Enantioselective acylation of a beta-lactam intermediate in the synthesis of loracarbef using penicillin G amidase

Enantioselective acylation of a beta-lactam intermediate in the synthesis of loracarbef using penicillin G amidase

Tehahedrcm Letters. Vo1.32, No.13. pp 1621-1622 1991 Priited in Great Britain ENANTIOSELECTIVE oo4o439i91 $3.00 + .oo Pergamlm Press plc ACYLATION ...

140KB Sizes 2 Downloads 124 Views

Tehahedrcm Letters. Vo1.32, No.13. pp 1621-1622 1991 Priited in Great Britain


oo4o439i91 $3.00 + .oo Pergamlm Press plc




Milton J. Zmijewski, Jr.‘, Barbara S. Briggs, Allen R. Thompson, and Ian G. Wright Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Indiana, 462854813, U.S.A.

Summary. Penicillin G amidase from E. rpli has been isomer of a cis, racemic azetkfinone intermediate used acyiation occurs using methyl phenylacetate (MPA) and enzyme displays similar enantfoselectivity with MPOA or

shown to selectively acylate, in an efficient manner, the (2R,3S) in a synthesis of ioracarbef, a carbacephalosporin antibiotic. The using methyl phenoxyacetate (MPOA) as the acylating agentsThe MPA.

Loracarbef 1, a member of the new carbacephalosporin

class of beta-lactam antibiotics, is a stable analog of the clinically

useful antibiotic cefaclor ([email protected]&

Like cefaclor, the synthesis of loracarbef presents interesting challenges. A key

intermediate in one of the syntheses to loracarbef is 9, a cis racemic [(2R,3S), (2S,3R)] azetfdinone. The (2R,3S) isomer of this compound can be elaborated in subsequent chemistry to the antibiotic.

A chemical resolution of this compound was

cumbersome and relatively low yielding. Penicillin G (PG) amidase is enantioselective

in its treatment of beta-lactams,

amides, and esters (l-6). We hoped that this selectivity could be put to use in the resolution of 8. Instead of using the enzyme in the normally used hydrolytic direction, we opted to acylate the substrate. In doing so, we envisioned that the precipitation of the water insoluble compound formed would be the driving force for the acylation and would protect the product from hydrolysis by the enzyme.




60s 5 R = Me; X = COCH,Ph 4 R = Me; X = COCH,OPh ZR=H ;X=COCH,Ph &R=H ;X=COCH,OPh Preliminary experiments suggested that pH 6.0 was optimal for carrying out the synthesis of susing

methyl phenylacetate

(MPA), PG amidase, and the oxalate salt of 9 at 28°C (7). Precipitation of the product did provide the driving force for the



reaction to




in the synthetic direction (yield = 41%, ee = 96%) (6). subsequent



in the


Since the phenylacetyl

we decided

to try

a side

(G) side chain was not





was found to be an excellent substrate in acylating 3 (Yield = 44%) (6) and purified fi had an ee =

97%. This side chain was compatible with later chemistry.

The formation of a precipitate in this reaction presented operational problems with the immobilized enzyme; therefore, another approach was


Compound 4 was prepared by base mediated hydrolysis of 9 and was used as a substrate

for the reaction. Product & from the enzymatic acylation of 4 using MPOA is water soluble; so, the potential driving force for the reaction could be lost due to subsequent hydrolysis of 8 by the enzyme. This was not the case since yields of compound B remained high and little hydrolysis was evident (final yield after 24 hours = 45%, ee = 100 %). Control kinetic studies in the formation of z demonstrated that with time, the enzyme did hydrolyze the Z formed (final yield after 24 hours = 31%. ee = 100%).

Studies with purified Z and & at pH 6.0 and the enzyme showed that & was hydrolyzed at only

6% of the rate of z. This could explain why yields were higher when MPOA was used as a substrate for the acylation of 4 than when MPA was used during the extended time course studies.

Employing this information, we have developed a process for the resolution of 4 using PG amidase and MPOA (9) that is compatible with the downstream chemical scheme and high yielding. results show that an ester of phenoxyacetate not appreciated

by recent investigators

Besides the applied aspect of our work,


is indeed a good substrate for the enzyme in the synthetic direction, a fact


Our results have also called into question the classification of this

penicillin amidase as “G” specific. These observations have encouraged us to explore further the side chain specificity of the enzyme in this acylation.

References 1) Rossi, D.; Calcagni, A.; Romeo A. J. Org. Chem,l 97 9, 44, 2222. 2) Rossi, D.; Calcagni, A. Expertentia, 19 8 5,41, 35. 3) Fuganti, C.; Grasselli, P. Tetrahedron Letters, 1986,

27, 3191.

4) Fuganti, C.; Grasselli, P.; Servi,S.; Lazzarini. A.; Casati, P. Tetrahedron, 1 9 88,44,


5) Waldman, l-f. Tetrahedron Letters, 1 9 6 9, 30, 3057. 6) Ogasa, T.; Sake, H.; Hashimoto, Y.; Sato, K.; Hirata, T. Chem. Pharm. Bulletin, 1989,

37, 315.

7) Acylation: The substrate to be examined (3 or 4) was dissolved in distilled water at a concentration of 10% and the pH adjusted to 6.0. The side chain ester (MPA or MPOA, 1.5 equivalents)

and enzyme (Sclavo penicillin G amidase

immobilized on Eupergit, 84 IU/gm amine) was added to start the acylation. The pH was maintained at 6.0 by the addition of 2.5 N ammonium hydroxide. Reactions were terminated when base uptake stopped or after 24 hours. 8) Yields were determined by HPLC of reaction mixtures or methylene chloride extracts. The enantiomeric (Chiralcel OJ column) was determined on purified compounds. Yields after purification by semi-preparative


HPLC for a

typical 4 hour reaction were 26.8%(5), 33.1%(6), 35.7%(7), 33.9%(8). All physical chemical data was consistent with structures shown. 9) Levy, J. et al., Lilly Research Laboratories, personal communication. 10) Kaufmann, W.; Bauer, K.; Offe, H.A. Antimicrobial Annual, 19 6 1 ,I. 11) Bondareva, N.S.; Levitov, M.M.; Rabinovich, M.S. Biokhimiya, 1969,

(Received in USA 3 December 1990)

34, 378.