Over 80 scientists met in the beautiful Mediterranean island of Capri last month to discuss their latest research on cell migration in development and disease. This EMBO workshop brought together developmental and molecular biologists, as well as cancer researchers, aiming to decipher and regulate the behaviour of cells in different contexts. Understanding the mechanisms of cell migration in development and disease is crucial for the improvement of current cancer therapies. The meeting's sessions covered a variety of topics, from the cellular and molecular mechanisms of migration to the latest findings in tissue engineering. Several talks highlighted the roles of adhesion molecules in cell migration and presented novel drug targets for cancer treatment. However, in vivo time-lapse imaging stole the show as visualizing how cells move during embryogenesis or metastasis is yielding exciting new insights into these processes.

Adhesion and Cell Migration

Richard Assoian, University of Pennsylvania, opened the first session on 'Cellular and molecular mechanisms of migration' presenting his work on the relationship between cell adhesion and proliferation. His laboratory has found that epithelial to mesenchymal transition (EMT) is associated with the formation of actin stress fibres and the activation of mitogen-activated protein kinase (MAPK), which leads to increased levels of cyclin D1. Furthermore, using acrylamide gels of varying stiffness, he showed that the compliance of the substrate influences both the organization of the cytoskeleton and the induction of cyclin D1. Re-seeding MCF10A cells on low stiffness gels after undergoing EMT prevents both the formation of stress fibres and the induction of cyclin D1. These findings indicate that in response to different tensional environments, adhesion receptors activate different signalling pathways to induce cyclin D1 and proliferation; in high tensional environments MAPK/ERK activation increases cyclin D1 expression, whereas in low tensional environments a Rac-dependent pathway ensures cell division.

The role of cell adhesion in migration and invasion was emphasized in two other talks. Gareth Jones, Kings College London, gave a talk on WIP, a Wiskott-Aldrich syndrome protein (WASP)-interacting protein that is required for the formation of podosomes — highly dynamic actin-based adhesive structures — in haematopoietic cells. WIP regulates both the stability of WASP, by regulating calpain-mediated cleavage of WASP, and its recruitment to sites of actin polymerization. The finding that 80% of WAS patients have mutations in the WIP-binding site of WASP reflects the importance of WIP function in this disorder. Podosomes are also found in endothelial cells and work from Elisabeth Genot's lab has identified a novel role for these structures in vascular remodelling. In ischemic tissue and sites of inflammation, TGF signals through the activin-like kinase receptor 1 (ALK1) and ALK5 receptors to induce podosome formation and extracellular matrix degradation therefore, increasing the cells' invasive potential.

Elisabetta Dejana, IFOM Milan, highlighted the role of vascular endothelial (VE)-cadherin in signal transduction from growth factor receptors during development of the vascular system. She showed that inactivation of VE-cadherin inhibits normal vascular development due to uncontrolled vascular-endothelial growth factor (VEGF) signalling. VE-cadherin is found in a complex that is anchored to the actin cytoskeleton and modulates VEGF receptor internalization through clathrin-coated pits. These findings demonstrate a role for adherens junction proteins in intracellular signalling that is independent of their function in cell-cell adhesion.

Cancer and Cell Invasion

EMTs are not only observed during embryonic development; hallmarks of EMTs have also been associated with cancer formation and metastasis. David Salomon spoke about the Nodal co-receptor Cripto-1 which regulates branching of the ductal system during mammary gland development and is overexpressed in most cases of human breast cancer. Cripto-1 signals through Src and phosphatidylinositol 3-kinase (PI3K) to induce mammary epithelial cell migration and invasion in adult mice. Cripto-1 reduces the expression of Netrin and exogenous Netrin was able to reduce Cripto-1-induced EMT and cell invasion. Whether Cripto-1 could be used as a therapeutic target in breast cancer remains to be seen.

Alternative anti-cancer strategies involve targeting secreted proteases which, by degrading the extracellular matrix, are able in some instances to promote cell invasion and metastasis. However, as Dylan Edwards explained, proteases have both pro- and anti-metastatic functions that could account for the disappointing performances of broad spectrum protease inhibitors in clinical trials. His talk emphasized the need to identify the specific proteases that are upregulated in particular cancers and determine their prognostic significance. For example, Hepsin is a transmembrane serine protease that is upregulated in prostate cancer and activates hepatocyte growth factor signalling leading to tumour growth. Selective Hepsin inhibitors could become valuable agents in the treatment of prostate cancer.

Intravital imaging of mouse tumours carried out in Erik Sahai's laboratory, Cancer Research UK, shows that even metastatic cancer cells are non-motile in vivo but transiently switch to a highly motile amoeboid-like behaviour. Upon reaching a metastatic site the cells then revert to a non-motile state. Furthermore, they do not use focal adhesions as has been observed in vitro. Acto-myosin contractility appears to be sufficient to deform collagen fibres in the absence of protease function in vitro. Consistent with this, inhibition of Rho kinases, which are key regulators of acto-myosin contractility, dramatically reduced amoeboid tumour cell motility in vivo.

Neuronal Migration and Guidance

Frank Gertler from MIT showed his ongoing characterization of Mena/Vasp/Evl triple knockout mice, which survive until just before birth. These proteins are well established actin anti-capping proteins. Ablation of all three proteins cause axon guidance defects in the spinal cord as well as a dramatic decrease in intermediate zone layer neurons in the cortex. In addition to this cell autonomous migration defect, triple knockout cortical neurons fail to elaborate neurites in culture. This effect is associated with the inabilityto form filopodia that prevents microtubules from engaging with actin bundles and driving neuritogenesis. The knockout mice also exhibit defects in vascular junction integrity, indicating that cell adhesion could also be compromised.

