The Role of Gonadotropin-Releasing Hormone (GnRH) in Endometrial Cancer

Endometrial cancer (EC) is one of the most common gynecological malignancies. Gonadotropin releasing hormone (GnRH) is a decapeptide first described to be secreted by the hypothalamus to regulate pituitary gonadotropin secretion. In this systematic review, we analyze and summarize the data indicating that most EC express GnRH and its receptor (GnRH-R) as part of an autocrine system regulating proliferation, the cell cycle, and apoptosis. We analyze the available data on the expression and function of GnRH-II, its putative receptor, and its signal transduction. GnRH-I and GnRH-II agonists, and antagonists as well as cytotoxic GnRH-I analogs, have been shown to inhibit proliferation and to induce apoptosis in human EC cell lines in pre-clinical models. Treatment with conventional doses of GnRH-agonists that suppress pituitary gonadotropin secretion and ovarian estrogen production has become part of fertility preserving therapy of early EC or its pre-cancer (atypical endometrial hyperplasia). Conventional doses of GnRH-agonists had marginal activity in advanced or recurrent EC. Higher doses or more potent analogs including GnRH-II antagonists have not yet been used clinically. The cytotoxic GnRH-analog Zoptarelin Doxorubicin has shown encouraging activity in a phase II trial in patients with advanced or recurrent EC, which expressed GnRH-R. In a phase III trial in patients with EC of unknown GnRH-R expression, the cytotoxic GnRH doxorubicin conjugate was not superior to free doxorubicin. Further well-designed clinical trials exploiting the GnRH-system in EC might be useful.


Introduction
Endometrial cancer (EC), derived from the epithelial lining of the uterine cavity, is one of the most common female cancers. Worldwide, 382,069 new cases were diagnosed in 2018 and 89,929 women died of this disease [1]. In Western Europe and North America, EC is the fourth most common malignancy in women and the most common malignant tumor of the female genital organs [2].
The prognosis of EC is rather favorable, as about 75% of cases are diagnosed in an early stage and can be cured by surgery and in more advanced cases by additional radiotherapy and/or chemotherapy [2] resulting in a general five-year cancer-specific survival of around 80% for all stages and histological types [1,2]. The majority of EC (85%) known as the so-called type 1 cancers develop due to prolonged exposure to endogenous or exogenous estrogens in the absence of sufficient progestogen activity [3][4][5]. These type 1 EC are hormone-dependent and can be treated in early stages without surgery by endocrine manipulation, including estrogen withdrawal and/or high dose progestogens in premenopausal women who have not yet completed their families. Later in their development, these hormone-dependent EC dedifferentiate, lose expression of estrogen receptors and/or progesterone receptors, and are no longer amenable to estrogen withdrawal, anti-estrogens, or progestogens [2]. Type 2 EC do not express estrogen receptors or progesterone receptors, and 29 fmol/10 6 cells, respectively [9]. Petersen et al., using radiolabeled native GnRH or [D-Lys 6 ]-GnRH, failed to detect GnRH-R in the HEC-1A cell line [10]. Using [ 3 H] GnRH as a ligand, Imai et al. detected GnRH-R in RL 95-2 and HHUA EC cancer lines (Kd 2.38 ± 0.86 nM) and in 16 of 18 well differentiated and 4 of 7 poorly differentiated EC specimens (Kd 5.89 ± 3.59 nM). Bmax = 1.8 ± 0.95 pmol/mg membrane protein [11]. In addition, they found the mRNA for the GnRH-R in both cell lines and all EC samples where specific binding sites were detected.
Chatzaki et al. found low levels of GnRH-R mRNA transcripts but no high affinity GnRH binding in Ishikawa and HEC-1A cells and low affinity binding in three of eight EC tissues [12]. Shibata and colleagues showed the expression of mRNA for GnRH-R in HHUA human EC cell line [13]. Borri et al. failed to detect the expression of mRNA for the GnRH-R in HEC-1A and HEC-1B cell lines by Northern blot analysis and RT-PCR [14]. Ohta et al. detected mRNA for GnRH-R in human EC cell lines Ishikawa, EIIL, HEC1A, HEC6, HEC50, and HEC59 [15]. Kim et al. demonstrated the expression of mRNA for GnRH-R in human EC cell line CUME-1 [16]. Noci et al. using RT-PCR detected GnRH-R in 5/7 EC specimens [17].
We sequenced the complementary DNA of the GnRH-R from position 31 to position 1204, covering the complete coding region (position 56 to position 1041) in Ishikawa and HEC-1A EC cell lines. There were neither mutations nor splice variants of the GnRH-R transcript [18]. We detected high affinity GnRH binding sites (Kd 1.5 to 8.2 nM) and mRNA for the GnRH-R in EC cell lines HEC-1B, KLE, and AN-3-CA. In the endometrial cancer cell line, MFE-296 GnRH-R mRNA was found but no specific binding of GnRH detected [19]. In five EC biopsy samples including type 2 cancers, we detected specific high-affinity/lowcapacity GnRH binding sites as well as the mRNA for the GnRH-R. In another Type 2 EC, neither GnRH-R nor their mRNA were found [19]. Using RT-PCR, Nagai et al. detected GnRH-R expression in the HEC-1A cell line and in 36 of 38 (95%) EC samples as well as in two samples of atypical endometrial hyperplasia (AEH), which is a precancer for EC [20]. Yang et al. found high affinity/low capacity binding sites in Ishikawa cells [21]. Engel et al. detected GnRH-R in HEC-1A and RL-95-2 EC cell lines by RT-PCR, Western blot analysis, and radio-ligand binding assays [22]. Using a monoclonal GnRH-R antibody, Jeon et al. analyzed paraffin-embedded tissue blocks from 141 endometrial cancer patients with a 10% cut off level for GnRH-R positivity. Seventy specimens (49.6%) stained as GnRH-R positive. Tumor grade, histio-type, and disease stage were not associated with GnRH-R positivity [23]. Jankowska et al. detected GnRH-R in 9 of 12 well differentiated EC using RT-PCR and immunohistochemistry [24]. In the context of a multicenter phase 2 clinical trial, we analyzed tumor samples from 48 patients with recurrent or metastatic EC for the expression of GnRH-R using immunohistochemistry. Forty-four (92.4%) of these EC were GnRH-R positive (range of positive cells: 30-90%, median: 70%, mean: 69 ± 21%) [25]. Ten of the 44 GnRH-R positive tumors were type 2 EC (serous or clear cell) [25]. The expression of GnRH-R in EC cell lines Ishikawa and ECC as well as in three EC tissue samples was demonstrated by Wu et al. using immunohistochemistry and knockdown with small interfering RNA [26,27]. Hao et al. found 16 of 20 cases of EC samples (80%) positive for GnRH-R by immunohistochemistry [28].

