1260-Plat The Assembly, Structure and Activation of Influenza a M2 Transmem- brane Domain Depends on Lipid Membrane Thickness and Composition Elka R. Georgieva, Haley D. Norman, Peter P. Borbat, Jack H. Freed. Cornell University, Ithaca, NY, USA. The influenza A M2 protein is single-pass transmembrane protein that assem- bles in a tetramer, forming a pH-activated proton channel. It is essential for viral function. The tetramer of M2 transmembrane domain (M2TM), residues 22-46, is the minimal unit for proton conductance. Although substantial knowl- edge about M2-membrane interactions was accumulated, understanding the en- ergy profile of the M2TM assembly and structural alterations upon channel activation in lipid membranes is incomplete. We utilized pulse ESR spectroscopy, DEER, with spin-labeling to study the AM2TM peptide (residues 21-49) in lipid membranes. A unique cysteine residue (L46C), introduced in M2TM, was spin-labeled with MTSL. The peptide was re- constituted in membranes of pure DLPC or DOPC, DLPC:DLPS or DOPC:POPS 85:15 mol%, and DLPC:DLPS:Cholesterol or DOPC:POPS:Cholesterol 65:15:25 mol% at peptide-to-lipid molar ratio 1:500. We measure inter-spin distances and distance distributions to monitor structural alterations in M2TM residing in different lipid environment, and also upon changing the acidity from pH 8 to 5.5. Based on the modulation depth of DEER signals and the reconstructed dis- tance distributions, our results indicate that membrane-associated M2TM exists in various oligomerization states, most likely dimers and tetramers, supposedly relevant to M2 folding. The length of lipid hydrocarbon tail and cholesterol affect M2TM oligomer assembly: The DEER modulation depth for M2TM in DLPC cor- responds to dimers, whereas the increased depth measured in POPC membranes and DLPC:DLPS:Cholesterol suggests a shifted equilibrium to tetramers. The dis- tances in DOPC membranes at both pHs were shorter (maximum at 22 A ˚ ), as ex- pected from a closed M2 conformation. Longer distances, 27-40 A ˚ , due to channel opening at low pH were observed clearly only in the presence of charged lipid, indicating the role of these lipids in the activation of M2 channel. 1261-Plat HSP70 Associates with Phosphatidylserine Membranes via the Peptide Binding Domain Antonio De Maio 1 , Gabrielle Armijo 2 , Victor Lopez 2 , Derek Gonzales 2 , Jonathan Okerblom 2 , Nelson Arispe 3 , David M. Cauvi 4 . 1 Departments of Surgery and Neuroscience, University of California, San Diego, La Jolla, CA, USA, 2 Initiative for Maximizing Student Development (IMSD) Program, University of California, San Diego, La Jolla, CA, USA, 3 Department of Anatomy, Physiology and Genetics, Uniformed Services University, Bethesda, MD, USA, 4 Department of Surgery, University of California, San Diego, La Jolla, CA, USA. The expression of heat shock proteins (hsp) is a natural response to an array of physiological, environmental and clinical stressors. These proteins participate in the repair and recovery from an insult and confer protection from subsequent stresses. The cytoprotective effect of hsp has been associated with their chaperone function within the cytosol. Recently, they have been found in the extracellular environment where they propagate the stress signal to avoid the dissemination of the insult. Hsp70, the major inducible form of the hsp family, does not contain any consensus secretory signal that predicts its export via the classical ER-Golgi secretory pathway. We have proposed that Hsp70 is exported by a novel mecha- nism that is initiated by the translocation of the protein into the plasma membrane. Furthermore, the protein is released associated with extracellular vesicles (ECV), which we speculate result in a robust activation of the immune system. We inves- tigated the mechanism of Hsp70 insertion into membranes using liposomes. We observed that Hsp70 insertion into lipid membranes was spontaneous and specific for negatively charged lipids, such as phosphatidylserine. Using a proteomic approach, we determined that the C-terminus end of the molecule, which contains the peptide binding domain, is inserted into the lipid bilayer. The N-ter- minus of the molecule, containing the ATP binding site, is exposed to the external part of the liposome. These results resemble our initial studies characterizing channel activity of Hsp70 observed in artificial lipid membranes. Supported by NIH R01GM098455. Platform: Computational Methods 1262-Plat Bayesian Structure Determination from Sparse Single Molecule X-Ray Diffraction Images Michal Walczak, Helmut Grubmueller. MPI for Biophysical Chemistry, Goettingen, Germany. X-ray free electron lasers used in single molecule experiments offer new pos- sibilities for molecular structure determination. We propose a Bayesian method capable of extracting structure information from sparse and noisy diffraction images. We investigate two different strategies. In the first, a ‘seed’ model is used to determine the molecular orientation for each of the collected diffraction images, and an improved molecular transform is obtained by averaging those images in three-dimensional reciprocal space. In the second approach, a real space structure model that fits best to the entire set of diffraction images is ob- tained, thus enabling distinction between different structures. We found that the achievable resolution increases with molecular mass as m 1/6 , which somehow unexpectedly suggests that, at a given resolution level, struc- ture determination is more challenging for small molecules. As a proof of concept, we have computed the electron density for a glutathione (molecular mass 307 Da) from 20,000 synthetic diffraction images, each with 82 recorded elastically scattered photons, and up to 50% additional background noise. Alternatively, and demonstrating the feasibility of the second approach, the structure of the same molecule was also determined in a Monte Carlo refine- ment simulation starting from random conformations. Further, the second approach is exemplified for a ribosomal structure (molecular mass about 2.5 MDa). Our results show that it is possible to distinguish between minute struc- tural changes associated with tRNA translocation. Overall, our results suggest that the proposed method allows for structure deter- mination at atomic resolution from sparse and noisy X-ray diffraction images in single molecule experiments for a broad spectrum of molecular masses. 1263-Plat xMDFF: Molecular Dynamics Flexible Fitting of Low-Resolution X-Ray Structures Abhishek Singharoy 1 , Ryan McGreevy 1 , Qufei Li 2 , Jingfen Zhang 3 , Eduardo Perozo 4 , Klaus Schulten 1 . 1 Beckman Institute, Univeristy of Illinois, Urbana Champaign, IL, USA, 2 1Department of Biochemistry and Molecular Biology, Univeristy of Chicago, Chicago, IL, USA, 3 Informatics, Univeristy of Missourie, Columbia, MO, USA, 4 Department of Biochemistry and Molecular Biology, Univeristy of Chicago, Chicago, IL, USA. X-ray crystallography remains one of the most versatile and dominant methods for solving the all-atom structure of biomolecules. However, for relatively large systems, availability of only medium to low-resolution diffraction data often limits the determination of atomic structures. We have developed a new molec- ular dynamics flexible fitting (MDFF)-based approach, xMDFF, for determining structures from such low-resolution crystallographic data. xMDFF employs a real-space refinement scheme that flexibly fits atomic models into an iteratively updating electron density map. The iterations continue until the fitted structure yields R-factors lower than a predefined tolerance. xMDFF addresses significant large-scale structural deformations of the initial model to fit the low-resolution density, as has been tested with synthetic low resolution maps of D ribose bind- ing protein. xMDFF has been successfully applied to re-refine six low-resolution protein structures of varying sizes that were already submitted to the PDB. An improvement in the R-factors is observed, and sterically and conformationally favored atomic geometries are achieved in all the six cases. Finally, via systematic refinement of a series of data form 3.6 to 7 A ˚ , xMDFF together with electrophysiology experi- ments confirmed the first all-atom structure of a voltage sensing protein Ci-VSP. 1264-Plat i-ATTRACT: a New Flexible Docking Approach for Investigating Protein Protein Interactions Christina Schindler, Martin Zacharias. Technische Universita ¨t Mu ¨nchen, Physics Department (T38), Garching, Germany. Many of the most important processes in the cell are carried out by large molec- ular machines built up from multiple proteins. However, structural data for a large fraction of known and putative complexes is still lacking. Computational docking methods aim at predicting protein complexes based on the structure of the individual constituents. A new protein-protein docking approach, i- ATTRACT has been evaluated on a large benchmark. The docking combines rigid body degrees of freedom and fully flexible interface residues in a simulta- neous potential energy minimization. To our knowledge this is the first docking method performing an energy minimization in degrees of freedom of multiple scale. Procedures combining Monte Carlo sampling and energy minimization were applied as well. Refinement of rigid body docking solutions from a system- atic search with unbound protein structures using ATTRACT [1] shows prom- ising results on a large number of cases. i-ATTRACT is able to significantly improve results for initial structural deviations of up to 8 A ˚ from bound geom- etries. Compared to molecular dynamics this refinement procedure comes at low computational cost but shows more efficient sampling by combining small-scale conformational rearrangements and large-scale center-of-mass displacements. Monday, February 17, 2014 249a