Betaglycan Gene (TGFBR3) Polymorphism Is Associated with Increased Risk of Endometrial Cancer

We investigated single nucleotide polymorphism (SNP) of the betaglycan gene (TGFBR3) encoding the TGFβ co-receptor in endometrial cancer (EC) and its association with betaglycan expression. The study group included 153 women diagnosed with EC and 248 cancer-free controls. SNP genotyping and gene expression were analyzed using TaqMan probes. Three out of the eight SNPs tested, i.e., rs12566180 (CT; OR = 2.22; 95% CI = 1.15–4.30; p = 0.0177), rs6680463 (GC; OR = 2.34; 95% CI = 1.20–4.53; p = 0.0120) and rs2296621 (TT; OR = 6.40; 95% CI = 1.18–34.84; p = 0.0317) were found to be significantly associated with increased risk of EC (adjusted to age, body mass index, menarche and parity). Among the analyzed SNPs, only rs2296621 demonstrated the impact on the increased cancer aggressiveness evaluated by the WHO grading system (G3 vs. G1/2, GT—OR = 4.04; 95% CI = 1.56–10.51; p = 0.0026; T—OR = 2.38; 95% CI = 1.16–4.85; p = 0.0151). Linkage disequilibrium (LD) analysis revealed high LD (r2 ≥ 0.8) in two haploblocks, constructed by rs2770186/rs12141128 and rs12566180/rs6680463, respectively. In the case of C/C haplotype (OR = 4.82; 95% CI = 1.54–15.07; p = 0.0116—Bonferroni corrected) and T/G haplotype (OR = 3.25; 95% CI = 1.29–8.15; p = 0.0328—Bonferroni corrected) in haploblock rs12566180/rs6680463, significantly higher frequency was observed in patients with EC as compared to the control group. The genotype-phenotype studies showed that SNPs of the TGFBR3 gene associated with an increased risk of EC, i.e., rs12566180 and rs2296621 may affect betaglycan expression at the transcriptomic level (rs12566180—CC vs. TT, p < 0.01; rs2296621—GG vs. TT, p < 0.001, GT vs. TT, p < 0.05). Functional consequences of evaluated TGFBR3 gene SNPs were supported by RegulomeDB search. In conclusion, polymorphism of the TGFBR3 gene may be associated with an increased EC occurrence, as well as may be the molecular mechanism responsible for observed betaglycan down-regulation in EC patients.


Introduction
Endometrial cancer (EC) is one of the leading female cancer-related causes of death with around 382,069 new cases and 89,929 deaths worldwide each year. Significantly higher incidence rate is observed in developed countries in contrast to less-developed ones with the world morbidity around polymorphism (SNP) investigations. In the current study, we examined eight SNPs within the TGFBR3 gene and their association with primary EC and their clinico-pathological variables, as well as potential impact on betaglycan expression.

Study Population
In the study, we enrolled only Caucasian women born and living in Poland. The case-control study involved 153 women who underwent surgery of EC and 248 healthy individuals which served as cancer-free controls. Biological material (cancer group-endometrial tissue samples and peripheral blood; cancer-free control-peripheral blood) was collected in the II nd Department of Gynecology, Lublin Medical University, Lublin, Poland and in the Department of Gynecological Oncology, Medical University of Lodz, Lodz, Poland, between 2012-2017. Inclusion criteria for case group, included women with diagnosed primary endometrial adenocarcinomas, who had not received neither hormonal therapy, radiation therapy nor chemotherapy prior surgery; whereas control group was recruited from non-related women during periodic health check-ups, who have never been diagnosed with endometrial cancer or other tumors.
Cancer tissue specimens, after the surgery, were divided into two portions; one was fixed in buffered formalin (pH 7.4) for routine histological assessment while the other was immediately placed at −70 • C. Clinical stage was assigned based on surgico-pathological findings according to the revised FIGO staging, while WHO classification was applied to determine the histological type and grade. Table 1 presents socio-demographic and clinical characteristics of the patients and examined samples. All studied cancer samples were classified as endometrioid cancers (type I)-endometrial adenocarcinomas.

Ethical Approval
The study was conducted in accordance with the ethical principles of the 1975 Helsinki Declaration and its later amendments. The local Independent Committees of Bioethics of Lublin Medical University, Medical University of Lodz and University of Lodz approved the tissues collections and study protocols. All methods in the study were performed in accordance with above-mentioned bioethical permissions. All participating subject gave written, informed consent prior to enrolment.

