germany) and photographed to measure the gap distance using the computing software Axio Vision (Carl Zeiss, Inc.). Each result is the mean of three independent experiments.
Tissue samples and immunohistochemistry analysis
A group of 23 PTCs, collected at the Anatomic Pathology of Department of Experimental Medicine of Sapienza University of Rome from 2008 to 2010, were used. Thyroid tissue sections were studied using the catalyzed signal amplification protocol (DAKO A/S, Glostrup, Denmark; Lazzereschi et al. 1997). Slides were incubated in a humidified chamber overnight at 41 °C with 1:100 dilution of polyclonal serum to SMAD4 (H552). The peroxidase-based LSAB2 Detection kit (DAKO A/S), followed by hematoxylin counterstaining, was used to visualize the reactions. Written informed consent was obtained from each patient according to Helsinki Declaration and approved by the local ethics committee.
Statistical analysis
All statistical analyses have been performed using JMP Software purchased by Statistical Discovery SAS Institute (Cary, NJ, USA). Data were analyzed by the following tests: Student's t-test; Fisher test across groups (P value referred to F value); non-parametric one-way ANOVA Wilcoxon/Kruskal–Wallis test (P value referred to χ2-value). P<0.05, statistical significance; P<0.001, high statistical significance.
Results
TGFβ signaling in TPC-1 and BCPAP
Expression of members of TGFβ pathway has been studied in TPC-1 and BCPAP cells. These cells represent a model of human papillary thyroid cell lines, whose thyroid origin has been confirmed by RT-PCR for the presence of RET/PTC rearrangement and PAX8 expression in TPC-1 and PAX8 and TTF1 expression in BCPAP (data not shown).
The control cells used in this study were a human follicular cell line, FTC133, and a normal rat thyroid cell line, FRTL-5. FTC133 retains the differentiated thyrocyte functions like responsiveness to TSH for the presence of TSH receptor, TTF1, and PAX8 confirmed by RT-PCR; FRTL-5 was widely used as a normal control (Pang et al. 1989, Shimura et al. 2001, Pomérance et al. 2006, Visconti et al. 2007, Kogai et al. 2008).
Western blot analyses performed on total cell lysates, using polyclonal antibodies anti-TβRII and anti-TβRI, demonstrated that both receptors were present in all examined cell lines (Fig. 1A), and they were not modified by TGFBβ1 treatment (data not shown). Since the expression of TβRI was lower in TPC-1 and in BCPAP with respect to the control cells, we assayed the level of SMAD7 and SMURF1 to assess whether there was an alteration in the mechanism of TβRI degradation. The expression of these proteins was comparable in all cell lines (Fig. 1A). Consistent with the finding that the expression of SMAD7 is induced by TGFβ, the level of this protein was increased by TGFβ1 treatment, demonstrating that this branch of signaling was functioning in PTC cell lines (data not shown). As an indication of the ability of the lower level of TβRI to phosphorylate SMAD2 and SMAD3, we performed western blot analyses with or without 10 ng/ml of TGFβ1 treatment for 30 min. Our data demonstrated that SMAD2 and SMAD3 were phosphorylated both in TPC-1 and in BCPAP, as in FTC133 and FRTL-5 control cells (Fig. 1B and C).
TGFβ signaling in TPC-1 and BCPAP. (A) Whole cell extracts (40 μg/lane) were analyzed by western blot, using antibodies against TβRII, TβRI, SMAD7, SMURF1, and β-actin as equal loading control. (B and C) Phosphorylation of SMAD2 and SMAD3 after treatment with 10 ng/ml of TGFβ1 for 30 min. Histograms of densitometric analysis of the SMAD2-P and SMAD3-P bands normalized to the respective unphosphorylated forms. (D) Cells were transiently transfected with 3TP-Lux reporter and grown for 24 h in the presence or the absence of 10 ng/ml of TGFβ1. Luciferase activity was expressed in fold induction ±s.d. Each result is the mean of four independent experiments (** indicates a high statistical significance (Student's t-test, P<0.001) correlate to untreated cells).
Citation: Journal of Molecular Endocrinology 45, 4; 10.1677/JME-10-0044
The transcriptional activity of the TGFβ cascade was evaluated using a TGFβ-responsive reporter construct, p3TP-Lux. The p3TP promoter contains three consecutive 12-O-tetradecanoyl phorbol 13-acetate (TPA) response elements, which are involved in the transcriptional responses of several genes to TGFβ, and a portion of the plasminogen activator inhibitor-1 promoter region (de Groot & Kruijer 1990, Keeton et al. 1991, Wrana et al. 1992). The results showed (Fig. 1D) that the increase of transcriptional activity, after 24 h of TGFβ1 treatment, was not significant both in TPC-1 (3.5-fold) and in BCPAP (3.1-fold), while it was significant both in FTC133 (10.3-fold, P<0.001) and in FRTL-5 (25-fold, P<0.001).
Altogether, our data demonstrated an important impairment of TGFβ signaling in papillary thyroid cell lines.
SMAD4 expression in TPC-1 and BCPAP
Based on the observation of frequent inactivation of SMAD4 in several types of tumors, we analyzed SMAD4 expression in TPC-1, BCPAP, and in a set of different human cell lines: MDA MB468, FRO, and human-immortalized B lymphocytes. FTC133 and FRTL-5 were used as a positive control, whereas MDA MB468, in which SMAD4 was deleted in homozygosis, was used as a negative control. Western blot in Fig. 2A shows that SMAD4 expression was strongly reduced in all PTC cells with respect to FTC133 and FRTL-5. Densitometric analysis of bands confirmed a significant (P<0.05) reduction with respect to FTC133 both in TPC-1 (66%) and in BCPAP (72%), and with respect to FRTL-5 both in TPC-1 (69%) and in BCPAP (75%).
Expression of SMAD4. (A) Whole protein lysates (80 μg/lane) from TPC-1, BCPAP, FTC133, FRTL-5, and a set of different human cell lines MDA MB468, FRO, and human-immortalized B lymphocytes were analyzed by western blot using an antibody against SMAD4. Densitometric evaluation of the SMAD4 signals was performed normalizing to the levels of β-actin (* indicates a statistical significance (Student's t-test, P<0.05) correlate to FTC133 and FRTL-5). (B) The upper part of panel shows a schematic representation of specific primers for SMAD4 used for the amplification. GAPDH was used as a housekeeping gene. (C) Semiquantitative real-time PCR analysis of SMAD4 mRNA levels. Data reported represent the mean of three independent experiments. FTC133, HepG2, and HepG2 transiently transfected with pEGFPC3-SMAD4 wt construct and MDA MB468 samples were used as controls in the experiments. (D) Total protein lysates (80 μg/lane) from TPC-1, BCPAP, FTC133, and FRTL-5, treated or untreated with MG132 (30 μM) for 24 h, were analyzed by western blot using an anti-SMAD4 antibody. Histogram of densitometric analysis of SMAD4 levels normalized to the levels of β-actin.