Expression of GnRH in EC
Using a specific antibody, we detected GnRH-immunoactivity in extracts of Ishikawa (426 ± 84 fmol/10 6 cells) and HEC-1A (368 ± 41 fmol/10 6 cells) EC cell lines [29]. GnRH bioactivity of these extracts was assessed in rat pituitary cells in the culture. The release of luteinizing hormone (LH) induced by these samples corresponded dose-dependently to that obtained with comparable amounts of authentic GnRH. The expression of mRNA for GnRH by Ishikawa and HEC-1A EC cells was demonstrated by RT-PCR, a restriction enzyme-, and Southern blot analysis [29]. Chatzaki et al. detected mRNA for GnRH in Ishikawa and HEC-1A cell lines, but not secretion of immunoreactive GnRH. In 1 of 10 primary cultures of EC tissues, GnRH mRNA and immunoreactivity were found [12]. Furui et al. detected GnRH (0.13 ± 0.074 pg/mg protein) in 12 EC samples using a radioimmunoassay and a high-performance liquid chromatography. Nine of these EC samples expressed GnRH-R [30]. Jankowska et al. demonstrated the presence of GnRH in 12 of 12 EC samples by means of RT-PCR and immunohistochemistry [24].

Effects of GnRH, GnRH-Agonists, and GnRH-Antagonists on the Proliferation of EC Cells
The proliferation of Ishikawa and HEC-1A EC cell lines was time and dose dependently inhibited by the GnRH agonist [D-Trp 6 ] GnRH (Triptorelin) and the GnRHantagonist Cetrorelix [9]. The anti-proliferative effects of both analogs were significant in the Ishikawa cell line at 10 pM (86%) and 1 nM in the HEC-1A cell line (80%) and maximal at 10 µM: Ishikawa 52% and HEC-1A 62% of the respective controls [9].
Kleinman et al. found that the GnRH-antagonist Cetrorelix inhibited basal and insulin like growth factor-I (IGF-I) induced growth of Ishikawa cells [31]. Peterson et al. observed no inhibition of growth of HEC-1A cells hetero-transplanted into nude mice by treatment with GnRH agonist leuprorelin [10]. Chatzaki et al. observed no effects of GnRH agonists triptorelin and goserelin on in vitro proliferation of Ishikawa and HEC-1A cells [12]. We found that GnRH agonist triptorelin dose-dependently reduced basal and epidermal growth factor (EGF) induced proliferation of HEC-1A EC cancer cell line under serumfree and phenol red-free conditions [32]. Sica and colleagues found the GnRH agonist triptorelin and leuprorelin to be ineffective on basal growth of the Ishikawa cell line but to counteract and even suppress estrogen-induced proliferation [33]. Shibata et al. reduced the proliferation of the HHUA human EC cell line by treatment with the GnRH agonist buserelin to 60% of control [13]. Borri et al. described significant anti-proliferative effects of GnRH agonist leuprorelin and of GnRH antagonist antide in HEC-1A and HEC-1B EC cell lines. [14]. Ohta et al. demonstrated an inhibition of proliferation by Triptorelin in human EC cell lines Ishikawa, EIIL, HEC1A, 6, 50, and 59 [15]. Noci and colleagues treated primary cultures from nine human EC for 96 h with 1 µM of GnRH agonist leuprorelin or GnRH antagonist cetrorelix and observed median growth inhibitions of 14.3% (leuprorelin, range 46.4-2.4%) and 27.5% (cetrorelix, 60.9-22.7%) [17]. Nagai et al. found a significant inhibition of proliferation of HEC-1A cells after treatment with 1 µM leuprorelin [20].
Zhao and colleagues observed a dose-dependent inhibition of proliferation of the HEC-1A cell line by triptorelin [34]. Park et al. described a dose-dependent inhibition of proliferation of HEC-1A EC cells with 10 and 100 nM native GnRH [35]. In ovarian cancer cell lines EFO-21 and EFO-27 that had been shown to express high affinity GnRH-R and to secrete GnRH, treatment with an antiserum to GnRH in neutralizing concentrations significantly increased proliferation by 21%, while treatment with native GnRH even at very low concentrations (1 pM) had no effect or an anti-proliferative effect. At higher concentrations of GnRH (≥10 pM), significant inhibition of proliferation was observed [36].

