Skip to Main Content
From Computational Biophysics to Systems Biology (CBSB12)
June 3-5, 2012  Knoxville, Tennessee
ORNL link
OU link
UT link
ORNL link

Sponsored by:
ORNL link
Participants

Emal Alekozai, Poster
Bachir Aoun, Poster
Jerome Baudry Invited Speaker
Workalemahu Berhanu, Poster
Nicholas Bodmer, Contributed Speaker
Jodian Brown, Contributed Speaker
Virginia Burger, Award Speaker
Nicholas Callahan, Poster
Derek Cashman, Poster
Xiaolin Cheng Invited Speaker
Mihir Date, Poster
Samuel DeLuca, Poster
Stephanie DeLuca, Contributed Speaker
Igor Drobnak, Contributed Speaker
Sally Ellingson, Poster
Jun Fan, Contributed Speaker
Dennis Glass
Garrett Goh, Award Speaker
Ewa Golas, Award Speaker
Hao-Bo Guo, Poster
Russel Hanson, Poster
Jason Harris, Poster
Andrew Hirsch, Poster
Liang Hong, Poster
Jun Hu
Ping Jiang, Poster
Alexander Johs, Poster
Sigurdur Jonsson, Contributed Speaker
Karan Kapoor
Andrea Kravats, Poster
Pawel Krupa, Poster
Yongjin Lee
Ron Levy, Keynote
Jianing Li, Poster
Peng Lian, Poster
Benjamin Lindner, Poster
Yanxin Liu, Poster
Gia Maisuradze, Contributed Speaker
Buddhadev Maiti, Poster
Jens Meiler, Invited Speaker
Yinglong Miao, Poster
Julie Mitchell, Invited Speaker
Barmak Mostofian
Magdalena Mozolewska, Poster
Sayak Mukherjee, Poster
Hai Nguyen, Poster
Gungor Ozer, Poster
Jerry Parks
Loukas Petridis
Medeliene Pincu, Poster
Xianghong Qi, Poster
Arvind Ramanathan
Demian Riccardi
Amit Roy, Contributed Speaker
Celeste Sagui, Invited Speaker
Reza Salari, Contributed Speaker
Amandeep Sangha
Andrej Savol, Poster
Michael Schneiders, Poster
Roland Schulz, Poster
Ester Sesmero, Poster
Piotr Setny, Poster
Jonathan Sheehan
Jana Shen, Invited speaker
Jeremy Smith
John Straub, Invited Speaker
Collin Stultz, Invited Speaker
Panourios Tamamis, Award Speaker
Kelly Theisen, Poster
Sam Tonddast-Navaei, Poster
Sahin Uyaver, Poster
Sandor Vajda, Invited Speaker
Arjan van der Vaart, Invited Speaker
Gregory Voth, Keynote
Huan Wang, Poster
Jeff Wereszczynski, Contributed Speaker
Thomas Wireki, Poster
Qin Xu
Xiaojun Xu, Poster
Chunli Yan, Award Speaker
Jianzhuang Yao
Fatih Yasar, Poster
Zheng Yi, Poster
Hang Yu, Poster
He Zhang, Poster
Zhe Zhang, Poster


Abstracts:

Keynotes and Invited Speakers

  • Exploring Landscapes for Protein Folding and Binding Using Replica Exchange Dynamics, Kinetic Networks and Markov State Models
    Ronald M. Levy

    Department of Chemistry and Chemical Biology, and BioMaPS Institute for Quantitative Biology, Rutgers University

    Advances in computational biophysics depend critically on the development of accurate effective potentials and powerful sampling methods to traverse the rugged energy landscapes that govern protein folding, binding and fitness. I will review work in my lab over the last few years concerning the construction of all-atom effective potentials for proteins and multi-scale methods for simulating their folding and binding on long time scales. Replica exchange (RE) is a generalized ensemble molecular simulation method for accelerating the exploration of free-energy landscapes which define many challenging problems in computational biophysics, including protein folding and binding. We have clarified some of the obstacles to obtaining converged thermodynamic information from RE simulations. I will discuss these issues and new multi-scale approaches to recover protein folding rates and pathways for folding and binding using the combined power of replica exchange, kinetic network models with flux analysis, and effective stochastic dynamics.
    Back to the top of the page

  • Theory and Simulation of Biomolecular Systems: Surmounting the Challenge of Bridging the Scales
    Gregory A. Voth

    Department of Chemistry, James Franck Institute, Institute for Biophysical Dynamics, and Computation Institute, University of Chicago

    A multiscale theoretical and computational methodology will be presented for studying biomolecular systems across multiple length and time scales. The approach provides a systematic connection between all-atom molecular dynamics, coarse-grained modeling, and mesoscopic phenomena. At the heart of the approach is a method for deriving coarse-grained models from protein structures and their underlying molecular-scale interactions. This particular aspect of the work has strong connections to the theory of renormalization, but it is more broadly developed and implemented for heterogeneous systems. A critical component of the methodology is also its connection to experimental structural data such as cryo-EM or x-ray, thus making it “hybrid” in its character. Applications this overall multiscale approach to study key features of large multi-protein complexes such the HIV-1 virus capsid, the entire HIV-1 immature virion, actin filaments, and protein-mediated membrane remodeling will be presented as time allows.
    Back to the top of the page

  • Back to the top of the page

  • Simulating the transport of a cellulose chain through the cellulase catalytic tunnel
    Xiaolin Cheng

    Department of Molecular Biophysics, Oak Rifge National Laboratory

    The understanding of emerging collective behaviors in biomolecular complexes represents a major challenge in modern biophysics. As a first step toward the study of such processes we have applied multi-resolution nonlinear dimensionality reduction and diffusion analysis to obtain reliable low-dimensional representations and models for the dynamics of apparently high-dimensional complex systems such as proteins in a biological environment. The results clearly show that the proposed methods can efficiently find low-dimensional representations of complex processes such as protein folding, and suggest strategies to simplify significantly the study of such processes.
    Back to the top of the page

  • A Computational Platform for Protein Structure Determination Integrating Limited Experimental Data
    Jens Meiler

    Computational Chemical and Structural Biology, Vanderbilt University

    I will present BCL::Fold, a new algorithm for de novo prediction of complex and large protein topologies by assembly of secondary structure elements. The method was designed to integrate experimental data from NMR, EPR, EM, SAXS experiments, or combinations thereof. These data sets often provide more readily restraints for regions of defined secondary structure. I will present examples for atomic-detail structure elucidation from medium resolution cryo-EM density maps (Figure 1) and paramagnetic restraints from NMR spectroscopy. Briefly: Computational de novo protein structure prediction is limited to small proteins of simple topology. The present work explores an approach to extend beyond the current limitations through assembling protein topologies from idealized α-helices and β-strands. The algorithm performs a Monte Carlo Metropolis simulated annealing folding simulation. It optimizes a knowledge-based potential that analyzes radius of gyration, β-strand pairing, secondary structure element packing, amino acid pair potential, amino acid environment, and loop closure. Discontinuation of the protein chain favors sampling of non-local contacts and thereby creation of complex protein topologies. The folding simulation is accelerated through exclusion of flexible loop regions further reducing the size of the conformational search space. The algorithm is benchmarked on 66 proteins with lengths between 83 and 293 amino acids. For 61 out of these proteins the best SSE-only models obtained have an RMSD100 below 8.0 Å and recover more than 20% of the native contacts. The algorithm assembles protein topologies with up to 215 residues and a relative contact order of 0.46. BCL::Fold includes a modified scoring function for the assembly of membrane proteins. BCL::Fold is typically combined with Rosetta refinement algorithms to arrive at proteins models accurate at atomic detail.
    Back to the top of the page

  • Knowledge-Based Structural Approaches for Predicting Hot Spots of Protein Binding and Allostery
    Julie Mitchell

    Departments of Mathematics and Biochemistry, University of Wisconsin - Madison

    Using information derived from protein structures, it is possible to predict amino acid positions where mutations will have a deleterious effect on protein binding or allosteric communication. The KFC2 model captures 80% of alanine scanning mutagenesis hot spots, which result in a binding energy increase of at least 2 kcal/mol. A unique feature of the model is a local plasticity feature that suggests whether a change in sequence can be accommodated through local sidechain rearrangements. A different plasticity measure, known as local structural entropy, is a dominant feature in our AlloSIND model for allosteric hot spots that lie between the effector and active sites of allosteric proteins. One possible interpretation is that rigidity of internal protein secondary structure prevents an allosteric protein from absorbing the impact of effector binding locally, resulting in longer range conformation effects.
    Back to the top of the page

  • Free energy landscapes of polyproline and polyglutamine peptides
    Celeste Sagui

    Department of Physics, North Carolina State University

    We use enhanced sampling techniques (the Adaptively Biased Molecular Dynamics (ABMD) method, multiple walkers, replica exchange, steered MD, and various combinations thereof) to study peptide systems whose conformational space cannot be sampled with regular MD simulations. These include transitions between the PPII and PPI forms of polyproline (polyP); proline-rich guest/host peptides; polyglutamine (polyQ), and polyQ-polyP systems. Several statistical techniques allow us to explore the atomic mechanisms that underlie various experimental observations: the apparent PPII propensity of guest amino acids in polyP-rich peptides; the properties that may favor aggregation in polyQ systems; and the suppression of aggregation of polyQ by the addition of a C-terminal polyP peptide.
    Back to the top of the page

  • Role of hydration and confinement in protein folding and aggregation
    John E. Straub

    Department of Chemistry, Boston University

    Reverse micelles provide an environment in which the number of water molecules and overall cavity size may be "tuned," by adjusting the water-to-surfactant ratio, allowing, in principle, the role of hydration and confinement on protein folding and aggregation to be systematically studied. We have used molecular dynamics simulations to explore the structure and dynamics of the alanine-rich AKA2 peptide and aggregation-prone fragments of amyloid proteins in reverse micelle environments. The dependence of the peptide-micelle interaction on capping of the N- and C-termini and the nature of the force field is explored. The time scales for fluctuations in the reverse micelle shape and surface area are characterized and compared with the results of more idealized spherical micelle models. The results suggest that an understanding of the detailed nature of protein-surfactant interactions can be essential to the interpretation of studies of protein folding and aggregation in reverse micelle environments.
    Back to the top of the page

  • "From Computational Docking to Exploration of Biochemical Pathways
    Jerome Baudry

    Department of Biochemistry & Cellular and Molecular Biology, University of Tennessee, Knoxville

    Virtual (in-silico) docking of small molecules in protein targets is a popular and successful approach to discover molecules capable of binding in proteins. We describe how docking is used to investigate biochemical pathways, focusing on how P450s can turn environmental molecules into estrogenic pollutants by increasing their binding to the estrogen receptor alpha target. We also describe ongoing developmental work to use supercomputing architectures efficiently to perform massive docking of very large chemical databases against a large number of protein targets.
    Back to the top of the page

  • Modeling disordered states of Proteins: Are structural models of the unfolded state correct?
    Collin Stultz

    EECS & HST, Massachusetts Institute of Technology

    The characterization of intrinsically disordered proteins is challenging because accurate models of these systems require a description of both their thermally accessible conformers and the associated relative stabilities or population weights. These structures and weights are typically chosen such that calculated ensemble averages agree with some set of pre-specified experimental measurements; however, the large number of degrees of freedom in these systems typically leads to multiple conformational ensembles that are degenerate with respect to any given set of experimental observables. Moreover, our recent work demonstrates that estimates of the relative stabilities of conformers within an ensemble are often incorrect when one does not account for the underlying uncertainty in the estimates themselves. Therefore, we have developed a method for modeling the conformational properties of disordered proteins that estimates the uncertainty in the weights of each conformer. A unique and powerful feature of the approach is that it provides a built-in error measure that allows one to assess the accuracy of the ensemble. Using this approach we constructed an ensemble that characterizes the accessible states of the IDPs, tau protein, alpha-synuclein and abeta; i.e., proteins that play a role in several neurodegenerative disorders. These data led to new insights into intramolecular interactions that may play a role in promoting IDP self-association - a process which has been linked to neuronal death and dysfunction in patients with Alzheimer's disease. We further demonstrate that these data may be used in the initial stages of a strategy to design ligands that prevent IDP aggregation. More generally, we derive an order parameter that quantifies the extent of disorder within a protein. Although protein disorder is normally thought of as a binary phenomenon (i.e., a protein is either disordered or not), we suggest that the concept of protein disorder should be treated like a continuous variable, and that not all unfolded states are created equal.
    Back to the top of the page

  • Recent development and application of continuous constant pH molecular dynamics
    Jana Shen

    Chemistry, University of Oklahoma

    Development of the constant pH molecular dynamics techniques has opened a door to atomically detailed studies of dynamic processes coupled to protonation/deprotonation. Here we discuss the most recent development of the continuous constant pH molecular dynamics (CpHMD) technique and application studies for gaining novel insights into ionization-coupled conformational phenomena in biology and chemistry. We show that CpHMD simulations offer, for the first time, thermodynamic description of coupled protonation and conformational equilibria for proteins. We will also discuss other applications such as pH titration of micelles and pH-dependent phase transitions.
    Back to the top of the page

  • Enhanced sampling simulations of DNA and protein-DNA complexes
    Arjan van der Vaart

    Department of Chemistry, University of South Florida

    The flexibility of long DNA is well-described by the worm-like chain model, but at the small-length scales this model breaks down. Moreover, several experiments suggest that short DNA has an increased flexibility. We performed enhanced sampling simulations of short DNA strands to assess its flexibility. Our results indicate that the stiffness of DNA decreases upon strong bending. We will also discuss enhanced sampling simulations to show how proteins and ligands modulate the DNA structure.
    Back to the top of the page