Citation: Journal of Molecular Endocrinology 45, 4; 10.1677/JME-10-0044
The expression of SMAD4 gene was studied by RT-PCR. As shown in Fig. 2B, all the primers, used for RT-PCR, amplified fragments of the expected size both in PTC cells and in human controls (FTC133, FRO, human-immortalized B lymphocytes, and MCF7). Since mutational inactivation of SMAD4 in different carcinomas has been reported, we sequenced the CDS region of SMAD4 cDNA to look for nucleotide alterations. Nucleotide sequencing did not show any changes in all examined cell lines (data not shown).
SMAD4 mRNA levels were evaluated by semiquantitative real-time PCR on cDNA of TPC-1 and BCPAP and a set of positive controls (FTC133, HepG2, and HepG2 transfected with pEGFPC3-SMAD4 wt) and a negative control (MDA MB468). The results obtained demonstrated that all cell lines showed a comparable level of SMAD4 transcript (Fig. 2C). Therefore, these data suggested that the cause of hypoexpression of SMAD4 protein in these cells could be an alteration in one of the post-transcriptional mechanisms.
To analyze the implications of proteasome in SMAD4 degradation, we treated cell lines for 24 h with MG132, an inhibitor of proteasome activity. As showed in Fig. 2D, MG132 treatment determined an increase of SMAD4 level in both cell lines, likewise in FRTL-5 and FTC133 (about 1.5-fold). These results indicated that the ability of proteasome to degrade SMAD4 in these cells was preserved, but not enhanced.
Shuttling of SMAD proteins
The molecular mechanism that controls subcellular localization and activation of SMAD proteins is crucial for transduction of TGFβ signal from transmembrane receptors into the nucleus, and it is not yet fully understood. The immunofluorescence experiments demonstrated that in unstimulated cells, SMAD2 was predominantly cytoplasmic, whereas after 1 h of TGFβ1 treatment, it was localized throughout the cytoplasm and the nucleus, as a result of its shuttling between these two compartments (Fig. 3A). The localization of SMAD2-P, clearly cytoplasmic in basal condition, became nuclear after TGFβ1 treatment in all cell lines (Fig. 3A).
Subcellular localization of SMAD2 and SMAD4. (A and B) TPC-1, BCPAP, and FRTL-5, treated or untreated with 10 ng/ml of TGFβ1 for 1 h, were fixed and immunolabeled for SMAD2, SMAD2-P, and SMAD4. Nuclei were counterstained with DAPI. Images are representative of three separate experiments. Scale bar, 20 μm.
Citation: Journal of Molecular Endocrinology 45, 4; 10.1677/JME-10-0044
The nuclear translocation of SMAD4 was investigated in the same cells and in the negative control represented by MDA MB468. SMAD4 not only accumulates in the nucleus in association with activated R-SMADs, but also undergoes continuous nucleocytoplasmic shuttling on its own, independently of TGFβ signaling (Pierreux et al. 2000, Watanabe et al. 2000). In accordance with these findings, the data presented in Fig. 3B demonstrated the presence of continuous nucleocytoplasmic shuttling both in basal condition and after TGFβ1 treatment. Therefore, the SMAD4 ability to translocate to the nucleus was preserved, and the weak cellular fluorescence signal observed in TPC-1 and in BCPAP was due to the hypoexpression of this protein.
Western blot performed on the cytoplasmic and nuclear extracts of TPC-1, BCPAP, and FRTL-5 demonstrated a clear nuclear translocation of SMAD4 and SMAD2-P, after 30 and 60 min of TGFβ1 treatment in all cell lines. The presence of SMAD2 and SMAD4 in the cytoplasmic fractions in all cell lines, after TGFβ1 treatment, confirmed the continuous nucleocytoplasmic shuttling of these proteins (Fig. 4).
Subcellular fractionation of SMAD2, SMAD2-P, and SMAD4. (A–D) Cells, treated with 10 ng/ml of TGFβ1 for indicated time periods, were separated into cytosol and nuclear fractions before analyss by immunoblotting with the SMAD2, SMAD2-P (A and B), and SMAD4 (C and D) antibodies. Western blots, performed using antibodies against the β-actin, cytoplasmic protein, and against Lamin B, nuclear protein, indicated that the extracts are virtually free from cross-contamination.
Citation: Journal of Molecular Endocrinology 45, 4; 10.1677/JME-10-0044
SMAD4 overexpression in TPC-1 and BCPAP
In order to study the effects of the SMAD4 overexpression, the cells have been transiently transfected with a construct containing human SMAD4 wt, tagged with GFP (pEGFPC3-SMAD4), or with a construct containing human SMAD2 wt tagged with FLAG (pCMV5 FLAG-SMAD2) or cotransfected with both plasmids. The transcriptional activity of SMAD4 was studied using the p3TP-Lux reporter construct in the presence or in the absence of 10 ng/ml of TGFβ1 for 24 h. The results, evaluated by Student's t-test, demonstrated a significant (P<0.05) increase of 3TP-Lux activity in TPC-1 transfected with SMAD4 or SMAD4-/SMAD2-cotransfected cells versus cells transfected with reporter construct. In BCPAP, the increase was highly significant (P<0.001) in SMAD4-transfected cells versus cells transfected with reporter construct, and the increase was significant (P<0.05) in SMAD4-/SMAD2-cotransfected cells versus cells transfected with reporter construct (Fig. 5A).
TGFβ signaling in TPC-1 and BCPAP overexpressing SMAD4. (A) TPC-1 and BCPAP overexpressing SMAD4 wt and/or SMAD2 wt were transiently transfected with 3TP-Lux reporter. Cells were stimulated with 10 ng/ml of TGFβ1 for 24 h, and the luciferase activity was expressed as arbitrary units. # and ## indicate a statistical significance (P<0.05 and P<0.001 respectively, P value referred to Student's t-test) of TPC-1 and BCPAP overexpressing SMAD4 versus cells transfected with reporter construct. * and ** indicate a statistical significance (P<0.05 and P<0.001 respectively, P value referred to F value, Fisher test across groups) of cells transfected with SMAD4 versus cells transfected with reporter construct and SMAD4-/SMAD2-cotransfected cells versus cells transfected with reporter construct. Each result is the mean of four independent experiments in triplicate ±s.e.m. (B) Immunofluorescence assays of GFP-tagged SMAD4 protein in TPC-1 and BCPAP overexpressing GFP-tagged SMAD4 wt and cells transfected with empty vector, treated or untreated with 10 ng/ml of TGFβ1 for 1 h. Scale bar, 20 μm. (C) Representative western blots performed using specific antibody anti-GFP in total extracts from transiently transfected cells.
