The EPH/Ephrin System in Gynecological Cancers: Focusing on the Roots of Carcinogenesis for Better Patient Management

Gynecological cancers represent some of the most common types of malignancy worldwide. Erythropoietin-producing hepatocellular receptors (EPHs) comprise the largest subfamily of receptor tyrosine kinases, binding membrane-bound proteins called ephrins. EPHs/ephrins exhibit widespread expression in different cell types, playing an important role in carcinogenesis. The aim of the current review was to examine the dysregulation of the EPH/ephrin system in gynecological cancer, clarifying its role in ovarian, endometrial, and cervical carcinogenesis. In order to identify relevant studies, a literature review was conducted using the MEDLINE and LIVIVO databases. The search terms ephrin, ephrin receptor, ovarian cancer, endometrial cancer, and cervical cancer were employed and we were able to identify 57 studies focused on gynecological cancer and published between 2001 and 2021. All researched ephrins seemed to be upregulated in gynecological cancer, whereas EPHs showed either significant overexpression or extensive loss of expression in gynecological tumors, depending on the particular receptor. EPHA2, the most extensively studied EPH in ovarian cancer, exhibited overexpression both in ovarian carcinoma cell lines and patient tissue samples, while EPHB4 was found to be upregulated in endometrial cancer in a series of studies. EPHs/ephrins were shown to exert their role in different stages of gynecological cancer and to influence various clinicopathological parameters. The analysis of patients’ gynecological cancer tissue samples, most importantly, revealed the significant role of the EPH/ephrin system in the development and progression of gynecological cancer, as well as overall patient survival. In conclusion, the EPH/ephrin system represents a large family of biomolecules with promising applications in the fields of diagnosis, prognosis, disease monitoring, and treatment of gynecological cancer, with an established important clinical impact.


Introduction
Erythropoietin-producing hepatocellular receptors (EPHs) constitute the largest subfamily of receptor tyrosine kinases (RTK) that bind membrane-bound proteins called ephrins [1]. EPHs are categorized into two subfamilies, EPHAs and EPHBs, based on their structural homology and preferential binding affinities to their ephrin-A and ephrin-B ligands, respectively [2,3]. In humans, nine EPHA receptors (EPHA1-8, 10) that bind five ephrin-A ligands (ephrin-A1-5), along with five EPHB receptors (EPHB1-4, 6) that interact with three ephrin-B ligands (ephrin-B1-3), have been described (Table 1) [4][5][6]. Ephrin-A ligands generally interact with EPHAs via a glycosylphosphatidylinositol anchor on plasma membranes, whereas EPHBs are bound by ephrin-B ligands and tethered to the membrane by a transmembrane domain [7]. EPH-ephrin interaction triggers a response in the cytoplasm of the receptor-expressing cell (called forward signaling) as well as a Ephrin-A ligands are anchored to the plasma membrane by a glycosylphosphatidylinositol anchor, whereas ephrin-B ligands contain a transmembrane domain and a short cytoplasmatic tail. EPHs/ephrins, upon activation, exert their physiological as well as their tumor-promoting and tumorsuppressive functions through complex molecular pathways inside the cytoplasm. Forward signaling is conducted through EPH's interaction with a number of different biomolecules and pathways, such as GTPases of the Rho and Ras family, focal adhesion kinase (FAK), and the pathways of the Janus kinase (JAK)-signal transducer and activator of transcription (STAT), as well as the phosphoinositide 3-kinase (PI3K). In backward signaling, upon the phosphorylation of ephrins, different proteins that contain Src Homology 2 (SH2) or PDZ domain, such as Grb4, interact with the ephrin and transmit the signal downstream [14]. Created with BioRender.com.
Pergaris et al., recently reviewed the clinical impact of the EPH/ephrin family's expression in solid tumors, reporting their key role in the development and progression of head and neck, thoracic, skin, gastrointestinal tract, urinary tract, gynecological, and pediatric cancer [15]. Ovarian, endometrial, and cervical tumors represent the main types of gynecological cancer (GC). More than 90,000 women were diagnosed with GC each year between 2012 and 2016 in the USA, with the incidence rate varying by cancer type and race [16]. In this review, we investigate the dysregulation of the EPH/ephrin system in gynecological tract neoplasia, incorporating research works that explored their role in carcinogenesis via both in vitro and in vivo studies. The literature review was conducted using the MEDLINE and LIVIVO databases. The search terms ephrin, ephrin receptor, ovarian cancer, endometrial cancer, and cervical cancer were employed and we were able to identify 57 studies focused on gynecological cancer and published between 2001 and 2021.