  • High throughput identification and druggability analysis of protein binding sites
    Sandor Vajda

    Department of Biomedical Engineering and Chemistry, Boston University

    Our lab has developed the computational solvent mapping method of determining binding hot spots of proteins (1). The method globally samples the surface of a target protein using small organic molecules as probes, finds favorable positions, clusters the conformations, and ranks the clusters on the basis of the average energy. The regions that bind several probe clusters predict the binding hot spots, in good agreement with experimental results. Solvent mapping can be used to solve two important problems. First, it achieves higher accuracy than any other method in the identification of ligand binding sites on unbound protein structures (2). Second, the mapping results provide information for assessing druggability, i.e., the ability of a protein to bind drug sized ligands with high affinity (3). While both applications were very successful, they required lengthy calculations and hence were originally demonstrated on small benchmark sets. We describe our work toward modifying the methods such that applications to large sets of proteins would become computationally feasible, enabling some general conclusions on the binding properties of proteins.
    Back to the top of the page

Award Speakers

  • Characterizing Pathways between Metastable States of Intrinsically Disordered Proteins
    Virginia Burger

    Computational Biology, University of Pittsburg

    Intrinsically disordered proteins (IDPs) lack well-definedsecondary and tertiary structure. These proteins makeup about 50% of signaling proteins and are implicated indiseases such as cancer and neurodegenerativedisorders. IDPs often undergo synergistic folding uponbinding, but range from being completely random coil tohaving a small degree of residual structure whileunbound. Capturing the unbound state of IDPs is aparticular challenge for traditional molecular dynamicssimulations. Increasingly, atomistic simulations canadequately sample the conformational landscape ofsmall peptides, but analysis techniques forcharacterization of metastable substates in IDPs arelacking. Here, we present a novel method for multi-scalespatio-temporal organization of long time-scale IDPsimulation data. We demonstrate our method on the nuclear co-activatorbinding domain (NCBD) of CBP (CREB binding protein).NCBD is a 59-residue molten globule peptide containingthree helices which unfold and refold to form diversebound configurations. Six experimental structures of thispeptide, both bound to co-activators and unbound, havebeen determined with three distinct folds. Tocharacterize the conformational space of NCBD, wegenerated 40-microseconds of all-atom explicit waterMD simulation of unbound NCBD at equilibrium, whichcomes within 1 - 4 Angstroms of each experimentalstructure. We clustered the simulated conformers intoconformationally similar states using a hierarchicalrandom-walk based grouping scheme. Analogous to Markov State Models, we use spectral methods to groupconformational states into kinetically similar metastablestates. However, here anharmonic conformationalanalysis is applied to obtain energetically coherentconformational states, and the number of states isdetermined implicitly by the landscape. The hierarchicalnature of our clustering scheme demonstratesmulti-scale conformational similarities, in that finelyseparated clusters with fast interconversions areidentified at early hierarchy levels, while coarselygrouped conformations with slow interconversions areprovided by later hierarchy levels. To characterize thetransition pathways between the clusters, we identifykinetic bottlenecks and short time-scale transitionsbetween the distinct conformational states using spectralperturbation theory. We find pathways and ratesdescribing the transition of unbound NCBD between itsconformational states and identify rate-limiting stepsseparating the six experimental conformations. Thismulti-scale spatio-temporal analysis can be used todescribe the conformational space of IDPs in general,e.g. for identifying druggable metastable states.
    Back to the top of the page

  • pH-coupled simulations of RNA in explicit solvent
    Garrett Goh

    Chemistry, University of Michigan

    From serving as catalytic residues in ribozymes totriggering pH-dependent conformation changes, the roleof protonated nucleotides in modulating RNA dynamicsand function is fast emerging as one of the keyunanswered questions in the study of RNA biology.Conventional molecular dynamics (MD) methods canonly partially address these questions because theprotonation states are modeled as fixed states, whichare reliant on prior knowledge of the identity and pKavalues of key residues. With the work presented here,we establish the framework and demonstrate the firstconstant pH MD simulations (CPHMD^MSλD) ofnucleic acids in explicit solvent, where the protonationstates are modeled as dynamic variables that arecoupled to the dynamics of the RNA viaλ-dynamics. We adopted a recently developedfunctional form, λNexp, for λ within theframework of multi-site λ-dynamics (MSλD)which improves sampling efficiency in the λ space10-fold over existing explicit solvent CPHMD methods.Calculated pKa values of simple nucleotides are in anexcellent agreement with experiment, with an averageunsigned error of 0.24 pKa units. Using lead-dependentribozyme as a model RNA system, we show thatCPHMD^MSλD simulations accurately reproducethe direction of pKa shifts and provide reasonablyaccurate recapitulation of experimental pKa values,demonstrating the potential for applying this approach toexplore pH-dependent processes in complex RNAmolecules. We anticipate that CPHMD^MSλDsimulations will be used as a powerful tool applied inconjunction with existing experimental techniques toinvestigate various pH-dependent phenomena of nucleic acids.
    Back to the top of the page

  • Molecular Dynamics of the Hsp70 Chaperone in response to nucleotide and substrate: a coarse-grained perspective
    Ewa Golas

    Molecular Modeling Group, Theoretical Chemistry, University of Gdansk, Poland

    The 70KDa heat-shock (Hsp70) proteins form a class ofchaperones recognized for their diverse and essentialroles in the domain of protein repair, folding assistance,and agglomerate prevention. The present studyexamines the mechanism of Hsp70 function viamolecular dynamics, employing the coarse-grainedUNRES model and forcefield in a series of canonicalLangevin molecular dynamics simulations. The effect ofimplicit nucleotide in the nucleotide-binding domain(NBD) and substrate in the substrate-binding domain(SBD) was investigated for the Hsp70 from E. Coli,DNAK (pdb 2KHO). Three binding pathways of the SBDto the NBD were observed, along with a dominatingeffect of substrate. Implications of both nucleotide andsubstrate towards the Hsp70 cycle are discussed.
    Back to the top of the page

  • Development of Regulatory Compounds for the Complement System by MD Simulations and Experiments
    Phanourios Tamamis

    Department of Physics, University of Cyprus, Cyprus

    The complement system (CS) provides the first line of defense against the invasion of foreign pathogens. Nevertheless, its inappropriate or excessive activation may cause or aggravate several pathological conditions, such as age-related macular degeneration (AMD). Therefore, the development of drugs tackling chemical pathways for the control of inappropriate or excessive complement activation is of significant medical interest. Compstatin and PMX53 are two peptide-based promising therapeutic compounds as they bind, respectively, on key proteins C3 and C5aR of CS and inhibit complement activation. Simulations and experiments have been employed to optimize the binding efficacy and inhibitory activity of both compounds. Using Molecular Dynamics (MD) simulations, we have recently suggested an interpretation for the species specificity of Compstatin [1] (its activity against human C3 and inactivity against lower-mammal C3), and succeeded in designing a modified “transgenic” mouse protein aiming at testing AMD disease models in non-primates [2]. Furthermore, using a combination of de novo drug design and MD simulations we recently proposed new compstatin analogs with optimized binding affinity and solubility, relative to known compstatin analogs [3]. The most promising compounds constitute the most potent inhibitors in completed and ongoing experimental studies. In addition, we have recently implemented a combined protocol, consisting of ligand-docking, implicit-membrane MD simulations and binding free energy calculations, to generate and assess an exhaustive ensemble of structural models for the complex between the key GPCR protein of the complement-system C5aR and its most potent antagonist, hexapeptide PMX53. The most promising complex contains interactions in line with available experimental data from site-directed mutagenesis [4] and provides insights into PMX53 residues which confer antagonist activity. Novel compounds based on PMX53 are now computationally and experimentally investigated aiming at optimizing the peptide inhibitory activity.
    1. P. Tamamis, D. Morikis, C. A. Floudas and G. Archontis. Species Specificity of the Complement Inhibitor Compstatin investigated by All-atom Molecular Dynamics Simulations. Proteins: Structure, Function and Bioinformatics 78, 2655-2667 (2010).
    2. P. Tamamis, P. Pierou, S. Mytidou, C. A. Floudas, D. Morikis and G. Archontis. Design of a modified mouse protein with ligand binding properties of its human analog by molecular dynamics simulations: the case of C3 inhibition by compstatin. Proteins: Structure, Function and Bioinformatics, 79: 3166-3179 (2011).
    3. P. Tamamis, A. López de Victoria, R. D. Gorham Jr., M. L. Bellows, P. Pierou, C. A. Floudas, D. Morikis and G. Archontis. Molecular Dynamics Simulations in Drug Design: New Generations of Compstatin Analogs. Chemical Biology & Drug Design, 79: 703-718 (2012).
    4. A. Higginbottom, S. A. Cain, T. M. Woodruff, L. M. Proctor, P. K. Madala, J. D. A. Tyndall, S. M. Taylor, D. P. Fairlie, and P. N. Monk. Comparative Agonist/Antagonist Responses in Mutant Human C5a Receptors Define the Ligand Binding Site. J. Biol. Chem. 280, 17831-17840 (2005).
    Back to the top of the page

  • Coupling of DNA binding and architectural remodeling drives the function of Human RPA
    Chunli Yan

    Chemistry, Georgia State University

    The eukaryotic single-stranded (ss) DNA-binding protein(SSB), replication protein A (RPA), plays a central role inreplication, recombination and repair. Human RPA is aheterotrimer with three subunits of ~70, 32 and 14 kDa,which are referred to as RPA70 (ABC), RPA32 andRPA14, respectively. In DNA-processing events, RPAalso interacts with many additional nuclear proteins, andthis interaction both regulates, and is regulated by, aninteraction with ssDNA. Binding of ssDNA is critical forshielding DNA strands from endonuclease activity andpreventing the formation of disruptive secondarystructures. Using additional binding surfaces, RPArecruits DNA processing factors, thereby providing aplatform for their organization and managing theiraccess to the ssDNA. Engaging ssDNA and subsequenttransfer from RPA to other DNA processing proteins arecrucial events for the progression of DNA processingmachinery; however, the physical basis for thesetransactions are not well understood. RPA binds to DNA with a specific polarity and has at least three major binding modes characterized by thelength of ssDNA that it contacts 8¨C10, 12¨C23 and28¨C30 nucleotides (nt). The first mode, which isconsidered to be a major one, is characterized by anoccluded binding site of ~30 nt. This binding modeexhibits high affinity and low cooperativity. The secondmode, which is less stable and may be a precursor forthe 30 nt mode, has an 8¨C10 nt binding site andexhibits a lower affinity and a higher cooperativity. Thetransition from the 8 to the 30 nt mode is thought to be afunctionally important event implicated in DNAunwinding. A more stable intermediate binding of 12¨C23nt mode occurs with the additional involvement ofRPA70C. The association constant of the bindingranges from 108 to 1011 M-1 depending on thesequence and length of the substrate.These three modes of binding ssDNA proposed forRPA, involving contacts to RPA70AB, RPA70ABC, andRPA70ABC/32D, respectively, should modify theorganization of RPA domains. Small angle X-ray andneutron scattering (SAXS/SANS) experiments withall-atom molecular dynamics simulation were performedto determine the effects of ssDNA binding on thearchitecture of RPA. The SAXS data combined withmolecular dynamics simulations revealed two not threetransitions as RPA binds ssDNA. There is no evidencefor an intermediate state corresponding to the ~20 ntbinding mode. The revised view of the DNA-boundstates combined with evidence of significant residualmotion leads to a new model for RPA-ssDNAinteractions and provides insight into binding and releaseby RPA in DNA processing pathways.
    Back to the top of the page

Contributed Speakers

  • Molecular Investigations into the Mechanics of a MuscleAnchoring Complex
    Nicholas Bodmer

    Chemistry, University of Cincinnati

    Titin governs elasticity within muscular contractile units.Telethonin plays a critical role in anchoring titin repeatsin the Z-disk of the sarcomere. Given the all b-strandstructure of the complex, hydrogen bonding is crucial forstructural stability, however it is unclear what part of thehydrogen bonding network is responsible for the functionof the complex(1). Moreover atomic force microscopyexperiments indicate that the titin-telethonin complexcan withstand the highest measured force in a globularprotein when pulled in the physiological C-terminaldirection (2). However, when pulled along otherdirections, the complex fails at forces corresponding tothe mechanical stability of typical immunoglobulindomains(3). Previous attempts to account for thisphenomenon in silico have met with limited success dueto a set up that deviates substantially from itsexperimental counterpart (4). We employed a different approach focused on a coarsegrained description of the protein to follow the dynamicsof the titin-telthonin complex under forces applied tomimic the AFM experiments(5) Due to the large size ofthe system and the long timescales involved, to be ableto follow the behavior of the complex on experimentallyrelevant hundreds of ms, we implemented thesimulations on graphics processing units (6). Our most exciting finding is that this system is finely tuned to theforce load regime such that an order of magnitudechange in the loading rate results in significant changesin behavior. Our results shed light on the structuralunderpinnings of the previous experimental results(2)and clarify the role of the hydrogen bond network in themechanical behavior of titin-telethonin. Moreover ourinvestigations show how a system comprised entirely of immunoglobulin proteins can display behavior that issignificantly different from its building blocks.
    1.Zou et al, 2006, Nature, 439:229
    2.Bertz et al, 2009, PNAS, 103:13307
    3.Rief et al, 1997, Science, 276:1109
    4.Lee et al, 2006, Structure, 14:497
    5.Hyeon et al, 2006, Structure, 14:1633
    6.Zhmurov et al, 2010, Proteins, 78:2984
    Back to the top of the page

  • Using Molecular Dynamics to Understand theMolecular Mechanisms Underlying the Application ofCombination Therapies to Inhibit Hepatitis C VirusPolymerase
    Jodian Brown