Citation: Journal of Molecular Endocrinology 45, 4; 10.1677/JME-10-0044
The response to TGFβ1 treatment between each group and across groups was evaluated by matched pair test. The Student's t-test between groups demonstrated that the transcriptional activity was significantly increased (P<0.001) in all conditions examined (Fig. 5A). Fisher test across groups (P value referred to F value) demonstrated that the luciferase induction after TGFβ1 treatment was statistically significant in all of the following conditions: TPC-1 cells transfected with SMAD4 versus cells transfected with reporter construct (P=0.0204) or versus cells transfected with SMAD2 (P=0.0034) and SMAD4-/SMAD2-cotransfected cells versus cells transfected with reporter construct (P=0.0144); BCPAP cells transfected with SMAD4 versus cells transfected with reporter construct (P=0.0010) or versus cells transfected with SMAD2 (P=0.0015). The results obtained in SMAD4-/SMAD2-cotransfected cells versus cells transfected with reporter demonstrated an increase of luciferase activity (P=0.0516; Fig. 5A). The Fisher test across groups in a matched pair test evaluates the difference in the response to TGFβ1, but does not take into account the baseline values induced by the overexpression of SMAD4. To determine whether the fold induction in 3TP-Lux reporter activity for TPC-1 and BCPAP was significantly increased by SMAD4 overexpression, we examined the data using the non-parametric one-way ANOVA Wilcoxon/Kruskal–Wallis test (P value referred to χ2-value). The results obtained demonstrated that the mean of fold induction of transcriptional activity, after TGFβ1 treatment in non-transfected TPC-1, was 3.6 vs 4.2 (P<0.0076) in SMAD4-transfected cells, and the mean of fold induction in non-transfected BCPAP was 3.1 vs 4.1 (P<0.0466) in SMAD4-transfected cells (Fig. 5A). These results showed that in transfected cells, the fold induction of luciferase activity was significantly increased after TGFβ1 treatment: the level of SMAD4 in the cells was causing this response.
To verify whether the reduction of SMAD4 shuttling was caused by hypoexpression of this protein, we performed immunofluorescences in transfected cells, using an antibody specific for GFP. The results shown in Fig. 5B demonstrated that the overexpression of SMAD4 determined a very strong nuclear translocation of SMAD4 after TGFβ1 treatment (upper panel), with respect to the control cells transfected with empty vector (lower panel). Western blot confirmed the overexpression of GFP-tagged SMAD4 protein in transiently transfected cells (Fig. 5C).
Taken together, these data indicated that the attenuated signaling pathway observed in TPC-1 and BCPAP could result from the low levels of SMAD4, and the hyperexpression of this protein was responsible for a significant enhancement of TGFβ responsiveness.
TGFβ growth inhibition in TPC-1 and BCPAP
The responsiveness to TGFβ growth inhibition has been evaluated in PTC cells by proliferation assay. TPC-1, BCPAP, and FRTL-5 control cells were grown with or without 10 ng/ml of TGFβ1, and counted at time 0 and at 48 h. The results, expressed as a percentage of growth inhibition with respect to relative untreated control, showed that TPC-1 was significantly inhibited by 32% (P<0.05), BCPAP was inhibited only by 8%, and FRTL-5 was highly significantly inhibited by 72% (P<0.001; Fig. 6A). Therefore, BCPAP was resistant to the TGFβ-mediated growth inhibition, while TPC-1 showed a significant inhibition of proliferation, although lower than FRTL-5.
Effect of TGFβ1 on cell growth in thyroid cell lines. (A) Cells were grown in the presence or the absence of 10 ng/ml of TGFβ1 for 48 h and were counted three times by two independent investigators. Inter-observer variation was below 5%. Values represent the mean of triplicate determination ±s.d. of three experiments, and the results were expressed as a percentage of growth inhibition (* and ** indicate a statistical significance (Student's t-test), P<0.05 and P<0.001 respectively) of treated versus untreated cells. (B and C) Cells transiently transfected with GFP-tagged SMAD4 wt or with the empty vector, grown in the presence or the absence of 10 ng/ml of TGFβ1 for 24 and 48 h, were counted three times by two independent investigators. Inter-observer variation was below 5%. Values represent the mean of triplicate determination ±s.d. of three experiments. (D and E) Histograms of the results expressed as a percentage of growth inhibition (* and ** indicate a statistical significance (Student's t-test), P<0.05 and P<0.001 respectively).
Citation: Journal of Molecular Endocrinology 45, 4; 10.1677/JME-10-0044
To investigate the role of SMAD4 in the control of cell growth, we performed proliferation assay on cells overexpressing SMAD4. TPC-1 and BCPAP, transiently transfected with GFP-tagged SMAD4 wt or with the empty vector, were counted at time 0, at 24 h, and at 48 h in the presence or in the absence of 10 ng/ml of TGFβ1. In basal conditions, SMAD4 overexpression did not cause a significant difference in growth rate when comparing SMAD4-transfected cells to empty vector-transfected cells (Fig. 6B and C). TGFβ1 treatment, instead, determined a considerable reduction of proliferation.
In TPC-1, the inhibition of proliferation was significant (P<0.05) at 24 h (59%) and highly significant (P<0.001) at 48 h (71%), while in the cells transfected with empty vector, the inhibition was 19% at 24 h and 35% at 48 h (P<0.05), as in untransfected cells (Fig. 6D). The results obtained in BCPAP overexpressing SMAD4 demonstrated a significant (P<0.05) recovery of TGFβ-mediated inhibition of proliferation only at 48 h (23%), when compared to the empty vector-transfected cells (6%; Fig. 6E). Student's t-test, which was performed comparing the percentage of growth inhibition of SMAD4-transfected PTC cell lines versus empty vector-transfected cells, demonstrated a significant increase (P<0.05) at 24 h, 59 vs 19%, and a highly significant increase (P<0.001) at 48 h, 71 vs 35%, in TPC-1, and it also demonstrated a significant increase (P<0.05) only at 48 h, 23 vs 6%, in BCPAP. The trypan blue test performed in all proliferation assays (data not shown) demonstrated that the growth inhibition observed in these cells was not due to reduction of cellular viability.
The above results provided evidence that the overexpression of SMAD4 caused a strong increase of the antiproliferative effect of TGFβ in human PTC cell lines.
