Anti-Müllerian Hormone Expression in Endometrial Cancer Tissue

Anti-Müllerian hormone (AMH) is a commonly known factor secreted by Sertoli cells, responsible for regression of the Müllerian ducts in male fetuses. AMH has also other functions in humans. In vivo and in vitro studies have shown that AMH inhibits cell cycle and induces apoptosis in cancers with AMH receptors. The aim of the study was to assess whether the tissue of pre-cancerous states of endometrium (PCS) and various histopathologic types of endometrial cancer (EC) exhibit the presence of AMH. We aimed to investigate whether the potential presence of the protein concerns menopausal women or those regularly menstruating, and whether is related to cancers with a good or a bad prognosis, as well as what other factors may influence AMH expression. The undertaken analysis was carried out on tissues retrieved from 232 women who underwent surgical treatment for PCS and EC. Tissues were prepared for immunohistochemical assessment with the use of a tissue microarrays method. AMH expression was confirmed in 23 patients with well differentiated endometrioid adenocarcinoma (G1), moderately differentiated endometrioid adenocarcinoma (G2), clear cell carcinoma (CCA) and nonatypical hyperplasia. AMH was not found in EC tissues in regularly menstruating women. An appropriately long mean period of breastfeeding in line with a prolonged period of hormonal activity had a positive effect on AMH expression. Our results may suggest that AMH is a factor which protects the organism against cancer, and should be further investigated as a potential prognosis marker and a therapeutic agent.


Introduction
Anti-Müllerian hormone (AMH) also known as Müllerian-Inhibiting Substance, is a well-studied regulatory molecule in reproductive functioning, especially in sexual differentiation during early were associations between the AMH expression levels in tissues of various endometrial cancer types in terms of comorbidities, tumor malignancy, stage, histological type and grade.
There was statistically significant differentiation of AMH protein expression between cancer types (Kruskal-Wallis ANOVA, H (7, N = 232) = 20.636, p = 0.004). Expression was observed in the tissues of (pre-cancerous state) nonatypical hyperplasia, G1 and G2 cancers with a good prognosis, and in clear cell carcinomas (CCA) with a generally poor prognosis; the highest expression was observed in the clear cell carcinomas, slightly lower in the case of nonatypical hyperplasia and the lowest in good-prognosis G1 and G2 cancers (   Table 1) (Kruskal-Wallis ANOVA, H (7, N = 232) = 20.636, p = 0.004); (b) in different clinical stages of endometrial cancer according to FIGO (International Federation of Gynecology and Obstetrics) 0-carcinoma in situ, IA -carcinoma limited to the inner lining of the uterus, IB-invasion less than half of the myometrium, IC-invasion equal to or more than half of the myometrium, IIA-invasion of the cervical glands, IIB -invasion of the cervical stroma, IIIA-involvement of the serosa or adnexa or both, IIIB-vaginal and/or parametrial involvement, IIIC-pelvic and/or paraaortic lymph node involvement IVA-Tumor invades bladder mucosa and/or bowel mucosa, IVB-Distant metastases (Kruskal-Wallis ANOVA, H (9, N = 231) = 12.819, p = 0.171); (c) group of patients without and with diabetes mellitus type 2 (Mann-Whitney U test, AMH: Z = 0.019, p = 0.985); (d) group of patients that used hormone replacement therapy (Wald-Wolfowitz runs test, Z = 2.240, p = 0.025). IRS-immunoreactive score of Remmele and Stegner; c.l.-confident limits.
