CA1341085C - Recombinant 2,5-diketogluconic acid reductase production - Google Patents

Abstract

The invention relates to purification of and recombinant production of 2.5-diketogluconic acid (2.5-DKG) reductase and the use of the reductase so produced in converting 2.5-DKG stereoselectively into 2-keto-L-gluconic acid (2 KLG), as well as to the production of a single recombinant organism capable of synthesizing 2-KLG. The 2-KLG
produced is a useful intermediate in the production of ascorbic acid (vitamin C).

Description

10 RECOMBINANT 2,5-DIKETOGLUCONIC ACID
REDUCTASE PRODUCTION
Background ~ 34' 08 5 The invention herein concerns aspects of a process for the production of ascorbic acid. It specifically relates to purification of a useful protein, to production of proteins using recombinant techniques and to the use of such proteins in chemical conversions. More particularly, the invention relates to purification of and recombinant production of 2,5-diketogluconic acid (2,5-DKG) reductase and the use of the reductase so pr~~duced in converting 2,5-DKG stereoselectively into 2-keto-L-gluconic acid (2 KLG), as well as to the production of a single recombinant organism capable of synthesizing 2-:KLG. The 2-KLG produced is a useful intermediate in the production of ascorbic acid (Vitamin C).
Ascorbic acid has become a major chemical product in the United States, and elsewhere in the world, due to its importance in health a maintenance. While there may be some controversy over its efficacy in ameliorating the tendency of individuals to contract certain minor illnesses, such as, for example, the common cold, there is no doubt that it is essential for human beings to ingest required amounts of vitamin C. It has become ~; matter of concern in recent years that "natural" foods may not provide adequate amounts of vitamin C. Accordingly, there has developed a large demand for ascorbic acid, both as an additive to foods which are marketed to the consumer with supplemented levels of this vitamin, and as a direct vitamin supplement. Furthermore, ascorbic acid is an effective' antioxidant and thus finds applications as a preservative both in nutritional and in other products.
There are a number of processes available, some commercially viable, for the production of vitamin C. Several of these result in the preliminary production of 2-keto-L.-gulonic acid (2-KLG) which can then be rather simply converted to ascorbic acid through acid or base catalyzed cyclization. Accordincily, 2-KLG has become, in itself, a material of considerable economic and industrial importance.
Means are presently available in t:he art to convert relatively plentiful ordinary metabolites, such a.s, for example, D-glucose, into 2,5-diketogluconic acid (2,5-DKG) by processes involving the metabolism of prokaryotic microorganisms. See, for example, U.S.
Patent 3,790,444 (February 5, 1974); ?.,998,697 (December 21, 1976);
and EPO Application Publication No. 0046284 published February 24, 1982. The availability of this 2,5-DKG intermediate offers a starting material 'which is converted t:o the desired 2-KLG only by the single step of a two electron reduction. The reduction can be effected chemically or catalyzed enzymatically. Various bacterial strains are known which are capable of effecting this reduction.
Such strains are found in the genera Brevibacterium, Arthrobacter, Micrococcus, Staphylococcus, Pseudomonas, Bacillus, Citrobacter and Corynebacterium. See, for example, U.S. 3,922,194 (November 25, 1975), U.S. 4,245,049 (January 13, 19~~1) and U.S. 3,959,076 (May 25, 1976). Suc~~ strains have indeed been used to effect this reduction;
however, use of such strains per se arid without enzyme purification does not permit certain alternative approaches available with the use of purified enzyme. Such a system would permit, for example, continuous production through immobilization of the enzyme on a solid support. Further, access to the genetic machinery to produce such an enzyme is of convenience making improvements in carrying out this process since this machinery may be manipulated and localized to achieve production of the enzyme at a site most convenient for the conversion of 2,5-DKG. Most important among such loci is a site within the same organism which is capable of effecting the production of 2,5-~DKG. Thus, a single organism would be able to use its own machinery to manufacture the ?_,5-DKG, and then convert this endogenous 2,5-DKG in sii~u into the desired product, using the 2,5-DKG reductase gene and appropriate control sequences to produce the catalyst.
It is helpful to undE~rstand the context into which the present 2p invention finds uJ:ility, by representing the process in terms of the relevant chemical conversions. An outline of a typical overall process for manuf<icture of ascorbic acid is shown in Reaction Sc heme 1.

