Synthesis of intermediates towards oseltamivir and theoretical prediction of their protein binding

Abstract

Gabaculine derivatives are the important intermediate from many reported total synthesis of oseltamivir phosphate or Tamiflu (R), the well known influenza neuraminidase inhibitor. The attempted synthesis of gabaculine derivatives were carried out via Birch reduction, condensation or Diels-Alder reaction. Although the products from some tested Birch reductions were successfully obtained, the reaction failed to give the desired gabaculine isomer. The condensation of pyrrolidine and diacetal to create the diene precursor for subsequent Diels-Alder reaction only afforded unidentifiable polymeric products. The direct cycloaddition on Boc-pyrrole gave the azabicyclic 85, though in rather low yield of 3.40%. Subsequent reduction of this compound had not yet yielded the expected product. The synthesis of gabaculine derivatives has led to the design of 34 oseltamivir-like structures for the prediction of the viral protein binding site. The binding prediction of these potential neuraminidase inhibitors revealed that there were only two energetically favorable analogs. From analysis of relative binding energy and ligand position with respect to the reference has suggested that the suitable stereoisomers of the inhibitors in the binding sites should have (S)-3-pentyl ether group configuration at C3 position and either (R)-NH2 or (R)-NH3+ configurations at C4 and C5 positions respectively on the 1-carboxy-cyclohexadienyl core structure of the neuraminidase inhibitors.