Yu Komatsu

Designing the Binding Surface of Proteins to Construct Nano-fibers

In the field of nanotechnology, a variety of applications have been anticipated. We have been trying to design nano-fibers using proteins while maintaining their native structures. We try to use Lac repressor two-helix protein (LARFH), sulerythrin, and 3-isopropylmalate dehydrogenase (IPMDH) as adaptors for constructing nano-fibers. By making use of the α-helices outside of respective proteins, we are trying to form binding site between proteins: two α-helices of one protein are designed to form four-helix bundle with two α-helices of another protein. In addition, by introducing mutations in amino acids at the binding sites, hydrophobic and electrostatic interactions can be modified. The fiber may be produced upon mixing the two kinds of proteins. By umbrella sampling simulation, we have found that in the combination of LARFH-/-LARFH, hydrophobic interaction is enhanced in wild type, and electrostatic interaction is enhanced in variant. We also found high stability of IPMDH-/-IPMDH.

Evaluation of the protein interfaces that form an intermolecular four-helix bundle as studied by computer simulation

Rational design of protein surface is important for creating higher order protein structures, but it is still challenging. In this study, we designed in silico the several binding interfaces on protein surfaces that allow a de novo protein–protein interaction to be formed. We used a computer simulation technique to find appropriate amino acid arrangements for the binding interface. The protein–protein interaction can be made by forming an intermolecular four-helix bundle structure, which is often found in naturally occurring protein subunit interfaces. As a model protein, we used a helical protein, YciF. Molecular dynamics simulation showed that a new protein–protein interaction is formed depending on the number of hydrophobic and charged amino acid residues present in the binding surfaces. However, too many hydrophobic amino acid residues present in the interface negatively affected on the binding. Finally, we found an appropriate arrangement of hydrophobic and charged amino acid residues that induces a protein–protein interaction through an intermolecular four-helix bundle formation.

Light absorption and excitation energy transfer calculations in primitive photosynthetic bacteria

In photosynthetic organisms, light energy is converted into chemical energy through the light absorption and excitation energy transfer (EET) processes. These processes start in light-harvesting complexes, which contain special photosynthetic pigments. The exploration of unique mechanisms in light-harvesting complexes is directly related to studies, such as artificial photosynthesis or biosignatures in astrobiology. We examined, through ab initio calculations, the light absorption and EET processes using cluster models of light-harvesting complexes in purple bacteria (LH2). We evaluated absorption spectra and energy transfer rates using the LH2 monomer and dimer models to reproduce experimental results. After the calibration tests, a LH2 aggregation model, composed of 7 or 19 LH2s aligned in triangle lattice, was examined. We found that the light absorption is red shifted and the energy transfer becomes faster as the system size increases. We also found that EET is accelerated by exchanging the central pigments to lower energy excited pigments. As an astrobiological application, we calculated light absorptions efficiencies of the LH2 in different photoenvironments.

The simulation study of protein-protein interfaces based on the 4-helix bundle structure

Docking of two protein molecules is induced by intermolecular interactions. Our purposes in this study are: designing binding interfaces on the two proteins, which specifically interact to each other; and inducing intermolecular interactions between the two proteins by mixing them. A 4-helix bundle structure was chosen as a scaffold on which binding interfaces were created. Based on this scaffold, we designed binding interfaces involving charged and nonpolar amino acid residues. We performed molecular dynamics (MD) simulation to identify suitable amino acid residues for the interfaces. We chose YciF protein as the scaffold for the protein-protein docking simulation. We observed the structure of two YciF protein molecules (I and II), and we calculated the distance between centroids (center of gravity) of the interfaces’ surface planes of the molecules I and II. We found that the docking of the two protein molecules can be controlled by the number of hydrophobic and charged amino acid residues involved in the interfaces. Existence of six hydrophobic and five charged amino acid residues within an interface were most suitable for the protein-protein docking.

Coarse-Grained Molecular Dynamics Simulation of IPMDH Proteins

protein to study the effect of mutation to the binding sites of protein. We replace original amino acids on the surface by negatively or positively charged amino acids. We also study the effect of solvent to the binding of proteins. Our computational study may be useful for future experimental study of protein bindings. Making nano-fibers with proteins (1-3) is an interesting and challenging field. We have tried to construct nano-fibers while retaining native structure. We started with proteins which form 4-helix bundle structure (4) by two proteins, each having two helices. In our previous studies (5-6) of molecular dynamics simulation of all-atom model, we have found that IPMDH (3-isopropylmalate dehydrogenase) (7-9) is a good candidate for a building unit of constructing nano-fibers because of the high stability of the mutated IPMDH. In this study, we have performed molecular dynamics simulation of coarse-grained model of IPMDH proteins. We have introduced several mutations on the surface of IPMDH and found that the mutated IPMDHs are likely to make larger clusters than the wild-type IPMDH. 2. Simulation Model and Method We used coarse grained model of IPMDH with MARTINI force field (10), where one interaction point represents on average four heavy atoms. There are four kinds of interaction points in MARTINI force field: polar, nonpolar, apolar and charged. Using coarse grained model, we can perform simulation of long-time behavior of large scale system. To check the model, we calculated radial distribution function for both all-atom model and coarse grained model, and obtained similar results.

Analysis of the histone protein tail and DNA in nucleosome using molecular dynamics simulation

We study the effect of the tails of H3 and H4 histones in the nucleosomes, where DNA and histones are packed in the form of chromatin. We perform molecular dynamics simulations of the complex of DNA and histones and calculate the mean square displacement and the gyration radius of the complex of DNA and histones for the cases with tails intact and the cases with tails missing. Our results show that the H3 tails are important for the motion of the histones. We also find that the motion of one tail is affected by other tails, although the tails are distanced apart, suggesting the correlated motion in biological systems.