Studies on selected molecular factors in endometrial cancers

Results. Expression of ERβ1 was augmented in EC type II, in poorly differentiated cancers and with growing clinical advancement. An augmented expression of ERα was noted in well-differentiated EC and at lower clinical stage. An increased expression of PR and decreased of MLH1 were detected in type I EC. The expression of ARID1A and c-MET proteins showed no differences between the types of EC, stages of clinical advancement or grading. In 51.6% patients with type II EC, a loss of BRCA1 expression was disclosed; in this group of cancers a decreased expression of ERα was noted.


Introduction
According to global statistics, endometrial carcinoma (EC) is diagnosed in 4.8% of the female population.In 2012, EC developed in 319,605 women, of which 76,160 cases were fatal.In Europe since 2005 a stable increase has been noted in morbidity and mortality due to EC.In 2012 almost 100,000 women developed EC, which accounts for 6.2% of all morbidities due to malignant tumors in women. 1,2ulti-year observations indicate that ECs vary in etiology, clinical course and prognosis.Since Bokhman's hypothesis was 1 st proposed, 2 types of EC have been distinguished. 3ype I is the most frequently diagnosed in around 80% women: mainly endometrioid, it is linked to an unbalanced estrogen stimulation and metabolic syndrome, manifests a slow course and good prognosis.It carries common receptors for estrogens (ER) and progesterone (P), while their expression depends on the degree of clinical advancement and histological grade.Most of the cancers are sporadic; around 3-5% are linked to mutations in the mismatch repair genes (MMR): MLH1, MSH2, MSH6 and PMS2.][6][7][8][9] Type II is a non-endometrioid cancer that manifests an aggressive biology, encompassing serous, clear-cell and poorly differentiated cancers.Frequent relapses cause an unfavorable course.It is a carrier of TP53 (p53) and HER2/neu mutations.][12][13][14][15] Uterine serous carcinoma (USC) is thought to represent a unique type of EC, which should be treated as a distinct morbid unit. 16he results of studies covered by the Human Cancer Genetic Program (543 unselected female endometrial cancers) and focused on MSI demonstrated that, even though MSI was documented in 21.7% of the studied cases, genetic testing for MMR (MLH1, MSH2, MSH6 and PMS2) detected Lynch syndrome in just 1.8% of cases. 17Another study showed that MSI was more frequent in metastatic than in primary EC, that it appeared late in tumor development and that it might promote progression. 18Brinton et al., when reporting the results of the Gynecologic Oncology Group (GOG 210 Trial), highlighted the similarity of endometrioid grading 3 (G3) cancers and type II cancers. 19The authors supported the hypothesis suggesting heterogeneity of EC type II cancers.
As stated above, the mutation of PIK3CA is linked to EC type I, although alterations in PIK3CA are also present in USC. 20Studies by Takeda et al. showed that mutations in the ARID1A suppressor gene induce an altered expression of many genes, including MLH1 and genes linked to the PI3K/AKT signaling pathway, and are therefore associated with both type I and type II EC. 21The PI3K signaling pathway also involves the MET proto-oncogene and its HGF ligand.Studies by Bishop et al. proved that the expression of c-MET occurred likewise in USC. 22n recent years USC has been found to be potentially associated with carriership of a mutated BRCA1 gene.Bruchim et al. subjected women with histologically documented USC to genotyping of 3 main mutations, including BRCA1 (185delAG and 5382insc) and BRCA2 (6174delT). 23hey found that over 25% of women with USC carried mutations in BRCA1/2.Similar observations were made in English women in 2013; 68% of patients with USC had suffered from breast cancer before being diagnosed with USC.The authors also proposed that at least a subgroup of USC should be recognized as hereditary breast/ovarian cancer, which might carry prophylactic implications (prophylactic adnexectomy) and therapeutic implications (inhibitors of poly adenosine diphosphate ribose polymerase -PARP). 24The heterogeneity of EC, particular of EC type II, requires further molecular studies.
This study aimed to estimate expression manifested by hormonal receptors (ERα, EBβ1, PR), expression of MSH2 and MLH1 proteins involved in the development of certain endometrial cancers belonging to the Lynch syndrome, expression of proteins linked to poor prognosis and metastases (ARID1A, c-MET), and expression of BRCA1 protein in type II EC, which might indicate that a proportion of the cancers are dependent on mutations in BRCA1 and represents a proportion of breast/ovarian cancer syndrome.

