Activation of receptor tyrosine kinases (RTKs) and G-protein coupled receptors GPCRs by their ligands leads to the activation of intracellular signaling cascades. While these pathways were initially thought to be distinct, recent data indicate an important role for RTK-GPCR transactivation in a number of physiological and pathological cellular responses. This form of receptor transactivation has been shown to regulate cell proliferation [22–24], migration and invasion [1, 25, 26] in various types of cancer, and our recent data indicate an important role for IGF-1R and CXCR4 transactivation in migration of MDA-MB-231 breast cancer cells . This IGF-1R-CXCR4 heterodimer appears to be linked with the metastatic phenotype of these cells as the related but non-metastatic MCF-7 breast cancer cell line does not express functional heterodimers. Therefore, it is important to understand how IGF-1R-CXCR4 transactivation facilitates migration of MDA-MB-231 cells. In the present study, three novel observations with respect to IGF1-R and CXCR4 transactivation were made. First, PI3Kγ is the major PI3K isoform involved in IGF-I-induced cell migration of the metastatic breast cancer cell line MDA-MB-231. Second, eEF2 is one of the downstream targets of PI3Kγ after this heterodimeric receptor transactivation. Third, IGF-1R-CXCR4 transactivation leads to PI3Kγ-dependent phosphorylation of eEF2. These findings indicate that PI3Kγ may promote breast cancer cell migration through a novel mechanism by deactivating eEF2 after IGF-1R-CXCR4 transactivation.
Activation of the class IA PI3Ks, PI3Kα, β and δ following ligation of IGF-1R by IGF-1 is well documented [27–30]. However, the two major PI3K isoforms known to be activated downstream of GPCRs and to play a role in cell migration in response to GPCR ligands are p110γ and p110δ [4, 9, 31–33]. Thus, we investigated the expression of these PI3K isoforms in metastatic MDA-MB-231 and observed that these cells express both p110γ and p110δ. Our previous data indicate that MDA-MB-231 cells express a functional IGF-R-CXCR4 heterodimer whereas MCF-7 cells do not . In fact, IGF-I signals directly through IGF-1R in MCF-7 cells to control migration of the cells, independently of CXCR4 . We therefore investigated the level of expression of these PI3K subunits in MCF-7 cells and found that while both MCF-7 and MDA-MB-231 cells express similar levels of p110δ, MCF-7 cells express a low level of p110γ.
PI3Kγ is generally activated by GPCRs, including chemokine receptors, such as CXCR4 , but to the best of our knowledge, has not been implicated in IGF-1R signaling. Here, we show that IGF-I stimulation leads to the membrane translocation of p110γ, an indicator of PI3K activation. Moreover, specific inhibition of PI3Kγ and knockdown of p110γ resulted in decreased phosphorylation of Akt and cell migration in response to IGF-I, whereas PI3Kδ did not appear to be involved in this response. Taken together, these data indicate that PI3Kγ is a major PI3K isoform regulating MDA-MB-231 breast cancer cell migration in response to IGF-I.
To shed light on the signaling pathways regulated by p110γ downstream of IGF-1R-CXCR4 transactivation, we performed a 2D DIGE proteomics experiment. We compared the cytosolic proteome from MBA-MB-231 control cells with MBA-MB-231 cells in which p110γ has been knocked down after 5 min of IGF-I stimulation. Importantly, this short stimulation time allowed us to focus on post-translational modifications to the MDA-MB-231 cell proteome as this time point was too short for effects on gene expression. These experiments identified eEF2 as one of the downstream effectors of PI3Kγ after receptor transactivation. eEF2 is known to play a critical role in regulating protein synthesis by mediating the ribosomal translocation from the A to the P-site in eukaryotic tissues, the reaction that induces movement of mRNA along the ribosome during translation . Phosphorylation of eEF2 prevents functional binding to the ribosome and delays the elongation step, thereby terminating translation . PI3Ks have previously been implicated in the regulation of the eEF2 downstream of proliferative signals , however whether specific PI3K isoforms or all PI3K isoforms regulate eEF2 signaling has not yet been determined.
Regulation of the eEF2 activity by PI3Kγ may be the result of multiple molecular mechanisms. Firstly, PI3Kγ may activate the eEF2 kinase through the mTOR/P70S6K pathway, which has been shown as a key pathway regulating eEF2 activity . Secondly, PI3Kγ may reduce the rate of eEF2 dephosphorylation through inhibiting the activity of a protein phosphatase such as protein phosphatase 2A (PP2A) [36, 37]. The first protein kinase substrate of PI3Kγhas been identified recently. PI3Kγphosphorylates SET, an endogenous inhibitor of PP2A, on two serine residues. eEF2 might therefore be a direct or indirect substrate of PI3Kγ.
Cells can respond to growth factors by either migrating or proliferating, but not both at the same time, a phenomenon termed “migration-proliferation dichotomy” , This is not only observed in cancer progression but also during wound healing and developmentand the underlying mechanism remains unknown. The proposed physiological basis for this phenomenon is that directional cell migration occurs along an increasing ligand gradient until migrating cells reach a zone in which they start dividing as a result of the presence of ligands that regulate proliferation. Thus limited protein synthesis occurs in migrating cells, which diverts energy to the process of migration. However, when cells stop migrating and start proliferating, protein synthesis is necessarily upregulated. Our results support the view that reduced proliferation is an integral part of migration and, more specifically, that in metastatic breast cancer cells the initiation of both processes might be regulated by PI3Kγ.
It should be noted that the data presented in this manuscript have been obtained using one breast cancer cell line and further experimentation will be required to determine the generality of our observation. This would include examining different cell lines (from breast and other cancers) and ideally, in cells from clinical tissue samples. With respect to the latter, unfortunately at present there is no means by which to identify cells in tissues in which IGF1R/CXCR4 transactivation occurs. However, ultimately, an improved understanding of the molecular mechanisms underlying IGF1R/CXCR4 transactivation, including the role of PI3K signal transduction pathways, in the progression of breast cancer metastasis and invasion may lead to development of more effective diagnostic and therapeutic strategies.
We have also observed that phosphorylation of eEF2 occurs upon stimulation of MDA-MB-231 cells with CXCL12 (the chemokine ligand of CXCR4) in a PI3Kγ-dependent manner (data not shown). Future studies could determine if phosphorylation of eEF2 generally occurs downstream of activated G-protein coupled receptors. Importantly, the results of a recent study implicate phosphorylation of eEF2 as an important link between the DNA damage response and translation of mRNAs . After activation of the DNA damage checkpoint, AMPK mediates activation of eEF2 kinase, which in turn phosphorylates eEF2. The authors conclude that because protein synthesis is energetically costly, stressed cells inhibit this process to devote resources to the stress response. That study, together with the observations in the present study, implies that phosphorylation of eEF2 to inhibit translation may be a general mechanism regulating energy consumption between important energy-dependent cellular processes.