Title of Invention

A PROCESS FOR PRODUCING 2-KETO-L-GULONIC ACID FROM L-SORBOSE AND/OR D-SORBITOL

Abstract ABSTRACT 1997/MAS/97 The present invention relates to a process for producing 2-keto-L-gulonic acid from L-sorbose and/or D-sorbitol, the said process comprises converting L-sorbose and/or D-sorbitol into 2-keto-L-gulonic acid with the aid of a biochemical action of a recombinant organism, said organism carrying a recombinant expression vector, said expression vector comprising one or more DNA molecules encoding an enzyme having an alcohol and/or aldehyde dehydrogenase activity, said enzyme comprising a recombinant polypeptide selected from the group consisting of SEQ ID NO:5, SEQ ID N0:6, SEQ ID N0:7, SEQ ID N0:8, chimeric recombinant enzymes between the polypeptides identified by said sequences, and polypeptides with at least 80% identity to said sequences.
Full Text



The invention relates to a process for producing 2-keto-L-gulonic acid from L-sorbose and/or D-sorbitol.
also The present invention/relates to recombinant enzyme preparations of
alcohol/aldehyde dehydrogenate(s) (hereinafter referred to as AADH or AADHs)
having alcohol and/or aldehydes dehydrogenize activity (activities). The present
invention also relates to novel recombinant DNA molecule(s) encoding AADH(s),
recombinant expression vector(s) containing said DNA(s), and recombinant
organism(s) containing said recombinant DNA molecule(s) and/or said recombinant
expression vector(s). Furthermore, the present invention relates to a process for
producing recombinant enzyme preparation(s) of AADH(s) and a process for producing
aldehyde(s), carboxylic acid(s) and ketone(s), especially, 2-keto-L-gulomc acid (herein
after referred to as 2KGA) by utilizing said recombinant enzyme preparation(s), and a
process for producing aldehyde(s), carboxylic acid(s) and ketene(s), especially, 2KGA
by utilizing said recombinant organism(s).
2-KGA is an important intermediate for the production of L-ascorbic acid (vitamin C). Numerous microorganisms are known to produce 2KGA from D-sorbitol or L-sorbose. Japanese Patent Publication No. 51-40154 (1976) discloses the production of 2KGA from D-sorbitol by microorganisms of the genus Acetobacter, Bacterium or Pseudomonas. According to Acta Microbiologic Sinica 21(2), 185-191 (1981), 2KGA can be produced from L-sorbose by a mixed culture of microorganisms, especially, Pseudomonas striate and Gluconobacter oxydans. European Patent Publication No. 0221 707 discloses the production of 2KGA from L-sorbose by Pseudogluconobacter sacchctroketogenes with and without concomitant bacteria. European Patent Publication No. 0278 447 discloses a process for the production of 2KGA from L-sorbose by a mixed culture, which is composed of strain DSM No.

4025 (Gluconobacter oxydans) and DSM No. 4026 (a Bacillus megaterium strain). European Patent Publication No. 88116156 discloses a process for the production of 2KGA from L-sorbose by Gluconobacter oxydans DSM No. 4025.
From G. oxydans DSM No. 4025, AADH was purified and characterized to catalyze the oxidation of alcohols and aldehydes, and was thus capable of producing the corresponding aldehydes and ketones from alcohols, and carboxylic acids from aldehydes (see European Patent Publication No. 606621). More particularly, the AADH catalyzed the oxidation of L-sorbose to 2KGA via L-sorbosone. The physic-chemical properties of the purified sample of the AADH were as follows:
a) Optimum pH: about 7.0 - 9.0
b) Optimum temperature: about 20°C - 40°C
c) Molecular weight: 135,000 +1- 5,000 dalton
(Consisting of two subunits in any combination of such a-subunit and P-subunit, each having a molecular weight of 64,500 +/- 2,000 and 62,500 +/- 2,000, respectively)
d) Substrate specificity: active on primary and secondary alcohols and aldehydes including L-sorbose, L-curbstone, D-sorbitol, D-glucose, D-mannitol, D-finites, DL-glycelaldehyde, ethanol, 1-propanol, 1-butanol, 1-pentanol, 1-hexanol, 1-heptanol, 2-propanol, 2-butanol, propionaldehyde, PEGIOOO, PEG2000, PEG4000, PEG6000 and polyvinyl alcohol
e) Prosthetic group: pyrroloquinoline quinone
f) Bioelectric point: about 4.4
Once the gene(s) coding for said AADH have been cloned, they can be used for the construction of a recombinant organism capable of producing a large amount of the

recombinant enzyme preparation of AADH or the various aldehydes, ketones and carboxylic acids, especially, 2KGA. However, there have been no reports so far of the cloning of such genes.
Therefore the present invention relates to novel recombinant enzyme preparation(s) of AADH(s) having alcohol and/or aldehyde dehydrogenate activity (activities). Comprised by the present invention are novel recombinant molecule(s) encoding the AADH(s); recombinant expression vector(s) containing said DNAs; recombinant organism(s) carrying said DNA(s) and/or recombinant expression vector(s); a process for producing the recombinant AADH(s); and a process for producing aldehyde(s), carboxylic acid(s) and ketone(s), especially, 2KGA utilizing the recombinant AADH(s) or the recombinant organism(s).
More particularly, an aspect of the present invention concerns a recombinant enzyme preparation having an alcohol and/or aldehyde dehydrogenate activity which comprises one or more enzymatic polypeptide(s) selected from the group consisting of polypeptides which are identified by SEQ ID NO 5, SEQ ID NO 6, SEQ ID NO 7 and SEQ ID NO 8 and chimerical recombinant enzymes between the polypeptides identified by SEQ ID NO 5, SEQ ID NO 6, SEQ ID NO 7 and SEQ ID NO 8 and functional derivatives of the polypeptides identified above which contain addition, insertion, deletion and/or substitution of one or more amino acid residue(s), wherein said enzymatic polypeptides have said alcohol and/or aldehyde dehydrogenase activity.
Such functional derivatives can be made either by chemical peptide synthesis known in the art or by recombinant means on the basis of the DNA sequences as disclosed herein by methods known in the state of the art and disclosed e.g. by Sambrook et al. (Molecular Cloning, Cold Spring Harbour Laboratory Press, New York, USA, second edition 1989). Amino acid exchanges in proteins and peptides which do not generally

activity of such molecules are known in the state of the art and are described, H. Neurath and R.L. Hill in "The Proteins" (Academic Press, New York, 1979, sec especially Figure 6, page 14). The most commonly occurring exchanges are: Ala/Ser, Val/De, Asp/Glu, Thr/Ser, Ala/Gly, Ma/Thr, Ser/Asn, AlaA'al, , Ser/Gly. Tyt/Phe, Ala/Pro, Lys/Arg, Asp/Asn, Leu/IIe, LeuA'al, Ala/Glu, Asp/Gly as
i
well as these in reverse.
Another aspect of the present invention concerns a recombinant DNA molecule encoding at least one enzymatic polypeptide selected from the group consisting of
polypeptides identified by SEQ ID NO 5, SEQ ID NO 6, SEQ ID NO 7 and SEQ ID
i NO 8 and chimeric recombinant enzymes between the polypeptides identified by SEQ
ID NO 5, SEQ ID NO 6, SEQ ID NO 7 and SEQ ID NO 8 and functional derivatives of
the polypeptides identified above which contain addition, deletion and/or substitution of
one or more amino acid residue(s), wherein said enzymatic polypeptides have said
alcohol and/or aldehyde dehydrogenate activity.
Furthermore the present invention is directed to DNA sequences encoding the polypeptides with alcohol and/or aldehydes dehydrogenate activity as disclosed e.g. in the sequence listing as well as their complementary strands, or those which include these sequences, DNA sequences which hybridize under standard conditions with such sequences or fragments thereof and DNA sequences, which because of the degeneration of the genetic code, do not hybridize under standard conditions with such sequences but which code for polypeptides having exactiy the same amino acid sequence.
"Standard conditions" for hybridization mean in this context the conditions which are generally used by a man skilled in the art to detect specific hybridization signals and which are described, e.g. by Sambrook et al., "Molecular Cloning" second edition.

Cold Spring Harbour Laboratory Press 1989, New York, or preferably so called stringent hybridization and non-stringent washing conditions or more preferably so called stringent hybridization and stringent washing conditions a man skilled in the art is familiar with and which are described, e.g. in Sam brook et al. (s.a.). Furthermore DNA sequences which can be made by the polymerase chain reaction by using primers designed on the basis of the DNA sequences disclosed herein by methods known in the art are also an object of the present invention. It is understood that the DNA sequences of the present invention can also be made synthetically as described, e.g. in EP 747 483.
Further aspects of the present invention conked a recombinant expression vector which carries one or more of the recombinant DNA molecule(s) defined above and a recombinant organism which carries the recombinant expression vector defined above and/or carries one or more recombinant DNA molecule(s) on a chromosome.
A further aspect of the present invention conches a process for producing a recombinant enzyme preparation having an alcohol and/or aldehyde dehydrogenate activity as defined above, which comprises cultivating a recombinant organism defined above in an appropriate culture medium and recovering said recombinant enzyme preparation.
Another aspect of the present invention concems a process for producing an aldehyde, ketone or carboxylic acid product from a corresponding substrate which comprises converting said substrate into the product by the use of a recombinant organism as defined above.

Moreover another aspect of the present invention concerns a process for producing 2-keto-L-gulonic acid which comprises the fermentation of a recombinant organism as defined above in an appropriate medium containing L-sorbose and/or D-sorbitol.
Another aspect of the present invention concerns a process for producing an aldehyde, ketone or carboxylic acid product from a corresponding substrate which comprises the incubation of a reaction mixture containing a recombinant enzyme preparation of the present invention.
Further more another aspect of the present invention concerns a process for producing 2-keto-L-gulonic acid which comprises the incubation of a reaction mixture containing a recombinant enzyme preparation defined above and L-sorbose and/or D-sorbitol.
It is also an object of the present invention to provide a process for the production of vitamin C from 2-keto-L-gulonic acid characterized therein that a process for the production of 2-keto-L-gulonic acid as described above is effected and die 2-keto-L-gulonic acid obtained by such process is transformed into vitamin C (L-ascorbic acid) bv methods known in the art

Accordingly, the present invention provides a process for producing 2-^eto-L-gulonic acid from L-sorbose and/or D-sorbitol, the said process ccm^ses (a) converting L-sorbose and/or D-sorbitol into 2-keto-L-guloiuc acid with the aid of a biochemical action of a recombinant (vganism, said or^nism carrying a recombinant ejqwession vector, said ejqaression vector comprising one or more DNA molecules encoding an enzyme having an alcohol and/or alddbyde dehydrogenase activity, said enzyme comprising a recombinant polypeptide selected from the group c(Misisting of SEQ ID NO 5, SEQ YD NO 6, SEQ ID NO 7, SEQ ID NO 8, chimeric recombinant enzymes between the polypeptides identified by said sequences, and polypeptides with at least 80% identity to said sequences, wherem the conversion is carried out under a pH value in the range of from 6.0 to 9.0 at a temperature in the range of from 10°C to 50°C, in the presence of an electron acceptor in a suitable buffer and with a substrate concentration in the range of from 1 to 200 g/1; and (b) isolating the resulting 2-keto-L-gulonic acid.
Before describing the present invention in more detail, a shOTt explanation with reference to the accompanying drawings, in which,
Figure 1 schematically illustrates the structures of the recombinant expression vectors each carrying the recombinant DNA molecule which encodes the recombinant Enzyme A or B of the present invention.
Z 1 MAY 2002 DUPLICATE

Figure 2 schematically illustrates the structures of the recombinant expression vectors each carrying the recombinant DNA molecule which encodes the chimeric enzyme of the present invention.
Figure 3 schematically illustrates the stmctures of the material plasmids each carrying the recombinant DNA molecule containing tandem structural genes of Enzyme A and Enzyme B for constracting the chimeras by a homologous recombination method.
Figure 4 illustrates the recombinant expression vectors each encoding the chimera Enzyme sA2, Enzyme sA21, Enzyme sA22, or Enzyme sB, using preferable codon usage, wherein Enzyme sA2 has the structure of "Enzyme A part of No. 1-135, Enzyme B part of No. 136-180 and Enzyme A part of No. 180 - 556", Enzyme sA21 has the structure of "Enzyme A part of No. 1-128, Enzyme B part of No. 129 -180 and Enzyme A part of No. 180 - 556", Enzyme sA22 has the stmcture of "Enzyme A part of No. 1-125, Enzyme B part of No. 126 - 180 and Enzyme A part of No. 180 - 556", and Enzyme sB has the stracture of "Enzyme A part of No. 1 - 95, Enzyme B part of No. 96-180 and Enzyme A part of No. 180 - 556". These numbers are the amino acid residue numbers of the mature enzyme amino acid sequences.
Figure 5 shows the alignment of the amino acid sequences of the mature Enzyme A and Enzyme B.
Figure 6 illustrates the constraction schemes of the recombinant genes encoding chimeric enzymes of the present invention.
Figure 7 shows the restriction map of the genes of Enzymes A and B.
Figure 8 illustrates the construction of chimeric genes by homologous recombination of two AADH genes in vivo at the conserved nucleotide sequences in both genes.
Figure 9 shows a site-directed mutagenesis to introduce a BaniHl site upstream of the Enzyme B gene.