SMAD4 overexpression in the control of migratory behavior of TPC-1 and BCPAP
Migration of tumor cells has an important role in specific processes for metastases formation (Chiang & Massagué 2008). The ability of TPC-1 and BCPAP to migrate was studied through the wound-healing assay. The experiments were performed in basal condition and after transient transfection either with pEGFPC3-SMAD4 wt or with pEGFPC3 empty vector. The cells were grown to confluence in monolayers, and their ability to migrate into and across a denuded area of the monolayer has been evaluated after 24 h. As shown in Fig. 7A, TPC-1 and BCPAP had a highly significant (P<0.001) migratory attitude, compared to the FRTL-5, which migrated quite slowly. The overexpression of SMAD4 strongly reduced the migratory behavior of these cells (Fig. 7B). The analysis of the data demonstrated that over a 24 h time course, TPC-1 overexpressing SMAD4 consistently migrated at a slower rate (21%; P<0.001) than empty vector control cells (50%). BCPAP overexpressing SMAD4 showed a similar behavior, migrating at a slower rate (14%; P<0.001) than empty vector control cells (38%). Stimulation with TGFβ1 did not cause any significant modification in motility in transfected cells, since SMAD4 overexpression alone was responsible for maximal activation of SMAD signaling (Fig. 7B). The inhibition of migration at 24 h was independent of the proliferative response because the overexpression of SMAD4 did not influence the growth rate in the same gap (Fig. 6B and C).
Migratory behavior of TPC-1 and BCPAP. (A) Wound-healing assay was performed in confluent monolayers for 24 h, and the extension of the area colonized by the cells was estimated using Axio Vision, Zeiss software (** indicates a statistical significance (Student's t-test, P<0.001) compared to FRTL-5). (B) Wound-healing assay in TPC-1 and BCPAP overexpressing SMAD4 wt, untreated or treated with 10 ng/ml of TGFβ1 for 24 h (** indicates a high statistical significance (Student's t-test, P<0.001) compared to empty vector control cells). Images are representative of three separate experiments. (C) Whole protein lysates (40 μg/lane) from TPC-1, BCPAP, FTC133, and FRTL-5 control cells, MCF7, and HEK293, were analyzed by western blot using antibodies against E-cadherin and N-cadherin. (D) Western blot of whole protein lysates (40 μg/lane) from TPC-1 and BCPAP transiently transfected with pEGFPC3-SMAD4 wt or pEGFPC3 empty vector performed using an antibody against E-cadherin. Bottom, densitometric evaluation of the E-cadherin bands normalized to the levels of β-actin (** indicates a high statistical significance (Student's t-test), P<0.001).
Citation: Journal of Molecular Endocrinology 45, 4; 10.1677/JME-10-0044
It has been demonstrated that E-cadherin and N-cadherin have an important role in progression-promoting pathway in invasive and migratory behavior in many human carcinomas (Christofori 2006, Shiou et al. 2007). The expression levels of E-cadherin and N-cadherin have been evaluated performing western blot analyses on total lysates of TPC-1, BCPAP, FTC133, and FRTL-5. MCF7 has been used as a positive control of E-cadherin and a negative control of N-cadherin; HEK293, instead, has been used as a positive control of N-cadherin and a negative control of E-cadherin (Hogan et al. 2004). The results obtained demonstrated that E-cadherin expression was lower in PTC cell lines with respect to FTC133, and N-cadherin expression was comparable in all thyroid carcinoma cell lines (Fig. 7C). TPC-1 and BCPAP overexpressing SMAD4 showed a highly significant (P<0.001) increase of the E-cadherin level (Fig. 7D).
In conclusion, our results strongly supported that, in thyroid papillary carcinomas, the overexpression of SMAD4 was responsible both for the increase of antiproliferative effect of TGFβ and for the reduction of cellular motility in vitro.
SMAD4 expression in human PTCs
To investigate whether the findings observed in cell lines were relevant to human thyroid carcinogenesis, a group of 23 human PTCs were examined for SMAD4 expression by immunohistochemistry and compared to normal tissues from the opposite lobe. Our results showed that 7 out of 23 (30%) PTC tumor samples, including 1 case of follicular variant of PTCs (Fig. 8), presented a weak and focal intensity of SMAD4 staining with SMAD4-negative areas compared to normal tissue, where a diffuse staining of SMAD4 was observable. In remaining samples, SMAD4 expression appeared to be retained and comparable to the normal tissues. This result was consistent with the data obtained in PTC cell lines, outlining the important role of SMAD4 in thyroid carcinogenesis.
SMAD4 expression in human papillary carcinomas. Immunohistochemistry, using antibody against SMAD4, was performed in thyroid samples from normal tissue (A), primary PTCs (B and C), and in one case of PTC follicular variant (D). Original magnification was 150× (A–C) and 100× (D).
Citation: Journal of Molecular Endocrinology 45, 4; 10.1677/JME-10-0044
Discussion
It is known that a lack of response to TGFβ inhibitory action in thyroid cells may lead to a loss of controlled growth and is responsible for some cases of multinodular thyroid goiter (Grubeck-Loebenstein et al. 1989, Bidey et al. 1999). In the last year, the role of TGFβ in the thyroid carcinogenesis has received a significant attention. Experimental evidences demonstrate that thyroid cancer cells can escape to tumor-suppressive action of TGFβ for the quantitative reduction of receptors and transduction elements. In a model of KRAS-transformed rat thyroid cells, the overexpression of TβRII reverts their malignant phenotype and strongly reduces the number of metastases, when transplanted in athymic nude mice (Turco et al. 1999). A deregulated expression with a reduced TβRII RNA level has been identified in a large number of thyroid carcinomas and correlates with a more undifferentiated histotype (Lazzereschi et al. 1997). Moreover, mutations and/or imbalances in expression due to alternative RNA splicing and altered localization of the SMAD4 protein have been demonstrated in a large group of thyroid cancer with different histotypes (Lazzereschi et al. 2005). It is known that TGFβ's role in human cancer appears both complex and context dependent. Depending on the tumor type and the stage of tumor progression, it can exhibit strong tumor-suppressive or tumor-promoting functions. More recently, it has been demonstrated that also in thyroid cells, as well as in the skin tumors, or in metastatic colon cancer (Padua & Massagué 2009), TGFβ can acts as a tumor-promoting factor. The expression of BRAF V600E in normal rat thyrocytes and in 50 cases of human PTCs determines a reduction of NIS expression and an increase of TGFβ secretion, suggesting an iperactivation of TGFβ signaling, responsible for the pro-tumorigenic activity (Riesco-Eizaguirre et al. 2009).
