Methylation Changes of Primary Tumors, Monolayer, and Spheroid Tissue Culture Environments in Malignant Melanoma and Breast Carcinoma

Epigenetic changes have major role in the normal development and programming of gene expression. Aberrant methylation results in carcinogenesis. The primary objective of our study is to determine whether primary tumor tissue and cultured tumor cells in 2D and 3D tissue culture systems have the same methylation signature for PAX5, TMPRSS2, and SBDS. These findings will play an important role in developing in vitro model system to understand the effect of methylation inhibitors on primary tumor tissue. In a previous study PAX5, TMPRSS2, and SBDS genes that we are investigating were reported to be methylated more than 60% in breast cancer and malignant melanoma cell lines. However, these genes have never been studied in primary tumor tissues. Thus, primary tumor tissues of breast cancer and malignant melanoma were first grown in 2D and 3D cultures. Then these two types of tumor tissues and their 2D and 3D cultures were investigated for changes considering methylation levels in PAX5, TMPRSS2, and SBDS genes using real-time polymerase chain reaction. No differences were observed in the primary tissues and culture systems for both PAX5 and TMPRSS2 in malignant melanoma tissues. We found that PAX5 gene was an efficient marker to measure the effects of methylation inhibitors for in vitro systems for malignant melanoma tissue.


Introduction
Malignant melanoma is a carcinoma of the skin that grows as a result of malignant transformation of melanocytes in the basal layer of the epidermis and it is the most aggressive of all skin cancers [1]. Cancerous cells develop due to the deterioration in skin cells damaged through UV radiation and sunshine or due to damage in DNA. Five percent of all skin carcinomas have been reported as malignant melanoma but it is the type that is most responsible for skin cancer deaths [2]. Recent studies showed that DNA methylation was associated with proliferative and immunologic processes in subgroups of melanoma [3][4][5]. Recently tumor suppressor genes (TSGs) have been shown to play a role in the pathogenesis of various cancer types. More than 100 genes contributing to the pathogenesis of malignant melanoma have been found to be abnormally hypermethylated. Abnormal methylation of TSGs is thought to be associated with aggressive clinicopathological features and poorer survival. Epigenetic changes may be a significant prognostic marker that benefits routine practice [6]. In general, the DNA methylation profile of cancer cells is characterized by global genomic hypomethylation accompanied by specific hypermethylation on tumor suppressor genes, resulting in metastasis and poor clinical outcomes [7].
Breast cancer develops as a result of the uncontrolled proliferation of cancer cells in breast tissue. The incidence of this disease is increasing worldwide. Germline mutations of breast cancer susceptibility genes are responsible for hereditary breast cancers [8]. Also, mutations and epigenetic changes affect the development of breast cancer [9]. In the analysis of whole genome gene expression microarray and DNA methylation microarray, genes with abnormal DNA methylation were investigated for early detection in breast cancer. Recent studies have shown that changes in DNA methylation at an early stage in the development of breast cancer (BC) may be clinically appropriate for therapeutic decisions [10].
To date, at least 50 different genes have been demonstrated to have the rearranging function of DNA methylation and histone modifications in malignant melanoma [11]. In the study by Echrich et al. (2008), a high rate of methylation was shown in PAX5, TMPRSS2, and SBDS genes of various cell lines. The authors found non-similar methylation patterns of breast cancer samples although strong cluster formation was observed for colon cancer, CNS, and melanoma when the relationship was analyzed between the cell lines [12]. PAX5 gene is basically involved in the development of B cells. In the absence of this gene, these cells can turn into T cells or natural killer cells [13]. PAX5 was found as a potential immunohistochemical marker in differential diagnosis of lymphoid neoplasms [14]. TMPRSS2 is included in the family of transmembrane serine protease type 2 genes and was found in the basal cells of prostate epithelium [15]. Although the exact function of gene is unknown, it was determined that it modulates sodium current in the epithelial sodium channel of lungs [16]. Although the function of SBDS is not fully understood, some studies have shown that it is involved in the maturation and processing of RNA and ribosomes [17,18].
Methylation inhibitors are not used in solid tumors because of their unknown adverse effects and its usage is only to hematologic malignancies [19]. Methylation inhibitors such as 5-azacytidin and 5-azadeoxycytidin reversibly inhibit DNA methyltransferases and thus, hypomethylation of DNA has occurred [4]. This activity may activate tumor suppressor genes that have the antitumor effect silenced by hypermethylation and oncogenes have oncogenic effect by hypermethylation. As a consequence, in vitro model systems are required in patients especially with solid tumors for measuring the adverse effects of methylation inhibitors. Hence, it is very important to measure antitumor effects of methylation inhibitors before using as chemotherapeutic agents for patients with solid tumors.
In the study by Harman et al. (2016), the effects of the DNA methyltransferase (DNMT) inhibitors such as 5azacytidine (5-AzaC) were examined on normal and tumoral mammary cell lines derived from dogs, cats, and humans whether they were useful models for human breast cancer. The study showed that the effects of 5-AzaC varied on gene expression between the different species and different tumor cell lines of the same species. The results of study were confirmed in primary malignant cells isolated from dog and cat adenocarcinomas. These findings suggest that three species, dogs, cats, and human, may be useful models for the preclinical evaluation of new therapeutics in human breast cancer studies [20].
Nowadays, it is known that 3D tumor tissue culture systems have 85% similarity with primary tumor tissue. Therefore, 3D tumor tissue culture systems can be used as in vitro model systems for these purposes. But, specific genes are needed to know as a biomarker for each specific tissue. These genes have to show the same features of methylation in tumor cell grown in 2D and/or 3D culture systems compared with primary tumor tissue.
The main goal of the study is to discover an in vitro model system and associated tools for measuring the effects of methylation inhibitors. For this reason, the primary objectives were to grow tumor cells in 2D and 3D tissue culture systems and to determine whether the cells grown in the culture systems and primary tumor cells had the same features of methylation in the genes PAX5, TMPRSS2, and SBDS.
In this study, the genes were selected from the cell line of associated tissues that have previously reported in literature (Ehrich, Turner et al. 2008). These genes were found to be hypermethylated in breast cancer and malignant melanoma cell lines. However, the methylation status of selected PAX5, TMPRSS2, and SBDS genes was examined in cell lines but not primary tumor tissue of patients with breast cancer and malignant melanoma in the literature. For this reason, these genes were selected to determine the methylation levels in primary tumor tissue and to find out the difference between the primary tumor tissue and the 2-and 3-dimensional cell cultures.
One of the most effective way to display tumor tissue changes is 3D culture systems. If the changes in 3D cultures are the same with primary tumor tissues, in vitro 2D and/or 3D tissue culture systems could be used as a model to detect methylation changes. In this way, the expression levels of genes converted to unmethylated levels with methylation inhibitors could be followed in vitro. Moreover, the interactions between cells and the interested gene could be understood with in vitro 3D tissue cultures. In vitro 3D culture systems could be used to estimate the benefits of methylation inhibitors in the treatment of tumors and also to predict the adverse effects of methylation inhibitors in patients.

