INDUSTRIAL ENZYMES

The present invention relates to an enzyme process for the one step conversion of cephalosporin C or a derivative thereof into 7-aminocephalosporanic acid or a corresponding derivative thereof. The one step conversion is effected using a cephalosporin C amidohydrolase enzyme derived from Pseudomonas vesicularis B965, or from any cephalosporin C amidohydrolase producing or potentially producing descendants thereof, or from any expression of DNA, particularly a recombinant DNA molecule, derived from Pseudomonas vesicularis B965 as described herein, or any cephalosporin C amidohydrolase producing or potentially producmg descendants thereof.

Cephalosporin C is the fermentation product of the cephalosporin biosynthesis pathway and although it has been shown to have some activity against gram-negative microorganisms as an antibiotic itself, the major commercial use of cephalosporin C is as a building block for other cephalosporin-like antibiotics. In particular, the D-α- aminoadipoyl side chain may be removed to give the highly useful intermediate 7- aminocephalosporanic acid (7-ACA) which is a precursor to a wide range of semi- synthetic cephalosporin antibiotics including cephalothin, cephaloridine and cefuroxime.

The current industrial process for producing 7-ACA from cephalosporin C is a chemical cleavage of the D-α-aminoadipoyl side chain. There are several different methods in use (see for instance A Smith in "Comprehensive Biotechnology: the Principles, Applications and Regulations of Biotechnology in Industry, Agriculture and Medicine", Volume 3 ("The Practice of Biotechnology: Current Commodity Products"), Eds. H W Blanch et.αl, esp. pp.163- 185, Pergamon Press, Oxford, UK, 1985), but all are essentially imino-halide processes with appropriate protection of amino and carboxyl groups. See, for example, the nitrosyl chloride cleavage developed by Morin et αl (J.Am.Chem.Soc, 84, 3400 (1962)) now superseded by the imino ether method developed by Ciba Geigy (Fechtig et αl, Helv. Chi . Acta., 51, 1108 (1968)).

These chemical processes have several disadvantages which include: the cost of the chemical reagents for protection and cleavage; the expense of providing the required

low operating temperatures (e.g. -20 C); the cost of the complex, often multi-step plant; the cost of the measures to contain the toxic chemical reagents (e.g. trimethyl silyl chloride, phosphorous pentachloride and chloroacetyl chloride); and the need to purify the highly impure cephalosporin C (or a derivative) starting material.

There is therefore a need to provide a means of converting cephalosporin C (or a derivative thereof) to 7-ACA (or a corresponding derivative thereof) which is cheap, involves simple technology, is environmentally friendly and is safe. Such criteria are met herein using an enzyme process.

The search for efficient microbiological or enzyme processes for converting cephalosporin C (or a derivative thereof) into 7-ACA (or a corresponding derivative thereof) has been largely unsuccessful. There have been reports in the literature of two-stage enzymatic processes for converting cephalosporin C to 7-ACA- These require the initial conversion of cephalosporin C into glutaryl-7-ACA using a D-amino acid oxidase (see for instance US Patent-Nos 3658649 or 3801458) followed by cleavage of the glut^l side chain to give 7-ACA. Two-stage enzymatic processes have, for example, been described by Shibuya et al, AgricBiol. Chem..45. 1561-1567 (1981), but all suffer from the disadvantage of reduced efficiency and increased complexity compared with the one step conversion.

There have also been reports of low activity single-step enzyme reactions. For example, EP0283218, EP0322032, and EP0405846, describe one step conversion of cephalosporin C to 7-ACA using enzymes derived from Arthrobacter viscosus, Bacillus Megaterium, and another Bacillus species respectively. EP0474652 describes an enzyme capable of a one step conversion of cephalosporin C to 7-ACA which is derived from Pseudomonas diminuta. Despite extensive work in this area it has not been demonstrated that an enzyme isolate as described above is able to be used for production of commercially relevant amounts of 7- AC A.

It is particularly unexpected therefore, that we have isolated a strain of Pseudomonas vesicularis which produces an amidohydrolase (also known as amidase or acylase) enzyme that can convert cephalosporin C (or a derivative thereof) into 7-ACA (or a corresponding derivative thereof) by way of a one .step conversion. The enzyme activity is only weakly observed in the wild type strain, however, useful activity is

- 3 -

achieved following partial purification of the enzyme. Furthermore, the gene coding for the amidohydrolase enzyme activity has also been isolated and sequenced (Sequence Identity Number 1) and can therefore be expressed in a recombinant host such as Kcoli to produce increased amounts of the enzyme. In addition it is a further feature of this invention that the amidohydrolase enzyme isolated is sufficiently robust that it may be immobilised whilst maintaining significant enzymatic activity. Immobilisation of the enzyme is a particularly important step in the adaption of an enzyme for use within large scale fermentation synthesis of 7-ACA. The use of an immobilised enzyme as compared to free enzyme in fermentation has obvious benefits in significantly reducing the requirement for enzyme and subsequent reduction in cost of 7-ACA manufacture.

Accordingly, there is provided in a first aspect of the present invention a process for the one step conversion of cephalosporin C or a derivative thereof into 7-aminocephalosporanic acid or a corresponding derivative thereoζ comprising treating said cephalosporin C or a derivative thereof with cephalosporin C amidohydrolase derived from Pseudomonas vesicularis B965, or from any cephalosporin C amidohydrolase producing or potentially producing descendants thereof, or from any expression of DNA, particularly a recombinant DNA molecule containing DNA of said Pseudomonas vesicularis B965, or any cephalosporin C amidohydrolase producing or potentially producing descendants thereof.

In an alternative aspect of the present invention, there is provided a process for the one step conversion of cephalosporin C or a derivative thereof of formula (I)

wherein

R ¹ represents a group selected from -CO(CH ₂ )3CH(NHR ⁵ )CO ₂ R ⁴ ,

-CO(CH ₂ ) ₃ CO ₂ R ⁴ or -CO(CH ₂ ) ₂ CO ₂ R ⁴ ;

R ² represents a carboxylic acid or a carboxylate group or a salt, an ester or a protected derivative thereof;

R ³ represents a hydrogen atom or a group selected from -CH ₂ OC(O)CH3, -CH ₂ OH,

-CH3, -CH ₂ OC(O)NH ₂ ^{or a} pyridiniummethyl group;

R represents a hydrogen atom or a carboxyl protecting group; and

R5 represents a hydrogen atom, a benzoyl group or an amino protecting group; into 7-aminocephalosporanic acid or a corresponding derivative thereof of formula (II)

wherein R ² and R ³ are as defined above; comprising treating a compound of formula (I) with a cephalosporin C amidohydrolase derived from Pseudomonas vesicularis B965, or from any cephalosporin C amidohydrolase producing or potentially producing descendants thereof, or from any expression of DNA, particularly a recombinant DNA molecule containing DNA, of said Pseudomonas vesicularis B965, or any cephalosporin C amidohydrolase producing or potentially producing descendants thereof.

In preferred aspect of the present invention the compound of formula (I) is cephalosporin C (i.e. R ¹ is the group -CO(CH ₂ )3CH(NH ₂ )CO ₂ H; R ² is a carboxylic acid group; and R ³ is the group -CH ₂ OC(O)CH3) and the compound of formula (II) is 7-aminocephalosporanic acid (i.e. R ² is a carboxylic acid group; and R ³ is the group -CH ₂ OC(O)CH ₃ ).

In a further aspect of the present invention, there is provided a recombinant DNA molecule for use in cloning a DNA sequence in a eukaryotic or prokaryotic host, said recombinant DNA molecule comprising a DNA sequence selected from DNA sequences from the DNA sequence set out in Sequence Identity Number 1, which code on expression for a cephalosporin C amidohydrolase enzyme or functional variations of the DNA sequence as depicted in Sequence Identity Number 1, which codes on expression for a cephalosporin C amidohydrolase enzyme.

In a further preferred aspect of the present invention, the DNA or the recombinant DNA comprises a DNA sequence selected from part or all of the DNA sequence set out in Sequence Identity Number 1 or functional variations thereof.

In an alternative aspect of the present invention, there is provided a recombinant DNA molecule for use in cloning a DNA sequence in a eukaryotic or prokaryotic host, said recombinant DNA molecule comprising the DNA sequence as depicted in Sequence Identity Number 1 herein or a functional variation of the DNA sequence as depicted in Sequence Identity Number 1, which codes on expression for a polypeptide having the biological activity of the enzyme cephalosporin C amidohydrolase.

