Participants

Invited Talks
Contributed Talks
Software Demonstrations
Posters
Other Participants

Invited Speakers

	Jan Antosiewicz (Department of Physics, Warsaw University)
	Binding of mRNA cap analogues to eIF4E protein: Simulations using SDA package.
The binding of five analogues of the 5'-end mRNA cap, differing in their electrostatic and hydrodynamic properties, to the eukaryotic initiation factor eIF4E, is simulated by means of Brownian dynamics methods, using the SDA package adapted to include nucleic acids fragments as encounter partners for protein receptors. Electrostatic and hydrodynamic models of eIF4E protein and the ligands were prepared using established molecular electrostatics and hydrodynamics simulation methods for predicting ionisation states of titratable groups, adequate for given experimental conditions, and for computing their translational and rotational diffusion tensors. Because factors influencing association rates computed by SDA were thoroughly investigated for several pairs of associating proteins, this investigation is mainly devoted to inspection of significance of those aspects of the simulations which are related to nucleic acids ligands. These include the choice of atoms participating in native contacts with the protein binding site and models for determining charge distribution within the cap analogues, which are subsequently used for computation of the effective charges for these molecules. The diffusional encounter rate constants obtained from simulations are compared with bimolecular association rate constants resulting from stopped-flow spectrofluorimeter measurements, based on analysis of the reaction time courses by numerical integration, assuming either a one-step or a two-step binding mechanism.

	Gianni De Fabritiis (Barcelona Biomedical Research Park (PRBB))
	Extending the time resolution: high performance and multiscale molecular simulations
A fundamental problem of computational approaches of macromolecular systems is the limited time resolution of atomistic based force fields. This computational limitation is due to at least nine orders of magnitudes spanning between atomistic and macromolecular time scales. In this talk, I present two different approaches which go in the direction of extending the time scale of molecular simulations. First, a multiscale method for coupling molecular dynamics (MD) and fluctuating-hydrodynamics (FH) is presented for sound waves reflected by a lipid monolayer. Secondly, the possibility to significantly speed-up molecular simulations codes of approximately 20 times compared to a standard processor by using specialized hardware like the novel Cell BE processor from IBM. References [1] G. De Fabritiis, Performance of the Cell processor for biomolecular simulations, in press Comp. Phys. Commun. (2007). http://www.iac.rm.cnr.it/~gianni/pub/cell_for_md.pdf [2] G. De Fabritiis, R. Delgado-Buscalioni and P. V. Coveney, Modelling the mesoscale with molecular specificity, Phys. Rev. Lett. 97, 134501 (2006). http://www.iac.rm.cnr.it/~gianni/pub/hybridmd.pdf

	Adrian Elcock (University of Iowa)
	Brownian Dynamics of Macromolecular Assemblies
This talk will describe current work in our group aimed at developing a tractable combined model for simulating large molecular assemblies comprising both rigid and flexible elements. The talk will focus in particular on problems that we have overcome and problems that remain to be solved.

	Jan Ellenberg (European Molecular Biology Laboratory)
	Dissecting the contribution of diffusion, interactions and molecular crowding to the mobility of nuclear proteins in living cells
TBA

	Dieter W. Heermann (Institute of Theoretical Physics)
	Lattice and Off-Lattice Simulation Methods for Diffusive Systems
One of the challenges of crowded systems, such as the cell, is how to handle the vastly different time scales that are involved. There are conformational changes of the DNA, packing processes and diffusive processes. In a quenched approximation one may regard the entire matrix as static and only look at diffusive processes neglecting the non static interaction with the matrix. Such a modelling is well suited for lattice Monte Carlo Methods. I will discuss various advantages and short-comings of Monte Carlo Methods based on this approach. In continuum it is some times advantageous to use multi-time step methods. I will contrast these with Monte Carlo Methods.

	Volkhard Helms (Saarland University)
	Diffusional Encounter of Barnase and Barstar
We present an analysis of trajectories from Brownian dynamics simulations of diffusional protein-protein encounter for the well-studied system of barnase and barstar. This analysis reveals details about the optimal association pathways, the regions of the encounter complex, possible differences of the pathways for dissociation and association, the coupling of translational and rotation motion, and the effect of mutations on the trajectories. We found that a small free-energy barrier divides the energetically most favorable region into a region of the encounter complex above the barnase binding interface and a region around a second energy minimum near the RNA binding loop. When applying the same analysis to several barnase mutants, we found that single mutations may drastically change the free-energy landscape and may significantly alter the population of the two minima. Therefore, certain protein-protein pairs may require careful adaptation of the positions of encounter and transition states when interpreting mutation effects on kinetic rates of association and/or dissociation. References Spaar, A. and Helms, V. (2005) J. Chem. Theor. Comp., 1, 723-736. Spaar, A., Dammer, C., Gabdoulline, R.R., Wade, R.C., and Helms, V. (2006) Biophys. J., 90, 1913-1924.

	Gary Huber (Howard Hughes Medical Institute)
	Generic Software Engineering Techniques for Simulations
One of the main obstacles to progress in the field of molecular simulations is the difficulty of modifying and sharing existing code. Ways around these difficulties using the rich templating facilities of C++ will be discussed.

	Jrg Langowski (DKFZ (German Cancer Research Center))
	Coarse-grained modelling of chromatin

	John McCaskill (Ruhr-Universitt Bochum)
	Mesoscale kinetic simulation: from reactive spin lattice models to multipole reactive DPD
TBA

	G. V. Shivashankar (National Center for Biological Sciences, TIFR-Bangalore)
	Spatio-temporal organization of chromatin assembly and function within single living cells
In eukaryotes, the genome is differentially compacted by complexation with histones and other nuclear proteins into euchromatin and heterochromatin which are highly organized within the cell nucleus. Accessing genetic information requires decompaction of chromatin and this is regulated by molecular machines as well as epigenetic codes located on the histones. Spatio-temporal organization and controlled unfolding of local chromatin structure within the nucleus is essential for cellular development and function, and yet the physical bases of these processes are poorly understood. In this context, recent progress in single-cell biophysics combined with systems level approaches provide a new paradigm in understanding chromatin structure and function. Using such methods, my laboratory has, in the recent past explored the physical principles underlying dynamics of self-assembled structures, fluctuations and kinetic barriers in gene regulation. Using the above tools, we are now focusing on probing the coupling between nanoscale chromatin assembly and nuclear architecture in defining cellular transcription control and its memory. I will discuss some of our latest results on the functional implications of the diffusion of histone proteins, transcription factories and gene locus within living cells.

