Molecular Dynamics Simulations of a Catalytic Multivalent Peptide–Nanoparticle Complex

Molecular modeling of a supramolecular catalytic system is conducted resulting from the assembling between a small peptide and the surface of cationic self-assembled monolayers on gold nanoparticles, through a multiscale iterative approach including atomistic force field development, flexible docking with Brownian Dynamics and µs-long Molecular Dynamics simulations. Self-assembly is a prerequisite for the catalysis, since the catalytic peptides do not display any activity in the absence of the gold nanocluster. Atomistic simulations reveal details of the association dynamics as regulated by defined conformational changes of the peptide due to peptide length and sequence. Our results show the importance of a rational design of the peptide to enhance the catalytic activity of peptide–nanoparticle conjugates and present a viable computational approach toward the design of enzyme mimics having a complex structure–function relationship, for technological and nanomedical applications.

molecule using a grid spacing of 1 Å . If during a simulation step, one molecule penetrates the exclusion volume of the other molecule, then the move is regenerated with a different random number until it omits the overlapping.
The empirical scaling factor for weighing between long range electrostatic interaction energy and short range electrostatic desolvation term is set as 1.67 for our calculation. For non-polar desolvation, the potential depends on a scaling factor (γ) such that γ = 1.0 for all the points of the second solute, lying within a minimum distance (a =3.10 Å ) from the surface of the first one, at this limit the surface of the first solute is completely occluded. Similarly, γ = 0.0 if the point is further than a maximum distance (b = 4.35 Å ) from the surface of the first solute, where the presence of the second solute does not affect solvation of the first one. Finally, the factor is linearly interpolated if the distance lies in between these two limits. We use γ = 0.5 in our calculation as per observations obtained from trial and error method, to get better estimation of the potential. The surface of solute 1 that is excluded to solute 2 is computed by increasing radii of the atoms by 1.77 Å for proteins of solute 1, which have solvent accessible surface area (SASA) more than the threshold. A probe of radius 1.4 Å is used as a representative of solvent (water) molecule to calculate the SASA of the solute. Spacing of the exclusion grid is considered ~ 0.5 Å to take into account the shape of the solute. The entire potential grid is multiplied by suitable factors to include the ionic contribution in potential grid.

=1
, after aligning all the simulated structures of the molecule with respect to its initial conformation using script in VMD 3 . Here ( ) is the position of th atom at time , and ( ) is that for initial configuration. N is total number of atoms of the polymer.
Here, we address radius of gyration ( ) through standard gmx gyrate command of GROMACS 4,5 for this analysis. The calculation is manifested as per the formula, , is mass of th atom and ( ) is coordinate of center of mass at time and We use standard g_sasa command of GROMACS 4, 5 to generate SASA of system of interest ( ) and the corresponding histogram ( ( )) over simulated data. A solvent representative sphere of radius is rolled over the envelop of the van der Waals surface of biotin, where this surface is composed of interlocking spheres of appropriate van der Waals radius consistent to the atom. Thus, SASA of an atom of radius is defined as area of sphere with radius = + , given the fact the solvent sphere is in contact with the respective atom without penetrating any other atoms.
We also use radial distribution function, the probability distribution of finding a neighbour atom around a central atom within a spherical shell of radii and + ∆ through standard gmx rdf command.
For cluster analysis, we use standard gmx cluster command with gromos algorithm of GROMACS 4, 5 .
This counts the number of neighbours using a threshold value of RMS ~ 0.1 nm of the structures with respect to a reference structure. The structure with largest number of neighbours with all its neighbours is considered as one cluster and after eliminating this particular cluster, the process is repeated for other structures of the trajectories. The structure with the smallest average distance to the others or the average structure or all structures for each cluster will be written to a trajectory file.
Mean squared displacement (MSD) of the center of mass of peptide over simulated trajectory is computed using g_msd command and by the algorithm as 6 Figure S3.