In a further alternative aspect of the present invention, there is provided a recombinant DNA molecule comprising a DNA sequence which hybridizes to the DNA sequence as depicted in Sequence Identity Number 1 or is related to the DNA sequence as depicted in Sequence Identity Number 1 by mutation.

It will be appreicated that the term a "functional variation of the DNA sequence as depicted in Sequence Identity Number 1" as used comprises a DNA sequence selected from:

(i) DNA sequences which are allelic variations of the DNA sequence as depicted in Sequence Identity Number 1 ;

(ii) DNA sequences which hybridize with the above sequences (i) or the DNA sequence as depicted in Sequence Identity Number 1; and

(iii) DNA sequences which are degenerate as a result of the genetic code to the above sequences (i) and (ii), or the DNA sequence depicted in Sequence Identity Number 1.

Preferably a "a function variation of the DNA sequence as depicted in Sequence Identity Number 1" will encode a polypeptide which comprises an amino acid sequence substantially corresponding to the sequence of native cephalosporin C amidohydrolase as found in Pseudomonas vesicularis B965 (Sequence Identity

Number 1) or contain one or more deletions, substitutions, insertions, inversions or additions of allelic origin or otherwise. The resulting polypeptide will have at least 80% and preferably 90% homology with the sequence of a native cephalosporin C amidohydrolase enzyme and retain essentially the same cephalosporin C amidohydrolase activity as herein defined or, more preferably, exhibit enhanced activity or enzyme kinetics in the presence of cephalosporin C substrate, for example, by increasing the affinity of the enzyme for its substrate (i.e. reduction in Km) or enhancing the rate of reaction (i.e. increase in K^t).

It will be appreciated that the cephalosporin C amidohydrolase-encoding recombinant DNA sequences of the present invention may comprise an operatively linked transcription and translation activating sequence that controls the expression of the cephalosporin C amidohydrolase-encoding recombinant DNA sequence. Such activating sequences for use in recombinant DNA expression are well known in the art.

It will also be appreciated that it is particularly desirable to combine the cephalosporin C amidohydrolase-encoding recombinant DNA sequence with a start codon such as GTG, CTG or, most preferably, ATG, by conventional methods.

Furthermore, the cephalosporin C amidohydrolase-encoding recombinant DNA sequences of the present invention may comprise regulatory signals at the 3' end of the coding sequence such as the stop codons TAG, TAA and TGA, and mRNA polyadenylation and processing signals.

It will be appreciated that a polypeptide with cephalosporin C amidohydrolase activity according to the present invention may be expressed as a fusion protein using a gene fusion where a cephalosporin C amidohydrolase-encoding recombinant DNA sequence of the present invention is joined to a coding sequence for another protein such that their reading frames are in phase. Such gene fusions may be constructed in order to link the polypeptide with cephalosporin C amidohydrolase activity to a signal peptide to allow its secretion by a transformed cell. Another use of this technique may be to enable the splicing of a cephalosporin C amidohydrolase-encoding recombinant DNA sequence behind a promoter sequence in the correct reading frame without the need to position it exactly adjacent to the promoter. A further use of this technique would be

to add an affinity tag to the polypeptide with cephalosporin C amidohydrolase activity to facilitate assaying for the polypeptide.

The term "cephalosporin C amidohydrolase-encoding recombinant DNA sequences'* is intended to incorporate those sequences which contain one or more modifications such as mutations, including single or multiple base substitutions, deletions, insertions or inversions and which code on expression for a polypeptide possessing cephalosporin C amidohydrolase activity as herein defined. Where modified DNA sequences are used, these may be those recombinant DNA sequences which are degenerate as a result of the genetic code which code on expression for a polypeptide with an a ino acid sequence identical to that of a native cephalosporin C amidohydrolase enzyme as found in Pseudomonas vesicularis B695.

Alternatively, a cephalosporin C amidohydrolase-encoding recombinant DNA sequence incorporating the modifications described above may code on expression for a polypeptide possessing cephalosporin C amidohydrolase activity as herein defined, in which one or more deletions, substitutions, insertions, inversions or additions of allelic origin or otherwise, results in improved expression of the enzyme and/or the enzyme having improved catalytic activity, for example, by increasing the affinity of the enzyme for its substrate (i.e. reduction in K^) or enh-ancing the rate of reaction (i.e increase in K _ca ^).

It will further be appreciated that the cephalosporin C amidohydrolase-encoding recombinant DNA sequences of the present invention may be combined with other recombinant DNA sequences.

For example, the cephalosporin C amidohydrolase enzyme of the present invention will accept glutaryl-7-ACA as a substrate, hence the cephalosporin C amidohydrolase- encoding recombinant DNA sequences may be combined with a D-amino acid oxidase- encoding recombinant DNA sequence. Alternatively, the cephalosporin C amidohydrolase enzyme of the present invention naturally accepts cephalosporin C as a substrate, hence the cephalosporin C amidohydrolase-encoding recombinant DNA sequences may be combined with a DAC acetyltransferase-encoding recombinant DNA sequence.

The recombinant DNA sequences of the present invention may be introduced directly into the genome of an antibiotic-producing microorganism in which case they may be used as such. Alternatively, the recombinant DNA sequences of the invention may be introduced by standard techniques known in the art by the use of recombinant DNA expression vectors. When introduced in this way it will be appreciated that the vector will, in general, additionally comprise promoter and/or translation activating sequences, operatively linked to the cephalosporin C amidohydrolase-encoding recombinant DNA sequences and preferably additionally using a selectable marker system for identifying the transformed host, for example, as described herein below.

There is thus provided in a further aspect of the present invention a recombinant DNA expression vector comprising a recombinant DNA sequence of the invention and additionally comprising a promoter and translation activating sequence operatively linked to the cephalosporin C amidohydrolase-encoding recombinant DNA sequence.

In an alternative aspect the recombinant DNA sequences of the present invention may be introduced into a non-antibiotic-producing microorganism, for example, Escherichia coli, either directly or as part of a recombinant DNA expression vector to provide the expression .and isolation of an enzyme with cephalosporin C amidohydrolase activity which may be useful in catalysing in vitro antibiotic preparation or the in vitro preparation of precursors for use in antibiotic production. Even if expression of a polypeptide with cephalosporin C amidohydrolase activity is not intended in Kcoli, a suitable recombinant DNA expression vectoζ should preferably be capable of replication in a prokaryotic host such as Kcoli to facilitate further genetic manipulation.

According to a further aspect of the present invention, there is thus provided a eukaryotic or prokaryotic host transformed with a recombinant DNA expression vector comprising a cephalosporin C amidohydrolase-encoding recombinant DNA sequence of the invention.

Preferred hosts include species of filamentous fungi such as Acremonium spp or Penicillium spp , species of streptomyces, or other bacteria such as Kcoli or Bacillus spp. Particularly preferred hosts are Acremonium chrysogenum or

K coli.

In a further aspect of the present invention, there is provided a method of expressing cephalosporin C amidohydrolase activity in a eukaryotic or prokaryotic host comprising:

a) transforming the host cell with a recombinant DNA expression vector that comprises (i) an expression control sequence that functions in the host cell, and (ii) a cephalosporin C amidohydrolase-encoding recombinant DNA sequence which is operatively linked to the expression control sequence;

b) culturing the host cell transformed in step (a) under conditions that allow for expression of cephalosporin C amidohydrolase activity.

It will be appreciated that the recombinant DNA sequence described herein may contain lengths of recombinant DNA which do not code on expression for cephalosporin C amidohydrolase. Although these non-coding sequences in no way detract from the usefulness of the whole recombmant DNA sequence, it may be convenient or desirable to utilise a smaller restriction fragment derived from the recombinant DNA sequences. The restriction fragments may be obtained, for example, by cleavage with suitable restriction endonucleases enzymes or specific exonuclease enzymes by methods well known in the art.

According to yet another aspect of the present invention, there is provided a recombinant DNA expression vector which comprises a gene coding on expression for a cephalosporin C amidohydrolase polypeptide having the a ino acid sequence depicted in Sequence Identity Number 1' or allelic or other functionally equivalent variations thereof in which one or more amino acids has or have been added, substituted or removed without substantially affecting the cephalosporin C amidohydrolase activity in the enzyme activity assay described herein.

Suitable recombinant DNA expression vectors may comprise chromosomal, non- chromosomal and synthetic DNA such as derivatives of known bacterial plasmids, for example, 'natural' plasmids such as ColEl or 'artificial' plasmids such as pBR322, pAT153, pUC18, pUC19, pACYC 184 or pMB9, or yeast plasmids, for example, 2μ.

