Methylation Degrees of NIS and TIMP-3 Promoters in Well-Differentiated Thyroid Tumors

The aim of this study was to investigate the DNA methylation pattern in CpG islands of NIS and TIMP-3 promoters in thyroid cancers and the matched non-tumoral tissues as well as multinodular goiter (MNG) tissues. This cross-sectional study was performed on total 64 patients including 28 papillary thyroid cancer (PTC), 9 follicular thyroid cancer (FTC) and 27 MNG cases. Method: The bisulte sequencing PCR technique was used to evaluate the promoter methylation pattern of the NIS and TIMP-3 genes.


Conclusion
Aberrant promoter DNA methylation suggests potential utility to differentiate benign and malignant PTC and FTC tissues for early diagnosis, and contribution for personalized clinical management and surveillance. According to the reversibility of DNA methylation, these events may be potential targets for demethylating treatments.

Background
Thyroid cancers (TCs) are the most common malignancies of endocrine system and comprise several subtypes with remarkably different biological characteristics. Among these subtypes, papillary and follicular thyroid cancers (PTC and FTC) are referred to well-differentiated thyroid cancers (WDTCs) accounting for more than 90% of all thyroid malignancies (1)(2)(3). WDTCs typically are well managed and represent favorable prognosis, however, sometimes, patients progress toward poorly-differentiated and more aggressive subtypes (4).
Nowadays, the molecular pathogenesis of TCs is not completely clear. However, environmental factors alongside the genetic alterations have been shown to play a putative role in the development of thyroid cancers (5,6). The effect of environmental factors exerts on the regulation of gene expression which are known as epigenetic alterations (5). These alterations are speci c to each histological subtype of TCs and de ne their gene expression patterns and phenotypic characteristics. Therefore, understanding the details of epigenetic alterations in TCs can provide effective therapeutic strategies.
Aberrant DNA methylation has been introduced as a major epigenetic event in tumorigenesis (7,8). Gene inactivation through hypermethylation of CpG islands located in promoter regions were described in many types of cancers (9,10). Promoter methylation of tumor suppressor genes which are thyroid speci c have been widely studied in WDTCs, however, to date, the majority of these reports are relied on the non-quantitative analysis of methylation speci c PCR (MSP) methodology (9,(11)(12)(13)(14).
Solute carrier family 5 member 5 (NIS also called SLC5A5) gene is located on chromosome 19 (19p13.11) and encodes an 80-90 kDa transmembrane glycoprotein that actively transports iodide from the bloodstream into the thyroid follicular cells (15). Thyroid-stimulating hormone (TSH) regulates NIS protein activity which consequently participates in the regulation of iodine uptake by thyroid follicular cells. In this process, a Na+/K+-ATPase pump generates a sodium gradient, transporting two Na + to an I − as the rst step in the thyroid hormones biosynthesis (16).
Tissue inhibitor of metalloproteinase-3 (TIMP-3), as a member of tissue inhibitors of matrix metalloproteinases (MMPs), is involved in different cellular processes like cell growth, angiogenesis, invasion, and metastasis in human cancers. TIMP-3 plays an important role in the regulation of vascular endothelial growth factor (VEGF)-mediated angiogenesis by direct binding to VEGF receptor-2 (VEGFR-2).
This process prevents downstream essential signaling pathways for cell stimulation. Therefore, according to the potential angiogenesis inhibitory role of TIMP-3, it has drawn particular attention for therapeutic purposes. It has been demonstrated that hypermethylation of TIMP-3 gene was associated with clinicopathological characteristics of PTC, including extrathyroid invasion, lymph node metastasis, and advanced disease stages (III and IV) (17). TIMP-3 blocks the binding of VEGF to the receptor, and inhibit downstream signaling pathways and angiogenesis (18). Beside this, the promoter hypermethylation of TIMP-3 gene and subsequently its down-regulation, have been observed in thyroid cancers and they are in association with aggressive pathological features (11,19).
According to the pivotal role of NIS in the process of WDTCs pathogenesis and also in the effectiveness of the current therapeutic strategies by radioactive iodine, the objective of this paper is an investigation on NIS promoter methylation and its expression in PTC and FTC patients. Furthermore, TIMP-3 is the other studied gene in the present study owing to its targeting can be one of the essential therapeutic strategies. In the present study, DNA methylation in the promoter of these two genes including has been assessed in PTC and FTC thyroid tissues compared to matched non-tumoral as well as other benign tissues, in Iranian population.

Demographic and clinicopathological characteristics
Page 4/20 The mean ± SD of age in PTC, FTC and MNG groups were 37.89 ± 12.17, 55 ± 18.15 and 48.67 ± 14.75 years, respectively. The mean ± SD of tumor size were 1.77 ± 1 and 3.25 ± 1.81 cm in PTC and FTC groups, respectively (Table 3).   Fig. 3.b). We could not found a signi cant difference for total methylation level of TIMP-3 promoter between PTC tumoral and the matched adjacent non-tumoral tissues (22.17 ± 1.23 vs.