Ovarian Cancer
Ovarian cancer (OC) is the leading cause of death from gynecological malignancies and the fifth most common cause of tumor-related deaths in women in the United States. For 2021, the American Cancer Society estimates the incidence of OC at 21,410 cases and OC-related deaths at 13,770 in the USA [17]. OC includes several histological types. Epithelial carcinomas comprise the vast majority of OCs, with high-grade serous OC representing the most common morphological subtype [18]. Lack of specific symptoms of the disease at its early stages is a significant factor contributing to the typical advanced stage of the tumor at diagnosis, after metastasis has already occurred [19]. Consequently, the five-year survival rate for women diagnosed with invasive epithelial OC in a distant Surveillance, Epidemiology, and End Results (SEER) stage is 31% [20]. In spite of substantial research efforts, the molecular mechanisms of OC's origin, initiation, and progression still remain largely unclear [21]. Given the lack of effective diagnostic tools and treatment strategies, it is hence of utmost importance to identify new molecular markers involved in the pathogenesis of OC, with a view to offering novel, targeted, biological therapeutic approaches.

The EPH/Ephrin System in OC Cell Lines and Human Xenograft Models
Many studies have investigated the role of the EPH/ephrin system in OC cell lines (Table S1) and human xenograft models. Cui et al., knocked down the EPHA1 gene using the CRISPR/CAS9 technique. The inactivation of EPHA1 suppressed many aggressive properties and resulted in G0/G1 cell cycle arrest, reduction of the cellular adhesion capacity, and inhibition of the migration capacity, proliferation, transwell invasion, and activity of the matrix metalloproteinase (MMP)-2 and c-MYC signaling pathways in SKOV3 and COV504 OC cells [22]. On the contrary, Jin et al., reported that EPHA1 expression was negative in HO8910 and weakly positive in A2780 OC cells, with the proliferation rate being significantly reduced in OC cells after transfection with EPHA1 plasmid compared with cells transfected with mock plasmid or untreated ones. An alteration in apoptosis could not, however, be detected in these groups [23].
EPHA2 has been described to promote the growth of OVK-18 cells [24]. Thaker et al., evaluated EPHA2 expression in OC cell lines by Western blot analysis. EG, 222, and SKOV3 OC cell lines overexpressed EPHA2, whereas A2780-PAR and HIO-180 exhibited low to absent EPHA2 expression [25]. Both OVCAR3 and SKOV3 cells demonstrated strong EPHA2 and ephrin-A1 mRNA expression, as detected through reverse transcription polymerase chain reaction (RT-PCR) and Western blot [26]. Moreover, EPHA2 overexpression promoted cell-extracellular matrix attachment in A2780 cells, increased anchorage-independent cell growth in vitro, promoted tumorigenesis in an orthotopic mouse model of OC, and resulted in enhanced microvessel density (MVD) [27].
EPHB2 demonstrated only small variations in RNA expression across OC cell lines. Promoter hypermethylation of EPHB2, EPHB3, and EPHB4 did not, however, seem to play an important role in ovarian tumors, as reported by Wu et al. [28]. Davidson et al., conducted affymetrix U133A array analysis for angiogenic gene expression in multiple OC cell lines. The genes encoding ephrin-B2 and EPHB4 were upregulated in mutant TP53 cells; ephrin-B2 was upregulated in the A2780 line; and EPHB4 was overexpressed in the larger pool of mutant TP53 lines. In contrast, EPHB2 showed high expression levels in wild-type TP53 cell lines, while overexpression of ephrin-A3 was induced by hypoxia [29]. EPHB4 was found to be highly expressed in OC cell lines and to regulate cell migration and invasion. Treatment of Hoc-7 cells with progesterone led to a dosedependent reduction in EPHB4 expression, while inhibition of EPHB4 by specific siRNA or antisense oligonucleotides resulted in reduced viability and led to apoptosis as well as activation of the death receptor caspase pathway in this OC cell line. Interestingly, EPHB4 synthetic antisense oligodeoxynucleotides significantly inhibited tumor growth in mice-bearing human OC xenografts [30]. Ma et al., transfected antisense EPHB4 and shRNA vectors into A2780 and SKOV3 cells. Co-transfection with both vectors could inhibit growth, induce apoptosis, and reduce invasive ability of OC cells, accompanied by downregulation of EPHB4 and the PI3K/Akt/mTOR pathway [31].
Ephrin-A1 expression was shown to be induced in OC cells by the pleotropic transcription nuclear factor kappa B (NFκB) after stimulation with tumor necrosis factor alpha (TNF-α) and interleukin 1 beta (IL-1β) [32]. Ephrin-A3 expression was upregulated by hypoxia and promoted endothelial cell migration and adhesion in OC cell lines [29]. Unlike ephrin-A1, Jukonen et al., suggested that endogenous ephrin-A5 is an inefficient activator of EPHA2-pY588 signaling and receptor internalization in OC cells [33].
Ephrin-B2 was found to be overexpressed in OC cell lines [29]. Table 2 summarizes the role of the EPH/ephrin system in OC cell lines and human xenografts.