Other suitable vectors may be phage DNA's, for example, derivatives of Ml 3, or bacteriophage lambda (λ), or yeast vectors such as derivatives of Yip, YEp or YRp.

It will be appreciated that apart from containing the recombinant DNA sequence or a restriction fragment derived therefrom as herein defined, the recombinant DNA expression vector may also comprise a promoter and translational activating sequence which not only functions in the chosen host cell, but also is operatively linked - meaning that it is positioned in the correct orientation and position to control expression of the cephalosporin C amidohydrolase-encoding recombinant DNA sequence. Such an activating sequence may, of course, be found in the full-length recombinant DNA molecule described herein.

Examples of useful promoter sequences which might be used in the recombinant DNA expression vectors of the invention include, for example, the gcbC promoter of isopenicillin N synthetase (IPNS) from Acremonium spp or Penicillium spp, the cefEF promoter of expandase hydroxylase (DAOCS DACS) from Acremonium spp, the penDE promoter of Acyl CoA:6APA acyl transferase from Penicillium spp, and numerous promoter sequences from Aspergillus spp including pgk. gpdA. pjά, amdS. argB. trpC. alcA. aldD and niaD promoter sequences.

Further examples of useful promoter sequences which might be used in the recombinant DNA expression vectors of the invention include the glycolytic promoters of yeaΛ (for example, the promoter of 3-phosphoglycerate kirrase, PGK), the promoters of yeast acid phosphatase (for example, Pho 5) or the alcohol dehydrogenase-2 (alcA) or glucoamylase promoter. Where expression is carried out in a bacterial host such as Kcoli useful promoters which might be used in the recombinant DNA expression vectors of the invention are well known in the art, and include the λP]JCτ857 system as well as other common promoters such as tac, lac. trpE and recA. Another suitable expression system would be the T7 polymerase system.

In addition, such recombinant DNA expression vectors may possess various sites for insertion of a cephalosporin C amidohydrolase-encoding recombinant DNA sequence of this invention. These sites are characterised by the specific restriction endonuclease which cleaves them. Such cleavage sites are well recognised by those skilled in the art.

The expression vector, and in particular the site chosen therein for insertion of a selected recombinant DNA fragment and its operative linking to an expression control sequence, is determined by a variety of factors including the number of sites susceptible to a given restriction enzyme, the size of the protein to be expressed, contamination or binding of the protein to be expressed by host cell proteins which may be difficult to remove during purification, the location of start stop codons, and other factors recognised by those skilled in the art. Thus the choice of a vector and insertion site for a recombinant DNA sequence is determined by a balance of these factors, not all selections being equally effective for a given case. Likewise, not all host/vector combinations will function with equal efficiency in expressing the recombin.ant DNA sequences of this invention. The selection is made, depending upon a variety of factors including compatability of the host and vector, ease of recovery of the desired protein and/or expression characteristics of the recombinant DNA sequences and the expression control sequences operatively linked to them, or any necessary post-expression modifications of the desired protein.

The recombinant DNA expression vector may also comprise a selectable marker which will facilitate screening of the chosen host for transformants. Suitable markers for use in fungal transformations are well known in the art and include the acetamidase (a dS) gene which confers upon transformants the ability to utilise acetamide as the sole source of nitrogen, antibiotic resistance, for example, the phosphotransferase genes which confer upon transformants resistance to aminoglycosidic antibiotics such as G418, hygromycin B and phleomycin, or resistance to benomyl by transformation with an altered tubulin gene. Another useful marker is the nitrate reductase (niaD) gene which confers upon nitrate reductase deficient mutants the ability to utilise nitrate as the sole nitrogen source. Other marker systems for filamentous fungi involve the reversions of arginine auxotrophy using the argB gene, tryptophan C auxotrophy using the trpC gene and uracil auxotrophy using the pyr-4 or pyrG genes.

It will be appreciated that the cephalosporin C amidohydrolase gene of the present invention and the chosen marker gene may conveniently be part of the same plasmid or alternatively they may be on different plasmids in which case the host must be co- transformed using any suitable technique well known in the art.

The recombinant DNA expression vector may also optionally be characterised by an autonomous replication sequence (ars) derived from a fragment of either chromosomal or mitochondrial DNA, preferably from the same species as that being transformed.

In bacterial, e.g. K coli, expression systems, selectable markers are very well established in the art. Of particular use are markers which confer antibiotic resistance to, for example, tetracycline or chloramphenicol.

It will be appreciated that the cephalosporin C amidohydrolase need not necessarily be totally pure, and that a substantial purification by techniques well known in the art should be sufficient to provide an enzyme composition with cephalosporin C amidohydrolase activity.

In a further aspect of the present invention, there is therefore provided a cephalosporin C amidohydrolase-containing enzyme composition capable of the one step conversion of cephalosporin C or a derivative thereof into 7-aminocephalosporanic acid or a corresponding derivative thereof, wherein said enzyme composition is derived from Pseudomonas vesicularis B965, or any cephalosporin C amidohydrolase producing or potentially producing descendants thereof, or from any expression of the genetic material of said Pseudomonas vesicularis B965, or any cephalosporin C amidohydrolase producing or potentially producing descendants thereof.

Reference herein to protection of amino and carboxyl groups refers to the use of conventional protecting groups, for example as described in "Protective Groups in Organic Synthesis" (2nd Edition) by Theodora Greene and Peter Wuts (John Wiley and Sons Inc. 1991).

The principles of using protecting groups are well established. Greene and Wuts propose a number of criteria for useful protective groups:

• It must react selectively in good yield to give a protected substrate that is stable to the projected reactions;

• It must be selectively removed in good yield by readily available, preferably non-toxic reagents that do not attack the regenerated functional group;

• It should form a crystalline derivative (without the generation of new stereogenic centres) that can easily be separated from side products associated with its formation or cleavage;

• It should have a minimum of additional functionality to avoid further sites of reaction.

Suitable protective groups would be readily recognised by a person of ordinary skill in the art.

Salts of the compounds of formulae (I) and (H) include inorganic base salts such as alkali metal salts (e.g. sodium and potassium salts).

Esters of the compounds of formulae (I) and (H) are preferably physiologically acceptable and include acyloxyalkyl esters, for example, lower alkanoyloxymethyl or - ethyl esters such as acetoxymethyl, acetoxyethyl or pivaloyloxymethyl esters, and alkoxycarbonyloxyethyl esters, for example, lower alkoxycarbonyloxyethyl esters such as the ethoxycarbonyloxyethyl ester.

As stated above, the process of the present invention comprises treating cephalosporin C or a derivative thereof with a cephalosporin C amidohydrolase of the present invention. In this context, the term "treating" means any conventional method of contacting the substrate with the enzyme.

Thus, for example, a partially purified solution or cell free broth of crude cephalosporin C or a derivative thereof may be utilised as the feed stream which is treated in a batch-wise or continuous manner with the cephalosporin C amidohydrolase enzyme composition of the present invention. Such a method may be effected with or without any prior purification of the substrate or the enzyme.

Alternatively, the cephalosporin C amidohydrolase of the present invention may be immobilised and used, for example, in a stirred tank reactor or as an enzyme column through which the cephalosporin C or a derivative thereof is passed. Such an

immobilised form of the cephalosporin C amidohydrolase enzyme composition constitutes a further aspect of the present invention.

Methods of enzyme immobilisation are well known in the art (see, for instance, "Biotechnology"; Volume 7a, Enzyme Technology (Ed. J F Kennedy), Chapter 7, pp347-404, VCH Publishers, Cambridge, UK, 1987). The enzyme may be bound to a support by, for example covalent, ionic or hydrophobic interactions or alternatively by physical adsorption, for example, by entrapment in gel or fibre, or by microencapsulation.

Suitable support materials include natural polymers such as cellulose, starch, dextran, agarose, alginate and protein; synthetic polymers such as polyacrylamides, polyacrylates and polymethacrylates and styrene dibenzene co-polymers; or minerals such as silica, bentonite, kieselgur, glass and metals.

A particularly preferred means of immobilisation is by covalent binding of the enzyme to a support using activation reagents such as cyanogen bromide or glutaraldehyde, or supports with reactive groups such as oxirane, hydrazine, N-hydroxysuccinimide, carbonyldϋmidazole or pyridyldithiopropyl. Particularly preferred supports include cross-linked polyacrylamides, agarose and silica.

