Page 57 - ASME_NEMB_2016_Program
P. 57

TRACK 4                                                 TRACK 4                      Technical Program




        moment, then those forces do not contribute to the dynamics and therefore   Technical Presentation. NEMB2016-6134
        may be eliminated from the model. The use of this method greatly reduces
        the forward dynamics computational burden without significant deviation in   Taha Goudarzi, Nahil A. Sobh, UIUC, Urbana, IL, United States
        results. This method has previously been applied to dynamic simulations of
        motor proteins and estrogen molecules.                  The analytical tools at our disposal for characterizing the mechanical behav-
                                                                ior of indented materials in a nanoindentation test with a spherical probe
        The model used in this work is the Ebola virus glycoprotein. This virally en-  (usually suitable for soft/bio materials) are limited to the Hertz’s contact solu-
        coded glycoprotein is bound in the virion envelope and has been shown   tion and its extensions. To overcome this deficiency many researchers have
        to play a pivotal role in the mechanism of infection. Simulations of the in-  used finite element simulations to model the nanoindentation tests.
        teraction of this protein with host cell receptors could greatly aid on-going   To characterize the behavior of indented materials in an inverse manner
        research for effective treatments.                      using the finite element simulation, one needs to run the nanoindentation
                                                                simulation with a guess for the material properties which are the unknown of
        The glycoprotein model has full atomic resolution and is assembled piece-  the problem and then compare the simulation results with the experimental
        wise from a user-input sequence of amino acids. Given the large number of   data to see how good of a guess has been made. The computational cost
        degrees of freedom, the Featherstone algorithm is used to numerically solve   of such an approach has been an obstacle in development of inverse com-
        the forward dynamics. This algorithm is ideal for biopolymer simulations   putational methods and tools for material characterization based on results
        as it precludes a symbolic model and is easily parallelizable. Active forces   of nanoindentation tests. We have developed an online software (https://
        include electrostatic and Lennard-Jones interaction potentials, viscous flu-  nanohub.org/tools/nanoindentation) that can characterize the mechanical
        idic damping, and stochastic Brownian motion arising from non-modeled   behavior of indented materials in an iterative approach using finite element
        collisions with the medium. This work is the first application of the multiscale   simulations. In order to maintain the computational cost under a reasonable
        method to a high-resolution protein model and a simulation using Feather-  level very effective optimization methods have been utilized.
        stone algorithms.                                       In this online tool that is the first to a series of tools focusing on more com-
                                                                plicated problems (e.g. inelastic constitutive behavior and boundary condi-
        4:20pm  Adhesion Mechanisms of Curli Subunit CsgA       tions) the concentration is on the nanoindentation of isotropic elastic materi-
                                                                als with spherical probes under arbitrary boundary conditions.
        Technical Presentation. NEMB2016-5997
                                                                5:00pm  Atomistic-to-continuum rod modeling of DNA mechanics
        Elizabeth P. DeBenedictis, Jenny Liu, Sinan Keten, Northwestern
        University, Evanston, IL, United States                 Technical Presentation. NEMB2016-5986