    Depertment of Chemistry and Biochemistry, University of Maryland, Baltimore

    The RNA-dependent RNA polymerase (NS5B)of the hepatitis C virus (HCV) is a vital component ofviral replication. In addition, there is no knownmammalian homolog of HCV NS5B enzyme, making it apromising target for clinical investigation. A majorchallenge of treating HCV is the emergence ofresistance to current treatment regiments. An approachto reducing the rate of drug resistance is to increase theinhibitory effects of small molecule inhibitors by usingthem in combination. This proposal focuses onunderstanding how multiple allosteric ligands can beused to synergistically inhibit the enzyme. The primarygoal is to use molecular dynamics simulation tounderstand the dynamic and thermodynamic changesthat result from the binding of multiple ligands to NS5B.Understanding the molecular mechanisms that mediate the binding of multiple inhibitors to NS5B may allow usto optimize the inhibitory activity of these compoundsagainst the enzyme. Furthermore, the knowledge gainedcan provide fundamental insight into how multipleligands bind to proteins. Such information has direct applications in the areas of drug discovery, regulation of metabolic pathways, and other signal transductionprocesses.
    Back to the top of the page

  • ROSETTAEPR: An Integrated Tool for Protein Structure Determination from Sparse EPR Data
    Stephanie DeLuca

    Center for Structural Biology, Vanderbilt University

    Membrane proteins remain a particular challenge instructural biology. Only about 1.5% of reported tertiarystructures and 60 unique membrane protein topologiesconsisting of more than one transmembrane span arerepresented in the PDB. However, these proteins makeup an estimated 30-40% of the entire proteome, andover half of all therapeutics target this group.Site-directed spin labeling electron paramagneticresonance (SDSL-EPR) is often used for the structuralcharacterization of proteins that elude other techniques,such as X-ray crystallography and NMR. However,high-resolution structures are difficult to obtain due touncertainty in the spin label location and sparseness ofexperimental data. ROSETTAEPR has been designed toimprove high-resolution protein structure prediction usingsparse SDSL-EPR distance data. The“motion-on-a-cone” spin label model is converted into aknowledge-based potential, which was implemented asa scoring term in ROSETTA. We have demonstrated thefeasibility of using ROSETTAEPR with soluble proteinsby benchmarking the method on T4-lysozyme.ROSETTAEPR increased the fractions of correctlyfolded models (RMSDCα < 7.5Å) and modelsaccurate at medium resolution (RMSDCα < 3.5Å)by 25%. After full-atom refinement, ROSETTAEPR yielded a 1.7Å model of T4-lysozyme, thus indicatingthat atomic detail models can be achieved by combiningsparse EPR data with ROSETTA. ROSETTAEPR wasalso benchmarked on a set of membrane proteins ofknown structure. If EPR experimental data were notavailable, simulated data were derived from the existingstructures. It was generally observed that de novofolding in the presence of EPR restraints enriched therecovery of the proteins’ correct topology compared towhen folding with no restraints.
    Back to the top of the page

  • Modeling Autotransporter Secretion
    Igor Drobnak

    Department of Chemistry and Biochemistry, University of Notre Dame

    Autotransporters are a class of proteins in Gramnegative bacteria that are secreted across the outermembrane and are often involved in pathogenicfunctions. They are transported across the innermembrane cotranslationally by the ATP driven Secmachinery. To cross the outer membrane, however, theyneed an alternative energy source, since ATP is notavailable outside of the cytosol. It has been proposedthat the autotransporter remains unfolded in theperiplasm, but folds on reaching the extracelluar space,which would prevent it from backsliding. We haveconstructed a computational model that uses astochastic simulation algorithm to simulate the kineticbehavior of an autotransporter system given differentequilibrium and rate constants for folding and outermembrane translocation. Computational modelingprovides useful insight into the secretion process even inthe absence of experimental data - the ability or inabilityof the folded autotransporter to cross the outermembrane is shown to be a crucial determinant of thesystem, along with the rate of folding in the periplasm.Extracellular protein folding can be used as a drivingforce for secretion only if translocation of the foldedprotein across the outer membrane is very slow. In thiscase, any folding in the periplasm represents a“dead-end” side reaction, the rate of which needs to beminimized if secretion is to be efficient. On the otherhand, a high rate of translocation for the folded proteinresults in equilibration of the system, with theconsequence that an external difference in free energy isneeded to favor secretion. No such energy source isknown to date. Efforts are currently underway toexperimentally determine the kinetic parameters thatcontrol autotransporter secretion.
    Back to the top of the page

  • Molecular dynamics simulation and coarse-graining study on cofilin remodeling actin filaments
    Jun Fan

    Department of Chemistry, University of Chicago

    Actin filaments consist of actin proteins, which areabundant in eukaryotic cells. These filaments formmesh-like structures to provide mechanical support anddetermine the shape of cells. Actin filaments also playimportant roles in cell mobility, cell division, endocytosisand intracellular transportation. Cofilin proteins, actindepolymerization factor, play key roles in the dynamicsof actin filaments. The binding of cofilin modifies thestructure, conformational dynamics and mechanicalproperties of actin filaments. We investigate thesemodifications using molecular dynamics (MD)simulations and coarse-grained (CG) analysis. In MDsimulations, we observe that twist, tilt, cross-over lengthof actin filaments differ due to cofilin binding, consistentwith experimental data. Moreover, mechanicalproperties, including the persistence length and torsionalstiffness, decrease significantly after cofilin binding, inagreement with experiments. This decrease is causedby the weakened longitudinal and lateral interactionswithin the filament. To quantify this, we apply the CGanalysis method to the MD data. CG analysis resultsreveal that the longitudinal distance between DNase-Ibinding loop and subdomain 1 of the neighboring subunitincreases twice as far thus the effective interactionsdecrease remarkably. Another longitudinal contactbetween subdomain 3 and subdomain 4 of theneighboring subunit also become weaker even thoughthe distance between these two subdomains does notchange dramatically. Meanwhile, the lateral contactdistances between hydrophobic loop and neighboringsubunit vary slightly while the effective interactionstrengths are still compromised with cofilin binding.These results provide a molecular interpretation for theeffect cofilin binding has on the structure and propertiesof an actin filament.
    Back to the top of the page

  • Distinct phases of free alpha-synuclein - a flat-histogram Monte Carlo study
    Sigurdur Jonsson

    Department of Astronomy and Theoretical Physics, Lund University, Sweden

    The alpha-synuclein protein, implicated in Parkinson'sdisease, shows conformational versatility. It aggregatesinto beta-sheet-rich fibrils, occurs in helicalmembrane-bound forms, is disordered as a freemonomer, and has recently been suggested to have afolded helical tetramer as its main physiological form.Here I present an implicit solvent all-atom Monte Carlostudy of the conformational ensemble sampled by thefree alpha-synuclein monomer [1]. We analyzesecondary-structure propensities, size and topologicalproperties, and compare with existing experimental data.Our study suggests that free alpha-synuclein has twodistinct phases. One phase has the expected disorderedcharacter. The other phase also shows largeconformational variability. However, in this phase, thebeta-strand content is substantial, and the backbone foldshows statistical similarities with that in alpha fibrils.Presence of this phase is consistent with data fromlow-temperature experiments. Conversion of disorderedalpha-synuclein to this fibril-like form requires thecrossing of a rather large apparent free-energy barrier.The presence of the free-energy barrier makessimulating this 140-residue protein a challenge. Totackle this problem, we use a two-step simulationprocedure based on the Wang-Landau and themulticanonical methods [2].
    1. Distinct phases of free alpha-synuclein -- a Monte Carlo study, S.A. Jónsson, S. Mohanty and A. Irbäck(submitted to Proteins)
    2. Accelerating atomic-level protein simulations by flat-histogram techniques, S.A. Jónsson, S. Mohantyand A. Irbäck, Journal of Chemical Physics 135,125102 (2011)
    Back to the top of the page

  • Protein folding by coarse-grained molecular dynamics
    Gia Maisuradze

    Chemistry and Chemical Biology, Cornell University

    In order to understand the kinetics and thermodynamics of protein folding it is important to know why proteins foldor do not fold, and what governs the way that proteinsfold. To answer this general question, there are manyaspects in folding which must be understood. Forexample, it is of interest to know (i) whether localfluctuations in a polypeptide chain play any role in themechanism by which the chain folds to the nativestructure of a protein; the correlations between local andglobal motions; (iii) the key residues playing importantrole in folding; (iv) correlation between the side chainand main chain motions in native state and folding. Allthese aspects are addressed in this presentation byanalyzing the folding and non-folding trajectories of the37-residue triple beta-strand WW domain from theFormin binding protein 28 (FBP) [PDB: 1E0L], B-domainof staphylococcal protein A [PDB: 1BDD (alpha; 46residues)], 46-residue alpha/beta model protein VA3[PDB: 1ED0]. Molecular dynamics trajectories weregenerated with the coarse-grained united-residue(UNRES) and all-atom (OPLS) force fields.
    Back to the top of the page

  • Long-distance correlations of rhinovirus capsid dynamics contribute to uncoating and antiviral activity
    Amitava Roy

    Medicinal Chemistry and Molecular Pharmacology, Purdue University

    Human rhinovirus (HRV), cuastive of common cold, and other members of the enterovirusgenus bind small-molecule antiviral compounds in acavity buried within the viral capsid protein VP1. Thesecompounds block the release of the viral protein VP4and RNA from inside the capsid during the uncoatingprocess. In addition, the antiviral compounds prevent“breathing” motions, the transient externalization of theN-terminal regions of VP1 and VP4 from the inside ofintact viral capsid. The site for externalization ofVP1/VP4 or release of RNA is likely between protomers,distant to the binding cavity for antiviral compounds.Molecular dynamics simulations were conducted toexplore how the antiviral compound, WIN 52084, altersproperties of the HRV 14 capsid through long-distanceeffect. We developed an approach to analyze capsiddynamics in terms of correlated radial motion and theshortest-paths of correlated motions. In the absence ofWIN, correlated radial motion is observed betweenresidues separated by as much as 85 °A, a remarkablylong distance. The most frequently populated pathsegments of the network were localized near the 5-foldsymmetry axis and included those connecting theN-termini of VP1 and VP4 with other regions, inparticular near 2-fold symmetry axes, of the capsid. Theresults provide evidence that the virus capsid exhibitsconcerted long-range dynamics, which have not been.previously recognized. Moreover, the presence of WINdestroys this radial correlation network, suggesting thatthe underlying motions contribute to a mechanistic basisfor the initial steps of VP1 and VP4 externalization anduncoating.

    Concerted motion is essential for proteins tocarry out their functions. Concerted motion implies correlated fluctuation of different parts of protein, whichcan be distance apart. Atomistic molecular dynamics(MD) simulation of proteins can be a powerful tool toreveal correlated motion in great details. Howeverlong-range correlation motion has so far been elusive inin-silico studies. The most widely used method toquantify correlated fluctuation is finding Pearson'scorrelation coefficient of displacement vector (DCC).DCC depends on the cosine of the angle between thevectors and is only sensitive to linear correlation. Radialcorrelation (RCC) and generalized correlation (GCC)coefficients have been used to quantify correlation toovercome shortcomings of DCC. However RCC isinsensitive to azimuthal fluctuation and might not beuseful where radial symmetry is not apparent. GCCrequires components of a displacement vector to beindependent of each other, a constraint which is usuallyviolated. In this article we discuss merits andweaknesses of using DCC, RCC and GCC in analysis ofMD trajectories of macromolecules and show a newcorrelation coefficient, distance correlation coefficient(DiCC) can capture both linear and non-linear correlationwithout imposing any assumption on the time series ofvectors. We also discuss how dimension reductionbased on multivariate analysis on linear covariances canbe misleading and show that distance covariances canreduce dimension more effectively.
    Back to the top of the page

  • Direct Observations of Shifts in the β-Sheet Register of a Protein-Peptide Complex Using Explicit Solvent Simulations
    Reza Salari

    Department of Chemistry, University of Pittsburg

    We report direct observations of rearrangements in anintermolecular protein-peptide β-sheet using explicitsolvent simulations. The β-sheet is formed between theFHA domain of cancer marker protein Ki67 (Ki67FHA)and a peptide fragment of the hNIFK signaling protein.Simulations of rearrangements from a misregisteredstate to the native state were generated using acombination of large-scale distributed computing andsupercomputing resources. We discuss a commonmechanism that is shared by our resultingrearrangements. To our knowledge, these simulationsprovide the first atomically detailed visualizations of amechanism by which nature might correct for errors inthe alignment of intermolecular β-sheets.
    Back to the top of the page

  • The calculation of free energies over vast length scales through conventional, enhanced sampling, and free energy molecular dynamics simulations
    Jeff Wereszczynski

    Chemistry and Biochemistry, University of California, San Diego

    The guided entry of tail-anchored proteins (GET)mechanism post-translationally targets tail-anchored(TA) protein to the endoplasmic reticulum membrane.At the center of this pathway is the ATPase Get3, ahomodimer of approximately 700 residues that shuttlesTA proteins from the cytosolic Get4/Get5 complex to thetransmembrane Get1/Get2 complex for insertion.Crystallographic studies have shown that binding of ATPand ADP molecules induces large scale conformationalrearrangements from the “open” state, which is observedin the nucleotide free structures, into “closed”conformations. Here, we present a series of all-atommolecular dynamics simulations that address thestability of these conformational states and themechanism of transition between them. By combingresults from conventional and accelerated MDsimulations with rigorous free energy calculations, thethermodynamic landscape along the dimensionsprimarily responsible for the opening/closing transitionare reconstructed for five possible nucleotide states.Results show good agreement with experiments on thepropensity of the open and closed forms in theno-nucleotide, two ATP, and two ADP bound states, andreveal their relative populations in the asymmetric oneATP and one ADP states. In addition, thenucleotide-free case is shown to exist in an equilibriumof configurations, the recently revealed “semi-open”state is observed as an energy minimum in multiplenucleotide bound cases, and we present evidence forthe novel “wide-open” conformation. Taken together,these results and have lead to a model of the nucleotidedependent mechanism of Get3 opening and closing andsuggest new insights into Get3's function.
    Back to the top of the page