Patient Groups and Preparation of Specimens.
Primary tumor tissues were obtained from 10 patients with breast cancer and from 6 patients with malignant melanoma with a presumed diagnosis of cancer which were admitted to the Surgical Unit of Clinical Oncology Department in Oncology Institute at Istanbul University. The study was approved by the Ethics Committee of Istanbul Medical Faculty at Istanbul University (2009/2524-7).
Two pieces of 0.5 cm 3 Tru-cut tissue biopsies were extracted from primary tumors during surgery for the analysis. Under sterile conditions, one of the tissue biopsies was placed in 10 mL RPMI 1640 solution including 10% phosphate-buffered saline (PBS) for tissue culture, and the other tissue biopsy was treated with RNAlater solution (Qiagen Inc.) for DNA/RNA studies. The tissue piece in the Table 1: Methylated primers and probes.

GENES SBDS (138bp) TMPRSS2 (189bp) PAX5 (230bp)
FORWARD RNAlater was stored for 1 night at 4 ∘ C and then transferred to a liquid nitrogen cryo-resistant tube and stored in a liquid nitrogen tank. The tissue piece in the RPMI 1640 (Biochrom) was shredded into pieces for primary tissue culture.

Preparation of Primary Cell Cultures.
In a sterile laminar flow, the shredded tissue was planted in a 25 cm 2 cell culture flask by adding 5 mL of DMEM-F12 medium (Biochrom). Flasks were placed into an incubator containing 5% CO 2 at 37 ∘ C. Flasks were left to culture for 5 days without being moved. At the end of the 5th day, flasks were examined with an inverted microscope to observe the cellular proliferation. The media in the tissue culture flask were first replaced on the 5th day and refreshed every 2 days thereafter. When flasks became confluent, 3D tissue culture processes were started. All 2D and 3D cultures were planted into 3 flasks against contamination risk and coefficient variations between flasks.

