A Novel Epimerization System in Fungal Secondary Metabolism Involved in the Conversion of Isopenicillin N into Penicillin N inAcremonium chrysogenum

Abstract

The epimerization step that converts isopenicillin N into penicillin N during cephalosporin biosynthesis has remained uncharacterized despite its industrial relevance. A transcriptional analysis of a 9-kb region located downstream of thepcbC gene revealed the presence of two transcripts that correspond to the genes named cefD1 and cefD2 encoding proteins with high similarity to long chain acyl-CoA synthetases and acyl-CoA racemases from Mus musculus,Homo sapiens, and Rattus norvegicus. Both genes are expressed in opposite orientations from a bidirectional promoter region. Targeted inactivation of cefD1 andcefD2 was achieved by the two-marker gene replacement procedure. Disrupted strains lacked isopenicillin N epimerase activity, were blocked in cephalosporin C production, and accumulated isopenicillin N. Complementation in trans of the disrupted nonproducer mutant with both genes restored epimerase activity and cephalosporin biosynthesis. However, when cefD1 orcefD2 were introduced separately into the double-disrupted mutant, no epimerase activity was detected, indicating that the concerted action of both proteins encoded by cefD1 andcefD2 is required for epimerization of isopenicillin N into penicillin N. This epimerization system occurs in eukaryotic cells and is entirely different from the known epimerization systems involved in the biosynthesis of bacterial β-lactam antibiotics. The epimerization step that converts isopenicillin N into penicillin N during cephalosporin biosynthesis has remained uncharacterized despite its industrial relevance. A transcriptional analysis of a 9-kb region located downstream of thepcbC gene revealed the presence of two transcripts that correspond to the genes named cefD1 and cefD2 encoding proteins with high similarity to long chain acyl-CoA synthetases and acyl-CoA racemases from Mus musculus,Homo sapiens, and Rattus norvegicus. Both genes are expressed in opposite orientations from a bidirectional promoter region. Targeted inactivation of cefD1 andcefD2 was achieved by the two-marker gene replacement procedure. Disrupted strains lacked isopenicillin N epimerase activity, were blocked in cephalosporin C production, and accumulated isopenicillin N. Complementation in trans of the disrupted nonproducer mutant with both genes restored epimerase activity and cephalosporin biosynthesis. However, when cefD1 orcefD2 were introduced separately into the double-disrupted mutant, no epimerase activity was detected, indicating that the concerted action of both proteins encoded by cefD1 andcefD2 is required for epimerization of isopenicillin N into penicillin N. This epimerization system occurs in eukaryotic cells and is entirely different from the known epimerization systems involved in the biosynthesis of bacterial β-lactam antibiotics. The molecular genetics of cephalosporin biosynthesis is an excellent model for secondary metabolism, since considerable information on the enzymology (reviewed in Refs. 1Demain A.L. Demain A.L. Solomon N.A. Antibiotics Containing the β-Lactam Structure. Springer-Verlag New York Inc., New York1983: 189-228Google Scholar and 2Martı́n J.F. Liras P. Adv. Biochem. Eng. Biotechnol. 1989; 39: 153-187PubMed Google Scholar), molecular genetics, and gene expression mechanisms (3Aharanowitz Y. Cohen G. Martı́n J.F. Annu. Rev. Microbiol. 1992; 46: 461-495Crossref PubMed Scopus (199) Google Scholar, 4Martı́n J.F. Gutiérrez S. Demain A.L. Anke T. Fungal Biotechnology. Chapman and Hall, Weinheim, Germany1997: 91-127Google Scholar, 5Martı́n J.F. J. Bacteriol. 2000; 182: 2355-2362Crossref PubMed Scopus (74) Google Scholar) has accumulated in the last few years. The biosynthesis of cephalosporins by Acremonium chrysogenum and cephamycins by Amycolatopsis(Nocardia) lactamdurans and Streptomyces clavuligerus (reviewed in Refs. 6Liras P. Antonie Leeuwenhoek. 1999; 75: 109-124Crossref PubMed Scopus (57) Google Scholar and 7Martı́n J.F. Appl. Microbiol. Biotechnol. 