Signal Transduction and Intracellular Actions of GnRH in EC
Kleinman et al. demonstrated inhibitory effects of GnRH antagonist cetrorelix on basal and Insulin like growth factor (IGF)-induced growth of Ishikawa EC cells. This GnRH analog also completely inhibited IGF-II release from these cells [31]. Estradiol (E 2 ) partially abolished the inhibitory effects of cetrorelix on the proliferation of Ishikawa cells. The growth of HEC-1A cells was not influenced by the antagonist. The GnRH agonist buserelin had no significant effect on both cell lines. The growth inhibitory effects of cetrorelix on the Ishikawa cell line was associated with an induction of apoptosis [37].
Our group showed that, in HEC-1A and Ishikawa EC cell lines, the GnRH-agonist triptorelin had no effects on GnRH-signaling mechanisms known from the pituitary gonadotrophs, i.e., phospholipase C, protein kinase C, and adenylate cyclase [38]. The proliferation of HEC-1A cells in serum/phenol red free medium was significantly increased by epidermal growth factor (EGF). This mitogenic effect of EGF was dose-dependently reduced by triptorelin. Expression of EGF-receptors was not influenced. However, net tyrosine phosphorylation induced by EGF (1 nM) was nearly completely abrogated by the addition of 10 µM triptorelin or pre-incubation with 100 nM of the GnRH agonist for 48 h. This inhibitory effect of triptorelin on EGF-induced net tyrosine phosphorylation was partly antagonized by sodium vanadate, which is an inhibitor of phosphotyrosine phosphatase. The EGF-induced activation of mitogen-activated protein kinase/extracellular signal related kinase (MAP-kinase/ERK) was completely suppressed in HEC-1A cells by pretreatment with triptorelin [32]. We could show that, in Ishikawa and HEC-1A cell lines, EGF-induced c-fos expression was dose-dependently reduced to base line levels by 100 nM of GnRH agonist triptorelin or antagonist cetrorelix [38]. Both the G protein αi and G protein αq were expressed in Ishikawa and HEC-1A cell lines and coupled to GnRH-R. Inhibition of EGF-induced c-fos expression through GnRH was, however, mediated through pertussis toxin (PTX) sensitive G protein αi. GnRH agonist triptorelin antagonized PTX-catalyzed APD-ribosylation of G protein αi. In HEC-1A and Ishikawa cells, triptorelin stimulated phospho-tyrosine phosphatase activity, which is an effect blocked by PTX. EGF induced autophosphorylation of EGF receptors was reduced by 45-63% after GnRH-agonist (100 nM) treatment.
This inhibitory effect of GnRH was abrogated using the phosphotyrosine phosphatase (PTP) inhibitor vanadate [18] (Figure 1). In addition, GnRH-analogs activated nucleus factor kappa-B, protecting the cancer cells from apoptosis and activator protein-1 (AP-1) through PTX-sensitive G protein αi and expression of c-Jun and activation of c-Jun phosphorylation [39] (Figure 1). The proliferation of estrogen receptor α (ERα) and estrogen receptor β (ERβ) positive EC cell lines (KLE and HEC-1B) but not of ERα negative/ERβ positive cell lines (Ishikawa, HEC-1A) was significantly stimulated by treatment with E 2 (10 nM). This mitogenic effect of E 2 was dose-dependently inhibited by triptorelin. E 2 induced activation of the serum response element (SRE) and expression of the immediate early response gene c-fos was blocked by the GnRH agonist [40] (Figure 1).
In our hands, native GnRH, GnRH-agonist triptorelin, and GnRH antagonist cetrorelix reduced apoptosis in the human EC cell line Ishikawa, induced by the cytotoxic agent doxorubicin and by UV light through activation of nucleus factor-κB (NFκB) [41]. Doxorubicin induced apoptosis in HEC-1A and Ishikawa human EC cell lines were markedly increased by knockdown of GnRH-R expression [42]. In ovarian cancer cell lines EFO-21 and EFO-27 that express GnRH and GnRH-R, treatment with GnRH-agonist triptorelin (100 nM) resulted in an increase of cells in the G 0/1 phase of the cell cycle and a decrease in the G 2/s phase. DNA synthesis was decreased to 45% of the control, likely meditated through JunD and AP-1 [43].
Takagi et al. demonstrated the presence of Gαi and Gαs proteins and the activation of phosphatase activity in plasma membranes from EC samples expressing GnRH-R [44]. Imai et al. showed that GnRH agonist induced protein dephosphorylation in plasma membranes from EC samples expressing GnRH-R. This effect was dose-dependently and completely inhibited by pertussis toxin [45]. Later, this group provided evidence that the GnRH-agonist leuprorelin stimulated intra-tumoral expression of the apoptosis-inducing Fas ligand in GnRH-R bearing EC. At 10 µM concentration, the GnRH agonist induced up to 90% reduction in the cell number preceded by Fas ligand production [46]. In HEC-1A EC cells and 36 human EC samples expressing GnRH-R mRNA, telomerase activity was not significantly altered by GnRH-agonist treatment. The expression of mRNA of human telomerase transferase, however, was markedly decreased [20]. Chatzaki and colleagues found no activation of GnRH-signal transduction mechanisms, known from pituitary gonadotropes including total inositol phosphate increase, cyclic AMP-production, and cytosolic Ca 2+ after GnRH or GnRH agonist treatment of Ishikawa and HEC-1A cell lines and EC tissues [12]. Shibata et al. found that inhibition of proliferation of the HHUA EC cell line by the GnRH-agonist was mediated through activation of protein kinase C and subsequent increase of annexin V, which is a Ca 2+ -dependent phospholipid binding protein [13]. Borri et al. showed clear anti-proliferative effects of GnRH agonist leuprorelin and GnRH antagonist antide in human EC cell lines HEC-1, inhibitory effects on [ 3 H] thymidine incorporation, and a slight accumulation of cells in the Go/G1 phase, but no effect on intracellular free calcium concentration [14]. Ohta et al. showed that GnRH-a inhibited telomerase activity but not mRNA expression for telomerase in six human EC cell lines [15]. Kim and colleagues demonstrated a cell cycle arrest at the G 1 -S transition of CUME-1 EC cells after treatment with GnRH agonist triptorelin [16]. Sica and coworkers found that two GnRH-agonists counteracted or even suppressed estrogen stimulated growth of Ishikawa cells and prevented estrogen-induced decrease in the level of estrogen receptors [33]. In our hands, native GnRH, GnRH-agonist triptorelin, and GnRH antagonist cetrorelix reduced apoptosis in the human EC cell line Ishikawa, induced by the cytotoxic agent doxorubicin and by UV light through activation of nucleus factor-κB (NFκB) [41]. Doxorubicin induced apoptosis in HEC-1A and Ishikawa human EC cell lines were markedly increased by knockdown of GnRH-R expression [42]. In ovarian cancer cell lines EFO-21 and EFO-27 that express GnRH and GnRH-R, treatment with GnRH-agonist triptorelin (100 nM) resulted in an increase of cells in the G0/1 phase of the cell cycle and a decrease in the G2/s phase. DNA synthesis was decreased to 45% of the control, likely meditated through JunD and AP-1 [43].
Park et al. showed that GnRH treatment induces integrin β3 expression and activation of focal adhesion kinase (FAK) and an activation of ERK 1/2 and p38MAPK [35]. Zhao et al. demonstrated that GnRH-agonist triptorelin inhibits proliferation of HEC-1A cells and induced apoptosis in a dose-dependent manner. This effect was augmented, when phosphatase and tensin homolog gene (PTEN) was knocked down [34]. Öztürk et al. found anti-proliferative effects of GnRH agonist leuprorelin and GnRH antagonist ganirelix in three primary EC cell lines associated with an increase of apoptotic cells [47].
The GnRH metabolite GnRH-(1-5), resulting from cleavage of the native decapeptide at the Tyr 5 -Gly 6 bond, has autonomous mechanisms of action including stimulation of proliferation of the Ishikawa EC cell line by increasing EGF release and the phosphorylation of EGF receptors. This effect is likely mediated through the orphan G protein-coupled receptor 101 (GPR 101) [48,49]. Using the peptidomimetic GnRH-antagonist AEZS-115, Engel et al. could show a marked time-dependent and dose-dependent inhibition of human EC cell lines HEC-1A and Ishikawa growth that expressed mRNA for the GnRH-R. They concluded, however, that the antitumor activity of AEZS-115 was not mediated through the GnRH-R in the tumor cells [50].