Other processes for contacting the substrate with the enzyme include the use of membrane reactors, or by simultaneous culture of Pseudomonas vesicularis B965, or any cephalosporin C amidohydrolase producing or potentially producing descendants thereof, with the cephalosporin C (or derivative) producing filamentous fungi.

As used herein, the term "cephalosporin C amidohydrolase producing, or potentially producing, descendants thereof is intended to incorporate all mutants of Pseudomonas vesicularis B965 which might arise through either spontaneous or induced mutation, other than those which do not produce cephalosporin C amidohydrolase. It should be noted however that some non-producers will still be "potentially producing descendants" due to the fact that their genetic material may still be capable of expressing cephalosporin C amidohydrolase as a result of the application of standard recombinant DNA techniques.

Methods of mutagenesis as a means of strain improvement are well known in the art. Modern recombinant DNA techniques provide a more directed means of strain improvement which may involve, for example, the expression of all or part of the genetic material of Pseudomonas vesicularis B965 in another microorganism, for example, Kcoli. The recombmant gene coding on expression for cephalosporin C amidohydrolase may or may not be reintroduced into the parent strain of Pseudomonas. Alternative expression systems well known in the art may be utilised to produce large quantities of cephalosporin C amidohydrolase for use in accordance with the present invention. The expression of all or part of the genetic material of Pseudomonas vesicularis B965, or any cephalosporin C amidohydrolase producing or potentially producing descendants thereof constitutes a further aspect of the present invention.

In the following non-limiting examples, included to facilitate further understanding of the present invention, the following abbreviations may be used:-

Ceph C cephalosporin C

7ACA 7-aminocephalosporanic acid

GL-7ACA 7-β-(4-carboxybutanamido)-cephalosporanic acid (glutaryl

7-ACA) kb kilobase

DTT dithiothreitol

EDTA ethylenediaminetetraacetic acid

MOPS 3-[N-morpholino]propanesulphonic acid

SDS Sodium lauryl sulphate

SDS PAGE SDS polyacrylamide gel electrophoresis

Tris-HCl tris(hydroxymethyl)aminoethane hydrochloride

TSB tryptone soya broth

DMAB p-dimethylaminobenzaldehyde

2ME 2-mercaptoethanol

PVDF polyvinylidene difluoride gac GL-7ACA and Ceph C amidohydrolase-encoding gene

Preparation 1: Recovery and Assav of Amidohvdrolase Enzyme Activity from Pseudomonas vesicularis

(a) Natural Isolate Screening System

Microbial isolates were recovered from a wide range of environmental samples using standard microbiological methods as typified in L R Hawker & A H Linton, "Micro- Organisms: Function, Form and Environment", (Edward Arnold Pubs. 1971); R G Board & D W Lovelock, "Sampling-Microbiological Monitoring of Environments", (Academic press, 1973); J M Lynch & N J Poole, "Microbial Ecology: A Conceptual Approach", (Blackwell Scientific, 1979); and "The Oxoid Manual, 5th Edition", (Oxoid Limited, Wade Road, Basingstoke, Hants, RG24 OPW).

Over 105,000 isolates from nature were subjected to an assay system involving around 260,000 assays in the screening for amidohydrolase enzyme activity. For this assay system, the isolates were grown on complex media solidified with agar in order to provide inoculum to an enzyme substrate mixture. During initial screening the enzyme substrate mixture typically contained lOOmM sodium phosphate (pH 7.0); 3mg ml Ceph C; 3mg/ml GL-7ACA; 13.5mg ml cloxacillin. A sample of freshly grown cells was incubated for 16-24 hours in enzyme substrate mixture. The 7ACA produced was detected by reaction with DMAB following the method of Balsingha et al (Biochim. Biophys. Acta., 276, 250-256).

From this initial screening, 38 isolates were identified as possessing amidohydrolase activity against GL-7ACA. Further testing was undertaken using the methods of preparation 1(d) and 1(e) described below to identify amidohydrolase activity against Ceph C. Of the 38 isolates with GL-7ACA amidohydrolase activity, only 1 strain (present in 4 sister isolates derived from a single environmental sample) had an amidohydrolase enzyme activity with Ceph C.

(b) Strain B965

The bacteriological characteristics of Pseudomonas vesicularis strain B965 are as follows:

1. Morphology

1.1 Colony colour (grown on nutrient agar or TSB agar) - egg yolk yellow, semi- translucent

1.2 Colony morphology - round, regular, entire, convex, smooth, shiny, 0.5mm diameter after 3 days, 30°C

1.3 Cell type - rods, 1.5-3μm length approx

1.4 Gram - negative

1.5 Motility - positive

1.6 Spores - negative

1.7 Growth temperature - optimum 30 C; 37 C + ; 41 C -

2. Biochemical Characteristics:

2.1 Reduction of nitrate to nitrite _

2.2 Reduction of nitrate to nitrogen _

2.3 Indole production from tryptophan _

2.4 Glucose acidification _

2.5 Arginine dihydrolase _

2.6 Urease -

2.7 β-Glucosidase hydrolysis. of aesculin +

2.8 Protease hydrolysis of gelatin _

2.9 β-Galactosidase activity _

2.10 Catalase +

2.11 Oxidase -

2.12 Production of lysine decarboxylase -

2.13 Acid from xylose _

2.14 Alkalisation of Simmons citrate _

2.15 Reaction in Litmus milk No Change

2.16 Assimilation of glucose + arabinose - mannose _ mannitol _

N-acetyl-glucosjunine . maltose - gluconate _ caprate _ adipate (+) malate _ citrate _ phenyl-acetate _

2.17 Cytochrome oxidase +

(c) Culture Conditions

The strain is maintained by the sub-culture of single isolated colonies onto agar plates. A number of complex agar media support good growth, in particular tryptone soya broth (TSB; Oxoid CM129) + 2.0% w/v agar. This media has a pH of 7.3 ± 0.2 and the following composition:

Component Blt

Pancreatic digest of caesin 17.0

Papaic digest of soybean meal 3.0

Sodium chloride 5.0

Dibasic potassium phosphate 2.5

Dextrose 2.5

Agar powder 20.0

Cells are streaked onto the surface of an agar plate which is incubated at 27.5°C for 2-6 days. Colonies remain viable for several weeks if the agar plates are stored at 4°C.

For long term preservation a single colony from an agar plate is inoculated into 10ml of liquid medium. Most complex media designed for bacterial growth are suitable, in particular L-broth can be used. This has the following composition:

Component g/L

Bacto tryptone (Difco) 10.0

Yeast extract (Oxoid L21 ) 5.0

Sodium chloride 10.0

Inoculated broth is incubated at 27.5°C with shaking at 220 rpm. After 24 hours the broth is centrifuged. The cell pellet is resuspended in 1ml of 5% glycerol in water and stored in aliquots at -20°C.

(d) Enzyme Activity Assay

Substrate buffer consists of 50mM sodium phosphate buffer pH 7.0 containing 13.5mg/ml cloxacillin and GL-7ACA (3mg ml) or cephalosporin C (lOmg ml). Cell suspensions or semi-purified enzyme samples were incubated in substrate buffer for up to 24 hours at 37 C. At a suitable time a lOOμl sample was withdrawn and mixed with 800μl 1M citrate-phosphate buffer (pH 4.2). A lOOμl aliquot of fluram reagent was then added, mixed and the sample incubated at room temperature for 1 hour. Fluram reagent consists of fluorescamine (Sigma Chemical Co) at lmg/ml in dry acetone. The presence of 7ACA in the reaction mix was detected after derivatisation of the free primary amine group by using fluorescence spectroscopy at an excitation wave length of 378nm and emission detection at 495nm in a flow injection assay system with 10% acetone in water as carrier fluid.

In addition, 7ACA production was also demonstrated in an HPLC system. The enzyme reaction mix was derivatised as above and a 20μl sample applied in a mobile phase consisting of 35% acetonitrile, 0.1% trifluoroacetic acid in HPLC-grade water onto a 15cm Hypersil ODS 5μ column at 32°C with a flow rate of 1.5ml/min. Detection by fluorescence spectroscopy was as above. In this system authentic 7ACA standard has a retention time of 6.7 minutes, with a minor degradation peak at 10.5 minutes.

In typical experiments using these conditions B965 cells sufficient to fill a lOμl inoculating loop produced 320μg ml 7 AC A in 18 hours incubation using GL-7ACA as substrate. Using cephalosporin C as substrate with double the volume of cells enabled production of 5μg/ml 7ACA in an overnight incubation.

