MAPK and SHH pathways modulate type 3 deiodinase expression in papillary thyroid carcinoma

Abstract

Type 3 deiodinase (DIO3, D3) is reactivated in human neoplasias. Increased D3 levels in papillary thyroid carcinoma (PTC) have been associated with tumor size and metast atic disease. The objective of this study is to investigate the signaling pathways involved in DIO3 upregulation in PTC. Experiments were performed in human PTC cell lines (K1 and TPC-1 cells) or tumor samples. DIO3 mRNA and activity were evaluated by real-time PCR and ion-exchange column chromatography respectively. Western blot analysis was used to determine the levels of D3 protein. DIO3 gene silencing was performed via siRNA transfection. DIO3 mRNA levels and activity were readily detected in K1 (BRAFV600E) and, at lower levels, in TPC-1 (RET/PTC1) cells (P!0.007 and PZ0.02 respectively). Similarly, DIO3 mRNA levels were higher in PTC samples harboring the BRAFV600E mutation as compared with those with RET/PTC1 rearrangement or negative for these mutations (P!0.001). Specific inhibition of BRAF oncogene (PLX4032, 3 mM), MEK (U0126, 10–20 mM) orp38 (SB203580, 10–20 mM) signaling was associated withdecreases in DIO3 expression in K1 and TPC-1 cells. Additionally, the blockage of the sonic hedgehog (SHH) pathway by cyclopamine (10 mM) resulted in markedly decreases in DIO3 mRNA levels.

Interestingly, siRNA-mediated DIO3 silencing induced decreases on cyclin D1 expression and partial G1 phase cell cycle arrest, thereby downregulating cell proliferation. In conclusion, sustained activation of the MAPK and SHH pathways modulate the levels of DIO3 expression in PTC. Importantly, DIO3 silencing was associated with decreases in cell proliferation, thus suggesting a D3 role in tumor growth and aggressiveness.

Introduction

Thyroid hormone (TH) influences a wide variety of biological processes including the balance between cell proliferation and differentiation. TH homeostasis is pro132-hormone thyroxine (T4) into the biologically active form triiodothyronine (T3) via outer ring deiodina- tion. In contrast, type 3 iodothyronine deiodinase (DIO3, D3) catalyzes the inactivation of T4 and T3 via inner ring deiodination (Maia et al. 2011). Several studies have shown an association between thyroidal status and tumor pathogenesis (Pinto et al. 2011, Kim & Cheng 2013) while several reports have documented changes in the expression of deiodinases in benign and malignant tumors (Huang et al. 2001, Murakami et al. 2001, de Souza Meyer et al. 2005, Dentice et al. 2007, 2012, Meyer et al. 2007).
D3, a known fetal protein, is reactivated in human neoplasias and has been associated with tumor behavior. The upregulation of DIO3 gene occurs at the transcrip- tional level, and it might be driven by disruption of several signaling pathways such as MAPK, transforming growth factor beta (TGFB), sonic hedgehog (SHH) and B-catenin (Dentice & Salvatore 2011). Increased levels of D3 were demonstrated in mouse and human malignant basal cell carcinoma (BCC). In BCC, the DIO3 induction occurs via activation of the SHH/GLI pathway and is associated with increased levels of cyclin D1 and keratinocyte proliferation (Dentice et al. 2007). High levels of DIO3 mRNA also occur in human intestinal adenomas and carcinomas (Dentice et al. 2012). Interest- ingly, the activating TH enzyme, D2, was downregulated by SHH and B-catenin activation. Thus, this opposite but convergent action of D2 and D3 enzymes seems to contribute to the local TH attenuation, leading to proliferative effects.

Papillary thyroid carcinoma (PTC) is the most common malignant thyroid tumor, occurring in 85–90% of cases (Pellegriti et al. 2013). Aberrant activation of MAPK pathway is a hallmark in PTC. The BRAFV600E mutation is the most common genetic event, observed in w50% of PTC, while RET/PTC rearrangement occurs in w20% and RAS mutations in 10–15% of cases (Kimura et al. 2003, Fagin & Mitsiades 2008). Increased levels of D3 were shown in PTC and associated with advanced disease (Romitti et al. 2012). Augmented D3 levels were also observed in follicular thyroid carcinoma but not in medullary or anaplastic thyroid carcinoma samples. The reasons for these distinct patterns of D3 expression is still lacking.