Material and methods
The study had a multi-center, retrospective character.The research material included archival histopathological preparations of endometrial carcinoma, obtained from 162 patients diagnosed and treated due to EC in 2007-2014 in 8 specialized centers treating the diseases from the oncological gynecology branch.The study group consisted of patients treated primary with surgery.Due to insufficient clinical data in 16 patients, further analyses were conducted on 146 patients.The mean age of the patients was 65.1 years.
The preparations were incubated in a water bath at 96°C in a citrate buffer, pH 6.0, for 50 min.The activity of endogenous peroxidase was blocked using 3% H 2 O 2 .The preparations were incubated with the antibody at room temperature for 60 min, followed by 10 min rinsing in tris-buffered saline (TBS).The tissue material was incubated with the EnVision system (DakoCytomation, K5007; Dako, Santa Clara, USA) for 30 min.In all preparations, 3,3'-diaminobenzidine (DAB-3.3)was used to visualize the reaction.Subsequently, the preparations were counterstained with Mayer's hematoxylin, then passed through a row of alcohol to xylene and finally closed under a coverslip.
In immunohistochemical tests the negative control involved a reaction with omission of the primary antibody.Using an Olympus BX 43 light microscope and XC 30 digital camera (Olympus, Shinjuku, Tokyo, Japan), 10 photographs were taken of every stained preparation with the immunohistochemical reaction.The photographs were taken at total magnification of ×400.
The expression of BRCA1 protein was evaluated in cancer tissue in patients with non-endometrioid cancer.In the cases where the BRCA1/MLH1/MSH2 expression were evaluated in the studied preparations, it was assumed that the presence of the color reaction indicated an absence  of mutations in the BRCA1/MLH1/MSH2 gene while the absence of a color reaction indicated the loss of protein expression, which may be the result of the BRCA1 gene mutation and MLH1/MSH2 mutations or hypermethylation of their promoter in the cancer tissue.The immunohistochemical reaction was detected in both cell nuclei and in the cytoplasm with the use of the ERβ1-specific antibody.The c-MET protein manifested a cytoplasmic reaction while a nuclear reaction was shown by reactions detecting MLH1, MSH2, BRCA1, PR, ERα and ARID1A (Fig. 1-5).
Statistical calculations were performed with STATIS-TICA v. 10 software (StatSoft Inc., Tulsa, USA).The Mann-Whitney, the Kruskal-Wallis and Spearman's tests were used.Statistical significance was set at p < 0.05.

Histological type
Based on histopathological diagnosis, the patients were divided into 2 groups: patients with 1) endometrioid type and 2) non-endometrioid type of cancer.The latter group included patients with serous cancer, clear-cell cancer and mucinous cancer.In patients with endometrioid type cancer, a decreased expression of both MLH1 protein and ERβ1 was observed (p = 0.013 and p = 0.035, respectively) (Fig. 6).A reduced expression of PR receptor was detected (intensity of the reaction of 3-2 vs 1-0) among patients with non-endometrioid EC (p = 0.041).The reduced immunohistochemical expression of PR receptor was seen mainly in patients with serous cancer (p = 0.022), while intensified immunohistochemical reactions for MLH1 and ERβ1 involved clear-cell cancers (p = 0.02184 and p = 0.00109).Following the subdivision of non-endometrial cancers to individual subtypes, the subgroup of patients with clear-cell cancers manifested a reduced expression in the immunohistochemical reaction specific for ERα receptors as compared to the expression noted in endometrioid cancers (p = 0.048) and a higher expression of immunohistochemical reaction for ERβ1 as compared to expression noted in serous cancers (p = 0.02658).No differences were detected in the expression of immunohistochemical reactions specific for the remaining proteins in the studied groups.

Grading
An increase in the expression of ERβ1 receptor was found in parallel to the decrease in histopathological differentiation (G3 vs G1 p = 0.003) (Fig. 7).Patients with G1 or G2 endometrial cancer manifested a higher expression of ERα receptors than patients with a G3 cancer (G1 + G2 vs G3) (p = 0.011).No differences were detected in the expression of MLH1, MSH2, PR, c-MET and ARID1A dependent on the histopathological grading of the cells.in the expression of MLH1, MSH2, PR, ERβ1, c-MET or ARID1A were detected among women with positive or negative BRCA1 staining (Fig. 9).

Discussion
Many endometrioid adenocarcinomas are thought to carry receptors for estrogens and progesterone.According to Reid-Nicholson et al., as many as 84% of endometrioid cancers (type I) manifest G1 and G2 maturity express ER receptors, as compared to ER expression in 9-54% of non-endometrioid (serous and clear-cell cancers). 6The expression of such receptors manifests a correlation exclusively with histological grading, but not with the clinical stage of the disease.
Estrogen receptors may be present in 2 isoforms: ERα and ERβ, which exhibit distinct functions. 25Estrogen receptor β is thought to function as a guardian of the endometrium; its disturbed expression has been described in most endometrial cancers. 26In our patients, the examination of ERβ1 expression demonstrated an increase in non-endometrioid cancer -type II EC (mainly serous carcinoma and clear-cell carcinoma) (Fig. 6) as compared to type I EC: 43.3 vs 21.9, respectively.The expression of ERβ1 increased parallel with histological grading: it was least pronounced in G1 -9.1, higher in G2 -26.0 and highest in G3 -38.5 (Fig. 7).It also increased with the clinical advancement of the cancer: FIGO IA 11.8 vs FIGO IB-IV 31.7.In studies by Chakravarty et al. a decrease was also noted in the expression of ERβ in endometrioid EC cancers, but no differences were detected in the expression as related to grade. 27n the other hand, on the basis of our studies it may be accepted that an increased expression of ERβ1 in type II EC,