GnRH-II and GnRH-II Receptors in EC
Apart from the well-known decapeptide GnRH-I, a second form, GnRH-II, was found to be universally present and preserved from jawed fish to humans [51]. Millar et al. cloned a type II GnRH-R from the marmoset monkey, which is highly specific for GnRH-II and found to be expressed in human tissues [52]. Neill et al. cloned the GnRH-II-R from the rhesus monkey [53]. The type II GnRH-R has a C-terminal and, in contrast to the type I GnRH-R, is rapidly internalized. In addition, GnRH-I antagonists had agonistic effects on the type II receptor [17]. We detected the expression of the mRNA for GnRH-II-R in HEC-1A and Ishikawa human EC cell lines using reverse transcriptase-polymerase chain reaction (RT-PCR) and Southern blot analysis. The proliferation of these cells was dose-dependently and time-dependently reduced by native GnRH-II, which was significantly more effective than GnRH-I agonist triptorelin. In the ovarian cancer cell line SK-OV3, that expresses GnRH-II-R but not GnRH-I-R, native GnRH-II but not GnRH-I agonist triptorelin had anti-proliferative effects [54]. GnRH agonist triptorelin, GnRH-antagonist cetrorelix, and native GnRH-II had anti-proliferative effects on human EC cell lines Ishikawa, HEC-1A, and HEC-1B that express receptors for GnRH-I and mRNA for receptors for GnRH-II. When the receptor for GnRH-1 was knocked out, the anti-proliferative effect of triptorelin was abrogated, while those of cetrorelix and GnRH-II persisted [55]. Since a functional human GnRH-II-R transcript has not been found so far, the mRNA detected by us is suspected to be non-functional because of a stop-codon in the sequence [24]. Using a specific anti-serum to human GnRH-II-R and Western blot analysis, we detected the expected 54-kDa band in ovaries from the marmoset monkey, where the expression of a GnRH-II-R had been shown. In human EC cell lines, Ishikawa and HEC-1A, as well as in human placenta, we obtained a 43-kDa band, which might represent a truncated, but functional 5 transmembrane domain human GnRH-II-R. Using a photo affinity labelling technique, with a 125 I labelled GnRH-II analog, we obtained a band of 43 kDa. GnRH-I antagonist cetrorelix and GnRH-II but not GnRH-I agonist triptorelin were able to displace this binding [56]. In the human EC cell line HEC-1A, EGF induced autophosphorylation of the EGF-receptor, activation of MAP-kinase, and expression of the immediate early gene c-fos were inhibited by treatment with the GnRH-II agonist [D-Lys 6 ] GnRH-II. After knock out of the GnRH-I-R expression, these effects of [D-Lys 6 ] GnRH-II persisted [57].
Neither GnRH-I and GnRH-I-agonists nor GnRH-I-antagonists induced apoptosis in endometrial cancer cell lines. Antagonists of GnRH-II, however, resulted in apoptotic cell death via dose-dependent activation of caspase-3. In nude mice, GnRH-II markedly inhibited growth of xenotransplants of the human EC cell line HEC-1B [58]. GnRH-II antagonists were shown to activate apoptosis in human endometrial cancer cells through GnRH-I-R via activation of stress-induced mitogen-activated protein kinases p38 and c-Jun NH 2 terminal kinase, leading to activation of pro-apoptotic protein Bax [59].
Kim and colleagues found that GnRH-II stimulates stress activated p38MAPK in an ovarian cancer cell line OVCAR-3, leading to an inhibition of proliferations and an increase of apoptosis [60].
Wu et al. inhibited proliferation and induced apoptosis in the human EC cell line Ishikawa by treatment with GnRH-II and provided evidence that this effect is mediated through the GnRH-I-R, activation of the ERK1/2 pathways, and induction of GADD45αsignalling [26]. Park et al. found that the inhibition of growth in the human EC cell line HEC-1A achieved by treatment with GnRH-I or GnRH-II is mediated through activation of integrin beta-3, focal adhesion kinase (FAK), and ERK1/2 and p38MAPK pathways. The effects of GnRH-II were more pronounced than those of GnRH-I [35]. Wu et al. found that, in Ishikawa and ECC-1 human EC cell lines, a GnRH-II agonist promoted cell motility through the GnRH-I-R and subsequent phosphorylation of ERK1/2 and JNK leading to activation of matrix metalloproteinase (MMP)-2 [27].