The present study focuses on the role of TGFβ pathway as tumor suppressor in two human PTC cell lines, TPC-1 and BCPAP, and in a group of 23 cases of human papillary carcinomas. In PTC cells, which were widely used by different groups (Baldini et al. 2004, Motti et al. 2005, Meireles et al. 2007, Visone et al. 2007, Schweppe et al. 2008, Salerno et al. 2010), we have studied the TGFβ pathway, demonstrating that the hypoexpression of SMAD4 protein is responsible for alteration in TGFβ signal transduction, as supported by the strong reduction of transcriptional activity of 3TP-Lux. The p3TP-Lux reporter is most commonly used to measure TGFβ responsiveness (Wrana et al. 1992, Carcamo et al. 1994). SMAD4 has been shown to be required for this transcriptional activity, since SMAD4-deficient cell lines are not responsive but can be rescued with SMAD4 expression (Lagna et al. 1996, de Caestecker et al. 1997). The reduction of 3TP-Lux activity observed in PTC cell lines could be explained by the low threshold of activation of TGFβ signaling, depending on TβRI or SMAD levels (Yingling et al. 1995). The TβRI level in TPC-1 and in BCPAP is lower than the control cells, but the ability of the receptor to phosphorylate SMAD2 and SMAD3 is fully maintained, as demonstrated in Fig. 1B and C. The stability of TβRI represents an important regulatory mechanism for TGFβ signaling, both in cell culture studies and in vivo models. TGFβ receptors are ubiquitinated and degraded through the action of several cooperating protein complexes containing E3 ligases, as well as other important regulators of protein degradation. The I-SMADs seem to have an important role in regulating many of these complexes, orchestrating both ubiquitination and de-ubiquitination (Lönn et al. 2009). The levels of SMURF1 and SMAD7, overexpressed in the anaplastic thyroid carcinoma cell line (Cerutti et al. 2003), are also unaltered in our model. Therefore, the lower level of TβRI observed in PTC cell lines cannot be attributed to SMAD7 and SMURF1 alterations. However, whether the receptors are degraded via the lysosomes, proteasomes or via both of these pathways is still an open question (Lönn et al. 2009). Thus, the hypoexpression of SMAD4 is responsible for the reduction of the threshold of induction of the signaling cascade, resulting in partial loss of TGFβ-transcriptional response. In fact, the overexpression of SMAD4 significantly increases the transcriptional stimulation of 3TP-Lux in the absence or in the presence of TGFβ1 treatment. Altogether, these data demonstrate that, despite the lower level of TβRI, the hypoexpression of SMAD4 is mainly responsible for the impairment of TGFβ signaling.
Since we previously demonstrated that SMAD4 mutations are frequent in PTCs (Lazzereschi et al. 2005), sequencing of the entire coding part of the SMAD4 gene was performed. It did not find any mutations in the SMAD4 gene; also mRNA levels are not modified; consequently, the alteration responsible for the hypoexpression of SMAD4 protein does not involve the transcriptional mechanism. Our data, in agreement with Heldin et al. (1999), led us to hypothesize that the cause of hypoexpression could be found in alteration of major components of translational machinery, which are frequently altered in human neoplasms (Rosenwald 2004). Moreover, recent observations demonstrate an important role of microRNA in the development of a wide range of human diseases, including cancer. Therefore, in our cells, we could hypothesize an alternative mechanism in the regulation of SMAD4 levels, involving an alteration in miRNA expression profiles (Erson & Petty 2008).
One of the important effects of SMAD4 hypoexpression in TPC-1 and BCPAP is the drastic reduction of nuclear accumulation of this protein. The molecular mechanism that controls subcellular localization and activation of SMAD proteins is crucial for TGFβ signaling, and it is not yet fully clarified. SMAD4 accumulates in the nucleus by association with activated R-SMADs (Liu et al. 1997, Hoodless et al. 1999, Watanabe et al. 2000). However, SMAD4 also undergoes continuous nucleocytoplasmic shuttling on its own, independently of TGFβ signaling (Pierreux et al. 2000, Watanabe et al. 2000, Hill 2009). Our results demonstrate that the SMAD4 ability to translocate into the nucleus is preserved, and the weak cellular fluorescence signal observed in TPC-1 and in BCPAP was due to the hypoexpression of this protein. Work over the past few years has emphasized the importance of SMAD nucleocytoplasmic shuttling in TGFβ signaling regulation. Tumorigenic SMAD4 alterations, as demonstrated in pancreatic human cancer, can eliminate nuclear accumulation of SMAD4, suggesting that defects in nuclear import may contribute to tumorigenicity (Morén et al. 2000). The overexpression of SMAD4 in TPC-1 and in BCPAP clearly reinstates the translocation of SMAD4 from the cytoplasm to the nucleus. The levels and the duration of residence in the nucleus of SMAD4 are important events for the response to TGFβ in the cells. Consistent with the notion that TGFβ can act as a morphogen that induces distinct cell fates along a concentration gradient in embryonic development, many studies demonstrate that the intensity and duration of the TGFβ-SMAD response are important determinants for signaling specificity and, therefore, the activity of SMADs is carefully regulated (Gurdon & Bourillot 2001, Hill 2001, ten Dijke & Hill 2004). Enhanced proteasomal degradation of SMAD4 is responsible for TGFβ resistance in breast cancer (Dupont et al. 2005). In our cells, the ability of proteasome to degrade SMAD4 is maintained and comparable to that of control cells. Therefore, different mechanisms of protein degradation seem to be involved.
Lack of responsiveness to TGFβ antiproliferative effects is a hallmark for the cancer, where TGFβ and its pathway function as oncosuppressors. Our data show that BCPAP is resistant to growth inhibition of TGFβ, while TPC-1, in accordance with the data of Matsuo et al. (2006), is still sensitive to its antiproliferative response, although lower than control cells. Here, we demonstrate that the level of SMAD4 protein is low both in TPC-1 and in BCPAP, and it is responsible for the weakening of TGFβ responsiveness in these cell lines. The overexpression of SMAD4 significantly increases the TGFβ growth inhibition in TPC-1 and partially restores this response in BCPAP. We think that the behavior observed in BCPAP could be caused by the different genetic and molecular profile, recently well characterized in these cell lines (Meireles et al. 2007, Schweppe et al. 2008).
Although the majority of thyroid cancers have an excellent prognosis, there are a small percentage of cases that show an extensive local invasion and distant metastases, which frequently do not respond to standard treatments and have a worsened prognosis. The genetic basis for thyroid cancer initiation and development is well characterized. It has been demonstrated that the activation of oncogenes like RAS, BRAF, RET/PTC, and PI3K/AKT plays an important role in thyroid tumorigenesis (Vasko et al. 2007), but the molecular mechanism involved in the thyroid cancer invasion needs clarification. The acquisition of a migratory phenotype is an essential property of invading and metastasizing cancer cells. TPC-1 and BCPAP cells have a strong and significant ability to migrate (Sancho et al. 2006, Scarpino et al. 2007). Genetic data indicate that loss of SMAD4 in cancer cells may correlate not only with the specific loss of TGFβ cytostatic response, but also with acquisition of an invasive phenotype (Takaku et al. 1998). In SW480 colon carcinoma cells, the re-expression of SMAD4 results in loss of tumorigenicity in nude mice, and is accompanied by the restoration of a more epithelioid morphology and induction of E-cadherin (Müller et al. 2002, Tian et al. 2009). Here, we demonstrate that the overexpression of SMAD4 in PTC cell lines induces a strong suppression of motility, as well as a significant increase of E-cadherin expression, indicating that the level of SMAD4 is a critical regulator of these processes.
