One of the most critical events required for propagation and maintenance of signals generated following binding of integrins by ECM is clustering of the integrin receptors, often at sites known as focal adhesions or focal contacts. Without the appropriate clustering, ligand occupation of the receptor is not sufficient to induce a full biological response (Miyamoto et al., 1996). The clustering is believed to facilitate interactions between the integrin cytoplasmic tails and adaptor proteins to allow focal adhesion complexes to assemble. Not surprisingly, immunoprecipitation of these FA complexes showed that the growth factor (GF) receptors are also found within these integrin-containing complexes. For example, both PDGF-BB and insulin receptors were immunoprecipitated in complexes isolated using antibodies against avft3 integrin (Schneller et al., 1997). The colocalization of GF receptors and integrins is believed to facilitate 'crosstalk' between ECM and GF receptors and coordinate or amplify the signals which may be independently generated by the ECM or soluble mitogens.
Clustering of integrins is not only required for integrin-induced migration or proliferation but is also essential for differentiation and tissue-specific gene expression in mammary epithelial cells on BM (Roskelley et al., 1994). This absolute dependence on clustering of integrins for either proliferation or differentiation underscores the importance of having structure imposed upon cytoplasmic signalling mediators.
In addition to structural changes at the level of the focal adhesion, adhesion to different ECMs can induce cells either to spread or to become rounded and polarized. It has become increasingly evident that cell-shape changes are a necessary and integral component of how cell--ECM interactions can generate tissue-specific architectures and gene expression, i.e. in order for ECM to evoke the appropriate response, cells must adopt an appropriate morphology (Roskelley et al., 1994). In general, cell spreading appears to be required for cells to proliferate while cell rounding is a prerequisite for growth arrest. For example, work by Ingber and colleagues has elegantly demonstrated that when endothelial cells are cultured on FN matrices they often adopt a spread morphology and proliferate. However, using micropatterned substrates which forced cells to become rounded while maintaining a similar degree of integrin mediated contact with the matrix, cells were unable to grow in the presence of mitogens (Chen et al., 1997). In contrast, cell rounding, which can be induced by BM-type ECM, is required for other functions such as the expression of the ft-casein gene by mammary epithelial cells (Roskelley et al., 1994).
Binding of cells to the ECM not only can ligate and cluster integrins to initiate signalling cascades, but also provides the cells with a morphology to sustain the appropriate response.
What has not been directly established in these studies is how the cell shape impacts on intracellular signalling cascades. It is entirely possible that a cell's shape can determine whether integrins recruit signalling intermediates which interact with the growth-promoting MAPK pathways. For example, perhaps cell rounding, which generally suppresses growth, might preclude recruitment of membrane-associated mediators such as caveolin-Shc complexes and thereby attenuate proliferative signals, whereas cell spreading may support this effect. In order to understand how cellular geometry influences intracellular signalling it is necessary to understand the dynamics of integrin-cytoskeletal interactions which underlie these morphological changes. These are discussed briefly below.
Was this article helpful?