This article is mentioned in:

Introduction

Endometrial cancer is one of the most common and lethal gynecological malignancies in Poland. However, the causes of endometrial carcinogenesis remain to be clarified. Possible causes include the imbalance of endogenous estrogen and progesterone levels, obesity, polycystic ovarian syndrome and estrogen replacement therapy (1,2). Since the 1980s, endometrial cancer has been classified into two main subsets of sporadic endometrial cancer that differ in molecular genesis and prognosis (3). Type I endometrioid cancers are estrogen-dependent cancers that develop from hyperplasia and are usually low grade with a favorable prognosis. Endometrioid cancers account for ~70–80% of cases, affecting mainly younger, pre- and post-menopausal females. In this type of tumor, specific molecular aberrations such as PTEN gene inactivation by mutation and/or promoter methylation, mutation of protooncogene H-RAS, mutation of CTNNB, and microsatellite instability (MSI) are common (4). Type II sporadic endometrial cancer is non-endometrioid endometrial carcinoma (NEEC), which progresses from an atrophic endometrium, occurs in older females and has a more aggressive course of disease. This type of cancer is estrogen-independent and characterized by frequent molecular alterations in oncoprotein HER2/neu, TP53 mutation and inactivation of CDH1 (4,5). Experiments conducted in this study focused on endometrioid adenocarcinoma, which is the type I endometrioid cancer.

Clinicopathologic variables such as age, FIGO stage, histological grade, myometrial invasion, metastasis to lymph node and histological type are crucial prognostic factors (6). However, new molecular markers should be identified to improve the prediction of therapy outcome and prognosis. Currently, the potential markers available remain controversial and intensely discussed (6).

WWOX is localized in the common fragile site, FRA16D (locus 16q23.3–24.1). It has been confirmed to be altered in various types of cancer including breast, lung, gastric, ovarian, Wilms' tumor and glioblastoma multiforme (7–13). Genetic and epigenetic alterations of this gene include loss of heterozygosity (LOH) and promoter methylation. The gene product is an oxidoreductase comprising two WW protein interaction domains. One of the roles of WWOX protein is participation in steroid hormone metabolism (14,15). Results of previous studies showed that WWOX is also associated with apoptosis, proliferation, adhesion and cell signaling pathways (16–19). Additionally, WWOX has been shown to bind to PPxY motif-containing proteins, and inactivate their transcription transactivation function by sequestering them in the cytoplasm (20,21). It is known that WWOX modulates the Ap2α/γ, p73, ErbB4, Met, Jun, Wnt signaling pathways. Moreover, Gourley et al identified its role in the decrease of integrin activity and adhesion of tumor cells to fibronectin (22).

The expression level of WWOX is known to be decreased in breast cancer cells and to correlate with poor prognosis (11). In a series of in vitro experiments, a WWOX-transfected MDA-MB-231 breast cancer cell line showed increased migratory ability. However, the results of a test of growth in Matrigel showed that the transduced cells had more ‘normal’ phenotype and formed mammary ducts. Control cells in Matrigel grew into spherical structures, typical of neoplastic cells (22). The improved differentiation evident in cells with an elevated WWOX level suggests its involvement in these type of processes. The tumor suppressor function of WWOX was confirmed in a soft agar growth test, where WWOX-transfected cells exhibited inhibition of anchorage-independent growth (23). MDA-MB-231 cells overexpressing WWOX were also significantly less tumorigenic in vivo (24). Experiments performed by Gourley et al on ovarian cancer cells confirmed that WWOX protein is an inhibitor of anchorage-independent growth also in this case. Moreover, WWOX silencing was found to result in enhanced adhesion to fibronectin (22).

No data is currently available on the role of the WWOX gene in endometrial cancer. The present preliminary qPCR-based study was conducted on 79 endometrial adenocarcinoma in relation to 28 tumor-free endometrial tissue samples. The aim of this study was to investigate the correlation of the expression levels of WWOX and nine other tumor-related genes: MKI67, BAX, BCL2, EGFR, CCNE1, CCND1, CDH1, TP73 and NCOR1. Additionally, the implications of loss of heterozygosity with regard to the regulation of WWOX expression in endometrial cancer were also analyzed.