In the present study, we sought to determine the signaling pathways involved in DIO3 upregulation in PTC as well as to investigate whether DIO3 induction is associated with neoplastic thyroid cell proliferation.

Material and methods

Cell culture

Studies evaluating DIO3 gene regulation were performed in two human PTC-derived cell lines, which endogenously express the DIO3 gene: K1 cell line (European Collection of Cell Cultures; ECACC, Porton Down, Salisbury, UK), which carries the BRAFV600E mutation, and TPC-1 cells (kindly provided by Dr J A Fagin to Dr E T Kimura, Memorial Sloan Kettering Cancer Center, New York, NY; Human Oncology and Pathogenesis Program, Memorial Sloan-Kettering Cancer Center, New York, NY, USA) harboring the RET/PTC1 rearrangement (Geraldo et al. 2012). The origin in human PTC and gene expression profile have already been demonstrated for both cell lines (Schweppe et al. 2008, Saiselet et al. 2012). K1 cells were grown in DMEM: Ham’s F12:MCDB 105 (2:1:1 (Life Technologies, Invitro- gen)) plus 2 mM glutamine and 10% fetal bovine serum (FBS) (Life Technologies, Invitrogen). TPC-1 cells were grown in DMEM containing 10% FBS. Additionally, we used a human medullary thyroid carcinoma (MTC) cell line carrying the RET C634Y mutation – TT cells (Carlomagno et al. 1995; American Type Culture Collec- tion, ATCC, Manassas, VA, USA) – to determine the effect of SHH on DIO3 reactivation. TT cells were grown in RPMI (Life Technologies, Invitrogen) medium supplemented with 10% FBS. All experiments were performed between passages 15–25 for K1 cells, 20–25 for TPC-1 cells and 10–15
for TT cells. All cells were maintained at 37 8C in a humidified atmosphere of 5% CO2 and 95% air, and the culture medium was changed three times a week. Experiments were performed in cell cultures at 60–80% confluence. All analyses were performed in triplicate in at least three independent experiments.

Human PTC samples

Neoplastic tissues were collected from 25 unselected PTC patients who underwent thyroidectomy at our institution. Non-tumor thyroid tissues adjacent to tumor were also collected. All PTC and normal tissue samples (nZ9) were examined by a pathologist to exclude the presence of necrosis and/or inflammation. Tumors were histologically classified according to WHO recommendations (Hedinger et al. 1989). The Ethical Committee of the Hospital de Cl´ınicas de Porto Alegre, Porto Alegre, Brazil, approved the study.

Fixed Tissue System (Promega Corporation), according to the manufacturer’s instructions. The BRAF exon 15 was amplified by PCR using specific oligonucleotides (Supple- mentary Table S1, see section on supplementary data given at the end of this article). For PCR amplification, we used 50–100 ng/ml of DNA in a reaction mix (25 ml) containing 20 mM Tris–HCl (pH 8.0), 50 mM KCl, 2 mM MgCl2, 0.2 mM dNTPs, 0.2 mM of each primer and 1.25 U Platinum Taq DNA polymerase (Invitrogen). BRAFV600E mutation genotyping was performed by direct sequencing (Life Technologies, Applied Biosystems).

For RET/PTC1 detection, total RNA was extracted from PTC samples using the Trizol Reagent, and cDNA was generated using the Super Script III First-Strand Synthesis System (Life Technologies, Invitrogen, Carlsbad, CA, USA). Detection of RET/PTC1 rearrangement was performed by RT-PCR. We used the forward primer for H4 gene and the reverse primer for the tyrosine kinase domain of RET (Supplementary Table S1). PCR conditions were performed according to a previously described protocol (Sapio et al. 2007). Afterwards, 10 ml of the PCR product were analyzed by electrophoresis in a 1.5% agarose gel. TPC-1 cells were used as a positive control. Positive samples were subjected to direct sequencing to confirm the presence of RET/PTC rearrangement.