Lifestyle Risk Factors
Study participants were interviewed during the examination about socio-demographic, health related information and reproductive history (parity and menarche). Body mass index (BMI) was calculated as current weight in kilograms divided by square of height expressed in meters. Any missing survey data were subsequently completed using patient's query.

Genomic DNA Isolation
Genomic DNA was extracted from peripheral blood collected in the presence of anti-coagulant (EDTA) using PureLink Genomic DNA Mini Kit (Thermo Fisher Scientific, Waltham, MA, USA) and stored at −70 • C. The quality and quantity of DNA was estimated spectrophotometrically with BioPhotometer plus (Eppendorf, Hamburg, Germany). DNA samples were characterized with A260 nm/A280 nm ratio, which was in the range of 1.8-2.0.

SNP Selection and Genotyping
Eight SNPs in the TGFBR3 gene were selected according to NCBI SNPs database: rs883873, rs2770186, rs12141128, rs12566180, rs6680463, rs1805110, rs1805113, rs2296621. All SNPs were supposed to have minor allele frequency (MAF) ≥5% and localization assigned as 5 regulatory region, intron or exon of the TGFBR3 gene, which is located on chromosome 1. Characteristics of studied SNPs are presented in Table 2.  Real-Time PCR method with TaqMan Genotyping Assays (Thermo Fisher Scientific, Waltham, MA, USA) was applied for SNPs genotyping. The characteristics and sequences of used TaqMan probes are shown in Supplementary Table S1. PCR amplifications were conducted in a total volume of 10 µL and consisted of 5 µL (2×) of TaqMan Genotyping Master Mix buffer (Thermo Fisher Scientific, Waltham, MA, USA), 0.25 µL (40×) TaqMan Genotyping Assay (Thermo Fisher Scientific, Waltham, MA, USA) and 10 ng of template DNA. Thermal conditions were as follows: initial denaturation at 95 • C for 10 min, followed by 40 cycles of sequential incubations at 95 • C for 15 s and at 60 • C for 1 min and final endpoint measurement of fluorescence. Real-Time PCR amplifications and allelic discrimination were performed using Mastercycler ® ep realplex (Eppendorf, Hamburg, Germany).

Expression of the TGFBR3 Gene
Total RNA was extracted from frozen endometrial tissues using PureLink RNA Mini kit (Thermo Fisher Scientific, Waltham, MA, USA) according to dedicated protocol. The amount and quantity of isolated RNA was assessed spectrophotometrically with BioPhotometer plus (Eppendorf, Hamburg, Germany) based on A260 nm/A280 nm ratio, which was in the range 1.8-2.0. Total RNA (1 µg) was transcribed using RevertAid™ H Minus First Strand cDNA Synthesis kit (Thermo Fisher Scientific, Waltham, MA, USA) according to manufacturer's recommendation. cDNA synthesis was performed in Thermocycler 2720 (Applied Biosystems, Foster City, CA, USA) with the following incubations: 10 min at 25 • C, 120 min at 37 • C and 5 min at 85 • C. The obtained cDNA was stored at −70 • C. Real time PCR was performed using TaqMan probes (Thermo Fisher Scientific, Waltham, MA, USA) in line with the manufacturer's protocol on Mastercycler ® Epgradient S Realplex (Eppendorf, Hamburg, Germany) in the presence of TaqMan Gene Expression Master Mix Thermo Fisher Scientific, Waltham, MA, USA). GAPDH served as a reference gene. The catalogue numbers of probes were Hs00234259_m1 for TGFBR3 and Hs99999905_m1 for GAPDH. The relative expression level was normalized to GAPDH and was calculated using the following equation: 2 −∆Ct × 1000.

Statistical Analysis
The genotype frequency was tested for agreement with Hardy-Weinberg equilibrium (HWE) and assessed by chi-square goodness-of-fit test. Case-control differences in genotype and allelic distribution were analyzed using Pearson's χ 2 (chi-square) or Fisher's exact tests against the homozygote of the common allele as the reference group (OR = 1.00). Dominant and recessive genetic models were also implemented in the analysis. Variants of homozygotes and heterozygotes were combined to evaluate the dominant effect. SNPs distribution and their association with the clinico-pathological parameters were evaluated by multiple logistic regression. Genotype and allelic associations with endometrial cancer risk were expressed as odds ratio (ORs) and 95% confidence interval (95% CI) in crude and multivariate model including age, BMI, parity, and age at menarche.
Linkage disequilibrium (LD) and haplotypes distribution analysis were performed using the powerful online platform SHEsis (http://analysis.bio-x.cn/myAnalysis.php) [29]. Haplotypes with frequency less than 0.03 were excluded from the analysis. Bonferroni correction was applied for multiple comparisons of SNPs haplotypes.
To assess inter-group differences of socio-demographic parameters (age, BMI, parity and menarche), as well as the TGFBR3 gene expression levels between respective genotypes of the analyzed SNPs, the first Shapiro-Wilk test was applied to determine the normality of obtained data. Following, the statistical significance of difference was evaluated using either Student's t-test, for normally distributed data, or Mann-Whitney test, for non-normally distributed data. p < 0.05 in a two-tailed test was considered statistically significant. A statistical analysis of obtained data was conducted using GraphPad Prism version 5.00 for Windows (GraphPad Software, La Jolla, CA, USA) and PQStat version 1.6.8 (PQStat Software, Poland).

Bioinformatic Analysis
The functional consequences of significant SNPs were examined in RegulomeDB, which is a public database dedicated for noncoding SNP and annotates SNPs with known and putative regulatory elements in non-coding regions of human genome, such as regulatory DNA elements including regions of DNAase hypersensitivity, binding sites of transcription factors, and promoter regions that have been biochemically characterized to regulate transcription. RegulomeDB annotations are based on an integration of data from ENCODE project and other published literature, combined together by self-developed score system ranging from 1-6. A higher rank corresponds to a less functional significance [30].

SNPs Association with EC
Eight single nucleotide polymorphisms in the TGFBR3 gene and their association with EC risk and invasiveness were evaluated. Three of analyzed polymorphisms, i.e., rs883873 (g.92380302A > G), rs2770186 (g.92378843T > C) and rs12141128 (g.92373747A > G) are located in 5 regulatory region, whereas the other five polymorphisms, i.e., rs12566180 (c.−114 Based on a comparison of 153 women diagnosed with endometrial cancer and 248 healthy controls, we found significant differences in the distribution of the three studied SNPs (Table 3) adjusted to the following covariates, i.e., age, BMI, menarche and parity. Significant differences between endometrial cancer patients and control women were noted in body mass index (BMI; p < 0.001), menarche (p < 0.001) and parity (p < 0.001). In addition, the subjects' age showed significant differences between case and control group when analyzed in 10 years subgroups (p = 0.009) ( Table 1). Accordingly, age, BMI, menarche and parity were selected as main covariates in further analysis. All studied SNPs were in Hardy-Weinberg equilibrium (HWE) except polymorphic site rs1805110 (p.Ser15Phe), which was excluded from further analysis. The obtained results indicate that the polymorphisms of the highest importance for the increased endometrial cancer risk are rs12566180 (c.