1998; 50: 1-15Crossref PubMed Scopus (98) Google Scholar) begins with the formation of the tripeptide δ-(l-α-aminoadipyl)-l-cysteinyl-d-valine (ACV) 1The abbreviations used for: ACV, δ-(l-α-aminoadipyl)-l-cysteinyl-d-valine; IPN, isopenicillin N; DAC, deacetylcephalosporin C; HPLC, high pressure liquid chromatograph; GITC, 2,3,4,6-tetra-O-acetyl-β-d-glucopyranosyl isotiocyanate. 1The abbreviations used for: ACV, δ-(l-α-aminoadipyl)-l-cysteinyl-d-valine; IPN, isopenicillin N; DAC, deacetylcephalosporin C; HPLC, high pressure liquid chromatograph; GITC, 2,3,4,6-tetra-O-acetyl-β-d-glucopyranosyl isotiocyanate. by the ACV synthetase (8Martı́n J.F. J. Antibiot. (Tokyo). 2000; 53: 1008-1021Crossref PubMed Scopus (52) Google Scholar), followed by cyclicization of ACV to isopenicillin N (IPN). IPN is later converted into penicillin N by an epimerase activity that has remained uncharacterized so far in A. chrysogenum. After the epimerization step, penicillin N is transformed by a deacetoxycephalosporin C synthase (expandase) into deacetoxycephalosporin C, and finally, deacetoxycephalosporin C is converted into deacetylcephalosporin C (DAC) and cephalosporin C by DAC synthase (hydroxylase) and DAC acetyltransferase, respectively (Fig.1). The genes pcbAB (9Gutiérrez S. Dı́ez B. Álvarez E. Martı́n J.F. J. Bacteriol. 1991; 173: 2354-2365Crossref PubMed Scopus (134) Google Scholar) and pcbC (10Samson S.M. Belagaje R. Blankenship D.T. Chapman J.L. Perry D. Skatrud P.L. van Frank R.M. Abraham E.P. Baldwin J.E. Queener S.W. Ingolia T.D. Nature. 1985; 318: 191-194Crossref PubMed Scopus (167) Google Scholar) encoding ACV synthetase and IPN synthase are linked together in chromosome VII in the so-called “early cephalosporin gene cluster” (11Gutiérrez S. Fierro F. Casqueiro J. Martı́n J.F. Antonie Leeuwenhoek. 1999; 75: 81-94Crossref PubMed Scopus (55) Google Scholar). The genescefEF, encoding the bifunctional expandase-hydroxylase (12Samson S.M. Chapman J.L. Belagaje R. Queener S.W. Ingolia T.D. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 5705-5709Crossref PubMed Scopus (36) Google Scholar), and cefG, encoding the DAC acetyltransferase (13Gutiérrez S. Velasco J. Fernández F.J. Martı́n J.F. J. Bacteriol. 1992; 174: 3056-3064Crossref PubMed Scopus (67) Google Scholar, 14Velasco J. Gutiérrez S. Campoy S. Martı́n J.F. Biochem. J. 1999; 337: 379-385Crossref PubMed Google Scholar), are linked together in the so-called “late cephalosporin cluster” in chromosome I (11Gutiérrez S. Fierro F. Casqueiro J. Martı́n J.F. Antonie Leeuwenhoek. 1999; 75: 81-94Crossref PubMed Scopus (55) Google Scholar). The isopenicillin N epimerization step still remains unclear. Demain and co-workers (15Konomi T. Herchen S. Baldwin J.E. Yoshida M. Hunt N. Demain A.L. Biochem. J. 1979; 184: 427-430Crossref PubMed Scopus (78) Google Scholar, 16Sawada Y. Baldwin J.E. Singh P.D. Solomon N.A. Demain A.L. Antimicrob. Agents Chemother. 1980; 27: 380-387Google Scholar) reported that isopenicillin N was converted into penicillin N by extracts of A. chrysogenum, although the epimerizing enzyme was extremely labile (17Baldwin J.E. Keeping J.W. Singh P.D. Vallejo C.A. Biochem. J. 1981; 194: 649-651Crossref PubMed Scopus (22) Google Scholar, 18Jayatilake G.S. Huddleston J.A. Abraham E.P. Biochem. J. 1981; 194: 645-647Crossref PubMed Scopus (35) Google Scholar), preventing purification of the protein. However, the isopenicillin N epimerase has been purified from S. clavuligerus (19Jensen S.E. Westlake D.W.S. Wolfe S. Can. J. Microbiol. 1983; 29: 1526-1531Crossref PubMed Scopus (33) Google Scholar, 20Usui S. Yu C.A. Biochim. Biophys. Acta. 1989; 999: 78-85Crossref PubMed Scopus (33) Google Scholar) and A. lactamdurans (21Láiz L. Liras P. Castro J.M. Martı́n J.F. J. Gen. Microbiol. 1990; 136: 663-671Crossref Scopus (27) Google Scholar), and the cefD gene encoding this protein was cloned from both microorganisms (22Kovacevick S. Tobin M.B. Miller J.R. J. Bacteriol. 1990; 172: 3952-3958Crossref PubMed Google Scholar, 23Coque J.J.R. Liras P. Martı́n J.F. Mol. Gen. Genet. 1993; 236: 453-458Crossref PubMed Scopus (41) Google Scholar). Repeated attempts to clone the homologous cefD gene of A. chrysogenum using as probe the bacterial cefD gene or oligonucleotides based on bacterial epimerase conserved amino acid regions with the A. chrysogenum preferred codon usage were unsuccessful. 