	Kathryn Thomasson (University of North Dakota)
	Brownian Dynamics Simulations of Glycolytic Enzyme/Actin Interactions
The association of glycolytic enzymes to F-actin may be one mechanism through which these enzymes are compartmentalized, and as result, may possibly play important roles in regulation of the glycolytic pathway and substrate channeling. In vitro experiments have shown that several glycolytic enzymes interact with F-actin including: lactate dehydrogenase (LDH), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and fructose-1,6-bisphosphate aldolase (aldolase). These interactions are dependent upon ionic strength. Herein, Brownian dynamics explores the energetics and docking of these three glycolytic enzymes including their ionic strength dependence. Electrostatics are calculated using the Poisson-Boltzmann theory, both the full (FPB) and linearized (LPB) forms. The first encounter snapshots provide clues to the binding modes of the enzymes. All enzymes bind subdomain 1 of the F-actin subunits. Muscle LDH binds using its broad shallow grooves created in the quaternary structure between subunits A/D and B/C. GAPDH, on the other hand, makes use of positive patches on the corners of the enzyme. Aldolase binds F-actin using grooves that are deeper and less broad than LDH, but are also created by subunits A/D and B/C. Thus, the quaternary structure of the enzymes is critical for binding F-actin. These kinds of interactions are observed for several animal species: rabbit, fish, and human. BD also confirms the experimental observations that the association of these enzymes diminishes as ionic strength increases. Both the FPB and LPB give qualitatively similar results, but FPB calculations are statistically more reliable especially at lower ionic strength. The specific residues responsible for salt bridge formation in stable complexes remain the same irrespective of ionic strength. Therefore, BD can reproduce the relative binding of the various enzymes to F-actin, uncover potential binding modes, suggest the importance of specific residues, and reproduce the ionic strength dependence for these macromolecular systems.

	Jose Garcia de la Torre (University of Murcia, Department of Physical Chemistry)
	Hydrodynamics and Brownian Dynamics of Rigid and Flexible Bioparticles in Solution: Theoretical and Computational Aspects
The hydrodynamic properties of macromolecules and nanoparticles in dilute solution are valuable sources of information on their overall conformation shape and, eventually, flexibility). For rigid particles, the connection between their solution properties and the details of their shape can be made using hydrodynamic bead models. Our group has been involved in the development of theoretical aspects, based on the low-Reynolds number hydrodynamics, and generalized-Einstein theories of suspensions of rigid bodies, and their implementation in computational methodologies that originated the HYDRO suite of computer programs. Such procedures include from low-resolution models applicable to large macromolecular complexes (typically studied by electron microscopy or small-angle x-ray scattering), to atomic-level models applicable to any structure of which atomic coordinates are available (including the atomic-level prediction of NMR relaxation of proteins). An overview of these methodologies for rigid particles will be presented. Brownian dynamics plays an essential role in a variety of situations that may be more complex than the mere prediction of simple transport coefficients of rigid bodies. Even for rigid entities, their behaviour in the presence more complex environments (e.g. in presence of obstacles, flows, fields or neighbour particles) may be beyond the reach of explicit theories and direct calculation. Still, the Brownian dynamics can be effectively simulated. I shall present examples of Brownian dynamics (BD) simulations of a single rigid particle in complex environments. When the diffusing particles are flexible, the interplay between conformational variability, internal and overall dynamics makes analytical predictions unpractical, but still knowledge about their Brownian dynamics is accessible through simulation. In our group we have paid attention to the BD simulation of flexible entities in dilute solution. This problem shares with rigid-body bead-model hydrodynamics the complicating but essential effect of hydrodynamic interaction. Actually, our BD simulations are usually carried out for coarse-grained models composed, again, by beads, joined by connectors of different kinds, and with a variety of intramolecular interactions. Indeed, there is an alternative, approximate but sometimes effective treatment of a flexible molecule as a collection of instantaneously rigid conformations, treated by a combination of Monte Carlo simulation and rigid-body hydrodynamics. These methodologies intended for flexible particles are useful in many problems, including synthetic flexible polymers, polyelectrolytes, continuously-semiflexible wormlike molecules (DNA), segmentally flexible macromolecules (antibodies), etc. With an adequate choice of model parameters, this approach can be applied even to atomic-level models of molecules in a continuous, implicit solvent, making possible Brownian (Langevin) dynamics that reach far beyond the time scale available in standard molecular dynamics simulation. Our applications of this approach to various problems has also motivated methodological developments that we are presently building into a suite of BD software. A summary of the methods and some examples will be presented. References * http://leonardo.fcu.um.es/macromol/ * Efficient, accurate calculation of rotational difusin and NMR relaxation of globular proteins from atomic-level structures and approximate hydrodynamic calculations, J. Am. Chem. Soc., 127, 12764-12765 (2005). * MULTIHYDRO and MONTEHYDRO: Conformational search and Monte Carlo calculation of solution properties of rigid and flexible macromolecular models Biophys. Chem., 116, 12-128, (2005) * Calculation of solution properties of flexible macromolecules. Methods and applications. Eur. Biophys. J. 32, 477-486 (2003).

	Joanna Trylska (ICM Warsaw University)
	Binding pathways of ligands to HIV-1 protease
To study ligand-HIV-1 protease binding pathways we performed multi-scale molecular dynamics simulations. These included coarse-grained Langevin and Brownian dynamics, as well as all-atom molecular dynamics in implicit solvent. We have recently developed a one-bead coarse-grained model which allowed for a microsecond time scale simulations. This model enabled us to see full opening of HIV-1 protease flaps which serve as a gate for the approaching peptide substrate. We analyzed the encounter of the substrate with the native HIV-1 protease, the mechanism of substrate incorporation in the binding cleft, and the dissociation of products after substrate hydrolysis. Simulations show that the flaps need to open for the peptide to bind and that the protease interaction with the substrate influences the flap opening mechanism, frequency, and interval. On the other hand, release of the products does not require flap opening because they can slide out from the binding cleft to the sides of the enzyme. Also, a smaller cyclic urea inhibitor (XK263) can reach the binding site when the flaps are not fully open which can explain its faster kinetic features. Based on the analysis of internal fluctuations and correlations of the native and complexed protease, we suggest some mechanistic principles for how the flexibility of the protein may be involved in ligand binding and release.

	Matthias Weiss (German Cancer Research Center)
	Probing the size-dependent diffusion of membrane inclusions and nonequilibrium fluctuations of lipid bilayers with mesoscopic simulations
Experimentally determined diffusion constants are often used to elucidate the size and oligomeric state of membrane inclusions, e.g. rafts. This approach critically relies on the knowledge of the size-dependence of diffusion. We have used mesoscopic simulations to elucidate the size-dependent diffusion of membrane inclusions. For small radii R we find that the lateral diffusion coefficient D is well described by the Saffman-Delbruck relation, yet for large radii we observe an asymptotic scaling D~1/R² which originates from the asymptotic hydrodynamics and the inclusion's internal degrees of freedom. In contrast, the rotational diffusion constant D_r follows the predicted hydrodynamic scaling D_r~1/R² over the entire range of considered radii. We also have studied the effect of active membrane inclusions on the non-equilibrium mechanics of lipid bilayers. The inclusions correspond to two-state integral membrane proteins that change between an open and a closed state, leading to a renormalization of the bending rigidity (as calculated from the fluctuation spectrum) and the area compression modulus. The results show that the observed softening of the membrane is logarithmically dependent on the activity of the inclusions and varies with their concentration. Also, the diffusion of the active inclusions is enhanced with their activity.

	Martin Zacharias (International University Bremen)
	Approximate inclusion of flexibility during Brownian dynamics and docking of proteins
Biomolecular recognition and association events are often coupled to conformational changes in the binding partners. In case of proteins these changes can involve local conformational transitions of amino acid side chains but also global motions of the backbone. For computational efficiency most current docking and Brownian dynamics approaches to study protein-protein association use rigid partner structures. A computationally rapid method has been developed that allows to approximately account for local and global conformational changes during systematic protein-protein docking. To account for possible global deformations the approach employs pre-calculated collective degrees of freedom as additional variables during protein-protein docking. The global collective degrees of freedom are obtained from normal mode analysis using a Gaussian network model for the protein. The results indicate that docking including global flexibility can significantly improve the agreement of near-native docking solutions with the corresponding experimental structures at a very modest increase of computational demands compared to rigid docking. Efforts to treat conformational flexibility using pre-calculated collective modes of proteins during Brownian dynamics simulations of protein-protein/ligand association will also be reported.