−114 were found to be more frequent as heterozygous variants in the study group as compared to the controls with respective frequencies 55.6% vs. 44.0% (p = 0.0177) and 58.2% vs. 45.2% (p = 0.0120), increasing the risk of endometrial cancer about 2.3 times. In turn, the polymorphic site rs2296621 (c.2285 − 99G > T) was found to be 6.4-fold more frequent in the case of study group compared to the controls (4.6% vs. 0.8%, p = 0.0317) as homozygous variant TT ( Table 3).
Linkage disequilibrium (LD) analysis revealed that among eight studied SNPs in the TGFBR3 gene, four of them were in high LD (r 2 ≥ 0.8) (Figure 1) and were arranged in two haploblocks constructed by rs2770186/rs12141128 and rs12566180/rs6680463. The frequency of haplotypes rs12566180/rs6680463 C/C and T/G were significantly higher in endometrial cancer patients as compared to healthy controls: for C/C haplotype 0.038 vs. 0.008 (OR = 4.82; 95% CI = 1.54-15.07; p = 0.0116-Bonferroni corrected) and for T/G haplotype 0.045 vs. 0.014 (OR = 3.25; 95% CI = 1.29-8.15; p = 0.0328-Bonferroni corrected), respectively (Table 4). gene, four of them were in high LD (r 2 ≥ 0.8) (Figure 1) and were arranged in two haploblocks constructed by rs2770186/rs12141128 and rs12566180/rs6680463. The frequency of haplotypes rs12566180/rs6680463 C/C and T/G were significantly higher in endometrial cancer patients as compared to healthy controls: for C/C haplotype 0.038 vs. 0.008 (OR = 4.82; 95% CI = 1.54-15.07; p = 0.0116-Bonferroni corrected) and for T/G haplotype 0.045 vs. 0.014 (OR = 3.25; 95% CI = 1. 29-8.15; p = 0.0328-Bonferroni corrected), respectively (Table 4).  Figure 2 shows the mRNA expression level of the TGFBR3 gene in 50 EC patients in relation to the genotypes of eight studied SNPs of theTGFBR3 gene. Six SNPs of the TGFBR3 gene were found to modulate its expression. The significant down-regulation of TGFBR3 mRNA was observed in the case of homozygous variant of rs2770186 for genotype CC (p < 0.05), rs12141128 for genotype GG (p < 0.05), rs1805110 for phenotype variant Phe/Phe (p < 0.01) and rs2296621 for genotype TT (p < 0.05) as compared to heterozygous variants. Furthermore, rs883873 polymorphism for genotype AG (p < 0.05), rs12566180 for genotype TT (p < 0.01) and rs2296621 for genotype TT (p < 0.001) demonstrated a statistically lower TGFBR3 mRNA level with regard to the wild-type carriers.   Figure 2 shows the mRNA expression level of the TGFBR3 gene in 50 EC patients in relation to the genotypes of eight studied SNPs of the TGFBR3 gene. Six SNPs of the TGFBR3 gene were found to modulate its expression. The significant down-regulation of TGFBR3 mRNA was observed in the case of homozygous variant of rs2770186 for genotype CC (p < 0.05), rs12141128 for genotype GG (p < 0.05), rs1805110 for phenotype variant Phe/Phe (p < 0.01) and rs2296621 for genotype TT (p < 0.05) as compared to heterozygous variants. Furthermore, rs883873 polymorphism for genotype AG (p < 0.05), rs12566180 for genotype TT (p < 0.01) and rs2296621 for genotype TT (p < 0.001) demonstrated a statistically lower TGFBR3 mRNA level with regard to the wild-type carriers. Figure 2. Impact of single-nucleotide polymorphisms related to the TGFBR3 gene on betaglycan mRNA expression in women with EC. Data are shown as scatter dot plots, horizontal lines represent median, whereas whiskers correspond to interquartile range. # p < 0.05, ## p < 0.01, ### p < 0.001. Figure 2. Impact of single-nucleotide polymorphisms related to the TGFBR3 gene on betaglycan mRNA expression in women with EC. Data are shown as scatter dot plots, horizontal lines represent median, whereas whiskers correspond to interquartile range. # p < 0.05, ## p < 0.01, ### p < 0.001.