Results
Primary tumor tissues from patients with breast cancer and malignant melanoma were cultivated in 2D and 3D culture systems ( Figure 1). Cells were collected from primary tumors and cultivated tumor tissue samples. DNA materials were isolated from tumor tissues and the 2D and 3D culture cells of 10 patients with breast cancer (n=10) whose mean age was 56.5±12 years and 6 patients with malignant melanoma (n=6), mean age 49±14 years, were used for methylation analysis. The methylation levels of PAX5, SBDS, and TMPRSS2 were examined according to ß-actin gene, which was the internal control gene.

Breast
Cancer. The methylation rates in tissue DNA of patients with breast cancer were found between 33% and 58% for PAX5 gene and between 49% and 55% for TMPRSS2 gene according to ß-actin gene (Table 4) (Figure 2). Methylation on SBDS gene was not detected in all groups of patients with breast cancer ( Table 5). The differences between cells grown in 2D-3D cultures and primary tumor tissues were statistically nonsignificant (P>0.05) for SBDS gene. In the patient group with breast cancer, PAX5 was found methylated in 7 out of 10 (70%) tumor tissues, only one sample (10%) was observed as methylated in 3D culture, and no methylation was observed in any 2D culture samples (Table 5). There was no statistically significant difference in methylation levels between 2D and 3D culture systems (P>0.05); however, the difference both between 2D culture system and tumor tissues (P= 0.003) and between 3D culture system and tumor tissue was found statistically significant (P=0.02). There was no significant difference (P=0.17) between both methylation levels of 2D culture and tumor tissues and between 3D culture and tumor tissues for TMPRSS2 gene. Two samples (20%) were found methylated in 2D and 3D cultures and 6 samples were also observed as methylated (60%) in primary tumor tissue for TMPRSS2 (P>0.05). The results were statically nonsignificant. TMPRSS2 could not be used as a marker for our in vitro model system because of differences of methylation levels in different types of tissue culture systems for breast cancer.

Malignant Melanoma.
Methylation rates were observed between 22 and 68% for PAX5 gene and between 37 and 51% of patients for TMPRSS2 gene in tissue of patients with malignant melanoma according to ß-actin gene (Figure 3) (  (Table 5). At the SBDS gene, methylation was detected in a single tumor tissue sample and was determined as 17% (Table 5). There was no methylation detected in the 2D and 3D culture systems for SBDS. When all 3 genes were considered together in the analyses, there was no statistically significant difference in the methylation levels of all culture methods (P>0.05). Table 4: Methylation rates (%) of three genes in patients with breast cancer (2D: two-dimensional tissue culture, 3D: three-dimensional tissue culture, T: primary tumor tissue, and BC: breast cancer).

Discussion
The main purpose of the study was to develop a biomarker to measure the in vitro effects of methylation inhibitors. For this purpose, our first aim was to cultivate the tumor cells in 2D and 3D tissue culture systems and to determine whether the primary tumor tissue and cultured cells in 2D and 3D tissue culture systems had the same methylation signature for PAX5, TMPRSS2, and SBDS. For this reason, appropriated primers and fluorescently labeled TaqMan probes were designed and methylation levels were investigated using real-time PCR.
Upon the presence of ß-actin control gene, Ct values were obtained with real-time PCR, and methylation differences in the genes in question were calculated. According to the results, it was shown that SBDS could be used as a negative control in in vitro model systems because the gene was completely unmethylated in primary breast tumor tissues and cultured breast tumor cells in 6 BioMed Research International Table 5: Methylation status of genes according to groups of patients and cultivation methods (2D: two-dimensional tissue culture, 3D: threedimensional tissue culture, and T: primary tumor tissue).