Contributed Talks

	Silvia Carlotto (Universit degli Studi di Padova)
	A coarse-grained approach to solvation of nanoparticle colloidal suspensions
Description of solvents as continuous fluids [1] can be adopted to determine the influence of media on properties of molecular solutes nano-sized particles. An useful approach is given by the non-linear fluctuating hydrodynamics approach, which includes the description of a hydrodynamic fluid [2] and the detailed time dependent description of the solute. In order to establish a coherent formalism it is of primary importance to include a coherent description of stochastic process based on fluctuation-dissipation arguments, which give as a result a non-linear fluctuating hydrodynamics description of the embedding solvent, which is a profitable tool for interpreting dynamical phenomena in liquid phases, rheological properties of emulsions and colloids or influence of a surrounding medium dynamics on micro and nano-probes. This general methodology is being tested to study dynamical properties of molecules in solvents. In particular we investigate the rheological behaviour of a suspension structure of aqueous TiO2 nanoparticle suspensions [3] dispersed in pure water. Another molecular system under investigation is the CuO nanofluid [4] having different mean particle size. We shall concentrate on the relationship between viscosity and shear rate and shear stress with different mean particles sizes. [1] J. Tomasi, B. Mennucci and R. Cammi, Chem. Rev., 105, (2005), 2999. [2] H.J. Ko and G.S. Dulikravich, Int. J. Non-Linear Mech., 35, (2000), 709. [3] W.J. Tseng and K.-C. Lin lMater. Sci. Engin. A, 55, (2003), 186. [4] H. Chang, C.S. Jwo, C.H. Lo, T.T. Tsung, M.J. Kao and H.M. Lin Rev. Adv. Mater. Sci., 10, (2005), 128.

	Razif Gabdoulline (BIOMS (Center for Modeling and Simulation in the Biosciences))
	Reduced protein models for multiparticle simulations
The development of efficient reduced models of proteins is needed for simulating long timescale Brownian dynamics of proteins. Non-spherical models are especially interesting, as they are able to describe protein rotations. In these models, the interaction between proteins has a simplified form, with parameters reproducing the interaction properties derived from all-atom representation of proteins with a tunable degree of accuracy.

	Michal Harel (Weizmann Institute of Science)
	On the dynamic nature of the transition state for protein-protein association as determined by double-mutant cycle analysis and computer simulation
The association reaction between two proteins starts with their random collision, followed by a rotational search for the binding orientation, a transition state and finally complex formation. Here we aim to experimentally characterize the nature of the transition state for protein-protein association of three different interactions (Barnase-Barstar, TEM1-BLIP and IFN2-IFNAR2), and use the data to model the transition state structures. To obtain distance constraints of the association transition state, Gint values between interface residues were determined using double mutant cycles. Significant Gint values were obtained only between residues on Barnase and Barstar. However, introducing specific mutations that optimized the charge complementarity between TEM1 and BLIP resulted in the introduction of significant Gint values also between residues of these two proteins. While electrostatic interactions make major contributions towards stabilizing the transition state, we show that for the IFN2-IFNAR2 interaction, steric hindrance exerts an effect as well. To model the transition state structures, we introduced a method for structure perturbation, searching for those inter-protein orientations that best support the experimental Gint values. After a cutoff analysis, two types of transition state were found: specific and diffusive, which were interchangeable by rational design. While the specific transition state can be structurally modeled, the diffusive transition state lacks preferred structures, which may result in a slower association. Furthermore, the Barnase-Barstar transition state region superimposed with the encounter complex region (previously determined by Spaar et al. by Brownian Dynamics). Together, these data provide a structural view of the mechanism allowing rates of association to differ by five orders of magnitude between different protein complexes.

	Karel Jelinek (Polymer Research, BASF Aktiengesellschaft)
	Simulation of Polymer Chain Collapse and Swelling Process
This study is inspired by an experimental measurement of single polymer chain dimensions of temperature-sensitive polymers during the temperature change. Temperature-sensitive polymers exhibit a thermally induced reversible phase transition. They are soluble in a solvent at low temperatures but become insoluble as the temperature rises above the lower critical solution temperature (LCST). It was observed experimentally that poly(N,N-isopropylacrylamide) chains in a very dilute aqueous solution pass through different transition conformations during the single polymer chain collapse induced by increasing temperature above LCST and swelling process during the subsequent cooling. We use the simulations for studying collapse and swelling dynamics of polymer chains in an explicit and implicit solvent. The polymer chain is represented by a simple bead-spring model and the explicit solvent molecules are monomeric with the same size as polymer beads. The FENE potential is used as a bonding potential between monomers and a modified Lennard-Jones potential with tunable minimum depth is used for nonbonded interactions. The collapse and swelling transitions are induced by changing the solvent quality that is controlled by the attractive part of monomer-monomer and/or monomer-solvent interaction potentials.

	Dirk Lebiedz (IWR)
	Modeling subdiffusion using reaction diffusion systems
Within living cells or within their membranes, diffusing species often do not follow Fick's laws, but instead show transient subdiffusive behavior. Formulating spatiotemporal models that take this behavior into account is a delicate matter, as one is usually faced with the choice of resorting either to fractional calculus or to microscopic descriptions. In this talk, we present a novel alternative concept based on multi-scale modeling ideas which is much easier to tackle than the existing approaches. Specifically, starting from the Continuous Time Random Walk model, we construct common linear reaction diffusion systems that can be used as a macroscopic model which captures the characteristic properties of subdiffusion. We show numerical results illustrating the strength of our new approach.

	Yaakov (Koby) Levy (Weizmann Institute of Science)
	Energy landscape of protein self-assembly: Lessons from native topology-based models
Many cellular functions rely on interactions among proteins and between proteins and nucleic acids. In the recent years, we have come to understanding the assembly of the individual actors in this drama thanks to many cooperative experimental and theoretical efforts. We now better understand the main principles of the self-assembly of a single protein chain into its unique structure (i.e., protein folding). However, knowing everything about monomeric proteins does not give a full picture of function. Function requires change of structure and specific recognition to form large assemblies. In my presentation, I will discuss the dynamics and the mechanisms of protein-protein assembly and of protein-DNA recognition that are studied using simplified computational models. Native topology based landscape models, which corresponds to a perfectly funneled energy landscape, reproduces many of the grosser and finer structural and kinetic aspects of various binding mechanisms found in the laboratory. Not only are our computational results consistent with experiments, they also demonstrate that protein plasticity, as envisioned by the fly-casting mechanism, is more fundamental in protein recognition than traditionally imagined. In protein-DNA recognition, both fly-casting effect and electrostatic forces contribute to an efficient binding. Keywords: Protein folding, protein-protein interactions, protein-DNA recognition, protein flexibility, energy landscape

	Domantas Motiejunas (EML Research gGmbH)
	Following protein-protein association towards the bound complex
We present a computational procedure to model protein-protein association and predict structures of protein-protein complexes. The initial sampling stage is based on an efficient Brownian dynamics algorithm that mimics the physical process of diffusional association. Relevant biochemical data can be directly incorporated as distance constraints in the sampling stage. Docked configurations are grouped with a hierarchical clustering algorithm into ensembles that represent potential encounter complexes in the process of protein-protein complex formation. Flexible refinement of selected structures is then done by molecular dynamics simulation. The protein-protein docking procedure was thoroughly tested on 10 structurally and functionally diverse protein complexes. Starting from X-ray crystal structures of the unbound proteins, in 9 out of 10 cases it yields structures of protein-protein complexes close to those determined experimentally with the percentage of correct contacts > 30 % and interface backbone RMSD < 6 . The talk will consist of three parts: 1) description of the computational procedure, 2) overview of the results for 10 test cases and 3) closer look into the docking of Human Leukocyte Elastase and its inhibitor.