Analysis of TIMP-3 mRNA expression and its promoter methylation in PTC patients
18.29 ± 1.74 respectively, P = 0.07). However, analyzing ot 15 CpGs sites of TIMP-3 promoter showed that the methylation degree in the 5th (-3620) and 6th (-3625) CpGs of the forward strand of TIMP-3 promoter was signi cantly increased in PTC samples compared to the matched adjacent non-tumoral tissues. The methylation level in tumoral and matched adjacent non-tumoral were found 38 ± 5.71 vs. 18 Association of NIS and TIMP-3 mRNA levels with clinicopathological characteristics The mRNA level of NIS was signi cantly decreased in PTC patients with lymph-node metastasis compared to those without metastasis (P = 0.026, Table 4). TIMP-3 mRNA level was signi cantly reduced in PTC patients with age ≥ 45 years compared to those with < 45 years (P = 0.036). TIMP-3 mRNA level was also signi cantly reduced in PTC patients who had positive lymph-node metastasis and exrathyroidal invasion (P = 0.035 and P = 0.044, respectively, Table 4).

Discussion
It is well accepted that increased methylation may induce dedifferentiation and decrease expression of thyroid tissue-speci c genes. NIS promoter is CpG rich and its methylation in these regions may cause to decrease the expression of NIS (20). The results of the present study did not show signi cant differences for methylation level of NIS promoter in PTC tissues compared to the corresponding none-tumoral tissues in none of the CpG sites, but the NIS mRNA level in tumoral tissues signi cantly decreased compared to the matched adjacent none-tumoral tissues. In addition, NIS promoter methylation level in PTCs had not signi cant difference compared to its methylation level in MNGs. NIS mRNA expression was not signi cantly different between these two groups. In previous studies, reduced expression of NIS mRNA has been shown in PTC tissues (21)(22)(23), and have been suggested as a consequent of increased promoter methylation (22). Subsequently, some studies were carried out to determine the methylation status of NIS promoter. However, nowadays, the association between the PTC development and hypermethylation of NIS promoter is controversial (24). In this regards, in an old study, Caillous et al. (25) demonstrated that absent of iodide-concentrating ability in patients with PTC was associated with the NIS methylation. XiaoGuang et al. demonstrated that NIS promoter was hypermethylated in 27% of PTC (16 of 60) samples compared to the control specimens (26). The methylation degree of NIS promoter in PTC tissues and in their matched adjacent non-tumoral tissues was 30.9% (47/152) and 6.58% (10/152) respectively (27). Galraro et al (15), showed that methylation of the NIS promoter is a common occurrence in malignant thyroid tumors (18 PTCs and 2 FTCs) and their surrounding tissues, but they did not nd a quantitative correlation between methylation levels and mRNA expression in any groups and described a heterogeneous pattern of methylation along the CpG islands of this gene. In a study carried out using real-time quantitative methylation-speci c PCR (qMSP), Stephen et al. (28) investigated 21 genes promoter methylation levels in 85 PTC, 90 FTC, 83 follicular adenoma and 65 normal groups. They did not nd a signi cant differential methylation of NIS promoter in PTCs compared to the follicular adenoma and normal tissues taken from healthy thyroid patients. Smith et al. evaluated the methylation degrees of the NIS promoter in 32 PTCs and they found that it is not a statistically common event in the PTC tumors (9). In a meta-analysis which carried out by Zhao et al. (24), seven articles with totally 360 cases and 268 controls, were selected to investigate the relationship between NIS promoter methylation levels and PTC development. Finally they concluded that the promoter methylation in NIS gene is related to the PTC development and its aggressive and metastatic behavior. According to our results, promoter methylation of the NIS gene is not common event in PTC tumors, their matched non-tumor and benign tissues and it could not be a possible mechanism leading to reducing of NIS expression. Other regulatory mechanisms may be affected on the expression of this gene during neoplastic transformations.
According to our results, the methylation levels of the NIS promoter were signi cantly higher in FTC tissues compared to the matched adjacent non-tumoral. These ndings were consistent with the hypermethylation of 3th, 6th and 7th CpG positions of forward strand of the NIS promoter. In a contrary manner, mRNA expression was reduced signi cantly in tumoral tissues. As well as, we found highly signi cant results with over 20% differences in methylation levels between tumoral and benign (MNG) tissues. To the best of our knowledge, no study has investigated methylation status of NIS gene in FTC tumoral tissues compared to matched-tumoral and benign tissues, however; Stephen et al (28) found a signi cant differences in methylation levels of NIS promoter between FTC and normal thyroid groups. These ndings imply that the NIS promoter hyper-methylation may result in reduction of its expression in FTC and can be a potential therapeutic target to restoration of iodine condensation.
In the present study, we also investigated the methylation status of TIMP-3 as a critical tumor suppressor gene in well differentiated thyroid cancers. The results demonstrated a signi cant increase in methylation degree at 5th (-3620) and 6th (-3625) CpG positions in the forward strand of TIMP-3 promoter in PTC tissues compared to their corresponding non-tumoral and MNG tissues. Subsequently, we found 50% decrease in TIMP-3 mRNA levels in PTC tumoral tissues compared to their corresponding non-tumoral tissues. In a study carried out on 231 PTCs by Hu et al. (11), increased methylation of TIMP-3 promoter gene was observed in 53% of patients and its association with high risk clinicopathlogical chracteristic of patients were also reported. In another study, Houqe et al (29) investigated the promoter methylation pattern of 11 genes in 23 PTC patiennts samples by qMSP technique. They reported a signi cant increase in methylation of TIMP-3 promoter in the PTC group compared to the noncancer tissues. Our results are similar to their ndings. In the study performed by Stephen (28), as well as a signi cant hyper-methylation NIS promoter, they also reported a signi cant increase in TIPM-3 promoter methylation in PTC tumors compared to the adenoma and normal tissues. Brait et al. (30) demonstrated 51% methylation of TIMP-3 promoter in PTC tissues, but they demonstrated that PTC tumor samples had overlapping frequencies of methylation for TIMP-3 with benign and normal thyroid samples (51% tumors vs. 42% benign vs. 27% normal). They concluded that this nding could limit the use of TIMP-3 as diagnostic biomarker for PTC and answer to whether hypermethylation present in the tumors are relevant to the neoplasia. However, the synchronization of hyper-methylation of the gene promoter and decrease of its mRNA levels in the present study and previous researches strengths this hypothesis that there is a relationship between TIMP-3 promoter methylation and PTC development.
Finally we found highly signi cant inversion proportion between TIMP-3 promoter methylation degrees and its mRNA expression in FTC tissues in comparison with non-tumoral and MNG tissues, more methylation in promoter lower expression in mRNA level. Our ndings in this eld are in accordance with Shephen et al (28) results. They reported a highly signi cant hyper-methylation in TIMP-3 promoter in FTC tissues compared to the adenoma and normal tissues. However, according to the limited number of studies and sample size in the mentioned area, it seems that more clinical researches with larger sample size and higher quality case-control studies are required to be taken in the future. To the best of our knowledge, this is the rst study determining the promoter methylation status of these two functional genes in the selected region in tumorigenesis of well-differentiated thyroid cancers. The lack of signi cance may be due to the low sample numbers. In this study, we also enlighten methylation status of CpG sites to reveal speci c citations for targeted therapy purposes with demethylated agents in future. However, time and budget constraints limited us to examine more CpG sites of these genes. Studying methylation levels of other CpG sites of these and other genes involved in pathogenesis of WDTCs and the mechanisms of epigenetics regulation will elucidate the role of epigenetic mechanisms in PTC and FTC development that lead to the improvement in designing effective therapies. This mater requires further investigations.