International Federation of Gynecology and Obstetrics gynecologic cancer staging system
For the analysis, patients with endometrial carcinoma were subdivided depending on the stage of clinical advancement of the disease into early stage of advancement (IA) and the late stage (IB-IV).Women manifesting stage IA showed a statistically significant higher expression of the ERα receptor (p = 0.04).Patients with a more advanced disease manifested an augmented expression of the ERβ1 receptor (p = 0.02) (Fig. 8).No differences were detected in the expression of MLH1, MSH2, PR, c-MET and ARID1A, which would depend on the stage of clinical advancement manifested by the disease.

Mutation in BRCA1
Patients with non-endometrioid cancer were subjected to an evaluation for the expression of BRCA1 protein in cancer tissue.Among 31 patients with type II EC, 16 proved to have negative expression of this protein (Table 2).
Among the patients with non-endometrioid type of EC lack of BRCA1, expression was correlated with reduced expression of ERα receptor (p = 0.02).No other differences in the presence of poor differentiation (G3) and a higher stage of clinical advancement, was associated with poor prognosis.
Studies by Kreizman-Shefer et al. demonstrated that early endometrioid cancers preserve their expression of ER and PR, while poorly differentiated and clinically advanced cancers manifested an absence of 1 or both receptors. 12imilarly, in our studies the expression of ERα was higher in well or moderately differentiated cancers than in G3 cancers (48.5 vs 36.5 -Fig.7), and it was also significantly higher in IA cancers according to FIGO (Fig. 8, FIGO 1A 55.8 vs FIGO IB-IV 34.6).We could not detect ERα expression in non-endometrioid type II cancers, although Sho et al. described it in 21.2% cases of USC, and this was linked to a poor prognosis. 14n studies by Togami et al., ER and PR expression in USC was associated with a good prognosis. 13In our studies the expression of the progesterone receptor PR was significantly more pronounced in type I endometrioid EC (89.8 in EV type I vs 75 in EC type II), which was consistent with the results of the study by Togami et al. 13 However, we failed to identify differences in PR expression which would depend on histological grading (G), as indicated in the studies by Reid-Nicholson et al. and those by Zhu et al. 6,28 Since, as mentioned above, around 3-5% of EC are linked to mutations in DNA-repair genes (MMR), we estimated the expression of the 2 main relevant proteins, products of MSH2 and MLH1 genes, responsible for 85% cases of Lynch syndrome in all our patients. 7,9,29We detected a lower expression of MLH1 protein in endometrioid cancer (type I 51.9 vs type II 77.4), which may indicate that the MLH1 gene mutation occurred more frequently in cases of type I EC.Berends et al. noted the loss of MLH1 expression among women with EC connected to Lynch syndrome. 30o abnormalities in the expression of ARID1A protein were revealed in our study.Other studies detected a loss of ARID1A expression in around 30% of EEC cancers, in the progression of atypical hyperplasia to cancer as well as in the induction of many genes, including MLH1. 21,31urthermore, the expression of the c-MET protein demonstrated no change in any of the parameters we examined, despite the evident association between c-MET and poor prognosis and metastasizing documented in other studies on EC. [32][33][34][35] In 16 of our 31 patients (51.6%) with a diagnosis of EC type II, immunohistochemical tests demonstrated an absent expression of BRCA1 protein, indicating a mutation in BRCA1.Such incidence was much higher than described by Bruchim et al. 23 In studies by Raffi et al., 50.5% of USC patients were found to develop breast cancer (17.5% before and 33% following diagnosis of USC), which, according to the authors, suggested that the cases represented a proportion of BRCA mutation syndrome. 24In our studies, the significantly reduced expression of the estrogen receptor α in this group of women might additionally indicate a relationship between some USC and BRCA mutation; it has been postulated that certain BRCA1 proteins inhibit ERα activity. 36

Conclusions
An augmented expression of ERβ1 in EC was linked to type II EC.Higher expression of ERα in EC cancers was associated with a lower histopathological grade.A decreased expression of MLH1 protein was estimated in EC type I, which may indicate a mutation in MLH1 gene in this type of cancer.Type II EC may be connected to BRCA1 mutation.

Fig. 3 .Fig. 4 .Fig. 5 .
Fig. 3.A pronounced cytoplasmic reaction with c-MET-specific antibody in a section of endometrial adenocarcinoma (A and B).No such reaction in cell nuclei.Magnification ×400

Fig. 8 .Fig. 9 .
Fig. 8.The values reflect the proportion of endometrial cancer cases manifesting expression of a given protein in cells of early (IA) or more advanced degree (IB-IV) of clinical stage according to FIGO (immunohistochemical reaction +++/++).* p < 0.05

Table 1 .
Characteristics of patients included in the study

Table 2 .
BRCA1 protein expression status in studied patients with endometrial carcinoma type II