	Dimitar Pachov (Brandeis University / Martin Fisher School of Physics)
	Adenylate Kinases (ADKs) at Pressures - Characterization of the Conformational Transition Energy Landscape by MD, TMD and NMR Experiments
Why and how do enzymes undergo conformational changes in order to perform their function? How can we extract information about these transitional pathways and states? The protein Adenylate Kinase (ADK) has two major conformations, the open and closed states. The conformational transition is important for the biological function of the protein in that, 1) the protein has to transform between the two conformations for catalytic function, and 2) the conformational transition is the rate limiting step during the catalytic cycle as shown by NMR experiments. The goal of our computational studies is to answer the questions about "why and how" these conformational transitions happen. We approach this problem indirectly by analyzing how different external pressure conditions affect the dynamics and functions of both P. profundum ADK (Padk), which lives at 700 atm pressure in the deep sea, and its homologue E. coli ADK (Eadk) that lives at ambient pressures. Using NMR, we found that at pressures above 1400atm Padk unfolds, while Eadk remains stable. At the same time, the rate of opening/closing transition in Padk increases with increasing pressures, which indicates that the protein possesses smaller partial molar volume in the transition state compared to its open and closed conformational states. These experimental results were used to evaluate pathways of transitions calculated from pressure MD simulations and TMD simulations. The MD and TMD trajectories provided atomic scale details about the residues and their interactions responsible for both the different adaptation of Padk and Eadk under pressure and the Padk's smaller transition state volume. We believe that the true and complete picture of protein function mechanism can be obtained only by rigorously examining and comparing computational with experimental results. On the other hand, conventional molecular simulations cannot access the micro- to milisecond transitional time scales and at the same time give satisfactory atomic description. Therefore, alternative pathways methods have to be tested on this system to provide better accuracy, prediction and analysis of the conformational regions.

	David Sept (Washington University)
	Multiscale Modeling of Actin Filament Dynamics
Cell motility is an inherently multiscale problem interactions on the molecular or microscopic level manifest themselves as the macroscopic phenomenon of cell movement. Although we have significant structural data showing the atomic detail of actin and many actin-associated proteins, using such information to probe protein dynamics and interactions on a cellular scale is a daunting task. We have developed a multiscale model for actin that is entirely parameterized from molecular dynamics simulations. This reduced model is simple enough to allow us to use a Brownian dynamics approach and simulate the dynamics of filaments with hundreds or thousands of monomers on experimentally attainable timescales. Without adding any additional factors we reproduce macroscopic filament properties such as the persistence length and the extensibility of filaments based on single molecule AFM experiments. The details and experimental comparisons our model will be presented. Keywords: actin, cell motility, Brownian dynamics, multiscale

Software Demonstration

	Razif Gabdoulline (BIOMS (Center for Modeling and Simulation in the Biosciences))
	SDA - Simulation of diffusional association of proteins
Protein association events are ubiquitous in biological systems. Some protein associations and subsequent responses are diffusion controlled in vivo. Hence, it is important to be able to compute bimolecular diffusional association rates for proteins. The Brownian dynamics simulation methodology may be used to simulate protein protein encounter, compute association rates, and examine their dependence on protein mutation and the nature of the physical environment (e.g., as a function of ionic strength or viscosity). SDA computes the rates of diffusional association of 2 proteins given the atomic structure of the bound complex of the 2 proteins. Experimental rates for a series of protein mutants and the dependence of rates on ionic strength can be reproduced well by Brownian dynamics simulations.

	Gary Huber (Howard Hughes Medical Institute)
	UHBD / SMOL

	Jrg Langowski (DKFZ (German Cancer Research Center))
	corchy++ - A Brownian dynamics / Monte Carlo simulation package for flexible polymer chains
TBA

	Kathryn Thomasson (University of North Dakota)
	MACRODOX

	Jose Garcia de la Torre (University of Murcia, Department of Physical Chemistry)
	The HYDRO Suite of Computer Programs for Macromolecular Hydrodynamics in Dilute Solutions.
NOTE: Title is temptative, and abstract will be sumitted later (soon), because this presentation could include work that is still in progress.

Posters

	Denitsa Alamanova (Center for Bioinformatics, University of Saarland)
	Brownian dynamics simulations on a lattice
Protein protein association involves processes on different length and time scales. At large distances, only the relative motion of the two centers of mass is important. In this regime, the free energy landscape can be sampled either by brownian dynamics or Monte Carlo simulations, or by a grid search. At close distances, the internal degrees of freedom and the molecular nature of the solvent become of crucial importance. So far, no efficient approaches exist that allow computing the interaction in this regime except for the final, bound state. Miyashita and co-workers [1] have determined the positions of encounter and transition states at those regions where the computed energy differences between wildtype and mutant proteins fit to experimental data on binding kinetics and strength. This approach allows, at least, to approximate the energy surface at close distances. In order to provide an efficient method to simulate macromolecular diffusion as well as association and dissociation events, we are aiming at a dual description. Here, the spatial dynamics of individual proteins takes place on a coarse grained lattice, while the association dynamics, which is governed by the local interactions, is mapped onto a protein association network, where association is described by adding a link and dissociation by link removal. The required association and dissociation rates may be fitted against experimental or computational data. Here we present results for the spatial component of this hybrid approach, i.e., the problem of how to correctly model diffusion on a coarse grid, given the microscopic diffusion coefficient. In contrast to the time independent random walk methods, our requirement is that transition probabilities are modeled correctly for jumps beyond the next neighbors, too. From the condition that such a propagation should reproduce the diffusive evolution of a free particle at finite times, we derive a condition for discretizing the stepsizes for arbitrary gridspacings, timesteps and diffusion coefficients. [1] Miyashita, O., Onuchic, J. N., and Okamura, M. Y., PNAS 101, 16174 (2004).