The EPH/Ephrin System in OC Patient Tissue Samples
The upregulation of EPHA1 in OC tissues reveals its implication in disease onset and progression [34]. EPHA1 expression has been reported to be increased by more than 10-fold in OC specimens through quantitative real-time (RT-PCR) [35] and to predispose OC patients to adverse clinical outcomes [36].
EPHA4 expression in OC patients correlated with early relapse and showed an adverse clinical association when expressed in both OC tumor cells and tumor-associated macrophages [36].
On the contrary, loss of EPHA5 immunohistochemistry (IHC) expression in OC cases was shown to be associated with tumor grade, International Federation of Gynecology and Obstetrics (FIGO) stage, and poor patient outcome [39].
Liu et al., employed a tissue microarray IHC analysis and detected high EPHA8 protein expression in 44.80% of OC tissues, but only 6.67-15% of normal or benign ovarian tissues. A high EPHA8 protein level correlated with older age at diagnosis, higher FIGO stage, positive lymph nodes (LNs), the presence of metastasis, positive ascitic fluid, and higher serum CA-125 levels [40].
Loss of EPHB1 protein expression in OC tissue specimens was associated with higher tumor grade, the presence of metastasis, a high proliferative index (assessed by Ki67 expression), and significantly worse overall survival (OS) [41]. OC patients older than 60 years of age exhibited higher EPHB2 expression and poorer survival compared with younger ones [42]. On the other hand, Gao et al., performed IHC staining of EPHB3 and underlined that EPHB3 protein expression is significantly reduced in OC specimens, as well as negatively associated with histological grade and FIGO stage [43]. EPHB4 expression was shown to be upregulated in OC and correlated with an adverse clinical outcome, advanced disease stage, the presence of ascites, poorer OS, and poorer response to chemotherapy [30,36,42,44].
The expression of EPHB6 was associated with grade and TNM stage, and negatively correlated with Ki67 expression levels in OC samples, while patients with low EPHB6 protein expression exhibited a poorer outcome [45].
Elevated levels of ephrin-A1 expression led to a more aggressive tumor phenotype and were associated with poor survival in OC specimens [42]. High ephrin-A5 expression correlated with a more aggressive subtype of OC, as well as with poor patient OS [33,35,46].
High-grade OCs showed the greatest ephrin-B expression, with a strong correlation found between ephrin-B expression and MVD, disease recurrence, and a decrease in OS [47]. Schaner et al., investigated the genes that are more highly expressed in ovarian than breast carcinomas and identified ephrin-B1 as one of the best discriminators more highly expressed in OC [48]. Ephrin-B2 was overexpressed in OC tissue samples, significantly increased with clinical stages, and correlated with poor survival [44]. Ephrin-B3 has been described to correlate with EPHB4 expression in ovarian tumor specimens [49]. Table 3 summarizes the role of the EPH/ephrin system in OC tissue samples.

