Increased expression of EphA7 correlates with adverse outcome in primary and recurrent glioblastoma multiforme patients

The incidence of brain tumors worldwide is about 7 in 100,000 per year [1, 2]. Glioblastoma multiforme (GBM), the most aggressive tumor among malignant gliomas, is the most common primary brain tumor in adults and represents a significant source of cancer-related death. GBM usually recurs despite the most aggressive treatment via surgical resection of the tumor followed by radiation and/or chemotherapy [1, 2]. The poor prognosis of patients with GBM (median survival ranging from 9 to 12 months, 5-year survival rate close to 0%) mandates the exploration of novel molecular mechanisms that might contribute to the pathogenesis of this disease and its resistance to therapy with the purpose of therapeutic targeting [1‐3].

Receptor tyrosine kinases (RTKs) are known to be important regulators of cellular growth controlling cell proliferation, differentiation and migration [4, 5]. The Eph receptors and their ligands, ephrins, represent the largest known family of RTKs. Their role has been largely studied during the development of nervous system. They are involved in the development of central nervous system, including axon guidance, axon fasciculation, neural crest cell migration, hindbrain segmentation, vasculogenesis and neuronal cell survival during embryonic development [6‐13]. Eph receptors and ephrin ligands are classified into A and B subfamily, on the basis of their sequence, homologies, structures, and binding affinities. EphA receptors bind the glycosylphosphatidylinositol (GPI)-anchored ephrin-A ligands, whereas EphB receptors bind the transmembrane ephrin-B ligands, whose cytoplasmic domain is capable to engage in various signaling activities; an exception is the EphA4 receptor that binds ephrin-B2 and ephrin-B3 as well as ephrin-A ligands [14‐16]. Moreover, these RTKs have the ability to induce both forward and reverse (bi-directional) signaling between adjacent interacting cells.

To date, various studies have investigated the involvement of the Eph-RTKs in several pathogenetic processes in the nervous system. EphB2 and ephrin-B2 signaling participate in the glial scarring process after spinal cord trauma [17]. The phosphorylation ratio of R-Ras was closely linked to the phosphorylation ratio of EphB2 in glioblastoma tissues [18]. Additionally, the phosphorylation ratio of EphB2 is an important mechanism that mediates glioma cell migration and invasion [19]. Ephrin-B2 and EphB4 were overexpressed by endothelial cells of human malignant gliomas [20]. Ephrin-B3 was also demonstrated as an important factor regulating glioma cell invasion through Rac1 GTPase [21]. EphA2 protein was overexpressed in GBM and anaplastic astrocytoma tissues and was identified as a novel target for the development of glioma vaccines [22, 23]. Another group confirmed overexpression of EphA2 expression in GBM cells, probably through decreased interaction between EphA2 receptor and its inhibitory ligand ephrin-A1 in malignant cells [19].

EphA7 (formerly known as Mdk1/Ebk/Ehk) is highly conserved in vertebrates from fish to human [24]. It is widely expressed in embryonic tissues, especially developing central nervous system [25]. EphA7 cooperates with other EphA receptors in cell signaling, but in contrast to other Eph receptors, it contains two developmentally regulated isoforms: a full-length version containing the intracellular tyrosine kinase domain and a truncated form that lacks this domain [26]. Immunoreactivity for the full-length wild type receptor is found in all cell populations expressing EphA7 mRNA in mouse embryo heads and developing brain, while the truncated EphA7 is absent in the embryos. Interestingly, both isoforms show striking distributions in adult mouse brain [27]. The full-length EphA7 is strongly expressed in neuropil; in contrast the truncated EphA7 is conspicuous on cell bodies and proximal dendrites of a limited number of neuronal types [28]. The truncated form of EphA7 acts as a dominant negative antagonist, suppressing tyrosine phosphorylation of the full-length EphA7 receptor and shifts the cellular response from repulsion to adhesion. Additionally, EphA7 is probably essential during closure of the neural folds, since EphA7-null mice displayed lack of the neural folds resembling anencephaly in man [29]. Moreover, EphA7 has been identified as an important molecular cue expressed after spinal cord injury, implicated in glial apoptosis [30]. Recent work indicated EphA7 as an important mediator of neural progenitor apoptosis during brain development [31]. However, little is currently known about its role in brain tumor angiogenesis and pathogenesis.

In the present study, we investigated the immunohistochemical expression of EphA7 and correlated it with clinical pathological parameters and tumor vascularity. We provide evidence that EphA7 is overexpressed in GBM and suggest that this receptor might be used as a new diagnostic and prognostic marker for further Eph/ephrin targeted molecular cancer therapy.