Finally, our results obtained in the model in vitro are strongly supported by the preliminary finding that 30% of human PTCs, examined by immunohistochemistry, show a strong reduction of the SMAD4 staining.
All together, our data demonstrate that the level of SMAD4 protein plays an important role in thyroid carcinogenesis; although the mechanism by which tumor cells achieve this hypoexpression needs further investigation, we think it could be considered as a new possible therapeutic target for thyroid cancer.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This work was supported by the Ministero Italiano dell'Università e della Ricerca (PRIN 2007) and by the Associazione Italiana Ricerca sul Cancro (AIRC).
We thank Prof. Liliana Attisano for plasmids, Prof. Alfredo Fusco for TPC-1 and BCPAP cell lines, and Prof. Salvatore Ulisse for FTC133 cell line. The authors would also like to thank Mr Daniel Kenton for the careful revisions made. This work is dedicated to the memory of Prof. Giulia Colletta.
References
Ambesi-ImpiombatoFSParksLCoonHG1980Culture of hormone dependent functional epithelial cells from rat thyroids. PNAS773455–3459.doi:10.1073/pnas.77.6.3455.
AttisanoLWranaJL2000Smads as transcriptional co-modulators. Current Opinion in Cell Biology12235–243.doi:10.1016/S0955-0674(99)00081-2.
BaldiniETollerMGrazianoFMRussoFPPepeMBiordiLMarchioniECurcioFUlisseSAmbesi-ImpiombatoFS2004Expression of matrix metalloproteinases and their specific inhibitors in normal and different human thyroid tumor cell lines. Thyroid14881–888.doi:10.1089/thy.2004.14.881.
BideySPHillDJEggoMC1999Growth factors and goitrogenesis. Journal of Endocrinology160321–332.doi:10.1677/joe.0.1600321.
de CaesteckerMPHemmatiPLarisch-BlochSAjmeraRRobertsABLechleiderRJ1997Characterization of functional domains within Smad4/DPC4. Journal of Biological Chemistry27213690–13696.doi:10.1074/jbc.272.21.13690.
CarcamoJWeisFMVenturaFWieserRWranaJLAttisanoLMassaguéJ1994Type I receptors specify growth-inhibitory and transcriptional responses to TGF-β and activin. Molecular and Cellular Biology143810–3821.
CarneiroCAlvarezCVZalvideJVidalADomínguezF1998TGF-β1 actions on FRTL-5 cells provide a model for the physiological regulation of thyroid growth. Oncogene161455–1465.doi:10.1038/sj.onc.1201662.
CeruttiJMEbinaKNMatsuoSEMartinsLMacielRMKimuraET2003Expression of Smad4 and Smad7 in human thyroid follicular carcinoma cell lines. Journal of Endocrinological Investigation26516–521.
ChiangACMassaguéJ2008Molecular basis of metastasis. New England Journal of Medicine3592814–2823.doi:10.1056/NEJMra0805239.
ChristoforiG2006New signals from the invasive front. Nature441444–450.doi:10.1038/nature04872.
CollettaGCiraficiAMDi CarloA1989Dual effect of transforming growth factor β on rat thyroid cells: inhibition of thyrotropin-induced proliferation and reduction of thyroid-specific differentiation markers. Cancer Research493457–3462.
CoppaAMincioneGMammarellaSRanieriACollettaG1995Epithelial rat thyroid cell clones, escaping from transforming growth factor β negative growth control, are still inhibited by this factor in the ability to trap iodide. Cell Growth and Differentiation6281–290.
CostamagnaEGarciaBSantistebanP2004The functional interaction between the paired domain transcription factor Pax8 and Smad3 is involved in transforming growth factor-β repression of the sodium/iodide symporter gene. Journal of Biological Chemistry2793439–3446.doi:10.1074/jbc.M307138200.
DepoortereFPirsonIBartekJDumontJERogerPP2000Transforming growth factor β(1) selectively inhibits the cyclic AMP-dependent proliferation of primary thyroid epithelial cells by preventing the association of cyclin D3-cdk4 with nuclear p27(kip1). Molecular Biology of the Cell111061–1076.
ten DijkePHillCS2004New insights into TGF-β-Smad signalling. Trends in Biochemical Sciences29265–273.doi:10.1016/j.tibs.2004.03.008.
DupontSZacchignaLCordenonsiMSoligoSAdornoMRuggeMPiccoloS2005Germ-layer specification and control of cell growth by Ectodermin, a Smad4 ubiquitin ligase. Cell12187–99.doi:10.1016/j.cell.2005.01.033.
ErsonAEPettyEM2008MicroRNAs in development and disease. Clinical Genetics74296–306.doi:10.1111/j.1399-0004.2008.01076.x.
FranzénÅPiekEWestermarkBten DijkePHeldinNE1999Expression of transforming growth factor-β1, activin A, and their receptors in thyroid follicle cells: negative regulation of thyrocyte growth and function. Endocrinology1404300–4310.doi:10.1210/en.140.9.4300.
GoretzkiPEFrillingASimonDRoeherHD1990Growth regulation of normal thyroids and thyroid tumors in man. Recent Results in Cancer Research11848–63.
de GrootRPKruijerW1990Transcriptional activation by TGF β1 mediated by the dyad symmetry element (DSE) and the TPA responsive element (TRE). Biochemical and Biophysical Research Communications1681074–1081.doi:10.1016/0006-291X(90)91139-J.
Grubeck-LoebensteinBBuchanGSadeghiRKissonerghisMLondeiMTurnerMPirichKRokaRNiederleBKassalH1989Transforming growth factor β regulates thyroid growth. Role in the pathogenesis of nontoxic goiter. Journal of Clinical Investigation83764–770.doi:10.1172/JCI113955.
GurdonJBBourillotPY2001Morphogen gradient interpretation. Nature413797–803.doi:10.1038/35101500.
HeldinNEBergströmDHermanssonABergenstråhleANakaoAWestermarkBten DijkeP1999Lack of responsiveness to TGF-β1 in a thyroid carcinoma cell line with functional type I and type II TGF-β receptors and Smad proteins, suggests a novel mechanism for TGF-β insensitivity in carcinoma cells. Molecular and Cellular Endocrinology15379–90.doi:10.1016/S0303-7207(99)00086-6.
HillCS2001TGF-β signalling pathways in early Xenopus development. Current Opinion in Genetics and Development11533–540.doi:10.1016/S0959-437X(00)00229-X.