It -C-OH -C-OH C=0 C=0 ~=0 HO-~

1 t I t 1 1~

-C-OH ' -C-OH ~ -C-OH
' -C-OH ' -C-OH
' C

s I 1 1 t -C-OH -C-OH -C-OH C=0 HO-C- HO-C

D-glucose D-gluconic2-Keto-D- 2,5-diketo-D- 2-keto-L-ascorbic acid gluconic gluconic gulonic acid acid acid acid 4 ~1 341 08 5 Reaction Scheme 1 The process conveniently begins i~ith a metabolite ordinarily used by a microorganism such as, for example, D-glucose which is the illustration chosen for Reaction Scheme 1. Through enzymatic conversions, the D-glucose undergoes a. series of oxidative steps to give 2,5-diketo-D-gluconic acid. It has been shown that this series of steps can be carried out in a single organism. (U. S. Patent 3,790,444, EPO Appl. A20046284 (supra); such organisms are, for example, of the genus Gluconobacter, A,cetobacter or Erwinia).
Alternate preparations of ascorbic acid have circumvented the 2,5-DKG intermediate by a combination of fermentative and chemical oxidations, and are clearly more cumbersome than the process shown.
Typical of these is the Reichstein synthesis which utilizes diacetone-2-keto-L-gulonic acid as a precursor to ~-KLG. This intermediate is generated through a se~ies of reductive and oxidative steps involving fermentation, hydrogenation, and, e.g., permanganate oxidation, and the required sequence is clearly more complex than that involved in the reactions shown. The conversion of 2,5-DKG into 2-KLG can also be carried out enzymatically (U. S.
Patent 3,922,194; 3,959,076 (supra); a.nd 4,245,049 (Jan. 13, 1981)).
h1eans are presently well known in the art to convert the resulting 2-KLG into ascorbic acid. This may be done either in the presence of dilute acid and heat according to the method of Yamazaki, or in a two-step process utilizing preliminary esterification in methanol, followed t~y lactonization in base.
Effective procedures are described in Crawford, T.C., et al., Advances in Carbohydrate Chemistry and Biochemistry, 37, 79-155 (1980). These alternatives are straightforward and take advantage of the greater stability and shelf life of 2-KLG over ascorbic acid. Thus, it is more desirable and convenient to stockpile the 2-KLG intermediate for subsequent conversion to the desired final product than to synthesize the ascorbic acid directly.