Posters

  • Multiscale Sampling of the CBM and Lk Properties in the Cel7A-Cellulose Interaction
    Emal M. Alekozai

    Center for Molecular Biophysics, University of Tennessee

    Cellulose, the most abundant biopolymer on earth(approx. 100 billion dry tons/ year) holds an enormouspotential as a renewable energy source. It consists ofsugar subunits which can be unlocked and fermented toproduce bioethanol. Cellulase enzymes, in particularCel7A, play an essential role in the cellulose degradationand carbon turnover in the biosphere. Cel7A consists ofa carbohydrate binding module (CBM) and a catalyticdomain which are hold together by a linker peptide (Lk).There is evidence that the CBM and the Lk are importantfor the Cel7A-cellulose interaction. In a two stepsimulation protocol Brownian (>40 ms) as well asmolecular dynamic simulations (>5.3 microseconds)were conducted. Our results suggest that hydrophobiccellulose fiber faces are thermodynamically andkinetically favored. Our simulation results complement alarge body of previous studies, providing detailedinsights into the mechanism of the Cel7A-celluloseinteraction.
    Back to the top of the page

  • nMOLDYN4 software
    Bachir Aoun

    Laue Langevin Isere, Grenoble France

    nMOLDYN is a modular program package for theanalysis of Molecular Dynamics trajectories, especiallydesigned for the computation and decomposition of neutron scattering spectra. The current release 3.11, itallows one to calculate the mean-square displacement,the velocity autocorrelation function as well as its Fourier Transform (the density of states) and its memoryfunctions, the angular velocity autocorrelation function and its Fourier transform, the reorientational correlationfunction. Moreover it can compute several quantitiesrelated to neutron scattering: the coherent and incoherent intermediate scattering functions with theirFourier transforms and their memory functions, and theelastic incoherent structure factor. Additionally, thenMOLDYN package allows one to construct modifiedtrajectories from an input trajectory ; rigid-bodytrajectories, in which the internal motions of themolecules (or parts thereof) are eliminated, angulartrajectories, which describe rigid-body motions bycenter-of-mass and orientational (quaternion)coordinates, frequency-filtered trajectories, from whichmotions outside a specified frequency interval are eliminated.
    Back to the top of the page

  • Structure and dynamics of models of amyloid-β segmental polymorphisms
    Workalemahu Berhanu

    University of Oklahoma

    Amyloid-beta (Ab) aggregates is considered to play a role in the pathogenesis of Alzheimer’s disease. Ab molecules form b-sheet structures with multiple interaction sites within each Ab molecules forming various polymorphism. Ab polymorphism gives rise to differences in morphology, physico-chemical property and level of cellular toxicity. Eisenberg laboratory has determined the microcrystal structures of short, self-complementing pairs of ß-sheets (steric zippers) from segments of Aß. Combining these atomic structures with previous NMR several fiber models have been proposed. We investigated the conformational stability of various segmental polymorphisms of Ab structures in solution using molecular dynamics simulations. The structural comparison among the segmental polymorphism in aqueous environment shows the inter-sheet side chain-side chain interaction, hydrophobic interaction among the strands (ß1 and ß2) and salt bridge are important in stabilizing the aggregates. The segmental polymorph with smaller size of steric zipper shows a larger structural fluctuation while the one with larger size of steric zipper at the interface is very stable. Despite some difference in their structural stability the segmental polymorphic models of Ab the retained U-shaped architecture with smaller fluctuation in b-sheet region during the simulation showed. Residues at the edge and loop region showed higher mobility. The inter-peptide salt bridges between Asp23 and Lys28 were strong compared to intra-chain salt bridge and there is as exchange of the inter-chain salt- bridge with intra-chain salt bridge. The simulation showed the segmental polymorphs of Ab retain U-shaped architecture with smaller fluctuation in b-sheet region. Residues at the edge and loop region showed higher mobility. This knowledge of structural stability and aggregation behavior of Ab polymorphic forms may help to develop therapeutics for Alzheimer's disease.
    Back to the top of the page

  • Mutual Information Study of Adenylate Kinase
    Nicholas Callahan

    Chemistry, The Ohio State University

    In the course of protein evolution, the interactionsbetween side chains are optimized for overall organismfitness. Anenzyme must both be stable and fulfill its metabolic rolewithin certain kinetic parameters. Because the preferredenvironment of organisms can vary widely acrossphylogenies, homologous enzymes can have similarbackbones butdifferent side chain packing, thereby fine-tuninginteraction strength and dynamic mechanisms forenvironmentalconditions. This evolution-driven varying of side chainpacking can create positional correlations in multiplesequencealignments. It has been shown that these correlationsmay reflect stabilizing interactions, but do notnecessarily do so.We present here correlation-guided mutations made inthe adenylate kinase enzyme of Bacillus subtilis bacteria(bsADK) which, in the context of a destabilized variant,affect kinetic activity without altering stability. The folded Lid domain of bacterial ADK closes over the active sitefollowing substrate binding. This domain is stabilized byeither azinc-chelating motif in gram-positive species or by anetwork of hydrogen bonds in gram-negative species.Mutatingthe four chelating residues of bsADK to theircounterparts in Escherichia coli ADK (ecADK),abolishing metalbinding,results in a severe drop in global stability and the loss ofenzymatic activity. Using mutual informationanalysis, we located two positions in the lid domainstrongly correlated to chelating motif. Further mutationsweremade in bsADK at these positions. These additionalmutations were found to have little effect on globalstability, butserved to partially rescue enzymatic activity. Although itis understood that loss and gain of stability can directlyalterthe dynamics of adenylate kinase, the fact that thesevariants do not vary in stability suggests that they couldserve asan experimental system for further studying the kineticpathway of this enzyme. We are currently pursuingbiophysicalstrategies to further characterize these variants andelaborate on their kinetic differences.
    Back to the top of the page

  • Homology Modeling in the Twilight Zone: Can Molecular Dynamics or Monte Carlo Simulations Improve the Quality of Homology Models?
    Derek Cashman

    UT/ORNL Center for Molecular Biophysics, Oak Ridge National Laboratory

    In order to assess the quality of homology modeling,models were constructed of the E. coli and T. maritime chemotaxis protein, CheW, two proteins with knownNMR structures. A homology model of E. coli CheW wasconstructed based on its sequence aligned to the T.maritima NMR structure, and a model of T. maritime CheW was constructed based on its sequence alignedto the E. coli NMR structure. In other words, we areasking the question of whether it is possible to build amodel of a protein of known structure using the templateof a homologous protein, and then to use simulationmethods to find the correct conformation of the originalknown protein that we are modeling. Three homologymodels of each protein were constructed and eachstructure was simulated in explicit solvent for 50 nsusing the NAMD2 molecular dynamics software. Each ofthe three models was also simulated using theLibrary-Based Monte Carlo (LBMC) software package(Zuckerman et al., 2009). While the molecular dynamicssimulations provide a good picture of the conformationalfluctuations of the CheW protein in a time-dependentmanner, the primary advantage of the Monte Carlosimulations lies in their random and time-independentnature, as well as in their ability to more rapidly samplethe broad conformational space of proteins. Simulationswere also performed using the starting NMR structuresas a control. The root-mean-square deviation wascalculated for each MD and LBMC trajectory individually,as well as between each individual structure in thehomology-modeled trajectory versus each individualstructure in the native trajectory. This measures theability of each homology-modeled structure to“find†the native protein structureas well as the other homology-modeled structures in theconformational space and therefore provide anassessment of how closely each homology-modeledensembles agree with the overall native proteinensemble. Our results indicate that there are significantfluctuations in the amine and carboxy termini of theproteins, both in the native and homology-modeledensembles. However, when the protein core is alignewithout these flexible regions, there is reasonably goodagreement between the homology models and the nativeprotein.
    Back to the top of the page

  • Effect of salt on protein fold stability: Study through implicit as well as explicit solvation technique.
    Mihir Date

    Department of Chemistry, Clemson University

    Solvent plays an important role in determining thestructure and function of biological molecules. Hence anaccurate representation of solvent is immenselyessential for the accurate assessment of stability anddynamics of proteins and enzymes and their interactionwith other solvated ions. Salt influences proteins and enzymes through thescreening of intra and inter molecular electrostaticinteractions as well as through Hofmeister effects. In thiswork, a continuum solvation model was developed toexplore the impact of ionic strength on protein stability.This model combined a Poisson-Boltzmann equation forcontinuum electrostatic solvation forces with a surfacearea dependent term, containing a new saltconcentration dependent microscopic surface tensionfunction, to capture hydrophobic effects. The model wasvalidated against experimentally determined salt effectdata for cold-shock protein B and 27 of its mutants. Itshowed good qualitative as well as quantitativeagreement in matching experimental data. The approachwas then applied to HIV-1 protease in order to explainthe origins of its experimentally observed increasedstability as a function of NaCl concentration. The effect of different ions in close vicinity to proteinsurfaces remains a valuable area of study, and mayprovide important information regarding protein foldingand evolution. One way of capturing the Hofmeistereffects on proteins is by quantifying how salts alter theassociation thermodynamics of waters and amino acidsside chains on protein surfaces. Molecular dynamicssimulations were performed on a model protein in aperiodic box of TIP3P water molecules and Hofmeistersalt ions at concentrations ranging from 0.5 to 3.0 mol/L.Radial distribution functions calculated for watermolecules and salt ions and calculated coordinationnumbers between the protein and salt ions show anagreement between the rank for an ion in the Hofmeisterseries and its influence on solvation shell waters aroundthe protein. On the other hand, the atom’s positionalcorrelation time calculated in different shells from proteinsurface and for different ions show that the rank of anion has no correlation and influence on long range waterstructure making and breaking properties. With thisresult, one of the most important and new aspects of theinfluence of ions on proteins emerges as the saltimpacts the structure of biological water and bulk waterdifferently and the effect of salt on the later is not centralto the Hofmeister effect.
    Back to the top of the page

  • A Novel Method for Guiding High Throughput Protein-ligand docking with QSAR-Derived Pharmacophore Maps
    Samuel DeLuca

    Chemistry, Vanderbilt University

    Currently, QSAR and computational ligand dockingstudies are valuable but independently used tools fordrug design. Data from Pharmacophore maps producedby tools such as COMFA are typically compared to theresults of docking simulations by hand in a qualitativemanner. RosettaLigand has been previously successfulat predicting binding poses with highresolution(Kaufmann, et. al, Proteins, 2009). We aredeveloping RosettaHTS, an extension to RosettaLigandwhich will integrate these two methods by usinginformation from QSAR derived pharmacophore maps toguide the low resolution phase of ligand docking.Discrete cartesian grids describing the hydrogenbonding ability, electrostatics and shape of the ligandbinding site are overlaid on the protein structure, andthese grids are used to score the initial placement of theligand prior to fine grained docking. Sampling of theligand within this grid is guided by the chemical i As thescoring grids are precomputed, ligand scoring isextremely fast, and thorough Monte Carlo sampling ofthe ligand binding site can be rapidly performed beforefine grained ligand docking using the high resolution Rosetta scoring function. This efficient and fast initialsampling makes it possible to distinguish between activeand inactive compounds with less fine grained sampling,decreasing the amount of CPU time necessary to predicta single binding interaction, and increasing thepracticality of structure based virtual High ThroughputScreening (vHTS). The integration of structure basedand ligand based vHTS techniques allows the full rangeof pharmacological information surrounding a target anddrug scaffold to be considered in a single approach. Thistechnique can be used to rapidly develop small focusedlibraries for High Throughput Screening, increasing thehit rate and decreasing the number of compounds thatneed to be purchased for testing.
    Back to the top of the page

  • High-throughput Virtual Molecular Docking on High-Performance Computers
    Sally R. Ellingson

    Genome Science and Technology and Center for Molecular Biophysics, UT/ORNL

    Many pharmaceuticals act by selectively binding to a specific protein and thus inhibiting a specific process relevant to a disease or illness. Because of this, the early stage of drug discovery consists of identifying potential compounds that bind to a protein of interest with a high affinity and specificity. Experimentally testing a very large number of these compounds is both costly and time consuming. Virtual high-throughput screening is an equivalent computational process that can reduce the time and cost of discovering new drugs. After a potential lead compound is identified in the drug discovery process further tests must be done in order to determine toxicity and side effects. Computational tools with the ability to virtually screen a lead compound against a very large number of different proteins to help predict these effects earlier in the drug discovery pipeline would be valuable. Here we discuss the screening of a million compound library on a petascale supercomputer and future directions to incorporate large libraries of proteins in high-throughput screens.
    Back to the top of the page

  • Structural Modeling of the Outer Membrane Protein Igni1226 Nonamer of Ignicoccushospitalis
    Hao-Bo Guo

