Background
The homeodomain transcription factor NKX2.5 is critical for differentiation and proliferation of the primitive pharyngeal endodermal cells [
1], being expressed up to the embryonic day 11.5 in the mouse thyroid primordium [
2]. However, NKX2.5 expression is discontinued just before the appearance of proteins that are crucial for thyroid hormone biosynthesis, such as thyroglobulin (Tg), thyroperoxidase (TPO), thyrotropin receptor (TSHR), sodium-iodide symporter (NIS) and dual oxidases (DuOx1 and 2) [
2‐
4].
Even though NKX2.5 expression seems to be restricted to heart in adults [
5], NKX2.5 has been described to be expressed in several types of tumors, including pediatric acute lymphoblastic leukemia [
6], skin squamous cell carcinoma [
7] and ovarian yolk sac tumor [
8]. However, the possible role of NKX2.5 in thyroid cancer remains elusive.
Thus, in the present work, we aimed at evaluating NKX2.5 expression by immunohistochemistry in 10 PTC samples. To assess NKX2.5 association with clinical prognostic, we have also included 51 patients from a well-characterized 10 years Brazilian cohort of PTC patients [
9]. We have also investigated the functional role of this transcription factor on the expression of thyroid differentiation markers and reactive oxygen species (ROS) production in normal thyroid cells (PCCL3). To our knowledge, this is the first study showing NKX2.5 expression in PTC samples.
Methods
Patients
To validate the expression of NKX2.5, we have included 10 patients admitted to the hospital in 2012, with the diagnosis of PTC, according to the International Classification of Diseases for Oncology (ICD-O) 80,503 (papillary thyroid carcinoma) in the register of the Cancer Hospital from the centre’s database. The eligible patients were submitted to thyroid surgery to treat PTC and were aged 18 years or more at the time of surgery.
To assess NKX2.5 association with clinical prognostic, we have also included 51 patients [
9] admitted to the hospital between January 1st, 1990, and December 31st, 1999, with a diagnosis of PTC as per the International Classification of Diseases for Oncology (ICD-O) codes 83,403, 82,603 or 80,503 (follicular variant of PTC, thyroid papillary adenocarcinoma (SOE), papillary thyroid carcinoma, respectively) in the register of the Cancer Hospital from the centre’s database, which was available in 2010. The eligible patients had undergone thyroidectomy as a type of thyroid surgical approach to treat PTC in the study period and were aged 18 years or more at time of initial surgery.
The three main outcome variables of interest were persistence, recurrence and PTC-free status. For this classification, we have only considered the first event experienced by the patient. Another outcome studied was death as a result of PTC and/or from any other cause. PTC persistence was defined as evident residual disease (active disease) until 12 months after initial surgical treatment. Furthermore, PTC recurrence was defined as having the first event of active disease occurring between 1 and 10 years of follow-up. Patients were considered PTC-free if they did not show active disease after the initial surgery, in the period between the initial surgery and ten years of follow-up. Active PTC disease was defined when one or more of the following was observed: (a) structural disease evidenced by positive imaging findings or after radioactive iodine (
131I, RAI) therapy; and (b) biochemical evidence of disease, with significant increase in serum thyroglobulin (Tg) levels during thyroid hormone treatment (levothyroxine, LT4) compared with previously stable levels and/or an increase in serum Tg levels after LT4 withdrawal (stimulated Tg). The clinicopathologic information in each case, including age, gender, treatment, pathologic stage and presence or absence of tumor recurrence or persistence, was obtained from medical records and tumor registries. The characteristics of the 51 patients selected for this study are summarized in Table
1.