Rüdiger Klein presented his most recent work on ephrin/Eph transendocytosis. Glial cells are able to phagocytose EphB2 receptors expressed on the surface of neurons suggesting that glial cells could have an important role in shaping neuronal morphology and even the response to guidance cues by regulating the levels of ephrin receptors. He also presented data on the modulation of EphA4-mediated repulsion of lateral motor column neurons away from the ventral limb mesoderm by glial cell line-derived neurotrophic factor (GDNF) and its receptor Ret. GDNF in the dorsal territory functions as an attractive guidance cue that reinforces the motor neuron's guidance decision in the hindlimb. Ephrin/Eph signalling at hindbrain boundaries is also known to lead to cell repulsion or invasion, and research from David Wilkinson's lab indicates that fibroblast growth factor (FGF) signaling can modulate these alternative cell responses to EphB activation.

Cell Migration in homeostasis

Cell migration is particularly important for leukocyte function and the inflammatory response. To reach sites of infection, leukocytes must exit the blood stream by crossing the vascular endothelium. Anne Ridley's latest work on tumour necrosis factor (TNF)-induced leukocyte migration through endothelial cells indicates that leukocytes can either take a para- or the trans- cellular route (through inter-endothelial cell junctions or through the endothelial cell cytoplasm, respectively). The transcellular route depends on the presence of caveolin in endothelial cells. The migrating cell will receive different stimuli depending on the route it takes, but how this affects its function is yet to be investigated.

Angela Nieto, Instituto de Neurociencias Alicante, gave a very interesting talk on Snail transcription factors that are involved in the induction of EMT, both during development and metastasis. Snail 1 does not only repress the levels of E- cadherin, a hallmark of EMT, it also reduces cell proliferation by decreasing cyclin D2 levels and increases cell survival. Using microarray analysis, her laboratory has found that Snail1 also reduces the expression of the kidney-specific cadherin-16. Snail expression is normally silenced in the kidney to maintain epithelial homeostasis and Snail activation in the adult kidney is sufficient to induce EMT and the development of fibrosis. These results identify a novel role for Snail and shed light on how a common feature of kidney disease can be brought about.

Cell Migration and Development

During development, the shape of C. elegans embryos changes from a relatively spherical shape to a long, thin, worm-like shape. Michel Labouesse presented his work on epithelial cell morphogenesis and elongation during C. elegans embryogenesis. His laboratory has identified Rga-2, a RhoGAP that inhibits Rho kinase and myosin II contractility therefore preventing the shortening of circumferential actin bundles. He proposed that by switching between an active and an inactive state, Rga-2 regulates epithelial cell relaxation and contraction during elongation.

Denise Montell, John Hopkins School of Medicine Baltimore, showed some remarkable time-lapse movies of D. melanogaster border cell migration in living organ culture. Using this system, it is possible to directly observe the effects of genetic manipulation on migration in vivo. Work from her laboratory suggests that rather than stimulating protrusion at the leading edge, guidance cues from the oocyte, such as PVF1, function to repress protrusion in the wrong direction. Ongoing screens are identifying novel regulators of border cell migration, some of which also contribute to ovarian cancer.

Further examples of in vivo time-lapse microscopy during development were presented by Frederic Rosa and Claudio Stern. Rosa's studies on endoderm formation in temperature-sensitive zebrafish mutants indicate that Casanova, a Sox-related protein that is essential for endoderm determination, mediates the transition from random to directed movement upon activation of Nodal signalling in endoderm precursor cells. Claudio Stern's 2-photon time-lapse imaging of chick gastrulation provides valuable new insights into cell behaviour during primitive-streak formation. In particular the vertical movement of cells up and down within a tight epithelium has never been observed in vitro.

Tissue engineering

The final session of the meeting focused on tissue engineering and Robert Tranquillo, University of Minneapolis, presented his work on nerve regeneration and growth of aligned cardiovascular tissues. Previous studies have shown that magnetically aligned collagen or fibrin fibrils guide regenerating nerves and that a similar alignment may occur via mechanically-constrained cell induced tissue contraction. Growing aligned tissue in this fashion (without magnetic force) will enable researchers to exploit contact guidance in the regeneration of both neuronal and cardiovascular tissues. Also in this session Christine Mummery spoke about the recent advances in differentiating human embryonic stem cells (hESCs) into cardiomyocytes which unlike adult cardiomyocytes, have a foetal phenotype and survive transplantation into host myocardium after infarction. Genome-wide transcriptional profiling of hESCs differentiating into cardiomyocytes shows that many of the changes in gene expression are similar to those that take place during heart development. These differentiated cells become integrated in host tissue, but further work regarding their long-term survival and the potential development of teratomas and arrhythmias is required before they can be used clinically. Finally, Giulio Cossu showed that transplantationof mesoangioblasts into a canine model of muscular distrophy leads to a partial recoveryof muscular function. Moreover, the migration of the transplanted cells could be modulated in vivo by several cytokines, therefore improving therapeutic efficacy.

This was a very interesting meeting that brought together researchers studying cell migration from different angles and in different species. The importance of cell migration for physiological and pathological process was central to the meeting and the incredible time-lapse movies that were presented suggest that we may still be in for a few surprises...

Author: Monica Hoyos-Flight