Figure 10 illustrate a scheme of the replacement of the promoter for the Enzyme B gene.
Figure 11 shows graphs illustrating the substrate specificity of chimeric enzymes of the invention.
The AADH genes of the present invention encode the AADH enzymes capable of catalyzing the oxidation of various alcohols and aldehydes as described above. Specifically speaking the particular genes of AADHs present in Gluconobacter were cloned and expressed. Alternative sources in addition to Gluconobacter may well be found among the other organisms by the man skilled in the art using the teachings of the present invention.
A specific and preferred Gluconobacter oxydans strain has been deposited at the Deutsche Sammlung von Mikroorganismen in Gottingen (Germany) under DSM No. 4025.
Moreover, a subculture of the strain has also been deposited in the Agency of Industrial Science and Technology, Fermentation Research Institute, Japan, under the deposit No.: FERM BP-3812. European Patent publication No. 0278 477 disclose the characteristics of this strain.
The AADH genes and the recombinant microorganisms utilized in the present invention can be obtained by the following steps:
(1) Cloning the AADH genes from a chromosomal DNA by colony- or plaque-hybridization, PCR cloning, Westem-blot analysis, Southem-blot hybridization and the like.

(2) Determining the nucleotide sequences of such AADH genes by usual methods and constructing recombinant expression vectors which contain and express AADH genes efficiently.
(3) Constructing recombinant microorganisms carrying recombinant AADH genes on recombinant expression vectors or on chromosomes by transformation, transduction, transconjugation and electroporation.
The materials and the techniques applicable to the above aspect of the present invention are exen^lified in details as described in the following:
A total chromosomal DNA can be purified by a procedure well known in the art (Marmur J., J. Mol. Biol. 2:208,1961). Then, a genomic library of the strain for such genes can be constructed with the chromosomal DNA and the vectors described below in detail. The genes encoding AADHs can be cloned in either plasmid or phage vectors from the total chromosomal DNA by the following methods:
(i) determining the partial amino acid sequences of the purified enzyme, according to the sequence information, synthesizing the oligonucleotides, and selecting the objective gene from the gene library by Southern-blot-, colony-, OT plaque-hybridization;
(ii) by amplifying the partial sequence of the desired gene by polymerase chain reaction (PCR) with the oligonucleotides synthesized as described above as the primers and with the PCR product as a probe, selecting the complete sequence of the objective gene from the gene library by Southern-blot-, colony-, or plaque-hybridization;
(iii) by preparing the antibody reacting against the desired enzyme protein by such a method as previously described, e.g. in Methods in Enzymology, vol. 22. P 46,1981,

and selecting the clone which expresses the desired polypeptide by immnunological analysis including Western-blot analysis;
(iv) by aligning the amino acid sequences of the homologs to the one of the desired enzyme, selecting the amino acid sequences which are well conserved, synthesizing the oligonucleotides encoding the conserved sequences, amplifying the partial sequence of the desired gene by PCR with the above oligonucleotides as the primers, and selecting the complete sequence as described above (ii).
The nucleotide sequence of the desired gene can be determined by a well known method such as the dideoxy chain termination method with the M13 phage (Sanger F., et al., Proc. NaU. Acad. Sci.USA, 24:5463-5467, 1977).
By using the information of the so determined nucleotide sequence (in consideration of the codon usage) a gene encoding evolutionally divergent alcohol and/or aldehyde dehydrogenases, can be isolated from a different organism by colony-or Southern-hybridization with a probe synthesized according to the amino acid sequence deduced from said nucleotide sequence or by the polymerase chain reaction with primers also synthesized according to said information, if necessary.
To express the desired gene or generally speaking the desired DNA sequence of the present invention efficiently, various promoters can be used; for example, the original promoter of said gene, promoters of antibiotic resistance genes such as the kanamycin resistant gene of Tn5 (Berg, D. E., and C. M. Berg. 1983. Bio/Technology 1:417-435), the ampicillin resistant gene of pBR322, a promoter of the beta-galactosidase gene of Escherichia coli (lac), trp-, tac- trc-promoter, promoters of lambda phages and any promoters which can be functional in the hosts consisting of

microorganisms including bacteria such as E. coli, P. putida, Acetobacter xylinum, A. pasteurianus, A. aceti, A. hansenii and G. oxydans, mammalian and plant cells.
Furthermore other regulatory elements, such as a Shine-Dalgamo (SD) sequence (for example, AGGAGG etc. including natural and synthetic sequences operable in the host cell) and a transcriptional terminator (inverted repeat structure including any natural and synthetic sequence operable in the host cell) which are operable in the host cell into which the coding sequence will be introduced can be used with the above described promoters.
For the expression of periplasmic polypeptides (AADHs) a signal peptide, which contains usually 15 to 50 amino acid residues totally hydrophobic, is indispensable. A DNA encoding a signal peptide can be selected from any natural or synthetic sequence operable in the host cell.
A wide variety of host/cloning vector combinations may be employed in cloning the double-stranded DNA. Cloning vector is generally a plasmid or phage which contains a replication origin, regulatory elemtents, a cloning site including a multi-cloning site and selection markers such as antibiotic resistance genes including resistance genes for ampicillin, tetracycline, kanamycin, streptomycin, gentamicin, spectinomycin etc.
Preferred vectors for the expression of the DNA sequences of the present invention in E. coli are selected from any vectors usually used in E. coli, such as pBR322 or its derivatives including pUC18 and pBluescript H, pACYC177 and pACYC184 (J. BacterioL, 124:1141-1156,1978) and their derivatives, and a vector derived from a broad host range plasmid such as RK2 and RSFIOIO. A preferred vector for the expression of the DNA sequences of the present invention

mGluconobacter including G. oxydans DSM No. 4025 and P. putida is selected from any vectors which can rephcate in Gluconobacter and/or P. putida, as well as a preferred cloning organism such as in E. coli. The preferred vector is a broad-host-range vector such as a cosmid vector like pVK102 and its derivatives and RSFIOIO and its derivatives, and a vector containing a replication origin functional in Gluconobacter and another origin functional in E. coli. Copy number and stability of the vector should be carefully considered for stable and efficient expression of the cloned gene and also for efficient cultivation of the host cell carrying the cloned gene. DNA molecules containing transposable elements such as Tn5 can be also used as a vector to introduce the DNA sequence of the present invention into the preferred host, especially on a chromosome. DNA molecules containing any DNAs isolated from the preferred host together with the desired DNA sequence of the present invention are also useful to introduce the desired DNA sequence of the present invention into the preferred host, especially on a chromosome. Such DNA molecules can be transferred to the preferred host by transformation, transduction, transconjugation or electroporation.
Useful hosts may include microorganisms, mammalian cells, and plant cells and the like. As preferable microorganisms, there may be mentioned bacteria such as E. coli, P. putida, A. xylinwn, A. pasteurianus, A. aceti, A. hansenii, G. oxydans, and any Gram-negative bacteria which are capable of producing recombinant AADHs. Functional equivalents, subcultures, mutants and variants of said microorganism can be also used in the present invention. A preferred strain is E. coli K12 and its derivatives, P. putida or G. oxydans DSM No. 4025.
The functional AADH encoding DNA sequence of the present invention is ligated into a suitable vector containing a regulatory region such as a promoter and a ribosomal binding site operable in the host cell described above using well-known

methods in the art to produce an expression plasmid. Structures of such recombinant expression vectors are specifically shown in Fig. 1,2,4, and 10.
To construct a recombinant microorganism carrying a recombinant expression vector, various gene transfer methods including transformation, transduction, conjugal mating (Chapters 14 and 15, Methods for general and molecular bacteriology, Philipp Gerhaidt et al. ed., American Society for Microbiology, (1994), and electroporation can be used. The method for constructing a recombinant organism may be selected from the methods well-known in the field of molecular biology. Usual transformation systems can be used for E. coli, Pseudomonas and Acetobacter. A transduction system can also be used for E. coli. Conjugal mating systems can be widely used in Gram-positive and Gram-negative bacteria including E. coli, P. putida and G. oxydans. A preferred conjugal mating method is described in WO89/06688. The conjugation can occur in liquid media or on a soUd surface. The preferred recipient is selected from E. coli, P. putida and G. oxydans which can produce active AADHs with a suitable recombinant expression vector. The preferred recipient for 2KGA production is G. oxydans DSM No. 4025. To the recipient for conjugal mating, a selective marker is usually added; for example, resistance against nalidixic acid or rifampicin is usually selected.
The AADHs provided by the present invention catalyze the oxidation of alcohols and/or aldehydes, and are thus capable of producing aldehydes, ketones or carboxylic acids fiom the corresponding substrates. More particularly, the AADHs provided by the present invention can catalyze the oxidation of L-sorbose to 2KGA via L-sorbosone and/or the oxidation of D-sorbitol to L-sorbose. More particularly, the AADHs provided by the present invention contain Enzyme A, Enzyme A', Enzyme A", and Enzyme B, which have the amino acid sequences shown in SEQ ID NO. ID NO. 5,6, 7, and 8, respectively.

Enzyme A, A', A", and B genes, which have the nucleotide sequences shown in SEQ ID NO. 1,2,3, and 4, respectively, and encode the polypeptides having the amino acid sequences shown in SEQ ID NO. 5,6,7, and 8, respectively can be derived from G. oxydans strain DSM No. 4025.
The AADHs including Enzymes A, A', A" and B provided by the present invention can be prepared independently by cultivating an appropriate organism, disrupting the cells and isolating and purifying them from cell free extracts of disrupted cells, preferably from the soluble fraction of the microorganism.
The recombinant organisms provided in the present invention may be cultured in an aqueous medium supplemented with appropriate nutrients under aerobic conditions. The cultivation may be conducted at a pH between about 4.0 and 9.0, preferably between about 6.0 and 8.0. While the cultivation period varies depending upon pH, temperature and nutrient medium used, usually 2 to 5 days will bring about favorable results. A preferred temperature range for carrying out the cultivation is from about 13°C to 45°C preferably from about 18°C to 42°C.
It is usually required that the culture medium contains such nutrients as assimilable carbon sources, digestible nitrogen sources and inorganic substances, vitamins, trace elements and other growth promoting factors. As assimilable carbon sources, glycerol, D-glucose, D-mannitol, D-fructose, D-arabitol, D-sorbitol, L-sorbose, and the like can be used.
Various organic or inorganic substances may also be used as nitrogen sources, such as yeast extract, meat extract, peptone, casein, com steep liquor, lu-ea, amino acids, nitrates, ammonium salts and the like. As inorganic substances, magnesium

sulfate, potassium phosphate, ferrous and ferric chlorides, calcium carbonate and the like may be used.
In the following, the properties of the purified recombinant AADH enzymes specifically from P. putida and the production method are summerized.
(1) Enzyme activity
The AADHs of the present invention catalyze oxidation of alcohols and aldehydes including D-sorbitol, L-sorbose, and L-sorbosone in the presence of an electron acceptor according to the following reaction formula.
Alcohol+ Electron acceptor -> Aldehyde + Reduced electron acceptor
Alcohol+ Electron acceptor -> Ketone + Reduced electron acceptor
Aldehyde + Electron acceptor -> Carboxylic acid + Reduced acceptor
Sugar alcohol + Electron acceptor -> Aldose + Reduced electron acceptor
Sugar alcohol + Electron acceptor -> Ketose + Reduced electron acceptor
Aldehyde ketose+Electron acceptor -> Ketocarboxylic acid+Reduced electron acceptor
Carboxylic acid+Electron acceptor -> Ketocarboxylic acid+Reduced electron acceptor
The enzymes do not utUize molecular oxygen as an acceptor. As an acceptor, 2,6-dichlorophenolindophenol (DCIP), phenazine methosulphate (PMS), Wurster's blue, ferricyanide, coenzyme Q or cytochrome c can be used.
One unit of enzyme activity was defined as the amount of enzyme which catalyzed the reduction of 1 ^mole of DCIP per minute. The extinction coefficient of DCIP at pH 8.0 was taken as 15 mM'l. The standard reaction mixture (1.0 ml) contained 0.1 mM DCIP, 1 mM PMS, 2 to 125 mM substrate, 50 mM Tris-malate-

NaOH buffer (pH 8.0), andlO |il of the enzyme solution. A reference cuvette contained all the above coniponents except the substrate.
(2) Properties of the AADHs
a) Substrate specificity and products of the enzymatic reaction
The Enzymes A, A', A" and B were characterized by their substrate specificities
as described above using 8 substrates: n-propanol, isopropanol, D-glucose, D-sorbitol,
L-sorbosone, D-marmitol, L-sorbose, and D-fiiictose. The results are indicated in
Table 1.
Table 1. Substrate specificity of the Enzymes A, A', A" and B

Enzyme B showed a high reactivity for D-glucose or D-mannitol, but relatively low reactivity for n-propanol and isopropanol. Enzyme A, Enzyme A' and Enzyme A" showed a high reactivity for n-propanol and isopropanol, but a low reactivity for D-glucose and D-mannitol; the enzymes showed similar substrate specificity patterns, except that the Enzyme A' had a very low reactivity for L-sorbose or D-finictose.