Table 1
Clinicopathologic characteristics of PTC patients (n = 51)
Gender
| Male | 10 | 19.6 |
Female | 41 | 80.4 |
Age
| ≤40 years | 23 | 45.1 |
41–60 years | 11 | 21.6 |
> 60 years | 17 | 33.3 |
Type of surgery
| Total or near total thyroidectomy | 42 | 82.4 |
Partial thyroidectomy | 7 | 13.7 |
Others | 2 | 3.9 |
Neck dissection
| Yes | 27 | 52.9 |
No | 24 | 47.1 |
Tumour size
| ≤4 cm | 43 | 84.3 |
> 4 cm | 8 | 15.7 |
Pathological stage of primary tumor size (pT)
| pT1 and pT2 | 22 | 43.1 |
pT3 and pT4 | 28 | 54.9 |
pTx | 1 | 2.0 |
Pathological stage of regional lymph node (pN)
| pN1 | 30 | 58.8 |
pN2 | 3 | 5.9 |
pNx | 18 | 35.3 |
Pathological stage of distant metastasis (pM)
| pM0 | 42 | 82.4 |
pM1 | 9 | 17.6 |
Pathological stage (pTNM)
| pTNM I | 22 | 43.1 |
pTNM II | 3 | 5.9 |
pTNM III | 4 | 7.8 |
pTNM IV | 15 | 29.4 |
Unknown | 7 | 13.8 |
Histopathological classification
| Classic papillary carcinoma | 40 | 78.2 |
Follicular variant | 5 | 9.8 |
Tall cell variant | 2 | 4.0 |
Clear cell variant | 1 | 2.0 |
Solid variant | 1 | 2.0 |
Oncocytic variant | 1 | 2.0 |
Others | 1 | 2.0 |
Extrathyroid extravasation
| Yes | 30 | 58.8 |
No | 18 | 35.3 |
Unknown | 3 | 5.9 |
Vascular or angiolymphatic invasion
| Yes | 23 | 45.1 |
No | 22 | 43.1 |
Unknown | 6 | 11.8 |
Morbidity outcome
| Without recurrence or persistence | 21 | 41.2 |
Recurrence | 9 | 17.6 |
Persistence | 21 | 41.2 |
Place of outcome
| Without recurrence or persistence | 21 | 41.2 |
Local and/or regional lesion only | 11 | 21.6 |
Distant metastasis only | 8 | 15.7 |
Both (Local and regional lesion and distant metastasis) | 10 | 19.6 |
Unknown | 1 | 1.9 |
Lethality within 10 years follow-up
| Yes | 8 | 15.7 |
No | 43 | 84.3 |
Cause of death
| PTC | 6 | 75 |
Others | 2 | 25 |
Immunohistochemistry analysis
Immunohistochemistry (IHC) analysis was performed on paraffin sections of the papillary thyroid carcinomas mounted on glass slides. Tissue sections of heart necropsy, obtained from the “Departamento de Patologia da Universidade Federal Fluminense”, served as NKX2.5 positive control. For antigen retrieval, the slides were incubated in a pH 6.0 solution (target antigen retrieval solution) for 45 min in a water bath, at 96 °C, followed by a washing step with phosphate-buffered saline (PBS). Incubations with the primary antibody against NKX2.5 (polyclonal anti-NKX2.5, SAB2101601, Sigma-Aldrich; diluted 1:800) were performed overnight at 4 °C. Samples were then incubated with biotinylated secondary antibodies using the streptavidin-biotin-peroxidase kit (Strep ABC complex/HRP Duet kit, DAKO Cytomation). The reactions were developed with a solution containing diaminobenzidine tetrahydrochloride chromogen, and the sections were counterstained with Harris’s hematoxylin. Negative and positive controls were included in all reactions.
All the sections were assessed independently by two pathologists, who met to resolve discordant interpretations and establish a consensus categorization. A binary classification (positive vs. negative) was used to score the IHC. The positive slides were evaluated semi-quantitatively by the distribution of the immunohistochemical positivity (1–49% and 50–100%) of neoplastic cells. Whenever the distribution was < 50%, the cases were classified as low expression, and cases with ≥50% of cells stained were classified as high expression. Positive slides were also classified according to the subcellular distribution of NKX2.5 in cytoplasmic and nuclear.
Cell culture
The non-tumoral rat thyroid cell line (PCCL3) was kindly donated by dr. Fusco, from the Department of Molecular Medicine and Medical Biotechnology, Naples University, who developed this cell line. PCCL3 was maintained in Coon’s modified Ham’s F-12 medium (HiMedia Laboratories, Mumbai, India), supplemented with 5% FBS and a six-hormone mixture (1 mU/ml TSH, 10 μg/ml insulin, 5 μg/ml transferrin, 10 nM hydrocortisone, 10 ng/ml somatostatin, and 10 ng/ml glycyl-L-histidyl-L-lysine acetate) and maintained in a humidified 5% CO
2 incubator at 37 °C, as previously described [
10]. The highly transfectable cell line derivate from human embryonic kidney 293 cells (HEK293T) were grown in Dulbecco’s modified Eagle’s medium (DMEM) (HiMedia Laboratories, Mumbai, India), supplemented with 10% FBS.