    UT/ORNL Center for Molecular Biophysics, Oak Ridge National Laboratory

    Ignicoccushospitalis, the smallest free-living Archaea,and Nanoarchaeumequitans, the smallest Archaea,form the smallest and the only purely archaeal symbioticsystem known to date. N. equitans acquires metabolitesand energy together with lipids and amino acids from itshost, I. hospitalis. Previous studies suggested that an85-residue protein Igni1226 may serve as the cell-to-celltransporter between the two Archaea. A recent electrontomography experiment showed that nine chains ofIgni1226 form a nonamer spanning a 2 nm pore, throughwhich materials such as the respiratory complexes andions may transport among two Archaea and theirenvironments. To understand how Igni1226 nonamerperform the transportation works, however, detailedstructural knowledge would be demanded from eitherexperimental or theoretical investigations, or both. In thepresent work, we constructed the undetermined, with yetunknown function Igni1226 pore structural model startingfrom its amino acid sequence. This model representsthe electron tomography and reserves the predictedsecondary structure within the presence of a membranemodel. Molecular dynamics (MD) simulations of theIgni1226 nonamer model in the membraneenvironments, together with the normal mode analysisusing the anisotropic network model (ANM), wereperformed to tackle the structural characteristics of thenonamer that may be correlated to the transportationbetween the two Archaea.
    Back to the top of the page

  • The use of immune epitope prediction and HLA subtype imputation to model cancer cell recognition by the immune system
    Russell Hanson

    Harvard Medical School, Children’s Hospital Boston

    It is well known that cancer is in part an immunologicaldisease. In this study we present novel algorithmic andcomputational results for some of the largest cancerimmunology studies done to date to predict and infer theimmunological reactivity of different cancer pointmutations as well as the susceptibility and protectionstatus for different patient HLA subtypes. These studiesare important for personalized medicine in that patientswith different protected vs. susceptible types may needor require different treatment regimes. We presentresults across the largest cancer genome databasesavailable for point mutations, HLA subtype imputationsfor from 3,000-4,000 patients, and indications for futurework in the area of cancer immunology and immunomestudies.
    Back to the top of the page

  • USING HIGH-PERFORMANCE SUPERCOMPUTING TO FIND ENDORCRINE DISRUPTORS: A FAST TRACK TO DISCOVERING NEW MEDICINES AND PROTECTING THE ENVIRONMENT
    Jason Harris

    Genome Science and Technology, University of Tennessee

    Although endocrine disrupting chemicals (EDCs) areonly a subset of the tens of thousands of chemicalsknown to mankind, their human health impacts can beenormous ─ ranging from reproductive disorders,obesity, diabetes, immune dysfunction,neurodegenerative diseases, and increased incidencesof some cancers (NIEHS 2010). Identifying EDCs occursthrough expensive testing programs costing between$7,000 and $30,000 per chemical, yet there are stillmany chemicals to be tested that either already exist orare being synthesized on a regular basis (EDSTAC1998). For this reason, it is critical to develop simple,fast and affordable experimental assays that can identifychemicals with endocrine disrupting activity before they.pose a health risk by entering the manufacturing stream.Moreover, new predictive models based on chemical binding to protein structures (e.g., protein docking),rather than models based only on the chemical structure(e.g., QSAR), are needed to identify undiscoveredchemical-protein interactions with novel chemicalfeatures. The goal of this project is to build and validate apredictive protein-docking model that quickly, efficiently,and affordably identifies chemicals that bind directly tothe human estrogen receptor alpha (hER-α) protein orindirectly through P450-mediated oxidation. Ourmulti-protein (hER-α and P450) approach integratescomputational prioritization and experimental testing ofcompounds via computational docking models andactivity measurements inbioluminescent-yeast-estrogen-receptor (BLYES)bioreporter assays. An additional advantage of thisapproach over previous methods such as QSAR(chemical similarity matching) is the ability to search forchemicals with novel binding modes and/orP450-mediated activity towards the hER-α protein.
    Back to the top of the page

  • A Multiscale Model for T7 RNA Polymerase Initiation on Supercoiled DNA Minicircles
    Andrew Hirsh

    Mechanical Engineering, University of Michigan

    Because T7 RNA polymerase (RNAP) is a single subunitRNAP and shares biochemical characteristics with muchlarger and more complex multisubunit RNAPs, it remainsa model system to study, mechanistically, fundamentalaspects of polymerase function including transcriptioninitiation, elongation, and termination. It is well knownthat many gene repression proteins can simultaneouslybind two operator sites and tightly loop the interveningDNA, about 100 base pairs (bp). Not only do theseDNA-repressor contacts interfere with transcription bydirectly blocking DNA, but the mechanics of thetopologically-constrained loop may also play animportant role in repression. These factors, although notcurrently well understood, may affect T7-RNAP duringinitiation and prevent effective transition to theelongation phase. To understand how T7-RNAPinitiation responds to highly stressed DNA, we constructa model for the RNAP complexed with a supercoiled(~100 bp) DNA minicircle. The current model representsthe minicircle DNA with a continuum elastic rod modelwhich relies heavily on boundary conditions derived froman atomistic model of DNA entering/exiting the RNAP.Preliminary results investigate two sets of boundaryconditions derived from the rigid crystal structure andreveal significant differences depending on the boundaryconditions. Therefore, in this study we describe amultiscale modeling approach in which the elastic rodmodel provides the conformation of the highly-stressedminicircle DNA as the initial condition to an all-atom MDsimulation which accounts for the flexibility of the RNAPwith the explicit goal of determining superior boundaryconditions for further modeling efforts.
    Back to the top of the page

  • Combining molecular dynamics simulation with incoherent neutron scattering to illustrate the hydration effect on internal protein motions
    Liang Hong

    UT/ORNL Center for Molecular Biophysics, Oak Ridge National Laboratory

    Internal motions are crucial for protein function andfolding, and depend on various external parameters,among which, hydration is of particular importance.Dehydrated enzymes lose bio-activity and the so-calleddynamical transition, manifesting as an anharmoniconset in the mean-squared atomic displacement around180 ~ 230 K, occurs only in hydrated proteins. Despitedecades of study, the detailed microscopic mechanismof how water boosting the protein dynamics is stilllacking. By combining incoherent neutron scattering andmolecular dynamics simulation, the present workdecomposed the internal motions of lysozyme on the ps-ns time scales into three components: localized single-well diffusion, methyl group rotation andnon-methyl jumps, and showed that whereasmethyl-group rotation is hydration independent andnon-methyl jumps have weak hydration dependence, thelocalized diffusion is significantly boosted by hydration,manifesting as increase of diffusional amplitude. Furtheranalysis revealed that the hydration effect on localizeddiffusion of protein atoms occurs on both the proteinsurface and in its dry core. This phenomenon can beattributed to the fact that these diffusive motions arestrongly correlated, thus enabling the hydration effect topropagate into the protein interior.
    Back to the top of the page

  • Structural characterization of intramolecular interactions of a metallochaperone domain in the mercuric reductaseMerA
    Alexander Johs

    Environmental Sciences Division, Oak Ridge National Laboratory

    Many microbes possess the ability to deal with heavymetal toxicity through elaborate metal resistancemechanisms. One well known example, bacterialmercury resistance, is mediated by the mer operon,which encodes specific genes involved in the transferand transformation of toxic Hg(II) species. The mercuricreductaseMerA is a key component of the mer operon.MerA is an NADPH-dependent flavin-disulfideoxidoreductase and catalyzes the reduction of Hg(II) to Hg(0), which is relatively inert and passively diffusesfrom the bacterial cell. Here, we combine experimentalbiophysics and computer simulation techniques toinvestigate structural features important for Hg(II)transfer in MerA. All MerA proteins consist of ahomodimeric catalytic core domain, and many alsopossess an N-terminal metallochaperone-like domainNmerA, which is tethered to the core by a ~30 aminoacid linker of unknown fold. Prior studies usingseparately expressed NmerA and core domains showedthat NmerA acquires Hg(II) from other mer proteins suchas the organomercuriallyase, MerB, and the membranetransport protein, MerT, and delivers it to the MerAcatalytic core for reduction. Here, we have appliedsmall-angle X-ray scattering (SAXS), small-angleneutron scattering (SANS) and molecular dynamicssimulations to characterize the interactions of NmerAand the core in full-length MerA in solution. Our datareveals the extent of spatial sampling of the two NmerAdomains relative to the homodimeric catalytic core andidentifies the inter-domain docking orientation thatoccurs during transient handoff of Hg(II) from a pair ofcysteine residues on NmerA to a pair of cysteines on theC-terminus of the catalytic core.
    Back to the top of the page

  • Modeling Structural Flexibility of Proteins with Go-Models
    Ping Jiang

    Department of Chemistry and Biochemistry, University of Oklahoma

    Structure-based model is an efficient tool of folding of proteins, as by construction their energy landscape is minimal frustrated. However, theirintrinsic drawback is a lack of structural flexibility as usually only one target structure is employed to construct the potentials. Hence,aGo-model may not capture differences in mutation-induced protein dynamics, if - as in the case of the disease-related A629P mutant of the Menkes protein ATP7A - the structural differences between mutant and wild type are small. We will introduce three implementations of Go-models that take into account the flexibility of proteins in NMR ensemble. Comparing the wild type and the mutant A629P of the 75-residue-large 6th domain of Menkes protein, we find that these new Go-potentials lead to broader distributions than Go-models relying on an arbitrarily chosen single member of the NMRensemble. This allowed us to observe in our simulations the transient unfolding of a loosely formedß1ß4-sheet in the mutant protein. The result is consistent with that of our previous simulations using physical force field in explicit solvents, and suggests a mechanism by which this mutation causes Menkesdisease.In addition, the improved Go-models also suggest differences in folding pathway between wild type and mutant, an observation that was not accessible in simulations of this 75-residue protein with a physical all-atom force field and explicit solvent.
    Back to the top of the page

  • Computer simulations of ATP-driven protein remodeling by the ClpY biological nanomachine
    Andrea Kravats

    Chemistry, University of Cincinnati

    Selective destruction of misfolded proteins by the proteinquality control system is critical to maintain cell viability.In bacteria, these activities are carried out by membersof the AAA+ superfamily (ATP-ases associated withdiverse cellular activities), known as Clp ATP-ases. Wefocus on ClpY, an unfoldase within this family which iswell characterized crystallographically. ClpY forms a homohexameric ring structure with anarrow central pore which fluctuates between "open"and "closed" conformations of 18Å and 8Å,respectively. ATP hydrolysis induces large scaleconformational changes of flexible central pore loopswith a highly conserved GYVG motif. Substrateunfolding and translocation are effected using multiplecycles of ATP-driven allostery within the ClpY ring. Dueto uneven binding affinity of subunits, ATP binding withinClpY is asymmetric. Experiments suggest non-concertedallostery; however, it remains unclear whether anordered or random mechanism is favored.To elucidate the coupling between ClpY-assistedunfolding and translocation of a substrate protein (SP),we developed a coarse grained model of ClpY and afour helix bundle SP (1). We determined that unfoldingand translocation occur on distinct timescales. Unfoldingis achieved byunraveling from the C-terminus of the tagged SP,forming an obligatory non-native intermediate structure,a three helix bundle. Although this intermediate structureis competent for initiation of translocation, we observemultiple translocation pathways following the initialunfolding event. Large portions of helical secondarystructural elements are translocated simultaneously,indicating a powerstroke mechanism. These results areconsistent with recent single molecule experimentssuggesting discrete steps in translocation of a SP. To investigate the effect of sequential intra-ring allosteric motions of ClpY, we performed Langevin dynamicssimulations of ordered (clockwise and counterclockwiseas determined on the proximal side of the pore) andrandom allostery. Our results suggest that clockwiseintra-ring transitions are most efficacious in handling thesubstrate, successfully passing it from active loop to thenext active loop during each allosteric transition (2). Thisresults in the most effective SP translocation. Inaddition, we find the minimum requirement for unfoldingand translocation is four active subunits, as predicted byexperiments.
    1. A. Kravats, M. Jayasinghe, G. Stan Proc. Natl. Acad. Sci. USA 108, 2234-2239 (2011).
    2. A. Kravats, S. Tonddast-Navaei, R. Bucher, G. Stan. Manuscript in preparation.
    Back to the top of the page

  • Optimizing the time step in coarse-grained force field (UNRES) molecular dynamics simulations.
    Pawel Krupa

    Molecular Modeling, University of Gdansk, Poland

    Force fields based on physical interactions arecommonly used to perform all-atom simulations. This intuitive approach has manyadvantages; however, it enables us to investigate only small systems.The time step in these classical MD simulations shouldbe smaller than the fastest vibration in the system; aconsensus value is 1 fs. However, this value can beincreased when the SHAKE/LINCS algorithms are usedto constrain bond lengths and/or when the multiple-timestep algorithms are applied. Typically, several millionsteps of molecular dynamics (MD) simulation per daycan be run for a medium-size protein with explicitsolvent, which translates into several nanosecond timeof simulation, which is too short to observe the mostinteresting properties of proteins, as e.g. folding.However, use of dedicated machines, such as, e.g.ANTON increases this time even to milliseconds.One way to increase the time- and length-scale of simulations is to use coarse-grained force field. Such anapproach not only reduces the CPU time per step (byorders of magnitude) because of reducing the number ofinteractions, but also enables us to use a greater timestep. Because of reducing the representation of thesystem under study, the optimal time step is no longerrelated to the period of atom vibrations. The purpose of our work was to determine the optimal length of the time step and to investigate the influence oftime step on ensemble averaged calculated with the useof Replica Exchange Molecular Dynamics (REMD) andMultiplexing REMD simulations of the three small model proteins: 1BDD, 1L2Y, 2HEP. We used two differentparameterizations of the UNRES force field and twoalgorithms to control the stability of the MD algorithm,namely the Variable Time Step (VTS) and AdaptiveMultiple Time Step (A-MTS). The Weighted HistogramAnalysis Method (WHAM) was used to calculateensamble averages (average RMSD from theexperimental structure and heat capacity) from theresults of the simulations. We found that the time stepcan easily be increased to 10 fs and to 15-20 fs with theuse of the A-MTS algorithm.
    Back to the top of the page