	Madhumalar Arumugam (Bioinformatics Institute)
	Computational characterizations of p53-MDM2 interactions
A.Madhumalar(1), Lee Hui Jun(1), David P.Lane(2) and Chandra Verma(1) (1) Bioinformatics Institute 30 Biopolis Street, #07-01 Matrix, Singapore 138671 (2) Institute of Molecular and Cell Biology Biopolis Drive, Proteos, Singapore 138673. p53 is a major tumor suppressor in mammals and is involved in a variety of cellular process like apoptosis, growth arrest and DNA repair. The MDM2 oncoprotein, a natural inhibitor of p53, is important for its degradation. MDM2 amplification has been reported to be a major factor for the inactivation of p53 in certain cancers. This interaction is modulated by phosphorylation of both, p53 and MDM2, at certain critical residues. Extensive MD simulations carried out in our lab showed the importance of the phosphorylation of Thr18 of p53 peptide in disrupting the binding to MDM2, in contrast to the other residues Ser15, Ser20. While this yielded information on the role of noncolavnet interactions it does not reveal quantitative data on the rate of association, i.e., the kinetic barrier(s) associated with the formation of the complex. Brownian dynamics is a computational method by which we aim to address the dynamics of association of phosphorylated /non-phosphorylated p53 peptide and the MDM2 protein

	Silvia Carlotto (Universit degli Studi di Padova)
	A coarse-grained approach to solvation of nanoparticle colloidal suspensions
S. Carlotto and A. Polimeno Universit di Padova, Dipartimento di Scienze Chimiche, Via Marzolo 1, I-35131 Padova, Italy Description of solvents as continuous fluids [1] can be adopted to determine the influence of media on properties of molecular solutes nano-sized particles. An useful approach is given by the non-linear fluctuating hydrodynamics approach, which includes the description of a hydrodynamic fluid [2] and the detailed time dependent description of the solute. In order to establish a coherent formalism it is of primary importance to include a coherent description of stochastic process based on fluctuation-dissipation arguments, which give as a result a non-linear fluctuating hydrodynamics description of the embedding solvent, which is a profitable tool for interpreting dynamical phenomena in liquid phases, rheological properties of emulsions and colloids or influence of a surrounding medium dynamics on micro and nano-probes. This general methodology is being tested to study dynamical properties of molecules in solvents. In particular we investigate the rheological behaviour of a suspension structure of aqueous TiO2 nanoparticle suspensions [3] dispersed in pure water. Another molecular system under investigation is the CuO nanofluid [4] having different mean particle size. We shall concentrate on the relationship between viscosity and shear rate and shear stress with different mean particles sizes. [1] J. Tomasi, B. Mennucci and R. Cammi, Chem. Rev., 105, (2005), 2999. [2] H.J. Ko and G.S. Dulikravich, Int. J. Non-Linear Mech., 35, (2000), 709. [3] W.J. Tseng and K.-C. Lin lMater. Sci. Engin. A, 55, (2003), 186. [4] H. Chang, C.S. Jwo, C.H. Lo, T.T. Tsung, M.J. Kao and H.M. Lin Rev. Adv. Mater. Sci., 10, (2005), 128.

	Anna Feldman-Salit (EML Research gGmbH)
	Protein Recognition Processes: The Cysteine Synthase Complex
Plants and bacteria can assimilate and incorporate inorganic sulfur into organic compounds such as the amino acid, cysteine. Cysteine biosynthesis involves a bienzyme complex, the cysteine synthase (CS) complex. The CS complex is composed of the enzymes Serine-Acyl-Transferase (SAT) and O-Acetyl-Serine-(Thiol)-Lyase (OAS-TL). The biological function of the CS complex and the mechanism of reciprocal regulation of the constituent enzymes are still poorly understood. We are investigating the SAT and OAS-TL enzymes from Arabidopsis thaliana mitochondria and their complexation using different computational techniques. The three-dimensional structures of the enzymes were modeled by homology modeling. The complex formation was first modeled with rigid-body Brownian Dynamics (BD), simulating enzymes diffusional association. The experimental data was incorporated into BD to increase the number of good orientation structures. The conformational changes, occurring upon tighter binding of the enzymes, have been then explored with flexible-body Molecular Dynamics (MD).

	Holly Freedman (University of Alberta)
	Entry of taxol through microtubule pores to access its interior binding site evaluated by geometric description of pores and estimation of the drugs mean rate of passage using Brownian dynamics
Microtubules are cytoskeletal components involved in maintaining cell structure and making up the mitotic spindle during cell division. They take on the form of long, hollow and cylindrical protein complexes assembled from a number of protofilaments, generally 13, composed of alpha/beta tubulin dimers. The anti-mitotic drug taxol, commonly used in cancer treatment, is known to bind to a site within the microtubule interior. Experimental evidence suggests that taxol may gain access to its tubulin binding site by entering through pores between individual tubulin dimers in the microtubule walls. Four different types of pores exist, two for each kind of lattice that can arise: the B lattice in which alpha and beta proteins of aligned protofilaments face similar units, and the A lattice with the alternate pattern of alpha and beta units respectively facing each other. Molecular modeling of these pores shows that cross-sectional dimensions are comparable in size to the taxol itself and indicates that taxol must take advantage of its flexibility to assume an extended conformation allowing for passage. Measurements of pore sizes are complicated by the existence of an assortment of isotypes for both beta and alpha protein units. Distributions of the different isotypes are dependent upon cell tissue type, so that the frequency of isotype expression may lead to microtubules with different physical properties, including taxol-sensitivity, in various tissues of the human body. It is important to understand on the molecular level correlations between taxol-sensitivity of microtubules and their isotype composition. In particular, we are investigating whether tubulin isotype composition may influence the rate of taxol entry through microtubule pores. This is significant in that it may demonstrate the effect of size on the design of more potent taxol analogs, as well as provide a knowledge of the effect of pore size on the localization of taxol in various cell types, useful for the consideration of harmful drug side effects. That size may be a limiting factor in drug binding has already been demonstrated by the cytotoxicity of the smaller taxol-like drug epothilone, known to bind to the same site as taxol, in taxol-resistant mutant cell lines where it is possible that diffusion of taxol through pores is more restricted. We are working on applying the flexible Brownian dynamics utility implemented in the University of Houston Brownian Dynamics program package to estimate the mean rate of passage of the taxol drug through microtubule pores. Such a computational study should allow this proposed means of entry to be evaluated, and compared for different tubulin isotypes.

	Tihamer Geyer (Center for Bioinformatics, Saarland University)
	Brownian Dynamics with explicit ions and their effects on the association of small proteins
Brownian Dynamics is often used to determine association rates or binding dynamics of biological molecules. For this, not only the solvent molecules are traced out, but also the interactions between the proteins have to be adapted to account for the physiological environment. By simulating the association of cytochrome c to a simplified membrane protein under physiological conditions we show that the assumption of a continuous distribution of counter ions is at least questionable, when the binding dynamics of small proteins are considered and how the explicit inclusion of the ions into the Brownian dynamics simulation modifies the binding rates and the relative density distributions compared to a description with the shielded Coulomb interaction of a continuous ion density.

	Magdalena Gruziel (University of Warsaw)
	Electrostatic properties of tRNA in the Poisson-Boltzmann model - comparison of methods and conditions
Magdalena Gruziel, Pawel Grochowski and Joanna Trylska Proper description of the electrostatic potential around biomolecules is crucial in simulations of molecular association, recognition and binding. For moderately charged molecules the Poisson-Boltzmann implicit solvent model has been widely used for these purposes. These studies aim at testing its applicability to highly charged systems such as nucleic acids. The sensitivity of the electrostatic potential to various conditions of the surrounding were tested on the transfer RNA (tRNA) molecule which crystallizes with numerous divalent metal ions and water molecules. We describe the effect of explicit divalent ions and water molecules, ionic strength of the dielectric medium, and the linear approximation of the Poisson-Boltzmann equation on the electrostatic potential around tRNA.

