NEWS AND VIEWS

6 8 Moreover Kirby et a/. have recently enzyme fumarate , strong evidence for of a number of focal adhesion compo­ observed an EM for an unusually favour­ the intervention of quantum-mechanical nents. Treating cells with tyrosine kinase able intramolecular proton transfer to tunnelling effects for a variety of enzyme­ inhibitors blocks focal adhesion as­ 4 9 carbon in excess of 6 x 10 M. Detection catalysed proton transfers , and confer­ sembly5, but the link between tyrosine of large EM values such as these improve ences and reviews devoted to the struc­ phosphorylation and assembly is yet to be the prospects for the rational design of tural enzymology10 of proton transfer resolved. Could tyrosine phosphorylation efficient proton transfer catalysts, but reactions all show the degree of current somehow affect the head-tail interaction also reveal deficiencies in the predictive interest. in vinculin? In normal cells the level of understanding of this most fundamen­ At least in principle, studies of catalytic phosphotyrosine in vinculin is minimal tal and thoroughly studied of chemical antibodies should hold a mirror up to and so it is unlikely that the actin-binding processes. nature. For the important class of proton site in vinculin is unmasked by direct Despite their simplicity and the wealth transfer reactions, Thorn et al. have tyrosine phosphorylation of vinculin of data available for them, proton trans­ shown us a first image. D itself. However, two of the proteins that fers remain a live issue for bio-organic become tyrosine phosphorylated in re­ chemistry and enzymology. In the past Daniel S. Kemp is in the Department of sponse to adhesion, paxillin and tensin, three years, a proposal of a proton trans­ Chemistry, Massachusetts Institute of bind vinculin. Significantly, paxillin's 7 fer crane , a detailed analysis of the Technology, Cambridge, Massachusetts binding site on vinculin has been mapped 6 catalysis of malic acid formation by the 02139, USA. to the tail domain , raising the interesting and testable idea that tyrosine phosphory­ CELLADHESION ~------lated paxillin (or some other protein) might bind to vinculin and thereby expose its actin-binding site. Cryptic sites in vinculin Another signalling molecule implicated in focal adhesion assembly is the small Andrew P. Gilmore and Keith Burridge GTP-bindingprotein, rho. Microinjection of activated rho into quiescent cells with FOR more than a decade there has been maintain that this interaction occurs and diminished focal adhesions and stress controversy over whether vinculin, an to provide evidence for it. Most recently, fibres stimulates the reformation of these 7 abundant cytoskeletal protein concen­ getting around the criticism of contami­ structures . Most recently, rho has been trated at sites of cell adhesion, interacts nants, they have used bacterially express­ shown to regulate phosphatidylinositol- directly with actin. Vinculin is found in ed fragments to show that the carboxy­ 4-phosphate 5-kinase, leading to the 2 cell-extracellular matrix adhesions in­ terminal tail of vinculin binds actin • This synthesis of phosphatidylinositol-4,5- 8 volving integrins (focal adhesions) and in finding was convincing, but difficult to bisphosphate (PtdinsP2) • Several actin­ cell-cell adhesions mediated by cadherins reconcile with the absence of binding binding proteins are regulated by (adherens junctions). Both of these re­ found by others with intact vinculin. PtdinsP2 and an interaction of vinculin 9 gions are major sites for the attachment of Vinculin has two principal domains, an with PtdinsP2 has been reported , again actin filaments to the plasma membrane, amino-terminal globular head and an raising the possibility that PtdinsP2 or and current evidence suggests that vincu­ elongated carboxy-terminal tail (in elec­ related phospholipids may regulate the lin is a big player in this attachment. The tron microscopic images, it looks like a accessibility of the cryptic site in vinculin. controversy is now resolved, with an in­ balloon on a string). Johnson and Craig Progress in understanding the organiza­ teresting twist, by Johnson and Craig on previously identified an intramolecular tion and assembly of focal adhesions has 1 page 261 of this issue . These authors interaction between the head and tail been slow. Johnson and Craig's identifi­ confirm that vinculin does contain an domains of vinculin, and showed that it cation of a cryptic actin-binding site in 3 actin-binding domain, but show that this decreases the affinity of vinculin for talin • vinculin may be a missing piece of the site is normally masked in the purified This study laid the groundwork for their puzzle. If this discovery leads to experi­ protein. reinvestigation of vinculin's interaction ments that connect adhesion-mediated Given vinculin's distribution in cell with actin. In agreement with the work of signalling pathways with the assembly of adhesions, it was assumed from the time Jockusch's group, they have identified an focal adhesions, then the significance of of its discovery that its job is to link actin actin-binding site in the tail of vinculin, this work will go far beyond the resolution to the membrane. This assumption but show that the head-tail interaction of the long-standing controversy of seemed to be confirmed when several labs normally masks the site in the purified whether vinculin does or does not bind demonstrated an interaction between vin­ protein. More than simply resolving a actin. D culin and actin. Not long after, however, it controversy, this work raises the question was shown that most of the actin-binding of whether the site is unmasked in focal Andrew P. Gilmore and Keith Burridge are activity in vinculin preparations was due adhesions and, if so, how the process in the Department of Cell Biology and to contaminants, which could be removed might be regulated. It is easy to Anatomy, University of North Carolina at by further purification. One of these con­ envisage how a conformational change Chapel Hill, North Carolina 27599, USA. taminants was identified as a fragment of that promotes interaction of focal tensin, another focal adhesion protein adhesion proteins could be critical to 1. Johnson, R. P. &Craig, S. W. Nature373, 261-264 (1995). that binds to vinculin and caps actin focal adhesion assembly. An appealing 2. Menkei,A. R. etai.J. Cei/Bio/.126, 1231-1240 filaments. In addition vinculin was found model is presented in Johnson and (1994). 3. Johnson, R. P. &Craig, S. W.J. bioi. Chern. 269, to bind two other prominent focal Craig's Fig. 3 on page 264. 12611-12619 (1994). adhesion proteins, £1'-actinin and talin, There is considerable interest in signal­ 4. Schaller, M.D. &Parsons,J. T. Trends Cell Bioi. 3, both of which interact with integrins and ling pathways initiated in response to cell 258-262 (1993). 5. Burridge, K., Turner, C. E. &Romer, L. H.J. Cei/Bioi.U9, actin in vitro. In many models of focal adhesion that, among other consequen­ 893-903 (1992). adhesions, vinculin's function was rel­ ces, may contribute to the development 6. Wood, C. K., Turner, C. E.,Jackson, P. &Critchley, D. R. J. Ce//Sci.107, 709-717 (1994). egated to that of linker between talin of focal adhesions. An early event, follow­ 7. Ridley,A.J. &Hall, A. Ce//70, 389-399(1992). and £1'-actinin. Nevertheless, Jockusch's ing adhesion to extracellular matrix pro­ 8. Chong, L. D., Traynor-Kaplan,A., Bokoch, G. M. & Schwartz, M.A. Ce//79, 507-513 (1994). group, which first identified the inter­ teins, is activation of the focal adhesion 9. Fukami, K., Endo, T., Imamura, M. & Takenawa, T.J. bioi. 4 action with actin, has continued both to kinase and the tyrosine phosphorylation Cham. 269, 1518-1522 (1994). NATURE · VOL 373 · 19 JANUARY 1995 197