WO2002072806A2 - Variant glutaryl amidase (cephalosporin acylase) and uses thereof - Google Patents

Abstract

Variants of glutaryl amidase are described and a process for the preparation of 7-aminocephalosporanic acid using the same.

Description

MUTANT GLUTARYL AMIDASE AND USES THEREOF

Field of the invention

The invention relates to mutated glutaryl amidase and its use in a process for the preparation of 7-aminocephalosporanic acid.

Background to the invention

7-amino-cephalosporanic acid (7-ACA) is the starting product for the semisynthetic preparation of various cephalosporin antibiotics. For large-scale production from fermentatively prepared cephalosporin C, either a peptide-chemical process or one of the enzymatic processes is used, the latter being preferred for economic and above all environmental reasons.

To obtain 7-ACA from cephalosporin C by an enzymatic route, there are three technically feasible processes. In the first process, cephalosporin C is firstly decarboxylated by chemical oxidation to form glutaryl-7-ACA. This substrate is then further processed to 7-ACA using the enzyme glutaryl amidase. This process is associated with a high level of environmental pollution and high costs.

In the second process, the oxidative decarboxylation is carried out enzymatically with the help of D-amino acid oxidase. Glutaryl-7-ACA is in turn then converted to 7-ACA, again with the enzyme glutaryl amidase. This process has a lower environmental impact and is used industrially.

A third and more commercially and environmentally attractive process would be to convert cephalosporin C directly to 7-ACA enzymatically. However, although an enzyme, cephalosporin C acylase has been described that can carry out this conversion (Ishii, Y. et al, 1995, Eur. J. Biochemistry 230: 773-778), the specific activities of this enzyme for industrial use are very poor. Attempts have been made to improve the enzymatic activity through mutagenesis of the enzyme. Despite this, an industrial process based on this single-enzyme process has not been described in the literature.

The crystal structure of a cephalosporin acylase at a 2.0A resolution has been determined by Kim et al, 2000, Structure 8: 1059-1068. Kim et al. deduced the binding cleft through a sequence homology comparison and superimposition on the structure of a related enzyme, penicillin-G-acylase (Duggleby et al, 1995, Nature 373: 264-268). The authors speculate that alteration of a number of residues in the substrate binding pocket may allow the design of an enzyme with an improved conversion rate of cephalosporin C to 7-ACA. However, no complex of cephalosporin C and the enzyme is described.

Summary of the invention The object of the invention is to improve the economic efficiency of the enzymatic

7-ACA process requiring two enzymes. This object is achieved by using a mutated glutaryl amidase which can cleave cephalosporin C directly to 7-ACA. Through the use of the mutated enzyme, it is possible to arrive at a single-enzyme process which is simpler in terms of process engineering and significantly cheaper. More specifically, the crystal structure of glutaryl amidase (GA) has been obtained and used to model the position of the substrate glutaryl-7-ACA in the substrate binding pocket of the GA. This has allowed a more accurate determination of the key residues within the subsfrate pocket which should be modified to enhance cephalosporin C binding and which amino acids should be included at these positions. The mutations in the amino acid sequence required for changing the substrate specificity of the enzyme are based on a precise analysis of the amino acids which participate in the glutaryl-7-ACA binding using a 3D structure of the enzyme combined with modelling-in of the substrate glutaryl-7-ACA and a comparison with the desired substrate cephalosporin C. For example, it has been possible to determine that a modified GA with improved cephalosporin C binding and hence cleavage of cephalosporin C directly to 7-ACA by GA should include in the substrate binding pocket a histidine or glutamate residue for binding the -amino adipyl moiety of a cephalosporin C substrate molecule and only one positively charged residue for binding the carboxylate group of the cephalosporin C. Accordingly, the present invention provides a non-naturally occurring variant of a glutaryl amidase which catalyses conversion of cephalosporin C (CPC) to 7-aminocephalosporanic acid (7-ACA), which variant comprises a histidine or glutamate in its substrate binding pocket which binds the α-amino adipyl moiety of a CPC substrate molecule. A single positively charged residue for binding the carboxylate group of cephalosporin C may be provided by the wild type ArgB57 or provided by mutating ArgB57 to an uncharged amino acid residue, such as histidine, and mutating another residue in the substrate binding pocket to arginine, such as LeuB24 or GlnB50. In a preferred embodiment, the variant GA comprises a histidine residue at position B177 and/or a glutamate residue at position A150. It is also preferred, in addition, for ThrB176 to be replaced with an aspartate residue.

In a further embodiment, additional mutations are made that stabilise or enlarge the pocket volume. Preferred examples include replacement of TyrAl 50 with alanine, GlnB50 with alanine or valine and/or TyrB153 with leucine.

All amino acid numbering is with respect to the A and B chains shown as SEQ ID Nos 1 and 2. However, the numbering applies to equivalent positions in other GA enzyme amino acid sequences as determined for example by sequence alignment. Preferably the variant enzyme is obtained by modification of a Pseudomonas GA, such as the Pseudomonas GA described in EP-A-504798, the nucleotide sequence of which is present in plasmid pCM145 deposited as accession no. DSM 6409.

The present invention also provides a nucleotide sequence encoding a variant enzyme of the invention and vectors and host cells comprising the same. The present invention further provides a method for producing a variant enzyme of the invention which method comprises culturing a host cell of the invention under suitable conditions which allow for expression of the variant GA enzyme.

In a highly preferred embodiment, purified enzyme can be obtained, without a chromatographic purification step, from a solution obtained by lysing/disrupting the host cells expressing the variant GA, by a method comprising (i) adjusting the potassium phosphate concentration of the solution to from 1 to 1.5M at a pH of from 7 to 8; (ii) concentrating the solution; and (iii) and allowing the variant GA protein to crystallise out from the concentrated solution at a temperature of about 4°C or less.

The mutant GA enzymes of the present invention can be used to produce 7-ACA directly from cephalosporin C. Accordingly, the present invention also provides a method for preparing 7-ACA from cephalosporin C which comprises reacting cephalosporin C with a variant GA enzyme of the invention under suitable conditions such that the variant GA enzyme cleaves the cephalosporin C directly to form 7-ACA.