	Michal Harel (Weizmann Institute of Science)
	On the dynamic nature of the transition state for protein-protein association as determined by double-mutant cycle analysis and computer simulation
The association reaction between two proteins starts with their random collision, followed by a rotational search for the binding orientation, a transition state and finally complex formation. Here we aim to experimentally characterize the nature of the transition state for protein-protein association of three different interactions (Barnase-Barstar, TEM1-BLIP and IFN2-IFNAR2), and use the data to model the transition state structures. To obtain distance constraints of the association transition state, Gint values between interface residues were determined using double mutant cycles. Significant Gint values were obtained only between residues on Barnase and Barstar. However, introducing specific mutations that optimized the charge complementarity between TEM1 and BLIP resulted in the introduction of significant Gint values also between residues of these two proteins. While electrostatic interactions make major contributions towards stabilizing the transition state, we show that for the IFN2-IFNAR2 interaction, steric hindrance exerts an effect as well. To model the transition state structures, we introduced a method for structure perturbation, searching for those inter-protein orientations that best support the experimental Gint values. After a cutoff analysis, two types of transition state were found: specific and diffusive, which were interchangeable by rational design. While the specific transition state can be structurally modeled, the diffusive transition state lacks preferred structures, which may result in a slower association. Furthermore, the Barnase-Barstar transition state region superimposed with the encounter complex region (previously determined by Spaar et al. by Brownian Dynamics). Together, these data provide a structural view of the mechanism allowing rates of association to differ by five orders of magnitude between different protein complexes.

	Jitka Havrankova (Charles University in Prague)
	Fluorescence Correlation Spectroscopy: Monte Carlo simulations
Fluorescence correlation spectroscopy (FCS) is a non-invasive technique that enables to study fluorescently labeled systems. A very small volume of a sample is illuminated by a laser and the autocorrelation function of the fluctuating fluorescence intensity is recorded. The fluctuations are driven mostly by the diffusion of fluorophores. The obtained signal is usually fitted to simple analytical models. We attempt to address more subtle features of experimental results by simulations considering processes not accounted for in the simple analytical models, including rotational and surface diffusion, triplet states, rate processes, etc. keywords: fluorescence correlation spectroscopy, Monte Carlo, simulation, rotational diffusion

	Sebastien Huet (European Molecular Biology Laboratory)
	Implications of molecular crowding on nuclear protein dynamics
In the nucleus of eukaryotes, the genetic material is organized into "stable" and functional membraneless subcompartments. Among them heterochromatin and nucleoli are the two most prominent, and these two structures contain large amounts of cosolutes such as DNA, proteins or RNA in comparison to bulk euchromatin. This dense nature is expected to result in a crowded microenvironment for nuclear proteins. In vitro, molecular crowding has been shown to significantly alter protein behavior in inducing volume exclusion, slowing down diffusion and enhancing association with binding partners. These molecular crowding consequences have been hypothesized to apply also in-vivo. To test this, we assayed the behavior of diffusive tracers or generic chromatin binders in heterochromatin or nucleoli in comparison to euchromatin using confocal microscopy, photoactivation and fluorescence correlation spectroscopy in live cells.

	Voltz Karine (DKFZ)
	Parameterisation of a coarse-grained model for the nucleosome
In eukaryotic cells, DNA is packed with histone proteins into chromatin. Because many biological processes require a free DNA as a substrate, the chromatin structure needs to transit between tight and loose states. To investigate this long time-scale dynamical process, we performed coarse-grained molecular dynamics on the nucleosome, the basic unit of chromatin. Here we present a method for parameterising the force-field for this model based on a Boltzmann inversion of the radial distribution function obtained from crystallographic and all-atom molecular dynamics data. This strategy has already been powerful in the case of the ribosome study.

	Peter Kosovan (Charles University in Prague)
	Molecular Dynamics Simulations of Weak Polyelectrolytes
Many biomacromolecules are polyelectrolytes and almost all of them are weak polyelectrolytes. The behaviour of weak polyelectrolytes is different from that of the strong ones. The degree of dissociation of weak polyelectrolytes is dependent on the solution pH, ionic strength and local permittivity. The charges on a weak polyelectrolyte are mobile, e.g. a COOH group of a polyacid can dissociate and re-associate. Estimates of dissociation rate constants suggest that this happens on the same time scale as the dynamics of the polymer chain. Therefore the dynamics of the dissociation has to be considered in the simulations of weak polyelectrolytes. Current Molecular Dynamics simulation methods do not provide a tool for simulation of the reversible dissociation process in a weak polyelectrolyte. In our work we present several ideas on how this could be done in a coarse-grained polymer model. We are at the very beginning of the project, but we can already present some preliminary simulation results. keywords: polyelectrolyte, simulation, coarse-graining, dynamics

	Thrse E. Malliavin (Unite de Bioinformatique Structurale/Institut Pasteur)
	The complex between protein EF from Bacillus anthracis and calmodulin: toward a simplified model
Elodie Laine, Julliane Hugenin, Arnaud Blondel et Thrse E. Malliavin Unite de Bioinformatique Structurale, Institut Pasteur, Paris terez@pasteur.fr Three proteins from Bacillus anthracis are required in order to infect macrophages. Among them, the oedema factor (protein EF) is an adenylyl cyclase. It catalyzes the production of cyclic AMP from ATP, and consequently induces the cell death. EF is active only when bounded to calmodulin, and the Xray-crystallographic structure of he EF-calmodulin complex have been solved (Drum et al, 2002). The interaction of EF with calmodulin is modulated by the level of Calcium complexation, and is optimal if the two lobes of the C terminal domain are complexed by Calcium ions. Three molecular dynamics simulations of the EF-calmoduline complexed have been realized using three different levels of Calcium complexation. One purpose of these simulations was to obtain a simplified description of the complex. The interaction energies between the different domains of EF and calmodulin were analyzed (Hamacher et al, 2006) in order to determine the mutual influence of the different parts of the complexe. Different influence patterns are obtained for the different levels of Calcium, and they can be qualitatively related to experimental observations made on the complex. The atomic fluctuations by residues were also analyzed in the frame of an elastic network (Bahar et al, 1997), by recalculating the Kirchof matrices corresponding to the fluctuations. The obtained matrices fit more or less to the elastic network model, according to the harmonicity of the molecular dynamics trajectory. Indeed, the trajectory recorded with 2 Calciums complexed to calmodulin visits a unique basin of the conformational hypersurface and produces a matrix fitting well to the elastic network model. On the other hand, the trajectory recorded with 4 Calciums visits several basins, and the fit of the matrix to the elastic network model is worse. These dynamics and energetic analyses permit to propose simplfied models for fluctuations and interactions inside the complex. References * Bahar I, Atilgan AR, Erman B. Direct evaluation of thermal fluctuations in proteins using a single-parameter harmonic potential. Fold Des. 1997;2(3):173-81. * Drum CL, Yan SZ, Bard J, Shen YQ, Lu D, Soelaiman S, Grabarek Z, Bohm A, Tang WJ. Structural basis for the activation of anthrax adenylyl cyclase exotoxin by calmodulin. Nature. 2002 Jan 24;415(6870):396-402. * Hamacher K, Trylska J, McCammon JA. Dependency map of proteins in the small ribosomal subunit. PLoS Comput Biol. 2006 Feb;2(2):e10. Epub 2006 Feb 17.