Discussion
Transforming growth factors β isoforms, i.e., TGFβ1, TGFβ2 and TGFβ3, belong to a large superfamily of cytokines, which were identified due to their important role in normal development and homeostasis. TGFβ pathway controls many opposed processes, which is known as the pleiotropic effect on cell and tissue physiology. TGFβ cascade is responsible for both suppression or induction of cell proliferation and apoptosis, as well as regulates autophagy, cell dormancy and senescence. Deregulation of TGFβ signaling, both at induction step and downstream signaling contributes to developmental anomalies and diseases, in particular fibrosis and cancer, which is associated with overexpression of TGFβ isoforms [31,32]. Moreover, in cancer cells, disturbed signal mediation in TGFβ pathway triggers its role from a tumor suppressor, early in neoplastic transformation, to a cancer-promoting and metastatic agent in advanced clinical stages of the disease [33].
In cancer cells, alteration of TGFβ signaling, which plays the pleiotropic role during carcinogenesis, may be influenced by gene polymorphism. Knowledge of the potential role of polymorphism of the TGFBR3 gene encoding betaglycan and its relation to development of EC is elusive. Risk of endometrial cancer development is highly associated with different lifestyle and socio-demographic factors including obesity, onset of menarche, reproductive history, ethnicity and patient's age [34][35][36][37][38][39][40]. Overweight, young age at menarche or nulliparity cause prolonged exposure to estrogens, which possess high proliferative potential, in particular to uterus lining. In the case of obesity, unopposed estrogen stimulation is the result of reduction of progesterone synthesis and higher levels of circulating estrogens. During pregnancy, the estrogen exposure is balanced by the shift toward progesterone signaling. For this reason, nulliparous women have a higher risk of developing endometrial cancer due to extended estrogen stimuli. Moreover, the increased number of births shows a protective effect on endometrial cancer occurrence [40][41][42]. Our study has demonstrated that rs12566180 (c.−114 + 2392C > T), rs6680463 (c.−114 + 7008C > G) and rs2296621 (c.2285 -99G > T) polymorphisms of the TGFBR3 gene are associated with increased risk of EC, both as crude or adjusted for socio-demographic risk factors, such as age, body mass index (BMI), menarche and parity. None of the patients had received EC related hormonal therapy prior to surgery; however, the patients' records concerning hormonal replacement therapy as well as Lynch syndrome history were not available. In spite of the fact, that Lynch syndrome significantly increases the risk of EC to 25-60%, its impact on development of EC can be excluded as its occurrence ranges between 0.5% and 4.6% of all EC cases [9,10].
The rs12566180 (c.-114 + 2392C > T) and rs2296621 (c.2285 − 99G > T) polymorphisms significantly altered in endometrial cancer are located within intronic regions of the TGFBR3 gene, which may indicate their potential impact on transcription and stability of the primary transcript. Obtained results suggest that studied SNPs are involved in the observed betaglycan down-regulation in endometrial cancer; however, the only one of them, i.e., rs2296621 (c.2285 − 99G > T), seems to be related to pronounced tumor aggressiveness. Bioinformatic analysis using RegulomeDB showed that rs12566180 (c.-114 + 2392C > T) and rs6680463 (c.−114 + 7008C > G) have a score of 4, whereas rs2296621 (c.2285 − 99G > T) has a score of 2 [30]. Taking into account SNPs case-control study, haplotype analysis and genotype-phenotype findings, the observed results indicate that rs12566180 (c.-114 + 2392C > T), rs6680463 (c.−114 + 7008C > G) and rs2296621 (c.2285 − 99G > T) could be regarded as potential markers for EC. However, further studies are required.
Furthermore, besides above-mentioned relationship between SNPs of the TGFBR3 gene and neoplastic transformation, the importance of polymorphism in betaglycan encoding gene was reported for other non-cancerous diseases, i.e., premature ovarian failure (POF), testicular dysgenesis, sickle cell anemia, pulmonary emphysema and primary open angle glaucoma [46][47][48][49][50][51]. It is suggested that the TGFBR3 gene polymorphism may play a potential role in determining bone mineral density, as well as optic disc area parameters [52][53][54].
Interestingly, in our study, six out of eight analyzed SNPs of the TGFBR3 gene were found to have an impact on betaglycan expression in EC. Altered betaglycan expression may be responsible for impaired TGFβ signaling initiated by TGFβ isoforms and simultaneous redirection of this signal to Smad-independent pathways. TGFBR3 gene downregulation has been stated in the case of different cancers, and the observed decline in the expression of the TGFBR3 gene appears to be correlated with cancer progression, when tumor cells demonstrate an increase invasiveness and metastatic potential [17][18][19][20][21][22][23][25][26][27][28].
As previously described by different research groups, TGFβ signaling induced by TGFβ isoforms may be engaged in the induction of epithelial-mesenchymal transition (EMT). EMT is a biological process characterized by the reorganization of the epithelial tissue structure and is manifested by the acquisition of the mesenchymal phenotype resulting in the loss of polarity and adhesion by cells together with ability to migration and invasion. This process plays a vital role during physiological events, i.e., embryogenesis, organogenesis and morphogenesis of different tissues, wound healing, as well as inflammation. In cancer progression, EMT is responsible for the development of drug resistance and metastasis due to the increased cancer cell motility [55][56][57].
In summary, our study has demonstrated for the first time the role of the TGFBR3 gene polymorphism and its association with the increased risk of EC development. Moreover, we have shown that the TGFBR3 gene SNPs may modulate betaglycan expression at the transcriptomic level. Our findings contribute to a better understanding of the importance of gene polymorphism in the TGFβ signaling, especially at the level of signal initiation through TGFβ2 isoform mediated exclusively by betaglycan. Along with our previous findings concerning the significance of allelic loss in the TGFBR3 gene, where LOH was reported in 52% of examined cancer samples, SNPs may be an additional mechanism responsible for betaglycan deregulation in EC [28,58]. What is more, obtained results strongly support the view of individual variability among EC patients and suggest the necessity of developing personalized diagnostic and/or therapeutic approach in the treatment of endometrial cancer. Table S1: Characteristics and sequences of TaqMan probes used for genotyping of TGFBR3 gene; Table S2: The association between significantly altered SNPs, i.e., rs12566180, rs6680463 and rs2296621 polymorphisms and clinico-pathological parameters of studied cancer samples.