The AMH protein was found in some stages in the clinical staging system of cancer according to FIGO (International Federation of Gynecology and Obstetrics) staging (Figure 2b). The AMH protein was absent in the tissues of II B, IIIA, and IIIB FIGO stages. There was no statistically significant differentiation between the mean expression of the AMH protein in the stages of cancer according to FIGO (Kruskal-Wallis ANOVA, H (9, N = 231) = 12.819, p = 0.171). Diabetes type 2 diagnosed before cancer did not affect the expression of AMH in EC tissues (Mann-Whitney U test, AMH: Z = 0.019, p = 0.985, Figure 2c). The presence of AMH protein was detected only in the group of patients who did not use hormonal replacement therapy (HRT; Figure 2d). Differences in AMH expression between both The AMH protein was found in some stages in the clinical staging system of cancer according to FIGO (International Federation of Gynecology and Obstetrics) staging ( Figure 2b). The AMH protein was absent in the tissues of II B, IIIA, and IIIB FIGO stages. There was no statistically significant differentiation between the mean expression of the AMH protein in the stages of cancer according to FIGO (Kruskal-Wallis ANOVA, H (9, N = 231) = 12.819, p = 0.171). Diabetes type 2 diagnosed before cancer did not affect the expression of AMH in EC tissues (Mann-Whitney U test, AMH: Z = 0.019, p = 0.985, Figure 2c). The presence of AMH protein was detected only in the group of patients who did not use hormonal replacement therapy (HRT; Figure 2d). Differences in AMH expression between both groups of women were statistically significant (Wald-Wolfowitz runs test, Z = 2.240, p = 0.025). AMH protein expression was observed in perimenopausal (o) and postmenopausal (m) women at similar levels ( Figure 3a). Expression was absent in women who had premenopausal cancer (p) (Figure 3a). There were no statistical differences in AMH expression levels between the three groups of women (Kruskal-Wallis ANOVA: H (2, N = 231) = 3.117, p = 0.210). Time from the first to the last menstrual bleeding of 40 or more years had an impact on the expression of AMH in EC tissues (Figure 3b), but the results were not statistically significant at the assumed level of I type error α = 0.05 (Mann-Whitney U test: Z = 1.854, p = 0.064). There were noticeable differences in the level of AMH protein expression depending on the presence or absence of arterial hypertension. Patients with arterial hypertension had a slightly higher expression of AMH ( Figure 3c). The results were statistically insignificant at the assumed level of error α = 0.05 (Mann-Whitney U test: Z = 1.880, p = 0.06). There were no significant relationships between the AMH expression and the number of childbirths (deliveries) (r s = 0.021, n = 232, p = 0.077, Figure 3d    It was found that the probability of AMH protein expression was significant depending on the average breastfeeding time (p = 0.004), the type of cancer (p = 0.006) and menstrual years (p = 0.045, Table 2). No significant interactions between the variables studied were found. The probability of AMH protein expression elevates with the increase in the average breastfeeding time, type of cancer (significantly increases the expression of type G2, G1 and CCA tumors) and depending on the years of menstruation (women with menstruation below 40 years had less expression of AMH, Table 3).  It was found that the probability of AMH protein expression was significant depending on the average breastfeeding time (p = 0.004), the type of cancer (p = 0.006) and menstrual years (p = 0.045, Table 2). No significant interactions between the variables studied were found. The probability of AMH protein expression elevates with the increase in the average breastfeeding time, type of cancer (significantly increases the expression of type G2, G1 and CCA tumors) and depending on the years of menstruation (women with menstruation below 40 years had less expression of AMH, Table 3).

Discussion
The normal levels of serum AMH in women between puberty and menopause amount to 1.4-5 ng/mL [15,59], and then it decreases to undetectable values [60]. The highest reported serum AMH concentration of 3205.93 ng/mL was found in a patient with sex cord tumor in whom remote metastases were present [50]. Sex cord tumors are rare and may also be analyzed in the context of Peutz-Jeghers syndrome [61]. Determination of serum AMH is used in diagnosing granulosa cell tumors [52]. A positive correlation was found between the AMH level and gross aggregate tumor mass determined by pathology, as well as between the AMH level and radiographic aggregate tumor mass [52]. In patients with this type of cancer, the serum AMH level reached 1200 ng/mL [52]. Serial measurement of serum AMH in granulosa cell tumor patients is performed in order to assess the efficacy of surgical treatment and to monitor possible relapses of the disease [62]. Elevated levels of AMH are observed also in patients with polycystic ovary syndrome (PCOS), where they are elevated up to 2-12 times [63,64]. This is connected to a higher number of small follicles in a group of these women [65]. Determination of serum AMH levels, apart from being a reliable assessment of the ovarian reserve, helps to individualize the dosing of follitropin alfa (rFSH) in artificial reproductive techniques (ART), which helps to reduce side effects of ovarian hyperstimulation [28,29,66].