Materials and methods

In total, 79 samples of endometrial carcinoma (endometrioid adenocarcinoma) were collected at the Department of Gynecological Oncology, Medical University of Lodz, Poland. The tumors were classified according to the FIGO (International Federation of Gynaecology and Obstetrics) classification system. The mean age of the patients was 61 years (median 60, range 36–83 years). The samples were examined histologically and stored at −80°C in RNAlater buffer (Ambion, Inc., Austin, TX, USA) until RNA extraction. Clinical characteristics of the patients are presented in Table I. Experiments involving human subjects were conducted according to the Declaration of Helsinki, and the study was approved by the Ethics Committee at the Medical University of Lodz. Control samples (n=28) were received from patients operated on for benign gynecologic disorders.

Table I Clinical characteristics of endometrial cancer patients.

Table I

Clinical characteristics of endometrial cancer patients.

Factor	No. of patients
FIGO stage
I	44
II	16
III	10
NS	9
Lymph node metastasis
Negative	65
Positive	8
NS	6
Histological grade
I	27
II	39
III	12
NS	1
Myometrial invasion
<1/2	41
>1/2	33
Without	4
NS	1

[i] NS, not specified.

qPCR

RNA was extracted from frozen tissues, stored at −80°C in RNAlater (Ambion), using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). cDNA synthesis was performed from 10 μg of total RNA at a volume of 100 μl using ImProm RT-II™ reverse transcriptase (Promega, Madison, WI, USA). Reverse transcription was carried out under the following conditions: incubation at 25°C for 5 min and 42°C for 60 min, and heating at 70°C for 15 min. cDNA samples were diluted with sterile deionized water to a total volume of 150 μl, and 2 μl was added to a PCR reaction. qPCR was performed using Rotor-Gene™ 3000 (Corbett Research, Australia). PCR products were detected using SYBR®-Green I and qPCR Core kit for SYBR-Green I (Eurogentec, Seraing, Belgium). Reactions were performed in duplicate. The relative expression levels of the following genes were analyzed: WWOX, TP73, CCND1, CCNE1, BCL2, BAX, MKI67, CDH1, EGFR, and NCOR1. The expression levels of the investigated genes were normalized to three reference genes (RPS17, H3F3A, RPLP0). Due to the presence of a relatively low level of WWOX mRNA, a semi-nested RT-PCR was used for the detection of WWOX expression levels. The first PCR reaction was performed with primers: 5′-TGCAACATCCTCTTCTCCAACGAGCTGCAC-3′ and 5′-TCCCTGTTGCATGGACTTGGTGAAAGGC-3′ in a 50 μl reaction volume. Subsequently, 2 μl of 200-fold diluted PCR product (171 bp) was used as a template for semi-nested PCR. The cycling protocol consisted of 2 min at 94°C, 30 sec denaturation at 94°C, 30 sec annealing at 63°C, 1 min extension at 72°C, repeated for 77 cycles, and additional extension for 7 min at 72°C.

The primer sequences, PCR reaction conditions and the length of obtained products are available upon request.

Relative gene expression was calculated based on the Roche company guidebook according to the previously published algorithm (25). Universal Human Reference RNA (Stratagene, La Jolla, CA, USA) composed of 10 cell lines was used as a calibrator.

The primers were designed to be intron-spanning to avoid amplification of genomic DNA. The detection temperature was determined above the non-specific/primer-dimer melting temperature.

Loss of heterozygosity analysis

LOH detection was performed using the high-resolution melting method in a LightCycler480 (Roche Molecular Systems, Penzberg, Germany). Genomic DNA was recovered after RNA isolation using back extraction buffer (BEB, 1 M Tris Base, 4 M guanidinium thiocyanate, and 50 mM sodium citrate) according to the manufacturer's instructions. Allelic losses were analyzed by PCR amplification with two sets of primers for microsatellites D16S518 (intron 1 of WWOX gene) and D16S3096 (intron 8). The primer sequences were obtained from the Genome database. PCR cycling programs included one cycle with 95°C for 10 min followed by 35 cycles consisting of 94°C for 30 sec, 56°C (for D16S3096) or 55°C (for D16S518) for 30 sec, 72°C for 60 sec. The high-resolution melting conditions involved a temperature increase of 50–95°C, ramp rate 0.01°C/sec and 40 acquisitions per °C.

