A possible role for selenoprotein glutathione peroxidase (GPx1) and thioredoxin reductases (TrxR1) in thyroid cancer: our experience in thyroid surgery

The study population was recruited in 2014 at the Department of Surgical Sciences of the Umberto I Hospital of Rome. Patients were between 24 and 68 years of age, with thyroid nodules classified as indeterminate (TIR3B), suspicious for malignant (TIR4) or malignant (TIR5) after cytologic examination, according to the “Italian Consensus for the classification and reporting of thyroid cytology, 2014” [18]. The Fine Needle Aspiration (FNA) to obtain samples for the cytological evaluation was performed only on thyroid nodules ≥ 1 cm with suspicious ultrasound (US) features, according to “Sonographic and Clinical Features of Thyroid Nodules and Recommendations for FNA” of the American Thyroid Association (ATA) [19]. Considering the cytological results (Table 1), all patients were candidates for thyroid surgery, according to the ATA 2009, the only one known at the time of the recruitment.

Surgical treatment was planned taking into account the cytological diagnosis, lymph node involvement and the size of the suspicious nodule (both evaluated through US of the neck), age of patients, and a history of prior head and neck irradiation. In cases with a cytologic diagnosis of malignancy (TIR4, TIR5), the patients underwent total thyroidectomy, according to ATA 2009 (recommendation 26). Thyroid lobectomy alone may be sufficient treatment for small (< 1 cm), low-risk, unifocal, intrathyroidal papillary carcinomas in the absence of prior head and neck irradiation or radiologically or clinically involved cervical nodal metastases, as suggested by ATA 2009. However, none of our TIR4/TIR5 patients were candidates for lobectomy. In fact, as previously described, only patients with suspicious nodules ≥ 1 cm underwent FNA and were included in the study. Table 1 reports the histological size of the tumor, measured “ex vivo” after thyroidectomy, which is often different from the pre-operative US measurements. Patients with clinically involved central compartment (level VI) lymph nodes underwent therapeutic central neck dissection along with total thyroidectomy (Table 1, cases 2, 4, 5, 8, 9, 10, 11, 12, 13, 14, 15, 17, 18, 19, 20). Patients with suspicion for lateral lymph node disease (compartments II–IV) on preoperative US or intraoperative exam also underwent lateral neck dissection (Table 1, cases: 15, 17, 19, 20; ATA 2009, recommendation 27). Patients 3, 6, 7 and 16, with indeterminate nodules (TIR3B), preferred to undergo total thyroidectomy, to avoid the possibility of requiring a future surgery on the contralateral lobe as provided by ATA 2009 (recommendation 25b). Informed consent was obtained from all participants. The criteria of exclusion from the study were the presence of debilitating diseases (i.e. advanced stage diabetes, immunological diseases or hematologic disorders), lack of informed consent, and/or lack of authorization form for the processing of personal data. All subjects recruited for the study underwent the following: electrocardiogram, chest X-ray, indirect laryngoscopy, and whole blood sampling. The following blood tests were obtained for each patient: hemochromocytometric analysis, routine chemistry tests, TSH, FT3, FT4, calcitonin, thyroglobulin, anti-thyroperoxidase antibody (TPOAb), and anti-thyroglobulin antibody (TgAb). In order to obtain a comparison between healthy thyroid tissue and cancer tissue in each individual patient, we selected only those patients (candidates for total thyroidectomy) with unilateral suspicious cytological nodules.

Antibodies were obtained from the following sources: polyclonal anti-glutathione peroxidase (GPx1) and polyclonal anti-thioredoxin reductase 1 (TrxR1) and RIPA buffer from Sigma–Aldrich (St Louis, MO, USA), monoclonal anti-actin from Santa Cruz Biotechnology (Santa Cruz, CA, USA), peroxidase-conjugated goat anti-mouse and peroxidase-conjugated rabbit anti-goat from Amersham (Arlington Heights, IL, USA). Nitrocellulose paper was obtained from Shleicher & Schuell Bioscience Inc. (Dassel, Germany). Enhanced chemiluminescent kit (ECL kit) and BCA Protein Assay Kit were purchased from Pierce (Rockford, IL, USA). The densitometric analyses of selenium protein were performed using the GS-900 densitometer (Bio-Rad) and the “Image J” Software. Spin trap 1-hydroxy-3-carboxy-pyrrolidine (CPH), 3-carboxy-proxyl radical (CP·) was purchased from ENZO Biochem (Laufelfingen, Switzerland).

A complete histological exam of the thyroid gland was performed after thyroidectomy to characterize the lesions and to obtain the definitive histological classification (Table 1). All patients in which the definitive histological exam revealed, in the uninvolved lobe, an incidental cancer lesion or signs of inflammation or other types of disease were excluded from this study. A sample of tissue containing the thyroid cancer and a sample of the contralateral lobe were taken from the gland at the time of surgery and conserved at − 80 °C for future analysis and measurement of selenoproteins, GPx1 and TrxR.

Samples prepared for Western Blot analysis were solubilized in 4× loading buffer, boiled for 5 min, and resolved by Sodium Dodecyl Sulphate–PolyAcrylamide Gel Electrophoresis (SDS–PAGE). Proteins were transferred to nitrocellulose paper at 35 V overnight. Blots were washed with TBS containing 0.05% Tween 20 (TTBS) and blocked with 3% bovine serum albumin in TTBS for 2 h. Washed nitrocellulose filters were incubated overnight at 4 °C with the anti-GPX1 or anti-TrxR1 antibody. After extensive washes in TTBS, the immunoreactive bands were detected by chemiluminescence coupled to peroxidase activity, according to the manufacturer’s specifications (ECL kit; Pierce).