  • Molecular dynamics insights into human adenosine A2A receptor activation and deactivation
    Jianing Li

    Chemistry, University of Chicago

    The human adenosine A2A receptor (AA2AR) is atypical class A GPCR modulated by organic compoundsacting as agonists or antagonists. Recently AA2ARligands have been extensively tested as potent agents totreat a variety of diseases, but dynamic details of theligand-dependent activation and deactivation are stillincomplete. Using crystal structures of AA2AR incomplex with three agonists and three antagonists, weconstructed the membrane protein models andperformed unbiased molecular dynamics andmetadynamics simulations totaling over 5 μs. Wehave captured the distinct structural characteristics ofthe active and inactive states at atomic detail, includingseveral key domains changing in a highly concertedmanner. In particular, six conformational states ofTrp246 induced by ligands with various efficacies havebeen revealed. Our findings suggest that duringactivation, Trp246 undergoes a rotameric transition,causing a series of coherent conformational changes toopen the G-protein-binding site. Further metadynamicssimulations have showed quantitative evidence for thismechanism, indicating that agonists and antagonistschange the relative energy and shift the equilibriumbetween the active and inactive states. Our analysis alsoidentified a number of crucial residues in the ligandbinding modes, enabling structure-based designtargeting three distinct regions of the ligand-bindingpocket. Generally our study provides a comprehensivepicture of AA2AR-ligand interactions as well as dynamicinsights of AA2AR activation and deactivation beyondcrystal structures, shining a light on the path to designmore effective and selective AA2AR ligands.
    Back to the top of the page

  • QM/MM Study on the Catalytic Mechanism of Cellulase TmCel12A
    Peng Lian

    UT/ORNL Center for Molecular Biophysics, Oak Ridge National Laboratory

    The efficiency of cellulase is the bottleneck of cellulosicethanol production. Finding of cellulase 12A fromThermotoga maritime (TmCel12A) makes it possible toaccelerate hydrolysis process via increasing the reactiontemperature up to as high as 95 °C. However, atomicdetails of catalytic mechanism and hyperthermophilicnature of TmCel12A were previously unknown. In thisstudy, free energy profile of the catalytic reaction wasexplored using umbrella-sampling method coupled withQM/MM (SCC-DFTB/MM) simulations. The retainingmechanism was confirmed, and free energy barrier forglycosylation and deglycosylation were 22.3 and 23.8kcal/mol, respectively. In both steps, the glucose ring atposition -1 inversed via E3 or 4H3 conformers. Thecharge population analysis suggested the existence ofoxocarbonium in both transition states. A specialcharacter of TmCel12A is the low-barrier hydrogen bond(LBHB) between E116 and general base E134. Thishydrogen bond decreases the energy barrier indeglycosylation step at the cost of increasing that ofglycosylation. However, like other cellulases, thedeglycosylation is still rate-limit step. Based on thesefindings we propose that the fact E116 interacts withE134 through LBHB may be an adaptation to thehyperthermophilic nature of TmCel12A.
    Back to the top of the page

  • Calculation of Scattering Intensities from MD Simulation of Cellulose using a Massively Parallel Computer
    Benjamin Linder

    UT/ORNL Center for Molecular Biophysics, Oak Ridge National Laboratory

    Cellulose is a major component in the plant cell wall. It’sabundance in nature makes it a prime target forsustainable production of biofuels. The crystalline qualityof cellulose in nature can vary greatly and it showsstrong correlation with it’s recalcitrance to hydrolysis.Scattering experiments allow the structural anddynamical characterization of bulk cellulose samples.The calculation of scattering from molecular dynamicssimulations allows to reconcile experiments withtheoretical models of cellulose. Scattering calculations for biologically relevant samples,like cellulosic biomass, require enormous computationalpower, because the scattering intensities have to beorientationally averaged and/or a significant number oftime points have to be analyzed. This requirementmotivated the development of the software Sassena,which implements efficient parallel algorithms tocompute the scattering intensities on a massivelyparallel computer in a timely manner. The software isshown to scale efficiently up to thousands of cores fordifferent types of scattering calculations.
    Back to the top of the page

  • Folding of a five-helix bundle protein from all-atom molecular dynamics simulations
    Yanxin Liu

    Physics, University of Illinois, Urbana-Champaign

    The five-helix bundle lambda-repressor fragment is a fast-folding protein. A length of 80 amino acid residuesputs it on the large end among all known microsecondfolders and its size poses a computational challenge formolecular dynamics (MD) studies. We simulated thefolding of lambda-repressor fast-folding mutants inexplicit solvent using an all-atom description. By meansof a recently developed tempering method, we observedreversible folding and unfolding of lambda-repressor in a10-microsecond trajectory. Starting from an extendedconformation, the folding kinetics was also investigatedthrough constant temperature MD simulation with morethan 100 microseconds duration. The protein was seento fold into a native-like topology and a slow-foldingpathway was identified. The simulation starting fromhigh-temperature and high-pressure denatured statefolded the protein in 30 microseconds, which confirmedthe existence of the T-jump and P-jump inducedfast-folding pathway. Our results also suggest newexperimental observables for better monitoring thefolding process, and a novel mutation expected toaccelerate lambda-repressor folding
    Back to the top of the page

  • Effect of Mutations and Calcium Ions Concentration onConnexin26 Hemichannel and Gap Junction Channel
    Buddhadev Maiti

    Department of Chemistry, Georgia State University

    Connexins or gap junction proteins, are a family ofstructurally related hundreds of intercellular communication channels that allow the passage ofmolecules such as ions, metabolites, nucleotides andsmall peptides. Each gap junction channel is composedby end-to-end docking of two hemichannels which arereferred to as connexons. Each hemichannelconstructed by six subunits with four transmembranehelices and two extracellular loops. Gap junctions havecrucial roles in many biological processes includingdevelopment, differentiation, cell synchronization,neuronal activity and immune responses. Mutations inconnexins thus cause several human diseases,including neurodegenerative diseases, skin diseases,deafness and developmental abnormalities. In thiscontext, we used all atom molecular dynamics (MD)simulations in an explicit solvent POPC membranesystem for connexin26 hemichannel and gap junctionchannel to show the effect of conformational change bythe mutation and calcium ion concentration. Our MDsimulations provided the effect of calcium binding sites during mutations
    Back to the top of the page

  • Coupled Dynamics Change in Cytochrome P450cam Substrate Binding Determined by Neutron Scattering, NMR and Molecular Dynamics Simulation
    Yinglong Miao

    UT/ORNL Center for Molecular Biophysics, Oak Ridge National Laboratory

    Neutron scattering and nuclear magnetic resonance(NMR) relaxation experiments are combined withmolecular dynamics (MD) simulations in a novelapproach to investigate the change in internal dynamicson substrate (camphor) binding to a protein (cytochromeP450cam). The MD simulations agree well with both theneutron scattering, which furnishes information on globalflexibility, and the NMR data, which provideresidue-specific order parameters. Decreasedfluctuations are seen in the camphor-bound form usingall three techniques, dominated by changes in specificregions of the protein. The combined experimental andsimulation results permit a detailed description of thedynamical change, which involves modifications in thecoupling between the dominant regions and concomitantsubstrate access channel closing, via specificsalt-bridge, hydrogen-bonding and hydrophobicinteractions. The work demonstrates how thecombination of complementary experimentalspectroscopies with MD simulation can provide anin-depth description of functional dynamical proteinchanges.
    Back to the top of the page

  • Molecular modeling of the interactions in the yeast Ssq1 - Jac1 - Isu1 system in the context of iron-sulfur cluster biogenesis.
    Magdalena Mozolewska

    Molecular Modeling, University of Gdansk, Poland

    Molecular modeling of the interactions in the yeast Ssq1- Jac1 - Isu1 system in the context of iron-sulfur cluster biogenesis.Yeast proteins Isu1 (Iron Sulfur protein) and Jac1(J-protein) are part ofhuge ATP system and both interact with the Ssq1molecular chaperone. In bacteria, the equivalent proteins are IscU (for Isu1) andHscB (for Jac1),respectively. Isu1 has influence on iron homeostasis inthe mitochondrionwhere it is involved in assembling of iron-sulfur proteins.It can also beinvolved in the repair of iron sulfur clusters. Jac1 isinvolved with Hsp70 and Isu1 in Fe-S cluster biogenesis inmitochondria.The iron-sulfur clusters are the most ancient co-factorsof proteinsinvolved in many essential processes such as catalysisand electron transfer. The release of the Fe/S cluster from Isu1, andits transfer and incorporation into recipient apoproteins (Apo) isfacilitated by late components of the ISC assembly machinery includingthe ATP-dependent Hsp70chaperone Ssq1, the DnaJ-like cochaperone Jac1, thenucleotide exchange factor Mge1, and the monothiolglutaredoxin Grx5.The disturbances of the balance of these processes canhave very serious and dangerous consequences. In Homo sapiens, suchdisturbances can cause serious diseases such as cerebellar ataxia, myopathy,Friedreich’sataxia, microcytic anemia, tumor suppressor. The aim of this work was to model the tertiary structure of Isu1 and to investigate the interactions between Isu1 and Jac1which may be crucial tounderstand the machinery of yeast mitochondrialchaperone system. We also wanted to obtain support of the results of earlierexperimental studies, which suggest that Jac1 interacts with Isu1 mainly byLeu105, Leu109 and Tyr163. WTo accomplish this, we modeled andassessed the stability of complexes of mutated Jac1 with Isu1, where we mutatedthe putative bindingIsu1-binding residues of Jac1. To carry out simulationsat a longer time scale, we used our coarse-grained UNRES force field.We used I-TASSER server and YASARA as tools tomodel, by means of homologymodeling, the tertiary structure of Isu1. Docking of Isu1to Jac1 was carried out by using the ZDOK server. The structures ofthe proteins and complexes were optimized by molecular dynamics (MD)simulations with explicit water. Analysis of interactions between Isu1 andJac1 provides information about the residues playing crucial role inprotein binding anddetermines the most probable conformations of the proteins.
    Back to the top of the page

  • Robustness of the Cell Signaling Network as a Means to Discriminate Among the Different Models of Itk Kinase Regulation in T cells.
    Sayak Mukherjee

    Nationwide Childern’s Hospital, Battelle Center for Mathematical Medicine, Columbus, Ohio

    The process that warrants the generation of self tolerantperipheral T cells is called the thymocyte selection.During this maturation process, the overtly self reactiveas well as un- responsive thymocytes are deleted fromthe cell population. The thymocytes equipped with T cellreceptors (TCRs), capable of responding moderately tothe self peptides are allowed to survive. Recently watersoluble second messenger, inositol(1,3,4,5)tetrakisphosphate (IP4), has been implicated to play acrucial role in thymocyte positive selection (Huang etal.). It has been suggested that these IP4 moleculesregulate the transient activation of the Tec- familyprotein tyrosine kinase Itk through a competing positiveand negative feedbacks. The exact molecularmechanism involved in this feedback is howeverunclear. It is possible to con- struct more than onemodel with contrasting molecular mechanisms to explainthe present body of experimental observations. This calls for criteria to choose among these models.Robustness in face of the variation of the parameters ina model has been ubiquitously used as a criterion formodel discrimination. Here we have used the maximumentropy, calculated with the constraints imposed by theexperiments as a measure of robustness. Our data indicates that the models which are maximally robustshare a cooperative allosteric mode of Itk regulationinvolving dimeric PH domains.
    Back to the top of the page

  • Improved Generalized Born Solvent Model Parameters for Protein Simulations
    Hai Nguyen

    Chemistry, Stony Brook University

    The generalized Born (GB) model is one of the fastestimplicit solvent models and it has become widelyadopted for Molecular Dynamics (MD) simulations. Thisspeed comes with tradeoffs, and many reports in theliterature have pointed out weaknesses with GB models.Because the quality of a GB model is heavily affectedby empirical parameters used in calculating solvationenergy, here, we have refit these parameters forGBNeck, a recently developed GB model, in order toimprove the accuracy of solvation energy and effectiveradii calculation. The data sets used for fitting weresignificantly larger than those used in the past.Comparing to other pairwise GB models like GB-OBCand GB-Neck, the new GB model(GBNeck2) has betteragreement to Poisson-Boltzmann (PB) in terms ofreproducing solvation energies for a variety of systemsranging from peptides to proteins. Secondary structurepreferences are also in much better agreement withexplicit solvent simulations, as are protein MDsimulations. We also obtain near-quantitativereproduction of experimental structure and thermalstability profiles for several model peptides.
    Back to the top of the page

  • Enhanced sampling through the use of adaptive steered molecular dynamics and statistical temperature molecular dynamics
    Gungor Ozer

    Department of Chemistry, Boston University

    Conformational sampling of important events ofbiomolecules---such as folding and binding---usingall-atom molecular dynamics has always been achallenge as in most cases atomistic-scale calculationsis cost-prohibitive to observe such events in feasiblesimulation windows. In this context, many enhancedmolecular dynamics algorithms have been developed tocompensate for the substantial amount of computationincluding but not limited to steered MD (SMD) [J. Chem.Phys. 120, (2004)] and statistical temperature MD(STMD) [J. Chem. Phys. 126, (2007)]. The studypresented herein aims to explore the foundations ofthese powerful algorithms as to develop and benchmarknovel methodologies: i) SMD---in conjunction withJarzynski's equality (JE)---is widely used to studyenergetics of the unfolding and binding/unbinding ofproteins assuming prior knowledge of the reactionpathway. A staged integration of the SMD alogrithm,adaptive steered molecular dynamics (ASMD) [J. Chem.Theory Comput. 6, (2010)], has been recently formulatedin which the reaction pathway is divided into stages sothat the work distribution always exhibit good statistics(Gaussian nature) and the resulting PMF betterrepresent the simulated nonequilibrium ensemble.ASMD is further explored, enhanced and benchmarkedon several systems such as helix-coil transformation ofdecaalanine in vacuum and in explicit solvent. ii) STMDis a powerful flat energy distribution algorithm thatupdates statistical temperature estimate at each iteration. Generalized replica-exchange has also been recently integrated with the STMD method. STMD hasbeen incorporated onto popular and fast MD integrators,CHARMM and NAMD, as to study more complexsystems. STMD is currently being tested on CHARMMand NAMD on various previously benchmarked unfoldingevents of ww-domain, albumin-binding domain, and neuropeptide Y.
    Back to the top of the page