HillCS2009Nucleocytoplasmic shuttling of Smad proteins. Cell Research1936–46.doi:10.1038/cr.2008.325.
HoganCSerpenteNCogramPHoskingCRBialuchaCUFellerSMBragaVMBirchmeierWFujitaY2004Rap1 regulates the formation of E-cadherin-based cell–cell contacts. Molecular and Cellular Biology246690–6700.doi:10.1128/MCB.24.15.6690-6700.2004.
HoodlessPATsukazakiTNishimatsuSAttisanoLWranaJLThomsenGH1999Dominant-negative Smad2 mutants inhibit activin/Vg1 signaling and disrupt axis formation in Xenopus. Developmental Biology207364–379.doi:10.1006/dbio.1998.9168.
ImamuraTTakaseMNishiharaAOedaEHanaiJKawabataMMiyazonoK1997Smad6 inhibits signalling by the TGF-β superfamily. Nature389622–626.doi:10.1038/39355.
IulianoRTrapassoFLe PeraISchepisFSamaIClodomiroADumonKRSantoroMChiariottiLVigliettoG2003An adenovirus carrying the rat protein tyrosine phosphatase eta suppresses the growth of human thyroid carcinoma cell lines in vitro and in vivo. Cancer Research63882–886.
KeetonMRCurridenSAvan ZonneveldAJLoskutoffDJ1991Identification of regulatory sequences in the type 1 plasminogen activator inhibitor gene responsive to transforming growth factor β. Journal of Biological Chemistry26623048–23052.
KimuraETKoppPZbaerenJAsmisLMRuchtiCMacielRMStuderH1999Expression of transforming growth factor β1, β2, and β3 in multinodular goiters and differentiated thyroid carcinomas: a comparative study. Thyroid9119–125.doi:10.1089/thy.1999.9.119.
KogaiTSajid-CrockettSNewmarchLSLiuYYBrentGA2008Phosphoinositide-3-kinase inhibition induces sodium/iodide symporter expression in rat thyroid cells and human papillary thyroid cancer cells. Journal of Endocrinology199243–252.doi:10.1677/JOE-08-0333.
LagnaGHataAHemmati-BrivanlouAMassaguéJ1996Partnership between DPC4 and SMAD proteins in TGF-β signalling pathways. Nature383832–836.doi:10.1038/383832a0.
LazzereschiDRanieriAMincioneGTaccognaSNardiFCollettaG1997Human malignant thyroid tumors displayed reduced levels of transforming growth factor β receptor type II messenger RNA and protein. Cancer Research572071–2076.
LazzereschiDNardiFTurcoAOttiniLD'AmicoCMariani-CostantiniRGulinoACoppaA2005A complex pattern of mutations and abnormal splicing of Smad4 is present in thyroid tumours. Oncogene245344–5354.doi:10.1038/sj.onc.1208603.
LiuFPouponnotCMassaguéJ1997Dual role of the Smad4/DPC4 tumor suppressor in TGFβ-inducible transcriptional complexes. Genes and Development113157–3167.doi:10.1101/gad.11.23.3157.
LönnPMorénARajaEDahlMMoustakasA2009Regulating the stability of TGFβ receptors and Smads. Cell Research1921–35.doi:10.1038/cr.2008.308.
MatsuoSELeoniSGColquhounAKimuraET2006Transforming growth factor-b1 and activin A generate antiproliferative signaling in thyroid cancer cells. Journal of Endocrinology190141–150.doi:10.1677/joe.1.06713.
MeirelesAMPretoARochaASRebochoAPMáximoVPereira-CastroIMoreiraSFeijãoTBotelhoTMarquesR2007Molecular and genotypic characterization of human thyroid follicular cell carcinoma-derived cell lines. Thyroid17707–715.doi:10.1089/thy.2007.0097.
MorénAItohSMoustakasAten DijkePHeldinCH2000Functional consequences of tumorigenic missense mutations in the amino-terminal domain of Smad4. Oncogene194396–4404.doi:10.1038/sj.onc.1203798.
MorosiniPTaccalitiADi LoretoCArnaldiGFaloiaEGiacchettiGManteroF1994Transforming growth factor-β 1 is more expressed in thyroid follicular adenoma than in normal tissue. Journal of Endocrinological Investigation17335–340.
MorrisJCIIIRanganathanGHayIDNelsonREJiangNS1988The effects of transforming growth factor-β on growth and differentiation of the continuous rat thyroid follicular cell line, FRTL-5. Endocrinology1231385–1394.doi:10.1210/endo-123-3-1385.
MottiMLCalifanoDBaldassarreGCelettiAMerollaFForzatiFNapolitanoMTaverniseBFuscoAVigliettoG2005Reduced E-cadherin expression contributes to the loss of p27kip1-mediated mechanism of contact inhibition in thyroid anaplastic carcinomas. Carcinogenesis261021–1034.doi:10.1093/carcin/bgi050.
MüllerNReinacher-SchickABaldusSvan HengelJBerxGBaarAvan RoyFSchmiegelWSchwarte-WaldhoffI2002Smad4 induces the tumor suppressor E-cadherin and P-cadherin in colon carcinoma cells. Oncogene216049–6058.doi:10.1038/sj.onc.1205766.
NakaoAAfrakhteMMorénANakayamaTChristianJLHeuchelRItohSKawabataMHeldinNEHeldinCH1997Identification of Smad7, a TGFβ-inducible antagonist of TGF-β signaling. Nature389631–635.doi:10.1038/39369.
NicolussiAD'InzeoSSantulliMCollettaGCoppaA2003TGF-β control of rat thyroid follicular cells differentiation. Molecular and Cellular Endocrinology2071–11.doi:10.1016/S0303-7207(03)00238-7.
NicolussiAD'InzeoSGismondiACoppaA2006Reduction of invasive potential in K-ras-transformed thyroid cells by restoring of TGF-β pathway. Clinical and Experimental Metastasis23237–248.doi:10.1007/s10585-006-9023-0.
PaduaDMassaguéJ2009Roles of TGFβ in metastasis. Cell Research1989–102.doi:10.1038/cr.2008.316.
PangXPHershmanJMChungMPekaryAE1989Characterization of tumor necrosis factor-α receptors in human and rat thyroid cells and regulation of the receptors by thyrotropin. Endocrinology1251783–1788.doi:10.1210/endo-125-4-1783.
PattersonGIPadgettRW2000TGF β-related pathways. Roles in Caenorhabditis elegans development. Trends in Genetics1627–33.doi:10.1016/S0168-9525(99)01916-2.
PekaryAEHershmanJM1998Tumor necrosis factor, ceramide, transforming growth factor-β1, and aging reduce Na+/I− symporter messenger ribonucleic acid levels in FRTL-5 cells. Endocrinology139703–712.doi:10.1210/en.139.2.703.