Western blot analysis

The tissue fragments were lysed in RIPA protein extraction buffer supplemented with protease, phosphatase inhibitor cocktail and PMSF (Sigma-Aldrich, St. Louis, MO, USA). The protein concentration was measured using the Bradford method (Bio-Rad Laboratories, Hercules, CA, USA), and 100 μg amounts were run on 10% SDS-PAGE gel electrophoresis and subsequently transferred to a PVDF membrane (Sigma-Aldrich). The membranes were blocked in 5% non-fat milk in TBST for 1 h at room temperature and then incubated for 19 h at 4°C with primary antibodies (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA). Following incubation, the membranes were washed three times with TBST and incubated with secondary antibodies conjugated with alkaline phosphatase (Sigma-Aldrich) for 1 h. Membranes were washed three times in TBST and developed using Novex® AP Chromogenic Substrate (Invitrogen). Glyceraldehyde-3-phosphate dehyrogenase (GAPDH) was used as a reference protein. The relative protein amount was assessed with ImageJ software (National Institutes of Health, USA) based on integrated density of bands.

Statistical analysis

Data analysis was performed by using Statistica version 8.0 (StatSoft, Tulsa, OK, USA). Gene expression correlation analysis was performed using the non-parametric Spearman's rank correlation test. The Mann-Whitney t-test was used to determine differences between the transcription levels of the WWOX gene in relation to its hemi/heterozygosity, as well as differences between quantities of WWOX in different types of tissue. Results were recognized as being statistically significant at a confidence level of >95% (P<0.05).

Results

LOH

Loss of heterozygosity in the region of the D16S3096 microsatellite marker (localized in intron 8 of WWOX gene) was found to be a common event in endometrial cancer, with observed hemizygosity at 38%. Moreover, hemizygous samples exhibited decreased WWOX gene expression levels compared with heterozygous samples (medians: 0.66 vs. 1.04, P=0.066). The percentage of LOH for the microsatellite marker, D16S518, was 40%; however, hemizygosity did not reveal any tendencies to correlate with decreased WWOX expression levels (Fig. 1).

Figure 1 Correlation of loss of heterozygosity (LOH) with WWOX expression levels. (A) Heterozygous sample melting peaks. (B) Hemizygous sample melting peaks. (C) Relative WWOX mRNA expression in heterozygotes and hemizygotes.

Expression correlation of WWOX and various tumor-related genes

The expression correlation between WWOX and nine other tumor-related genes is assessed. MKI67 was connected with the rate of proliferation, BAX, BCL2 and TP73 are involved in the course of apoptosis, CCNE1 and CCND1 encode cyclins crucial for cell cycle progression, EGFR and NCOR1 act as receptors and transcriptional factors, while CDH1 is a gene of one of proteins responsible for cell-cell adhesion. A positive correlation was observed between the expression of WWOX and BCL2 (Rs = 0.3822; P=0.0005), CCND1 (Rs = 0.3821; P=0.0005) as well as with the BCL2/BAX anti-apoptotic ratio (Rs = 0.4496; P=0.0001). WWOX expression correlated inversely with BAX (Rs = −0.2302; P=0.0412), CDH1 (Rs = −0.4126; P=0.0002) and NCOR1 (Rs = −0.3061; P=0.0064) expression levels. Details of the expression correlations between WWOX and the nine investigated genes are presented in Table II.

Table II Correlation analysis between the expression levels of WWOX and other tumor-related genes.

Table II

Correlation analysis between the expression levels of WWOX and other tumor-related genes.