Product(s) fonned from a substrate in the reaction with Enzyme A, Enzyme A', Enzyme A" or Enzyme B was analyzed by thin layer chromatography (TLC) and/or high perforaiance liquid chromatography (HPLC) with authentic compounds. Enzyme A, Enzyme A' and Enzjmie A" (designated A group) converted D-sorbitol, L-sorbose, L-sorbosone, D-mannitol, and D-fructose to D-glucose with L-gulose, L-sorbosone with 2KGA, 2KGA, D-mannose, and 2-keto-D-gluconic acid (2KD), respectively. Enzyme B (designated B group) converted D-glucose, D-sorbitol, L-sorbosone, D-mannitol, L-idose, glycerol, D-gluconic acid, D-mannoic acid to D-gluconate, L-sorbose, 2KGA, D-ftuctose, L-idonic acid, dihydroxyacetone, 5-keto-D-gluconic acid, and 5-keto-D-mannoic acid, respectively. In the analogy to the reactivity for L-sorbosone, D-glucosone can be converted to 2KD by all of above mentioned AADHs; actually A group enzymes produced, 2KD from D-fructose whose possible direct product is D-glucosone. All of the enzymes showed the activity for alcohols including sugar alcohol such as D-sorbitol and D-mannitol, and aldehydes including aldose such as D-glucose and ketose such as L-sorbosone.
b) Optimum pH
All the enzymes have their optimal point at pH 8.0 - 8.5 as shown in Table 2. The Enzymes A" and B have a relatively wide pH range toward a lower pH, compared with the Enzymes A and A'.




Table 4. Thermal stability of the enzymes
(Relative activity, %)

Temperature

Enzyme A

EnzymeA'

Enzyme A"

Enzyme B


4°C 100.0 100.0 100.0 100.0
20 °C 91.5 100.8 96.0 97.2
30 °C 78.0 103.6 86.1 95.4
40 °C 19.9 78.9 72.8 84.6
50 °C 4.1 0.6 26.6 29.2
60 °C 2.9 0.0 13.3 0.0

e)

Effect of metal ions and inhibitors

Remaining activities after the treatment of the enzymes with various metals and inhibitors are shown in Table 5. MgCl2 and CaCl2 were nearly inert to the enzymes, while the other metal ions, especially CuCl2, significantly affected the reactivity. EGTA and EDTA inhibited the Enzymes A, A' and A", remarkably. However, Enzyme B was less inhibited than the A group enzymes by EDTA and EGTA.
Table 5. Effect of metals and inhibitors on activities of the Enzymes A, A', A" and
B.
(Relative remaining activity)

Enzyme
A"
B
Compound
Substrate L-Sorbose n-Propanol L-Sorbose D-Sorbitol
16.6 7.9 46.9 23.6
0.0 0.0 0.0 0.0
1.5 6.1 19.2 0.0
96.3 85.3 78.8 100.0
98.8 95.3 123.0 102.9
0.0 45.7 0.0 0.0
16.6 0.0 0.0 5.9
7.8 0.0 44.7 0.0
42.7 59.7 90.3 79.4
5mM C0CI2 5mM CuCl2 5mM ZnCl2 5mMMgCl2 5mM CaCl2 5mMMnCl2 5mM FeCl2 5mM FeCls 5mMNiS04


f) Molecular weight and subunit
Enzymes A, A', A" and B purified fi"om P. putida transconjugants consist of one type of unit with the molecular weight of about 64,000, 62,500, 62,500 and 60,000, respectively, as measured by sodium dodecyl sulfate polyacrylamide gel electrophoresis. They can be heterodimers consisting of any two units of Enzymes A, A', A" and B when Enzymes A, A', A" and B genes/DNA sequences are expressed in the same host.
g) N-terminal amino acid sequence
N-terminal sequences of mature Enzymes A and B are
Enzyme A : Gln-Val-Thr-Pro-Val-Thr—-
Enzyme A" : Blocked N-terminal residue
Enzyme B : Gln-Val-Thr-Pro-Ile-Thr-Asp-Glu-Leu-Leu-Ala—. , N-tenninus of the mature Enzyme A' is not determined because of an insufficient purity of the sample.
(3) Production of the AADHs

Cells are harvested from the fermentation broth by centrifugation or filtration. The cells are suspended in the buffer solution and disrupted by means of a homogenizer, sonicator or treatment with lysozyme and the like to give a disrupted solution of cells.
AADHs are isolated and purified from a cell free extract of disrupted cells, preferably from the soluble fraction of the microorganisms by usual protein purification methods such as anamonium sulfate precipitation, dialysis, ion exchange chromatographies, gel filtration chromatographies, and affinity chromatographies.
(4) Enzyme reaction
Enzyme reaction was performed at pH values from about 6.0 to about 9.0 at the temperature of about 10°C to about 50°C, preferrably of 20°C to 40°C in the presence of an electron acceptor, for example, DCIP, PMS, Wurster's blue, fenicyanide, coenzyme Q, cytochrome c and the like in a buffer such as Tris-HCl buffer, phosphate buffer and the like. The concentration of the substrate in a reaction mixture can vary depending on the other reaction conditions but, in general, is desirable to be about 1 - 200 g/1, most preferably from 1 -100 g/1.
In the enzyme reaction, AADHs may also be used in an immobilized state with an appropriate carrier. Any means of inomobilizing enzymes generally known to the art may be used. For instance, the enzyme may be bound directiy to membrane granules or the like of a resin having functional group(s), or it may be bound through bridging compounds having functional group(s), for example, glutaraldehyde, to the resin.
The polypeptides provided by the present invention also include the derivatives prepared from the genes of AADHs consisting of Enzyme A, Enzyme A', Enzyme A",

and Enzyme B and the relatives prepared from the gene-homologues resulting fix)m degeneracy of the genetic codon or any sequence of natural, synthetic or recombinant origin which has significant homology to the AADH genes. The derivatives can be functional mutants of the polypeptides identified by SEQ ID N05, SEQ ID N06, SEQ ID N07 and SEQ ID NO 8 which contain addition, deletion and/or substimtion of one or more amino acid residue(s), wherein the said enzymatic polypeptide have alcohol and/or aldehydes dehydrogenase activity. The mutant genes can be prepared by treating AADH genes with a mutagen such as uluraviolet irradiation, X-ray irradiation, y-ray irradiation or contact with a nitrous acid, N-methyl-N'-nitro-N-nitrosoguanidine (NTG), or other suitable mutagens, or isolating a clone occurring by spontaneous mutation or by standard methods of in vitro mutagenesis known in the art.
The derivatives of the AADH polypeptides also include chimeric recombinant enzymes between polypeptides identified by SEQ ID N05, SEQ ID N06, SEQ ID N07 and SEQ ID NO 8. The chimeras can be prepared by combining two or more parts of DNA sequences of the present invention in vitro at the conserved restriction site in both sequences with restriction enzymes and T4-ligase as shown in Fig. 6, or by recombining two AADH genes in vivo at the conserved nucleotide sequences in both genes as shown in Fig. 8.
The derivatives of the AADH polypeptides also include polypeptides with additional polypeptides at the N-terminal, C-terminal and/or inside region of the AADH polypeptides. Enzyme B, Enzyme A/B25, and Enzyme A/B3 fused with cytochrome c polypeptides (17 -18 kDa) of G. oxydans DSM 4025 at the C-terminus showed comparable AADH activities with Enzyme B described in Example 4 in the conversion of D-sorbitol to L-sorbose, Enzyme A/B25 and Enzyme A/B3 both described in Example 14 in the conversion of L-sorbose to 2KGA. Thus, a relatively long

polypeptide can be added or inserted to the AADHs provided by the present invention with remaining AADH activity.
The derivatives of the AADH polypeptides described above possibly show preferrable characteristics such as a desired substrate specificity, higher affinity to a substrate, lower affinity to an inhibitory compound, higher stability against temperature and/or pH, and higher catalytic speed. As described in the working examples below, such derivatives would improve the productivity of the desired products.
The enzymatic polypeptides of the present invention are usually produced in the form of dimers. Such dimers contain homodimers of Enzyme A, A', A" or B, or the derivatives including chimeras, and heterodimers consisting of two different enzymatic polypeptides mentioned above. Thus the recombinant enzyme preparation of the present invention also contain one or more of said homodimers and/or heterodimers.
The recombinant organisms provided by the present invention are highly useful for the production of the recombinant enzyme preparations of AADHs having an alcohol and/or aldehyde dehydrogenase activity. Said organisms are also useful for the production of aldehydes, carboxylic acids and ketones, especially, 2KGA by utilizing said recombinant enzyme preparations, and by utilizing said recombinant orgamsms.
The production of 2KGA with the said recombinant organisms can be performed in the fermentation with the medium and culture conditions as described above. The production of 2KGA may be performed with the recombinant organisms described above together with the concomitant organisms such as E. coli, P. putida and Bacillus megaterium.

Examples
Example 1. Cloning of AADH genes
(1) Construction of a genomic library of G. oxydans DSM No. 4025
Chromosomal DNA was prepared as follows. G. oxydans DSM No. 4025 was cultivated on an agar plate containing 20 ml of NS2 medium consisting of 5.0% D-mannitol, 0.25% MgS04-7H20,1.75% com steep liquor, 5.0% baker's yeast (Oriental Yeast Co., Osaka, Japan), 0.5% CaC03,0.5% urea (sterilized separately) and 2,0% agar (pH 7.0 before sterilization) at 27°C for 3 days. The cells were collected from the agar plate, washed with 10 ml of 10 mM Tris-HCl buffer (pH 8.0) containing ImM EDTA and resuspended in 5 ml of 10 mM Tris-HCl buffer (pH 8.0) containing 20 mM EDTA. The cell suspension was treated with lysozyme (Sigma Chemicals Co., SL Louis, Mo., USA) at a final concentration of 400 fig/ml at 37°C for 30 minutes, then with pronase (400 units) at 37°C 30 minutes and with 1% SDS at 37°C for 1 hour. Chromosomal DNA was treated with phenol and RNase A (Boheringer Mannheim, GmbH, Mannheim, Germany) according to the method described by Maniatis et al. (Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y., (1982). Chiomosomal DNA (200 ^ig) was digested with 168 units of Sail (Boehringer Mannheim) at 37°C for 5 to 90 minutes. The resulting partially digested fragments of 15 - 35 kb were isolated by preparative agarose gel electrophoresis (agarose: 0.7%); the gel piece containing the desired fragments was cut out and the DNAs were electro-eluted from the gel into TAE buffer consisting of 40 mM Tris-acetate and 2 mM EDTA. Thus, 40 ^g of the DNAs were obtained. In parallel, 8 ^ig of the cosmid vector pVK102 (ATCC 37158) was completely digested with Sail and treated with calf intestine alkaline phosphatase (Boehringer Mannheim)

according to the supplier's recommendation. pVK102 (0.4 |j.g) was ligated with the 15-35 kb San. fragments (0.2 - 2 |ig) by the ligation kit (Takara Shuzo Co. Ltd., Kyoto, Japan) at 26°C for 10 minutes. The ligated DNAs were then used for in vitro packaging according to the method described by the supplier (Amersham): mixing the ligated DNAs with the phage coat protein parts. The resulting phage particles were used to infect E. coli ED8767 (Murray, N. E., W. J. Brammar and K. Murray. Mol. Gen. Genet., JiQ:53-61,1977). About 3,000 Km^ TcS colonies were obtained and all of the colonies tested (24 colonies) possessed the insert DNAs; its average size was 26.5 kb. Another cosmid library of G. oxydans DSM No. 4025 containing 55,000 clones was constructed by using chromosomal DNA of G. oxydans DSM No. 4025 partially digested with EcdRl and inserting them into the£coRI site of pVKKX) by almost the same method described above. All of the colonies tested (24 colonies) possessed insert DNAs (average size; 27 kb).
These two cosmid libraries in E. coli ED8767 were then transferred into E. coli S 17-1 (Tra+, Bio/Technology, 1:784-791,1983) by using the mixture of recombinant plasmid DNAs extracted from E. coli ED8767 libraries. About 4,000 Km^ transformants of E. coli S17-1 were picked up, cultivated individually in microtiter plates containing 1(X) |il of LB consisting of 10 g/1 of Bactotrypton (Difco), 5 g/1 of yeast extract (Difco) and 5 g/1 of NaCl supplemented widi 50 |ig/ml kanamycin at 37°C, and stocked with 15% glycerol at -80°C as cosmid libraries in E. coliSll-l.
The G. oxydans DSM No. 4025-Sfl/I and -£coRI cosmid libraries were constmcted in E. coli SI7-1. From the library, 1,400 clones were individually transferred from £. co/i S17-1 intoF.pMn^iaATCC 21812 by conjugal mating. 1,400 cultures stocked in microtiter plates at -80°C were thawed and transferred to microtiter plates containing 100 |il of fresh LB medium in each well with a plate transfer cartridge (Nunc) and cultivated at 37°C overnight. Nalidixic acid resistant (Nal'^ P. putida

