A drawback of membrane-proteins is the fact that they only stay monodisperse in solution within a non-ordered detergent layer (Boekema 1991), which makes projections fuzzy at the circumference. State of the art in single particle EM At present, single particle EM has its highest impact in large multi-subunit structures that Dibutyryl-cAMP nmr cannot be crystallized easily, either in 3D check details (X-ray crystallography) or 2D (electron crystallography). In the field of photosynthesis, 2D maps of photosynthetic membrane proteins are very helpful in analysis of the peripheral antenna
complexes (reviewed in Dekker and Boekema 2005), although many complexes have not yet been analyzed below 10–15 Å. Nevertheless, there is yet a very useful application at medium
resolution, which is the combination of EM and X-ray CH5183284 supplier structures. Over the last decade, docking of atomic resolution X-ray structures into the molecular envelopes derived by cryo-EM became popular (reviewed by Unger 2001 and Stahlberg and Walz 2008). At a resolution of about 15 Å, pseudo-atomic structures can be derived that tell about the interactions on the level of α-helices of specific subunits (Heinemeyer et al. 2007); a higher resolution (10 Å or slightly better) is necessary to predict interaction at the atomic level. The use of rapid freezing devices in cryo-EM enables to study structural changes within the millisecond range in protein complex during Teicoplanin catalysis. The ribosome is probably
the best studied example of conformational changes studied by single particle EM (Mitra and Frank 2006). Another example of the hybrid X-ray-EM approach is the worm hemoglobin, already presented earlier. It was crystallized more than 60 years ago, at a time when crystallization was just a method to purify a protein! However, to solve such a large structure from X-ray diffraction patterns, phases need to be generated. The phase problem in structure determination by X-ray diffraction was solved by taking information from a low-resolution 3D model by EM, similar to the one presented in Fig. 3c, d, and this finally helped to solve the structure to atomic resolution (Fig. 3e, f) (Royer et al. 2006). Because EM has the unique property to see individual molecules, it has another almost non-explored possibility: to work with partly purified proteins, or even non-purified particles from solubilized membranes (the possibility to work on non-purified proteins will be discussed in the last section). In order to correlate structures to specific proteins, however, biochemical techniques and mass spectrometry analysis are needed for final assignment (Arteni et al. 2005). This type of application is still at its infancy, but no doubt, the combination of mass spectrometry and EM will provide us with structural insight on the level of membranes and cellular complexity.