Immunohistochemistry studies

D3 expression was evaluated by immunohistochemistry studies in matched normal thyroid tissue, primary PTC tumor and lymph node metastases from two PTC patients (nZ18 and nZ19; Table 1). Immunohistochemistry analysis of D3 was performed on 6-mm sections of formalin-fixed and paraffin-embedded tissues. The immu- nohistochemical technique comprised tissue deparaffini- zation and rehydration, antigenic recovery, inactivation
of endogenous peroxidase and blockage of unspecific reactions. D3 anti-rabbit polyclonal antibody (Novus Biologicals, Littleton, CO, USA) was incubated overnight at a temperature of 4 8C, at dilution of 1:50 (the same antibody concentration was also used for positive control), followed by subsequent incubation with biotinylated
secondary antibody, streptavidin–HRP conjugate (LSAB; Dako, Carpinteria, CA, USA) and diaminobenzidine tetrahydrochloride (Kit DAB; Dako). Placenta tissue was used as a positive control. The negative control was obtained by omission of the primary antibody. The slides were examined using an Olympus BX51 microscope with an Olympus QColor 5 camera. The QCapturePro software was used to capture the images.

Real-time PCR

Total RNA was extracted from K1 and TPC-1 cells using the RNeasy minikit (Qiagen). Trizol Reagent was used for total RNA extraction from PTC samples and surrounding thyroid tissues. One microgram of RNA was reverse transcribed into cDNA using the SuperScript III First- Strand Synthesis System for RT-PCR (Life Technologies, Invitrogen), following the manufacturer’s protocol for the oligo (dT) method. RT-qPCR experiments were performed in a 7500 Fast Real-Time PCR System Thermal Cycler with 7500 FAST System Sequence Detection 1.4 Software (Life Technologies, Applied Biosystems). Experiments were performed by real-time monitoring of the increase in fluorescence of SYBR Green dye. The oligonucleotides used are described in Supplementary Table S1, see section on supplementary data given at the end of this article. Each sample was assayed in triplicate, and a negative control was included in each experiment. Standard curves representing five-point serial dilution of cDNA were analyzed and used as calibrators of the relative quantifi- cation of product generated in the exponential phase of the amplification curve. The R2 was greater than 0.99, while the amplification efficiency was higher than 98%. Quantification of DIO3, DIO2 and GAPDH cDNA were performed by relative quantification using the compara- tive DDCT method and expressed relative to the reference gene (cyclophilin A). Changes in gene expression were expressed as relative fold difference (n-fold change) or as arbitrary units (AUs).

Inhibition of MAPK and SHH signaling

To evaluate the effect of MAPK signaling activation on DIO3 induction in K1 and TPC-1 cell lines, we performed studies using specific inhibitors to the signaling effectors
MEK (U0126: 10–20 mM; Sigma–Aldrich), p38 (SB203580: 10–20 mM; Sigma–Aldrich) and BRAF mutated (PLX4032: 3 mM; Selleck Chemicals, Houston, TX, USA). Addition- ally, to assess the role of the SHH pathway on DIO3 regulation, we used the specific inhibitor of the smooth- ened (a SHH signaling effector), cyclopamine (10 mM, Sigma–Aldrich). The recombinant human SHH (1 mg/ml, R&D Systems, Minneapolis, MN, USA) was used to induce SHH activation in TT cells. Controls were incubated with mediumCvehicle (1% DMSO, Sigma–Aldrich). Cells were incubated during 24 h and then were harvested and processed for total RNA or total protein extraction.

D3 activity assay

D3 activity was determined in PTC cells by ion-exchange column chromatography (Wajner et al. 2011). After concluding the experiments, 300 ml of medium was collected, and the reaction was stopped with 200 ml horse serum and 100 ml 50% TCA, which was followed by centrifugation at 12 000 g for 2 min to precipitate the non-metabolized (125I)T3. The supernatant was used to determine (125I)T2 and (125I)T1 levels. The Sephadex LH-20 column was equilibrated with 0.1 M HCl, and an equal volume of 0.1 M HCl was added to 500 ml samples and then mixed. Stepwise elution was performed by successive application of 2X1 ml of 0.1 M HCl (for 125I release), 6X1 ml of 0.1 M NaOH–ethanol (8:1 v/v (125I) for T1 release) and 4X1 ml of 50% ethanol in 0.1 M NaOH (1:1 v/v (125I) for T2 release). The 1-ml fractions were collected and counted for radioactivity. The D3 activity was calculated by multiplying the fractional conversion by the T3 concentration in the media and expressed as T3 inactivation (fmol/mg protein per 24 h).