Transient transfection assays
NKX2.5 overexpression
The
NKX2.5 plasmid (7036 bp) (p
NKX2.5) contains the full length of
NKX2.5 coding region (RefSeq NM_008700), cloned into pcDNA3.1 expression vector (Invitrogen, Carlsbad, California) [
11]. PCCL3 cells (1.2 × 10
5) were seeded in 24-well plate and 1 μg of p
NKX2.5 or pcDNA3.1 (empty vector, control) were transfected, using Lipofectamine LTX combined with PLUS reagent (Invitrogen, Carlsbad, California) diluted in Ham’s F-12 complete medium. All procedures were performed following the manufacturer’s recommendations.
The efficiency of the transient transfection was evaluated 24 h later by real time PCR with specific oligonucleotides for NKX2.5 gene and immunoblotting analysis using NKX2.5 primary antibody (SAB2101601-Sigma-Aldrich, St. Louis, MO, USA). Concomitantly, PCCL3 cells were transfected with the vector encoding the green fluorescent protein, in order to confirm the transfection efficiency, using the same procedure described above.
Real time PCR
Total RNA from cell line was extracted using the RNeasy® Plus Mini Kit (Qiagen, Valencia, California), following the manufacturer’s instructions and subsequently quantified by NanoVue™ Plus spectrophotometer (GE Healthcare, Sweden). Total RNA (0.5–1 μg) was reversely transcribed using MultiScribe™ Reverse Transcriptase (Applied Biosystems, Foster City, CA), in accordance to the manufacturer’s instructions. Reactions for the quantification of mRNA by real-time PCR were performed in an ABI Prism 7500 Sequence Detection System from Applied Biosystems, using 6 μl Maxima SYBR Green qPCR Master Mix (Thermo Scientific, Rockford, IL, USA), 0.5 μl specific oligonucleotides (150 nM), 2.5 μl DEPC water and 3 μl diluted cDNA. The oligonucleotides for real-time PCR were purchased from Applied Biosystems, designed with PrimerQuest software (Integrated DNA Technologies, San Diego, CA, USA) and are listed in Additional file
1.
RPL4 was used as internal control. Relative gene expression was determined by subtracting cycle threshold (CT) for the gene of interest from CT for the reference gene, calculated using the 2
-ΔΔCT method, as previously described [
12] and expressed as relative to control.
Western blot
Cells were homogenized in lysis buffer containing 135 mM NaCl, 1 mM MgCl2, 2.7 mM KCl, 20 mM Tris, pH 8.0, 1% Triton, 10% glycerol and protease and phosphatase inhibitors (0.5 mM Na3VO4, 10 mM NaF, 1 mM leupeptin, 1 mM pepstatin, 1 mM okadaic acid, and 0.2 mM phenylmethylsulfonyl fluoride), and then syringed five times. An aliquot was used to determine the concentration of protein by BCA protein assay kit (Pierce, Rockford, IL, USA), as recommended by the manufacturer. Samples were then subjected to SDS/PAGE electrophoresis, transferred to PVDF membranes, and probed with the following antibodies: 1:2000 polyclonal anti-NKX2.5, Sigma-Aldrich; 1:4000 monoclonal anti-GAPDH, Millipore; 1:2000 anti-rabbit IgG HRP-linked antibody and 1:4000 anti-mouse IgG HRP-linked antibodies from Cell Signaling. Detection of the proteins was performed using ECL (Thermo Scientific, Rockford, IL, USA).
Cell viability assay
As an index of cell viability, we used the commercially available MTT assay (Sigma-Aldrich, St. Louis, MO, USA), according to the recommendations of the manufacturer. The assay is based on the cellular conversion of the tetrazolium salt into formazan that is soluble in culture medium and is directly measured at 490 nm, in a 96-well plate, using a spectrophotometer. Absorbance is directly proportional to the number of living cells in culture. PCCL3 cells were transfected with pcDNA3.1 (control) or pNKX2.5 and 0, 24, 48 and 72 h later MTT assay was performed. Cells were incubated with MTT (0.5 mg/ml) for 3 h at 37 °C in a humidified 5% CO2 atmosphere. Then, cells were lysed with DMSO (PA). All determinations were made in triplicates and the results were expressed as relative to pcDNA3.1 in initial time (0 h).
Iodide uptake assay
Iodide uptake assay was performed as described by Souza et al. [
10]. Briefly, PCCL3 cells (1 × 10
5) were grown in 24-well plates, transfected with pcDNA3 empty vector or p
NKX2.5, as described above, and 24 h later they were incubated for 45 min at 37 °C in 1 ml Hank’s balanced salt solution (HBSS) containing 0.1 μCi carrier-free Na
125I and 100 μM NaI. For each experimental condition, a well also received 10 μM KClO
4, a competitive inhibitor of
NIS, in order to determine the nonspecific radioiodide uptake. After incubation, cells were washed once with ice-cold HBSS and lysed with 0.1 M NaOH, and the radioactivity was determined in a gamma counter (Compu-Gama, 1214, LKB Wallac). An aliquot of each sample was used to determine the protein concentration, using BCA protein assay kit (Pierce, Rockford, IL, USA), as recommended by the manufacturer. Specific iodide uptake value was obtained by subtracting iodide uptake in the absence and in the presence of KClO
4 and related to protein concentration. Results were expressed as specific units of iodide accumulation relative to control.