ATCC21812 was cultivated at 30°C overnight in 100 ml of MB medium consisting of 2.5% mannitol, 0.5% yeast extract (Difco Laboratories, Detroit, Mich.) and 0.3% Baaotryptone (Difco). Fifty |Xl of the P. putida culture was individually added to the 1,400 wells containing cultures of the cosmid library. The 1,400 cell mixtures were spotted with plate transfer cartridges onto nitrocellulose filters placed on the surface of FB agar medium consisting of 5% finctose, 1% yeast extract (Difco), 1% polypeptone (Daigo Eiyo, Japan) and 1.8% agar and cultivated at 27°C overnight. Nalidixic acid was used for the counter-selection of transconjugants against donor E. coli. The cells grown on the filters were individually streaked onto MB agar medium containing 50 |ig/ml of nalidixic acid and 50 (Xg/ml of kanamycin hereinafter referred to as (MNK agar plate) and incubated for 4 days at 27°C for the selection of transconjugants. The resulting colonies were purified by streaking on MNK agar plates as mentioned above. Thus, 1,400 transconjugants of P. putida [gene library of G. oxydans DSM No. 4025 in P. putida] were prepared.
(2) Immunological screening of clones of the AADH gene of G. oxydans DSM No. 4025.
At first, 350 transconjugants (175 from Sail library and 175 firom fcoRI library) maintained MNK agar plates were individually cultivated in test tubes containing 5 ml of MNK medium. The cells were collected from 1.5 ml of each broth and treated for Westem-blot analysis as follows. The cells were suspended in 50 |il of Laemmli buffer consisting of 62.5 mM Tris-HCl, pH 6.8,10% glycerol, 5% mercaptoethanol and 2% SDS. The cell suspension was boiled for 3 minutes, and 10 |J1 of the cell lysate was applied on SDS-PAGE. The resulting protein bands were then electro-blotted to a nitrocellulose filter by an electroblotting apparatus (Marysol Industrial Co., Ltd.) operated at 40 V, 200 mA for 16 hours in 2.5 mM Tris-19.2 mM

glycine buffer, pH 8.6, containing 20 % methanol. The filter was, then, incubated for 1 hour in 3% gelatin in TBS buffer consisting of 20 mM Tris, pH 7.5, and 500 mM NaCl. After a brief rinse in TTBS buffer consisting of 20 mM Tris, pH 7.5,500 mM NaCl and 0.05% Tween 20, the filter was incubated for 1 hour with a first-antibody which contained l:500-diluted anti-AADH antibody in TTBS buffer containing 1% gelatin. The anti-AADH antibody had been prepared by mixing the AADH proteins purified fi-om G. oxydans DSM No. 4025 with incomplete adjuvant, injecting the resulting mixture into a white rabbit twice with 2 weeks' interval, collecting whole blood 1 week after the second injection and preparing the serum firaction as the anti-AADH antibody. Then, the filter was washed twice (5 min each) in TTBS buffer and incubated for 1 hour in a second-antibody (goat anti-rabbit IgG-horseradish peroxidase conjugate) solution which contained l:3,000-diluted second antibody in TTBS containing 1% gelatin. After washing in TTBS buffer twice and in TBS once, the filter was inmiersed in a color developing solution until blue bands became visible with Konica Immunostaining HRP Kit IS-50B (Konica, Tokyo, Japan) according to the supplier's recommendation. For an actual screening, five cell lysates were mixed and applied to one well for the first Westem-blot screening. Out of 70 mixtures, 14 exhibited positive bands; nine samples had immuno-reactive proteins of approximate Mr 64,000, but two of these exhibited weak signals; one had an immuno-reactive protein of approximate Mr 60,000; and four samples had immuno-reactive proteins of Mr 55,000.
Seven mixture samples showing strong signals at Mr 64,000 were individually subjected to a second Westem-blot screening to identify the clone in each mixture. One positive clone per one mixture samples was identified; plasmids of the seven clones were designated as p6E10, pl6C8, pl6F4, pl7E8, plE2, p24D4, and p26C3, respectively. By restriction enzyme analysis, it was found that four plasmids, p6E10, pl6C8, pl6F4, and pl7E8, carried the same DNA region and the other three carried different regions from that of the former four plasmids.

(3) Screening of the AADH genes from the cosmid libraries by colony-blot and
Southem-blot hybridization
To find the other AADH genes besides the genes obtained by the immunological screening as described above, the whole cosmid libraries of G. oxydans DSM No. 4025 in E. coli ED 8767 (Sa/I-library and EcoRI-libraries) were screened by colony-and Southem-blot hybridization with a 0.9 kb SaR fragment of p24D4. The 0.9 kb SaII. fragment hybridized with a oligonucleotide probe,
ATGATGGT(GATC)AC(GATC)AA(TC)GT synthesized according to an intemal amino acid sequence of the natural AADH enzyme purified from G. oxydans DSM No. 4025, MetMetValThrAsnValAspValGlnMetSerThrGlu, which was obtained by digestion and sequenced by automatic gas-phase sequencer (Applied Biosystems 470A). The cells of the cosmid libraries were appropriately diluted and spread on LK agar plates, and the resulting colonies were blotted onto nylon filters and were analyzed by hybridization with the 32p-labeled 0.9 kb Sali fragment About 1% of the colonies showed positive signals; 41 colonies were selected from the SaU. library and 20 from EcoRI. library, and they were subjected to restriction enzyme analysis, followed by Southem-blot analysis. Six different AADH gene-related DNA regions were isolated in this screening as follows: four already-isolated regions carried on p24D4, plE2, p26C3 and, pl7E8, and two new regions carried on two separate plasmids designated as pSS31 and pSS53. The other plasmid pSS33 carried both of the two regions which were carried on p24D4 and pSS31.
(4) Immunological and enzymatic characterization of AADH clones

Westem-blot analysis of cell lysates of P. putida carrying p24D4, plE2, p26C3, pSS3l and pl7E8 showed that the five clones encoded proteins with molecular weights of about 64,000, 62,500, 62,500, 60,000 and 62,000, respectively. Plasmid pSS33 encoded two immuno-reactive proteins with molecular weights of about 64,000 and 60,000, whereas pSS53 did not produce any immuno-reactive proteins.
Enzyme activities of each clone (cell free extract, soluble fraction and membrane fraction) were measured by photometric analysis. The cells of each clone were inoculated in 5 ml of MB medium in a test tube and cultivated at 30°C for 24 hours. The resulting broth was transferred into 200ml of fresh MB medium in 500 ml flask and the flask was shaken on the rotary flask shaker at 30°C for 24 hours. The cells were collected by centrifugation at 6,(X)0 x g for 10 minutes and washed with 40 ml of cold buffer consisting of 50 mM Tris-HCl, pH 7.5, 5mM MgCl2 and 0.5 mM phenylmethylsulfonyl fluoride and suspended with the same buffer to prepare cell suspension of 1 g wet cells per 5ml. The cell suspension was subjected twice to a French press cell disrupter (1,5(X) kg/cm^) and the resulting homogenate was centrifuged at 6,000 x g for 10 minutes to remove cell debris. Thus obtained cell free extract (CFE) was centrifuged at 100,000 x g for 60 minutes. The resulting supematant and pellet were collected as the cytosol fraction and the membrane fraction, respectively and subjected to PMS-DCIP assay as follows. The enzyme reaction mixture (1.0 ml) contained 100 ^iM DCIP, ImM PMS, 50 mM Tris malate-NaOH buffer, pH 8.0, a substrate and the enzyme (10 ]xl). Substrate-dependent decreasing rate of absorbance of DCIP at 600 nm was measured at 25°C by using a Kontron spectrophotometer UVIKON 810. Table 6 shows the level of enzyme activities in the cell free extract and the soluble fractions of the clones. According to the substrate specificity, the enzyme encoded on each plasmid was classified into large three groups, A-, B- and C-groups: A-group catalyzes the oxidation of L-sorbose, D-sorbitol and 1-propanol; B-group

v. '
catalyzes the oxidation of D-glucose and D-sorbitol; C-group showed no clearly detectable activities on the substrates used. In the A-group, there were three types. A, A' and A" each of which was distinguished from each other by their physical map of the DNA canied on each plasmid. B- or C-group each consisted of only one type of protein derived from one region of the chromosomal DNA.

Example 2. Nucleotide sequencing
Nucleotide sequences of the genes for Enzymes A, A', A" and B were determined with the plasmids, p24D4, plE2, p26C3, and pSS31, respectively, by the

dideoxynucleotide chain teimination method using M13mpl8 and M13mpl9 (Boehringer Mannheim). One open reading firame (ORF) for each gene was found; the nucleotide sequences of the four genes arc shown in the sequence list SEQ ID No. 1 to 4 and the amino acid sequences deduced fiom the nucleotide sequences were shown in the sequence list SEQ ID No. 5 to 8. The ORFs for Enzymes A, A', A" and B genes are 1737,1737,1734, and 1737-bp long and encode 579,579,578 and 579 amino acid residues all including 23 amino acid of signal sequences. The homologies between Enzymes A, A', A" and B are shown in Table 7.

Figure 5 shows the amino acid sequences of mature Enzyme A and Enzyme B which are aligned so as to be comparable.
Homology search of Enzymes A, A', A" and B rcvealed that Enzymes A, A', A" and B showed rather low homology (26-31% homology through the polypeptides) with several quino-proteins including alcohol dehydrogenase of Acetobacter aceti (T. Inoue et al., J. Bacteriol. 121: 3115-3122) or Acetobacter polyoxogenes (T. Tamaki et al., B. B. A., 1088:292-300'). and methanol dehydrogenase oiParacoccus denitrificans (N. Harms et al., J. Bacteriol., 169: 3966-3975), Methylobacterium organophilum (S. M. Machlin et al., J. Bacteriol., JLZQ: 4739-4747), as: Methylobacterium extorquens (D. J. Anderson et al.. Gene 2Q: 171-176).

Example 3. Subclonin^ of AADH genes
Enzyme A gene was originally cloned as a cosmid clone of p24D4 which has about 25 kb insert in £coRI site of pVKlOO. Then, it was further subcloned to use as an Enzyme A gene cassette. The 2.7 kb EcoRV fragment which includes ORF of Enzyme A gene with about 500 bp of non-coding regions at the both ends was excised from 3.4 kb Nrid fragment, which was isolated from p24D4 in Ml 3 mp 18, and was ligated to Hindm site of pUC18 with HindHI linker (CAAGCTTG). The resulting plasmid was designated pSS A202. Enzyme A gene cassette (2.7kb HindM fragment) was then inserted at Hindni site of pVK102 to produce pSS A102R. The plasmid pSSA102R was introduced into nalidixic acid resistant P. putida [ATCC 21812] by a conjugal mating method as described in Example 1-(1). The transconjugant of P. putida carrying pSSA102R was selected on MB agar medium containing 50 p-g/ml nalidixic acid and 10 |ig/ml tetracycline (MNT agar medium) and subjected to a mini-resting cell reaction. The reaction mixture (100 \il) consisting of 20 g/l L-sorbose, 3 g/1 NaCl, 10 g/1 ClaCOs and the cells collected from the MNT agar culture with a toothpick
was incubated at room temperature with gentie shaking for 24 hours. The reaction mixture was assayed with TLC and 2KGA was identified as the product, while no 2KGA was observed by the same resting cell reaction with the host, nalidixic acid resistant P. putida [ATCC 21812].
Enzyme B gene was originally cloned as a cosmid clone of pSS31 which has about SOkb insert in Sail site of pVK102. It was subcloned as 6.5kb BglR fragment into Bgm site of pVKlOl (ATCC 37157) to obtain pSSB 102. Then, it was further subcloned to use as a Enzyme B gene cassette. The 6.5 kb BglR fragment was cloned into BamiU site of pUC18 to obtain pSSB202. Then, 2.3 kb XhoTl fragment was excised from pSSB202. The 2.3 khXhoR fragment includes ORF of Enzyme B with 120 bp of 5'-noncoding region and about 500 bp of 3'-noncoding region. The fragment was treated with Klenow fragment to fill-in the cohesive ends and cloned into

