A structural foundation of utilizing Bacillus cereus phosphopentomutase in the synthesis of didanosine.

摘要

In this document I investigate a phosphopentomutase (PPM) from Bacillus cereus as a model system and provide the first structural and biochemical characterization of a prokaryotic PPM. PPMs are members of the alkaline phosphatase superfamily and interconvert α-D-ribose-1-phosphate (ribose-1-phosphate) and D-ribose-5-phosphate (ribose-5-phosphate). PPM can be used in the chemoenzymatic synthesis of didanosine, an inhibitor of HIV reverse transcriptase, by converting 2,3-dideoxyribose-5-phosphate into 2,3-dideoxyribose-1-phosphate. However, the wild type enzyme does not catalyze the reaction with 2,3-dideoxyribose-5-phosphate efficiently. To guide protein engineering that will increase turnover of 2,3-dideoxyribose-5-phosphate by PPM, I determined its structure and biochemical properties. PPM contains two domains with a di-manganese catalytic center located at the domain interface and the catalytic nucleophile, Thr-85, located adjacent to the metal ions. Comparison of the structure of PPM to structures of other members of the alkaline phosphatase superfamily demonstrated that one domain has homology to the alkaline phosphatase core domain, while the second, cap domain, is unique. Prior to my studies it was known that members of the alkaline phosphatase superfamily become covalently phosphorylated during the turnover cycle, however, here I demonstrate that PPMs require phosphorylation by a bisphosphate molecule before turnover of ribose-5-phosphate can occur. Furthermore, I show that in contrast to other studied phosphomutases in the alkaline phosphatase superfamily, phosphoryl transfer by PPM is intermolecular and proceeds through the formation of a ribose-1,5-bisphosphate intermediate. PPMs can disseminate between mono and bisphosphorylated ligands based on the position of a single lysine residue, Lys-240, that serves to regulate affinity for bisphosphate. Additionally, the relative orientation of the core and the cap domain may change during the turnover cycle and contribute to the reorientation of the bisphosphate intermediate. Lastly, I compare 2,3-dideoxyribose-5-phosphate binding, the desired substrate in the didanosine synthetic scheme, to ribose-5-phosphate binding and show that both bind to the same site. When taken together, these studies provide a solid basic science foundation upon which we can engineer a PPM useful for the synthesis of didanosine.