(e) Enzyme Recovery

B965 cells were cultured on TSB agar plates essentially by the method of preparation 1(b). The cells were harvested into enzyme buffer, washed and finally resuspended in 10ml chilled enzyme buffer. Enzyme buffer is lOOmM sodium phosphate buffer (pH 7.0) cont.aining ImM DTT and ImM benzamidine. The cells were lysed by sonication using a 650W output sonicator (Model W-380, Heat System Ultrasonics Inc, 1938 New Highway, Farmingdale, NY 11735, USA) on power settings 2-3, 50% cycle, for 12 minutes. The extract was clarified by centrifugation at 18,000g for 15 minutes at 4 C. The supernatant was recovered, adjusted to 10ml volume in enzyme buffer, then solid ammonium sulphate was added slowly with gentle mixing at 4 C to 30% saturation. The solution was clarified by centrifugation as above and the supernatant adjusted to 60% saturation in ammonium sulphate by slow addition of solid as above. The precipitated material was harvested by centrifugation as above and stored over saturated ammonium sulphate.

Further purification of the amidohydrolase was undertaken using an FPLC system (Pharmacia LKB Biotechnology, S-751 82 Uppsala, Sweden) in order to define the properties of the enzyme.

Material recovered from the 30-60% saturation ammonium sulphate precipitate was resuspended in MOPS buffer (lOmM MOPS pH8.0; ImM DTT; ImM benzamidine). The solution was clarified by brief centrifugation followed by filtration through a 0.2 micron filter. A 5ml aliquot of enzyme mixture was applied to an HR5/5 MonoQ column pre-equilibrated with MOPS buffer. Amidohydrolase activity was recovered in the eluate between 2-4ml after loading. This fraction was adjusted to 30% saturation in ammonium sulphate and loaded to an HR10/30 phenylsuperose column. Elution of protein was achieved using a salt gradient regime varying from 30% - 0% saturation in ammonium sulphate with a running buffer consisting of lOOmM sodium

phosphate pH7.0, ImM DTT. Amidohydrolase activity was recovered at between 18.5% - 15.0% saturation of ammonium sulphate under these conditions.

Following these purification steps the following typical values of amidohydrolase activity were obtained using the method of preparation 1(c):

(f) Cephalosporin C Amidohydolase Enzyme Properties

Physical

1. Apparent molecular weight by gel filtration: 70000 ± 4000. 2. Subunit molecular weight by SDS-PAGE: α 30,000, β 60,000. 3. Soichiometry apparent: αl βl. 4. Isoelectric point: pi 10.5 ± 0.2. 5. Hydrophobicity: does not bind to butyl-, hexanyl-, octyl- or decyl-derivatised agarose; binds to phenyl-derivatised agarose with elution at 650-800mM ammonium sulphate.

6. Stable in presence of ImM DTT. 7. Precipitates from solution at or above 50% saturation ammonium sulphate.

Kinetic

1. pH optimum: 8 (effective range 7 - 9). 2. Temperature optimum: 40 C (stable to 45 C). 3. K m, GL-7ACA substrate at 1.2-7.2mM, K m = 2-4mM;

Ceph C substrate at 0.5-1.5mM, K _m = 0.012mM;

Ceph C substrate at 3.6-132mM, K _m = 60-lOOmM.

4. Inhibitors: ammonium sulphate > 6% w/v. 5. Specific activity maximum O. Hlμmol 7ACA/min/mg enzyme using Ceph C substrate.

Example Substrates:

^a ) Assay response is given as product yield standardised to the 7ACA produced from GL-7ACA under equivalent reaction conditions. For the comparison it is assumed that products have molar extinction coefficients equivalent to 7ACA, such that the response was calculated directly from product peak area on HPLC. The standard reaction used lOOμl enzyme prepared by the procedure of preparation 2(d) from NM522/pVS44 cells, with lOmg/ml substrate in overnight reaction at 27 C.

(") Assay response obtained using lOμl of above enzyme extract, with 3 mg ml substrate in 2 hour reaction at 27 C. The actual response was then extrapolated to equivalent conditions to those used with the more slowly hydrolysed substrates.

Preparation 2: Cloning and Identification of Amidohydrolase-Encoding Gene

(a) Isolation of ^' P. vesicularis Genomic DNA

The organism was grown to late-log phase in 100ml L-broth at 28 C. The cells were pelleted by centrifugation (4000g, 10 minutes) and resuspended in 5 ml of TNES buffer (200mM Tris-HCl (pH 8.5); 250mM EDTA; 0.5% SDS). After 1 hour at 37°C, proteinase K was added to 50μg/ml and incubation continued at 37 C for 1 hour. An equal volume of phenol/chloroform/isoamyl alcohol (50:49:1), saturated with TNES buffer, were added and the suspension vortexed for 15 seconds. .After 10 minutes, at ioom temperature, the phases were separated by centrifugation (15,000g> 10 minutes 20°C). Two volumes of absolute ethanol was slowly added to the supernatant, and high molecular weight DNA spooled on to a glass rod. The DNA was dissolved in TE buffer (lOmM Tris-HCl, ImM EDTA, pH 8.0).

(b) Construction of the P. vesicularis Gene Library

Approximately 150μg P.vesicularis DNA was digested in a 900μl reaction for 45 minutes with 5 units Sau3AI. The DNA was precipitated using ethanol, resuspended in 500μl buffer and loaded on to a 38ml 10-40% sucrose gradient made in 1M NaCl, 20mM Tris-HCl (pH8.0), 5mM EDTA. After centrifugation overnight (26000rpm in SW28 rotor, Beckman L8-M ultracentrifuge, 15°C) 500μl fractions were taken, the

DNA ethanol precipitated, and analysed by agarose gel electrophoresis. Fractions in the size range 5-12kb were retained.

Approximately 200ng dephosphorylated BamHI-digested pSU18 vector was prepared by standard techniques. Plasmid pSU18 was obtained from Dr F de la Cruz (Departmento de Biologio Molecular, Universidad de Cantabria, 39001 Santander, Spain) and is derived from pSU2713 (Martinez et al, Gene, 68 (1988) pl59-162). The prepared vector was ligated with approximately 500ng size fractionated P. vesicularis DNA. Ligation was performed at 14 C overnight in a volume of 20μl with 1U T4 DNA ligase. The ligation mix was transformed into the widely available K coli strain NM522. Transformants were selected on L agar plus chloramphenicol (30μg/ml).

(c) Identification of Amidohydrolase-Encodin|g Clones

Single colonies were spread onto L agar plates containing chloramphenicol at 30μg/ml and grown at 27.5 C for 2-4 days. Aliquots of culture were recovered using a lOμl inoculating loop then resuspended in 0.5ml substrate buffer and further treated following the procedure of preparation 1(c). Isolates with amidohydrolase activity can be identified by these means.

(d) Amidohydrolase Activity of Clones

A clone identified by the method of preparation 2(c) carries an active gac gene on a recombinant plasmid. A further derivative plasmid was constructed by the subcloning of a 8.1kb PstI fragment carrying the gac gene into pSU18 by standard methods. This plasmid is termed pVS42. A restriction map of the P. vesicularis DNA portion of pVS42 is illustrated in Figure 1(a). Amidohydrolase activity using GL-7ACA or cephalosporin C as substrate is approximately equivalent between £.co/t/pVS42 and the original P. vesicularis isolate using whole cells by the method of preparation 1(c).

E. coli B1113 was grown on TSB agar plates following the method of preparation 1(b). The confluent cell growth was processed for amidohydrolase recovery essentially following the procedure of preparation 1(d) to the cell-free extract stage.

Amidohydrolase activity was further purified by application of the cell-free extract to a DE52 cellulose column (Whatman Ltd, Maidstone, Kent ME 14 2LE). The majority of proteins in the extract bind to the matrix and amidohydrolase activity is not retained.

Following this purification step the following typical values of amidohydrolase activity were obtained using the method of preparation 1(c).

Further subclones were derived from pVS42 by partial digestion of pVS42 plasmid DNA with the restriction enzyme Kpnl. ligation of the resultant mixture of reaction products and transformation into Kcoli strain NM522 using standard methods. Screening of the resultant isolates using procedures of preparation 1(b) .and 1(c) identified a number of clones with enhanced cephalosporin C amidohydrolase activity. One such isolate was found to contain a plasmid, termed pVS44. A restriction map of the P. vesicularis DNA portion of pVS44 is illustrated in Figure 1(b).

