Toward a general strategy for the synthesis of 3,4-dihydroxyprolines from pentose sugars

A general strategy is proposed, wherein a pentose sugar gamma-lactone can be converted, via a series of nine reactions, to a 3,4-dihydroxyproline, suitably protected for use in peptide synthesis. Thus, D-ribonolactone (6) has been converted to N-fluorenylmethoxycarbonyl-3,4-di-O-tert-butyldimethylsilyloxy-D-2,3-cis-3,4-cis-proline (7) in 18.9% overall yield. Likewise, L-arabinonolactone (11) has been converted to N-fluorenylmethoxycarbonyl-3,4-di-O-tert-butyldimethylsilyloxy-L-2,3-cis-3,4-trans-proline (36) in 13.7% overall yield and L-lyxonolactone (12) to N-fluorenylmethoxycarbonyl-3,4-di-O-tert-butyldimethylsilyloxy-L-2,3-trans-3,4-cis-proline (37) in 11.2% overall yield. These building blocks have also been fully deprotected to give the free amino acids. We believe that this series of reactions ought to be applicable to the synthesis of any of the eight stereoisomers of 3,4-dihydroxyproline, by judicious selection of the pentose starting material.