In order to measure GPx1 and TrxR1 expression levels, the intensities of specific bands, corresponding to the selenoproteins of interest, were measured. The densitometric analysis of the blots was done using the Image J Software [20]. In brief, the blot images were imported into the software and the contrast was adjusted such that the bands were clearly visible on the blot image. The area around each band was selected and the background intensity was subtracted from the blot image. The intensities of the selected bands were then expressed as a numeric value (arbitrary unit, a.u.) that could be used for carrying out further statistical analyses.

Free radicals are highly reactive species with half-lives usually less than 1 ms and are hardly detectable by direct EPR technique alone. This limitation is overcome by combining the EPR technique with spin probing. In brief, an organic compound (in our case 1-hydroxy-3-carboxy-pyrrolidine (CPH), capable of reacting with the radical species present, is added into the study system. The reaction product is a nitroxide radical (CP·), which is much more stable and therefore easily detectable by EPR technique [21]. Spectra were measured on a Bruker ECS 106 spectrometer equipped with a variable-temperature unit (ER4111VT). Samples were drawn up into a gas-permeable Teflon tube with 0.81-mm internal diameter and 0.05-mm wall thickness (Zeuss Industrial Products, Raritan, NJ, USA). The Teflon tube was folded four times, inserted into an open (air) quartz tube, and fixed to the EPR cavity (4108 TMH). The temperature was 37 °C. All ESR spectra were corrected for baseline drift by a linear function and double integrated and simulated using the software supplied by Bruker (ESP 1600 data system). Spectrometer conditions common to all spectra were modulation frequency, 100 kHz; microwave frequency, 9.4 GHz; microwave power, 20 mW.

All data were expressed as the mean (± standard deviation) of at least three measurements and analyzed with the Student’s t test or Two-Way ANOVA with Bonferroni post-test using the statistical software Graph Pad Prism 4.0 (GraphPad Software, Inc., CA, USA).

All patients underwent total thyroidectomy. Therapeutic lymph node dissection was performed when cervical lymph node metastases were suspected or shown preoperatively or intraoperatively by the clinical or radiological examination. For all patients, the recovery was uneventful, and discharge was on the third postoperative day. No significant decrease was detected in blood calcium levels after surgery.

Histology of the thyroid tissue demonstrated papillary thyroid cancer in all patients: 16 patients with classical variant (4 with multifocal cancer), and 4 patients with follicular variant. Cervical neck dissection was performed on 15 patients with clinically involved lymph nodes, and lymph node metastases were found in 10 patients. Table 1 summarizes the preoperative cytological diagnosis, the post-operative histological classification, the surgical approach, and the staging.

The overall protein composition of each patient’s thyroid tissue was determined by SDS-PAGE, and the protein content was verified by a specific protein staining (data not shown). The expression of GPx1 and TrxR1 was detected by the appropriate antibodies, able to identify only these two proteins in the thyroid protein lysate. Figures 1 and 2 show the Western blotting results referred to GPx1 and TrxR1, respectively. For each patient, two samples were tested for the presence of GPx1 and TrxR1: one sample from the healthy thyroid tissue and one sample from the cancer tissue. As demonstrated in Fig. 1, the intensity of the band detected by the anti-GPx1 antibody in thyroid cancer tissue was lower for all patients than those detected in healthy tissue. Similar results were obtained using the anti-TrxR1 antibody. In fact, all patients showed a decrease of the TrxR1 expression in cancer tissue compared to the healthy tissue. These findings support the hypothesis that an impairment in the ability of the antioxidant system to counteract oxidative stress may be implicated in thyroid cancer.

Western blot analyses showed a decrease in GPx1 and TrxR1 expression in thyroid cancer, without a real quantification of the difference. In order to better quantify our results, we performed a densitometric analyses of the bands representing GPx1 and TrxR1 in the western blot. To be sure that these results were not due to a different protein content between the two types of tissue, we decided to normalize our results with respect to actin, a ubiquitous protein present in all eukaryotic cells. Taking all patients together, the densitometric analysis of the GPx1 expression showed a significant lower mean concentration in cancer tissue, compared to healthy tissue. GPx1 expression in cancer tissue (red line, Fig. 1a) was 44.4 ± 2.8% less than the average values detected in healthy tissue (black line, Fig. 1a) (p < 0.001, range + 27 to − 74%). Furthermore, the difference of the GPx1 expression between healthy and cancer tissue, measured for each patient, was statistically significant for 16 of the 20 patients studied (Fig. 1b). The average TrxR1 expression was also decreased in tumor tissue when compared to healthy tissue, a difference of 49 ± 1.2% (p < 0.001, range − 3 to − 63%). Moreover, the mean values of the peaks obtained by the densitometric analysis of TrxR1 expression in thyroid cancer tissue (red line, Fig. 2a) were significantly lower than those measured in healthy tissue (black line, Fig. 2a), 98,965 ± 103,326 vs. 1,963,000 ± 91,242 a.u., p < 0.0001, respectively. Comparing individual patients’ tumor tissue to their own healthy tissue, the difference of TrxR1 expression was statistically significant for everyone (Fig. 2b).

As previously described [20] EPR-spin technique has proved very useful in being able to detect free radicals within some human tissues. Using this technique, we detected a significant increase in the production of free radicals in all of the thyroid tumor tissue samples, compared to the healthy tissue of the same patient. In particular, the mean value detected in the cancer tissues was 12.07 ± 0.97 a.u., while that detected in healthy tissues was 5.58 ± 0.89 a.u. p < 0.01). These data indicate, unequivocally, that the process of carcinogenesis is closely related to the physiological mechanisms responsible for the balance between oxidants and antioxidants species (Fig. 3).