Endometrial Cancer
With an estimate of approximately 66,570 newly diagnosed cases of cancer of the uterine body and about 12,940 deaths from cancers of the body of the uterus in 2021, endometrial cancer (EC) represents the most common malignant tumor of the female genital tract in the USA [50]. Depending on their development from atypical endometrial hyperplasia or atrophic endometrium, ECs can be divided into estrogen-dependent type I and type II nonendometrioid EC, respectively [51]. Even though the combination of surgery and chemotherapy as the first-line therapeutic regimen has showed high response rates, the duration of response does not last long, with a grim prognosis for patients with advanced EC and low five-year survival rates in the case of distant metastasis [52,53]. As such, a better understanding of the mechanisms controlling endometrial carcinogenesis is imperative in order to facilitate the development of novel prognostic and therapeutic tools and improve the clinical management of EC.
EPHs/ephrins that have been shown to play an important role in EC include EPHA2, EPHB3, EPHB4, ephrin-A1, and ephrin-B2. Figure 3 presents some of the oncogenic actions of the EPH/ephrin system in EC.

The EPH/Ephrin System in EC Cell Lines and Human Xenograft Models
EPHA2 protein expression was reported to be upregulated in EC cell lines [54]. EPHA2 has also been speculated to influence tumor cell lysis and contribute to susceptibility to Vδ1 γδ T cells' cytotoxic reactivity [55].
EPHB3 and EPHB4 were shown to be overexpressed in EC cell lines through genomewide microarray-based comparative genomic hybridization (aCGH) technology [54].
Ephrin-A1 induced intercellular dissociation, stimulated cell attachment, and inhibited cell aggregation through the EPHA receptor pathway in human EC-derived Ishikawa cells [56,57]. Table 4 summarizes the role of the EPH/ephrin system in EC cell lines (Table S1) and human xenografts.

The EPH/Ephrin System in EC Patient Tissue Samples
EPHA2 protein expression was reported to be upregulated in EC tissue samples, while its overexpression was significantly associated with high tumor stage and grade, increased depth of myometrial invasion, low estrogen receptor (ER) and progesterone receptor (PR) expression, high Ki67 index, high vascular endothelial growth factor (VEGF) expression, high MVD counts, and shorter disease-specific survival [58,59].
Ephrin-B2 has been reported to be overexpressed in EC cells and to correlate with higher histological grade, adverse clinical disease stage, dedifferentiation, myometrial invasion, and poor survival rates [60,[62][63][64]. Furthermore, ephrin-B2 activated EPHB4 receptor kinase through autocrine and/or paracrine activation [63]. Table 5 summarizes the role of the EPH/ephrin system in EC tissue samples.

Cervical Cancer
Cervical cancer (CC) used to be one of the most common causes of cancer death in female patients, but the death rate dropped significantly due to vaccination against HPV, as well as increased use of screening technology such as PAP and HPV tests. For 2021, the American Cancer Society estimates the CC incidence at 14,480 cases and CC-related deaths at 4290 [65]. Even though the tumorigenesis of CC is mostly associated with persistent infection with high-risk human papillomavirus (HPV) strains, growing evidence suggests that the pathophysiology of invasive CC involves dysregulation of a range of normal factors and cervical epithelium deterioration [66]. Treatment options vary by CC stage and range from radical hysterectomy to radiation with or without chemotherapy [67]. Nevertheless, the five-year survival rates for advanced CC patients are far from satisfactory, with the percentage of women diagnosed with invasive CC in a distant SEER stage reaching 17% [68]. Therefore, novel molecular markers that improve the clinical outcome could be of the highest value in developing individualized therapeutic approaches to CC. EPHA2, EPHB2, EPHB4, ephrin-A1, and ephrin-B2 have been shown to play an important role in CC.