In a preferred embodiment, the enzymatic reactions are carried out in a pH range from pH 7.5 to 8.5, preferably at pH 8.2, as a result of which autocatalytic decomposition of cephalosporin C is minimised. The concentration of cephalosporin C in the reaction mixture typically lies in the range from 20-200 mM, 50 to 120 mM being preferred. In the reactions, an enzyme concentration of 0.25 to 10 mg/ml is used in the reaction mixture, preferably 2 mg/ml. The temperature range can typically be varied between 15 and 45°C.

To further reduce costs, the mutated enzyme can also be used carrier-fixed in the reactions, the carrier materials used in industrial enzymology being able to be used (K. Buchholz and N. Kasche, Biokatalysatoren und Enzymtechnologie, NCH Nerlag, Weinheim, 1^st edition 1997).

The present invention further provides 7-aminocephalosporanic acid produced by the method of the invention

In a further aspect, the present invention provides a method for modifying a glutaryl amidase enzyme to enhance the cephalosporin C acylase activity of said enzyme which method comprises introducing a non-naturally occurring histidine or glutamate amino acid residue in the substrate binding pocket of said enzyme which histidine or glutamate binds the α-amino adipyl moiety of a CPC substrate molecule.

In one embodiment, the method further comprises substituting or deleting a naturally occurring positively charged amino acid residue in said substrate binding pocket and introducing a positively charged amino acid residue which binds the carboxylate group of the CPC.

The present invention also provides a modified glutaryl amidase enzyme produced by said method.

Detailed description of the invention

Although the enzyme glutaryl amidase has a high specific activity with respect to glutaryl-7-ACA, it cannot convert cephalosporin C directly to 7-ACA. Therefore, it has been used only as a glutaryl-7-ACA-cleavage enzyme in industrial processes. The good operational stability and the high conversion rates therefore make this enzyme particularly suitable for industrial reactions.

To make the enzyme glutaryl amidase available for use in a single-enzyme process, mutated enzymes of glutaryl amidase are described in the present application which recognise cephalosporin C as substrate and cleave it directly to 7-ACA. Starting from the crystal structure of the enzyme, this aim can be achieved by protein engineering without using evolutionary strategies.

The enzyme glutaryl amidase consists of two sub-units of different sizes which are formed from a precursor protein comprising 720 amino acids through processing steps talcing place in the periplasm of the expression organism and associate there to form a functional enzyme. The small sub-unit (A chain) comprises 160 or 161 amino acids in the functional enzyme, the large sub-unit (B chain) 522 amino acids. The protein sequences of the A and B chains are shown in SEQ ID NOs:l and 2, respectively. Numbering of residues throughout is with reference to the amino acid sequences shown as SEQ ID

NOs:l and 2. However, it will be understood that where other GA enzymes are used, the numbering applies to the functionally corresponding residues in those sequences which have not have exactly the same numbering as the protein sequences shown in SEQ ID NOs:l and 2. As discussed above, the mutations in the GA amino acid sequence required for changing the substrate specificity of the enzyme are based on a precise analysis of the amino acids which participate in the glutaryl-7-ACA binding.

According to the invention, glutaryl-7-ACA has been modelled into the putative binding cleft of the glutaryl amidase and compared with the orientation of the phenylacetic acid radical in the binding cleft of the penicillin-G-acylase. Taking the cleavage mechanism proposed by Duggleby et al, supra, for serine proteases, as a basis, the substrate glutaryl-7-ACA was then modelled such that SerBl can engage with the carbonyl atom at the glutaryl radical, and NalB70 and AsnB244 can form the oxyanion hole. The carboxylate side chain of the glutaryl radical was modelled such that it can interact with ArgB57. The same orientation was used for the side chain as is orientated for the phenylacetic acid in penicillin in the enzyme penicillin-G-acylase. The side chain binding cleft is hydrophobic in penicillin-G-acylase, whereas the glutaryl amidase is hydrophilic with ArgB57 at this point, and thus can interact with the glutaryl side-chain carboxylate. If glutaryl-7-ACA is now replaced by the spatial structure of cephalosporin C in the binding cleft, unfavourable interactions are obtained which explain why glutaryl amidase cannot catalyse the cleavage of CPC. The superimposition of the sequence of cephalosporin C acylase by that of the glutaryl amidase leads to the conclusion that improved binding and catalytic conversion of CPC by GA can be achieved by including in the substrate binding pocket a histidine or glutamate residue for binding the α-amino adipyl moiety of cephalosporin C and exactly one positively charged residue for binding the carboxylate group of cephalosporin C. For example, with reference to the amino acid sequence shown in SEQ ID NO: 1, PheB177 is replaced with a histidine residue and/or TyrAl 50 is replaced with a glutamate. Preferably, in addition, where a PheB177— »His mutation is used, residue ThrB176 is replaced with aspartate. The additional mutation Thrl76 Asp should have a favourable effect on the basic state of the histidine and the stabilising of the substrate binding.

With respect to the positively charged residue which binds the carboxylate group of cephalosporin C, the wild type ArgB57 can be used. However, as alternatives, the double mutation ArgB57His and LeuB24Arg or the double mutation ArgB57His, GlnB50Arg can be used. Other mutations which may be used to enlarge the binding cleft to provide a better fit for the larger substrate cephalosporin C may advantageously be included, such as TyrAl 50Ala and/or TyrB 153Leu.

According to the invention, the following mutations in the glutaryl amidase lead to enzymes with changed substrate binding which in turn effect direct catalysis of the cephalosporin cleavage to 7-ACA. Preferred embodiments of the present invention comprise the following combination of mutations:

(1) The combination of Phe B177 His and Thr B176 Asp to promote binding of the α-amino adipyl moiety of CPC, together with binding of the carboxylate group either by the natural Arg B57, LeuB24 Arg or Gin B50 Arg. In the latter two cases, Arg B57 must be mutated to an uncharged residue, preferably His. Further mutations may be made to enlarge the binding pocket. Preferred combinations of mutations are selected from the group consisting of:

1. PheB 177His, ThrB 176 Asp 2. PheB 177His, ThrB 176Asp, TyrAl 50Ala

3. PheB177His, ThrB176Asp, TyrA150Ala, TyrB33Ser

4. PheB 177His, ThrB 176Asp, LeuB24Arg, ArgB57His

5. PheB177His, ThrB176Asp, LeuB24Arg, ArgB57His, TyrA150Ala

6. PheB177His, ThrB176Asp, LeuB24Arg, ArgB57His, TyrA150Ala, TyrB33Ser 7. PheB 177His, ThrB 176Asp, GlnB50Arg, ArgB57His

8. PheB 177His, ThrB 176Asp, GlnB50Arg, ArgB57His, TyrAl 50Ala

9. PheB177His, ThrB176Asp, GlnB50Arg, ArgB57His, TyrA150Ala, TyrB33Ser (2) The binding of the α-amino adipyl moiety of CPC through Tyr A150 Glu, together with binding of the carboxylate group either by the natural Arg B57, LeuB24 Arg or Gin B50 Arg. In the latter two cases, Arg B57 must be mutated to an uncharged residue, preferably His. Further mutations may be made to enlarge the binding pocket. Preferred combinations of mutations are selected from the group consisting of:

1 TyrAl 50Glu

2 TyrA150Glu, PheB177His 3 TyrAl 50Glu, LeuB24Arg, ArgB57His 4 TyrA150Glu, LeuB24Arg, ArgB57His, PheB177His 5 TyrAl 50Glu, LeuB24Arg, ArgB57His, TyrB33Ser 6 TyrA150Glu, PheB177His, TyrB33Ser 7, TyrAl 50Glu, GlnB50Arg, ArgB57His 8 TyrAl 50Glu, GlnB50Arg, ArgB57His, PheB177His

(3) The binding of the α-amino adipyl moiety of CPC through Phe B177 His, together with binding of the carboxylate group by LeuB24 Arg. Arg B57 must be mutated to an uncharged residue, preferably His. Further mutations may be made to enlarge the binding pocket. Preferred combinations of mutations are selected from the group consisting of:

1. PheB 177His, LeuB24Arg, ArgB57His

2. PheB177His, LeuB24Arg, ArgB57His, GlnB50Nal

3. PheB 177His, LeuB24Arg, ArgB57His, GlnB50 Ala 4. PheB177His, LeuB24Arg, ArgB57His, GlnB50Nal, NalB70His

5. PheB 177His, LeuB24Arg, ArgB57His, GlnB50Ala, NalB70His

6. PheB177His, LeuB24Arg, ArgB57His, GlnB50Nal, NalB70His, TyrB153Leu

7. PheB 177His, LeuB24Arg, ArgB57His, GlnB50Ala, NalB70His, TyrB 153Leu

8. PheB177His, LeuB24Arg, ArgB57His, TyrB33Ser 9. PheB177His, LeuB24Arg, ArgB57His, TyrB33Ser, NalB70His

The invention thus provides, in a preferred aspect, a mutated glutaryl amidase which comprises a combination of mutations selected from the group consisting of: 1 PheB177His, LeuB24Arg, ArgB57His

2 PheB177His, LeuB24Arg, ArgB57His, GlnB50Nal

3. PheB177His, LeuB24Arg, ArgB57His, GlnB50Ala

4 PheB177His, LeuB24Arg, ArgB57His, GlnB50Nal, ValB70His

5, PheB177His, LeuB24Arg, ArgB57His, GlnB50Ala, ValB70His

6 PheB177His, LeuB24Arg, ArgB57His, GlnB50Nal, NalB70His, TyrB153Leu

7 PheB177His, LeuB24Arg, ArgB57His, GlnB50Ala, ValB70His, TyrB153Leu

8 PheB177His, ThrB 176 Asp

9 PheB177His, ThrB176Asp, TyrA150Ala

10 PheB177His, ThrB176Asp, LeuB24Arg, ArgB57His

11 PheB177His, ThrB176Asp, LeuB24Arg, ArgB57His, TyrA150Ala

12 PheB177His, ThrB176Asp, GlnB50Arg, ArgB57His

13 PheB177His, ThrB176Asp, GlnB50Arg, ArgB57His, TyrAl 50Ala

14 TyrAl 50Glu

15 TyrAl 50Glu, PheB 177His

16 TyrAl 50Glu, LeuB24Arg, ArgB57His

17 TyrAl 50Glu, LeuB24Arg, ArgB57His, PheB177His

18 TyrAl 50Glu, GlnB50Arg, ArgB57His

19 TyrAl 50Glu, GlnB50Arg, ArgB57His, PheB177His

20 PheB177His, TlιrB176Asp, TyrA150Ala, TyrB33Ser

21 PheB177His, ThrB176Asp, LeuB24Arg, ArgB57His, TyrA150Ala, TyrB33Ser

22 PheB177His, ThrB176Asp, GlnB50Arg, ArgB57His, TyrA150Ala, TyrB33Ser

23 TyrA150Glu, PheB177His, TyrB33Ser

24 TyrAl 50Glu, LeuB24Arg, ArgB57His, TyrB33Ser

25 PheB177His, LeuB24Arg, ArgB57His, TyrB33Ser

26 PheB177His, LeuB24Arg, ArgB57His, TyrB33Ser, ValB70His

Mutations in the glutaryl amidase gene sequence may be introduced, for example, using standard site directed mutagenesis techniques, such as those described in Ausubel, F. et al, Current Protocols in Molecular Biology, Greene Publishing Associates and Wiley Interscience, 1994-1998 and Supplements. The amino acid sequence of a Pseudomonas glutaryl amidase is shown in SEQ ID Nos 1 and 2. The nucleotide sequence of this glutaryl amidase gene is present in plasmid T415 described in US Patent No 5766881 and EP-A-708180. The GA sequences present in pT415 are derived from pCM145 which has been deposited under accession no. DSM 6409.

Mutations in the glutaryl amidase gene present in plasmid T415 may conveniently be introduced using cloned part fragments from the glutaryl amidase gene, which are flanked on both sides by unique sequences in the T 415 expression plasmid so that a back transfer of the mutated fragment into the vector can be carried out quickly and without difficulty. The location-specific mutations for the replacement of a coding codon in the gene fragments can be carried out in each case with the help of two mutagenic DNA primers and verified before back-cloning into the expression vector by DNA sequencing. Thus, the present invention also provides polynucleotides encoding a variant GA enzyme of the invention. Polynucleotides of the invention may comprise DNA or RNA. They may also be polynucleotides which include within them synthetic or modified nucleotides. A number of different types of modification to oligonucleotides are known in the art. These include methylphosphonate and phosphorothioate backbones, addition of acridine or polylysine chains at the 3' and/or 5' ends of the molecule. For the purposes of the present invention, it is to be understood that the polynucleotides described herein may be modified by any method available in the art. It will be understood by a skilled person that numerous different polynucleotides can encode the same polypeptide as a result of the degeneracy of the genetic code.

