A high performance system for the molecular dynamics simulation of biological molecules was constructed by combining a software package, Peach, with a special-purpose computer, Grape. The resultant simulator "Peach-Grape system" was used to analyze several important biological molecules including the Hin/DNA complex, the trp-Repressor/DNA complex, and Calmodulin. In addition to those simulations performed by the Peach-Grape system, other simulation studies of biomolecules by special-purpose computers are briefly reviewed.

View full abstract

We speculate that structures of genes might have been produced not only by the origination of the gene by itself, but also by the combination of several parts of structures from other species, through events, such as gene transfer, fusion of DNA sequences, and rearrangement of DNA sequences, to create new functions in evolution. To study the relationship between the 3D structure and DNA sequence of proteins comprised of multi-domains, six genes in Escherichia coli, including those coding for Flavodoxin (ferredoxin) reductase, Flavin oxidoreductase, Integrase/recombinase xerD, Endonuclease III, Heat shock protein (grpE), and Elongation Factor Tu (tufB), were chosen, since the 3D structures of these proteins had been determined by X-ray crystallography. The DNA sequences were analyzed by probability as calculated by using a species-specific inhomogeneous Markov model and divided into regions. The probability was calculated by the GeneMark program from the third-order matrix of Class III genes (Escherichia coli horizontally transferred genes) produced with the Markov model by Borodovsky M. et al.. The probability indicates the degree of similarity to the DNA sequence of the genes that were used for producing the matrix with the Markov model. Moreover, the probability indicates whether the origin of the divided DNA sequence would be different from that of the Class III genes. The divided regions, except for those of Flavin oxidoreductase, corresponded to the structural domains classified by CATH. Namely, the divided regions, suggested to be of different origins by analysis with the Markov model, showed correspondence to the domains and the gross content of the 3D structure. We can ascertain which DNA sequences would have been incorporated into the structure of a gene as parts of the structure in evolution. We propose a hypothesis that in evolution, the structure of genes might have been arisen by the incorporation of DNA sequences from other species. Key Words: DNA sequence analysis, Markov model, evolution, structural evolution Area of Interest: Bioinformatics and Bio Computing

View full abstract

The three-dimensional structure of Flavodoxin reductase in Escherichia coli had been determined by X-ray crystallography and was found to consist of multi-domains. Then the DNA sequence of Flavodoxin reductase was analyzed by using Markov models, and the sequence was divided into seven regions (I-VII), which corresponded to the structural domains classified based on "Class, " derived from the secondary structure, "Architecture, " which was derived from the gross orientation of the secondary structures and "Topology, " as defined by CATH. In addition, the domains defined by CATH are equivalent to the domains identified by analysis of the X-ray crystallography. A FASTA homology search was done against the DDBJ ALL database with each divided region. Region I+II belongs to the FAD domain, which binds FAD, and its DNA sequence showed homology to that of the FAD binding site of NADPH:ferredoxin reductase of Azotobacter vinelanndii. It is reported that the overall 3D structure was very similar to that of Flavodoxin reductase, since the RMSD between these two proteins was 1.49 Å and the amino acid sequences at the N-termini also showed high homology. Thus there is a correspondence between the 3D structures, amino acid sequences, and the divided sequences of both sequences. The region from IV to VII belongs to the NADP domain, which binds to NADP/NADPH. The DNA sequences of regions IV, V+VI and VII showed respective homology to conjugal transfer gene E (traE) of Escherichia coli pKM101, the czcB gene (cation-proton antipoter) from Ralstonia sp. CH34 pMOL30, and orf17 of Streptococcus pneumoniae Bacteriophage Cp-1, which was predicted as a tail protein. Thus it is suggested that the NADP domain would have been constructed by the fusion of a tail protein from a bacteriophage, the czcB gene and the conjugal transfer gene from a plasmid. Moreover, NADPH:ferredoxin reductase of Azotobacter vinelanndii which showed homology to the DNA sequence in the region of the FAD domain, was coded on the chromosome. On the contrary, genes homologous to the DNA sequence of the region of the NADP domain were coded on a plasmid or bacteriophage. The structurally divided regions of Flavodoxin reductase were supposed to be comprised of several fragments of DNA sequence whose imprinted structural information might be incorporated as parts of structure from other species.

View full abstract

Ras protein is a kind of small G protein that functions as a molecular switch or a timer alternating between the inactive GDP-bound and active GTP-bound form. Ras requires Mg²⁺ as a cofactor for its full activity. Recently, in another small G protein family, the release of Mg²⁺ coordinated by GDP was found to play an important role in binding the guanine nucleotide exchanging factor (GEF) that promotes the GDP/GTP exchange reaction. Here we calculated Mg²⁺-bound and Mg²⁺-unbound conformations of the GDP-bound Ras by molecular dynamics (MD) simulation. In the Mg²⁺-unbound conformation, significant conformational changes in the switch 1 region were observed, in which the switch region opened to expose the nucleotide-binding site. The distance between the switch 1 and switch 2 regions increased, resulting in the appearance of a groove. These conformations of the switch regions were very similar to that of Ras bound to its GEF: SOS (Son Of Sevenless). These structural changes were not observed in the Mg²⁺+-bound conformation. These results demonstrate the regulatory role of Mg²⁺ in GEF binding and suggest that the GDP dissociation occurs by a stepwise mechanism; (1) Releasing of Mg²⁺; (2) Conformational changes of the switch regions (semi-open form); (3) GEF binding (complete-open form); and (4) GDP dissociation. We suggest that the concentration of Mg²⁺ ions may regulate the binding between small G-proteins and GEFs.

View full abstract

Wrong:[abstract in Japanese]
Right:[abstract in Japanese]

View full abstract