AMH expression was found in mitosing cells of the endometrium of reproductive-age women [53]. Its expression increases in presence of both sex steroid hormones-progesterone and estradiol [53]. AMH derived from endometrium has the potential to elicit apoptosis and decrease viability of endometrial cells [53]. Serum AMH present in referential concentrations until menopause may exert protective action on female organisms and inhibit the development of EC. This was confirmed by the average age of disease onset at 62.5 years in the group of women analyzed in this study. The most important number of cases of this type of cancer was observed in postmenopausal women (80.49%), when AMH levels dropped to undetectable values. In our series, AMH protein was rarely found in the analyzed EC samples. Although the age alone did not have a significant impact on AMH expression in EC cells, our study confirmed that AMH protein was absent in women who developed EC at the premenopausal age (0/24). Another author did not find a correlation between serum AMH levels and diagnosing EC at the premenopausal age [47]. Possibly, AMH derived from both tissues-ovary and endometrium-may negatively influence EC development while working together. In analysis of EC etiology one should also consider other factors including obesity, hypertension and diabetes, as their correlation with EC has already been proven [67][68][69]. Although hypertension was related to elevated AMH expression in EC cells (statistically not significant), neither type 2-diabetes, nor BMI correlated with AMH expression. Due to increased levels of AMH in PCOS [63,64,70], a correlation between EC and PCOS is questionable [70]. This doubt is supported by epidemiological data-annually, some 4000 new cases of EC are diagnosed in the UK, while the number of PCOS patients in this country is estimated to be 500,000 to 1 million [70].
The widely known negative relationship between parity and EC indicates that multiparity is a factor protecting women against EC. In our study the presence of the AMH protein in EC cells did not correlate with the number of child births and birth weight of newborns. The AMH protein was detected only in the group of patients who did not use HRT. However, as the patients' history revealed, only 15 patients used HRT for more than 6 months. Yet, a correlation was observed between elevated AMH expression and the length of life hormonal activity, that is the time from the first to the last menstrual period. This phenomenon was not observed when the time of estrogen activity was shorter than 40 years. Among women who menstruated 40 years and more, 82.92% were patients with diagnosed G1 and G2 endometrioid carcinomas. In conjunction with AMH expression, this coincides with a good prognosis in the hormone-dependent type of cancer. Similarly, a longer average period of breastfeeding in conjunction with G1, G2 or CCA histopathological type is a factor increasing AMH expression. We found no differences between the stages of cancer according to FIGO and AMH expression. Although the protein was not detected in IIB, IIIA and IIIB stages, it was present in IIIC and IV stages. This observation seems to correlate with the determined elevated levels AMH concentration in cases of cancer which spread outside the uterus, in contrast to low levels of AMH in patients with cancer limited only to the uterus [71].
In analysis of the histopathological type of cancer, the probability of detecting AMH in cells increased in cases of nonatypical hyperplasia, as well as G1 and G2 endometrioid type of EC and, surprisingly in CCA. Patients with the CCA and positive AMH expression demonstrated IA-IB clinical stage of disease. None of them were obese (BMI 17.1-25.9) and their average age was 71.6 years. This justifies undertaking further research on AMH expression and a 5-year survival period in patients with clear cell carcinomas. The presence of AMH in type II cancers according to Bokhman's taxonomy might be the reason behind their biological diversity and a better than average survival rate in this type of cancer.