Cephalosporin C amidohydrolase enzyme activity was recovered from NM522/pVS44 cells using procedures of preparation 1(b) .and 1(c) to the cell free extract stage followed by chromatography on DE52 cellulose as described above. In a typical assay using the method of preparation 1(c), a lOμl extract produced 12μg/ml 7 AC A from 50mg/ml cephalosporin C in a 3 hour incubation at 27 C. This represents approximately 30-fold improvement over the level of activity obtained using Bl 113 strain extract at an equivalent stage of purification.

(e) Additional Subcloning to Improve Amidohydrolase Activity

The size of the P. vesicularis-deήved DNA fragment encoding the active gac of pVS44 was reduced by progressive exonuclease HI enzyme digestion. The methodology was as described by Henikoff ("Methods in Enzymology", 1_55_ (1987) 156- 165) . Unidirectional deletion into the P. vesicularis DNA portion of p VS44 was initiated at the Xhol site, while the pSU18 DNA portion was protected from

exonuclease HI attack by cutting at the PstI site forming the boundary between the P. vesicularis and K cσΛ'-derived DNAs. These sites are indicated in Figure 1. Following enzyme treatments and self-ligation of reaction products using the methods of Henikoff (Joe. cit.), transformation of K coli produced a population of colonies carrying variant plasmids with deletions of differing sizes. The population was screened for amidohydrolase activity by the method of preparation 1(c). A whole cell assay was used for screening. A 10 μl inoculation loop full of cells grown for 2 days at 27.5°C was suspended in 1 ml substrate buffer (lOmg ml cephalosporin C) and incubated overnight at 37°C. One of the isolates recovered from this screen with increased amidohydrolase activity carried the plasmid pVS4461. The P. vesicularis-deήved portion of pVS4461 is illustrated in Figure 1(c).

The P. vesicularis portion of pVS4461 was recovered as a EcoRI-Hindiπ fragment using restriction sites derived from the pSU18 multiple cloning region into which the P. vesicularis DNA was originally inserted. ThisEcoRI-HindlU fragment (of about 2.8 kb) was cloned between the EcoRI and Hindu! sites of the commercially available phagemid vector pTZ18R (Pharmacia P-L Biochemicals Inc, Milwaukee, Wisconsin 53202, USA) to yield the phagemid termed pTZ 1861. DNA of pTZ 1861 was cut at the single Seal site within the β-lactamase gene. A plasmid derivative of the widely available vector pACYC184 was used as the source of a DNA fragment encoding the tetracycline resistance gene. This fragment was converted to blunt ends, isolated and ligated to Seal-cut, diphosphorylated pTZ1861 DNA following standard methods. Following transformation, isolates selected on the basis of tetracycline resistance were .screened for amidohydrolase activity. One such active tetracycline-resistant isolate was selected and the recovered phagemid termed pTT1861. The P. vesicularis- derived portion of pTT1861 is identical to that in pVS4461 and is illustrated in Figure 1(c).

K coli strains carrying the various amidohydrolase plasmids were screened for amidohydrolase activity by the method of preparation 1(c) using a lOμl inoculation loop full of well grown cells, suspended in substrate buffer containing 50 mg/ml Ceph C.

Ceph C amidohydrolase activity was assessed following 3 hours incubation at 27.5°C. In comparative testing the relative activities of isolates carrying the plasmids and phagemids described are as follows:

(f) Amidohvdrolase Gene Sequence

The DNA sequence of the gac gene was derived by standard dideoxynucleotide incorporation methods. DNAs of the plasmids pVS42, pVS44 and subclones thereof were used as templates for sequencing reactions. Sequencing was initiated from either commercially available oligonucleotide primers annealing to the lacZ' portion of the plasmids or from oligonucleotides designed to anneal within the gac gene. Double stranded DNA was purified using Sephacryl S-400 Spun Columns following the manufacturer's instructions (Pharmacia P-L Biochemicals), then mixed with primer and treated with l/5vol IN NaOH at 37°C for 10 minutes. The preparations were neutralised with IN HCI and used in sequencing reactions following standard methodology (Sequenase Version 2.0 Kit; US Biochemical Corp). Some highly GC- rich regions caused the appearance of sequencing artefacts through sequence compressions or premature reaction termination. These were resolved through a combination of using deoxyinosine triphosphate in place of deoxyguanosine triphosphate and additional reaction of sequencing products with terminal deoxynucleotidyl transferase prior to resolution by PAGE, following the manufacturer's recommended procedures (US Biochemical Corp) and the method of Fawcett and Bartlett (BioTechniques, 9. 46 - 49 (1990)). Some of the sequence was derived from deletion subclones. The deletions were constructed using exonuclease in digestion following the methods of Henikoff (loc. cit.). The derived DNA sequence and deduced amino acid sequence thereof are shown in Sequence Identity Number 1.

(g) Amidohydrolase Amino Acid Sequence

The amidohydrolase enzyme specified by the gac gene encoded by the pTT1861 plasmid was purified to homogeneity for amino acid sequence analysis. The E. coli strain NM522/pTT1861 was grown on TSB medium supplemented with lOμg/ml tetracycline and solidified with 1.5% w/v agar. Growth was for lδhours at 37°C. About lg cell paste recovered by scraping from the agar surface was washed and resuspended in 4ml ice cold lysis buffer (lOmM MOPS pH 8.0; ImM DTT; ImM benzamidine). Cells were lysed by sonication (650W output sonicator for 4 pulses of 2 minutes at power setting 2; sample chilled on ice). The lysate was clarified by centrifugation at 15,000 rpm for 30 minutes. Enzyme purification was achieved using a Pharmacia FPLC system. The clarified lysate containing about lOOmg soluble protein was applied to an HR5/5 MonoQ ion exchange column pre-equilibrated with lysis buffer. Amidohydrolase activity was recovered in eluate fractions between 2-4 ml after loading. This material was adjusted to 30% saturation in ammonium sulphate and applied to an HR 10/30 phenylsuperose hydrophobic interaction column. Elution of protein was achieved using a non-linear gradient varying from 30% - 0% saturation in ammonium sulphate in a lOmM sodium phosphate buffer containing ImM DTT. Eluate fractions containing amidohydrolase activity were pooled and applied to an HR 10/30 Superose 12 gel filtration column in the phosphate buffer. Amidohydrolase activity was recovered in fractions equivalent to protein of molecular weight 70,000 ± 4000. At this stage, 3 protein bands were present following SDS- PAGE (Pharmacia Phastsystem). The eluate fractions containing amidohydrolase activity were concentrated using a centrifugal membrane filter device (Centricon 30 Microconcentrator, Amicon Division of W R Grace,Danvers MA 01923, USA). About 3mg protein was recovered by this procedure. .Aliquots of protein were treated with SDS and 2ME and resolved by SDS-PAGE using ultrapure reagents following standard methods as described by Ha es and Rickwood in "Gel Electrophoresis of Proteins: A Practical Approach" (IRL Press). Major peptide species of about 30,000 and 60,000 molecular weights were recovered by blotting to PVDF membrane (P Matsudaira, J Biol Chem., 262. 10035-10038 (1987)). The material was subjected to automated amino acid analysis by Edman degradation in an Applied Biosystems 477A Protein Sequencer (Applied Biosystems Inc., FosterCity, CA 94404, USA).

The following N-terminal amino acid sequences were derived:

α-peptide (30,000 molecular weight):

Thr lie Gly Asn Ser Ser Ser/Val Pro Ala Thr Ala Pro Thr -?- Val Ala Gly Leu Ser Ala Pro He

β peptide (60,000 molecular weight):

Ser Asn Ala Trp Thr Val Ala Gly Ser Arg Thr Ser Thr Gly -?- Pro He Leu -Ala Asn Asp Pro His Leu

These sequences .are identified and located in Sequence Identity Number 1. The α- peptide sequence identified by amino acid sequencing represents a fusion protein between the N-terminus of the lacZ α-peptide derived from the plasmid vector pTT1861 and an internal portion of the α-peptide sequence derived from the P. vesicularis gac gene. The precise linkage of the lacZ-gac fusion is detailed in Figure 3. The linkage brings the two peptide sequences together as a precise in- frame fusion at the Kpnl site of the DNA sequence. As this Kpnl site forms the boundary of the gac gene in the sequence, it follows that from the methods of construction outlined in Preparation 2(d) and 2(e), the construct pVS44 and all derivatives including pVS4461 and pTT1861 represent similar lacZ-gac fusions. Expression of amidohydrolase activity in all such constructs utilises a recombinant system, rather than the natural P. vesicularis gac promoter.