GnRH Receptor in EC as a Target for Cytotoxic Therapy
Apart from pituitary gonadotropes and reproductive organs, most other tissues and hematopoietic stem cells do not express the GnRH-R [61]. In almost all cases of EC, ovaries, fallopian tubes, and the uterus are surgically removed during primary treatment. The GnRH-R in tumor residues or metastases of EC could, therefore, be used to deliver a targeted therapy with improved anti-tumor efficacy and fewer side effects [61,62]. A number of GnRH-analogs covalently linked to a cytotoxic agent have been developed that bind specifically and with high affinity to GnRH-R in tumors including EC [61][62][63]. After internalization of the ligand receptor complex, the cytotoxic moiety is liberated and acts within the tumor cells [61][62][63]. We could show that the cytotoxic GnRH-agonist Zoptarelin Doxorubicin (AEZS-108, AN-152), in which the GnRH-agonist [D-Lys 6 ] GnRH is covalently coupled to the cytotoxic agent doxorubicin selectively binds to GnRH-R in human EC cell lines. This is internalized and results in accumulation of doxorubicin in the nuclei of these cells. No binding and uptake of Zoptarelin Doxorubicin was detected in cell lines that did not express GnRH-R or in the presence of an excess of a GnRH-agonist competitively blocking the receptors [63]. Proliferation of cells was more potently inhibited by Zoptarelin Doxorubicin than by equimolar concentrations of free doxorubicin in cell lines expressing GnRH-R [63]. In nude mice, bearing xenografts of human EC cell line HEC-1B, tumor volumes were significantly reduced by treatment with Zoptarelin Doxorubicin, while treatment with equimolar concentrations of free doxorubicin only arrested tumor growth, but did not reduce their volume. No toxic side effects were observed with Zoptarelin Doxorubicin treatment, while free doxorubicin caused loss in body weight and death in the animals. The growth of tumors not expressing GnRH-R was not affected by treatment with Zoptarelin Doxorubicin [61]. In human EC cell lines, HEC-1A and Ishikawa Zoptarelin Doxorubicin induced apoptosis significantly more potently than doxorubicin without activating the multi-drug resistance 1(MDR-1) system [64].
Nechushtan et al. constructed a chimeric toxin based on GnRH and Pseudomonas exotoxin that bound specifically to human EC cell primary cultures and induced marked cell death [65]. Palyi and colleagues induced apoptosis in Ishikawa cells by treatment with conjugates of GnRH III-analogs and a copolymer [66]. Engel et al. showed that HEC-1A and RL-95-2 human EC cell lines express functional GnRH-R. Athymic mice bearing xenografts of these cell lines were treated with Zoptarelin Doxorubicin or another cytotoxic GnRHagonist (AN-207). Tumor growth was significantly suppressed. The cytotoxic radicals (doxorubicin or 2-pyrrolinodoxorubicin) had no effects [22]. Yang et al. demonstrated that a conjugate of [D-Lys 6 ] GnRH-Pro 9 -ethymamide with Pokeweed antiviral protein (PAP) was cytotoxic to the EC cell line Ishikawa. PAP alone, that cannot enter cells, had no effects. The cytotoxic activity of the GnRH-PAP conjugate was inhibited by an excess of a GnRH-agonist [21].
Li and colleagues developed a GnRH analog LHRH' that binds with high affinity to GnRH-R but does not interfere with gonadotropin secretion. LHRH' coupled to cecropin B, which is an antimicrobial peptide, significantly inhibited cell viability of EC cells HEC-1A but not of normal eukaryotic cells [67].
Our group developed a GnRH-R targeted gene therapy system for gynecological tumors, including EC that worked in vitro and in athymic mice bearing xenografts of HEC-1B human RC cells [68].

Clinical Applications of GnRH-Analogs in EC and Atypical Endometrial Hyperplasia (AEH)
GnRH-agonists and antagonists have been used in the treatment of EC and its precancer AEH based on two rationales: 1. Suppression of pituitary gonadotropin secretion leading to suppression of ovarian function including estrogen and progesterone production, which is a state called reversible medical castration [6]. This strategy has been applied for the conservative therapy of young women with AEH and early EC wishing to preserve their fertility. 2. Exploitation of the direct anti-tumor effects of GnRH-analogs on GnRH-R on the tumor cells. A third approach has been the application of cytotoxic GnRH-analogs for targeted chemotherapy of tumors expressing GnRH-R.