Tissue Samples
Patients who underwent surgical resection of thyroid tumor at Erfan and Atiyeh hospitals in Tehran, Iran, were selected from 2015-2016. Written informed consent was obtained from all of the patients. An experienced pathologist reviewed all tissue samples and con rmed the PTC, FTC and MNG subjects based on pathological evidence and clinical outcomes. The pathologist also separated the non-tumoral adjacent slices from tumoral samples. Overall, 64 patients met the eligibility criteria, consisting of 28 PTCs, 9 FTCs cases (tumoral tissues and the matched non-tumoral thyroid tissues from the same patients) and 27 benign cases (non-tumoral tissues from patients with multinodular goiter). Tissue samples were collected in RNase and DNase free tubes and immediately frozen in liquid nitrogen (-198 °C), and then stored at -80 °C for the later uses. Demographic and clinicopathological data of the patients were summarized in Table 1. This study was approved by the Ethics Committee of the Research Institute of Endocrine Sciences of the Shahid Beheshti University of Medical Sciences (IR.SBMU.ENDOCRINE.REC.1395.216). Genomic DNA extraction Genomic DNA was isolated according to the manufacturer's instructions from the fresh frozen thyroid tissue specimens using the FavorPrep Tissue Genomic DNA Extraction Mini Kit, USA. Then, DNA quality and quantity were determined using the NanoDrop 1000c (Thermo Scienti c, USA) with considering the 260/280 absorbance ratio. All the extracted DNAs were stored at − 80 °C.

Design of primers
Met-primer online Software was used to design primers for hot-start PCR. These primers were designed for the promoter regions of NIS and TIMP-3 genes. In summary, 1000 nucleotides upstream of the start codon were selected from genome databases. Islands in the promoters were detected according to the speci c inclusion criteria including island size > 100, GC percent > 50.0, and obs/exp > 0.6, CpG. (primers were detailed in Table 1). Finally, we found 7 and 15 CpG sites for NIS and TIPM-3 promoters, respectively and investigated methylation status in these sites. In order to evaluate the overall expression of NIS and TIMP-3, primers for qRT-PCR were designed using GeneRunner software (version 4.0) and checked in NCBI Primer Blast (Table 2).

Consent for Publication
Not applicable.

Availability of Data and Material
The data that support the ndings of this study are available from the corresponding authors, MH and ZN, upon reasonable request.

Competing of Interest
The authors declare no con ict of interest, nancial or otherwise.    Data are presented as Mean±SEM.