The Application Of Molecular Modeling On Bioorganic Phosphorus Chemistry

Molecular modeling can be used not only for constructing models of already known molecules, but also investigating unknown or not yet synthesized molecules. "Running molecular modeling program" is just like "doing chemistry experiment in computer". The modern techniques combined with traditional organic chemical methods have been becoming powerful tools in the study of the relationships between molecular structure and biological activity and thus essential in chemistry, biochemistry and rational drug design. In this thesis molecular modeling techniques were used to investigate the molecular interaction mechanism for a series of scientific problems in the field of bioorganic phosphorus chemistry. It will help us to get better understand on some experimental phenomenon arose from the early experimental studies of our Laboratory. N-phosphoryl amino acid, as a model compound linking bioorganic phosphorus chemistry and life science, has been proved a successful starting point for investigating life science problems from the point of view of a chemist. A number of results have been achieved by using this model in the studying of origin of life, MS techniques in structural proteomics and synthesis of amino acid-nucleoside complexes in our Laboratory. The most exciting achievement resulting from N-phosphoryl amino acid study is the discovery of that Ser-His dipeptide can cleave DNA. In this thesis molecular modeling studies were performed on DNA model molecule in complex with Ser-His dipeptide or other modified dipeptides exhibiting different cleaving activities in order to investigate the key structure features for DNA cleavage activity. Novel nonpeptide L/D serine/histamine amides designed had shown high DNA cleavage activities.Our previous studies on N-phosphoryl amino acid have also proved that many biomimic reactions are preceded through a pentacoordinate phosphorus intermediate. Based on the understanding of the pentacoordinate phosphorus structure, with the help of data mining techniques, molecular dynamics simulations were performed on modeled kinase-substrate complexes in an attempt to establish relationship between structure features and binding ability of the complexes. We found that the monitored distance between substrate nucleophilic oxygen (OG) and attacked phosphorus (PG) of ATP analog correlated closely with the binding affinity. It is reasonable to consider the OG-PG distance as a potential criterion to prefigure kinase-substrate binding specificity and the simple computational techniques may work as an easy approach to distinguish good substrates from week or poor substrates.Molecular modeling combined with traditional chemical experimental methods plays important role in the study of bioorganic phosphorus chemistry. Of course, theoretical modeling can never replace experimental practice and it is also restricted by the reliability of theoretical model and computation ability. However, molecular modeling has become an indispensable part of modern life science and the methods have been implemented in many fields including chemistry, biology and even pharmacy companies. The hierarchy of computational modeling methods combined with traditional experimental methods is becoming a powerful tool by organic chemists for searching for, rationalizing and predicting structure and reactivity of organic, bio-organic molecules and the interacting mechanism among the molecules.