AMH is a natural substance which induces cell cycle arrest and apoptosis, with its activity limited to a few tissues. Thus, it was conceived that AMH represents a non-toxic substance which may be potentially useful in treating cancers exhibiting AMH receptors [8,16,[72][73][74].
The efficacy of controlling the development of mouse ovarian carcinoma (MOVCAR) cells was confirmed with recombinant human AMH, with no symptoms of toxicity during a 11-week treatment, equivalent to a continuous 7-year treatment in humans [75].
The neoplastic process in the endometrium engages some 1000 genes-362 up-regulated and 638 down-regulated ones [76]. It was shown that applying AMH in EC tissue changes the activity of 2688 genes engaged in regulating the cell cycle and apoptosis [77]. Expression grows in, among others, apoptotic protease activating factor-1 (APAF-1), β-catenin-interacting protein (ICAT), Rb related protein 130 (p130), while it decreases in, among others, cyclin-dependent kinase 2 (CDK2) and phospho-c-Jun [77]. Understanding the mechanisms leading to proapoptotic and cell cycle arrest functions of AMH is of key importance in order to use this substance as a therapeutic protein agent.
AMH uses various signaling pathways and inhibits cell divisions or programmed cell death in particular tissues in distinct ways. In the tissue of endometrial cysts of the ovary, AMH increases the concentration of p53-dependent p21 protein (cyclin-dependent kinase inhibitor-CDK inhibitor), as well as p107 and p130 proteins from Retinoblastoma family, while it decreases the level of transcription factor E2F1 [78]. AMH exhibits a similar activity in EC [43]. In the ovarian carcinoma, it increases the level of CDKs inhibitors: p16 and p21 [79]. Due to the activity of AMH, the levels of p16, p107 and p130 increase in cervical carcinoma cells [80].
The new classification of EC according to The Cancer Genome Atlas Research Network (TCGA) distinguishes four EC groups [81,82]. Group 4, labelled "copy-number high (serous-like)," encompasses cancers with the most serious prognosis [81,82]. Analyzing mutations from which the neoplastic process derives, it may be concluded that about 92% of cancers in this group present mutations of the p53 gene [81,82]. Theoretically, AMH-as an adjuvant-could increase the efficacy of treatment in the worst types of EC mediating the increased levels of p21.
In breast cancer and prostate cancer, AMH activates the pathway of NFκB (nuclear factor kappa-light-chain-enhancer of activated B cells), which induces the IEX gene (immediate early gene), encoding the protein regulator of the cell cycle [83][84][85]. It was demonstrated that in the T47D estrogen-positive line of breast cancer cells AMH causes selective expression of mRNA IEX-1S splice variant, and IEX-1L variant, which is responsible for the survival of colonies of cells, was absent [45,86]. AMH causes a similar response in the estrogen-negative line of breast cancer (MDA-MB-231) [38]. Transcripts of both IEX variants were determined, but only IEX-1S reached biologically significant levels, which resulted in 50% cell cycle arrest [45].
AMH sensitizes malignant ovarian cells to chemotherapy, increasing its efficacy [46,87,88]. Current precise onco-therapy encompasses blocking receptor tyrosine kinases (RTKs) [89,90]. The development of resistance in EC cells to modern medication relies on downregulation of PHLDA1 (pleckstrin homology-like domain family A member 1), a protein regulating apoptosis [91,92]. PHLDA1 expression is responsible for basal apoptosis, impedes the growth of neoplastic cells and sensitizes the neoplastic tissue to chemotherapeutics [93,94]. Activation of the NFκB pathway leads to upregulation of PHLDA1 [91]. It is not known whether the activation of this pathway by AMH has any effect on the level of PHLDA1. If it does, AMH could support the activity of RTKs antibodies and counteract the development of resistance towards them.