Extracellular H
2O
2 generation was quantified by the Amplex red/horseradish peroxidase assay, which detects the accumulation of a fluorescent oxidized product, as previously described [
13]. 24 h after transfection with pcDNA3.1 or p
NKX2.5, PCCL3 cells (1 × 10
5) were incubated in Dulbecco’s PBS (D-PBS) containing CaCl
2, MgCl
2, D-glucose (1 mg/ml), ionomycin (1 μM), superoxide dismutase (100 U/ml), horseradish peroxidase (0.5 U/ml), and Amplex red (50 μM), and the fluorescence was immediately measured in a microplate reader (VictorX4) for 30 min (excitation wavelength = 530 nm and emission wavelength = 595 nm). Hydrogen peroxide concentration was determined using standard calibration curves and the result was expressed as nmol H
2O
2 per hour per 10
5 cells. The measurements were made in the presence and in the absence of ionomycin, a calcium ionophore, since dual oxidases are calcium-dependent enzymes, and H
2O
2 generation obtained in the presence of ionomycin were subtracted from that obtained in the absence of ionomycin.
For gene reporter assay, we used the
DuOx1 promoter (pDuOx1) plasmid (1 μg), containing the proximal 5′-flanking region of
DuOx1 gene and the luciferase reporter gene cloned in the PgL3 vector [
14]. It was transiently transfected in combination with wild type or mutated
NKX2.5 (Ile
183➔Pro) expression vectors [
11] or the corresponding empty vector, pcDNA3.1 (500 ng) in HEK293T cells (2 × 10
5 cells/well), using Lipofectamine LTX combined with PLUS reagent (Invitrogen, Carlsbad, California), as described above. pRL-CMV, which contains renilla cDNA, was used to correct for transfection efficiency (Promega, Madison, WI, USA). After 24 h, cells were harvested and collected for luciferase and renilla activity by the Dual-Luciferase reporter assay system (Promega, Madison, WI, USA). Luminescence was measured in a Victor X4 Multilabel Plate Reader (PerkinElmer, Norwalk, CT, USA). Results were expressed as relative activity, compared to the control (pcDNA3.1) in each experiment.
Statistical analysis
For the statistical analysis, we used the R program (Free Software Foundation, USA) and commercially available software SPSS 20.0 (SPSS Inc., Chicago, IL). Descriptive statistics were used in a preliminary analysis of the relation between baseline variables and outcome events. The PTC-free group was considered as the reference group. The association between the clinic pathological variables and NKX2.5 immunoexpression were analyzed by the Fisher’s exact tests. In addition, the Kaplan-Mayer method was used to evaluate whether the presence of NKX2.5 impacts on morbidity outcomes (persistence/recurrence) and lethality. Tests were considered statistically significant when the p-value was < 0.05.
All in vitro results were expressed as mean ± SEM and were analyzed by the non-parametric Mann Whitney’s test (when comparing two groups) or by the non-parametric Kruskal-Wallis test followed by Dunn’s multiple comparison test (when comparing three or more groups). Statistical analyses were performed using the software Graphpad Prism (Version 5, Graphpad Software Inc., San Diego, USA) and the difference was considered significant when p < 0.05.
Discussion
The regulation of cell proliferation and differentiation is crucial for both normal development and carcinogenesis, therefore it is expected that genes critical for thyroid organogenesis could also play a role in thyroid cancer [
19]. In this study, we reported that NKX2.5 is highly expressed in PTC samples. The expression of NKX2.5 has been described in a variety of tumors, being correlated to a malignant transformation of the neoplastic cells [
6‐
8]. In agreement with that, we have observed a nuclear immunostaining of NKX2.5 in normal adjacent thyroid tissue, which suggest that the expression of this transcription factor could be an early event of thyroid carcinogenesis. Accordingly, our data revealed that NKX2.5 overexpression decreased mRNA levels of the thyroid differentiation markers (
TPO, NIS and
TSH receptor) in normal thyroid cells, which reinforces our hypothesis that NKX2.5 promotes a dedifferentiation phenotype in thyroid cells. Above all, we have found a reduction in iodide uptake, which could be due to oxidative damage of NIS, since NKX2.5 overexpression induced an increment in hydrogen peroxide generation, and/or to a direct effect of NKX2.5 down regulating NIS. It is important to underline that NIS regulation is of great relevance not only for thyroid physiology but also for the management of thyroid diseases, since radioiodine therapy is used to treat thyroid cancer and the loss of NIS is associated with poor prognosis of thyroid cancer patients [
20,
21].
Elevated amounts of ROS have been related to genomic instability and tumorigenesis in thyroid cells [
22] and H
2O
2-generating activity of DuOx1 seems to contribute to these events [
23]. Overexpression of NKX2.5 in PCCL3 cells resulted in enhanced H
2O
2 generation, which seems to be due to increased DuOx1 activity, since mRNA levels of
DuOx1 were up regulated by NKX2.5 and
NOX4 mRNA was not detected in PCCL3 in our assay conditions. Even though
DuOx2 and
DuOxA2 mRNA levels were reduced in cells transfected with
NKX2.5, it is well known that
DuOx1 is the main source of H
2O
2 in PCCL3 cells [
18]. Interestingly, our results suggest that NKX2.5 upregulates
DuOx1 at the transcriptional level, leading to increased ROS production and thus reinforcing the idea that NKX2.5 could predispose thyroid cells to carcinogenesis.
Herein, we have found a predominantly cytoplasmic expression of NKX2.5 in PTC samples. NKX2.5 activation and subcellular localization have been shown to be regulated by factors such as matrix stiffness and sumoylation [
24,
25]. Since sumoylation machinery, which is known to activate NKX2.5 [
25], is reduced in PTC [
26], we hypothesize that the cytoplasmic localization of NKX2.5 in PTC could be due to a reduction in NKX2.5 sumoylation.
Changes in DNA methylation have been shown to induce tumor initiation and progression [
27], and PTC exhibits global hypomethylation when compared to normal thyroid [
28]. Interestingly, the expression of NKX2.5 is regulated by methylation of the promoter region [
29] and hence, an epigenetic dysregulation of
NKX2.5 promoter region might be a plausible mechanism underlying NKX2.5 overexpression in PTC samples.
Although the vast majority of PTC patients has good prognosis, 1/3 of all cases persist or relapse [
30]. In our study, the absence of NKX2.5 was associated to the persistence of PTC, suggesting that NKX2.5 expression does not seem to contribute to clinic aggressiveness of the disease. Furthermore, treated patients who express NKX2.5 have lower rate of persistence/recurrence, indicating that its expression might be associated to a less aggressive tumor behavior. So, NKX2.5 could be useful as a molecular target to help predicting the outcome of PTC patients, clinically relevant information to decide the best therapeutic approach. Moreover, in PTC patients, the absence of NKX2.5 was associated to the pathological stage of primary tumor size pT1 and pT2, despite the fact that PCCL3 cell viability was not affected by
NKX2.5 transfection. Thus, in tumor cells, NKX2.5 might impact cell proliferation and/or survival. In fact, NKX2.5 has been suggested to have a role enhancing survival of leukemic T-cells [
31].
Our data suggest that NKX2.5 has a dual role in thyroid. During thyroid organogenesis NKX2.5 seems to be important during the beginning of the organogenesis, but latter, NKX2.5 have to disappear before thyroid differentiation markers can be expressed [
2]. In a similar way, during thyroid carcinogenesis, NKX2.5 might play a role during tumour initiation, inducing dedifferentiation, as shown in PCCL3 overexpressing NKX2.5. However, thyroid cancer progression might require the disappearance of NKX2.5, thus explaining the correlation between the levels of NKX2.5 expression and the better prognosis. The mechanisms underlying this shift in NKX2.5 expression during thyroid cancer progression could involve many factors, such as changes in NKX2.5 sumoylation, which is known to regulate NKX2.5 [
25], changes in the tumour microenvironment, changes in the expression of other transcription factor that could regulate NKX2.5, increased expression of miRNA targeting NKX2.5, among others. Literature data have shown low frequency of BRAF mutations in distant metastases, in comparison with the paired primary tumours [
32], thus suggesting that this mutation could play a role in thyroid cancer initiation but not progression. Thus, NKX2.5 could have a similar effect, contributing for thyroid cancer initiation but not for progression.
Acknowledgements
We are grateful for the technical assistance of Norma Lima de Araújo Faria, Advaldo Nunes Bezerra, Wagner Nunes Bezerra, Walter Nunes Bezerra and José Humberto Tavares de Abreu.