Laboratory of Structural Membrane Biochemistry
Our laboratory studies the structures of membrane proteins. We try to understand, based on structure, what goes wrong in disease and how the protein function is altered in diseased states. Our primary focus is on proteins in the blood-brain barrier. The long-standing question in our laboratory is how the thousands of membrane channels and transporters that exist in the cell membrane work together to help cells maintain homeostasis. With that question in mind, we study membrane proteins that are involved in nutrient, ion, and water uptake; waste removal; signaling; and communication.
Our laboratory is multidisciplinary. Over the past decade we have employed structural biology techniques, such as cryo-electron microscopy (cryo-EM), x-ray crystallography, nuclear magnetic resonance (NMR), and molecular dynamics simulations, and we have used membrane biochemistry and biophysics to understand the function of the proteins of interest. Within electron microscopy we have used electron tomography, single-particle reconstructions, and electron crystallography; however, our specialty lies in electron diffraction.
Part of our laboratory is also devoted to method development in cryo-EM. In recent years we have developed two important methods in electron diffraction: the fragment-based phase extension and MicroED.
Some of our recent studies are outlined below.
The dynamic regulation of water channels. Water channels, or aquaporins, form specialized channels in membranes for water permeation. These are extremely efficient channels that allow millions of water molecules to permeate the pore per second. Because they are channels, the cell can regulate their activity dynamically to help maintain homeostasis. The eye lens water channel aquaporin-0 (AQP0) can be regulated by at least four known mechanisms that we studied over the past decade. The first is irreversible and involves the cleavage of the C-terminal domain of AQP0. The cleavage results in complete pore closure, and AQP0 ceases to act as a water channel. Instead it becomes an adhesive protein mediating cell-to-cell adhesive junctions (Figure 1).
Full-length AQP0 is dynamically modulated by three mechanisms: pH, calcium/calmodulin (Ca2+/CaM), and protein phosphorylation. We recently showed that the binding of Ca2+/CaM to AQP0 results in partial pore closure (Figure 2). The net effect is that the permeability through AQP0 halves in the presence of Ca2+/CaM. Conversely, we showed that phosphorylation of AQP0 by anchored protein kinase A (AKAP2/PKA complex) abolished CaM binding, keeping AQP0 in the open conformation and functioning at maximal activity.
Our studies of channel phosphorylation led us to discover a new protein in the eye lens called AKAP2. Our biochemical and structural studies indicate that AKAPs anchor PKA onto substrate and provide the kinase a sphere of action in which the kinase could phosphorylate substrates in a cAMP-independent way. This exciting observation helps explain how fast phosphorylation can occur, as seen, for example, in heart cells. Moreover, we showed that inhibition of phosphorylation of AQP0 in the lens results in cataract formation. Essentially, we recapitulated the lens disease ex vivo by inhibiting protein phosphorylation (Figure 3).
Membrane protein complexes. Our structure of the AQP0/CaM complex is the first for any full-length membrane channel in complex with this ubiquitous secondary messenger (Figure 4). Current efforts in the laboratory are focused on understanding how Ca2+/CaM binds to and modulates the activity of other channels such as ion channels.
We are also trying to understand more about the AQP-AKAP system. In particular, we are trying to assemble the AQP2/AKAP18/PKA complex and AQP0/AKAP2/PKA complex for structural studies. Intrinsically disordered regions of proteins are widespread in nature, yet the mechanistic roles they play in biology are underappreciated. Such disordered segments can act simply to link functionally coupled structural domains or they can orchestrate enzymatic reactions through a variety of allosteric mechanisms. The regulatory subunits of PKA provide an example of this important phenomenon where functionally defined and structurally conserved domains are connected by intrinsically disordered regions of defined length with limited sequence identity. Our studies show that this seemingly paradoxical amalgam of order and disorder permits fine-tuning of local protein phosphorylation events. The anchoring of PKA by AKAP affords the kinase a sphere of action in which multiple targets can be phosphorylated fast in a cAMP-independent way (Figure 5).
Membrane Transporters Involved in Nutrient Uptake
Sugar uptake. The major facilitator superfamily (MFS) of membrane proteins is the largest collection of structurally related membrane proteins that transport a wide array of substrates. The proton-coupled sugar transporter XylE is the first member of the MFS that has been captured and structurally characterized in multiple transporting conformations, including both the outward and inward facing states. We determined the crystal structure of XylE in a new inward-facing open conformation. Structural comparison of XylE in this conformation with its outward-facing, partially occluded conformation reveals how this transporter functions through a nonsymmetrical rocker-switch movement of the N-terminal domain as a rigid body and the C-terminal domain as a flexible body. Molecular dynamics simulations were employed to help describe how XylE transitions in a lipid membrane to facilitate sugar transport (Figure 6).
Nitrogen uptake. Nitrate is the preferred nitrogen source for plants on which all higher forms of life ultimately depend. Plants and microorganisms evolved to assimilate nitrate efficiently. Despite decades of effort, no structure was available for any nitrate transport protein. The mechanism by which nitrate is transported remained largely obscure until we reported the structure of the bacterial nitrate/nitrite transport protein NarK, from Escherichia coli, with and without substrate. The structures revealed a positively charged substrate-translocation pathway lacking protonatable residues, suggesting that NarK functions as a nitrate/nitrite exchanger and that protons are unlikely to be cotransported. Conserved arginine residues form the substrate-binding pocket, which is formed by association of helices from the two halves of NarK. We identified key residues that are important for substrate recognition and transport and related them to extensive mutagenesis and functional studies. We proposed that NarK exchanges nitrate for nitrite by a rocker-switch mechanism facilitated by interdomain H-bond networks (Figure 7).
Method Development in Cryo-EM
Fragment-based phase extension. In electron crystallography, membrane protein structure is determined from two-dimensional (2D) crystals where the protein is embedded in a membrane. Once large and well-ordered 2D crystals are grown, one of the bottlenecks in electron crystallography is the collection of image data to directly provide experimental phases to high resolution. We developed a new approach to bypass this bottleneck, eliminating the need for high-resolution imaging. We used the strengths of electron crystallography to rapidly obtain accurate experimental-phase information from low-resolution images and accurate high-resolution amplitude information from electron diffraction. The low-resolution experimental phases were used for the placement of α-helix fragments and extended to high resolution using phases from the fragments. Phases were improved by density modifications, followed by fragment expansion and structure refinement against the high-resolution diffraction data. Using this approach, we determined structures of three membrane proteins rapidly and accurately to atomic resolution without high-resolution image data (Figure 8).
MicroED: three-dimensional electron crystallography of protein microcrystals. We demonstrated that it is feasible to determine high-resolution protein structures by electron crystallography of three-dimensional crystals in a cryo-electron microscope (cryo-EM). Lysozyme microcrystals were frozen on an electron microscopy grid, and electron diffraction data were collected to 1.7 Å resolution. We developed a data collection protocol to collect a full-tilt series in electron diffraction to atomic resolution. A single tilt series contains up to 90 individual diffraction patterns collected from a single crystal with tilt angle increment of 0.1–1° and a total accumulated electron dose less than 10 electrons per angstrom squared. We indexed the data from three crystals and used them for structure determination of lysozyme by molecular replacement followed by crystallographic refinement to 2.9 Å resolution (Figure 9).
In 2014, we further enhanced the MicroED method. First, we developed continuous rotation, an improved data collection protocol for MicroED. Microcrystals are continuously rotated during data collection, yielding more complete data and allowing data processing with the crystallographic software tool MOSFLM, resulting in higher resolution for the model protein lysozyme to 2.5 Å.
Second, we used the improved MicroED protocols for data collection and analysis to determine the structure of catalase. Bovine liver catalase crystals that were only ~160 nm thick were used for the structure analysis. A single crystal yielded data to 3.2 Å resolution, enabling rapid structure determination. Current efforts include new phasing methods, automation, and program development. These reports pave the way for the broad implementation and application of MicroED in structural biology.
Computational Design of Genetically Encoded Self-Assembling Cages and Membrane Channels
In collaboration with David Baker (HHMI, University of Washington), we are designing genetically encoded self-assembling proteins for cellular microcircuitry.
We are developing general computational methods for designing proteins that self-assemble to a desired symmetric architecture. Protein-building blocks are docked together symmetrically to identify complementary packing arrangements, and low-energy protein-protein interfaces are then designed between the building blocks to drive self-assembly. Here we use trimeric protein-building blocks to design a 24-subunit, 13 nm diameter complex with octahedral symmetry and two related variants of a 12-subunit, 11 nm diameter complex with tetrahedral symmetry. The designed proteins assembled to the desired oligomeric states in solution, and crystal structures of the complexes revealed that the resulting materials closely match the design models. The method can be used to design a wide variety of self-assembling protein nanomaterials (Figure 10). So far we have developed nanomaterials with tetrahedral, octahedral, helical, and icosahedral symmetries.
As of October 21, 2014