Molecular analysis of two-dimensional protein crystallization

Helical crystallization of soluble and membrane binding proteins

Helical protein arrays offer unique advantages for structure determination by cryo-electron microscopy (cryo-EM). A single image of such an array contains a complete range of equally spaced molecular views of the underlying protein subunits, which allows a low-resolution, isotropic three-dimensional (3D) map to be generated from a single helical tube without tilting the sample in the electron beam as is required for two-dimensional (2D) crystals. Averaging many unit cells from a number of similar tubes can improve the signal-to-noise ratio and consequently, the quality of the 3D map. This approach has yielded reconstructions that approach atomic resolution [Miyazawa et al., 1999, 2003; Sachse et al., 2007; Unwin, 2005; Yonekura et al., 2005]. Proteins that naturally adopt helical protein arrays, such as actin and microtubules, have been studied for decades. The wealth of information on how proteins bind and move along these cytoskeletal tracks, provide cross-talk between tracks, and integrate into the cellular machinery is due, in part, to multiple EM studies of the helical assemblies. Since the majority of proteins do not spontaneously form helical arrays, the power of helical image analysis has only been realized for a small number of proteins. This chapter describes the use of functionalized lipid nanotubes and liposomes as substrates to bind and form helical arrays of soluble and membrane-associated proteins.

Anisotropic diffusion of point defects in a two-dimensional crystal of streptavidin observed by high-speed atomic force microscopy

The diffusion of individual point defects in a two-dimensional streptavidin crystal formed on biotin-containing supported lipid bilayers was observed by high-speed atomic force microscopy. The two-dimensional diffusion of monovacancy defects exhibited anisotropy correlated with the two crystallographic axes in the orthorhombic C 222 crystal; in the 2D plane, one axis (the a-axis) is comprised of contiguous biotin-bound subunit pairs whereas the other axis (the b-axis) is comprised of contiguous biotin-unbound subunit pairs. The diffusivity along the b-axis is approximately 2.4 times larger than that along the a-axis. This anisotropy is ascribed to the difference in the association free energy between the biotin-bound subunit-subunit interaction and the biotin-unbound subunit-subunit interaction. The preferred intermolecular contact occurs between the biotin-unbound subunits. The difference in the intermolecular binding energy between the two types of subunit pair is estimated to be approximately 0.52 kcal mol(-1). Another observed dynamic behavior of point defects was fusion of two point defects into a larger defect, which occurred much more frequently than the fission of a point defect into smaller defects. The diffusivity of point defects increased with increasing defect size. The fusion and the higher diffusivity of larger defects are suggested to be involved in the mechanism for the formation of defect-free crystals.

Influence of the roughness, topography, and physicochemical properties of chemically modified surfaces on the heterogeneous nucleation of protein crystals

In this study, the influence of some factors on the heterogeneous nucleation of hen egg-white lysozyme (E.C. 3.2.1.17) on a series of chemically modified surfaces has been investigated. Microbatch crystallization experiments were conducted on the microscope glass slides that were treated with poly-L-glutamic acid (PLG), poly(2-hydroxyethyl methacrylate) (P2HEMA), poly(methyl methacrylate) (PMMA), poly(4-vinyl pyridine) (P4VP), and (3-aminopropyl)triethoxysilane (APTES). An optical microscope with a heating/cooling stage was employed to measure the induction time of heterogeneous nucleation. The surface topography and roughness were characterized by atomic force microscopy. Contact angles for crystallization solution on the investigated surfaces were measured by a contact angle meter. From the theoretical analysis, the energetic barrier to heterogeneous nucleation was found to increase at higher contact angles and to decrease at higher roughness. Experimentally, a qualitative increase of the induction time of the heterogeneous nucleation on P2HEMA, APTES, and PMMA surfaces with the contact angle was observed. Such surfaces as P2HEMA, PLG, and APTES, which were of higher roughness, were shown to promote the heterogeneous nucleation. In addition, the surface with specific topography is expected to increase the possibility of the formation of a critical nucleus. Finally, the P4VP surface appeared to suppress the heterogeneous nucleation as a result of the electrostatic interaction between the lysozyme and P4VP molecules.

