Posted by: Dan | May 4, 2006

Mechanisms of Metastasis: Epithelial-Mesenchymal Transition

I often get the question “What’s important about cell migration?” I’ve noted a few areas of medicine and biology that cell migration impacts in my cell migration primer, but the one that seems to catch people’s attentions the most is metastasis, and the reasons are simply: cancer is one of the top causes of death in the world today, and of those cancer-related deaths, about 90% of patients die from the invasion of malignant cells to secondary tumors. Surgeons can, in most cases remove the original tumor, but an invasive cancer can evade and escape elimination.

So what events does the process known as metastasis comprise? The standard answer goes something like this: Metastasis is the process by which a cancer cell leaves the primary tumor, travels through and proteolytically dissovles surrounding connective tissues, invades the lymphatic system and blood vessels, travels to and attaches at a distant site of the body via the circulatory system, and colonizes a new tumor.

Here, I’ll focus on the first part – variously called “malignant transformation” or the “epithelial-mesenchymal transition,” where a cancer cell ceases to conform to the architecture of the tissue which gave rise to it (the epithelium, in this case), and takes on amoeboid (or mesenchymal, fibroblastoid, etc.) characteristics.


The cadherin family of Ca2+ -dependent adhesion molecules plays a critical role in regulating and maintaining cell-cell interactions through tight junctions, adherens junctions, desmosomes and gap junctions. Specifically, the re-organization of adherens junctions includes disruption of complex stability; the disintegration of tight junctions leads to loss of epithelial polarity, destroying the barrier function of epithelial cell sheets; and the loss of desmosomal structures disrupts the continuity of the intermediate filament network,relieving cells from strictly ordered epithelial sheets. Accordingly, changes in the regulation of intermediate filament subunits is frequently observed during EMT, involving a progresive decrease of cytokeratins, and a concomitant increase in vimentin expression.

Among the correlated molecular changes in EMT, one of the most prominent is the loss of E-Cadherin expression, or aberrant expression of other cadherins. Switching from E- to N-Cadherin expression has also been observed in a variety of cancers undergoing EMT, reminiscent of some developmental processes. And further, aberrant expression of N-Cadherin can have a dominant effect, enhancing the motility of tumor cells even in the presence of E-Cadherin in breast cancer cells.

In general, changes or switches in cadherin expression that has been observed in many tumors seems to be a mechanism to “make new friends”, provoking a shift to a more dynamic adhesion state. This change may disrupt normal cell morphology and tissue structure in the epithelia, as well as promoting interactions with the stroma and endothelia. However, the mechanisms underlying these changes in expression in cancer (and development) remain unclear, although the heterogeneous pattern of switching between cadherins observed in different cancers suggests that environmental cues and/or specific signaling pathways might underlie cadherin switching.

The family of Rho GTPases also plays a critical role in the control of cell migration by regulating arrangement of the actin cytoskeleton. Various studies suggest that RhoA may be involved in E-cadherin clustering during initial stages of junction formation, while Rac1/Cdc42 GTPases regulate the linkage of actin filaments to adherens junctions. The functional link of Rho GTPases to E-Cadherin-associated catenins, which are major constituents of cell-cell complexes, may provide a molecular mechanism that coordinates loss of cell adhesion and enhances cell motility.

A number of growth factors, cytokines, and even ECM components have also been suggested to induce EMT via activation of receptor tyrosine kinases (RTKs), when preceded or accompanied by an independent stimulus for cell proliferation for EMT to proceed. Most frequently, the contitutive activation of RTKs and their downstream signaling effectors such as mitogen-activated protein kinase (MAPK) or phosphatidyl-inositol kinase (PI3K) are crucial events in advancing hyperplastic/pre-malignant lesions. In particular, the cytokine Transforming Growth Factor-beta (TGF-beta), which stimulates cell cycle arrest and apoptosis in normal epithelial cells, has been shown to induce and maintain EMT in cooperation with active Ras signaling. The phenotypical changes associated with this TGF-beta/Ras collaboration include acquisition of a metastatic growth behavior, functional de-differentiation, resistance to apoptosis, and modulations of transcription factors involved in the expression of junctional proteins.

Close attention has also been paid to transcription factors with the ability to induce EMT. In mammary epithelial cells, c-Jun causes epithelial depolarization, whereas c-Fos induces loss of cell adhesion and an invasive phenotype which is accompanied by nuclear translocation of the trancription factor beta-catenin/LEF. Many other transcriptional factors have also been highlighted, but this area remains immensely complex, and worthy of an extended discussion all its own.

While the loss of cell-cell adhesions has been a major area of focus in EMT and metastasis, cell-ECM interactions also play a role, although this role is usually more associated with post-EMT cancer cell behavior. Integrin-switching is common, and quite complex, depending upon the tissue from which the tumor originated from, the tumor’s expression profile, and the stage of progression of the disease. This appears to occur in a quasi-Darwinian scheme, where neoplastic cells tend to lose the integrins that secure their adhesion to the basement membrane and help them to remain in a quiescent, differentiated state; and they maintain or overexpress the integrins that foster their survival, migration and proliferation during tumour invasion and metastasis. Although cell-type-dependent changes in integrin signalling make it impossible to rigidly assign each of the integrins to the ‘anti-neoplastic’ or the ‘pro-neoplastic’ category, present evidence indicates that alpha2beta1 and alpha3beta1, at least in some cases, suppress tumour progression, whereas alphavbeta3, alphavbeta6 and alpha6beta4 often promote it.

Also, Integrins and RTKs appear to coordinate pro-migratory signals through Focal Adhesion Kinase (FAK) and Src-family kinases (SFKs). These signals exert their effect by orchestrating changes in the cytoskeleton and by inducing gene expression. Both Rac and Cdc42 activate Wiskott–Aldrich syndrome protein (WASP)-family proteins and p21-activated kinase (PAK), which then activate the ARP2/3 complex and LIM kinase (LIMK), respectively, to induce actin polymerization. Myosin light chain kinase (MLCK), and the Rho effectors Rho kinase (ROCK) and mammalian diaphanous (mDIA), regulate bundling and contraction of actomyosin fibres. PAR6 and protein kinase C (PKC)zeta function downstream of Cdc42 to control cell polarity during migration. Jun amino-terminal kinase (JNK) and extracellular signal-regulated kinase/mitogen-activated protein kinase (ERK/MAPK), which can be activated by SHC or FAK, promote cell migration by activating activator protein-1 (AP-1)-dependent gene expression. Signalling through Ras–ERK/MAPK also cooperates with transforming growth factor-beta (TGF-beta)–SMAD signalling to induce epithelial–mesenchymal transition. Finally, the activation by FAK of ETK tyrosine kinase is also important for cell migration.

And the last major area of study in t


%d bloggers like this: