Nicolas Coudray

Inward-facing conformation of the zinc transporter YiiP revealed by cryoelectron microscopy

YiiP is a dimeric Zn2+/H+ antiporter from Escherichia coli belonging to the cation diffusion facilitator family. We used cryoelectron microscopy to determine a 13-Å resolution structure of a YiiP homolog from Shewanella oneidensis within a lipid bilayer in the absence of Zn2+. Starting from the X-ray structure in the presence of Zn2+, we used molecular dynamics flexible fitting to build a model consistent with our map. Comparison of the structures suggests a conformational change that involves pivoting of a transmembrane, four-helix bundle (M1, M2, M4, and M5) relative to the M3-M6 helix pair. Although accessibility of transport sites in the X-ray model indicates that it represents an outward-facing state, our model is consistent with an inward-facing state, suggesting that the conformational change is relevant to the alternating access mechanism for transport. Molecular dynamics simulation of YiiP in a lipid environment was used to address the feasibility of this conformational change. Association of the C-terminal domains is the same in both states, and we speculate that this association is responsible for stabilizing the dimer that, in turn, may coordinate the rearrangement of the transmembrane helices.

Sparse and incomplete factorial matrices to screen membrane protein 2D crystallization.

Electron crystallography is well suited for studying the structure of membrane proteins in their native lipid bilayer environment. This technique relies on electron cryomicroscopy of two-dimensional (2D) crystals, grown generally by reconstitution of purified membrane proteins into proteoliposomes under conditions favoring the formation of well-ordered lattices. Growing these crystals presents one of the major hurdles in the application of this technique. To identify conditions favoring crystallization a wide range of factors that can lead to a vast matrix of possible reagent combinations must be screened. However, in 2D crystallization these factors have traditionally been surveyed in a relatively limited fashion. To address this problem we carried out a detailed analysis of published 2D crystallization conditions for 12 β-barrel and 138 α-helical membrane proteins. From this analysis we identified the most successful conditions and applied them in the design of new sparse and incomplete factorial matrices to screen membrane protein 2D crystallization. Using these matrices we have run 19 crystallization screens for 16 different membrane proteins totaling over 1300 individual crystallization conditions. Six membrane proteins have yielded diffracting 2D crystals suitable for structure determination, indicating that these new matrices show promise to accelerate the success rate of membrane protein 2D crystallization.

Structure of the SLC4 transporter Bor1p in an inward-facing conformation.

Bor1p is a secondary transporter in yeast that is responsible for boron transport. Bor1p belongs to the SLC4 family which controls bicarbonate exchange and pH regulation in animals as well as borate uptake in plants. The SLC4 family is more distantly related to members of the Amino acid‐Polyamine‐organoCation (APC) superfamily, which includes well studied transporters such as LeuT, Mhp1, AdiC, vSGLT, UraA, SLC26Dg. Their mechanism generally involves relative movements of two domains: a core domain that binds substrate and a gate domain that in many cases mediates dimerization. To shed light on conformational changes governing transport by the SLC4 family, we grew helical membrane crystals of Bor1p from Saccharomyces mikatae and determined a structure at ∼6 Å resolution using cryo‐electron microscopy. To evaluate the conformation of Bor1p in these crystals, a homology model was built based on the related anion exchanger from red blood cells (AE1). This homology model was fitted to the cryo‐EM density map using the Molecular Dynamics (MD) Flexible Fitting method and then relaxed by all‐atom MD simulation in explicit solvent and membrane. Mapping of water accessibility indicates that the resulting structure represents an inward‐facing conformation. Comparisons of the resulting Bor1p model with the X‐ray structure of AE1 in an outward‐facing conformation, together with MD simulations of inward‐facing and outward‐facing Bor1p models, suggest rigid body movements of the core domain relative to the gate domain. These movements are consistent with the rocking‐bundle transport mechanism described for other members of the APC superfamily.

Deducing the symmetry of helical assemblies: Applications to membrane proteins

Helical reconstruction represents a convenient and powerful approach for structure determination of macromolecules that assemble into helical arrays. In the case of membrane proteins, formation of tubular crystals with helical symmetry represents an attractive alternative, especially when their small size precludes the use of single-particle analysis. An essential first step for helical reconstruction is to characterize the helical symmetry. This process is often daunting, due to the complexity of helical diffraction and to the low signal-to-noise ratio in images of individual assemblies. Furthermore, the large diameters of the tubular crystals produced by membrane proteins exacerbates the innate ambiguities that, if not resolved, will produce incorrect structures. In this report, we describe a set of tools that can be used to eliminate ambiguities and to validate the choice of symmetry. The first approach increases the signal-to-noise ratio along layer lines by incoherently summing data from multiple helical assemblies, thus producing several candidate indexing schemes. The second approach compares the layer lines from images with those from synthetic models built with the various candidate schemes. The third approach uses unit cell dimensions measured from collapsed tubes to distinguish between these candidate schemes. These approaches are illustrated with tubular crystals from a boron transporter from yeast, Bor1p, and a β-barrel channel from the outer membrane of E. coli, OmpF.

Structure of the Borate Transporter Bor1p by cryo-EM

Boron is an essential micronutrient for plants and animals. Transport of boron regulates uptake and protects against high levels of boron in plants and yeast. The yeast boron transporter Bor1p belongs to the SLC4 Anion Exchanger family of transporters that also includes the human chloride/bicarbonate exchanger. We have overexpressed and purified the Bor1p orthologue from Saccharomyces mikatae using Saccharomyces cerevisiae as a host. Helical crystals were produced by reconstituting the transporter together with cardiolipin and electron micrographs of frozen-hydrated tubes were recorded with a field emission gun at 200 kV using a direct electron detector. Crystals adopted two different helical symmetries from which two independent reconstructions were produced from 75 particles at a resolution of ∼7 A using a Fourier-Bessel approach. In order to interpret the dimeric densities revealed by the reconstructions, we built a homology model for Bor1p based on the crystal structure of the uracil transporter (UraA), which is the closest relative to Anion Exchangers within the APC superfamily of transporters. Initial homology models were produced from 10 alternative alignments of these distantly related proteins. The alignments were refined by mapping sequence conservation onto these 3D structures. The best model was then fitted to the cryo-EM map by adjusting the position of individual helices and the resulting structure was equilibrated in a lipid bilayer for 200 ns using molecular dynamics. The resulting structure suggests conformational changes relative to UraA in which helices at the dimer interface are tilted relative to the transporter domain, thus providing access to the substrate binding site from the extracellular side of the membrane. Comparison of our Bor1p structure with that from UraA are likely to reflect the structural changes that accompany the alternating access mechanism employed by this family of transporters.

Classification and mutation prediction from non–small cell lung cancer histopathology images using deep learning

Visual inspection of histopathology slides is one of the main methods used by pathologists to assess the stage, type and subtype of lung tumors. Adenocarcinoma (LUAD) and squamous cell carcinoma (LUSC) are the most prevalent subtypes of lung cancer, and their distinction requires visual inspection by an experienced pathologist. In this study, we trained a deep convolutional neural network (inception v3) on whole-slide images obtained from The Cancer Genome Atlas to accurately and automatically classify them into LUAD, LUSC or normal lung tissue. The performance of our method is comparable to that of pathologists, with an average area under the curve (AUC) of 0.97. Our model was validated on independent datasets of frozen tissues, formalin-fixed paraffin-embedded tissues and biopsies. Furthermore, we trained the network to predict the ten most commonly mutated genes in LUAD. We found that six of them—STK11, EGFR, FAT1, SETBP1, KRAS and TP53—can be predicted from pathology images, with AUCs from 0.733 to 0.856 as measured on a held-out population. These findings suggest that deep-learning models can assist pathologists in the detection of cancer subtype or gene mutations. Our approach can be applied to any cancer type, and the code is available at https://github.com/ncoudray/DeepPATH.A convolutional neural network model using feature extraction and machine-learning techniques provides a tool for classification of lung cancer histopathology images and predicting mutational status of driver oncogenes

Stalk-mediated communication in the dynein motor domain

The movement of a molecular motor protein along a cytoskeletal track requires communication between enzymatic, polymer-binding, and mechanical elements. Such communication is particularly complex and not well understood in the dynein motor, an ATPase that is comprised of a ring of six AAA domains, a large mechanical element (linker) spanning over the ring, and a microtubule-binding domain (MTBD) that is separated from the AAA ring by a ~135 A coiled-coil stalk. We identified mutations in the stalk that disrupt directional motion, have microtubule-independent hyperactive ATPase activity, and nucleotide-independent low affinity for microtubules. Cryo-electron microscopy structures of a mutant that uncouples ATPase activity from directional movement reveal that nucleotide-dependent conformational changes occur normally in one half of the AAA ring, but are disrupted in the other half. The large-scale linker conformational change observed in the wild-type protein is also inhibited, revealing that this conformational change is not required for ATP hydrolysis. These results demonstrate an essential role of the stalk in regulating motor activity and coupling conformational changes across the two halves of the AAA ring.The movement of a molecular motor protein along a cytoskeletal track requires communication between the ATPase and polymer-binding sites. While these sites are located in close proximity in kinesin and myosin, dynein has a ~135 Å coiled-coil stalk that separates the microtubule-binding domain (MTBD) from the AAA ATPase ring. An analysis of 534 unique dynein sequences revealed that the length of the stalk is highly conserved. A panel of mutants based on our analysis revealed three regions of the stalk for which altering the length resulted in diffusional motion along microtubules. Two diffusive mutants also showed hyperactive ATPase activity, and were no longer sensitive to microtubules. Structural analysis of one of these mutants using cryo electron microscopy revealed a previously uncharacterized open conformation of the AAA ring. This conformation is lowly populated in the wild-type protein and thus may be an on-pathway catalytic intermediate. Our results reveal how changes in the stalk can affect the conformation and activity of the AAA ring, and modulate dynein motility.

Two-Dimensional Crystallization of Membrane Proteins: Screening Strategies

Membrane proteins play fundamental roles in a broad range of biological processes and represent more than half of the current drugs targets [1]. Electron microscopy has been effective in producing high resolution representations, but the low signal-to-noise ratio inherent to the imaging process requires averaging of thousands of perfectly aligned proteins. Two-dimensional (2D) crystallization constitutes an advantage over single particle analysis, as the proteins are already aligned within the crystal. Furthermore, as opposed to X-ray analysis of three-dimensional crystals, the proteins are inserted into a lipid bilayer which makes the reconstitution closer to the native environment.