Western blot analysis

Cultured cells were lysed and prepared for western blot analysis, as previously described (Fuziwara & Kimura 2014). Briefly, 30–50 mg protein of each sample was fractionated by 8–12% SDS–PAGE and blotted onto an Immobilon PVDF membrane (Millipore, Billerica, MA, USA). Non-specific binding sites were blocked by incu- bation with 5% nonfat dry milk in Tris-buffered saline 0.1% Tween-20. The following primary antibodies were used: anti-D3 (1:400; Novus Biologicals), anti-ERK1/2 (1:400; Santa Cruz Biotechnology), anti-phospho-ERK1/2 (1:400; Santa Cruz Biotechnology), anti-p38 (1:500; Cell Signaling), anti-phospho-p38 (1:200; Santa Cruz Biotechnology), anti-cyclin D1 (1:400; Santa Cruz Biotechnology), anti-Gli1 (1:400; Cell Signaling Tech- nology), anti-a-tubulin B7 (1:500; Santa Cruz Bio- technology) and anti-b-actin (1:10 000; Sigma–Aldrich). The antigen–antibody complexes were visualized using HRP-conjugated secondary antibody and an enhanced chemiluminescence system (GE Healthcare, Pittsburgh, PA, USA). Expression was quantified using image densito- metry with ImageJ analysis software.

siRNA transfection

siRNA studies were performed to evaluate the specific effects of DIO3 inhibition on cell proliferation. The siRNAs were as follows: Silencer Select GAPDH siRNA (#4390849, Life Technologies, Ambion Inc., Austin, TX, USA) was used as positive control; Silencer Select Negative Control (#4390843, Life Technologies, Ambion Inc.) and Silencer Pre-designed DIO3 siRNA (#AM16708, Life Technologies, Ambion Inc.) were used in the experiments. Transfection studies were performed using Lipofectamine RNAiMAX reagent, according to the manufacturers’ instructions (Life Technologies, Invitrogen). A total of 15X104 cells/well (K1 and TPC-1) were plated in six-well plates and transfected with 40 pmol of GAPDH siRNA, 100 pmol of silencer negative and 100 pmol of DIO3 siRNA. All analyses were performed in triplicate and in at least three independent experiments.

Cell proliferation assays

Absolute cell number count and flow cytometry were performed to evaluate cell proliferation. Initially, 15X104 cells/well (K1 and TPC-1) were plated in six-well plates, transfected with 100 pmol of DIO3 siRNA and incubated for 48 h. The cells were trypsinized, and the absolute number was counted using the Neubauer chamber. To evaluate the effect of DIO3 expression on the cell cycle status, K1 cells were incubated with DIO3 siRNA. After 48 h, cells were washed with PBS and then resuspended in 50 mg/ml
cycle stage is shown as frequency. To compare D3 levels among the groups, we used t-test or one-way ANOVA. c2 test was used to compare the differences in the proportion of cells in different stages of cell cycle. The SPSS 18.0 and Prism 5.0 software were used for all analyses, and P!0.05 was considered statistically significant.

Results
MAPK activation induces DIO3 levels in PTC cell lines

To evaluate the role of MAPK activation in DIO3 regulation, experimental studies were performed in two distinct human PTC cell lines, K1 cells carrying the BRAFV600E mutation and TPC-1 cells harboring the RET/PTC1 rearrangement. We observed that, although DIO3 expression was readily detected in both cell lines (Fig. 1A), the levels of DIO3 mRNA and activity were significantly higher in K1 cells as compared to those observed in TPC-1 cells (0.32 vs 0.11 AU, PZ0.007; 14.9 vs 8.1 fmol/mg.prot.24 h, PZ0.02 respectively; Fig. 1A and B).