2S. Gutiérrez and J. F. Martı́n, unpublished results. 2S. Gutiérrez and J. F. Martı́n, unpublished results. These results suggested that the fungal IPN epimerization system was different from the bacterial one. In this article, we describe that two genes contain genetic information required for the epimerization step converting isopenicillin N into penicillin N in A. chrysogenum. The corresponding gene products act in a two-protein system described for the first time in antibiotic biosynthesis. The two genes (named cefD1 andcefD2) are completely different from the cefD gene of A. lactamdurans and S. clavuligerus. A. chrysogenum C10 (ATCC 48278), a strain producing high levels of cephalosporin C, was used for the genetic studies. Escherichia coli ESS2231, a β-lactam supersensitive strain, was used for routine cephalosporin bioassays. Escherichia coli DH5α andE. coli WK6 were used for plasmid purification (double-stranded or single-stranded DNA, respectively). All A. chrysogenum strains were first grown in seed medium (24Shen Y.Q. Wolfe S. Demain A.L. Bio/Technology. 1986; 4: 61-64Google Scholar) for 48 h. Ten ml of seed culture was used to inoculate 100 ml of defined production medium (MDFA) (24Shen Y.Q. Wolfe S. Demain A.L. Bio/Technology. 1986; 4: 61-64Google Scholar) in 500-ml triple-baffled flasks. The cultures were incubated at 25 °C in a rotary shaker at 250 rpm. The levels of cephalosporin antibiotics were determined by bioassay using E. coli ESS2231 as the test strain in plates with penicillinase from Bacillus cereus UL1 as described previously (13Gutiérrez S. Velasco J. Fernández F.J. Martı́n J.F. J. Bacteriol. 1992; 174: 3056-3064Crossref PubMed Scopus (67) Google Scholar). Several constructions named pCD4.5, pCD1, pCD2, and pCD1+2 were obtained from plasmid pJL43 (9Gutiérrez S. Dı́ez B. Álvarez E. Martı́n J.F. J. Bacteriol. 1991; 173: 2354-2365Crossref PubMed Scopus (134) Google Scholar). pD4.5 was obtained from pJL43 by inserting a 6.8-kb SalI fragment bearing the cefD1, cefD2, and bidirectional promoter region (see Fig. 5). SalI andBglII restriction sites were introduced by in vitro mutagenesis in the cefD1 and cefD2 genes, respectively. Finally, the hygromycin resistance (hph) cassette, subcloned from pAN7-1 (25Punt P.J. Dingemanse M.A. Kuyvenhoven A. Soede R.D. Pouwles P.H. van den Hondel C.C.A.A. Gene (Amst.). 1990; 93: 101-109Crossref PubMed Scopus (235) Google Scholar), was inserted in the BglII site to inactivate the cefD2 gene. PCD1+2 was obtained by inserting a 5.8-kbBamHI-EcoRV fragment bearing the cefD1 and cefD2 genes into the BamHI-EcoRI site of pC43 (26Gutiérrez S. Marcos A.T. Casqueiro J. Kosalková K. Fernández F.J. Velasco J. Martı́n J.F. Microbiology. 1999; 145: 317-324Crossref PubMed Scopus (41) Google Scholar). pCD1 bearing the cefD1 gene was constructed by partial digestion of pCD1+2 with SacI and religation. pCD2 contains a 3.5-kb BamHI-SalI fragment bearing the cefD2 gene inserted into theBamHI-EcoRI site of pC43. Transformation of A. chrysogenum protoplasts was performed as described previously (27Gutiérrez S. Dı́ez B. Alvarez E. Barredo J.L. Martı́n J.F. Mol. Gen. Genet. 1991; 225: 56-64Crossref PubMed Scopus (54) Google Scholar). DNA fragments were subcloned into pBluescript SK+, and a nested set of fragments was generated with the Erase-a-base procedure (Promega) by digestion with exonuclease III (28Henikoff S. Gene (Amst.). 1984; 28: 351-359Crossref PubMed Scopus (2831) Google Scholar). The fragments were sequenced by the dideoxynucleotide chain termination method (29Sanger F. Nicken S. Coulson A.R. Proc. Natl. Acad. Sci. U. S. A. 1977; 74: 5463-5467Crossref PubMed Scopus (52497) Google Scholar). To elucidate the presence of putative introns in the DNA sequence of thecefD2 gene, the DNA region containing the expected intron splicing sites was amplified by RT-PCR (Invitrogen) using RNA ofA. chrysogenum 48-h cultures as template with the primers 4A (TATCCGGCAAGCTTGGTCGTAGAG) and 4B (CTTGCGTGAGGGGCGGATGC). The amplified region was sequenced to confirm