Suppression of Ovarian Steroid Production by GnRH-Analogs as a Treatment for AEH and Early EC
Pérez-Medina et al. treated 10 patients with AEH with the progestogen norethisterone acetate for three months and monthly injections of Triptorelin for six months. At a five-year follow up, regression was observed in 84.2% of patients, persistence in 5% of patients, recurrence in 5% of patients, and progression in 5% of patients [69]. Minig et al. treated 20 patients with AEH and 14 with early EC with a levonorgestrel-intrauterine device (LNG-IUD) for one year plus the GnRH-analog for six months. They observed a complete response in 95% of patients with AEH and 57.1% in women with EC G1. Progression was observed in 5% (AEH) and in 28% (EC) of patients, respectively. Four of 20 patients with AEH and 2 of 24 with early EC had recurrences. Nine women had 11 spontaneous pregnancies [70]. Pashov and colleagues treated 13 women with AEH with leuprorelin acetate (six monthly injections) and an LNG-IUD Eleven women with stage IA well differentiated EC were treated with nine injections of GnRH and a LNG-IUD for at least 12 months [71]. They reported that the therapy was effective for all patients. Three women with EC became pregnant. Pronin et al., prospectively, treated 32 women with G1 EC with LNG-IUD and a GnRH agonist for at least six months. Twenty-three (72%) of the patients had a complete remission (CR). Two had a recurrence after the complete response [72]. Zhou et al. treated 29 patients with well-differentiated EC or AEH with monthly injections of a depot GnRH agonist combined with a LNG-IUD or the aromatase inhibitor letrozole (2.5 mg/day). 88.2% of the EC patients and 12 women (100%) in the AEH group had a complete response. 8% with AEH and 6% of women with EC had recurrence after CR [73]. Zhang et al. treated six young obese patients with EC with monthly injections of depot GnRH-agonist and aromatase inhibitor. The CR rate was 100%. After a median follow-up of four years, no recurrence was observed. The pregnancy rate was 50% and the life birth rate was 75% [74]. Tock et al., retrospectively, analyzed 18 women younger than 41 years with grade 1 EC (39%) or AEH (50%), who received endometrial resection and, subsequently, a GnRH agonist for three months. They observed a CR in 12 patients (67%) and three relapses (25%) after CR. Eight women had 14 pregnancies [75].
Analyzing the data from 127 patients with AEH (75%) or early EC (25%) after different fertility sparing treatments, Yin et al. found that the treatment with a GnRH agonist with or without an LNG-IUD/aromatase inhibitor might be preferable for obese patients, resulting in fewer relapses after primary CR [76] (Table 1).

Clinical Studies on Direct Anti-Tumor Effects of GnRH-Analogs in EC
Gallagher et al. treated 17 patients with EC that had recurred after surgery, radiotherapy, and progesterone treatment with monthly injections of a GnRH agonist, achieving an objective remission in six patients (35%). This continued for a median of 20 months with no adverse effects [77]. Jeyarajah et al. reported on these 17 patients and an additional 15 patients with recurrent EC with progressive, symptomatic, and measurable disease, treated with monthly injections of Leuprorelin or Goserelin. An objective response was obtained in 9 of 32 patients (28%) including pelvic as well as distant sites of recurrence. The response rate did not correlate with differentiation of the tumors (15 patients with G3 EC). Duration of responses was 4 to 86 months [78].  [79]. Lhommé et al. treated 24 eligible patients with advanced or recurrent EC with monthly injections of the GnRH agonist triptorelin depot. They observed one complete and one partial response (response rate 8.7%) and five disease stabilizations, often of long duration. Toxicity was very low [80]. Noci et al. reported on a patient with EC where a surgical treatment was not possible due to comorbidities. The tumor expressed mRNA for GnRH-R. Under treatment with leuprorelin, she remained stable for six years [81]. Asbury et al. treated 40 patients with advanced or recurrent EC with monthly s.c. implants of the GnRH agonist goserelin. They observed two complete and three partial remissions (12%). No severe or life threatening toxicities occurred because of the GnRH agonist [82]. (Table 2)

Clinical Studies on Cytotoxic GnRH-Analogs in EC
In a phase I study, we determined the maximal tolerated dose of Zoptarelin Doxorubicin and characterized dose limiting toxicities in 17 women with ovarian, endometrial, or breast cancers for which standard curative or palliative measures could not be used. In each patient, the primary tumor or the metastatic lesion expressed GnRH-R as determined by immunochemistry. The maximal tolerated dose of Zoptarelin Doxorubicin in the absence of supportive medication was 267 mg/m 2 , equimolar to 76.8 mg/m 2 of free doxorubicin. One EC patient in the 160 mg/m 2 and 267 mg/m 2 groups each, which were considered to be therapeutically effective, had a complete remission. Toxicity was lower for Zoptarelin/Doxorubicin than for free doxorubicin. Pituitary gonadotropin secretion was partly suppressed but other pituitary cells were not influenced [83]. In a subsequent Phase II trial, 43 eligible patients with advanced or recurrent EC that expressed GnRH-R were treated with Zoptarelin Doxorubicin (267 mg/m 2 , 2-h infusion every 21 days). Two patients (5%) had a complete remission and eight patients (18%) had a partial remission. Stable disease for at least six weeks was observed in 44% of patients, resulting in a clinical benefit rate of 67%. There was no difference in response between patients with type 1 or type 2 EC. The most frequently reported grade 3 or 4 adverse effects were neutropenia (12%) and leucopenia (9%) [25]. Based on these results, a phase III trial on 511 patients with advanced, recurrent, or metastatic EC who had failed to prior platinum and Taxane therapy was performed (ClinicalTrials.gov Identifier: NCT01767155). GnRH-R status was not determined. Patients were randomized to obtain Zoptarelin Doxorubicin (267 mg/m 2 ) or doxorubicin (60 mg/m 2 ) every 21 days for up to nine cycles. The median OS for Zoptarelin Doxorubicin was 10.9 months compared to 10.8 months for patients treated with doxorubicin.Progression free survival (PFS) was 4.7 months for both groups. The objective response rate was 12% vs. 14% and the clinical benefit rate was 54% vs. 52%. Side effects were not different between the two treatments [84,85].

