Eph receptor and mTOR pathway crosstalk: implications for cancer

Eph receptors constitute the largest family of receptor tyrosine kinases in the mammalian genome containing at least fourteen members. Unlike most other receptor kinases that bind secreted proteins, Eph receptors bind membrane-anchored ligands called ephrins. Eph receptors are divided into two classes based on their sequence homology and ligand binding properties. Those that bind the glycosylphosphatidylinositol (GPI) linked ephrin-A ligands are called the EphA group while those that bind transmembrane ephrin-B ligands are called the EphB group. These binding preferences have some exceptions: EphA4 can bind to ephrin-As as well as to ephrin-B2 and ephrin-B3 ligands, while EphB2 binds ephrin-A5 at high concentrations. The activation of Eph receptors is a multi-step process that includes ligand-dependent receptor clustering on the cell membrane, Eph autophosphorylation as well as phosphorylation of signaling proteins intracellularly. Eph receptors can regulate various signaling pathways in cells via interaction with a large number of downstream effectors, including GTPase regulating proteins, non-receptor kinases and phosphates as well as adaptor proteins. Another interesting feature of the Eph/ephrin signaling pathway is that ephrin ligands are also capable of transducing signals into the ligand-expressing cells¹.

Since Eph receptor EphA1 was originally cloned in a screen looking for oncogenic tyrosine kinases^{2, 3}, growing evidence suggests a role for Ephrins and Eph receptors in cancer. Expression levels of Eph receptors and ephrins are modulated in many tumors, and mutations in Eph receptors have been identified in various cancers ^{4, 5}. However, the role of Eph receptors and ephrins in cancer is far from clear. Previous studies have focused on the interaction between Eph receptor forward signaling and regulation of intracellular pathways, such as Ras-ERK and PI3K-Akt pathways. However, the consequences of these interactions have been difficult to put in context. We recently identified the mTOR pathway as a crucial mediator of EphA signaling in neurons⁶. Activation of EphA receptors leads to ERK inhibition (probably though RasGAP) and activation of TSC1/2 protein complex without modulating Akt signaling in neurons. Thus, Ephrin-A stimulation leads to mTOR inhibition via TSC1/2, negatively regulating local protein synthesis within the axon compartment of neurons and contributing to control of growth cone dynamics and guidance. This observation not only provides a cellular substrate for the physiological role ephrins play in neuronal motility, but could also have important implications for understanding Eph signaling in cancer.

Deregulated protein synthesis is known to play a crucial role in human cancer. Almost every component of the pathways that regulate protein translation is frequently altered in cancer^{7, 8}. Furthermore, inherited genetic syndromes that affect this pathway, such as Cowden disease, Peutz-Jeghers syndrome and tuberous sclerosis complex, commonly lead to increased risk of cancer. Finally, miRNAs, another regulator of translation are becoming recognized as oncogenes. While it is clear that deregulated protein translation contributes to cancer, it is not known which mRNAs are preferentially translated in various tumors. Identification of such mRNAs can provide both biomarkers and potential therapeutic targets specific to tumor type.

It appears that Eph receptors may be linked to the translational machinery in different ways in different tissues and cell types. For example in C. elegans, a yeast two-hybrid screen identified PTEN/DAF-18 as an interactor of Eph receptor (VAB-1)⁹. Eph/VAB-1 phoshorylates and negatively regulates PTEN/DAF-18 at the protein level. Whether Eph/VAB-1 activation leads to increase in mTOR activity and protein synthesis were not investigated in this study. However, if the interaction between Eph receptors and PTEN occurs in a similar manner in some vertebrate cells, then one would expect an increase in mTOR activity – the opposite of what we observed in response to EphA activation in rodent neurons. This raises the interesting possibility that Eph receptors could modulate the mTOR pathway differentially in different situations. VAB-1 is equally similar to the EphA and EphB receptor subfamilies, but one can imagine certain Eph receptors being able to interact with PTEN while others do not. Similarly, certain Eph receptors may signal to ERK alone while others may signal to ERK and PTEN/Akt. The ability of a particular Eph receptor pathway to signal differentially to ERK, Akt and PTEN may promote different cellular phenotypes. Since PTEN is commonly mutated in cancer cells, loss of PTEN function may modulate the effect of Eph in tumor formation, possibly contributing to the confusing roles that Eph receptors have been assigned in tumorigenesis.

The Eph/ephrin signaling not only plays a role in neural development and cancer, but also in vascular development ¹⁰. It is reasonable to postulate the Eph-mTOR crosstalk will play a role in vascular changes contributing to tumor formation. The exact interactions between Eph receptors and the translational machinery warrant detailed analysis under different normal and pathological conditions since small molecule and peptide modulators of the Eph/ephrin signaling are rapidly being discovered¹¹ and may play important roles in developing novel and successful therapeutic strategies.

Figure. Eph receptors may be linked to the translational machinery in different ways.

(A) In rodent neurons, EphA activation leads to inhibition of ERK with resultant activation of TSC1/2. This suppresses mTOR activity and protein synthesis locally in neuritis (based on ⁶). (B) In C. elegans, Eph homolog VAB-1 interacts with and inhibits PTEN activity. If this interaction were to occur in mammalian cells, it would increase Akt phosphorylation of TSC2, leading to increased mTOR activity and protein translation (based ⁹). Whether such a mechanism occurs in mammalian cell is not yet known. Thick lines indicate activated interactions, and thin lines inactivated interactions.