  • Reactions of Cellobiose and Ionic Clusters
    Medeliene Pincu

    Chemistry, University of California, Irvine

    Ab initio molecular dynamics (DFT / BLYP withdispersion) is used to analyze the interaction betweenconformers of cellobiose (CB), cis and trans, with singleproton and with micro-hydrated clusters of acids ofdifferent strengths, at ~300K. These are analyzed interms of probabilities for proton transfer, degree ofionization as compared with that of the acids alone in thewater cluster. Detailed information on the preferredprotonation site of cellobiose and the competitionbetween the sugar and water protonation are presented.The importance of conformer selectivity of reactionprocesses is discussed.
    Back to the top of the page

  • Swimming motility is a key role in describing the cell clumping phenomenon
    Xianghong Qi

    UT/ORNL Center for Molecular Biophysics, Oak Ridge National Laboratory

    Flagellum swimming motility has been shown tomodulate the initial interaction of bacteria with surfacesor with other cells and contribute to the emergence ofmacroscopic patterns. While the role of swimmingmotility in surface colonization has been analyzed insome detail, a quantitative physics analysis of transientinteractions between motile cells is lacking. We presenta physical model based on the motile bacteriumAzospirillumbrasilense to examine the dynamics of cellsin a crowded environment. Especially, the effect of theswimming motility of the cells on the cell clumping isstudied using simulations of motorized adhesiveBrownian particles subjected to hydrodynamicinteraction. We incorporate both equilibrium and activefeatures, such as mechanical interaction (¡°stickiness'¡±)between cells, thermal noise, and swimming velocity intoour model and investigate the clumping dynamics underseveral environmental and intrinsic factors. Our resultsshow that the modulation of active motions of the cells isrequired for the initial aggregation of cells to occur at arealistic time scale. Slowing down flagellar motor rotation(and thus swimming speed) is correlated to the degreeof clumping, which is consistent with the experimentalresults obtained for A. brasilense.
    Back to the top of the page

  • Conformational Substate Discovery for Fast Folders
    Andrej Savol

    Computational and Systems Biology, University of Pittsburgh

    The kinetic relationship between metastableconformational intermediates and protein folding rateshas challenged both computational and experimentalmethods. The present paradigm recognizes a rough freeenergy landscape (FEL) with minima corresponding tostable states and conformational transitions as energeticbarriers. However, exceptions to this simple abstractionare necessary for both intrinsically disordered proteins(IDPs), whose FELs are necessarily 'flatter', and also'ultra-fast' folders, which lack significant energy barriersduring folding. Long-run all-atom simulations are increasingly capableof accessing the timescales required to observe foldingevents. Our approach provides the needed subsequentanalysis: what conformational states were accessed,how long were they visited, and which ones facilitated orinhibited transition to the native state? We studied longmolecular dynamics simulations of villin headpiece(VHP), one such 'ultra-fast folder', and identifiednon-native metastable states that function as waypointswithin diverse folding trajectories. Rather than adoptstructural similarity measures (RMSD, Rg) to probestructural transitions, we considered an embeddeddihedral angle subspace where structural alignment biasis eliminated and overall dimensionality is drasticallyreduced. As documented, more dihedral-based modesare necessary to capture an equal quantity of varianceas its Cartesian-based counterpart, but these internal(dihedral) modes provided excellent sensitivity tonear-folded intermediates and rarely visited states. Weadditionally observed that commonly used foldingorder-parameters overlooked non-native contacts, evenwhen such 'misfolding' is reproducibly necessary forachieving the eventual native state. Taken together, ourresults address several organizing principles for theconformational landscape of fast-folders.
    Back to the top of the page

  • Refinement of DNA and Protein Structures from Neutron Crystallography Experiments: The Importance of Computational Biophysics
    Michael Schneiders

    Biomedical Engineering, University of Texas, Austin

    Oak Ridge National Laboratory is recognized as the world's leading center for Neutron Science due to thecombination of the Spallation Neutron Source (SNS),High Flux Isotope Reactor (HFIR) and the Oak RidgeElectron Linear Accelerator Pulsed Neutron Source(ORELA). In particular, the SNS provides the mostintense pulsed neutron beams in the world forperforming neutron diffraction experiments on biologicalcrystals. Unlike X-ray diffraction, the diffracted intensityof neutrons from hydrogen and deuterium atoms is asstrong as that from heavier organic elements. Therefore,neutron diffraction is complementary to X-ray diffractionand provides critical information on protonation state,tautomerization and hydrogen-bonding networks. In thispresentation, I’ll discuss the joint X-ray/neutronrefinement of DNA and protein crystals assisted by theprior chemical information in the polarizable atomicmultipole AMOEBA force field evaluated usingparticle-mesh Ewald (PME) summation. As CBSB12brings together leaders in neutron diffraction, force fielddevelopment and high-performance electrostaticsalgorithms, this meeting represents an ideal opportunityto discuss the role of computational biophysics in theapplication and interpretation of neutron diffractionexperiments.
    Back to the top of the page

  • Molecular Dynamics Simulations of Lignocellulose on a Petascale Computer
    Ronald Schulz

    UT/ORNL Center for Molecular Biophysics, Oak Ridge National Laboratory

    The understanding of biomass recalcitrance is central tothe efficient production of 2nd generation bio-ethanol.The performance and scaling of molecular dynamics(MD) simulation of multimillion-atom biological systemsare important for the simulation of realistic models ofbiomass. In the era of petaflop supercomputers, suchsimulations are limited by the parallel efficiency of theMD algorithms. The bottleneck for highly-parallelall-atom simulations is the computation of theelectrostatic interactions.Our performance results show improved scaling for bothreaction field and particle mesh Ewald and theimportance of threading, IO, and load-balancing. Thesimulations show the mechanism of lignin collapse atlow temperatures, and the origin of the preferredaggregation of lignin with crystalline cellulose.
    Back to the top of the page

  • Using Dynamic Importance Sampling to Explore Conformational Space in HCV Polymerase
    Ester Sesmero

    Chemistry and Biochemistry, University of Maryland, Baltimore

    Hepatitis C virus (HCV) is a wide spread health concernand causes approximately 35,000 new infections in theU.S. each year. Though there are treatments available, they cause many unpleasant effects and are not completely effective. HCV contains a positive sensesingle-stranded RNA genome and replicates with the aidof RNA dependent RNA polymerase (RdRp). Thispolymerase is known to have two differentconformations: an open inactive conformation and aclosed active conformation. Only the open conformationhave been seen with an inhibitor bound. Our goal is tounderstand how this transition occurs in order todetermine how allosteric inhibitors stop the replication ofHCV. These inhibitors are termed “allosteric” becausethey bind to the enzyme at locations other than theactive site. To accomplish this goal we employ theDynamic Importance Sampling Algorithm (DIMS). DIMSis a pathway finding algorithm that gives informationabout the intermediate states between defined startingand ending points. In our case of study this starting andending points are the coordinates of the open and closeconformations obtained from the MD simulations weperformed of RdRp previously. The DIMS algorithm willallow us to sample the conformations of intermediatesbetween the open and close conformations so we canget a better understanding of how this transition takesplace, what motions facilitate the transition and what roleit plays in enzyme inhibition.
    Back to the top of the page

  • Protein-RNA docking with coarse grained force field
    Piotr Setny

    Physics, Technical University of Munich, Germany

    It is becoming increasingly evident that functional, non-coding (nc) RNAs play important roles in multiplecellular activities, such as RNA processing andmodification, protein trafficking, chromosomemaintenance or enzymatic catalysis. To perform theirfunction, ncRNA molecules typically unite with proteinpartners, making ribonucleoprotein complexes. Structural insight into such assemblies is essential forour understanding of their mechanism of action and theability to design new therapeutic strategies, yet theavailable structural data is still relatively sparse.Computational docking methods can complementparticularly demanding ribonucleoprotein X raycrystallography and provide means for the refinementand integration of low resolution data coming fromrapidly advancing methods such as cryoelectronmicroscopy. We will present a new coarse-grained forcefield forprotein-RNA docking. It is implemented within theframework of ATTRACT program, widely used forprotein-protein docking. Complex structure prediction isbased on energy minimization in rotational and translational degrees of freedom of binding partners, with possible extension to include structural flexibility. The coarse-grained representation allows for fast andefficient systematic docking search without any priorknowledge about complex geometry.The method gives very good results when both bindingpartners are submitted in their bound conformations, andreasonable predictions for unbound geometries. We willpresent different approaches to account for structuralflexibility of the binding partners and show how they canimprove predictions in unbound docking scenario.
    Back to the top of the page

  • Multiscale modeling of the nanomechanics of microtubule protofilaments
    Kelly Theisen

    Chemistry, University of Cincinnati

    Several microscopy studies have shown thatmicrotubule (MT) depolymerization, which takes placeon a milliseconds to seconds time scale, starts with theoutward curling of protofilaments, forming "ram's horn"structures(1). The depolymerization products includecircular structures formed by collections of many tubulindimers(1,2), which may be caused by non-motilekinesins, such as kinesin-13(3). Conversely, MTsevering by AAA-ATPases appears to result in theremoval of a single tubulin dimer(4). We used molecularsimulations of a self-organized polymer (SOP) model(5)of an MT protofilament to quantify structural changesassociated with the stretching and bending thataccompany protofilamentdepolymerization. Oursimulations reveal that substantial bending of theprotofilament can occur in both the outward direction,and towards the interior of the MT cylinder. Also,atomistic simulations revealed that the thermal bendingof protofilaments has no directional preference(6). Mostimportantly, we found that only the bending in theoutward direction leads to the detachment of longprotofilament fragments, which suggests a mechanismfor kinesin-13(3). In contrast, interior bending leads tofragmentation of the protofilaments into individual tubulindimers, which agrees well with the proposed mechanismof MT severing proteins(4). If multiple attachment pointare used(7), the magnitude of the force required to severthe protofilaments compares well with the experimentallyobserved force applied by severing proteins(3,8).
    1. Mandelkow, E., et al., 1991. J. Cell Biol., 114:977
    2. Nogales, E., et al., 2003. Curr. Op. Struc. Biol.,13:256
    3. Moores, C., et al., 2002. Molec. Cell, 9:903
    4. Diaz-Valencia, J., et al., 2011. Biophys. J., 100:2440
    5. Hyeon, Changbong et al., 2006. Structure, 14:1633
    6. Grafmuller, A. and Voth, G.A., 2011. Structure,19:409
    7. White, S., et al., 2007. J. Cell Biol., 176:995
    8. Maier, B., et al., 2002. Proc. Natl. Acad. Sci. USA,99:16012
    Back to the top of the page

  • Substrate protein remodeling by the p97 nano-machine: Computational studies
    Sam Tonddast-Navaei

    Chemistry, University of Cincinnati

    The p97/VCP nanomachine, a double ring member ofthe AAA+ superfamily, is involved in substrate protein(SP) unfolding within the proteasomal degradationpathway. P97 has a homo-hexameric structure with twonucleotide binding domains, D1 and D2, per subunit thatencloses a central pore. ATP hydrolysis leads to largescale conformational changes in D2 domain, which affects the topology of its pore. Conserved loops at the entrance of D2 pore aresuggested to enable SP propagation through the porevia ATP-driven “paddling motion” of Trp551 and Phe552residues. Two other essential residues inside the D2pore, Arg586 and Arg599, contribute to the p97 function. However, it is unclear how p97 interacts with itssubstrate. We hypothesize that the SP, which entersthrough the D1 pore, binds to the Arg599 sites on the D2 cavity lining. Then, repetitive ATP-driven cycles of p97mediate the complete translocation of the SP into the D2pore. To test this hypothesis, we perform implicit solventsimulations of the ssrA-ssrA peptide threading throughp97. Our results confirm the role of Arg599 residues asthe binding sites in the ATP bound state. Thesesimulations reveal that these Arginines interact with thesubstrate primarily via electrostatic interaction with thepeptide's backbone, indicating an interactionindependent of the SP's sequence. By contrast, in the ADP bound state Glu554 has the primary role inreleasing the SP by narrowing the pore radius whichweakens the SP-Arg599 interaction (1). Using the results from implicit solvent simulations, wedevelop a coarse-grained model that extends oursimulations to biologically relevant timescales. Weinvestigate the coupling between ATP-DRIVENconformational changes in the D2 domain of the p97 andremodeling of the SP ( four helix bundle protein fusedwith ssrA). The mechanism of binding to Arg599 and releasing followed by the force exerted by the D2 loops allows the unfolding and translocation of the SP throughthe pore (2).
    (1) S. Tonddast-Navaei, G. Stan (manuscript inpreparation)
    (2) S. Tonddast-Navaei, A. Kravats, G. Stan (Manuscriptin preparation)
    Back to the top of the page

  • Monte Carlo Simulation of a Polyelectrolyte Chain with Permuted Charge Distribution
    Sahin Uyaver