PierreuxCENicolásFJHillCS2000Transforming growth factor β-independent shuttling of Smad4 between the cytoplasm and nucleus. Molecular and Cellular Biology209041–9054.doi:10.1128/MCB.20.23.9041-9054.2000.
PoméranceMQuillardJChantouxFYoungJBlondeauJP2006High-level expression, activation, and subcellular localization of p38-MAP kinase in thyroid neoplasms. Journal of Pathology209298–306.doi:10.1002/path.1975.
Riesco-EizaguirreGRodríguezIDe la ViejaACostamagnaECarrascoNNistalMSantistebanP2009The BRAFV600E oncogene induces transforming growth factor β secretion leading to sodium iodide symporter repression and increased malignancy in thyroid cancer. Cancer Research698317–8325.doi:10.1158/0008-5472.CAN-09-1248.
RosenwaldIB2004The role of translation in neoplastic transformation from a pathologist's point of view. Oncogene233230–3247.doi:10.1038/sj.onc.1207552.
SalernoPDe FalcoVTamburrinoANappiTCVecchioGSchweppeREBollagGSantoroMSalvatoreG2010Cytostatic activity of adenosine triphosphate-competitive kinase inhibitors in BRAF mutant thyroid carcinoma cells. Journal of Clinical Endocrinology and Metabolism95450–455.doi:10.1210/jc.2009-0373.
SanchoMVieiraJMCasalouCMesquitaMPereiraTCavacoBMDiasSLeiteV2006Expression and function of the chemokine receptor CCR7 in thyroid carcinomas. Journal of Endocrinology191229–238.doi:10.1677/joe.1.06688.
ScarpinoSDi NapoliAMelottiFTalericoCCancriniARucoL2007Papillary carcinoma of the thyroid: low expression of NCAM (CD56) is associated with downregulation of VEGF-D production by tumour cells. Journal of Pathology212411–419.doi:10.1002/path.2183.
SchutteMHrubanRHHedrickLChoKRNadasdyGMWeinsteinCLBovaGSIsaacsWBCairnsPNawrozH1996DPC4 gene in various tumor types. Cancer Research562527–2530.
SchweppeREKlopperJPKorchCPugazhenthiUBenezraMKnaufJAFaginJAMarlowLACoplandJASmallridgeRC2008Deoxyribonucleic acid profiling analysis of 40 human thyroid cancer cell lines reveals cross-contamination resulting in cell line redundancy and misidentification. Journal of Clinical Endocrinology and Metabolism934331–4341.doi:10.1210/jc.2008-1102.
ShiYMassaguéJ2003Mechanisms of TGF-β signaling from cell membrane to the nucleus. Cell113685–700.doi:10.1016/S0092-8674(03)00432-X.
ShimuraHSuzukiHMiyazakiAFuruyaFOhtaKHaraguchiKEndoTOnayaT2001Transcriptional activation of the thyroglobulin promoter directing suicide gene expression by thyroid transcription factor-1 in thyroid cancer cells. Cancer Research613640–3646.
ShiouSSinghABMoorthyKDattaPKWashingtonMKBeauchampRDDhawanP2007Smad4 regulates claudin-1 expression in a transforming growth factor-β-independent manner in colon cancer cells. Cancer Research671571–1579.doi:10.1158/0008-5472.CAN-06-1680.
SuzukiCMurakamiGFukuchiMShimanukiTShikauchiYImamuraTMiyazonoK2002Smurf1 regulates the inhibitory activity of Smad7 by targeting Smad7 to the plasma membrane. Journal of Biological Chemistry27739919–39925.doi:10.1074/jbc.M201901200.
TakakuKOshimaMMiyoshiHMatsuiMSeldinMFTaketoMM1998Intestinal tumorigenesis in compound mutant mice of both Dpc4 (Smad4) and Apc genes. Cell92645–656.doi:10.1016/S0092-8674(00)81132-0.
TatonMLamyFRogerPPDumontJE1993General inhibition by transforming growth factor β 1 of thyrotropin and cAMP responses in human thyroid cells in primary culture. Molecular and Cellular Endocrinology9513–21.doi:10.1016/0303-7207(93)90024-E.
TianXDuHFuXLiKLiAZhangY2009Smad4 restoration leads to a suppression of Wnt/β-catenin signaling activity and migration capacity in human colon carcinoma cells. Biochemical and Biophysical Research Communications380478–483.doi:10.1016/j.bbrc.2009.01.124.
TurcoACoppaAAloeSBaccheschiGMorroneSZupiGCollettaG1999Overexpression of transforming growth factor β-type II receptor reduces tumorigenicity and metastastic potential of K-ras-transformed thyroid cells. International Journal of Cancer8085–91.doi:10.1002/(SICI)1097-0215(19990105)80:1<85::AID-IJC17>3.0.CO;2-P.
VaskoVEspinosaAVScoutenWHeHAuerHLiyanarachchiSLarinASavchenkoVFrancisGLde la ChapelleA2007Gene expression and functional evidence of epithelial to mesenchymal transition in papillary thyroid carcinoma invasion. PNAS1042803–2808.doi:10.1073/pnas.0610733104.
ViscontiRFedericoACoppolaVPentimalliFBerlingieriMTPallantePKruhofferMOrntoftTFFuscoA2007Transcriptional profile of Ki-Ras-induced transformation of thyroid cells. Cancer Investigation25256–266.doi:10.1080/07357900701206406.
VisoneRRussoLPallantePDe MartinoIFerraroALeoneVBorboneEPetroccaFAlderHCroceCM2007MicroRNAs (miR)-221 and miR-222, both overexpressed in human thyroid papillary carcinomas, regulate p27Kip1 protein levels and cell cycle. Endocrine-Related Cancer14791–798.doi:10.1677/ERC-07-0129.
WatanabeMMasuyamaNFukudaMNishidaE2000Regulation of intracellular dynamics of Smad4 by its leucine-rich nuclear export signal. EMBO Reports1176–182.doi:10.1093/embo-reports/kvd029.
WranaJLAttisanoLCarcamoJZentellaADoodyJLaihoMWangXFMassaguéJ1992TGF β signals through a heteromeric protein kinase receptor complex. Cell711003–1014.doi:10.1016/0092-8674(92)90395-S.
WyllieFSDawsonTBondJAGoretzkiPGameSPrimeSWynford-ThomasD1991Correlated abnormalities of transforming growth factor-β1 response and p53 expression in thyroid epithelial cell transformation. Molecular and Cellular Endocrinology7613–21.doi:10.1016/0303-7207(91)90255-Q.
YinglingJMWangXFBassingCH1995Signaling by the transforming growth factor-β receptors. Biochimica et Biophysica Acta1242115–136.
*(S D'Inzeo and A Nicolussi contributed equally to this work)