Hindm site of pUC18 with Hindm linker to produce pSSB203. The Enzyme B gene cassette (2.3 kb Hindm fragment) was inserted at HindHI site of pVK102 to make pSSBlOSR. The plasmid pSSBlOSR was introduced into nalidixic acid resistant P. putida [ATCC 21812] by a conjugal mating method, and the transconjugant of P. putida carrying pSSBlOSR was selected on MNT agar medium and subjected to a mini-resting cell reaction. P. putida carrying pSSBlOBR showed the Enzyme B actiuvity (L-sorbose formation from D-sorbitol) in the resting cell reaction. (Incidentally, XhoII fragment was found not to be SiXholI-XhoIl fragment, hut SiXhoIl-XhoI fragment as a result of nucleotide sequencing. Xhol might be present in the XhoU preparation.)
Enzyme A' and Enzyme A" genes were originally cloned as a cosmid clone of plE2 and p26C3 which have about 30 kb insert in Sail site of pVK102 and further subcloned basically as described above. Enzyme A' gene in 3.5 kb Xholl fragment was subcloned in BglR site of pVK102 to construct pSSA'lOlR, and Enzyme A" gene in 2.7 kb EcoRY fragment was first subcloned into M13mpl9 and then re-subcloned between HindDI andBglU sites of pVK102 to construct pSSA"102.
Example 4. Isolation and characterization of AADHs from transconiugants of P.
pmida-
(1) Cultivation of microorganisms
P. putida [ATCC 21812] carrying cosmid vector pVK102 containing the Enzyme A, A', A" and B genes; pSSA102R, plE2, p26C3 and pSSB103R, respectively, were cultivated in MB broth in the presence of antibiotic. Antibiotics added into medium were as follows; 5 |i.g/ml tetracycline for pSSA102R (Enzyme A) and pSSB103R (Enzyme B), 25 ng/ml kanamycin for plE2 (Enzyme A') and p26C3 (Enzyme A"). From the agar plate of MB containing the respective antibiotic, tiie cells

were inoculated in 10 test tubes containing 5 ml MB medium with the respective antibiotic and cultivated with shaking at 30°C. After 2 days of cultivation, the cells were transferred to ten 500 ml-Erlenmeyer flasks containing 100 ml of the same medium and cultivated with shaking at 30°C. After 1 day of cultivation, the seed cultures were combined and transferred to 18 liters of tiie medium in 30 L jar fermenter (Marubishi) and cultivated for 18 hours with 300 rpm agitation and 1.0 wm aeration at 30°C. The cells were harvested by centrifuge at 6,000 x g for 10 minutes, washed once witii 1.5 liters of 25 mM Tris-HCl, pH 7.5, containing 5 mM CaCl2,1 mM MgCl2, 0.2 M NaCl, 2.5% sucrose, and 0.5 mM PMSF and stocked at -20'C until use. As a result, about 150 g wet weight cells were obtained.
(2) Purification of the cloned Enzymes A, A', A", and B.
Purifications of the Enzymes A, A', A" and B were carried out by the same procedure with almost the same scale. All operations were carried out at 4 - 10°C unless otherwise stated. The enzyme activity determination for Enzyme A, A', A" and B were carried out with the substrates, L-sorbose, n-propanol, n-propanol and D-glucose, respectively, by spectrophotometric assay as described in Example 1 throughout the purification steps. The cells (about 100 g wet weight cells containing 8 -10 g of total proteins) were thawed and suspended in about 200 ml of 25 mM Tris-HCl, pH 8.0, and disrupted by passing tiirough French press (1500 kg/cm^) twice. Then, DNase and MgCl2 were added to the suspension at the final concentration of 0.01 mg/ml and 1 noM, respectively, to reduce viscosity of the solution due to DNA. Cell debris was removed by centrifugation at 6,000 x g for 10 minutes. The suspension was filled up to 240 ml with the 25 mM Tris-HCl buffer, pH 8.0, and centrifuged at 100,000 X g for 90 minutes to remove insoluble membrane firaction. The soluble supematant was filled up to 240 ml with the Tris buffer and, then, pyrroloquinoline quinone (PQQ) and CaCl2 were added at the final concentration of 12.5 |iM and 5 mM,

respectively, and the solution was stirred vigorously for 15 minutes at room temperature. The soluble fraction prepared as above was fractionated by (NH4)2S04. The fraction 35 - 60%-saturated (NH4)2S04 was precipitated and resuspended in 100 ml of 25 mM Tris-HCl buffer, pH 8.0, containing 5 mM CaQ2, and 5% sucrose and, then, PQQ was added again at the final concentration of 12.5 |iM. The enzyme solution was dialyzed against 1000 ml of the same buffer (without PQQ) overnight. Twenty grams of solid polyethylene glycol #6000 was added to the dialysate slowly with gentie stirring. After stirring for 30 minutes, precipitates were removed by centrifugation at 10,000 X g for 20 minutes, and the supernatant was filled up to 200 ml with the buffer indicated as above.
The enzyme solution prepared as above was purified by following three chromatography steps.
The first step: DEAE-Toyopearl 650M
The cmde enzyme solution was subjected to a column of DEAE-Toyopearl 650M (2.5x 40 cm) which had been equilibrated with 25 mM Tris-HCl buffer, pH 8.0, containing 5 mM CaCl2, and 5% sucrose. The column was washed with 400 ml of the same buffer and the enzyme was eluted by 2,000 ml of 0 - 0.5 M NaCl linear gradient in the buffer at a flow rate of 150 ml/hour. The enzyme active fractions were pooled and diluted 2-fold with the buffer without NaCl.
The second step: Q-Sepharose (Fast Flow)
The enzyme solution was subjected to a column of Q-Sepharose (Fast Flow) (1.5 X 20 cm) which had been equilibrated with the buffer without NaCl. The column was washed with 200 ml of the buffer containing 0.2 M NaCl and the enzyme was

eluted by 600 ml of 0.2 - 0.6 M NaCl linear gradient in the buffer at a flow rate of 50 ml/hour. The enzyme active fractions were pooled and concentrated to 2.5 ml by using ultrafilterrAmicon, PM-30 under N2 gas.
The third step: Sephacryl S-300 HR (gel filtration)
The concentrated enzyme was filtrated by a column of Sephacryl S-300 HR (2.5 X 100 cm) which had been equilibrated with 25 mM HEPES, pH 7.5, containing 5 mM CaCl2, 5% sucrose, and 0.2 M NaCl. The column was developed by the same buffer at a flow rate of 20 ml/hour. The enzyme active fractions were pooled and concentrated to below 1 ml by the ultrafilter mentioned above and, then, stocked at -SO°C. The enzymes concentrated in the HEPES buffer was stable for at least 2 months at -80°C.
Consequently, 26.0 mg of Enzyme A, 0.35 mg of Enzyme A', 0.41 mg of Enzyme A", and 5.0 mg of Enzyme B were obtained.
(3) Properties of the Enzymes A, A', A" and B.
a) Molecular weight and subunit.
The Enzymes A, A', A" and B were eluted at the same position from the same gel filtration column on Sephacryl S-300HR under the same condition. The molecular weight of the enzymes was estimated as approximately 135,000 comparing with the molecular weight standard proteins (SDS-PAGE Standards, Low Range, Bio-Rad Laboratories, Richmond, CA, USA). The Enzymes A, A', A" and B showed homogeneous single bands on SDS-PAGE analysis with molecular weights of 64,000, 62,500,62,500 and 60,000, respectively. All the Enzyme bands A, A', A" and B were detected on Western blotting analysis using anti-AADH rabbit serum. Therefore, it was

concluded that the enzymes consisted of two identical subunits as an homo-ditneric form.
b) N-terminal amino acid sequence and amino acid composition.
N-terminal amino acid sequences of the mature Enzymes A, A" and B were analyzed with automatic gas-phase sequencer (470A; Applied Biosystems) by Edman metiiod [Acta Chem. Scand., 4,283-293, {1950)]. The analysis of the Enzyme A' was not done because of an insufficient purity of the sample. The results were as follows:
Enzyme A : Gln-Val-Thr-Pro-Val-Thr—-
Enzyme A" : Blocked N-terminal residue
Enzyme B : Gln-Val-Thr-Pro-ne-Thr-Asp-Glu-Leu-Leu-Ala—. The determined sequences of Enzyme A and B were identical to the sequences (starting from the twenty-fourth residues) deduced from the nucleotide sequences described in SEQ ID No. 5 and 8; these results indicate that the initial 23 residues of the enzymes are the signal sequences. By analogy of the Enzymes A and B, the first 23 residues of Enzyme A' and A" are also deduced to be the signal sequences.
The amino acid composition of the Enzyme A was determined. The protein was hydrolyzed with 6 N HCl at 110°C for 24 hours or 4 M methanesulfonic acid (after oxidation with peiformic acid) at 115°C for 24 hours. Amino acid analysis was performed by using Kontron amino acid analyzer (ninhydrin system). The analytical data were compared with the amino acid composition deduced fix)m the DNA sequence of Enzyme A gene. It indicated that the purified Enzyme A was certainly a product of the Enzyme A gene.
c) Substrate specificity

The Enzymes A, A', A" and B were characterized by their substrate specificities on PMS-DCIP assay as described above using 8 substrates, n-propanol, isopropanol, D-glucose, D-sorbitol, L-sorbosone, D-mannitol, L-sorbose, and D-fructose. The resuhs were indicated in Table 1.
d) Physicochemical property
Physicochemical studies of optimal pH, pH stability and thermal-stability, of the Enzymes A (as L-sorbose dehydrogenase activity). A' (as n-propanol dehydrogenase activity). A" (as L-sorbose dehydrogenase activity) and B (as D-sorbitol dehydrogenase activity), were performed by the PMS-DCIP assay.
Table 2 summarizes the results of optimal pH of the enzymes. The enzyme activity was assayed by the PMS-DCIP spectrophotometric assay using various pH buffers. The buffers were 50 mM Tris-malate-NaOH, pH 6.0, 6.5, 7.0,7.5, 8.0 and 8.5; 50 mM glycine-NaOH, pH 9.0 and 9.5. The extinction coefficients of DCIP at pH 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, 9.0 and 9.5 were taken as 10.8, 13.2, 14.5, 14.9, 15.0, 15.1,15.1 and 15.1, respectively. All the enzymes showed their optimal points at pH 8.0 - 8.5. The Enzymes A" and B had relatively wide pH range toward lower pH, compared with the Enzymes A and A'.
Table 3 indicates the results of pH-stabilities of the enzymes. The enzyme (about 0.01 mg/ml) was incubated with 50 mM buffer containing 5% sucrose, 0.2 M NaCl, and 5mM CaCl2 at 25°C for 3 hours and assayed by PMS-DCIP spectrophotometric method. The buffers were Na-acetate, pH4 and 5, Tris-malate-NaOH, pH 6,7 and 8, glycine-NaOH, pH 9 and 10. The values in tiie table are expressed as relative activity against that obtained by no incubation at pH 8.0. The

substrates used for the enzymes were 125mM L-sorbose for Enzymes A and A", 50 mM n-propanol for Enzyme A', and 125 mM D-sorbitol for Enzyme B. Profiles of pH-stabilities of Enzymes A, A', A", and B were almost the same; they were stable at the range of pH 6 to 9.
Table 4 indicates the results of thermal-stabilities of the enzymes. The enzyme (about 0.05 mg/ml) in 25 mM HEPES buffer, pH 7.5, containing 5% sucrose, 0.2M NaCl, and 5 mM CaCl2 was incubated at temperature indicated in the table (4 - 60°C) for 5 minutes, cooled in ice bath and assayed by PMS-DCIP spectrophotometric method. Remaining activity was expressed as relative activity against that obtained by 4°C incubation. The substrates used for the enzymes were 125mM L-sorbose for Enzyme A and Enzyme A", 50 mM n-propanol for Enzyme A', and 125 mM D-sorbitol for Enzyme B. After the treatment of the enzymes at 40°C for 5 min, the residual activity of Enzyme A was 20%, and those of Enzymes A', A", and B were 70 - 85%.
e) General inhibitors
The enzyme (about 0.05 mg/ml) in 25 mM HEPES buffer, pH 7.5, containing 5% sucrose was incubated with metal or inhibitor for 30 minutes at 25 °C. Remaining activity was assayed by PMS-DCIP spectrophotometric assay as described in Example 1. Remaimng activity is expressed as relative activity against blank incubation. Effects of metal ions on the enzymes are listed in Table 5. MgCl2 and CaCl2 were nearly inert to the enzymes, while the other metal ions, especially CuCl2, significantly affected. Effects of inhibitors on the enzymes are also included in Table 5. EGTA and EDTA inhibited the Enzyme A, A' and A", remarkably. However, Enzyme B was less inhibited than the A group enzymes by EDTA and EGTA.