Preparation 3: Immobilisation of Amidohydrolase

(a) Amidohydrolase Purification

E. coli str.ain NM522/pVS44 was grown on TSB medium supplemented with lOμg/ml choramphenicol and solidified with 1.5% w/v agar. Growth was for 2 days at 27.5°C. Cell paste was recovered by scraping from the agar surface, resuspended and washed in lysis buffer (lOmM MOPS pH 8.0; ImM DTT; ImM benzamidine). About 65g cell paste was lysed in 350ml ice cold buffer by sonication (650W output, power setting 4, 5 passes through a continuous flow head at 35ml/min flow, sample

chilled on ice). The lysate was clarified by centrifugation (15,000 rpm, 30 min) to yield 3120mg soluble protein. This material was passed over a 300ml column of pre- equilibrated DE52 cellulose. Fractions eluting between 100ml-500ml after loading contained amidohydrolase activity in a total of 398mg protein. The protein was concentrated using a centrifugal membrane device (Centriprep 30 concentrator, Amicon).

(b) Enzyme Immobilisation

The concentrated extract (300mg protein, 12.54ml) was adjusted to lOOmM phosphate buffer pH 8.0: ImM DTT; ImM benzamidine, and mixed with 3.0g (dry weight) oxirane-linked polyacrylic matrix (Eupergjt C: Rohm Pharma GmbH, D-6100 Darmstadt Germany). Immobilisation, monitored by protein removal from the liquid phase, was complete by 116 hours. About 120mg protein became covalently linked to the support. Excess protein was removed by washing in the phosphate buffer.

(c) Immobilised Enzyme Activity

Immobilised amidohydrolase activity was monitored by allowing the enzyme catalysed reaction to occur using an automated control system set to detect and maintain the reaction pH at a set point (a pH stat system). A pH monitor was set such that a fall in pH from the set point triggered a peristaltic pump to allow regulated dosing of alkali titrant.

A reaction volume of 50ml contained 50mg/ml substrate (Ga -7ACA or Ceph C) in water maintained at 25°C. The reaction mix was adjusted to pH 8.0 using 2M NaOH as titrant before addition of about 3g dry weight (12g wet weight) immobilised enzyme to initiate the reaction. The suspension was stirred using an overhead paddle type stirrer. Reaction progress was monitored by titrant addition rate and sampling at intervals. The samples were assayed for 7ACA by the DMAB method as in preparation 1(a). By this method the following activities were measured:

The maximum conversion with cephalosporin C (22 hours) was 1.7%, with GL-7ACA as substrate a conversion of 9.0% was achieved.

SEQUENCE LISTINGS

SEQ ID NO: 1

SEQUENCE TYPE: Nucleotide with corresponding peptide

SEQUENCE LENGTH: 2391

STRANDEDNESS: Single

TOPOLOGY: Linear

MOLECULAR TYPE: Genomic DNA

ORIGINAL SOURCE ORGANISM: Pseudomonas vesicularis B965

EXPERIMENTAL SOURCE: Plasmid pVS42 mKcoli strain Bl 113

FEATURES: Sequence location indicated in Figure 1

PROPERTIES: Encodes polypeptide sequence oϊ P. vesicularis cephalosporin amidohydrolas enzyme activity

ATG GACGGCCCGCCGCCGAGGGG TT A GCC TGTGGCGG ACCCACGAC48 Met Asp Gly Pro Pro Pro Arg Gly Leu Ala Cys Gly Gly Thr His Asp

5 10 15

GAA AGGATAGCAATGCTC GACCGGCGCeTATTT CTGCTT GGTTCGGCG96 Glu Arg De Ala Met Leu Asp Arg Arg Leu Phe Leu Leu Gly Ser Ala 20 25 30

GCG ACCGTGCTGGTGTCT GCCGCCTGCCACGGCCAGTCGGTACCGGCG144 Ala Thr Val Leu Val Ser Ala Ala Cys His Gly Gin Ser Val Pro Ala 35 40 45

ACA GCACCGACCCGCGTGGCGGG CTT TCG GC A CCG ATC GAGATC ATC 192 Thr Ala Pro Thr Arg Val Ala Gly Leu Ser Ala Pro Be Glu De He 50 55 60

GAC GATCGCTGGGGCGTGCCGCAT ATC CGCGCGCAGACGAAGGCGGAT240 Asp Asp Arg Trp Gly Val Pro His De Arg Ala Gin Thr Lys Ala Asp 65 70 75 80

GCG TTT TTC GGACAGGGCTAT GTC GTGGCGCGCGACCGACTGTTC CAG288 Ala Phe Phe Gly Gin Gly Tyr Val Val Ala Arg Asp Arg Leu Phe Gin

85 90 95

ATC GACCTGGCGCATCGCCGCGAACTGGGCCGGATGGCGGAAGCGTTC336

He Asp Leu Ala His Arg Arg Glu Leu Gly Arg Met Ala Glu Ala Phe 100 105 110

GGG CCGGATTTT GCCAAGCATGATGCCGTCGCCCGACTGTTC CATTAT 384 Gly Pro Asp Phe Ala Lys His Asp Ala Val Ala Arg Leu Phe His Tyr 115 120 125

CGC GGCGACCTGGACGCC GAGCTGGCGCGC GTACCC AAGGAAGTT CGC432 Arg Gly Asp Leu Asp Ala Glu Leu Ala Arg Val Pro Lys Glu Val Arg 130 135 140

GAC TGC GTGGCC GGATAT GTC GCAGGC ATC AAT GCGCGGATC GCC GAG480 Asp Cys Val Ala Gly Tyr Val Ala Gly De .Asn Ala Arg De Ala Glu 145 150 155 160

GTC GAGAAGGACCCGAGCCTATTGCCGCCGGAATAT CGCATC CTGGG 528 Val Glu Lys Asp Pro Ser Leu Leu Pro Pro Glu Tyr Arg De Leu Gly

165 170 175

GTG ACGCCGCTGCGCTGGGACATCCGCGATCTGGTGCGCGCGCGGGGC576 Val Thr Pro Leu Arg Trp Asp De Arg Asp Leu Val Arg Ala Arg Gly 180 185 190

AGC TCA ATC GGC AAT GCC GAC GACGAGATC CGC CGT GCGAAACTG GCC 624 Ser Ser De Gly Asn Ala Asp Asp Glu De Arg Arg Ala Lys Leu Ala 195 200 205

GCG CTC GGCATGCTGGAGCTGGACGCGGTG ATC GCGCCGCTGCGGCCG672 Ala Leu Gly Met Leu Glu Leu Asp Ala Val De Ala Pro Leu Arg Pro

210 215 220

ACG TGGAAGCTGGCCGTGCCGGA GG CTGGACCCGTCAAAGGTGAGT720 Thr Trp Lys Leu Ala Val Pro Glu Gly Leu Asp Pro Ser Lys Val Ser 225 230 235 240

GAT GCGGAT CTG GGCGTGCTT CAGCTG GGACGGCTG CCGTTC GGACCC 768 Asp Ala Asp Leu Gly Val Leu Gin Leu Gly Arg Leu Pro Phe Gly Pro

245 250 255

GAC ACGCCGACGCGTGAGCCAGAAGAGGATCTGGACCGCGCGCAGGCG816 Asp Thr Pro Thr Arg Glu Pro Glu Glu Asp Leu Asp Arg Ala Gin Ala 260 265 270

GGG TCG AACGCCTGGACC GTC GCT GGCAGCCGC ACC AGC ACT GGCCGC 864 Gly Ser Asn Ala Trp Thr Val Ala Gly Ser Arg Thr Ser Thr Gly Arg

275 280 285

CCC ATT CTG GCC AAT GAT CCG CAT CTGGG ATC GGCGG TTC GG CCG912 Pro De Leu Ala Asn Asp Pro His Leu GHy De Gly Gly Phe Gly Pro 290 295 300

CGA CAT GTGGCGCAT CTG ACC GCGCCGGGTCTC GACGTG ATT GGCGGT960 Arg His Val Ala His Leu Thr Ala Pro Gly Leu Asp Val De Gly Gly 305 310 315 320

GGC GCGCCT GGACTGCCCGGCATCATGCAGGG CAT ACC GAC CGT TTC 1008 Gly Ala Pro Gly Leu Pro Gly De Met Gin Gly His Thr Asp Arg Phe