    Computer Science, Istanbul Commerce University

    The effect of a permuted charge distribution of a polyelectrolyte (PEL) chain is investigated by Monte Carlo simulations. The cascade transition of the PEL chain is studied at various charging degrees and in different solvent regimes. The scaling laws are applied. The effect of permuted charge distribution is investigated in comparison to the other charge distribution in the literature. It is observed that very compact structures occur in poor solvent regimes if the system is poorly charged. For some moderate charged case, globular structure is deformed, the charges are moved to a tail occurring. For higher charged case, PEL chain is quite stretched. The effect of permuting charge distribution is observed that pearl-necklace structures, which can occur in a poor solvent if the degree of charging is a light value or a moderate value, are most likely replaced with a deformed globule, a globule having with a tangling tail.
    Back to the top of the page

  • Molecular dynamics simulations of the intra-ringallosteric mechanism of ClpY
    Huan Wang

    Chemistry, University of Cincinnati

    ClpATPases associate with its peptidase partner todegrade proteins in repetitive cycle of ATP hydrolysis. ClpATPases use conformationalchanges coupled with ATP hydrolysis to effect the substrate protein to unfoldand translocate through the central pore to thechambered peptidase for degradation.ClpY ,one of the best structurally characterized memberof ClpATPase, is a hexameric ring-shaped AAA+ motor with central pore formed by a highly conserved loop(GYVG motif) from each subunit. Structural andbiochemical studies of AAA+ ATPases proteins havesuggested several models (rotary, concerted andstochastic) of the progression of allosteric transitionaround the ring.We performed implicit solvent molecular dynamicssimulations to investigate the effect of the ATP-drivenconformational change of intra-ring coupling allostericmechanism[1]. The largest effect of this allosteric motionis found on the neighboring subunit nearest to the activeATP binding pocket. High correlation at subunit-subunitinterface near the active ATP binding site is due tonetworks of salt bridges formed by charged amino acidsR393, E286, E80, E68, R321 residing in the interfaces.Based on these results, we propose that the intra-ringallostery proceeds in the clockwise direction (asdetermined by the axial view point from the proximalentrance of the ClpY pore).To elucidate the substrate remodeling mechanisms, weperformed implicit solvent simulations that describeinteraction between peptides with diverse secondarystructure ( alpha helix/beta turn/random coil) covalentlyattached to a degradation tag (ssrA) and the centralchannel of ClpY. We find that substrate binding ismediated by strong interaction at the Tyr91 site of ClpYand substrate release is effected by competinginteractions at the neighboring Val92 site. In addition, wefind that Tyr91 mutations (Y91A, Y91F, Y91W) reduceaffinity of ClpY for the substrate due to the loss ofcapability to form transient hydrogen bonds. We find thatthe secondary structure of peptides is unraveled duringtranslocation starting from both termini. Translocation ofstructured substrate proteins requires significantly longertimescale than unstructured substrate protein withunfolding as the rate-limiting step. In addition, theclockwise allosteric mechanism is found to be mosteffective due to strong coupling between steps.
    1. Wang, H.; Jayasinghe, M. ; Stan, G. Probing thedirectionality of ClpY’s intra-ring allosteric mechanism, paper in preparation 2012
    Back to the top of the page

  • Testing of the performance of Hybrid Monte Carlo methods
    Tomasz Wirecki

    Department of Molecular Modeling, University of Gdansk, Poland

    One of the most serious limitations of the canonicalmolecular dynamics (MD) method is the small size of thetime step, which is limited by the period of the fastestvibrations of the system under study and usually is of theorder of 1-2 fs. The canonical Metropolis Monte Carlomethod, which is also used to study the behavior of dense systems (in particular biological macromolecules) does not suffer, in principle, from this shortcoming; however, the acceptance rate severely decreases withincreasing the perturbation step size. Recently, methodsresulting from combination of the MD and MCapproaches, termed hybrid Monte Carlo methods,received considerable attention because they combinethe best features of the MD (following theminimum-action direction) and MC (a possibility to take astep back if the chosen move leads to energy increase).In these methods, the Metropolis criterion is applied tothe new configuration of a system obtained from the oldconfiguration after taking several large MD steps. In thiswork, a comparative study was carried out of threevariants of the hybrid Monte Carlo methods: the originalHybrid Monte Carlo (HMC), the Shadow Hybrid MonteCarlo (SHMC) and the Separable Shadow Hybrid MonteCarlo (S2HMC) methods. A simple system composed of108 Lennard-Jones particles in a periodic box waschosen for this purpose. The ergodicity of simulationsand the dependence of the acceptance rate on thenumber of MD steps between Metropolis tests wasstudied. The acceptance rate decreases with anincrease of the dt value. HMC and SHMC behave quitesimilarly, while S2HMC exhibits a significantly higheracceptance rates for higher dt values; consequently, itseems to be the best of the three approaches. S2HMC,however, has a higher acceptance rate than the othertwo methods. All the tested HMC methods presentpromising results. Implementation of the S2HMCalgorithm in the united residue (UNRES) force field forlarge-scale simulations of protein structure anddynamics is pending.
    Back to the top of the page

  • Structure and Dynamics of Repair Complexes of Flap Endonuclease 1 with PCNA and the Checkpoing Clamp 9-1-1
    Xiaojun Xu

    Chemistry, Georgia State University

    Processivity clamps such as PCNA and the checkpointsliding clamp Rad9/Rad1/Hus1(9-1-1) act as versatilescaffolds in the recruitment of proteins involved inreplication, cell-cycle control and DNA repair.Structurally, both PCNA and 9-1-1 are composed ofthree subunits forming closed ring-shaped structuresaround DNA. While PCNA is a homotrimer, the 9-1-1complex is heterotimeric, reflecting the differentialinvolvement of the two clamps with protein partners andtheir distinct roles in coordinating DNA processing. Atrimeric ring can provide multiple binding sites forreplication and repair factors. Furthermore, competitionamong these bound proteins can lead to conformationalswitching and handoffs of these proteins. These are keyprocesses in PCNA biology, which are incompletelyunderstood from a structural perspective. Herein, wehave chosen an integrative computational andexperimental approach to model the assemblies ofFEN1 with its double-flap DNA substrate and each of thetwo clamps. Fully atomistic models of the ternaryDNA/hFEN1/h9-1-1 and DNA/hFEN1/hPCNA complexeswere developed. The models were simulated withmolecular dynamics (MD) in explicit solvent for 100ns toexpose the conformational dynamics of the systems.Clustering analysis of the trajectories revealed the mostdominant conformations accessible to the complexes.The cluster centroids were subsequently used inconjunction with single particle electron microscopy (EM)to generate an EM map of the h9-1-1/Fen1/DNAassembly to 18 Å resolution. Finally, the atomisticmodels were refined by flexible fitting into the EM densityresulting in a 3D structure of the 9-1-1/Fen1/DNAassembly.
    Back to the top of the page

  • The Investigation of the Secondary Structure Propensities and Free-Energy Landscapes of Peptide Ligands by Replica Exchange Molecular Dynamics Simulations
    Fatih Yasar

    Physics Engineering, Haceteppe University, Ankara, Turkey

    The conformational states of three peptide sequences that bind to Staphylococcal enterotoxin B (SEB) aresampled by Replica-exchange molecular dynamicsimulations (REMD) in explicit water. REMD simulationswere treated with 52 replicas in the range of 280-501 Kfor each peptide. The conformational ensembles of eachpeptide are dominated by random coil, bend and turnstructures with a small amount of helical structures foreach temperature. In addition, while an insignificantpresence of β-bridge structures were observed forall peptides, the β-sheet structure was observedonly for peptide II and III. The results obtained fromsimulations at 300 K are consistent with theexperimental results obtained from circular dichroism(CD) spectroscopy. From the analysis of REMD results,we also calculated hydrophobic and hydrophilic solventaccessible surface areas (SASA) for all peptides and weobserved that the hydrophobic segments of the peptidestend to form bend or turn structures. We have alsoobtained the free-energy landscapes of each peptide byprincipal component analysis, to understand how thesecondary structural properties change according totheir complex space. According to free-energy analysis,we have found several minima for each peptide whenthe temperature decreased. For these obvious minima ofeach peptide, it was observed that the random coil, bendand turn structures are still dominant and the helix,β-bridge or β-sheet structures can appear ordisappear with respect to minima. On the other hand,when we compare the results of REMD withconventional MD simulations for these peptides, theconfigurations of peptide II might be trapped in energyminima during the conventional MD simulations.
    Back to the top of the page

  • Probing Conformational Dynamics with the Integration of Markov State Model and Dynamic Neutron and X-Ray Scattering
    Zheng Yi

    UT/ORNL Center for Molecular Biophysics, Oak Ridge National Laboratory

    The conformational changes in biomolecules and theassociated atomic motions up to the microsecond timescale can be directly probed by neutron and X-rayscattering experiments by measuring thetime-dependence of the scattering signal and inferringrelaxation times for the molecular processes. However,many underlying dynamic processes often exist on thesame timescale, which makes it difficult to assigntimescales seen in the experiment to particular structuralrearrangements. Here, molecular dynamics simulationand Markov state modeling was used to connect theconformational changes directly to the exponential decayfunctions in the scattering spectra. This feat has beenaccomplished in two steps: First, Markov State Modelingwas used to decompose the conformational state spaceinto a number of long-lived metastable states, eachhaving a distinct relaxation time. Then a mathematicalframework was established for the direct computation ofthe intermediate scattering function from theeigenvectors and eigenvalues of the Markov model. Thisprocedure allows the establishment of a complete set ofthe exponential decay functions and the fulldecomposition into their individual contributions, e.g. thecontribution of every atom to each relaxation processes,which could be used to guide experimental designs ofselective deuterium substitution to amplify certainprocesses.
    Back to the top of the page

  • Computer Simulation of Membrane Tubulation by EFC F-BAR Domain Lattices
    Hang Yu

    Biophysics and Computational Biology, University of Illinois, Urbana-Champaign

    Cells are dynamically sculpted into many types of compartments by cellular membranes, in some caseswith the help of BAR domain proteins. BAR domainproteins act under in vitro conditions are found to induceformation. of tubules. We have seen in coarse-grained molecular dynamics simulation stretching over 100 microseconds how a flat membrane is curved into a tube when F-BAR domain proteins are arranged on the membrane surface as a regular lattice of parallel rows.The simulations could also characterize the membranebending properties of F-BAR domains in different latticearrangements, showing membrane curvatures with radiiranging from 25 to 100 nm. Lastly, the simulations reveal two key structural featuresof F-BAR domain that facilitate efficient binding tomembranes and membrane curving: (1) Curving ispromoted by close contact between phosphoserine lipidhead groups and clusters of cationic residues along themembrane facing surface of F-BAR domains, namelylysine and arginine residues 30, 33, 110, 113, 114, and139, 140, 146, 150, respectively. (2) Within the 100 nsof contact, the F-BAR domain hinge region, through a 20degree rotation of the helix moment of inertia,establishes a close contact between protein andmembrane. (1) and (2) result in membrane bending on amicrosecond-to-millisecond time scale.
    Back to the top of the page

  • A Lattice Monte Carlo Model of Biomass Degradation Systems
    He Zhang

    Department of Chemistry, University of Wisconsin-Madison

    It has been decades for researchers to search for highly efficient cellulose deconstruction systems, from singleenzymes to organisms. The complexity of these systemsarises from the recalcitrance of biomass. Lignocellulosicbiomass, as a sustainable source of renewable energy,is composed of crystalline cellulose fiber, cross linked byhemicelluloses, coated by lignin and further surroundedby proteins and other biomolecules. Because there is nosingle enzyme can perform such a degradation jobalone, nature has developed many biomassdeconstruction systems involving hundreds of enzymeswith different modularity and even higher level moleculararchitectures. In cellulose deconstruction problem,synergism involves combination of enzymatic activities,cooperation between catalytic modules and non-catalyticmodules, and the way how they form a complexmolecular architecture to increase the rate and yield ofglucose released from cellulose. To study the synergyeffect from molecular perspective, we built a LatticeMonte Carlo Model by calculating Reaction-diffusionprocesses with Metropolis algorithm. We investigatedhow cellulose morphology affects the synergy betweendifferent cellulases. The molecular basis for initial burstreaction is also studied.
    Back to the top of the page

  • In Silico Modeling the Effects of Disease-Causing Missense Mutations
    Zhe Zhang

    Department of Physics, Clemson University, Universite Paris Diderot, Paris

    Synder-Robinson Syndrome (SRS, OMIM, 309583) is amental retardation disorder caused by three missensemutations (G56S, V132G and I150T) in sperminesynthase protein (SMS). The molecular mechanism ofthese three missense mutations causing SRS will bepresented. In addition, the mutability of the missensemutation sites was also explored. The results suggestthat the mutability depends on the details of thestructural and functional factors, and thus can’t bepredicted only based on the evolutionary informationalone. Even the disease-causing sites can still harborharmless missense mutations. Another typical X-linkedmental retardation is creatine deficiency syndrome(CDS, OMIM, 300352), which is associated with thedeficiency on creatine transporter (CRT) encoded bygene SLC6A8. CRT is a transmembrane protein takingup the creatine, which is playing a key role for energysupply in cardiac and skeletal muscle cells, from theoutside of the cell. So far 17 missense mutations havebeen found clinically, 15 of which have been consideredas disease-causing. Different from SMS, there is noexperimental 3-D structure available, so we built the 3-Dmodel of CRT in silico to investigate the moleculareffects of the above mentioned missense mutations.During the above research, we developed a newmethodology named sMMGB to predict protein stabilitychanges due to a single point mutation. By testingsMMGB on a large database of 1109 mutants withexperimental folding free energy change, our protocolachieved RMSD = 1.78 kcal/mol and slope = 1.04 of thelinear regression between the predicted andexperimental values.
    Back to the top of the page