	Domantas Motiejunas (EML Research gGmbH)
	To be submited separately (Poster)
To be submited separately (Poster)

	Georgi Pachov (EML Research gGmbH)
	Simulation of linker histone in chromatin fiber
The dynamics and interaction between protein-protein and protein-DNA molecules appearing on different time and length scales in the cell are of fundamental interest. In chromatin, linker histone binds to the highly charged nucleosome particle driven by electrostatic attraction. However, its position and orientation with respect to the nucleosome core as well as its binding sites are unresolved problems. Using implicit representation of the surrounding molecules, rigid-body docking of linker histone to the nucleosome particle (SDA software) has been performed revealing how the protein binds to DNA on the nucleosome. Simulation of diffusional motion of both molecules for computation of the association rates and encounter complex formation is under study with SDA. The results will be integrated in higher-order simulation of chromatin fiber.

	Horacio Emilio Prez Snchez (Institut fr Nanotechnologie, Forschungszentrum Karlsruhe)
	Dynamics of bent DNA studied through Brownian dynamics
Horacio Emilio Prez Snchez(1), Jose Gins Hernandez Cifre(2), Francisco Guillermo Daz Baos(2), Jos Garca de la Torre(2) (1) Institut fr Nanotechnologie, Forschungszentrum Karlsruhe GmbH, Postfach 3640, D-76021 Karlsruhe (Germany) (2) Departamento de Qumica-Fsica, Facultad de Qumica, Campus de Espinardo, 30100, Espinardo, Murcia (Spain) In some nucleic acids permanent bends have been identified. In addition, it is known that nucleic acids show a certain degree of flexibility. Different experimental techniques, based on solution properties, for example transient electric birefringence, have been used to characterize the rotational dynamics of DNA and RNA in solution. But full interpretation of experimental results need of complex analytical theories which in some cases are not available. In those cases methodologies based on computer simulation appear as a valuable tool [1,2,3]. We present a simulation methodology using Brownian dynamics and rigid body hydrodynamic modelling that can help us in the interpretation of experimental data to characterize permanent bends and flexibility of nucleic acids. This general methodology is applied, using a hydrodynamic model to B-DNA with possible flexible bends. We discuss the interpretation of experimental data obtained by electro-optical techniques like transient electric birefringence in comparison with our simulations. [1] Garcia de la Torre J, Perez Sanchez HE, Ortega A, Hernadez JG, Fernandes MX, Diaz FG, Lopez Martinez MC, 'Calculation of solution properties of flexible macromolecules. Methods and applications', Eur. Biophys. J., 32, 477 (2003). [2] Perez Sanchez HE, Garcia de la Torre J, Diaz FG,'Influence of field strength and flexibility on the transient electric birefringence of segmentally flexible macromolecules', J. Phys. Chem. B, 106,6754, (2002). [3] Perez Sanchez HE, Garcia de la Torre J, Diaz FG,'Transient electric birefringence of wormlike macromolecules in electric fields of arbitrary strength: a computer simulation study', J. Chem. Phys., 122, 124902 (2005).

	Julia Romanowska (Warsaw University, ICM)
	Dynamics of aminoglycosidic binding site in the small ribosomal subunit
Julia Romanowska(1), Piotr Setny(1,2), Joanna Trylska(2) (1) Department of Biophysics, Faculty of Physics, Warsaw University, Zwirki i Wigury 93, 02-089 Warsaw, Poland (2) ICM, Warsaw University, Pawinskiego 5a, 02-106 Warsaw, Poland Most of aminoglycosidic antibiotics bind to the 16S RNA in the small ribosomal subunit. They interfere with the process of translation by decreasing the accuracy of codon recognition which results in blocking proper peptide synthesis. Aminoglycosides suffer from toxicity and low bioavailability. Moreover, bacterial resistance to antibiotics is limiting their effectivenes in medical therapy, therefore, there is a widely recognized need to understand their binding mechanism and improve their selectivity and efficiency. The sequence of the binding cleft differs in procaryotic and eucaryotic 16S RNA. The major difference is an adenine into guanine substitution in position 1408. The aim of this work is to aid our understanding of how this substitution prevents aminoglycosides from binding to eucaryotic ribosome by studying the differences in dynamic and electrostatic properties between procaryotic binding region and its eucaryotic counterpart. We study the conformational flexibility of a 22-nucleotide RNA fragment which mimics the aminoglycoside binding site in the small ribosomal subunit. It was proven experimentally that aminoglycosides bind in the same manner to this oligonucleotide as to the entire subunit. Molecular dynamics is performed with explicit solvent and ions which allows us to analyze solvent dynamics in the binding region, as it is known that some contacts betweeen antibiotics and RNA are mediated by water molecules.

	Outi Salo-Ahen (EML Research gGmbH)
	Searching for the Hot Spots in the Dimer Interface of the Human Thymidylate Synthase
Thymidylate synthase (TS) is an essential enzyme for DNA synthesis as it catalyses the reductive methylation of dUMP to dTMP. TS inhibitors have been developed and used as anticancer agents. Unfortunately, this homodimeric enzyme often develops resistance toward inhibitors, for example, by interrupting the autoregulatory inhibition of TS synthesis. In an attempt to avoid the cellular drug resistance of TS, one possible strategy to inhibit TS is by interfering with the enzyme dimerization. We are aiming to identify small molecules or peptides that could modulate the protein-protein interactions at the dimerization interface of TS. Here we describe the analysis of the TS dimer interface for hot spots. Among others, MolSurfer [1], FoldX [2] and PASS [3] results are discussed. [1] Gabdoulline RR, Wade RC, Walther D. Nucleic Acids Res. 31: 3349-3351, 2003 [2] Schymkowitz J, Borg J, Stricher F, Nys R, Rousseau F, Serrano L. Nucleic Acids Res. 33: W382-388, 2005 [3] Brady GP Jr, Stouten PFW. J. Comp.-Aided Mol. Des. 14: 383-401, 2000

	Alexander Spaar (Department of Cell Biology and Oncology, Consorzio Mario Negri Sud)
	Exploring the free energy landscape of protein-protein association using BD simulations
The computed rates and relative trends of Brownian dynamics (BD) simulations for protein-protein association are generally in good agreement with experimental results. Here, we report the computation of the free energy landscape for the encounter process by a novel analysis, where the individual positions and orientations are recorded along the trajectories of BD simulations and stored in so-called occupancy maps. These occupancy maps can be interpreted as a probability distribution which allows the calculation of the entropy landscape. From the free energy landscape of protein-protein encounter, obtained by summing the energy and entropy contributions, we deduced details of the association process like the optimal association pathways and the encounter complex, which is defined as the state of minimal free energy before the formation of the bound complex. We applied this analysis to study the association of barnase to barstar, a well characterized model system of electrostatically steered diffusional encounter, and of water-soluble cytochrome c to membrane-bound cytochrome c oxidase. For the first system, we found that for low and medium ionic strength, a small free-energy barrier divides the energetically most favorable region into a region of the encounter complex and a region near the RNA binding loop. However, single mutations may drastically change the free-energy landscape and may significantly alter the population of the two minima. For the second system, we investigated the effect of the membrane on the association behavior. Interestingly, the presence of the membrane had only little effect on computed association rates and on the free energies of the encounter states. As expected, the optimal association pathways largely overlapped for both cytochrome c species associating towards soluble COX and COX embedded in an uncharged membrane. Remarkably, after switching on the lipid partial charges, both cytochrome c species were strongly attracted by the inhomogeneous charge distribution of the zwitterionic lipid head groups and showed preferential diffusion parallel to the DPPC membrane surface.