Example 5. Efficient production of Enzvme B in E. coli
The signal peptide region of the Enzyme B was replaced with that of maltose binding protein (malE) of £. coli as follows. Two oligonucleotides (SEQ ID No. 9 and 10) were synthesized with Applied Biosystem 381A DNA synthesizer and annealed to generate a double-stranded DNA fragment encoding a amino acid sequence (SEQ ID No. 11), MetLysIleLysThiGlyAlaArglleLeuAlaLeuSerAlaLeuThrThrMetMetPheSer AlaSerAlaLeuAla(Gln), which was, then, treated with T4 polynucleotide kinase [J. Biol. Chem., 252.10606-10613, (1984)]. pSSB203 (see Example 3) was digested with the restriction enzyme Sphl, treated with T4 DNA polymerase and digested with BstPl. The resulting 1.72 kb DNA fragment carrying Enzyme B gene without the region coding for the original signal sequence and the first amino add residue (Ghi) of the mature Enzyme B was isolated from an agarose gel after agarose gel electrophoresis. The E. coli expression vector, pTrc99A (Phaimada Co., Uppsala, Sweden), which was digested with the restriction enzymes Ncol (at ATG start codon) and SmdL was ligated with above two DNA fragments. The resulting plasmid was designated as pTrcMal-EnzB and used to transform E. coli JM109. The transfonnant was grown in two 2-liter flasks each containing 600 ml of LB with 100 \ig/wl ampicillin at 28°C and IPTG was added to 0.1 mM when cell concentration reached at about 1.5 OD600. Following the addition of IPTG, the cells were cultivated for an additional 3-4 hours. The cells were harvested by centrifugation (4,000 x g) at 25°C for 10 minutes, suspended with 500 ml of 30 mM Tris-HCl, pH 8.0, containing 20% sucrose at 25°C. After EDTA was added to 1 mM into the cell suspension, the cells were incubated with gentle shaking for 5 minutes at 25*'C and collected by centrifugation (8,000 x g) at 4°C for 15 minutes. The cells were resuspended with 500 ml of ice cold 5 mM MgS04 solution and incubated with gentle shaking for 5 minutes at 4°C. The cell suspension was centrifuged at 8,000 x g for 10 minutes at 4°C to obtain a supernatant as a cold osmotic shock extract, which was found to contain the

Enzyme B protein (a molecular weight of 60,0(X)) with the purity more than 50 - 60% by SDS-PAGE analysis. The supernatant was first supplemented with Tris-HCl, pH 8.0, to 20 mM, and incubated at 25°C firstly with EDTA at 10 mM final concentration for 10 min, secondly with CaCl2 at 20 mM final concentration for 10 minutes and lastly with PQQ at 25 μM final concentration. For stabilization of the enzyme, a-methyl-D-glucoside (a competitive inhibitor) was added to 20 mM final concentration in the supematant. The Enzyme B was completely purified by following two chromatographies. At first, the supematant was loaded onto a Q-Sepharose column (1.6 X 12 cm) which had been equilibrated with 20 mM Tris-HCl, pH 8.0, containing 1 mM CaCl2 and 20 mM a-methyl-D-glucoside, and the Enzyme B was eluted with 600 ml of 0 - 0.4 M NaCl linear gradient in the same buffer. A red protein peak eluted at about 0.25 M NaCl was collected and concentrated to about 0.5 ml by Centricon-30 (Amicon). Finally, the Enzyme B was passed through a SephacrylS-300HR column with 20 mM HEPES, pH 7.8, containing 0.2 M NaCl, 1 mM CaCl2 and 20 mM a-methyl-D-glucoside. A red protein peak eluted around a molecular weight of 135,000 daltons position was collected as the final purified Enzyme B. Consequently, about 8 mg of the purified Enzyme B was obtained from 1.2 liters cultivation broth of E. coli.
Example 6. Host-vector system
A host-vector system for G. oxydans [DSM No. 4025] was established by using the conjugal mating system with a broad-host-range cosmid, pVK102. Initially, only one transconjugant was isolated from G. oxydans [DSM No. 4025] having nalidixic acid resistance. A new host, GOS2, was isolated from the transconjugants, G. oxydans[DSM No. 4025] carrying pVK102 by curing pVK102. A second host, GOS2R, was then derived from the G0S2 by adding rifampicin resistance (100μg/ml), which enables easy selection of the transconjugants from the donor E. coli. The

plasmid transfer frequency into G0S2R was 10'^ ~ 10'^ transconjugants/recipient The 2KGA productivity of GOS2R, however, was about 10% lower than that of G. oxydans [DSM No. 4025]. The third host, GORS6-35, was obtained from G. oxydans [DSM No. 4025] by selecting the strain with rifampicin resistance, high 2KGA productivity and relatively high competence through a series of experiments, including the conjugation, curing and 2KGA fermentation.
(1) Isolation of G0S2
Resistance to nalidixic acid was added to G. oxydans [DSM No. 4025]. Cells of G. oxydans [DSM No. 4025] were streaked onto Trypticase Soy Broth (BBL, Becton Dickinson Microbiology Systems Cockeysville, MD. USA) (T) agar medium with 50 |ig/ml of nalidixic acid (TN agar medium) and incubated at 27°C for 5 days. The resulting colonies were again streaked on the same agar plates to obtain a nalidixic acid-resistant G. oxydans DSM No. 4025, GON. The broad-host-range cosmid pVK102 (Km^, Tc^ was transferred from E. coli carrying pVK102 into the GON strain by the tri-parental conjugal mating as follows. A helper strain, E. coli carrying pRK2013 and a donor strain carrying pVK102 were cultivated in LB medium with 50 jig/ml of kanamycin at 37°C overnight. The cultures were transferred to fresh LB medium with kanamycin and incubated for 5-6 hours. Recipient strain, GON, was cultivated in TN liquid medium at 30°C overnight E. coli and GON strains were separately centrifuged and re-suspended in equal- and one tenth- volumn of fresh T medium, respectively. One hundred |il of each cell suspension was mixed together and 30 fil portion of the mixture was spotted onto a nitrocellulose filter placed on the surface of a NS2 agar plate. Transconjugants were selected on the T agar medium containing 50 |xg/ml of nalidixic acid and 50 |ig/ml of kanamycin (TNK agar medium). Several colonies were obtained on the selection plates where many spontaneous mutants ofE. coli (Nal'", Km^) colonies also appeared. The plasmid and chromosomal DNAs of the

transconjugant candidates were prepared and compared with the authentic pVK102 and chromosomal DNA of G. oxydans DSM No. 4025 by restriction analysis and Southern-blot hybridization. Consequently, one transconjugant of G. oxydans [DSM No. 4025] carrying pVK102, G0N8-1, was identified. The plasmid DNA prepared from GON8-1 was identical to that of pVK102 and replicable in E. coli. The chromosomal DNA of G0N8-1 was identical to tiiat of G. oxydans [DSM No. 4025].
To isolate strains that could work as hosts with higher competence for conjugal mating, the transconjugant G0N8-1 was cured of the plasmid pVK102. G0N8-1 was cultivated in T broth without antibiotics at 30°C for 2 days, 2% of the culture was transferred into fresh T broth. After three such cultivation cycles, the cells were spread on T agar plates, incubated at 27 °C for 4 days, and the resulting colonies were picked onto TNK and TN agar plates to select Km^ strains. One of the Km^ strains was designated as G0S2 and was confirmed by Southem-blot hybridization not to be carrying any DNA region of pVK102. Then, pVK102 was transferred into strain GOS2 by a conjugal mating; this strain showed 10^ ~ 10-^ fold higher competence (namely 10'^ ~ 10'^ transconjugants / recipient) than G. oxydans [DSM No. 4025] did.
(2) Isolation of G0S2R, a rifampicin-resistant mutant of G0S2.
Rifampicin resistant (RiPO mutants from G0S2 were isolated through repeated transfer of G0S2 cells onto T agar medium containing 20 ~ 100 |ig/ml rifampicin; one of the Rif^ strains was designated as G0S2R. Strain GOS2R showed very high competence; lO"^ ~ 10"^ and lO'^transconjugants /recipient on TRK agar (T agar medium containing 100 M.g/ml rifampicin and 50 |ig/ml kanamycin) plate and on TRT agar (T agar medium containing 100 p.g/ml rifampicin and 3 |ig/ml tetracycline) plate, respectively.

2KGA productivity from L-sorbose by GOS2R was compared with that of G. oxydans [DSM No. 4025]. The cells maintained on NS2 agar medium were inoculated into 5 ml of the seed culture medium consisting of 8% L-sorbose (sterilized separtely), 0.05% glycerol, 0.25% MgS04-7H20, 1.75% com steep liquor, 5.0% baker's yeast, 1.5% CaC03, and 0.5% urea (sterilized separately) (pH 7.0 before sterilization) and incubated at 30°C for 24 hours. The resulting seed culture (5 ml) was inoculated into a 5(X)-ml Erlenmeyer flask containing 50 ml of the production medium PMSIO consisting of 10% L-sorbose, (sterilized separtely), 0.05% glycerol, 0.25% MgS04-7H20, 3% com steep liquor, 6.25% baker's yeast, 1.5% CaC03, and 1.6% urea (sterilized separately) (pH 7.5 before sterilization) and incubated at 30°C for 4 days with shaking (180 rpm). The quantitative determination of 2KGA was assayed by high performance liquid chromatography. G0S2R and G. oxydans [DSM No. 4025] produced 87.3 and 97.3 g/1 of 2KGA, respectively.
(3) Isolation of GORS6-35 as a host with high 2KGA productivity
To evaluate the self-cloning of AADH genes in the strain with the same productivity of 2KGA from L-sorbose as G. oxydans [DSM No. 4025], a new host was constructed by (i) adding rifampicin-resistance (2(X) |J.g/ml), (ii) introducing and curing pVK102, and (iii) selecting 2KGA high producer from L-sorbose. Thus obtained GORS6-35 shows the following two characteristics: (i) almost the same 2KGA productivity (about 100 g/12KGA from 10% L-sorbose) as the parent G. oxydans [DSM No. 4025]; and (ii) a competence (10"^ ~ 10"^ transconjugants /recipient

Example 7. Construction of promoter-replaced Enzvme B gene.
The promoter of Enzyme A gene (PA) is likely strong in G. oxydans [DSM No. 4025] because Enzyme A was found to be one of the highest-e;q)ressing proteins in amount in the cell when total cell firee extract of G. oxydans [DSM No. 4025] was subjected to SDS-polyacrylamide gel electrophoresis and the resulting gel was stained with Coomassie Brilliant Blue R-250. The PA and another promoter, a promoter of kanamycin resistant gene of Tn5 (PTn5), which could express the kanamycin resistance in G. oxydans [DSM No, 4025], were attached to the structure gene with the SD sequence of Enzyme B gene as shown in Fig. 10.
Enzyme B gene-containing 2.3 kb HindUI fragment was inserted in M13 nq)18 and the resulting phage DNA was subjected to site-directed mutagenesis carried out with T7-GENTM in Vitro Mutagenesis Kit (TOYOBO Co., Ltd., Osaka Japan) according to the recommendations by the supplier (Fig. 9). To insert various promoters upstream of Enzyme B gene instead of Enzyme B promoter, BamHI site was created upstream of the SD sequence. A primer for the mutagenesis, GTTAGCGCCXJTGGATCCCCATTGGAGG (27-mer including 5amHI site, SEQ-IDNo. 12), were synthesized with Applied Biosystems 381A DNA synthesizer. The resulting BaniHI-Hindni fragment carries Enzyme B SD and structural genes without the Enzyme B promoter (PB).
Then promoter of Enzyme A gene (PA) was subcloned by PCR method using primers tagged with the sequences for the HindlB. and BamHI sites. The PCR reaction was carried out with GeneAmpTM DNA Amplification Reagent Kit (Takara Shuzo, Kyoto, Japan) with a thermal cycler, Zymoreactor n (Atto Corp., Tokyo, Japan). The reaction consists of pre-treatment before adding enzyme (94°C, 5 minutes.); 30 cycles of denaturation step (94°C, 1 minute.), annealing step (60°C, 1 minute.), and synthesis

step (72 'C, 1minute.); and post-treatment (72°C, 5 minutes.). Plasmid pSSA202 (pUC18-Enzyme A gene in 2.7kb HindIII was used as the template DNA. The reaction mixture contained 200μM of dNTPs, 1 μM of each primer, 1 ng of template DNA and 2.5 u of AmpliTaqTM DNA polymerase in the buffer supplied. Consequently, 300 bp fragment upstream from the SD sequence was amplified. The PCR product was inserted into pUC18 between HindIII and BamHI sites and used for nucleotide sequencing; the amplified sequences do not have any mutations caused by misincorporation in PCR.
The promoter of the kanamycin resistant gene, PTn5, was first obtained as a HindUI-Pstl fragment from the plasmid pNeo (Pharmacia Co., Uppsala, Sweden). The HindUl-Pstl fragment was then inserted into the multicloning site of pUC18, and finally the PTn5 was excised as a HindDI-BamHI fragment.
The HindIII -BaniHI fragments containing the PA and PTn5 promoters were inserted in the HindIII site of pUC18 together with BarhHI-HindIII fragment containing the PB promoter-removed Enzyme B structural gene. The HindIII fragments from the resulting plasmids were subcloned into pVKlOO to produce pSSAP-B and pSSPTn5-B, which were transferred into G0S2R by conjugal mating as described in Example 6.
Example 8. 2KGA production bv transconiugants of GQS2R in flask
(1) 2KGA production from L-sorbose by Enzyme-A gene-amplified transconjugant in single culture fermentation in flask.
The Enzyme A plasmid, pSSA102R, and the vector plasmid, pVK102, were introduced into G0S2R by a conjugal mating method as described in Example 6. The resulting transconjugants were maintained on NS2 agar medium containing 30 μg/ml

tetracycline and subjected to 2KGA fermentation from L-sorbose. The cells of the transconjugants were inoculated into 5 ml of the seed culture medium described in Example 6 and incubated at 30°C for 24 hours. The resulting seed culture (5 ml) was inoculated into a 500-ml Erlenmeyer flask containing 50 ml of the PMS10 production medium described in Example 6 or the PMS 12 production medium consisting of 12% L-sorbose, (sterilized separtely), 0.05% glycerol, 0.25% MgS04-7H20, 3% com steep liquor, 10 % baker's yeast, 1.5% CaC03, and 2% urea (sterilized separately) (pH 7.5 before sterilization) and incubated at 30°C for 4 or 5days with shaking (180 rpm). As a result, G0S2R (pSSA102R) and G0S2R (pVK102) produced 92.2 and 89.1 g/1 2KGA , respectively, from 10% L-sorbose in 4 days, and 105.7 and 99.9 g/12KGA, respectively, from 12% L-sorbose in 5 days.
(2) 2KGA production from D-sorbitol by GOS2R (pSSB 103R) in single culture fermentation in flask.
The Enzyme B plasmid, pSSB103R, and the vector plasmid, pVK102, were introduced into GOS2R by a conjugal mating method as described in Example 6. The resulting transconjugants were maintained on NS2 agar medium containing 30 |ig/ml tetracycline and subjected to 2KGA fermentation from D-sorbitol. The cells of the transconjugants were inoculated into 5 ml of the seed culture medium consisting of 8% D-sorbitol, 0.05% glycerol, 0.25% MgS04-7H20, 1.75% com steep liquor, 5.0% baker's yeast, 1.5% CaC03, and 0.5% urea (sterilized separately) (pH 7.0 before sterilization) and incubated at 30°C for 24 hours. The resulting seed culture (5 ml) was inoculated into a 500-ml Erlenmeyer flask containing 50 ml of three production media shown in Table 8 and incubated at 30°C for 3 days with shaking (180 rpm). As a result, GOS2R (pSSB103R) produced about 61.5, 71.5 and 73.0 g/1 of 2KGA from 8%, 10% and 12% D-sorbitol, respectively, while G0S2R (pVK102) produced 19.5, 25.4 and 30.2 g/l 2KGA, respectively.