Expression of GnRH in EC
The production of GnRH by EC-cell lines and in 75-100% of primary tumor samples was shown by immunoassay, bioassay, high performance liquid chromatography, immunohistochemistry, RT-PCR, restriction enzyme analysis, and Southern blot analysis [12,24,29,30], so that this finding is also well accepted.
A common finding made by all groups was that GnRH-antagonists had the same antiproliferative effects on EC cells as GnRH-agonists, suggesting that the dichotomy between GnRH-agonists and antagonists, known from the pituitary, is not valid in EC cells [6]. To elucidate the effects of GnRH produced and secreted by the tumor cells, we treated cell cultures of human ovarian cancer cell lines EFO-21 and EFO-27 that had been shown to express high affinity GnRH-R and to secrete GnRH with neutralizing concentrations of antiserum to GnRH. This resulted in a significant stimulation of proliferation. Native GnRH in low concentrations had no or little anti-proliferative effect [36].
Thus, it seems reasonable to conclude that the majority of human EC express GnRH and GnRH-R as an autocrine system reducing their proliferation. Both GnRH agonists and antagonists have dose-dependent anti-proliferative effects.

Signal Transduction and Intracellular Actions of GnRH in EC
In the pituitary gonadotrope GnRH-R that have bound the GnRH couple to G-protein αq and induce activation of phospholipase (PLC), protein kinase (PKC), and adenylyl cyclase (AC) [12,14,32]. These enzymes and pathways are present in human EC cell lines but are not involved in the mediation of anti-proliferative effects of GnRH agonists [32]. Upon binding of a GnRH agonist or antagonist, the GnRH-R rather couples to G-protein αi and activates a phosphotyrosine phosphatase that reduces EGF induced tyrosine phosphorylation of the EGF-R [18,32,38,[44][45][46]. This results in a suppression of the Ras/MAPK/ERK pathway and an inhibition of c-fos expression [18,32,38] (Figure 1). This mechanism has been shown in human breast and ovarian cancer cell lines [86]. In addition, through this mechanism, GnRH-agonists can inhibit E 2 induced cell proliferation of ERα-positive human endometrial, ovarian, and breast cancer cell lines (Figure 1) [40]. In breast cancers, signalling of membrane-bound G-Protein-coupled estrogen receptor 1 (GPER) through transactivation of EGF-R could also be inhibited by GnRH-agonist treatment (Figure 1) [86]. Thus, E 2 -induced proliferation of ERα-negative but GPER-positive breast cancer cells could be prevented by treatment with GnRH agonists.
GnRH agonists stimulated via G-protein αi the JNK/AP-1 pathway, leading to inhibition of the cell cycle ( Figure 1) [39]. Native GnRH, GnRH-agonists, and GnRH-antagonists did not induce apoptosis but rather protected endometrial cancer cells from apoptosis induction by UV-light or the cytotoxic agent doxorubicin through activation of NF-κB ( Figure 1) [42]. Thus, we conclude from our data and those of others, that the expression of GnRH and GnRH-R in human EC is part of an autocrine system, counteracting the proliferative effects of growth factors and estrogens, and increasing AP-1 expression, leading to cell cycle arrest. In addition, this GnRH system reduces apoptosis (Figure 1). Other authors, however, found induction of apoptosis caused by GnRH-agonists and/or antagonists, the activation of protein kinase C, and other mechanisms including FasL and FAS, telomerase transferase, telomerase activity, and annexin V [13,15,16,20,34,35,46,47]. The reasons for these discrepancies are unknown. It is not unreasonable to speculate that the GnRH-R in tumor cells can couple to multiple signal-transduction pathways depending on the cellular milieu in different cell lines and even different passages of the same cell line [86].

GnRH-II and GnRH-II-Receptors in EC
The data from our group suggest that a truncated, but functional 5-transmembrane human GnRH-II-R, is expressed in human endometrial and ovarian cancers. After binding of the GnRH-antagonist, native GnRH-II, or GnRH-II-agonist, it couples to G protein αi and activates the signal transduction mechanisms described for the GnRH-I-R in human cancers including inhibition of autophosphorylation of the EGF-receptor and phosphorylation of JNK [54,55,57,58] (Figure 2). Our findings suggest that the antiproliferative effects of GnRH-I antagonists are mediated through GnRH-II-R. The anti-proliferative effects of native GnRH-II mediated through GnRH-II-R were much more pronounced than those of GnRH-I agonist mediated through GnRH-I-R [54][55][56][57][58].
GnRH-I, GnRH-I-agonists, GnRH-I-antagonists, and GnRH-II-agonists did not increase apoptosis in EC, ovarian, or breast cancer cell lines, but rather protected cells from programmed cell death [58,59]. Antagonists of GnRH-II, however, potently induced apoptosis in human endometrial, ovarian, and breast cancer cells, mediated through activation of stress-induced mitogen activated protein kinases p38 and c-JunNH2-terminal kinase, leading to activation of proapoptotic protein Bax [58,59] (Figure 2). GnRH-II antagonists potently inhibited the growth of human endometrial and ovarian cancers in nude mice [58].
Other groups provided evidence that no functional GnRH-II-R is expressed in the human and that the effects of GnRH-II are mediated through GnRH-I-R [26,27,35,39] ( Figure 3).

GnRH-R in EC as a Target for Cytotoxic Therapy
The in vitro and in vivo experimental data available suggest that GnRH-R can be utilized as a target to specifically introduce toxic molecules into EC cells [21,[61][62][63][64][65]. Of particular interest is the finding that cytotoxic GnRH-molecules can bypass the multi-drug resistance 1 system that eventually makes tumor cells refractory to chemotherapy [64].