The EPH/Ephrin System in CC Cell Lines and Human Xenograft Models
EPHA2 protein expression exhibited upregulation in CC cell lines, leading to increased cell proliferation and migration of CC. EPHA2 knockdown increased the cellular susceptibility to epirubicin in CC cells and suppressed CC growth in a xenograft mouse model [69].
EPHB2 was also highly expressed in CC cell lines and promoted CC progression by inducing cell migration and invasion as well as the epithelial-mesenchymal transition (EMT) through R-Ras activation [70]. EPHB2 was identified as a direct target of microRNA-204 in CC cells [71]. Table 6 summarizes the role of the EPH/ephrin system in CC cell lines (Table S1) and human xenografts.

The EPH/Ephrin System in CC Patient Tissue Samples
EPHA2 protein expression exhibited upregulation in CC tissue samples, while its overexpression was associated with cyclin-dependent kinase 6 protein expression, invasion depth, LN metastasis, clinicopathological disease stage, and OS [69,72,73].
EPHB2 was highly expressed in CC tissue specimens and positively correlated with R-Ras expression [70,74].
Two large studies investigated the expression of EPHB4 through IHC and real-time RT-PCR in CC tissue samples and reported that high EPHB4 expression correlated with adverse clinical disease stage, larger tumor size, LN metastasis, high MVD, and poor patient OS [75,76].
Ephrin-A1 seems to be overexpressed in CC tissue samples and associated with poorer OS [72,73,77].
High ephrin-B2 expression correlated with adverse clinical disease stage, larger tumor size, LN metastasis, high MVD, and poor survival in CC tissue samples [75,76]. Table 7 summarizes the role of the EPH/ephrin system in CC tissue samples.

The EPH/Ephrin System as a Treatment Target in GC
As studies continue to prove the tumorigenic properties of various EPH/ephrin system members in GC, several researchers have focused on the development of novel therapeutic strategies targeting their carcinogenic actions over the past decades. In terms of OC treatment, restoration of miR-520d-3p decreased EPHA2 protein levels, thus suppressing tumor growth and migration both in vitro and in vivo, while dual inhibition of EPHA2 in vivo using 1,2-dioleoylsn-glycero-3-phosphatidylcholine (DOPC) nanoliposomes loaded with miR-520d-3p and EPHA2-siRNA had synergistic antitumor effects and greater therapeutic efficacy than either monotherapy alone [78]. The combination of EPHA2-targeting siRNA-DOPC with paclitaxel, on the other hand, resulted in a drastic reduction of tumor growth in vivo in comparison with treatment with paclitaxel and a nonsilencing siRNA [79], while platinum triggered an oncogenic EPHA2-S897 phosphorylation in vivo associated with platinum resistance [80]. Novel EPHB4-based therapeutic approaches against OC include the bidirectional ephrin agonist peptide (BIDEN-AP), which was found to inhibit endothelial migration and tube formation and to effectively suppress invasion and epithelial-mesenchymal transition in OC cell lines. In vivo, BIDEN-AP and its nanoconjugate CCPM-BIDEN-AP significantly hindered the growth of orthotopic ovarian tumors and compromised angiogenesis [81]. JI-101, an oral multikinase inhibitor targeting EPHB4, was tested in a clinical trial to assess its efficacy in OC. Unfortunately, no patient demon-strated a response according to predetermined criteria [82]. PF-06647263, an anti-ephrin-A4 antibody-drug conjugate, achieved sustained tumor regression in OC patient-derived xenografts in vivo [83]. In a phase I study, response to treatment with PF-06647263 did not, however, seem to correlate with ephrin-A4 expression levels [84]. [ 76] In vivo therapy experiments in mouse orthotopic models with monoclonal antibodies targeting EPHA2 resulted in significant MVD reduction, decreased proliferation, increased apoptosis, and a lower incidence of distant metastasis in EC [58,85]. Table 8 presents the different therapeutic agents targeting the EPH/ephrin system in GC.