It was found that single nucleotides polymorphism (SNP) can be responsible for the development of cancers, including EC [95][96][97][98][99]. Overexpression of the murine double minute 2 gene (MDM2) leads to inhibiting the activity of p53 protein, which, in consequence, causes an increased risk of cancer [100][101][102]. Nucleotide 309 polymorphism (SNP309) in the first intron of the MDM2 gene (rs2279744) is a risk factor for EC among Caucasian and Asian women, as a result of an increased level of MDM2 [103,104]. Worth considering is whether applying AMH in the adjuvant EC therapy in cases of SNPT309G, in the presence of AMHERII, may increase the efficacy of treatment in this type of cancer.
Belonging to the TGFβ superfamily, AMH acts through signaling pathways related to the SMAD protein and engages slightly different mechanisms and transmitter proteins than the constitutive proteins of TGFβ [105][106][107][108]. These proteins activate a non-canonical signaling pathway which leads to activating proapoptotic p38/mitogen-activated protein kinase (p38/MAPK) [109,110]. The TGFβ superfamily does not show, however, an antiproliferative effect in the case of constitutive overexpression or transient overexpression of MDM2 [111,112]. It is not known whether AMH, using other pathways than the classic TGFβ proteins [87,113], is also lacking this function.
It is believed that epigenetic modulations have a growing role in the neoplastic process. There are two opposed systems which maintain or suppress the activity of genes through remodeling of chromatin: Polycomb (PcG) and Trithorax (TrxG) [114][115][116]. An important component of Polycomb which is responsible for the development of EC is an enhancer of zeste homolog 2 (EZH2), which impedes the activity of suppressor genes [117,118]. Knockdown of the EZH2 gene leads to apoptosis of EC cells because of an increase in the level of caspase-3 and caspase-9 [117]. AMH also increases the level of caspase-3 [87]. In such cases AMH could also find a therapeutic application.
Among natural substances, not only AMH has a beneficial effect on the reduction of EC cells. Hesperidin (a flavonoid from Citrus species) induces apoptosis through p38/mitogen-activated protein kinase [119]. Eupatilin (from Artemisia princemps) increases the level of the p21 protein and inhibits the growth of EC cells in the G2/M stage [120].
Summarizing, as the knowledge on carcinogenesis and its molecular basis increases, novel or modified therapeutic solutions appear in modern oncology. A member of the TGFβ family, AMH represents a substance which should be focused on in 21st century medicine because of its unique properties and safety profile.

Ethics Statement and Research Material
The experimental material was collected at the Clinical Ward of Gynecology, Obstetrics and Oncological Gynecology at the Regional Specialist Hospital in Olsztyn, Poland. The study protocol was approved by the Bioethics Committee of the Warmia-Mazury Medical Chamber (OIL.164/15/Bioet; 2 April 2015) in Olsztyn, Poland. Case history reviews were collected for all patients in order to record demographic details and their whole medical history. The classification of patients according to confirmed postoperative histopathological type is presented in Table 4. For each patient medical data such as, the type of cancer, cancer stages according to FIGO staging, hormonal status of women, menstrual activity, presence of type 2 diabetes, hypertension, use of HRT, number of births, time of breastfeeding, age of women and body mass index (BMI), were collected. According to the World Health Organization (https://www.who.int/topics/obesity/en/) the BMI values correspond to underweight (BMI < 18.5 kg/m 2 ), normal weight (BMI = 18.5-25), overweight (BMI = 25-30), and obese (BMI > 30) [121]. According to the British guidelines (National Collaborating Centre for Women's and Children's Health, 2015) the perimenopausal period is the time of irregular menstruations and vasomotor symptoms [122]. The menopausal period means that a woman has not had menstruation for at least 12 months and she does not use hormonal contraception [122].