There is currently an urgent need for development of alternative, effective diagnostic and therapeutic approaches to GBM. The survival of patients with GBM may depend on the identification of novel targets. EphA2 receptor has already been recognized as a potential molecular marker and target in GBM for the development of novel biological therapeutic agents [22, 23, 33]. Whereas several studies in recent years have clearly indicated that altered expression of Eph receptors and ephrin ligands is associated with increased potential for tumor growth, angiogenesis, metastasis and adverse outcome [34‐42], few studies have addressed the role of EphA7 in tumor pathogenicity. By employing immunohistochemical techniques we have found that EphA7 protein is predictive for the outcome of patients with GBM, independent of MVD expression. The data in the present study revealed for the first time a strong correlation between EphA7 overexpression and patient survival.

Hafner C, et al. found that EphA7 is highly expressed in kidney vasculature [43]. The mRNA of EphA7 is strongly upregulated in hepatocellular carcinoma as compared with healthy liver tissue and is downregulated in colon carcinomas. EphA7 is also transcriptionally activated in lung cancer [44]. Furthermore, overexpression of EphA7 protein is frequently found in younger patients and in patients with advanced gastric carcinoma [45]. EphA7 expression is frequently silenced in human colorectal carcinoma by aberrant promoter methylation [46]. EphA7 is located on 6q16.1, a region in close proximity to the chromosome 6 breakpoint found in various types of cancer [47]. Although our findings are not consistent with Wang et al, who found a significant downregulation of EphA7 in colorectal carcinoma [46], they are in line with previous reports reporting a tumor promoter role in lung cancer and hepatocellular carcinoma [43, 44], implicating tumor type-specific function for different Eph family members. Eph receptors expressed in different cell types may have opposite effects due to cell-type specific intracellular signaling pathway [4]. Indeed, EphB4 receptor has been identified as a tumor suppressor in breast cancer, through activation of Abl-Crk antioncogenic pathway [48], while the same receptor presented a tumorigenic potential in mesothelioma, favoring uncontrolled cell growth, migration, and tumor progression [49]. Moreover, membrane-bound ephrins trigger Eph receptor phosphorylation, while soluble forms can bind to Eph receptor, but do not trigger receptor activation [50]. Murine and human peripheral lymphocytes secrete a truncated form of EphA7 [51]. Truncated Eph receptors retaining their ligand-binding capacity have been shown to block activation of the full-length receptor [52]. Promoter hypermethylation and silencing of EphA7 in mature B-cell lymphomas may serve to eliminate the inhibitory activity of secreted EphA7 on tumor-promoting EphA7 receptor signaling, thus enhancing tumor cell spread and recruitment of accessory cells able to promote tumor growth [51]. A recent study on signaling pathways involved in EphA7 RTK reported that direct EphA7 knockdown can result in attenuation of ERK1/2 phosphorylation and induce apoptosis of leukemia cells, suggesting the impact of EphA7 on the growth of tumor cells [45]. It is of interest that positive EphA7 expression was closely associated with increased age of patients (>55 years). Whether this is a random finding or not deserves further investigation.

The unfavorable prognostic influence of EphA7 in GBM could be attributed to the well-recognised role of Eph RTKs in tumor angiogenesis. Indeed, in this study a statistically significant correlation between expression of EphA7 and MVD was noted in GBM specimens. Another important observation was EphA7 overexpression in both vasculature as well as tumor cells. The process of angiogenesis plays a central role in tumor growth and in the development of distant metastases by facilitating the entry of cells into circulation [52‐55]. A vast biochemical and genetic evidence has implicated the critical role of Eph/ephrin signaling in angiogenesis, despite of VEGFR2 and Tie2 receptors long been recognized as key players in this process [53, 54]. Angiogenetic activity can be measured histologically by MVD, which has been shown to be an independent prognostic parameter in various malignancies including gliomas [56‐58]. However, other studies on angiogenesis of glioblastomas suggested the limited usage of MVD as prognostic parameter due to the complexity of the microvascular network in GBM [59, 60]. Although no correlation between MVD and overall survival or recurrence-free survival was found in our study, we observed a statistically significant correlation between lower MVD and tumor recurrence. Further prospective studies with large numbers of patients are, however, needed to fully clarify the clinical implications of MVD in GBM recurrence.

Taken together, our data illustrated that EphA7 could be a potential candidate as a prognostic tumor marker and a new targeted therapeutic assessment in primary and recurrent GBM. Based on our findings, there might be a possible relationship between EphA7 and tumor neovascularization. Recent data demonstrating inhibition of angiogenesis through EphA receptor blockade in two different animal tumor models are consistent with our observation [52]. Additional experimental work is necessary to unveil the biologic pathway linking Eph/ephrins with tumor growth in cancer cells and tumor-associated vessels of GBM and further studies are needed before EphA7 becomes established as an important prognostic and predictive tool in GBM. Ultimately, specific EphA7 inhibitors may prove to be of therapeutic value.