325 330 335

GCC TTT GGCCGGACC AATTTC CAT ATC GACCAGCAGGATCTGTTC GTC 1056 Ala Phe Gly Arg Thr Asn Phe His De Asp Gin Gin Asp Leu Phe Val 340 345 350

CTG GAGCTG GAT CCC AAC GAT CCC GAGCGGTAC CGC C AC GAC GGC GGC 1104 Leu Glu Leu Asp Pro Asn Asp Pro Glu Arg Tyr Arg His Asp Gly Gly 355 360 365

TGG AAGCGGTTC GAGCGGGTGGA GAGACGATC CCGGTCAAGGACGGT1152 Trp Lys Arg Phe Glu Arg Val Glu Glu Thr De Pro Val Lys Asp Gly 370 375 380

CCG CCGCAG AAGGTGGTC CTG CGC TAC GCGGTA CAGGGCCCGGTG ATC 1200 Pro Pro Gin Lys Val Val Leu Arg Tyr Ala Val Gin Gly Pro Val De 385 390 395 400

ATG CACGATCCGGCGGCGCGCCGGGCGACGGTGCTC GG TCG ATC GG 1248 Met His Asp Pro Ala Ala Arg Arg Ala Thr Val Leu Gly Ser De Gly

405 410 415

ATG CAGCCC GGCGG TTC GGATCGTTC GCGATGGTGGCGATC AACCTG 1296 Met Gin Pro Gly Gly Phe Gly Ser Phe Ala Met Val Ala De Asn Leu 420 425 430

TCG CGCGACTGGAACAGCCTG AAGGAAGCGTTC AAGCTGCACCCGTCG 1344 Ser Arg Asp Trp Asn Ser Leu Lys Glu Ala Phe Lys Leu His Pro Ser 435 440 445

CCC ACCAACCTGCATTAT GCCGACGTTGACGG AACCACGGCTGGCAA1 92 Pro Thr Asn Leu His Tyr Ala Asp Val Asp Gly Asn His Gly Trp Gin 450 455 460

GTG ATC GGCTTC GTT CCGCAGCGCAAGAAGGGCGACGG CTG ATGCCG 1440 Val De Gly Phe Val Pro Gin Arg Lys Lys Gly Asp Gly Leu Met Pro 465 470 475 480

GTG CCGGGCGACGG CGCTACGACTGGAACAGCTAT CGCGATTTC CGC 1488

Val Pro Gly Asp Gly Arg Tyr Asp Trp Asn Ser Tyr Arg Asp Phe Arg

485 490 495

GTG CTGCCGAGCGAGTTC AACCCGTCC AAGGGCTGGTTC GCATCT GCC 1536 Val Leu Pro Ser Glu Phe Asn Pro Ser Lys Gly Trp Phe Ala Ser Ala 500 505 510

AAC CAGAACAACCTGCCGGCGAATTGGCCGCGCGATCGCATT CCGGCG1584 Asn Gin Asn Asn Leu Pro Ala Asn Trp Pro Arg Asp Arg De Pro Ala 515 520 525

TTC TCG TTC CGC GACCCT TAT CGT TAC GAGCGC GTC GCC GAGGTGCTG 1632 Phe Ser Phe Arg Asp Pro Tyr Arg Tyr Glu Arg Val Ala Glu Val Leu 530 535 540

GCA TCGCAGCCGCGCCATTCGGTGGCGGACAGCGTGGCATTGCAGTTC 1680 Ala Ser Gin Pro Arg His Ser Val Ala Asp Ser Val Ala Leu Gin Phe 545 550 ^' 555 560

GAC ACGCTGTCGACGCCGGCCAAGCAGTTC CTG GCGTTG CTA CCC AAG 1728 Asp Thr Leu Ser Thr Pro Ala Lys Gin Phe Leu Ala Leu Leu Pro Lys

565 ^• 570 575

CAG CCGTCC GCGGGTGCGGCCCCGGCGGTGAAGATGCTGTCGGG TGG1776 Gin Pro Ser Ala Gly Ala Ala Pro Ala Val Lys Met Leu Ser Gly Trp 580 585 590

GAC GCGAAGCTGGACAAGGATAGCGGCGCGGCGGCGCTG TAC GAGATC 1824 Asp Ala Lys Leu Asp Lys Asp Ser Gly Ala Ala Ala Leu Tyr Glu De 595 600 605

_*

GTG TGGCGT GACCTGGGCAAGCGGATGCTGGCGGCC ATC GTGCCGGAG1872 Val Trp Arg Asp Leu Gly Lys Arg Met Leu Ala Ala De Val Pro Glu 610 615 620

CAG GCGAAGGAGCTG GTGGACGA ATC GCACCATCC GAGCTG CTG CGC 1920 Gin Ala Ly Glu Leu Val Asp Glu De Ala Pro Ser Glu Leu Leu Arg 625 630 635 640

CGT GCGGCAAGCCGGCCGGCGATGGTGGACGAGGCGCTGGCGAGCGGC 1968 Arg Ala Ala Ser Arg Pro Ala Met Val Asp Glu Ala Leu Ala Ser Gly

645 650 655

TGG GCCGAGGCGCAGCGACTGATGGGCAGCGATCCATCAGCCTGGCGC2016 Trp Ala Glu Ala Gin Arg Leu Met Gly Ser Asp Pro Ser Ala Trp Arg 660 665 670

TGG GGC ACGTTG CAC CAGGTGCGGATC GCGCAT CCGCTG TCG AGC ATC 2064 Trp Gly Thr Leu His Gin Val Arg De Ala His Pro Leu Ser Ser De 675 680 685

CCG GCC ATC GCCGCGGCGTTC CCC CCG ATT GAGGGCGAGGGATCG GGC 2112 Pro Ala De Ala Ala Ala Phe Pro Pro De Glu Gly Glu Gly Ser Gly 690 695 700

GGC GAC AGCTAT ACC GTC ATGGCGCGT TGGCTG GGC AAT GGCCCGGGC2160 Gly Asp Ser Tyr Thr Val Met Ala Arg Trp Leu Gly Asn Gly Pro Gly 705 710 715 720

TGG CGC ACC GGCGGCGG GCGAGCTAC CTGCATGTGATC GACGTGGGC2208 Trp Arg Thr Gly Gly Gly Ala Ser Tyr Leu His Val De Asp Val Gly

725 730 735

GAC TGGGACAAATCGGTGATGCTGAACCTGCCGGG CAGTCGAACGAC2256 Asp Trp Asp Lys Ser Val Met Leu Asn Leu Pro Gly Gin Ser Asn Asp 740 745 750

CCG CGC TCG CCG CAT TAC CGC GAT C AATAT GCGCCG TGG ATC AAGGGC2304 Pro Arg Ser Pro His Tyr Arg Asp Gin Tyr Ala Pro Trp De Lys Gly 755 760 765

GAA ATGCAGCCGATGCCGTTC AGCCGCGCCGCGGTGGACGGTGGCGTC2352 Glu Met Gin Pro Met Pro Phe Ser Arg Ala Ala Val Asp Gly Gly Val 770 775 780

AGC CGTTCGACGCTGACGCCGCAGGG AAGGGCAAGTGA Ser Arg Ser Thr Leu Thr Pro Gin Gly Lys Gly Lys *** 785 790 795

SEQ ID NO: 2

SEQUENCE TYPE: Nucleotide with corresponding peptide

SEQUENCE LENGTH: 93

STRANDEDNESS: Single

TOPOLOGY: Linear

MOLECULAR TYPE: Genomic DNA

ORIGINAL SOURCE ORGANISM: Pseudomonas vesicularis B965 and E.coli

EXPERIMENTAL SOURCE: Plasmid pVS42 derived from Kcoli strain Bl 113

FEATURES: Sequence features N-terminus lacZ α-peptide, T7 polymerase promoter part of synthetic pUClδ multiple cloning she region and part of gac gene sequence

PROPERTIES: Encodes fusion polypeptide sequence of lacZ α-peptide wit cephalosporin amidohydrolase enzyme α-peptide

ACAGCT ATG ACC ATG ATT ACG AAT TTA ATA CGACTC ACT ATA GG AAT 48 Met Thr Met De Thr Asn Leu De Arg Leu Thr De Gly Asn

5 10

TCGAGCTCGGTACCGGCGACAGCACCGACCCGCGTGGCGGG

Ser Ser Ser Val Pro Ala Thr Ala Pro Thr Arg Val Ala Gly 15 20 25