Gene	Spearman's rank correlation	P-value
BCL2/WWOX	0.3822	0.0005
BAX/WWOX	−0.2302	0.0412
CCND1/WWOX	0.3821	0.0005
CDH1/WWOX	−0.4126	0.0002
NCOR1/WWOX	−0.3061	0.0064
BCL2_BAX ratio/WWOX	0.4496	0.0001
EGFR/WWOX	0.1004	0.3786
CCNE1/WWOX	−0.0162	0.8872
MKI67/WWOX	−0.1132	0.3314

No correlation was found between WWOX mRNA levels and clinicopathological factors such as grade, FIGO stage, lymph node metastasis or myometrial invasion (Table III). However, the highest expression of WWOX was observed in normal endometrium tissue (NT) in comparison to tumor tissue (median expression 2.826, NT vs. G1, P=0.003; NT vs. G2, P<0.0001; NT vs. G3, P=0.002).

Table III Dependence of the WWOX expression levels and clinical characteristics of endometrial cancer patients.

Table III

Dependence of the WWOX expression levels and clinical characteristics of endometrial cancer patients.

Factor	n	Median of WWOX mRNA level (range)	P-value
FIGO stage
I	44	1.004 (0.678–1.613)	0.701 (I/II)
II	16	0.716 (0.357–1.370)	0.527 (II/III)
III	10	0.871 (0.554–4.267)	0.982 (III/I)
NS	9
Lymph node metastasis
Negative	65	0.870 (0.613–1.295)	0.646
Positive	8	0.871 (0.554–1.436)
NS	6
Histological grade
G1	27	1.125 (0.494–1.824)	0.386 (I/II)
G2	39	0.732 (0.613–0.974)	0.756 (II/III)
G3	12	0.722 (0.306–1.613)	0.523 (III/I)
NS	1
Myometrial invasion
<1/2	41	1.050 (0.641–1.563)	0.279
>1/2	33	0.856 (0.410–1.226)
Without	4
NS	1

Western blot analysis

To assess the level of WWOX protein in normal and cancer tissues, a western blot assay was conducted, which revealed that the protein amount was greater in normal endometrium tissue compared with cancer samples, although this tendency did not achieve statistical significance (P>0.05). There was no difference between tumor grades. Such results are reflected in the amounts of mRNA. However, in the qPCR analysis of gene expression the differences between normal and cancer tissues were statistically significant. The smaller differences in protein between normal and tumor tissues may depend on the low level of the WWOX protein, which hindered western blot sensitivity. The densitometric analysis and protein bands are presented in Fig. 2.

Figure 2 Expression of WWOX protein in representative normal endometrium tissues and grade 1–3 cancers.

Discussion

Apoptosis is a natural process for the elimination of senescent cells from a layer of endometrium during late secretory and menstrual phases. The mRNA levels of apoptosis-related genes change during phases of the ovulation cycle. The expression level of the antiapoptotic Bcl2 gene increases during the late proliferative phase but decreases during the late secretory and menstruating phases when the expression of the Bax proapoptotic gene increases. Anomalies in the expression of apoptosis-related genes may lead to pathological changes including endometriosis and cancer (26).

In the present study, positive correlations were found between WWOX and the antiapoptotic BCL2 gene as well as between WWOX and the BCL2/BAX ratio. Findings of a previous study demonstrated the increasing expression of the proapoptotic BAX gene and decreasing expression of the antiapoptotic BCL2 gene during progression from endometrial hyperplasia to cancer (27). Geisler et al (28) demonstrated that a higher expression level of the Bcl2 protein correlates with favorable clinicopathological variables such as well-differentiated tumor cells, reduced FIGO stage, lack of invasion into lymph node and superficial myometrial invasion (28,29). Chao et al observed a correlation of BAX overexpression in endometrial cancer specimens in relation to normal endometrium and premalignant lesions (30). A negative correlation between WWOX and BAX gene was also identified in the present study. These results are similar to our previously reported gene expression analysis conducted on breast cancer (11) and glioblastoma multiforme patients (9). Investigations on apoptosis of an A2780 ovarian cancer cell line transfected with the WWOX gene demonstrated a decreased ability for anchorage-independent growth with a simultaneous increase of apoptosis (31).

An important regulator of the cell cycle is cyclin D1. Expression of CCND1 can be regulated by several signaling pathways, such as RAS or PTEN. In type I endometrioid cancer, expression of the CCND1 gene is connected with the proliferation process Wnt signalling pathways (32). Findings of a previous study identified an increase in expression of cyclin D1 from normal endometria to hyperplasia and carcinoma (33). Moreno-Bueno et al suggest two different causes of cyclin D1 overexpression: an amplification of the gene in NEEC and a microsatellite instability in endometrioid cancer (34). In the present study, a positive correlation of WWOX and CCND1 was observed.

