Two-dimensional Crystallization of a Membrane Protein on a Detergent-resistant Lipid Monolayer Luc Lebeau 1 *, Franck Lach 1 , Catherine Ve  nien-Bryan 2 , Anne Renault 3 Jens Dietrich 4 , Thomas Jahn 5 , Michael G. Palmgren 5 Werner Ku È hlbrandt 4 and Charles Mioskowski 1 * 1 Laboratoire de Synthe Áse Bioorganique associe  au CNRS Universite  Louis Pasteur 67401 Illkirch, France 2 Institut de Biologie Structurale CEA-CNRS, 38027 Genoble Cedex 1, France 3 Laboratoire de Spectrome Âtrie Physique, Universite  Joseph Fourier, 38402 Saint Martin d'He Áres, France 4 Max-Planck-Institut fu Èr Biophysik, 60528 Frankfurt am Main, Germany 5 The Royal Veterinary and Agricultural University, 1871 Frederiksberg C, Copenhagen Denmark Two-dimensional crystals of a membrane protein, the proton ATPase from plant plasma membranes, have been obtained by a new strategy based on the use of functionalized, ¯uorinated lipids spread at the air- water interface. Monolayers of the ¯uorinated lipids are stable even in the presence of high concentrations of various detergents as was estab- lished by ellipsometry measurements. A nickel functionalized ¯uorinated lipid was spread into a monolayer at the air-water interface. The overex- pressed His-tagged ATPase solubilized by detergents was added to the subphase. 2D crystals of the membrane protein, embedded in a lipid bilayer, formed as the detergent was removed by adsorption. Electron microscopy indicated that the 2D crystals were single layers with dimen- sions of 10 mm or more. Image processing yielded a projection map at 9A Ê resolution, showing three well-separated domains of the membrane- embedded proton ATPase. # 2001 Academic Press Keywords: lipid layer crystallization; 2D crystal; membrane protein; nickel-chelating lipid; ¯uorinated lipid *Corresponding authors Introduction Genome sequences indicate that roughly 20-40 % of the proteins in any living cell are membrane proteins (Frishman & Mewes, 1997; Jones, 1998). To date, the three-dimensional structures of approximately 10,000 soluble proteins have been determined by X-ray crystallography. However, only the structures of roughly 20 membrane pro- teins have been solved. To a large extent this is due to the dif®culty of growing well-ordered mem- brane protein crystals in view of the amphiphilic nature of these macromolecules. The use of deter- gents (Helenius & Simons, 1975; Tanford & Reynolds, 1976) to mask the hydrophobic surface areas, and more recently the use of lipidic cubic phases (Landau & Rosenbusch, 1996) have yielded some excellent 3D crystals of membrane proteins. However despite these breakthroughs, progress in growing well-ordered 3D crystals of membrane proteins suitable for X-ray crystallographic analysis has been slow. As most membrane proteins are oli- gomeric complexes with a molecular weight great- er than 50 kDa, even without accounting for bound detergent, they are well outside the reach of current solution NMR techniques. Alternative approaches to structure determi- nation are essential, such as electron microscopy. This method relies on the ability to incorporate the membrane protein into a bilayer. Favourable pro- tein-protein interactions may lead to large, well ordered 2D crystals suitable for structural studies by electron microscopy (Henderson et al., 1990; E-mail addresses of the corresponding authors: lebeau@aspirine.u-strasbg.fr; mioskowski@bioorga.u-strasbg.fr Present addresses: A. Renault, GMCM, Ba à t. 11A, Universite  Rennes 1, Campus de Beaulieu, 35042 Rennes Cedex, France; C. Ve Ânien-Bryan, Laboratory of Molecular Biophysics, South Parks Road, Oxford OX1 3QU, UK. Abbreviations used: His-tag, histadine tag; Ni-NTA, Ni 2 -chelating nitrilotriacetate; LE, liquid expanded. doi:10.1006/jmbi.2001.4629 available online at http://www.idealibrary.com on J. Mol. Biol. (2001) 308, 639±647 0022-2836/01/040639±9 $35.00/0 # 2001 Academic Press