The biopsy specimens were obtained from 232 patients during surgical interventions consisting of hysterectomy with bilateral salpingo-oophorectomy, lymphadenectomy (except pre-cancerous states of endometrium (PCS) cases) and peritoneal washing. The surgical biopsies from the affected area were preserved in 4% buffered formalin/formaldehyde immediately after surgical removal in the form of a phosphate-buffered solution (Chempur, PiekaryŚląskie, Poland). The volume of fixative to tissue ratio was at least 10:1. The specimens were fixed by immersion for 6 to 12 h before further processing. The preserved tissues were placed in cassettes and the batches of specimens were loaded onto a tissue processor (Leica ASP 300S, Leica Biosystems, Nussloch, Germany) before being molten in wax. Embedding was performed 10 h later on the platform (Leica EG1160). Paraffin blocks were cut Leica SM 2000R microtome into 6 µm-thick sections. Following in time flattening and straightening on the surface of warm water (37 • C) they were picked up onto microscope slides. After thorough drying (overnight/37 • C) specimens were stained in the hematoxylin and eosin (H&E) automated stainer (Leica ST5020) for visualization of particular structures. After staining, the sections were covered with a glass coverslip and evaluated by a qualified pathologist. The Formalin-Fixed Paraffin-Embedded (FFPE) specimens were processed according to the College of American Pathologists (CAP) criteria described in The Practical Guide to Specimen Handling in Surgical Pathology [123].

Tissue Microarrays (TMAs)
The prepared slides were scanned using the Panoramic MIDI II (3DHistech, Budapest, Hungary) histological scanner. Panoramic Viewer (3DHistech) software was used to select manually three representative areas (each with a surface of 1.5 mm 2 ) from regions of EC previously indicated by the pathologist. Three representative cores of 1.5 diameter mm were taken from each archival tumor sample of an EC and embedded in paraffin as described above to create tissue microarrays (TMAs) using TMA Grand Master (3DHistech, Budapest, Hungary) in line with the manufacturer's instructions.

Evaluation of IHC Reactions
All immunohistochemical reactions were evaluated by two pathologists using a BX-41 light microscope. For evaluation an immunoreactive score of Remmele and Stegner was applied [124] ( Table 5).

Statistical Analysis
The expression of AMH was measured with the rank scale IRS immunoreactive score (IRS) of Remmele and Stegner (Table 5) in three samples retrieved from three sites of cancer tissue from each patient. There were no differences in the overall AMH expression in the three collected tissue samples (F-test with Greenhouse-Geisser correction, ἐ = 0.714, p = 0.501). The mean value of AMH expression for each patient was computed, and the mean values were used in the whole analysis. Differences in AMH expression between the type of cancer, cancer stages according to FIGO, hormonal status of women were tested with the Kruskal-Wallis test. Results of comparisons between individual groups were based on the post hoc nonparametric multiple comparison tests. The comparison of AMH expression in two groups of patients, e.g., years of menstrual activity (<40 years old, 40 or more years old), presence of type 2 diabetes, presence of hypertension, the use of hormonal replacement therapy (HRT), was performed with the Mann-Whitney U test or the Wald-Wolfowitz runs test. The relationship between AMH expression and values of metric traits (number of births, time of breastfeeding, BMI, age of women) was tested using Spearman's rank correlation coefficient. The multidimensional comparison of AMH expression (binary data: expression positive-1, and negative -0) and an independent category and continuous variables used in the study were tested in Generalized Linear Model (GLZ) with logit linking function and searching for the best fitted models with the Akaike criterion. The parameters were estimated using the highest likelihood method. In the case of variable redundancy, the parameter was not estimated. The p-value < 0.05 was defined as statistically significant. Statistical analysis was performed using Statistica 13.0 software (TIBCO Software Inc. 2017. Statistica (data analysis software system), version 13. http://statistica.io, Krakow, Poland)).

Conflicts of Interest:
The authors declare no conflict of interest.

MDPI Multidisciplinary Digital Publishing Institute DOAJ
Directory of open access journals TLA Three letter acronym LD linear dichroism