Polynucleotides of the invention can be incorporated into a recombinant replicable vector. The vector may be used to replicate the nucleic acid in a compatible host cell. Preferably, a polynucleotide of the invention in a vector is operably linked to a regulatory sequence that is capable of providing for the expression of the coding sequence by the host cell, i.e. the vector is an expression vector. The term "operably linked" refers to a juxtaposition wherein the components described are in a relationship permitting them to function in their intended manner. A regulatory sequence "operably linked" to a coding sequence is ligated in such a way that expression of the coding sequence is achieved under condition compatible with the control sequences.

The vectors are typically plasmid vectors provided with an origin of replication, optionally a promoter for the expression of the said polynucleotide and optionally a regulator of the promoter. The vectors may contain one or more selectable marker genes, for example a chloramphenicol resistance gene in the case of a bacterial plasmid. Vectors may be used in vitro, for example for the production of RNA or used to transfect or transform a host cell. Promoters/enhancers and other expression regulation signals may be selected to be compatible with the host cell for which the expression vector is designed. For example, prokaryotic promoters may be used, in particular those suitable for use in E. coli strains. Particularly preferred vectors that give high levels of expression are described in US Patent No 576688 l/ ΕP-A-708180 and US Patent No. 5830743/ EP-A-504798. Such vectors may be transformed or transfected into a suitable host cell using standard techniques to provide for expression of a polypeptide of the invention. This process may comprise culturing a host cell transformed with an expression vector as described above under conditions to provide for expression by the vector of a coding sequence encoding the polypeptides, and optionally recovering the expressed polypeptides.

Thus, production and purification of recombinant GA protein can be performed using suitable techniques known in the art, such as those described in US Patent No 5766881/ EP-A-708180 and US Patent No. 5830743/ EP-A-504798, the contents of which are incorporated by reference. For example, vector T415 described in US Patent No 5766881/ EP-A-708180 allows high constitutive expression of functional enzymes and is therefore useful for the expression of the mutated enzymes. The growth of the recombinant E. coli clones using this vector system is typically carried out at 28°C which has been shown previously to increase yields of GA expressed using the T415 vector. However, expression can also take place via inductive systems as described for example US Patent No. 5830743/ EP-A-504798. Moreover, intracellular expression as described for example by Lee and Park, 1998, J. Bacteriol. 180, 4576-4582 (1998) may be used.

A further aspect of the invention relates to a process is which the simple, swift and cheap preparation of large quantities of enzyme through crystallisation without using expensive chromatographic purification processes. According to this aspect of the invention, the supernatant obtained by lysing cultured cells expressing the enzyme and removing cell debris by centrifugation is used directly for crystallisation of pure enzyme in the presence of high-molar salt buffers, preferably 0.8-1.5 M potassium phosphate buffer. The pH range can typically be varied from pH 7.5-pH 8.2.

By way of an example, recombinant E. coli clones expressing variant GA of the invention is typically cultured to 10 litre scale according to the method described in US Patent No 5766881/ ΕP-A-708180. The cells are initially concentrated by centrifugation, taken up in small volume with 100 mM phosphate buffer, pH 7.2, and disrupted by means of a French press. To separate off the cell debris, the suspension is centrifuged at 10,000 rpm for 15 mins, and the supernatant removed.

The enzyme-containing supernatant solution is then is rebuffered in an Amicon ultrafiltration cell or by diafiltration against 1-1.5 M potassium phosphate buffer pH 7.8 at 4°C with simultaneous concentration. If the concentrated solution is left to stand in cold conditions, the enzyme crystallises out. The crystals may be enriched by gentle centrifugation and washed with the above-mentioned buffer.

The variant GA of the invention may be used for the conversion of cephalosporin to 7-ACA, particularly on an industrial scale.

It is preferred that the enzymatic reactions are carried out in a pH range from pH 7.5 to 8.5, preferably at pH 8.2, as a result of which autocatalytic decomposition of cephalosporin C is minimised. The concentration of cephalosporin C in the reaction mixture typically lies in the range from 20 to 200 mM, 50 to 120 mM being preferred, such as 75 mM. In the reactions, an enzyme concentration of 0.25 to 10 mg/ml is typically used in the reaction mixture, preferably 2 mg/ml. The temperature range can typically be varied between 15 and 45°C, for example 37°C. A reaction time of 1 hour is typical with the conversion to 7-ACA being monitored by, for example, HPLC.

The present invention will now be described with reference to the following Examples which are non-limiting and illustrative only.

EXAMPLES

Example 1 - Introduction of LeuB24Arg, PheB177His mutations

DNA plasmid T415 described in US Patent No 5766881 and EP-A-708180 is completely digested with the restriction enzymes BamHl and Hindlll, and the approx. 1900 bp and 4800 bp fragments are isolated after agarose gel electrophoresis and electroelution. First of all, the approx. 1900 bp fragment is ligated into the vector pUC19 which was cleaved with BamHl and Hindlll (New England Biolabs GmbH, Briiningstr. 50, 65926 Frankfurt).

After transformation into the E. coli MC 1061 strain, the plasmid DNA is isolated and subjected to a site-directed mutagenesis mediated by mutagenic primers. The LeuB24Arg mutation is carried out using primer pair no. 1 and confirmed after the analysis of the isolated plasmid DNA by DNA sequencing. This DNA mutagenesis is followed by the introduction of the PheB177His mutation using mutagenesis primer pair no. 2. The presence of both mutations in the BamHl -Hindlll fragment is confirmed after isolation of the plasmid DNA by DNA sequencing of the complete fragment. From the DNA of the vector which contains the fragment mutated at the two described points, the approx. 1900 bp fragment is released using restriction enzymes BamHl and Hindlll, separated by gel electrophoresis and isolated after electroelution. The isolated fragment is ligated into the 4800 bp BamHl -Hindlll cleaved fragment described above, and the ligation mixture transformed into E. coli W3110. Recombinant clones are selected by their ability to grow in the presence of 12 μg/ml chloramphenicol and to contain the desired mutations.