Additionally, results of the present study demonstrate that WWOX expression level correlates negatively with the NCOR1 (nuclear receptor corepressor 1) ERα corepressor gene. The corepressor suppressess transcription estrogen-responsive genes by modeling chromatin structures by incorporating histone deacetylases (HDAC) (35). Using a microarray method Moreno-Bueno et al noted a 2-fold higher expression of NCOR1 in endometrioid cancer in comparison to NEEC samples (36). However, such differences between tumor groups were not observed by Kershah et al, although they demonstrated the upregulation of nuclear receptor coregulators including NCOR1 in a malignant endometrium, as compared to a normal one. The ratio between coactivators SRC-1 and SRC-2 and corepressor NCOR1 decreased in malignant tissues. No significant differences were identified between tumor groups regarding NCOR expression, categorized on the basis of such clinical parameters as grades or stages (37). Upregulation of NCOR1 was also observed in breast cancer, however, a low expression of this gene was associated with worse prognosis and serves as a potential predicting factor for tamoxifen treatment in estrogen receptor α-positive breast cancer (38). In a previous study on ER-positive breast cancer patients, a positive correlation was noted between the expression level of WWOX and NCOR1 (data not shown).

The CDH1 gene encoding the cell-cell adhesion protein E-cadherin is located near WWOX on chromosome 16 (CDH1 locus 16q22.1, WWOX locus 16q23.3–24.1). As previously shown, CDH1 expression is often reduced or completely inactivated by promoter methylation. A low E-cadherin expression is associated with worse prognosis, higher stage and greater metastatic potential (39). Our previous experiments conducted on breast and colon cancer lines confirm that an increase of WWOX expression level results in changes in cell behavior (23, unpublished data). Cancer cells with a high WWOX express a higher motility, which has an effect on improved migration through the basal membrane. Additionally, they are less malignant due to the suppression of anchorage-independent growth. This change in cell motility may explain the observed correlation of WWOX expression with the reduced expression of the cell adhesion gene CDH1. A previous in vivo study revealed the role of WWOX protein in the attachment and adhesion of ovarian cancer cells. WWOX-transfected PEO1 cells showed a decrease in the adhesion to fibronectin in comparison to vector-transfected control cells, which suggests a WWOX influence on processes such as tumor invasiveness and spread (40). Gourley et al also confirmed these results on the ovarian cancer cell line, A2780, and showed that WWOX overexpression reduces adhesion through membranous integrin α3 protein (22).

In previous studies, a decrease in the expression of the WWOX gene was found to be associated with loss of heterozygosity (LOH) in gastric (7), pancreatic (18), esophageal (41) and lung cancer (13). In the present study, the percentage of hemizygosity at two analyzed loci of the WWOX gene was ~40%. LOH in microsatellite marker D16S3096, exhibited a tendency towards a correlation with the reduced expression level of the WWOX gene. This observation suggest that this process is involved in the regulation of the WWOX mRNA level.

In conclusion, to the best of our knowledge, this is the first study to demonstrate the potential role of WWOX in endometrial cancer through the regulation of Wnt (CDH1 and CCND1), apoptosis (BCL2 and BAX) and estrogen-related genes (NCOR1). Results of the present study have also shown that WWOX mRNA and protein levels decrease in transformed endometrial tissue. The fact that no significant differences exist between tumor grades suggests that WWOX silencing is an early event in endometrial cancerogenesis. Similar to other types of cancer, WWOX expression in endometrial adenocarcinoma correlates with the expression level of apoptosis and cell cycle regulators. The results of our preliminary experiment have shown that additional investigations should be conducted that may enable the better elucidation of the role of WWOX in endometrial cancer. Future experiments are to be conducted on endometrial cell lines, which may shed some light on the functions performed by the WWOX protein and its relevance to endometrial cancer promotion and progression.

Acknowledgements

This study was funded by the National Center of Sciences N N407 168940.

References