p'_ -5- 1 341 08 5 Because of the improvements of the present invention, alternate, superior means are available to effect: certain aspects of this overall conversion. In one approach, because the enzyme responsible for the conversion of 2,~> DKG into 2-t;LG has been isolated and purified, the reduction step can be carried out under more controlled conditions, including those whereby the enzyme is immobilized and the solution substrates are fed continuously over the immobilized catalyst. In addition, the availability of recombinant techniques makes possible the production of large amounts of such enzyme available for ready purification. Further, recombinant techniques permit the coding sequences and necessary expression control mechanisms to be transformed into suitable host organisms with improved characteristics. Thus, simply focusing on the conversion of 2,5-DKG to 2-KLG, three levels of improvement are attainable: 1) si:ricter control over variables; 2) availability of continuous processing; and 3) selecti~~n .of host organism for the enzyme which has desirable qualities pertinent to the reduction reaction.
2p The scope of -improvement permitted by the effective cloning and expression of the 2,5-DKI; reductase is, however, even broader.
Because of the availability of the appropriate genetic machinery, it is possible, as well as desirable, to transform an organism which is capable of producing the 2,5-DKG with the gene encoding the reductase. Thus, the same organism can effect the entire process of converting, for example, glucose or other suitable metabolite into the stable, storable intermediate 2-KLG.
Summary of the Invention The present invention effects dramatic improvements in the process for converting a commonly available metabolite such as glucose into 2-KLG, a stable storable precursor for ascorbic acid.
The pathway of the process described by the present invention encompasses the step of converting ?,5-Df;G into 2-KLG. The current processes for formation of the 2-KLG intermediate involve, at best, the deployment of at least two organisms or killed cultures thereof, do not permit regulation of the enzyme levels available, and are limited to batchwise processes.
A major aspect of the present invention is a process for preparing 2,5-DKG reductase in substantially pure form by a series of chromatographic steps resulting in a homogeneous (by HPLC) product. Further facets of this aspect of the invention include the purified enzyme itself and the use of this purified enzyme in the conversion of 2,5-DKG to 2-KLG. Such conversion may, preferably, be carried out using the enzyme in immobilized form.
Another major aspect of the invention is a process for the construction of a recombinant expression vector for the production of 2,5-DKG reductase. Other facets of this aspect include the expression vector so produced, cells and cell cultures transformed with it, and the product of such cells and cell cultures capable of effecting the reduction of 2,5-DKG stereospecifically to 2-KLG.
Still another facet of this aspect of the invention is a process for converting 2,5-Dt:G to 2-KLG using recombinant reductase.
Finally, the invention also relates to a process for converting glucose or other ordinary microbial metabolite into 2-KLG by fermentation by a single recombinant organism, and thereafter to ascorbic acid. It also relates to the recombinant organism capable of carrying out this process. Such an organism is conveniently constructed by tra~nsformi~ng a host cell capable of effecting the 30' conversion of the initial metabolite to 2,5-DKG with an expression vector encoding and capable of expressing the sequence for the 2,5-DKG reductase. Alternatively, such a recombinant organism is constructed by transform ing an organism already producing the 2,5-DKG reductase with vectors encoding the enzymes responsible for the oxidation of metabolite to 2,5-DKG. In either event, use of _,_ proper inducible F~ror~oters and control' systems within the construction of the exare~ssion vectors permit the regulation of enzymatic levels t:o optimize the rate at which the desired conversion steps take place.
Brief Description of the Drav;inas Figure 1 shows an exF~ression vector for the 2,5-DKG reductase gene.
Figures 2 and 3 show the construction of an alternative expression vector for the 2,5-DKG reductase gene.
Figure 4 shows a sequence including the 2,5-GKG reductase gene and control regions of the pTrpl-35 expression vector.
Figure 5 shovrs a stained gel of a protein extract from Erwinia herbicola (ATCC 21998) transformed wii:h the 2,5-DKG reductase expression vector having the sequence of Fig. 4.