Primer pair 1 :

5 ' -CAGAACCCGCACCGCTCCTGGACGACG-3 ' (SEQ ID NO : 3 )

3 ' -GTCTTGGGCGTCGCGAGGACCTGCTGC-5 ' (SEQ ID NO : 4) Primer pair 2:

5 ' - CAGGTGCCGACCCACAACATCGTCTAC - 3 ' (SEQ ID NO : 5 ) 3 ' -GTCCACGGCTGGGTGTTGTAGCAGATG- 5 ' ( SEQ ID NO : δ )

Example 2 - Introduction of LeuB24Arg, PheB177His, ArgB57His, GlnB50 (Val or Ala) mutations

Starting from the approx. 1900 bp BamHl -Hindlll fragment mutated and isolated at the positions LeuB24Arg and PheB177His, the additional mutations ArgB57His and GlnB50Nal and GlnB50Ala respectively are introduced analogously to Example 1 into the fragment already mutated at the above-specified two positions. For this purpose, the mutagenic primer pairs no. 3 and no. 4 and no. 3 and no. 5, respectively, are successively used. Recombinant E. coli W3110 clones which have been isolated after cloning into the 5290 bp vector fragment of the two respective mutagenised fragments, verified by DNA sequencing, are selected because of their ability to grow on 12 μg/ml of chloramphenicol and their ability to convert cephalosporin C into 7-aminocephalosporanic acid tested as described in example 26.

Primer pair 3:

5 ' -CTGCCGGTCATCCACTTCGCCTTCACC-3 ' (SΕQ ID NO: 7) 3 ' -GACGGCCAGTAGGTGAAGCGGAAGTGG-5' (SΕQ ID NO: 8)

Primer pair 4: 5 ' -TATGGCGCGACCGTGATCGGCCTGCCG-3 ' (SΕQ ID NO: 9) 3 ^■ -ATACCGCGCTGGCACTAGCCGGACGGC-5' (SΕQ ID NO: 10)

Primer pair 5:

5 ' -TATGGCGCGACCGCGATCGGCCTGCCG- 3 ' ( SΕQ ID NO : 11 )

3 ' -ATACCGCGCTGGCGCTAGCCGGACGGC- 5 ' ( SΕQ ID NO : 12 )

Example 3 - Introduction of LeuB24Arg, PheB177His, ArgB57His, GlnB50 (Val or Ala), ValB70His mutations

Analogously to Example 2 the additional mutations are introduced into the BamHl -Hindlll fragment named in Example 1, the mutations confirmed by DNA sequencing, and recombinant clones prepared. To replace NalB70His, the mutagenic primer pair no. 6 is used. The selection and testing of the recombinant E. coli W3110 clones took place as in Example 2.

Primer pair 6:

5 ' -ATCACCAATACCCACAACGGCATGGTG- 3 ' ( SEQ ID NO : 13 ) 3 ' -TAGTGGTTATGGGTGTTGCCGTACCAC- 5 ' ( SEQ ID NO : 14 )

Example 4 - Introduction of LeuB24Arg, PheB177His, ArgB57His, GlnB50 (Val or Ala), ValB70His, TyrB153Leu mutations.

Analogously to Example 3, the additional mutations of the BamHl -Hindlll fragments named in Example 3 are introduced, the mutations confirmed by DNA sequencing and recombinant clones were prepared. To replace TyrB153Leu, the mutagenic primer pair no. 7 is used. The selection and testing of the recombinant E. coli W3110 clones is performed as described in Example 2.

Primer pair 7:

5 ' -ATGCTCGAGCAGCTCTTCGACATGATC- 3 ' ( SEQ ID NO : 15 ) 3 ' -TACGAGCTCGTCGAGAAGCTGTACTAG- 5 ' ( SEQ ID NO : 16 )

Example 5 - Introduction of PheB177His, ThrB176Asp mutations

Analogously to Examples 1 to 4, these mutations are carried out and tested using primer pairs 8 (PheB177His) and 9 (ThrB176Asρ). Primer pair 8:

5 ' -GTGCCGACCCACAACATCGTCTAC-3 ' (SEQ ID NO: 17) 3 ' -CTCGGCTGGGTGTTGTAGCAGATG-5 ' (SEQ ID NO : 18 )

Primerpair 9:

5 ' -ATGGTGCCGGACCACAACATCGTC- 3 ' ( SEQ ID NO : 19 ) 3 ' -TACCACGGCCTGGTGTTGTAGCAG- 5 ' ( SEQ ID NO : 2 0 )

Example 6 - Introduction of PheB177His, ThrB176Asp, TyrA150Ala mutations

Analogously to Example 5, these mutations were carried out and tested using the primer pairs of the mutated DNA from Example 5 (PheB177His, ThrB176Asp) using the primer pair 10 (TyrAl 50Ala).

Primer pair 10:

5 ' -ATGAACTTCCTCGCTGTCGCGTCGCC - 3 ' ( SEQ ID NO : 21 ) 3 ' -TACTTGAAGGACCGACAGCGCAGCGG- 5 ' ( SEQ ID NO : 22 )

Example 7 - Introduction of PheB177His, ThrB176Asp, LeuB24Arg, ArgB57His mutations

Analogously to Example 5, these mutations are carried out and tested using the fragment already mutagenised in Example 5 and with successive use of primer pairs 1 and 3. Example 8 - Introduction of PheB177His, ThrB176Asp, LeuB24Arg, ArgB57His, TyrA150Ala mutations

Analogously to Example 7, these mutations are carried out and tested using the fragment already mutagenised in Example 7, followed by use of primer pair 10 from Example 6.

Example 9 - Introduction of PheB177His, ThrB176Asp, GlnB50Arg, ArgB57His mutations

Analogously to Example 5, these mutations are carried out and tested using the fragment already mutagenised in Example 5 and with successive use of primer pairs 11 and 3.

Primer pair 11 :

5 ' -TATGGCGCGACCCGGATCGGCCTGCCG- 3 ' ( SEQ ID NO : 23 )

3 ' -ATACCGCGCTGGGCCTAGCCGGACGGC- 5 ' ( SEQ ID NO : 24 )

Example 10 -Introduction of PheB177His, ThrB176Asp, GlnB50Arg, ArgB57His, TyrAl 50 Ala mutations

For the introduction of the TyrAl 50Ala mutation, isolated vector DNA from the already mutated plasmid from Example 9 (PheB177His, ThrB 176 Asp, GlnB50Arg, ArgB57His mutations) is completely digested with the Sail enzyme and the resulting three approx. 750 bp, 1550 bp and 4350 bp fragments isolated after electrophoretic separation by means of electroemtion. The 750 bp fragment is cloned into the Sail cleavage site of the vector puC19, and the TyrAl 50Ala mutation is introduced into the fragment using primer pair 10, and verified. From isolated plasmid DNA the mutagenised approx. 750 bp fragment was completely digested and then cleaved with the Sail restriction enzyme and isolated as described above. This was followed by a ligation reaction together with the already isolated 1550 bp and 4350 bp fragments.

