Like many other pathogens, enteropathogenic Escherichia coli (EPEC), a diarrhoea-causing bacterium responsible for the death of millions in developing countries, has evolved some cunning tricks for infecting host cells: it subverts normal cellular processes for its own ends. Attempts to understand this molecular trickery shed light on not only the pathogenic processes of EPEC infection but also the underlying cellular machinery. EPEC, which infects intestinal cells, binds to the surface of the cells and triggers the highly localized reorganization of the underlying cytoskeleton to form actin-rich pedestals thought to be crucial for EPEC infection. This picture depicts these pedestals in a HeLa cell exposed to EPEC, with the bacteria pseudocoloured in green, actin in red and bacterial and cellular DNA in blue. Elsewhere in this issue (Nature Cell Biol. 1, 389–391; 1999), Daniel Kalman and colleagues provide the first insights into the cellular mechanisms used by EPEC to induce pedestals, and identify at least two of the host factors involved.

They discover that cellular WASP proteins, known to be involved in actin reorganization in other systems, are specifically localized to the sites of EPEC attachment and that this is absolutely necessary to induce actin pedestals. WASP recruits the so-called Arp2/3 complex and stimulates its actin-polymerization activity, ultimately resulting in pedestal formation. So there are striking similarities and differences between how EPEC and some intracellular pathogens, such as Listeria monocytogenes and Shigella flexneri, activate actin polymerization.

But how does EPEC recruit WASP? Interestingly, the authors show that a domain of WASP that can bind members of the small-GTPase family is both required and sufficient for WASP localization to the sites of EPEC attachment, indicating that an as-yet-unknown GTPase may be involved in this process. They cite preliminary evidence suggesting that a recently identified cellular GTPase named Chp may be the culprit. Identifying this cellular GTPase, as well as other factors that link the bacteria to the host-cell machinery, will be exciting. We are certain to learn a lot more in the future about the interesting molecular devices used by EPEC to infect its host cells, and, more fundamentally, about the control of the actin cytoskeleton itself.