the presence of the intron. The nucleotide sequence is deposited in the GenBank™/EBI Data Bank under the accession number AJ507632. In vitro mutagenesis was performed with the QuikChange site-directed mutagenesis kit (Stratagene) by following the manufacturer's instructions. Oligonucleotides A1 (5′-CCGCTGCCTACCGTCGACGCCAAGC-3′) and A2 (5′-GCTTGGCGTCGACGGTAGGCAGCGG-3′) were used in the formation of aSalI site in the cefD1 gene. Similarly, a BglII site in the cefD2 gene was obtained with the oligonucleotides B1 (5′-GGCGCCATAGATCTGCCAAGAGCATGC-3′) and B2 (5′-GCATGCTCTTGGCAGATCTATGGCGCC-3′). Genomic DNA of A. chrysogenum was isolated as described previously (27Gutiérrez S. Dı́ez B. Alvarez E. Barredo J.L. Martı́n J.F. Mol. Gen. Genet. 1991; 225: 56-64Crossref PubMed Scopus (54) Google Scholar). Three μg of genomic DNA fromA. chrysogenum C10 or its transformants were digested withEcoRI and the fragments were resolved in a 0.7% agarose gel. The gel was blotted onto Hybond-N membranes (Amersham Biosciences) and hybridized with different probes (see “Results”) labeled with [32P]dCTP (30Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). Total A. chrysogenum RNA was isolated with the RNeasy Kit (Qiagen). Total RNA was resolved by agarose-formaldehyde gel electrophoresis and blotted onto Hybond-N membranes (Amersham Biosciences) as described by Sambrook et al. (30Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). Hybridization probes were labeled with [α-32P]dCTP by nick translation and purified by filtration were as described by al. (30Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). The of the was determined by using a C analysis was performed in a with a as described previously S. Velasco J. Marcos A.T. Fernández F.J. Fierro F. Dı́ez B. Barredo J.L. Martı́n J.F. Appl. Microbiol. Biotechnol. PubMed Scopus Google Scholar, K. Marcos A.T. Martı́n J.F. J. Microbiol. Biotechnol. 27: PubMed Scopus Google Scholar). of isopenicillin N and penicillin N was performed as described by et al. N. J. Demain A.L. Queener S.W. J. Antibiot. PubMed Scopus Google Scholar). N and penicillin N were with 2,3,4,6-tetra-O-acetyl-β-d-glucopyranosyl in for at with The was performed in a with a with a of with a of The IPN epimerase was performed as described by et al. Wolfe S. Demain A.L. Appl. Microbiol. Biotechnol. 1986; Scholar). IPN epimerase activity was determined by the in vitro of isopenicillin into penicillin N. The results were by a by the of IPN into cephalosporin C extracts were from A. chrysogenum cultures grown for h. The activity was expressed in of protein. is the activity of in the The activity is as of protein. Total protein in the was by the the genes encoding proteins involved in biosynthesis are in two we that the gene encoding the involved in the of isopenicillin N into penicillin N located in of the two cephalosporin gene we reported that a located in the cephalosporin downstream from encoding a protein a of cephalosporin C production G. Casqueiro J. Gutiérrez S. Martı́n J.F. Mol. Gen. Genet. PubMed Scopus Google Scholar). To for genes located on the of the downstream from the pcbC gene, a of this region was using RNA from of A. chrysogenum grown for 48 in since at this time the expression of genes is known to high in this strain J. Gutiérrez S. Fernández F.J. Marcos A.T. Martı́n J.F. J. Bacteriol. PubMed Scopus Google Scholar). and downstream of the pcbC gene were the presence of to a with the known of the of the pcbC gene (26Gutiérrez S. Marcos A.T. Casqueiro J. Kosalková K. Fernández F.J. Velasco J. Martı́n J.F. Microbiology. 1999; 145: 317-324Crossref PubMed Scopus (41) Google Scholar). and a RNA of and probes and a of indicating the presence in this region of two named and A fragment containing and was cloned in the pBluescript and sequenced on both analysis revealed the presence of two has and is by the presence of This gene was cloned from a previously constructed J. Gutiérrez S. Casqueiro J. Fierro F. Campoy S. Martı́n J.F. Appl. Microbiol. Biotechnol. Scholar), and the sequence the presence of the encoded a protein of amino with a molecular of The encoded protein similarity to long chain acyl-CoA synthetases from amino Rattus and Mus A. T. K. Y. N. T. T. J. PubMed Scopus Google Scholar). The protein has the of the acyl-CoA involved in the an of the of or amino K. M. M.A. Mol. Microbiol. 