	Dmitry Tworowski (Biona Biotech Research Company Ltd., and Weizmann Institute of Science)
	Diffusional Association of tRNA with Aminoacyl-tRNA Synthetases and Elongation Factors: Continuum Electrostatics and Brownian Dynamics Study
tRNA substrates bind to a broad range of target macromolecules that play crucial role in the maintenance of the genetic code. Among them are 20 types of aminoacyl-tRNA synthetases (aaRS), elongation factor Tu (Ef-Tu), and ribosomal sites. For the most part these macromolecules are negatively charged at physiological conditions, as also are tRNAs. Nevertheless, there are driving forces that make possible the fast diffusional association of like-charged molecules, thus favouring the formation of correct encounters and docking events. 3D-isopotential surfaces generated by aaRSs and EfTu at different contour levels, reveal the presence of positive patches (blue spaces), complementary to tRNA molecule. Certain charged residues on the protein-tRNA interface contribute to the long-range potential that first attracts tRNA nonspecifically. Further Brownian Dynamics (BD) simulations have confirmed the existence of electrostatic funnels towards tRNA binding sites. Moreover, BD approach correctly reproduces the experimental values of the rate association constants obtained for PheRS, GluRS and EfTu.

	Peter Winn (EML Research gGmbH)
	PROSURF - Potein Surface Docking With Brownian Dynamics
Recent combinatorial biotechnologies have shown that the molecular recognition capability of proteins can be specifically oriented toward inorganic surfaces. However, at present the principles regulating protein-surface interactions are poorly understood, thus hindering the effective exploitation of such specificity in scientific and technological application. What features of the surface and of the proteins (electronic, structural, morphological) determine which protein is able to bind to a given surface and how? PROSURF (http://www.s3.infm.it/prosurf/) will address these issues and develop computational tools to enable rational design of protein-surface associations. Here we discuss the use of Brownian Dynamics simulations to simulate protein surface simulations. Surfaces of interest range from biological surfaces, such as membranes, to metallic surfaces such as gold.

	Tomasz Wocjan (Biophysics of Macromolecules, DKFZ Heidelberg)
	Brownian dynamics simulation of DNA unrolling from the nucleosome
In eukaryotic cells DNA is compacted into chromatin. The basic packing unit, the nucleosome, consists of a histone protein around which ~ 160 bp of DNA are wrapped in two tight turns. The mechanism by which this structure can be opened, giving access to DNA-processing enzymes is of fundamental biological importance. Here we develop a model for the attachement of DNA on the histone core and simulate the unwinding of DNA from the nucleosome by mechanical forces, a process which has been analyzed experimentally [1,2]. We use a Brownian dynamics simulation [3] with a coarse-grained model in which the linear DNA is represented by a chain of rigid segments interacting through potentials for torsion, bending and stretching. A renormalized Debye-Hckel potential includes electrostatic interaction between the chain segments, while hydrodynamics are treated using the Rotne-Prager tensor. The interaction between the negatively charged DNA and the positively charged surface of the histone octamer (with cylindrical geometry) is modeled by a short-ranged potential. The dynamic structure of the nucleosome is studied in dependence on the extension or force applied to the DNA, which allows to analyze the transition from the conformation with the DNA helically coiled around the histone core to the state of fully extended DNA. The rupture force is obtained as function of the stretching velocity of the DNA. [1] B.D. Brower-Toland, C.L. Smith, R.C. Yeh, J.T Lis, C.L. Peterson, M.D. Wang, PNAS 99, 4 (2002) [2] L.H. Pope, M.L. Bennink, K.A. van Leijenhorst-Groener, D.Nikova, J.Greve, J.F. Marko, Biophys. J. 88, 3572-3583 (2005) [3] K. Klenin, H. Merlitz, J. Langowski, Biophys. J. 74 780-788 (1998)

	Mirco Zerbetto (Universita' degli Studi di Padova)
	Stochastic modeling of CW-ESR spectra of fullerene adducts with nitroxide probes
M. Zerbetto, A. Polimeno, C. Corvaja, L. Franco, M. Maggini Electron spin resonance measurements performed on macromolecules via site directed spin labeling are highly sensitive to molecular motions [1] and in particular, spectra from multi-probe modified molecules are very reach in informations, both structural and dynamic. However, in order to analyze in full the experimental observations, support of advanced theoretical models is required. We present here a protocol for the interpretation of continuous wave electron spin resonance spectra of bi-radical species in solution, based on the Stochastic Liouville Equation formalism [2] that provides a method to directly include motional dynamics in the form of stochastic (Fokker Planck / diffusive) operators in the spin super Hamiltonian H governing the time evolution of the system, in connection with integrated ab-initio QM determination of structural and magnetic properties. [3] As case studies, we applied the model to the interpretation of CW-ESR spectra in toluene of the five isomers of the [60]fulleropyrrolidine bis-adduct with nitroxide addends and of the reduced D3 [60]fulleropyrrolidine tris-adduct with nitroxide addends, in the temperature range of 270 350 K. Modeling is based on the definition of the spin Hamiltonian which includes exchange and dipolar interactions and in the introduction of motions in the form of a Smoluchowski diffusion operator describing the overall tumbling of the molecule with respect to the inertial laboratory frame. [1] W.L. Hubbel, H.S. Mchaourab, C. Altenbach, M.A. Lietzow, Structure, 4, 779 (1996); H.J. Steinhoff, Frontiers in Bioscience, 7, 97 (2002) [2] D.J. Schneider, J.H. Freed, Adv. Chem. Phys., 73, 387 (1989); A. Polimeno, J.H. Freed, J. Phys. Chem., 99, 10995 (1995); A. Polimeno, J.H. Freed, Adv. Chem. Phys., 83, 89 (1992) [3] A. Polimeno, V. Barone, Chem. Phys. Phys. Chem., 8, 4609 (2006); V. Barone, M. Brustolon, P. Cimino, A. Polimeno, M. Zerbetto, A. Zoleo, J. Am. Chem. Soc., 128, 15865 (2006)

Other Participants

Bolaji Akanbi (Unversity of Ilorin)
Mithun Biswas (IWR, University of Heidelberg)
Vlad Cojocaru (EML Research gGmbH)
Francesca De Rienzo (Dipartimento di Chimica - Universit di Modena e Reggio Emilia)
Sulaiman Faisal (EML Research gGmbH)
Divita Garg (EML Research gGmbH)
Stefan Henrich (EML Research)
Jos G. Hernndez (Universidad de Murcia)
Jana Humpolickova (J. Heyrovsky Institute of Physical Chemistry )
Tamar Ickitidze (New Challenges for Sustainable Development, NCSD)
Konstantin Klenin (Forschungszentrum Karlsruhe)
Xiaofan Li (Cancer Research UK London Research Institute)
Thomas Maeke (BioMIP, Ruhr Universitt Bochum)
Michael Martinez (EML Research gGmbH)
Stefan Richter (EML Research gGmbH)
Matthias Stein (EML Research gGmbH)
Filip Uhlik (Charles University in Prague)
Rebecca Wade (EML Research gGmbH)