Synchrotron radiation diffraction from two-dimensional protein crystals at the air/water interface

Protein structure determination by classical x-ray crystallography requires three-dimensional crystals that are difficult to obtain for most proteins and especially for membrane proteins. An alternative is to grow two-dimensional (2D) crystals by adsorbing proteins to ligand-lipid monolayers at the surface of water. This confined geometry requires only small amounts of material and offers numerous advantages: self-assembly and ordering over micrometer scales is easier to obtain in two dimensions; although fully hydrated, the crystals are sufficiently rigid to be investigated by various techniques, such as electron crystallography or micromechanical measurements. Here we report structural studies, using grazing incidence synchrotron x-ray diffraction, of three different 2D protein crystals at the air-water interface, namely streptavidine, annexin V, and the transcription factor HupR. Using a set-up of high angular resolution, we observe narrow Bragg reflections showing long-range crystalline order in two dimensions. In the case of streptavidin the angular range of the observed diffraction corresponds to a resolution of 10 A in plane and 14 A normal to the plane. We show that this approach is complementary to electron crystallography but without the need for transfer of the monolayer onto a grid. Moreover, as the 2D crystals are accessible from the buffer solution, the formation and structure of protein complexes can be investigated in situ.

Two-dimensional streptavidin crystals: macropatterns and micro-organization

Two dimensional crystals of streptavidin grown on lipid monolayers can be viewed as model systems for the study of phase transitions and morphology. These crystals form a variety of macroscopic morphologies associated with different microscopic crystal structures. Observed morphologies are similar to those found in two-dimensional lipid systems, and growth of the protein arrays is somewhat analogous. Such solid state physical processes as nucleation, transformation between crystal phases, crystal phase coexistence, and roughening have been observed in the streptavidin system. In this review, we highlight observations that cause streptavidin to remain an interesting model system exhibiting a variety of intriguing phenomena.

Two-dimensional crystallization of streptavidin: in pursuit of the molecular origins of structure, morphology, and thermodynamics

The streptavidin two-dimensional (2D) crystallization model has served as a paradigm for molecular self-assembly at interfaces. We have developed quantitative Brewster angle microscopy for the in situ measurement of spatially resolved relative protein surface densities. This allows investigation of both the thermodynamics and morphologies of 2D crystal growth. For crystal structure analysis, we employ TEM on grown crystals transferred to solid substrates. Comparison of results between commercially available streptavidin, recombinant streptavidin, and site-directed streptavidin mutants has provided insight into the protein protein and protein-lipid interactions that underlie 2D crystallization.

Electron crystallographic analysis of two-dimensional streptavidin crystals coordinated to metal-chelated lipid monolayers

Coordination of individual histidine residues located on a protein surface to metal-chelated lipid monolayers is a potentially general method for crystallizing proteins in two dimensions. It was shown recently by Brewster angle microscopy (BAM) that the model protein streptavidin binds via its surface histidines to Cu-DOIDA lipid monolayers, and aggregates into regularly shaped domains that have the appearance of crystals. We have used electron microscopy to confirm that the domains are indeed crystalline with lattice parameters similar to those of the same protein crystallized beneath biotinylated lipid monolayers. Although BAM demonstrates that the two-dimensional protein crystals grown via metal chelation are distinct from the biotin-bound crystals in both microscopic shape and thermodynamic behavior, the two crystal types show similar density projections and the same plane group symmetry.

Electron crystallography of macromolecules

Numerous technical advances in electron crystallography have facilitated determination of the three-dimensional structures of macromolecules, especially those that form two-dimensional or helical periodic arrays. Several recent studies have demonstrated the utility of this technique for visualizing secondary structure such as alpha-helices and beta-sheets of membrane proteins and, in one case, the entire polypeptide backbone. Electron crystallography, therefore, has great potential as a tool for studying structural problems that are relevant to both molecular biology and biotechnology.