After transformation of E. coli W3110 with the ligation mixture, clones are selected which were able to grow at 28°C in the presence of 12 μg/ml chloramphenicol and which contained vector DNA with the glutaryl amidase gene mutated completely in the correct protein reading direction. All mutations are verified by restriction analysis and DNA sequencing, and recombinant clones tested for conversion of cephalosporin C into

7-ACA.

Example 11 - Introduction of PheB177His, LeuB24Arg, ArgB57His mutations Plasmid T415 is completely digested with the restriction enzymes BamHl and

Hindlll, and the approx. 1900 bp and 4800 bp fragments are isolated after agarose gel electrophoresis and electroelution. First of all, the approx. 1900 bp fragment is ligated into the vector pUC19 cleaved with BamHl and Hindlll (New England Biolabs GmbH, Briiningstr. 50, 65926 Frankfurt). After transformation into the E. coli MC 1061 strain, the plasmid DNA is isolated and subjected to a site-directed mutagenesis mediated via mutagenic primers. The LeuB24Arg mutation is carried out using primer pair no. 1 and confirmed by DNA sequencing after analysis of the isolated plasmid DNA. This DNA mutagenesis is followed by the introduction of the PheB177His mutation using the mutagenic primer pair no. 2. The presence of both mutations in the BamHl-Hindlll fragment was confirmed after isolation of the plasmid DNA by DNA sequencing of the complete fragment. In a third modification round, ArgB57 is mutated to ArgB57His with the help of mutagenic primer pair 3.

From the DNA of the vector which contains the fragment mutated at the described three sites, the approx. 1900 bp fragment is released by digestion with restriction enzymes BamHl and Hindlll, separated by gel electrophoresis and isolated after electroelution. The isolated fragment is ligated into the 4800 bp BmHI-Hindlll cleaved fragment described above, and the ligation mixture is transformed into E. coli W3110.

Recombinant clones are selected on the basis of their ability to grow in the presence of 12 μg/ml chloramphenicol and are tested for conversion of cephalosporin C into 7-aminocephalosporanic acid as described in example 26.

Example 12: - Introduction of the TyrA150Glu mutation

For the introduction of the TyrA150Glu mutation, isolated plasmid T415 DNA is completely digested with the Sail restriction enzyme, and a 750 bp fragment is separated from the 1550 bp and 4350 bp fragments by agarose gel electrophoresis and isolated by electroelution. Then the fragment is cloned into the single cleavage site of the vector pUC19, and the TyrA50Glu mutation is introduced by site-directed mutagenesis using mutagenic primer pair 12, and confirmed by DNA sequencing. From isolated plasmid DNA the approx. 750 bp mutated fragment is isolated as described in Example 10. The subsequent procedure corresponds to that described in Example 10.

Mutagenic primer pair 12:

5 ' -ATGAACTTCCTCGAGGTCGCGTCGCC- 3 ' ( SEQ ID NO : 25 ) 3 ' - TACTTGAAGGAGCTCCAGCGCTGCGG- 5 ' ( SEQ ID NO : 26 )

Example 13 - Introduction of TyrA150Glu, PheB177His mutations

Isolated plasmid DNA from Example 12 which contains the TyrAl 50Glu mutation is completely digested with restriction enzymes BamHl and Hindlll, and the approx. 1900 bp and 4800 bp subfragments are isolated by electroelution after electrophoretic separation in agarose gel. The PheB177His mutation is introduced into the smaller fragment by means of site-directed mutagenesis using the mutagenic primer pair 2, and the presence of the mutations confirmed by DNA sequencing. The subsequent procedure corresponds to that described in Example 2.

Example 14 - Introduction of TyrA150Glu, LeuB24Arg, ArgB57His mutations

Analogously to Example 13, the approx. 1900 bp BamHl-Hindlll fragment is isolated from the vector described in Example 13 and by means of site-directed mutagenesis at first using the mutagenic primer pair 1 the mutation of LeuB24 is introduced to LeuB24Arg and verified by DNA sequencing. Then ArgB57 is mutated to ArgB57His using mutagenic primer pair 3 and verified by DNA sequencing.

From the DNA of the vector which contains the fragment mutated at the described sites, the mutated 1900 bp fragment was isolated after the digestion of the DNA with BamHl -Hindlll. The subsequent procedure corresponds to that described in Example 2.

Example 15 - Introduction of TyrA150Glu, LeuB24Arg, ArgB57His, PheB177His mutations

Isolated plasmid DNA from the vector in Example 14 which already contains the TyrAl 50Glu, LeuB24Arg, ArgB57His mutations is cleaved analogously to Example 14 with the restriction enzymes BamHl and Hindlll, the 1900 bp subfragment is isolated analogously to Example 14 and modified by site-directed mutagenesis at the PheB 177 site using mutagenic primer pair 2. The subsequent procedure for the reconstitution of the expression vector and the testing of the recombinant clones is as described in Example 2. Example 16 - Introduction of TyrA150GIu, GInB50Arg, ArgB57His mutations

Analogously to Example 15, the approx. 1900 bp BamHl -Hindlll fragment is isolated from the DNA of the vector from Example 12, subcloned, and the GlnB50Arg and ArgB57His mutations were introduced at the GlnB50 and ArgB57 sites with the help of the mutagenic primer pairs 11 and 3 and confirmed by DNA sequencing. Analogously to Example 15, the expression vector is reconstituted and tested.

Example 17 - Introduction of TyrA150Glu, GInB50Arg, ArgB57His, PheB177His mutations

From the vector described, analogously to Example 16, the additional PheB177His mutation is introduced into the approx. 1900 bp BamHl -Hindlll fragment with the already existing GlnB50Arg, ArgB57His mutations by means of site-directed mutagenesis with the help of the mutagenic primer pair 2. Analogously to Example 16, the expression vector is reconstituted and tested.