        Naturally occurring and making up a large portion of the world’s biomass,   Wonmuk Hwang, Xiaojing Teng, Texas A&M University, College
        biofilms are present in a number of industrial and medical applications. They   Station, TX, United States
        comprise bacterial communities housed in a scaffold of proteins and poly-
        saccharides, which along with water make a slime-like substance that is dif-  Mechanical properties of DNA is crucial for genome organization and inter-
        ficult to remove. Because of their strong adhesion and persistence, biofilms   action with proteins. A challenge in modeling DNA mechanics arises due
        are often associated with pathogenicity and nuisance, yet possess enviable   to the co-existence of atomistic behaviors and mesoscale behavior of DNA
        adhesive and survival properties. Curli is a specific fiber that grows on the   as a chain molecule. For the latter, order parameters must be identified that
        surface of E. Coli and plays a structural role in its biofilms, where it has been   properly describe elastic deformation of DNA. We apply theory of elastic rod
        shown to be critical for adhesion. Additionally, it has been shown that genet-  to the atomistic molecular dynamics simulation of B-DNA oligos, to identify
        ically engineered biofilms can be created to exploit the curli self-assembly   local principal axes of bending. To achieve this, we perform a total of 0.9 mi-
        process and to synthesize materials with molecularly precise features. Yet,   crosecond all-atom molecular dynamics simulation of DNA oligos possess-
        so far there is no clear link between the sequence, chemistry, and topolo-  ing repeating sequences. For each base pair, we assign a local coordinate
        gy of curli and resulting mechanical properties of their biofilms. Here, we   basis (triad) and follow its trajectory relative to the triad of the neighboring
        investigate the adhesion mechanisms of CsgA, the beta-helical subunit that   dinucleotide step. In this way, major and minor bending axes are identified,
        self-assembles into curli fibers. All-atomistic Molecular Dynamics simulations   where the distributions of bending angles are well-approximated by Gauss-
        of the protein subunit adsorbing on hydrophobic and charged surfaces in   ian. For a given dinucleotide step (there are 10 possible dinucleotide steps
        explicit water solvent are employed to assess adhesion. Using these, we ex-  formed by G, C, A, T), we calculate the four elastic stiffness of the major/mi-
        plore the connection between cooperative motion among residues, change   nor bending, twist, and extension. DNA is the most compliant in major bend-
        in secondary structure, and surface contact. Groups of residues in the struc-  ing, and we find that traditional description based on helicoidal parameters
        ture are seen to experience correlated changes in secondary structure and   do not adequately capture conformational fluctuation around the equilibrium
        flexibility, which can guide mutation location choices to induce changes in   conformation. This is because the 16 helicoidal parameters, although they
        subunit mechanical properties. While loss of secondary structure can result   describe atomic structure of DNA well, they are not based on principal axes
        in reduced stiffness, it often occurs in protein regions in contact with the   of DNA. For any given oligo sequence, we can calculate the corresponding
        surface. We find adhesion to be mediated by different mechanisms: aromatic   persistence length that is consistent with experimental measurement. Based
        residues facilitate adhesion onto graphene, while negatively charged resi-  on our calculated parameters, we build coarse-grained model of DNA that
        dues play a larger role in adhesion to silica. Additionally, rows of mutations   effectively captures its sequence-dependent elastic properties. The coarse-
        are incorporated within the subunit, with various compositions and positions.   grained simulation of DNA oligos further reveals a finite size effect where
        These are investigated in regard to changes in structural dependencies   the persistence length measured by following the end to end distance fluc-
        between residues, adhesion energy, and alignment between aromatic   tuation and applying the wormlike chain model, is shorter than one for an
        mutations. These results can be applied to determine optimal mutation lo-  ideal, infinitely long DNA that is calculated based on their bending stiffness
        cation and type for (1) enhancing adhesion and (2) facilitating pi-stacking for   and sequence composition. This is due to the nonzero intrinsic curvature.
        electron conduction and can be used as a guide for choices in genetically   We also decompose elastic energy in 1381 x-ray structures of protein-DNA
        modifications. These findings lend insight into the mechanisms governing   complexes. In most cases, DNAs are only mildly deformed with the total
        the impressive adhesive properties of curli, which can be applied to devel-  elastic energy less than two times the thermal energy. In this case, twist is
        opment of strong adhesives. Results concerning mutations set the stage for   the dominant mode of deformation. For structures with total elastic energy
        development of methods to enhance adhesion and use curli’s self-assem-  greater than about four times thermal energy, major bending becomes dom-
        bling power to create ordered nanostructures and microbial nanowires.  inant. The present results elucidate the atomistic origin for the elastic be-
                                                                havior of DNA and how it is used for interaction with proteins. Our approach   57
        4:40pm  Computational characterization of materials response   is applicable to DNAs in different conformational states as well as to other
        based on results of nanoindentation tests               filamentous proteins.
   52   53   54   55   56   57   58   59   60   61   62