Clinical Application of GnRH Analogs in EC and Atypical Endometrial Hyperplasia (AEH)
Estrogen withdrawal by the GnRH-agonist induced reversible medical castration, which has been used in combination with systemic or local progestogen application as a fertility sparing treatment for AEH and early EC [69][70][71][72][73]76]. Alternatively, GnRH agonists were combined with aromatase inhibitors for more intensive suppression of estrogens [73,74,76] or given alone after complete endometrial resection [75]. In these small series, complete response rates of about 70% (54-100%) were observed for early EC. Complete response rates were higher in AEH (80-100%). Pregnancy rates were about 50%. A relevant number of patients had progression during therapy or relapsed after complete remission [69][70][71][72][73][74][75][76]. Estrogen withdrawal through reversible medical castration by GnRH analogs alone or in combination with a LNG-IUD or an aromatase inhibitor seems to be preferable in obese patients [76] (Table 1).
A series of small phase II trials has assessed the efficacy of conventional doses of GnRH-agonists in the treatment of recurrent or advanced EC no longer amenable to surgery or radiotherapy, including patients with G3 tumors. Objective responses were observed in 0% to 35%, which tended to be long lasting. Toxicity was low [77][78][79][80][81][82] (Table 2). It has to be considered that the serum concentrations of GnRH-agonists achieved with standard depot-preparations are in the nanomolar range and sufficient to down-regulate pituitary GnRH-R [6]. For optimal direct effects of GnRH-I-agonists or antagonists on EC cells, much higher concentrations (µM) are needed [9,[13][14][15]17,20,31,32,34,35]. Clinical trials with higher doses of GnRH-I-analogs have not been performed. GnRH-II analogs, especially GnRH-II antagonists that are much more potent in vitro and in vivo [58,59], should be tested in EC patients. [75] case series EEC (n = 9) 3.6 mg Goserelin-depot s.c./month for 3 months EEC: 56% CR 44% SD 11% recurrence 14 pregnancies Abbreviations: AEH, atypical endometrial hyperplasia. EEC, early endometrial cancer. CR, complete remission. PR, partial remission. SD, stable disease. NETA, norethisterone acetate. LNG-IUD, levonorgestrel-intrauterine device. i.m., intramuscular. s.c., subcutaneous. p.o., per os.  Phase I and II trials with the cytotoxic GnRH analog Zoptarelin Doxorubicin resulted in an objective response rate of 23% and a clinical benefit rate of 67% in patients with advanced or recurrent EC that expressed GnRH-R [25]. Toxicity was low. In a phase III trial with patients with recurrent or metastatic EC that had failed to prior platinum chemotherapy, an equipotency of Zoptarelin Doxorubicin and free Doxorubicin was found. However, the GnRH-R-status of these tumors was unknown. Considering that Zoptarelin Doxorubicin had been shown to be ineffective in EC not expressing GnRH-R [61][62][63] and that 5-50% of EC have no GnRH-R [20,[23][24][25], it is reasonable to speculate that the Zoptarelin Doxorubicin is more efficacious in patients with EC that express GnRH-R. A series of small phase II trials has assessed the efficacy of conventional doses of GnRH-agonists in the treatment of recurrent or advanced EC no longer amenable to surgery or radiotherapy, including patients with G3 tumors. Objective responses were observed in 0% to 35%, which tended to be long lasting. Toxicity was low [77][78][79][80][81][82] Table  2). It has to be considered that the serum concentrations of GnRH-agonists achieved with standard depot-preparations are in the nanomolar range and sufficient to down-regulate pituitary GnRH-R [6]. For optimal direct effects of GnRH-I-agonists or antagonists on EC cells, much higher concentrations (µ M) are needed [9,[13][14][15]17,20,31,32,34,35]. Clinical trials with higher doses of GnRH-I-analogs have not been performed. GnRH-II analogs, especially GnRH-II antagonists that are much more potent in vitro and in vivo [58,59], should be tested in EC patients.
Phase I and II trials with the cytotoxic GnRH analog Zoptarelin Doxorubicin resulted in an objective response rate of 23% and a clinical benefit rate of 67% in patients with advanced or recurrent EC that expressed GnRH-R [25]. Toxicity was low. In a phase III trial with patients with recurrent or metastatic EC that had failed to prior platinum chemotherapy, an equipotency of Zoptarelin Doxorubicin and free Doxorubicin was found. However, the GnRH-R-status of these tumors was unknown. Considering that Zoptarelin Doxorubicin had been shown to be ineffective in EC not expressing GnRH-R [61][62][63] and that 5-50% of EC have no GnRH-R [20,[23][24][25], it is reasonable to speculate that the Zoptarelin Doxorubicin is more efficacious in patients with EC that express GnRH-R.

Conclusions and Perspectives
Treatment with conventional doses of GnRH-agonists that suppress pituitary gonadotropin secretion and ovarian estrogen production has become part of fertility preserving therapy either alone or in combination with LNG-IUD (52 mg) or aromatase inhibitors in young patients with AEH or early EC (EEC).
There is convincing evidence that EC express GnRH-I-R and GnRH-I as a negative autocrine system, limiting cell proliferation and likely apoptosis. It is unclear whether a functional GnRH-II-R exist in humans. The GnRH-I-R and, if existent, the putative GnRH-II-R can be targeted by analogs of GnRH-I and GnRH-II to inhibit proliferation and to induce apoptosis. Clinical trials to exploit these direct anti-tumor effects have, so far, been performed with conventional doses of GnRH agonists, resulting in marginal efficacy, but low toxicity. Trials using higher doses of GnRH-I analogs or the more potent GnRH-II analogs still have to be performed. The cytotoxic GnRH-analog Zoptarelin Doxorubicin has been shown to have meaningful activity in animal models and a phase II trial. A welldesigned phase III trial with patients with advanced or recurrent EC expressing GnRH-R is warranted.

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