Detailed Description A~ Definitions As used herein, "2,5-~DKG reductasE~" refers to a protein which is capable of catalyzing the conversion of 2,5-DKG stereoselectively to 2-KLG. In the specific example herein, the particular form of this enzyme present in Corynebacterium was purified, cloned, and expressed. However, other bacterial species, such as, for example, those from the genera Brevibacterium, Arthrobacter, hlicrococcus, Staphylococcus, Ps.eudomonas, Citrobaci:er, and Bacillus are also known to synthesize an enzyme with the same activity as this enzyme. These genera are illustrative of potential sources for an -1 341 ~8 5 enzyme similar to that present in Corynebacterium which may be available to catalyze this conversion. Alternate sources in addition to these naturally occurring ones in the prokaryotic kingdom may well be found. In addition, as the invention herein discloses and makes available the genetic sequence encoding such enzymes, modifications of the sequence which do not interfere with, and may, in fact, improve the performance of this enzyme are also available to those knowledgeable in the art. Such modifications and altered sequences are included in the definition of 2,5-DKG
reductase as used in this. specification. In short, the term 2,5-DKG
reductase has a functional definition and refers to any enzyme which catalyzes the conversion of 2,5-DKG to 2-~:LG.
It is well understood in the art that many of the compounds discussed in the instant specification, such as proteins and the acidic derivatives; of sac:charides, may exist in variety of ionization states depending upon their surrounding media, if in solution, or on the solui:ions from which they are prepared if in solid form. The use of a term such a>, for example, gluconic acid, to designate such molecules is intended to include all ionization states of the organic molecula referrE~d to. Thus, for example, both "D-gluconic acid" and "D-gluconate" refer to the same organic moiety, and are not intended to specify particular ionization states. It is well known that D-gluconic acid can exist in unionized form, or may be availabla as, for example, the sodium, potassium, or other salt,. The ionized or unionized form in which the compound is pe rtineni: to the disclosure will either be apparent from the context i:o one :>killed in the art or will be irrelavent.
Thus, the 2,5-DKG reductase protein itself may exist in a variety of ionization states depending on pH. A'11 of these ionization states are encompassed by the tE~rm "2,5-DKG reductase".
Similarly, "cE~lls" and "cell cultures" are used interchangeably unless the context: makes it clear that one or the other is referred to. Transformation of cells or.of a cell culture amounts to the same activity; it is clear, of course, that it is the organisms themselves which take up the transforming material although it is a culture of them that is treated v;ith t:he cloning vehicla or other transforming agent. The cells and microorganisms of this invention are defined to include any bacterial, or prokaryotic, organism.
"Expression vector" includes vectors which are capable of expressing DtJA sequences contained therein where such sequences are operably linked to other sequences caFable of effecting their expression. It is implied, although not explicitly stated, that expression vectors must be replicable in the host organisms either as episomes or as an integral part of a chromosomal DtJA; clearly a lack of replicability would render them effectively inoperable. In sum, "expression vector" is also given a functional definition.
generally, expression vectors of utility in recombinant techniques are often in the form of "plasmids," which term refers to circular double stranded DNA molecules which, in their vector form, are not bound to the chromosome. Other effective vectors commonly used are phage and non-circularized DNA. In the present specification, ~~Plasmid" and "vector" are often used interchangeably; however, the invention is intended to include such other forms of expression vectors which serve equivalent functions and which are, or subsequently become, known.
"Recombinant cells" refers to cells which have been transformed with vectors constructed using recombinant DNA techniques. "Host"
cells refers to such cells before they have been transformed. In general, recombinant cells produce protein products encoded by such recombinant vectors and which they would not ordinarily produce 30~ without them; however, the, definition also includes cells which have been transformed with vectors encoding proteins which are, coincidentally, encoded in the bacterial chromosome or otherwise endogenously expressed in the recipient cell. The definition includes any cell which is producing the product of a xenogeneic sequence by vi rtue of recombinant tech;ni ques.