Conclusions
It is established that members of the EPH/ephrin system are implicated in a multitude of molecular procedures, exerting their role in different stages of GC. Consequently, up-or downregulation of each individual EPH/ephrin in gynecological tract tumors influences numerous clinicopathological parameters such as tumor grade, invasion depth, tumor cells' proliferation rate (assessed by Ki67 index), FIGO stage, hormone receptor status, patient survival, and response to chemotherapy. Therefore, the aforementioned family of biomolecules emerges as potent biomarkers, useful in the fields of diagnosis, prognosis, and disease monitoring in GC. All investigated ephrins exhibited upregulation in OC, EC, and CC, whereas EPHs showed either significant overexpression (OC, EC, CC) or extensive loss (OC) depending on the particular receptor. However, discrepancies between the results reported by different research groups do exist, with studies sometimes reaching contradictory conclusions on the role of some EPHs/ephrins in certain types of GC. EPHA1 expression, for instance, has been reported to be upregulated in OC by several authors [22,[34][35][36], whereas Jin et al., concluded that EPHA1 expression is decreased in OC [23]. Of note, Jukonen et al., used seven independent OC datasets, as well as the Cancer Genome Atlas (TCGA), for result validation, and reported variable gene expression for different EPHs/ephrins in high-grade OC [33]. In recent years, large cancer genome studies, such as the TCGA, have searched exhaustively for genes, thus offering invaluable data in terms of results validation. Further research is, therefore, necessary in order to shed light on the impact of certain EPHs/ephrins on carcinogenesis, while researchers may make use of the TCGA and similar datasets to systematically investigate EPHs/ephrins in cancer entities. Given their tumor-promoting or tumor-suppressive properties, EPHs/ephrins could be targeted by therapeutic agents, blocking those members enhancing carcinogenesis or inducing the expression of those suppressing it. Several researchers have already investigated the use of agents targeting members of the EPH/ephrin system, alone or in combination with traditional chemotherapeutic agents, via both in vivo and in clinical studies [58,69,78,79,[81][82][83][84][85] and found promising antitumor effects for GC. Anti-ephrin-A4 calicheamicin conjugates, for example, effectively target OC-initiating cells to result in sustained tumor regression [83]; the preclinical data for EC treatment using the antibody drug conjugate MEDI-547 show substantial antitumor activity [85], while the level of EPHA2 expression positively correlates with the therapeutic resistance of CC to epirubicin [69]. Nevertheless, further clinical trials with larger numbers of patients are needed to verify the clinical utility and safety of the agents targeting members of the EPH/ephrin system in the treatment of women with GC, to investigate possible adverse effects following their administration to patients, as well as to assure their effectiveness depending on EPH/ephrin expression by GC cells. A limitation of this review is the nonsystematic methodology in terms of study selection. Even though systematic literature reviews offer the most accurate strategy to detect relevant research works, following rigorous rules and standards, this approach requires a narrow research question that does not cover broader topics, such as the role of the EPH/ephrin system in GC. Another limitation is the possible evidence selection bias, arising from publication bias, given that data from statistically significant studies are more likely to be published. Despite these limitations, the current work represents an extensive and concise review of the literature and includes all the studies conducted on the role of the EPH/ephrin system in GC pathogenesis.
Author Contributions: Literature analysis and conceptualization, I.P., A.P. and S.T.; original draft preparation and writing, I.P.; artwork, A.P.; review and supervision, K.V., G.T. and S.T. All authors have read and agreed to the submitted version of the manuscript.