(3) 2KGA production from D-sorbitol by G0S2R (pSSAP-B) and G0S2R (pSSPTn5-B) in single culture fermentation in flask.
The cells of G0S2R (pSSAP-B), G0S2R (pSSPTn5-B) and G0S2R (pSSBlOSR), G0S2R (pVKlOO) were cultivated in the PMSLIO production medium in Erlenmeyer flasks at 30°C for 3 days as described in Example 8 (2). The amounts of 2KGA produced were shown in Table 9.


Example 9 2KGA production fix)m D-sorbitol in 3-L Jar fermentations by single microorganism
(1) Single culture fermentation by G0S2R (pSSB 103R)
Five ml portions of the seed culture prepared in test tubes as described in Example 8-(2) were transferred to four 500-ml Erlenmeyer flasks containing 50 ml of the same seed culture medium and incubated at 30°C for 24 hours with shaking (180 ipm). The resulting broth (200 ml of the seed culture) was inoculated into 3-L jar fermentor containing 18(X) ml of the PMSLIO production medium containing 3ml of antifoam. The fermentor was operated at 30°C, 700 rpm and 0.5vvm. D-Sorbitol was fed in ways: (i) 200 ml of 50% D-sorbitol was fed in 6 hours from tiie 24* to the 30* hour, or (ii) 280 ml of 50% D-sorbitol was fed in 8.4 hours from tiie 24* to tiie 32.3* hour. As a result, 99.0 and 103.4 g/12KGA were produced by the fed-batch fermentations (i) and (ii), respectively in 51 hours.
Example 10. 2KGA production fiom D-sorhitol bv Enzvme B gene-ampUfied GOS2R in mixed culture fermentation with E. coli in flask
(1) Mixed-culture fermentations with 5. megcaeriwn, E. coli and P. putida.
B. megaterium [DSM No. 4026], E. coli HBlOl and P. putida [ATCC 21812], growth factor suppliers, were cultivated in 150 ml of the seed culture medium consisting of 0.3% yeast extract (Difco), 0.3% beef extract (Kyokuto Seiyaku, Tokyo, Japan), 3% com steep liquor, 1% polypeptone (Kyokuto), 0.1% urea, 0.1% KH2PO4, 0.02% MgS04-7H20,2% L-sorbose, 0.1% CaCOs (pH 7.1 before sterilization) for 24 hours at 37, 37, and 30°C, respectively. Strain G0S2R (pSSB103R) was cultivated in two test tubes containing 5 ml of the seed culture medium as described in Example 8-
(2) at 30°C for 24 hours. Four ml of G0S2R (pSSB103R) seed cultures and 3.5 ml of

growth factor supplier seed culture were inoculated to a 500-ml of Erlenmeyer-flask containing 50 ml of the production medium for mixed culture fomentations consisting of 8% D-sorbitol, 0.01% MgS04-7H20,1% com steep liquor, 0.1% KH2PO4, 0.6% CaCOs, 1.5% urea (sterilized separately) and antifoam (one drop per flask) (pH 7.0 before sterilization) and the flask was shaken at 30°C for 46.5 hours. As a result, mixed culhire with B. megaterium DSM No. 4026, E. coli HBlOl and P. putida ATCC 21812 produced 49.9, 54.1, 31.3 g/12KGA, respectively.
(2) Mixed culture fermentation of G0S2R (pSS AP-B) with E. coli in flask.
Mixed culture fermentations by G0S2R (pSSAP-B) with E. coli was performed in the same manner as described above except for the seed culture medium for E. coli containing 2% D-sorbitol instead of 2% L-sorbose. From 10% of D-sorbitol, G0S2R (pSSAP-B) produced 73.7 g/12KGA in 48.5 hours.
Example 11. 2KGA production by recombinant AADH
A reaction mixture containing 1.7 mg/ml of purified Enzyme A (purified according to Example 4), 50 mM Tris-HCl, pH 7,5,5 mM CaCl2,8 mg/ml bovine serum albumine (BSA), ImM PMS, 20 |xg/ml PQQ, and 4% L-sorbose was incubated at 30°C with genfle shaking for 20 hours. As a result, about 2 g/i 2KGA (TLC assay) was produced.
The other reaction mixture containing 2.4 mg/ml each of purified Enzyme A and Enzyme B (purified according to Example 4), 50 mM Tris-HCl, pH7.5,5 mM CaCl2, 8 mg/ml BSA, ImM PMS, 20 ng/ml PQQ, and 2% D-sorbitol was incubated at 30°C with gentle shaking for 20 hours. As a result, 0.25 g/12KGA (HPLC assay) and about 5 g/1 L-sorbose (TLC assay) were produced.

Example 12. Production of aldehydes from alcohos. ketones from alcohols or carboxvlic aicds and carboxvlic acids from aldehydes.
Enzyme reactions with purified Enzyme A or Enzyem B and various substrates were performed as described in Example 11. The resulting products were identified by TLC and/or HPLC as shown in Table 10.

Enzyme A converted D-fructose to 2KD; this means that D-glucosone was also a product formed from D-fructose as the intermediate.
Example 13. 2KGA and L-sorbose production by a transconjugant of P. putida
A resting cell mixture (2 ml) containing 1% CaCOs, 0.3% NaCl, ImM PMS, 5 |ig/ml PQQ, 2% L-sorbose and 10 OD600 unit-cells of nalidixic acid resistant (NaF) P. putida [ATCC 21812] carrying pSSA102R or pVKlOO was incubated at 30°C with gentle shaking for 17 hours. As a result, Nal^ P. putida [ATCC 21812] carrying pSSA102R or pVKlOO produced 18.9 or 0.0 g/1 of 2KGA, respectively.

A resting cell mixture (2 ml) containing 1% CaCOs, 0.3% NaCl, ImM PMS, 5 Hg/ml PQQ, 2% D-sorbitol and 10 OD600 unit-cells of Nal^ P. putida [ATCC 21812] with pSSB103R or with pVKlOO was incubated at 30°C with gentle shaking for 17 hours. As a result, Nal^ P. putida [ATCC 21812] carrying pSSB103R or with pVKlOO produced 7.8 or 0.0 g/1 of L-sorbose, respectively.
Example 14. Construction and characterization of chimera AADH enzymes
(1) Construction of chimera AADH enzymes
To alternate substrate specificity of AADH enzymes, a variety of chimera enzymes between Enzymes A and B were constructed.
(i) Figure 2 shows the structure of the chimera genes by strategy I (restriction and ligation method). The restriction sites conserved in both genes, Ava I (nucleotide No. 603 of Enzyme A gene), EcoRL site (nucleotide No. 1084), and Sail site (nucleotide No. 1470) (Fig. 7) were used for the construction. First, Enzyme A and B gene cassettes (2.7 kb and 2.3 kb Hind EI fragments, respectively) were subcloned in the same direction in this order on pUC18 to produce the plasmid pSSAB201, and Enzyme B and A gene cassettes were also subcloned in the same direction in this order on pUCl8 to produce pSSBA201 (Fig. 3). After partial digestion of these plasmids with each restriction enzyme, resulting digests were hgated and used to transform E. coli JM109. Ampicillin resistant transformants were analyzed for their plasmids, and Enzyme A gene-headed and Enzyme B-headed chimera gene cassettes with the expected HindUl fragment size of 2.7 kb and 2.3 kb, respectively, were selected. Thus constructed chimera gene cassettes were introduced into HindUl site of pVK102 to produce pSSA/BlOlR, pSSA/B102R, pSSA/B103R, pSSB/AlOlR, pSSB/A102R,

and pSSB/A103R which encode Enzyme A/Bl, EnzymeA/B2, EnzymeA/B3, EnzymeB/Al, EnzymeB/A2, and EnzymeB/A3, respectively, as shown in Fig. 2. These six plasmids were introduced into Nalr P. putida by a conjugal mating method as described in Example 1.
(ii) Figure 8 shows the scheme for constructing chimera genes by strategy 11; in vivo homologous recombination method to construct chimeras recombinated at random positions for altering the substrate specificity of AADH enzymes. The principle of this method is as follows: (i) Locate two homologous genes to be recombinated tandem in one plasmid with selective marker, (ii) Cut it at restriction sites between the two genes, and transform rec A+ E. coli cell with the linearized plasmid; (iii) Select transformants showing selective marker which carry circularized DNAs by recombination between the two genes at various positions. Two plasmids pSSAB201 and pSSBA201 which have Enzyme A and Enzyme B genes on pUC18 (Fig. 3) were linearized with pairs of restriction enzymes as shown in Fig. 8. E. coli JMlOl(rec A"*") was transformed with these linearized DNAs. Transformants were obtained at frequency of 10^-10^ l\ig DNA. To begin with, DNA size was determined to remove illegitimate recombinants. As a result, correct recombinants were obtained at ratio of 30%. Xhol-Ball fragment in which Enzyme A gene lost about two-third of C-terminus was efficient to obtain chimeras recombinated within one-third of N-terminus. Next, the recombinants were classified into recombination site groups bordered by restriction sites of three Smal, Sphl, Sail and Ball (Fig. 7). Thus constructed chimera genes were subcloned on pVKlOO as HindUl cassette and the plasmids were introduced into Nal^ P. putida by a conjugal mating method.
(2) Characterization of chimera AADH enzymes
(i) Characteristics of the chimeras obtained by restriction and ligation method

The chimera enzymes expressed in Nal'" P. putida were characterized enzymatically by using soluble fractions of the cells of the transconjugants as described in Example 1. Eight substrates were used for the evaluation as shown in Fig. 11. Enzymes A/Bl and A/B3 showed Enzyme A-type substrate specificity, although their expression level was lower than that of Enzyme A. On the other hand, Enzymes B/Al, B/A2, and B/A3 showed Enzyme B-type substrate specificity, although activity on n-propanol (Enzyme A type activity) became higher in accordans with the increase of the region fiom Enzyme A; the expression level of Enzyme B/Al gene was about 2-fold higher than that of wild Enzyme B gene. As a result from the chimeras obtained by recombination and ligation method, it was concluded that N-terminal one third region of Enzyme A or Enzyme B determines its substrate specificity basically.
(ii) Characteristics of the chimeras obtained by homologous recombination method.
Among the chimeras obtained as above, seven out of eighteen chimera enzymes obtained from the chimera genes recombinated between Small and Sail sites illustrated in Fig. 7 showed preferable substrate specificity. The seven chimera enzymes converted D-sorbitol to L-sorbose, not to D-glucose produced by Enzyme A, and converted L-sorbose to 2KGA like Enzyme A. The recombination sites were determined by nucleotide sequencing as described in Example 2. This type of chimeras having an approximate structure of" N-terminal 2/9 of Enzyme A + C-terminal 7/9 of Enzyme B " was classified as Enzyme superA-type. There were three Enzyme superA-type enzymes according to the recombinated site: Enzyme A/B21 (chimera consisting of Enzyme A part of amino acid residue No. 1 -128 and Enzyme B part of No. 129 -556), Enzyme A/B22 (chimera consisting of Enzyme A part of amino acid residue No. 1 -125 and Enzyme B part of No. 126- 556) and Enzyme A/B25 (chimera consisting of Enzyme A part of amino acid residue No. 1-135 and Enzyme B part of No. 136 -