Genes Patients Cell culture method Total number of patients Absence of Methylation Presence of Methylation Number of patients (n) % Number of patients (n) % PAX5
Breast ca  2D  10  10  100  0  0  3D  10  9  90  1  10  Primary tumor tissue  10  3  30  7   Methylation rates were found between 33 and 58% for PAX5 gene and 49 and 55% for TMPRSS2 gene, respectively. Four of 10 breast cancer primary tumor tissues were found as methylated for both PAX5 and TMPRSS2 (bc22, bc27, bc21, bc15). Methylation was not seen in 2D and 3D tissue culture systems for samples of bc22, bc24, bc27, bc21, bc16, bc38, bc10. Methylation in PAX5 gene was not seen in any types of tissues of bc12 tumors. 2D: two-dimensional tissue culture, 3D: three-dimensional tissue culture, TD: primary tumor tissue, and BC: breast cancer.
2D/3D culture systems. In our study groups, we found that methylation rates of primary breast cancer tissues were 70% (7/10) for PAX5 and 60% (6/10) for TMPRSS2, respectively. It was shown that PAX5 and TMPRSS2 were not suitable genes to measure the effects of methylation inhibitors in breast cancer tissue. Briefly, in this part of the study, we found that there was a requirement for new methylated genes other than PAX5 and TMPRSS2 to measure the effects of methylation inhibitor in breast cancer tissue in in vitro models. In our study, PAX5 gene was found methylated in all patients with malignant melanoma for primary tumor tissue and 2D/3D tissue culture systems. Therefore, PAX5 gene was  and 51% for TMPRSS2 gene, respectively. Methylation in PAX5 gene was seen in all malignant melanoma primary tissue and also grew tumor cells in 2D and 3D tissue culture systems (mg08-mg015). Two of 6 patients with malignant melanoma in 2D and 3D tissue culture systems did not show methylation for TMPRSS2 gene (mg08 and mg013). 2D: two-dimensional tissue culture, 3D: three-dimensional tissue culture, TD: primary tumor tissue, and mg: malignant melanoma.

8
BioMed Research International concluded to be the most appropriate gene that could be used to measure the effects of methylation inhibitors in in vitro models for malignant melanoma. Methylation of TMPRSS2 gene was observed in some cases of malignant melanoma, but not in all. Thus, this gene was not an appropriate gene for use in in vitro systems. In addition, it was revealed that SBDS gene could not be used for the measurement of effects in methylation inhibitors in in vitro models because the methylation of SBDS was only detected in a single primary tumor tissue and was not observed in cells grown in 2D and 3D tissue culture systems. Methylation was observed in all investigated genes in other studies performed on cell lines; however, no methylation was observed in SBDS gene in our study. This situation suggests that cell lines may not reflect the real characteristics of primary tumor tissues for methylation. In our study, PAX5 gene was found methylated in primary malignant melanoma tumor tissues for the first time. PAX5 gene could also be used as a potential biomarker in our in vitro tissue culture systems to measure the effects of methylation inhibitors when malignant melanoma tissues are treated with methylation inhibitors such as 5-azacytidine and 5-azadeoxycytidine. The in vitro tissue culture systems presented in this article could be used for measuring the effects of methylation inhibitors and these systems could be used as in vitro diagnostic and prognostic tests to understand the effects of methylation inhibitors before giving them to patients with solid tumors. They could also be used to investigate the biologic effects of genes that change the methylation status with methylation inhibitors in different biologic pathways. The effects of methylation inhibitors such as 5-azacytidine and 5-azadeoxycytidine could be calculated by comparing the patient's clinical parameters and gene expression levels in different pathways, and the effectiveness of the drugs could be estimated for survival rates.
This was a preliminary study to understand whether solid tumor tissues were being grown in 2D and 3D culture environment. We also investigated whether PAX5, SBDS, and TMPRSS2 could be used as biomarkers by evaluating the stability of primary tumor cells grown in 2D or 3D culture media. For these purposes, 16 patients' materials for two tumor types and 3 different genes were chosen to test the feasibility of the method investigated in the study. We are planning to examine effects of 5-azacytidine (5-Aza C) in further research using genetic markers found here in this study.
In the future, we are planning to use PAX5 gene as a biomarker for in vitro 2D/3D culture systems to examine the role of PAX5 gene and/or other genes in the pathogenesis and biologic pathways of malignant melanoma in larger patient groups. We will also continue to research new biomarkers for breast cancer tissues to measure the effects of methylation inhibitors for in vitro systems.

Conclusion
Our results show that PAX5 gene is an efficient marker to measure the effects of methylation inhibitors such as 5azacytidine and 5-azadeoxycytidine at in vitro systems for malignant melanoma because the methylation levels are all the same in primary tissue and culture systems. Also, the results show that SBDS gene can be used as a negative control at in vitro model systems, since the gene was completely unmethylated in primary tumor tissue of breast and cultured breast tumor cells. The finding suggests that SBDS could be a suitable gene for the application of DNA demethylation agents.
Further research is needed for effective and high throughput analysis for detecting other genes and biomarkers for other types of cancer to measure the effects of methylation inhibitors for in vitro systems.

Data Availability
The data used to support the findings of this study are included within the article.

Conflicts of Interest
There were no conflicts of interest for any of the authors.