Next, we evaluated the oncogenic effects of BRAFV600E mutation on DIO3 reactivation. The incubation of K1 cells with the specific BRAF-mutated inhibitor, PLX4032 (3 mM), decreased ERK phosphorylation in w30% (Fig. 2A) as well as the DIO3 mRNA levels (w2.2-fold, P!0.001; Fig. 2B). Similar results were obtained with higher PLX4032 doses, indicating that PLX4032-mediated decreases in ERK phosphorylation is not dose dependent (not shown). On the other hand, the incubation of K1 cells with increasing doses of MEK inhibitor (U0126, 10–20 mM) for 24 h resulted in a dose-dependent reduction in ERK phosphorylation (Fig. 3A) and in DIO3 expression (10- to 12.5-fold, P!0.001; Fig. 3B). Likewise, the incubation of K1 cells with SB203580, a specific inhibitor of p38 protein, caused a dose-dependent reduction of phospho-p38 levels propidium iodide and 0.1% Triton X-100 in sodium citrate solution. Cells were incubated on ice for at least 15 min. Marked cells were analyzed using Attune NxT Acoustic Focusing Cytometer (Life Technologies, Applied Biosystems). The data generated were analyzed using the FlowJo software. All experiments were performed in triplicate.

Blockage of SHH pathway prevents DIO3 upregulation in PTC

Next, we evaluated whether the SHH pathway and its downstream effector, GLI1, could be involved on DIO3 regulation in PTC. The SHH signaling was blocked using cyclopamine (10 mM). We observed a reduction in the GLI1 protein (Fig. 6A and C for K1 and TPC cells respectively), followed by a marked decrease in the DIO3 levels in K1 (w12-fold, P!0.001; Fig. 6B) as well as in TPC- 1 cells (w2.5-fold, P!0.001; Fig. 6D).

To identify whether MAPK and SHH cooperate in promoting DIO3 induction in PTC, we evaluated GLI1 protein expression under MEK and p38 proteins inhi- bition. We observed a reduction in GLI1 levels in K1 and TPC-1 cells after MAPK blockage with U0126 (Fig. 6E and G) or SB203580 (Fig. 6F and H), indicating that these signaling pathways might cooperate on DIO3 regulation in PTC. Of note, however, the treatment of cells with cyclopamine has no effect on ERK or P38 phosphorylation (data not shown).

We then sought to explore the role of SHH signaling pathway on DIO3 reactivation. We have previously shown that MTC, a C-cell originated tumor, does not express D3 enzyme (Romitti et al. 2012). Low levels of DIO3 mRNA were detected in MTC-derived TT cells while a marked increase in DIO3 expression was observed in recombinant SHH-treated TT cells (2.5-fold, P!0.001; Fig. 6I). In control. K1 and TPC-1 cells were transfected with the siRNAs and maintained for 48 h. The efficiency of siRNA transfection was established by the efficiency in silencing GAPDH gene (inhibition of 95% in K1 and w90% in TPC-1 cells, P!0.001). DIO3 gene knockdown resulted in a w90% blockage of DIO3 transcripts in both cell lines (P!0.001; Fig. 7A and C). Interestingly, DIO3 inhibition was associated with significant reduction in the absolute cell number, as compared with control (w30%, P!0.01; Fig. 7B and D) and decreased levels of cyclin D1 (Fig. 7E). Accordingly, further experiments showed that DIO3 silencing was associated with augmented proportion of cells in the G1 phase and a proportional reduction in the percentage of cells in S and G2 phases of cell cycle (w30%, P!0.005; Fig. 7F).

Discussion

In the present study, we have demonstrated that genetic alterations in the MAPK pathway effectors, such as BRAFV600E mutation and RET/PTC1 rearrangement, increase DIO3 levels. The SHH pathway is also involved in DIO3 upregulation once its signaling inhibition significantly reduces the DIO3 expression. Interestingly, siRNA-mediated DIO3 gene silencing decreases cyclin D1 levels while increasing the proportion of cells in the G1 phase of the cell cycle, reducing the proliferation of malignant thyroid cells.