-l0 1 3 4 1 8 8 5 "Ordinary metabolite" refers to such carbon sources as are commonly utilized by bacteria for growth. Examples of such metabolites are glucose, galactose, lactose, fructose or other carbohydrates which are readily available foodstuffs for such organisms. Such metabolites are defined herein to include enzymatic derivatives of such foodstuffs which are convertible into 2,5-diketo-D-gluconic acid. Such derivatives include D-gluconic acid, D-mannonic acid, L-gulonic acid., L-idonic acid, 2-keto-D-gluconic acid, 5-keto-D-gluconic acid, and 5-keto-D-mannonic acid.
B. General Description of Preferred Embodiments B.1 Preparation of Substantially Pure 2,5-DKG Reductase A preferred genus from which an organism is selected for preparation of pure 2,5-D~KG reductase is Corynebacterium. However, bacterial taxonomy is sufficiently uncertain that it is sometimes difficult to ascertain the correct designation between related 9e~era. However, many of those speciE~s which have been found to contain the reductase are members of i:he coryneform group, including the genera Corynebacterium, Brevibacterium, and Arthrobacter; hence it appears by virtue of present knowledge that the preferred source for the enzyme is a member of the coryneform group.
In a preferred mode of preparation, the cell culture is grown under suitable conditions dependent on the strain of bacterium chosen, to an OD550 of about 20 or greater. The culture is then centrifuged and the resulting cell paste (pellet) is treated to lyse 30~ the cells. This paste is.preferably washed, preliminarily, in buffer solution to remove contaminating medium, and the washed pellet treated with, for.example, lysozyme, or by sonication, or by mechanical means t.o break open the cells. The resulting extracts are then subjected to purification by ion exchange chromatography, Preferably using a cellulose based support for an anion exchanger such as, for example, DEAE cellulose. Other anion exchange resins, such as, for example, QAE or DEA,E sephadex* of course may also be used. Elution is accomplished by means known in the art, typically, and preferably by increasing the ionic strength of the eluting solution by increasing the concentration of dissolved salts, preferably sodiurn chloride. The fraction of eluate containing 2,5-DKG
reductase activity is them further purified by adsorption onto an affinity chromatographic support -- i.e., a support system to which it is covalently bound, a dye, or other organic li;gand which is similar to the enzyme substrate or its cofactor.
If the solution to be treated with the affinity support contains substantial amounts of solutes, a preliminary dialysis is desirable. A pa~~icularly effective affinity chromatography support is Amic~on MatrexR gel blue A which exhibits an affinity for enzymes utilizing 1\IADH or NADPH. Elution from such columns may be accomplished by increasing the concentration in the eluting solution of the material for which the enzyme exhibits an affinity, in this case NADP. Fractions containing the desired DKG reductase activity are then pooled for recovery of the pure protein.
The specific activity of the pure protein is greater than 5 units/mg.
Final verification of purification is achieved by size separation using, for example, sephadex gels, polyacrylamide gels, or TSK sizing gels using HPLC.
For the enzyme contained in Co ..rynebacterium sp ATTC 31090, separation by the TSKlHPLC method results in a peak corresponding to molecular weight 45,000 containing the entire complement of activity. However, when the enzyme is subjected to SDS-PAGE, either under reducing or non-reducing conditions, the protein migrates corresponding to a MW of 34,000. Additional characteristics of the protein of this preferred embodiment are given in Example 2.
In summary, the; purification of the enzyme involves the steps of cell lysis, 3 0 anion exchange chromatography, affinity chromatography and verification by size separation. With the * Trade mark exception of cell lysis, the steps may be performed in any convenient order, and thEr transition between steps monitored by assaying the activity according to the procedure in Example 2.