Example 18 - Introduction of PheB177His, ThrB176Asp, TyrA150Ala, TyrB33Ser mutations

Analogously to Examples 1 to 4, using the plasmid DNA from Example 6, the BamH 1 -Hindlll fragment mutated at the PheB 177His, ThrB 176Asp sites is isolated, subcloned, and with the help of the mutagenic primer pair 13, the additional TyrB33Ser mutation is introduced by means of site-directed mutagenesis and confirmed by DNA sequencing. After recloning of the mutagenised fragment into the isolated large BamHl - Hindlll expression vector fragment from Example 6, the recombinant clones are tested as in Example 2.

Mutagenic primer pair 13:

5 ' -GACTACTTCACCTCCTACGAGGCGCAT-3 ' (SEQ ID NO : 27 ) 3 ' -CTGATGAAGTGGAGGATGCTCCGCGTA-5 ' (SEQ ID NO : 28 ) Example 19 - Introduction of PheB177His, ThrB176Asp, LeuB24Arg, ArgB57His, TyrA150Ala, TyrB33Ser mutations

Analogously to Example 18 using isolated plasmid DNA from Example 8, the additional TyrB33Ser mutation is introduced into the already mutated BamHl -Hindlll fragment using the primer pair 13, and the recombinant clones are tested.

Example 20 - Introduction of PheB177His, ThrB176Asp, GlnB50Arg, ArgB57His, TyrA150Ala, TyrB33Ser mutations

Analogously to Example 19 using isolated plasmid DNA from Example 10, the additional TyrB33Ser mutation is introduced into the already mutated BamHl-Hindlll fragment using the primer pair 13, and the recombinant clones are tested.

Example 21 - Introduction of TyrA150Glu, PheB177His, TyrB33Ser mutations

Analogously to Example 20 using isolated plasmid DNA from Example 13, the additional TyrB33Ser mutation is introduced into the already mutated BamHl-Hindlll fragment using the primer pair 13, and the recombinant clones are tested.

Example 22 - Introduction of TyrAl 50Glu, LeuB24Arg, ArgB57His, TyrB33Ser mutations Analogously to Example 21 using isolated plasmid DNA from Example 14, the additional TyrB33Ser mutation is introduced into the already mutated BamHl -Hindlll fragment using the primer pair 13 and the recombinant clones are tested.

Example 23 - Introduction of PheB177His, LeuB24Arg, ArgB57His, ValB70His, TyrB33Ser mutations

Analogously to Example 22 using isolated plasmid DNA from Example 11, the additional mutations ValB70His and TyrB33Ser are introduced, by means of primer pairs 6 and 13 respectively, into the BamHl -Hindlll fragment mutated at the positions PheB177His, LeuB24Arg, ArgB57His, and the recombinant clones are tested. Example 24 - Introduction of PheB177His, LeuB24Arg, ArgB57His,TyrB33Ser mutations

Analogously to example 18, but using the Bam HI/Hindlll fragment of the plasmid DNA of example 11, the additional TyrB33Ser mutation is introduced by means of the mutagenic primer pair 13. Further cloning steps and testing of the recombinant plasmid is done according to example 18.

Example 25 - Alternative in vivo selection of recombinant E. coli clones, which can convert cephalosporin C into 7-ACA. The strain E. coli ESS 2231, obtainable from Prof. A. Demain, MIT Cambridge,

USA, can grow on agar plates with LB medium in the presence of 7-ACA (0.5-2 μg/ml, temperature 28°C), whereas the strain cannot grow on the corresponding medium if 7-ACA is replaced by cephalosporin C in a concentration of 0.5-2 μg/ml. If E. coli ESS 2231 that have been made competent are now transformed using transformation methods with recombinant vectors based on T415 which code for a cephalosporin C acylase activity after mutation of the glutaryl amidase and which form a corresponding functional enzyme in transformed E. coli, these clones can grow in the presence of cephalosporin C, as cephalosporin C is converted into 7-ACA.

This allows a direct positive selection on such clones as can cleave cephalosporin C into 7-ACA. The chloramphenicol resistance introduced with the vector is an additional selection criterion that allows growth of the clones in the presence of 2.5 μg chloramphenicol/ml.

Example 26 - Direct enzymatic conversion of cephalosporin C to 7-ACA To identify clones with mutated glutaryl amidase, which can convert cephalosporin C directly to 7-ACA starting from single colonies, recombinant cells are incubated at 220 rpm overnight at 28°C in 200 ml of a complex medium consisting of 1% bactotryptone, 0.5% yeast extract and 0.5% NaCl, pH 7.2 in an Erlenmeyer flask. The cells are then centrifuged off at 4°C, the cell pellet washed in physiological salt solution and the cells taken up in ca. 1 ml Tris/HCl buffer pH 8.2, to which cetyltriniethyl ammonium chloride or toluene in combination with EDTA or lysozyme is added as a detergent for permeabilisation and release of enzyme activity. After 30 mins incubation time at 4°C, the cells are centrifuged off and aliquots of the supernatant are used for an enzymatic assay, as described by Ishii et al, 1995, supra. The formation of 7-ACA from cephalosporin C is tracked by means of HPLC.

Example 27 - Expression of recombinant GA and industrial conversion of cephalosporin to 7-ACA

Recombinant E. coli clones with cephalosporin acylase activity are cultured to 10 litre scale according to the method described in EP-A-708180. The cells are initially concentrated by centrifugation, taken up in small volume with 100 mM phosphate buffer, pH 7.2, and disrupted by means of a French press. To separate off the cell residues, the suspension is centrifuged at 10,000 rpm for 15 mins, and the supernatant removed. The enzyme-containing solution is rebuffered in an Amicon ultrafiltration cell or by diafiltration against 1 to 1.5 M potassium phosphate buffer pH 7.8 at 4°C with simultaneous concentration. If the concentrated solution is left to stand in cold conditions, the enzyme crystallises out. The crystals are enriched by gentle centrifugation and washed with the above-mentioned buffer.

The protein crystals are dissolved in 0.1 M Tris/HCl buffer or 0.1 M potassium phosphate buffer pH 8.2 for industrial conversion and the enzyme used in a concentration of 2 mg/ml for the conversion of cephalosporin to 7-ACA. Substrate concentration: 75 mM, temperature: 37°C. Reaction time: 1 hour. The conversion is analytically tracked by means of HPLC .

All publications mentioned in the above specification are herein incorporated by reference. Various modifications and variations of the described methods and system of the invention will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been described in connection with specific preferred embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments. Indeed, various modifications of the described modes for carrying out the invention which are apparent to those skilled in molecular biology or related fields are intended to be within the scope of the following claims.