1992; PubMed Scopus Google Scholar). of and is by the presence of intron. RT-PCR using RNA from 48-h cultures ofA. chrysogenum C10 the presence of the intron. A DNA of the expected the splicing of the was obtained by The nucleotide sequence of the amplified DNA that the intron been at the splicing sites corresponding to from the translation encoded a protein of amino with a molecular of The encoded protein similarity to from amino M. and R. R. G. K. Mol. Scholar, E. Biochem. J. PubMed Scopus Google Scholar). To the of the proteins encoded by and in cephalosporin C inactivation of both genes was performed with the Nature. PubMed Scopus Google Scholar) for inactivation in A. chrysogenum G. Casqueiro J. Gutiérrez S. Martı́n J.F. J. Bacteriol. PubMed Scopus Google Scholar). To inactivation of genes, plasmid pD4.5 was constructed this plasmid an by of the hygromycin resistance and by a introduced in pD4.5 contains the resistance 5). pD4.5 was transformed into A. chrysogenum and transformants were by the resistance to hygromycin B. The transformants were for a indicating that transformants of the 250 the All of transformants were for cephalosporin production by the transformants a in the cephalosporin production, and transformants no cephalosporin production at as with A. chrysogenum that inactivation in the To confirm that inactivation at the the transformants and two with in cephalosporin production were by chrysogenum C10 and with no in cephalosporin production were used as The DNA of strains was digested with SalI and hybridized with a probe to the bidirectional promoter region located and that the A. chrysogenum C10 hybridized with a genomic DNA of in transformants with no cephalosporin production and the 6.8-kb was converted into a of as by a In transformants with a in cephalosporin production and the is converted into a that a at the in the with no in cephalosporin production two of and were an of the pD4.5 To in the of inactivation of the and transformants named and and in Fig. were To proteins encoded by and were involved in the of isopenicillin N into penicillin the epimerase activity was in extracts of disrupted transformants and and the strain A. chrysogenum that the epimerase activity was in the both at and chrysogenum C10 a high of IPN epimerase activity of at h. These results that the proteins encoded by the and are involved in the epimerization step in cephalosporin C biosynthesis. the corresponding genes been and cefD2 to β-lactam gene since the cefD is used to describe the bacterial epimerase gene J.F. Ingolia T.D. Queener S.W. R. Molecular New Scholar). cefD1 and cefD2 the IPN the and isopenicillin N. that no cephalosporins were by bioassay in of the disrupted in the culture high cephalosporin production was in A. chrysogenum C10 or in the by of the culture that the disrupted strains no of cephalosporin C, in A. chrysogenum C10 cephalosporin C is The culture from the disrupted and and the A. chrysogenum C10 were with (see and the were resolved by that chrysogenum C10 culture the corresponding to both isopenicillin N and penicillin N were However, in the disrupted the corresponding to penicillin N is and at the time is an of isopenicillin the indicating that disrupted strains are blocked in isopenicillin N To elucidate both the acyl-CoA synthetase and the encoded by cefD1 and cefD2 or of were involved in the of isopenicillin N into penicillin a of strains or both of genes was this pCD1, pCD2, and pCD1+2 bearing cefD2, or both genes, were constructed These were transformed separately in the mutant, and transformants were by resistance to transformants obtained with the pCD1 were and genomic DNA digested with SacI was hybridized with a fragment of bidirectional promoter region as that transformants with a DNA of in to the to the disrupted gene in the indicating that the of the plasmid pCD1 was Similarly, transformants obtained with the pCD2 plasmid were and DNA was digested with a and and hybridized with the probe of the promoter as a