6.2 Conversion of 2,5-Df;G into 2-E;LG Usina Purified Enzyme The conversion may be carried out with the purified enzyme either in solution, or, preferably, in im;nobilized forn. As the desired reaction is a reduction, a source of reducing equivalents is required; the enzyme is specific for P~ADPH, and thus at least a catalytic amount of the coenzyme must be present and the reduced form constantly regenerated during the process. Sources of electrons for the reduction of the coE~nzyme may be provided by any reduced substrate in contact with an enzyme for its oxidation, such as, glucose/glucos:e dehydrogenase; fo rmate/formate dehydrogenase; or glutamate/glutamat:e dehydrogenase. The considerations in choosing a suitable source of reducing equivalen~s include the cost of the substrate and the specificity of the oxidation catalyzing enzyme which must be compatible with the NADP requirement of the purified 2,5-DKG reductase. Other systems for regenerating NADPH cofactors are known in the a.rt using, for example H2 as the source of reducing equivalents and lipoamide dehydrogenase and hydrogenase or ferredoxin reductase and hydrogenase as catalysts, as described by along, C.H. et al. J. Am. Chem. Soc., :103: 6227 (1981). Additional systems applicable to large scale processes are described by Light, D., et al. 1983 in "Organic Chemicals from 8iomass", D.L. Wise, ed., pp. 305-:358.
In a typical conversion, the starting solution will contain 30~ 2,5-DKG in the concentration range of about 1-200 g/L preferably around 10-50 g/L arith thE~ pH of the medium controlled at about 5-7.5, preferably around 6.4. The pH may be maintained using suitable buffers such as, for example, phosphate buffer. The temperature range is about 15°C to 37"C, preferably around 25°C.
The concentration of reducing cofactor NADPH typically around 0.001 mM to 2 mM, preferably around 0.01-0.03 mtd with sufficient source of reducing equivalents to maintain such concentrations during the reaction.
If the enzyme is supplied in solution, its concentration is of the order of 10 mg/L of substrate medium, although, of course, the preferred concentration used vrill be depehdent upon the desired rate of conversion and the specific enzyme chosen. If immobilized enzymes are used, the above-described substrate solution is passed over a solid support containing adsorbed or covalently bound 2,5-DKG
reductase. Ideally, the solid support nnill also contain a suitable catalyst as described above for conversion of the source of reducing equivalents in amounts sufficient to rnaintain the concentration of ~ADPH in the solution. for example, 'the solution in a typical conversion Will contain am approximatcaly equimolar amount of glucose, formate, glutamate or dissolved hydrogen to the 2,5-DKG
concentration and the solid support will contain sufficient reducing catalyst to recycle continuously the tdADP formed by the desired conversion to 2-KLG.
B.3 Cloning and Expression of 2, ti-DKG Reductase Both the availability of large amounts of purified 2,5-OKG
reductase and its ability to be generated in situ in an organism which makes 2,5-DK.G is greatly aided by the process of the invention vrhich provides a means for cloning and expression of the gene for the reductase enzyme. The general procedure by which this is accomplished is summarized as follows" and a specific example of such procedures is outlined herein below in Example 3.
The gene encoding 2,5-DKG reductase is cloned in either plasmid or phage vehicles from a genomic library created by partially digesting high molecular weight DtdA f tom Corynebacterium or other suitable source using a restriction enzyme. For 2,5-DKG
reductase, a suitable restriction enzyme is Sau 3A. (Alternatively, '_' -1 ~+-a limit digest with a restriction enzyme having greater specificity, such as BamHI or _Ps_tI, may be used.) 'The restriction digest is then ligated to either plasmid vectors replicable in suitable bacterial hosts, or into phade sequences capable of propagation in convenient bacterial cultures., The a~esulting plasmid and phage libraries are then screened using probes constructed based on the known partial sequence of the 2,'.>-DKG rE~ductase protein (See Example 2). The efficiency of probe design may be improved by selecting for probe consf ruction those codons ~~rhich are known to be preferred by bacterial hosts. identification of the desired clones from the plasmid.~nd phage i,ibraries is best effected by a set of probes such that the desired gene will hybridize to all of the probes under suitable stringency conditions and false positives from the use of only one probe elirninated. Upon recovery of colonies or phage successful in hybridizing with the oligonucleotides provided as probes, identity of the sequence with the desired gene is confirmed by direct sequencing of the DNA and by in vivo expression to yield the desired enzyme..
The complete functional gene is ligated into a suitable expression vector containing a promoter and ribosome binding site operable in the host cell into which the coding sequence will be transformed. In tine current state of the art, there are a number of promotion/ control systems and suitable prokaryotic hosts available which are appropriate to the present invention. Similar hosts can be used both for cloning .and for expression since prokaryotes are, in general, preferred for cloning of D1JA sequences, and the method of 2-KLG production is most conveniently associated with such microbial systems. E. coli K12 strain 294 (ATCC No. 31446) is particularly useful as a cloning host. Other microbial strains 30~
which may be used include E. coli strains such as E. coli B, E. coli X1776 (ATTC No. 31537) and E. coli DH-1 (ATCC No. 33849). For expression, the aforementioned strains, as well as E. coli 4!3110 (F-, a-, prototrophic, ATTC No. 27325), bacilli such as Bacillus subtilus, and other enterobacteriaceaE~ such as Salmonella typhimurium or Serratia _marcesans, and various Pseudomonas species may be used. A particularly preferred group of hosts includes those cultures which are capable of converting glucose or other commonly available metabolite to 2,5-DKG. Examples of such hosts include Erwinia herbicola ATTC No. 21998 (also considered an Acetomonas albosesamae in U.S. Patent 3,998,697); Acetobacter melanoaeneum, IFC
3293, Acetobacter cerinus_, IFO 3263, a;nd Gluconobacter rubiainosus, IFO 3244.
In general, plasmid expression or cloning vectors containing replication and control sequences which are derived from species compatible with t:he host cell are used in connection with these hosts. The vector ordinarily c<irries a replication site, as well as marking sequences: which are capable of providing phenotypic selection in transformed cells. For c=_xample, E. coli is typically transformed using pBR322" a plasmid derived from an E.
coli species (Bolivar, et al., Gene 2.. 95 (1977)). pBR322 contains genes for ampicillin and tet:racycliine resistance and thus provides easy means for identifying transformed cells. For use in expression, the pBR322 plasmid, or other microbial pl~asmid must also contain, or be modified to contain, prornoters which can be used by the microbial organism for expression of its ovm proteins. Those promoters most commonly used in recombin ant DNA construction include the s-lactamase (penicillinaae) and lactose promoter systems (Chang et al, Nature, 27~i: 615 (1978); Itakura, et al, Science, 198: 1056 (1977); (Goeddel, et al Nature 281: 544 (1979)) and a tryptophan _ (trp) promoter system (Goeddel, et al, Nucleic Acids Res., 8: 4057 (1980); EPO Appl Publ No. 0036776). k~hile these are the most commonly used, otiner microbial promoters have been discovered and utilized, and details coincerning their nucleotide sequences have been published, enabling a skilled worker to ligate them functionally in operable relationship to genes in transformation vectors (Siebenli:;t, et al, Cell 20: 269 (1980)).

