Enzymes and the Phosphate-Phosphonate-Rearrangement

a-Substituted phosphonic acids are a large group organophosphorus compounds with various biological propeties. Analogues of amino acids block enzymes, which make them interesting structures on the border between organic chemistry and biochemistry. Two approaches are combined to prepare a large set of compounds. The first approach is based on hydrolytic enzymes, which allow the kinetic resolution of a-acetoxyphosphonates. The optically active a-hydroxyphosphonates are transformed into other a-substituted phosphonates and the corresponding phosphonic acids. The second entry, a noval one, is the rearrangement of phosphates and S-alkyl thiophosphates to a-hydroxy- and a-mercaptophosphonates, respectively. ß-N,N-Dibenzylaminoalcohols derived from a-amino acids will be phosphorylated and isomerised to ß-amino- a-hydroxyphosphonates. The isomerisation of S-alkyl thiophosphates, an unknown reaction, will be investigated thoroughly. The reaction conditions, the configurational stability of the intermediate thiocarbanions, and the stereochemistry (retention, inversion) of the rearrangement will have to be addressed. Various bases will be evaluated for the enantioselective synthesis of a-substituted phosphonates. Some compounds will serve special purposes such as the synthesis of phosphonic acid analogues of aspartic and glutamic acid and their respective amides. An a-hydroxyphosphonate being also chiral at phosphorus allows the determination of the stereochemistry of the phosphonate-phosphate-rearrangement at phosphorus. Enantiomerically pure a-hydroxy- ß-oxopropylphosphonic acids will assist the isolation of the P-C bond cleaving enzyme involved in the biodegradation of the clinically used antibiotic fosfomycin. a-Aminophosphonic acids derived form t- butylglycine will be tested as auxiliaries in chiral HPLC. At last, sulfur containing a-aminophosphonic acids will be prepared and evaluated as inhibitors (potential herbicides) of the phenylalanine ammonia-lyase. Part of the applications will be studied in cooperation with Prof. Lindner - Österreich, Prof. Amrhein - ETH Zürich, and Dr. Quinn - Großbritannien.

The aim of the project was the synthesis of chiral phosphonates and phosphonic acids containing a carbon atom carrying a phosphorus and another hetero-atom such as oxygen, nitrogen or sulfur as substituents. The starting materials for these compounds were racemic 1- and some 2-hydroxyphosphonates which were resolved as their chloroacetates using hydrolytic enzymes such as lipases and proteases in a biphasic system. The lipases hydrolysed the (S)-esters preferentially, the proteases the (R)-esters. The (S)-1-hydroxyphosphonates of high enantiomeric excess were converted to 1-azidophosphonates using the Mitsunobu reaction and to 1-thiocyanatophosphonates via the 4-nitrobenzenesulfonates, in both cases with inversion of configuration. The azides were reduced and deblocked to give 1-aminophosphonic acids. We have prepared quite a few, some of which are phosphonic acid analogues of proteinogenic amino acids. Analogues of phenylalanine are potential inhibitors of the phenylalanine ammonia lyase and were tested in cooperation with Prof. Amrhein and Prof. Retey for their herbicidal activity. The 1- thiocyanatophosphonates were reduced to 1-mercaptophosphonates under very mild conditions to minimize racemisation. The deblocking of the phosphonate group was so far not possible, although the corresponding 1- mercaptophosphonic acids could be metal chelating compound inhibiting certain enzymes. The chiral 1-hydroxy-3-butenylphosphonate obtained by a sigmatropic rearrangement was found to be a very versatile starting material for a variety of aminophosphonic acids. Different approaches to 1-mercapto- and 1-aminophosphonates were the S-alkyl thiophosphate-1- mercaptophosphonate and phosphoramidate-1-aminophosphonate rearrangement. The former was discovered here and investigated thoroughly. Its scope and stereochemistry were addressed. The chiral, deuteriated S-hexyl thiophosphate was rearranged to show that the metalation has a high primary kinetic isotope effect and the migration of phosphorus from sulfur to carbon atom is stereospecific. The phosphoramidate-1-aminophosphonate rearrangement was used to prepare chiral, tertiary 1- aminobenzylphosphonic acids from chiral secondary benzylamines for the first time. These rearrangements are induced by strong bases and proceed via configurationally stable, short-lived 1-thio- and 1-aminoalkyllthiums, which isomerise with retention of configuration. The full potential of these rearrangements, especially the enantioselective versions thereof, will be addressed in a follow-up project.