of in indicating that the of plasmid pCD2 is The to the in the disrupted clone the A. chrysogenum a Finally, the DNA of transformants obtained with the pCD1+2 plasmid was digested with a of and and hybridized with the probe as that the probe hybridized with a of indicating that the of the plasmid pCD1+2 was in an The to the in clone To of the two were for the epimerization step, the IPN epimerase activity was in transformants and that in A. transformants cefD1 andcefD2 are was the epimerase activity In transformants and of the genes was no epimerase activity was These results that both cefD1 and cefD2 proteins are involved in the epimerization of isopenicillin N into penicillin N epimerase activity in transformants and from by with cefD1 and cefD2 genes, of chrysogenum chrysogenum chrysogenum strains were grown in medium and were at h. Three different from strain were with results. in a The strains were grown in medium and were at h. Three different from strain were with results. C with of the and transformants were that in transformants the cephalosporin production was restored to levels to of A. chrysogenum In a of cephalosporins production was of cephalosporin in A. chrysogenum in no cephalosporin was at analysis that transformants cephalosporin C as in A. chrysogenum are of J. Biochem. 236: PubMed Scopus Google Scholar, J.F. Gutiérrez S. J.F. J. of Microbiology. Inc., Scholar) and is known the of amino acid epimerization that to the In epimerization is performed by an epimerization into the synthetases (8Martı́n J.F. J. Antibiot. (Tokyo). 2000; 53: 1008-1021Crossref PubMed Scopus (52) Google Scholar, T. M.A. Microbiol. PubMed Google Scholar), in the are by the of and J. 2000; PubMed Scopus Google Scholar). In this we that the epimerization step converting isopenicillin N into penicillin N in A. chrysogenum is by a two-protein system described in the biosynthesis of antibiotics or secondary These proteins are encoded by two genes expressed in opposite from a bidirectional promoter cefD1, high similarity to long chain acyl-CoA synthetases K. M. M.A. Mol. Microbiol. 1992; PubMed Scopus Google Scholar), high similarity to acyl-CoA racemases R. E. J. Biochem. PubMed Scopus Google Scholar, R. E. J. Biochem. PubMed Scopus Google Scholar). A of action of the A. chrysogenum IPN epimerase system on the of of the and proteins with known epimerization systems that of the as been reported to involved in the of acid and in the of in acid occurs as a of the and the and is converted into acid PubMed Scopus Google Scholar, R. E. J. 1993; PubMed Scopus Google Scholar). The and for of acid act on K. S. P. PubMed Scopus Google Scholar, E. J. Biochem. PubMed Scopus (33) Google Scholar), and a step is required for In are as by a to the protein encoded R. E. J. Biochem. PubMed Scopus Google Scholar, R. E. J. Biochem. PubMed Scopus Google Scholar). is the epimerization of the an of A acid is the of to its J. S. Biochem. PubMed Scopus Google Scholar). The for this begins with the of the as an acyl-CoA by an acyl-CoA synthetase the protein encoded by that is later by a and is by a R.D. Biochem. 1991; PubMed Scopus Google Scholar, J. 1993; PubMed Google Scholar). The epimerase has been purified from J. 1993; PubMed Google Scholar), that is to racemases E. J. Biochem. PubMed Scopus (33) Google Scholar). in this the protein encoded by the cefD2 has high similarity to and that of isopenicillin N to by an epimerization to occurs by a results that the of both genes cefD1 andcefD2 are for epimerase activity and for cephalosporin the and epimerization Finally, the required of the has been reported to in a by different R.D. Biochem. 1991; PubMed Scopus Google Scholar). The Streptomyces clavuligerus and lactamdurans isopenicillin N to by an entirely different since are proteins with a molecular of the A. or the S. clavuligerus that a of the amino at of the followed by in the (19Jensen S.E. Westlake D.W.S. Wolfe S. Can. J. Microbiol. 1983; 29: 1526-1531Crossref PubMed Scopus (33) Google Scholar, L. Liras P. Castro J.M. Martı́n J.F. J. Gen. Microbiol. 1990; 136: 663-671Crossref Scopus (27) Google Scholar). R. The and for isopenicillin N and penicillin N.