556). P.putida transconjugant expressing genes of Enzyme A/B21, Enzyme A/B2Tor Enzyme A/B25 converted D-sorbitol to L-sorbose and did not convert D-sorbitol to D-
j
glucose. The other type of chimera Enzyme A/B31 (Enzyme A part of amino acid
residue No. 1-95 and Enzyme B part of No. 96 - 556) converted D-sorbitol to L-
sorbose efficiently and did not convert L-sorbose to 2KGA; this chimera showed
i Enzyme !B-type activity. Expression level of above mentioned chimeras was higher
than that of wild Enzyme B because it was found that Enzyme B gene contains many
I
rare codons but Enzyme A does not when the genes were analyzed with the program, Codon Preference (Wisconsin Sequence Analysis Package^^, Genetics Computer Group).
(3) Improvement of codon usage in chimera genes
To further improve the chimeras. Enzyme A/B21, Enzyme A/B22, Enzyme A/B25 and Enzyme A/B31 in the view point of the preferable codon usage, the C-terminal two thirds consisting of Enzyme B residues were replaced with the C-terminal two thirds consisting of Enzyme A residues. Enzyme A/B21, Enzyme A/B22, Enzyme A/B25 and Enzyme A/B31 genes were used for constructing new chimera genes of Enzyme sA21 (Enzyme A part of amino acid residue No. 1 -128, Enzyme B part of No. 129 - 180 and Enzyme A part of No. 180 - 556), Enzyme sA22 (Enzyme A part of amino acid residue No. 1 - 125, Enzyme B part of No. 126 - 180 and Enzyme A part of No. 180 - 556), Enzyme sA2 (Enzyme A part of amino acid residue No. 1 - 135, Enzyme B part of No. 136 -180 and Enzyme A part of No. 180 - 556) and Enzyme sB (Enzyme A part of amino acid residue No. 1 - 95, Enzyme B part of No. 96 - 180 and Enzyme A part of No. 180 - 556) (Fig. 4). Actually, the replacement experiments for Enzyme sA2 and Enzyme sB were performed by partially digesting the plasmids, pUClS carrying Enzyme sA gene and Enzyme B/Al gene and pUC 18 carrying Enzyme A/B31 gene and Enzyme B/Al gene with Aval, ligating the resulting digests.

transforming E. coli JM109, analyzing the plasmid structure of the transformants by restriction analysis, and detennining the nucleotide sequence to confirm the expected recombination site, Aval. The replacement experiments for Enzyme sA21 and Enzyme sA22 were performed by replacing the HindlE-SspI fragment of pSSsA2 encoding N-terminal part of Enzyme sA2 with the corresponding HindBI-Sspl fragment containing recombinated site of Enzyme A/B21 or Enzyme A/B22 gene (Fig. 4).
(4) Kinetic properties of chimera enzymes
Tables 11 and 12 summarizes the kinetic properties of chimera enzymes. Enzyme sA2 and Enzyme sB in comparison with Enzyme A and Enzyme B, respectively.

Products from L-sorbose in product assay with Enzyme sA2 and Enzyme A were 2KGA. Products from D-sorbitol in product assay with Enzyme sA2 and Enzyme A were L-sorbose with trace amount of D-glucose and D-glucose only, respectively. Thus, Enzyme sA2 showed desired characteristics for 2KGA production from D-sorbitol; L-sorbose/L-sorbosone dehydrogenase activity to produce 2KGA from L-sorbose like Enzyme A and D-sorbitol dehydrogenase activity to produce L-sorbose from D-sorbitol like Enzyme B.


In comparison with Enzyme B, Enzyme sB showed higher affinity to D-sorbitol and lower affinity to L-sorbose which is the oxidation product of D-sorbitol and inhibitor in the conversion of D-sorbitol to L-sorbose.
Example 15. 2KGA production from D-sorbitol by G0S2R derivative strain amplified with chimera AADH enzymes
For evaluating Enzyme sA2 and Enzyme sB, GOBAK and G0I13 strains were constructed. GOBAK was made from GOS2R by deleting the whole Enzyme B gene and instead inserting 1.28 kb Km^ gene cassette isolated from pUC4K [4.1 kb, Km^, Amp'^; Pharmacia, Uppsala, Sweden; Viera, J., and Messing, J., Gene 12:259, (1982)] by using a suicide vector pSUP201 [Amp^, Cm^, mob"*", a derivative of pBR325, BioA'echnology, 1:784-791, (1983)].
GOI13 was constructed from GOBAK by replacing wild Enzyme A gene with Enzyme sB gene and deleting wild Enzyme A" gene replaced with gentamicin (Gm) resistant gene cassette with the suicide vector pSUP202 [Amp^, Cm^, Tc^, mob+, a derivative of pBR325, Bio/Technology, 1:784-791, {1983)]. The Gm^ gene cassette was designed to have PstI site at both ends by PCR amplification with the DNA fragment Tn5-GM [Sasagawa et al.. Gene 56: 283-288, (1987)] as the template, and the resulting PCR product was inserted into Pstl site of pUC4K to produce pUC8G; Gm^ gene can be isolated firom pUC8G by digesting with EcoRI, BaniHl, Sail, or Pstl.
(1) Effect of Enzyme sA2 amplification in 2KGA production

Plasmid pSSsA2, pVKlOO with 2.7 kb HindUI cassette containing Enzyme sA2 gene, and its control plasmid pSSA102R, pVK102 with 2.7 kb HindlU cassette containing Enzyme A gene, were introduced into GOI13 by a conjugal mating method as described in Example 6. The resulting transconjugants were cultivated in PMSLIO medium at 30°C for 4 days as described in Example 8. G0I13 carrying pSSsA2 and pSSA102R produced 66.3 and 38.5 g/1 of 2KGA, respectively, and 8.4 and 25.9 g/1 of 2KD (by-product of 2KGA produced from D-sorbitol via D-glucose and D-gluconate), respectively.
, (2) Plasmids pSSsA21 and pSSsA22, which are pVKlOO with 2.7 kb HindUI
cassettes containing Enzyme sA21 and Enzyme sA22 genes, respectively(Fig. 4), were i introduced into GOI13 by a conjugal mating method as described in Example 6. The : resulting transconjugants were cultivated in PMSLIO medium at at 30°C for 4 days as described in Example 8. G0I13 carrying pSSsA21 and pSSsA22 produced 66?!and T14g/1 of 2KGA, respectively, and 0.3 and 0.4 g/1 of 2KD, respectively.
(3) Effect of Enzyme sB in 2KGA production
i
I I
Plasmid pSSsB, pVKlOO with 2.7 kb Hindin cassette containing Enzyme sB gene (Fig. 4) and its control plasmid pSSB103R, pVK102 containing 2.3 kb Enzyme B gene, were ino-oduced into GOBAK by a conjugal mating method. GOBAK Carrying pSSsB, GOBDK carrying pSSB103R, and GOBAK were cultivated in PMSL8 medium as described in Example 8 (2) and produced 52.0,46.8, and 1.1 g/I of 2KGA, respectively, and 6.9, 9.3, 32.3 g/1 of 2KD, respectively.
GOn3, which carrys one copy of Enzyme sB on the chromosomal DNA without wild genes of Enzyme B, Enzyme A, and Enzyme A", was also cultivated in PMSLIO medium in 2 days. It produced 79.3 g/1 of L-sorbose.























Glu Ser He Asp Glu Asn Gly He Val Thr Val Asn Glu Asp Ala
355 360 365
He Leu Lys Glu Leu Asp Val Glu Tyr Asp Val Cys Pro Thr Phe
370 375 380
Leu Gly Gly Arg Asp Trp Pro Ser Ala Ala Leu Asn Pro Asp Ser
385 390 395
Gly He Tyr Phe He Pro Leu Asn Asn Val Cys Tyr Asp Met Met
400 405 410
Ala Val Asp Gin Glu Phe Thr Ser Met Asp Val Tyr Asn Thr Ser
415 420 425
Asn Val Thr Lys Leu Pro Pro Gly Lys Asp Met He Gly Arg He
430 435 440
Asp Ala He Asp He Ser Thr Gly Arg Thr Leu Trp Ser Val Glu
445 450 455
Arg Ala Ala Ala Asn Tyr Ser Pro Val Leu Ser Thr Gly Gly Gly
460 465 470
Val Leu Phe Asn Gly Gly Thr Asp Arg Tyr Phe Arg Ala Leu Ser
475 480 485
Gin Glu Thr Gly Glu Thr Leu Trp Gin Thr Arg Leu Ala Thr Val
490 495 500
Ala Ser Gly Gin Ala He Ser Tyr Glu Val Asp Gly Met Gin Tyr
505 510 515
Val Ala He Ala Gly Gly Gly Val Ser Tyr Gly Ser Gly Leu Asn
520 525 530
Ser Ala Leu Ala Gly Glu Arg Val Asp Ser Thr Ala He Gly Asn
535 540 545
Ala Val Tyr Val Phe Ala Leu Pro Gin
550 555

























WE CLAIM:
1. A process for producing 2-keto-L-gulonic acid from L-sorbose and/or D-sorbitol the
said process comprises
(a) converting L-sorbose and/or D-sorbitol into 2-keto-L-gulonic acid with the aid of a
biochemical action of a recombinant organism, said organism carrying a recombinant
expression vector, said expression vector comprising one or more DNA molecules en¬
coding an enzyme having an alcohol and/or aldehyde dehydrogenate activity said enzyme
comprising a recombinant polypeptide selected from the group consisting of S£Q ID MO
5, SEQ ID NO 6, SEQ ID NO 7 and SEQ ID NO 8, chimerical recombinant enzymes
between the polypeptides identified by said sequences, and polypeptides with at least 80 %
identity to said sequences,
wherein the conversion is carried out under a pH value in the range of from 6.0 to 9.0 at a temperature in the range of from 10 C to 50, m the presence of an electron acceptor in a suitable buffer and with a substrate concentration in the range of from 1 to 200 g/1; and
(b) isolating the resulting 2-keto-L-gulonic acid.
2. The process according to claim 1, wherein the enzyme comprising a recombinant poly¬
peptide is a chimerical enzyme, such as Enzyme A/Bl, Enzyme A/B2, Enzyme A/63, Enzyme
B/Al, En2)'me B/A2, Enzyme B/A3, Enzyme sA2, Enzyme sA21, Enzyme sA22, Enzyme .sB
and a functional derivative thereof.
3,The process according to claim 1 wherein the DNA molecule encoding the enzyme having an alcohol and/or aldehyde dehydrogenate activity is present in the form selected from the group consisting of a linear DNA, a circular DNA, and an insertion DNA frag¬ment on a chromosome.
4. The process according to claim I wherein the recombinant expression vector comprises one or more DNA molecule(s) as defined by claim 3.
5. The process according to claim 4 wherein said DNA molecule(s) is(are) functionally linked to one or more genetic control sequence(s) and are capable of expression of the en¬zymatic polypeptide(s) as defined in claim 1 or 2 in an appropriate host cell.
6. The process according to claim 5 wherein the recombinant expression vector comprising said DNA molecule(s) is selected from the group consisting of pSSAl02R, pSSA VlOlR, p$SA"102, pSSBl03R, pSSAP-B, pSSA/BlOlR, pSSA/Bl02R, pSS.'\/Bl.03R, pSSB/Al0lR, pSSB/A102R, pSSB/A103R, pSSsA2, pSSsA21, pSSsA22 and pSSsB.

7. The process according to claim 1, wherein the recombinant organism is selected from
the group consisting of microorganisms, mammalian cells and plant cells.
8. The process according to claim 7, wherein the microorganism is selected from the group
consisting oi Escherichia coli, Pseudomonas putida, Acetobacter xyliuum, Acetobacter
pastetirianus, Acetobacter aceti, Acetobacter Hansen, and Gluconobacter oxydans.
9. The process according to claim 8, wherein the microorganism is Gluconobacter oxydans
[DSMNo.4025].
10. The process for producing 2-keto-L-gulonic acid according to claim 1, wherein incu¬
bating under known conditions a reaction mixture containing (i) an enzyme having an
alcohol and/or aldehyde dehydrogenate activity, said enzymes comprising a recombinant
polypeptide selected from the group consisting of SEQ ID NO 5, SEQ ID NO 6, SEQ ID
NO 7 and SEQ ID NO 8, chimerical recombinant enzymes between the polypeptides
identified by said sequences, and polypeptides with at least 80 % identity to said sequences;
and (ii) L-sorbose and/or D-sorbitol.
11. A process for producing 2-keto-L-gulonic acid from L-sorbose and/or D-sorbitol,
substantially as herein described with reference to the accompanying drawings.


Documents:

1997-mas-97 abstract.pdf

1997-mas-97 claims.pdf

1997-mas-97 correspondence others.pdf

1997-mas-97 correspondence po.pdf

1997-mas-97 description (complete).pdf

1997-mas-97 drawings.pdf

1997-mas-97 form-1.pdf

1997-mas-97 form-26.pdf

1997-mas-97 form-4.pdf

1997-mas-97 form-6.pdf

1997-mas-97 petition.pdf


Patent Number 190888
Indian Patent Application Number 1997/MAS/1997
PG Journal Number 30/2009
Publication Date 24-Jul-2009
Grant Date 11-Mar-2004
Date of Filing 09-Sep-1997
Name of Patentee M/S. F. HOFFMANN-LA ROCHE AG
Applicant Address 124 GRENZACHERSTRASSE CH-4070 BASLE
Inventors:
# Inventor's Name Inventor's Address
1 M/S. F. HOFFMANN-LA ROCHE AG 124 GRENZACHERSTRASSE CH-4070 BASLE
PCT International Classification Number C07C 59/105
PCT International Application Number N/A
PCT International Filing date
PCT Conventions:
# PCT Application Number Date of Convention Priority Country
1 NA