A potential role of D3 in tumorigenesis has been postulated due to DIO3 upregulation in several benign and malignant tumors (Pallud et al. 1999, Huang et al. 2005, Dentice et al . 2007, 2012, Romitti et al. 2012). Studies performed in primary cultures of human fibroblasts showed that the DIO3 gene is induced by TGFB via SMAD and MAPK-dependent pathway (Huang et al. 2005). PTCs are known for carrying genetic alterations that lead to an aberrant and constitutive activation of MAPK pathway (Romitti et al. 2013). BRAFV600E is an oncogenic protein with markedly elevated kinase activity that overactivates the MAPK pathway, especially ERK signaling transduction (Xing 2005, Chakravarty et al. 2011). Here, we show that BRAFV600E mutation is associated with the highest levels of DIO3 expression, as compared with RET/PTC1 (Fig. 1A and B), which seems to be mediated by the stronger ERK phosphorylation induced by this oncogene (Fig. 3A and B). On the other hand, we also observed that the p38 inhibition led to a more pronounced decrease in the DIO3 levels in TPC-1 than in K1 cells (Fig. 3G and H; 3E and F respectively), thus suggesting that RET/PTC1 rearrangement induces the DIO3 levels mainly through this pathway. These observations are in agreement with previous studies which showed that RET/PTC1 rearrangement exerts its effects mainly by p38 phosphoryl- ation (Hayashi et al. 2000, Mariggio et al. 2007). Consist- ently, we also observed a stronger induction of DIO3 expression in human PTC samples carrying the BRAFV600E mutation, as compared to those harboring the RET/PTC1 or none of these genetic alterations (Fig. 4). Interestingly, similar D3 upregulation was observed in patient-matched tumor and lymph node metastasis, suggesting that the molecular mechanism responsible for D3 upregulation also occurs in the metastatic sites.

SHH signaling is critical for embryogenesis and other cellular processes such as proliferation and differentiation (Lum & Beachy 2004). Disruption in SHH signaling results in various human diseases and seems to contribute to neoplastic processes. The SHH reactivation occurs in up to 25% of human tumors and has been associated with DIO3 induction (Dentice et al. 2007, Aw et al. 2014). Here, we show that the SHH signaling also modulates DIO3 expression in PTC cells. The use of the SHH pathway inhibitor, cyclopamine, promoted a marked decrease in DIO3 transcripts, indicating a direct effect of SHH/GLI1 signaling on gene regulation (Fig. 6A, B, C, D). Moreover, recombinant SHH treatment induced the DIO3 expression in TT cells (Fig. 6I). Interestingly, the blockage of the MAPK pathway, using MEK or p38 inhibitors, reduced the levels of GLI1, thus suggesting that the DIO3 expression in PTC might be regulated by cooperation between the MAPK and SHH pathways, as previously demonstrated in other xenograft tumor, while the T3 addition promoted differ- entiation (Dentice et al. 2007, 2012). Nevertheless, the mechanisms underlying D3 and/or hypothyroidism induction of proliferation are still open to debate. Previous studies have shown that T3 is involved in regulation of cyclin D1 expression (Nyga˚rd et al. 2003, Garcia-Silva & Aranda 2004, Furumoto et al. 2005, Perri et al. 2014). In embryonic carcinoma cells, T3 treatment decreases the proliferation rate by arresting cells in the G1 and S phases while downregulates E2F1, both key regulators of prolifer- ation. These T3-mediated effects depend on active TRs (Nyga˚ rd et al. 2003). Accordingly, studies in mouse harboring a dominantly negative mutant TH receptor (TRbPV/PV) suggest that the absence of TRb, and consequent lacking T3 signaling, induces cyclin D1 expression, leading to the activation of the CDK/retino- blastoma (RB)/E2F pathway that mediates, at least in part, the aberrant proliferation of thyroid cells (Furumoto et al. 2005). In this study, we show that the siRNA-mediated DIO3 gene attenuation decreases the number and cyclin D1 levels (Fig. 7A, B, C, D, E), followed by a partial G1 phase cell cycle arrest (Fig. 7F). These results further support a pivotal role of DIO3 overexpression in neoplastic cell proliferation.

In conclusion, we have demonstrated that DIO3 expression is modulated by MAPK genetic alterations in M Romitti et al. Type 3 deiodinase regulation 23 :3 135–146
in PTC. SHH pathway activation might also be involved in DIO3 upregulation by a signaling cooperation with MAPK pathway. Finally, the reduction on cell proliferation rate in D3-depleted PTC cells supports the hypothesis that the decrease in intracellular TH levels might be associated with tumor growth and aggressiveness.