Claims

1. A non-naturally occurring variant of a glutaryl amidase (GA) which binds cephalosporin C (CPC) as a substrate and catalyses conversion of the CPC to 7-aminocephalosporanic acid (7-ACA), which variant comprises a histidine or glutamate in its substrate binding pocket which binds the α-amino adipyl moiety of the CPC.

2. A variant according to claim 1 which comprises a histidine residue at position B177 of the GA sequences shown as SEQ ID Nos:l and 2, or their equivalent in other GA sequences.

3. A variant according to claim 2 which further comprises an aspartate residue at position B176 of the GA sequence shown as SEQ ID No:2 or its equivalent in other GA sequences.

4. A variant according to claim 1 which comprises a glutamate residue at position A150 of the GA sequences shown as SEQ ID Nos:l and 2, or their equivalent in other GA sequences.

5. A variant according to any one of claims 1 to 4 wherein the substrate pocket comprises only one positively charged residue for binding the carboxylate group of cephalosporin C.

6. A variant according to claim 5 wherein the positively charged residue is inserted by mutating LeuB24 or GlnB50 to a positively charged residue, and mutating Arg B57 to an uncharged residue, as numbered according to the GA sequence shown as SEQ ID No:2 or its equivalent in other GA sequences.

7. A variant according to claim 5 or claim 6, wherein the positively charged residue is Arginine.

8. A variant according to any one of the preceding claims wherein additional mutations are present that stabilise or enlarge the substrate pocket volume.

9. A variant according to claim 8 wherein the mutations are selected from

TyrAl 50-*- Ala, GlnB50-»Val or Ala and TyrB153-»Leu, as numbered according to the GA sequences shown as SEQ ID Nos: 1 or 2 or their equivalents in other GA sequences.

10. A non-naturally occurring variant of a glutaryl amidase (GA) which binds cephalosporin C (CPC) as a substrate and catalyses conversion of the CPC to 7-aminocephalosporanic acid (7-ACA), which variant comprises a combination of mutations selected from the group consisting of: PheB 177His, LeuB24Arg, ArgB57His;

PheB177His, LeuB24Arg, ArgB57His, GlnB50Val;

PheB177His, LeuB24Arg, ArgB57His, GInB50AIa;

PheB177His, LeuB24Arg, ArgB57His, GlnB50Val, ValB70His;

PheB177His, LeuB24Arg, ArgB57His, GlnB50Ala, ValB70His; PheB 177His, LeuB24Arg, ArgB57His, GlnB50Val, ValB70His, TyrB 153Leu;

PheB177His, LeuB24Arg, ArgB57His, GlnB50Ala, ValB70His, TyrB153Leu;

PheB 177His, ThrB 176Asp;

PheB177His, ThrB176Asp, TyrAl 50Ala;

PheB177His, ThrB176Asp, LeuB24Arg, ArgB57His; PheB 177His, ThrB 176Asp, LeuB24Arg, ArgB57His, TyrAl 50Ala;

PheB177His, ThrB176Asp, GlnB50Arg, ArgB57His;

PheB177His, ThrB176Asp, GlnB50Arg, ArgB57His, TyrAl 50Ala;

TyrAl 50Glu;

TyrAl 50Glu, PheB177His; TyrAl 50Glu, LeuB24Arg, ArgB57His;

TyrAl 50Glu, LeuB24Arg, ArgB57His, PheB177His;

TyrAl 50Glu, GlnB50Arg, ArgB57His;

TyrAl 50Glu, GlnB50Arg, ArgB57His, PheB177His;

PheB177His, ThrB176Asp, TyrA150Ala, TyrB33Ser; PheB177His, ThrB176Asp, LeuB24Arg, ArgB57His, TyrA150Ala, TyrB33Ser;

PheB177His, ThrB176Asp, GlnB50Arg, ArgB57His, TyrA150Ala, TyrB33Ser;

TyrA150Glu, PheB177His, TyrB33Ser;

TyrAl 50Glu, LeuB24Arg, ArgB57His, TyrB33Ser; PheB177His, LeuB24Arg, ArgB57His, TyrB33Ser; and PheB177His, LeuB24Arg, ArgB57His, TyrB33Ser, ValB70His.

11. A polynucleotide encoding a variant GA according to any one of claims 1 to 10.

12. A nucleic acid vector comprising a nucleotide according to claim 11.

13. A host cell comprising a nucleotide according to claim 11 operably linked to regulatory control sequence which direct expression of the variant GA in said host cell.

14. A method for producing a variant GA of any one of claims 1 to 10 which method comprises culturing a host cell according to claim 13 under suitable conditions which allow for expression of the variant GA.

15. A method according to claim 14 which further comprises

(i) lysing/disrupting the host cells expressing the variant GA to obtain a solution comprising said variant GA in solution; (ii) adjusting the potassium phosphate concentration of the solution to from 1 to 1.5M at a pH of from 7 to 8; (iii) concentrating the solution; and

(iv) allowing the variant GA protein to crystallise out from the concentrated solution at a temperature of about 4°C or less.

16. A method for preparing 7-ACA from cephalosporin C which comprises reacting cephalosporin C with a variant GA enzyme according to any one of claims 1 to 10 under suitable conditions such that the variant GA enzyme cleaves the cephalosporin C directly to form 7-ACA.

17. A method according to claim 16 wherein the reaction is performed at a pH of from pH 7.5-8.5.

18. A method according to claim 16 or claim 17 wherein the concentration of cephalosporin C in the reaction mixture is from 20 to 200 mM.

19. A method according to any one of claims 16 to 18 wherein concentration of variant GA enzyme in the reaction mixture is from 0.25 to 10 mg/ml is used in the reaction mixture,

20. A method according to any one of claim 16 to 19 wherein the reaction temperature

21. A method for modifying a glutaryl amidase enzyme to enhance the cephalosporin C acylase activity of said enzyme which method comprises introducing a non-naturally occurring histidine or glutamate amino acid residue in the substrate binding pocket of said enzyme which histidine or glutamate binds the α-amino adipyl moiety of a CPC substrate molecule.

22. A method according to claim 21 which further comprises substituting or deleting a naturally-occurring positively charged amino acid residue in said substrate binding pocket and introducing a positively charged amino acid residue which binds the carboxylate group of the CPC substrate molecule.