By suitable cleavage and ligation of DNA sequences included in the aforementioned vectors and promoters with gene sequences prepared as outlined above encoding 2.,5-DKG reductase, and by deleting any unnecessary or inhibitory sequences, prokaryotic host cells are transformed sv as to be caused to produce the enzyme. The enzyme may then either be purified as outlined above, the intact or broken cells used directly as catalysts, or, alternatively, the host may be chosen so that once transformed it is capable of effecting the entire conversion of glucose or other suitable metabolite to the desired 2-KLG product.
B.4 Conversion of Glucose or Other Metabolite to 2-KLG by a Single Recombinant: Organism The availability of recombinant techniques to effect expression of enzymes in foreign hosts. permits the achievement of the aspect of the invention which envisions production of 2-KLG in a single host organism from a readily available metabolite. This method has considerable advantage over presently used methods in that a single viable organism fermentation is substituted for two fermentations, and there is at least a partial balance of the oxidizing and redu~~ing equivalents required for this conversion. At present there is no naturally occurring organism which is kno~m to be capable of catalysis of this entire sequence of steps.
Organisms are, however, known which effect the conversion of glucose or other ordinary metabolic substrate, such as, for example, galactose or fructose into 2,5-DKG. Another group of organisms is known which effects the conversion of the 2,5-DKG into 2-KLG, the latter conversion, of course, being catalyzed by a single enzyme within that organism, but utilizing the power of that organism to supply reducing equivalents.
One approach to producing a single organism conversion that is included in this invention comprises construction of an -m- 1 341 08 5 expression vector for 2,5-DKG reductase as outlined above, and transformation of this vector into ce'ils which are capable of the initial conversion of ordinary metabolites into the 2,5-DKG
substrate for this enzyrne. As outlined in Example 3 below, this transformation results in a single organism 2-KLG factory. The details of the vector construction, transformation, and use of the resultant organism in the transformation are outlined in the herein specification.
An alternative approach is to clone and express the genes encoding..the enzymes known to effect t;he conversion of glucose or other ordinary metabolite to 2,5-Df;G from the organisms known to contain them (as enumerated above) to construct expression vectors containing these cloned gene sequences, and using such vectors ~5 transform cells which normally producE~ the reductase. A third approach is to transform a neutral host with the entire sequence of enzymes comprising the ordinary metabolite to 2-KLG scheme. This last approach offers the advantage of choice of host organism almost at will, for whatever desirable growth characteristics and 20 nutritional requirements it may have. Thus, the use as host cells of organisms which have the heritage of a reasonable history of experience in their culture and gro~rth, such as _E. coli and Bacillus, confers the advantage of uniformity with other procedures involving bacterial production of enzymes or substrates.
Once the organism capable carrying out the conversion has been created, the process of the invention may be carried out in a variety of ways depending on the nature of the construction of the expression vectors for the recombinant. enzymes and upon the growth characteristics of the host. Typically, the host organism will be grown under conditions which are favorable to production of large quantities of cells and under conditions which are unfavorable for the expression of any foreign genes encoding the enzymes involved in the desired conversion. When~a large number of cells has accumulated, suitable inducers or derepressors are added to the ~ 341 ~8 5 medium to cause the promoters supplied with such gene sequences to become active permitting the transcription and translation of the coding sequences. Upon suitable expression of these genes, and hence the presence of the desired catalytic quantities of enzyme, the starting material, such as glucose, is added to the medium at a level of 1-500 g/L and the culture maintained at 20°C to about 40°C, preferably around 25-37°C for several hours until conversion to 2-KLG is effected. The starting material concentration may be maintained at a constant level through continuous feed control, and the 2-KLG produced is recovered from the medium either batchwise or continuously by means known in the art.
C. General Metho~Js Employed in the Invention In the examples below, the following general procedures were used in connection with probe construction, screening, hybridization of probe to desired material and in vector construction.
C.1 Probe Preparation Synthetic DNA probes were prepared by the method of Crea, R. and Horn, T., Nucleic Acids Res., 8: 2231 (1980) except that 2,4,6-triisopropylbenzenesulfonyl-3-nitro-1,2,4-triazole (TPS-NT) was used as coupling agent (de Rooij, J. et al, Rec. Tray. Chim. Pads-Bas, 98: 537 (1979)).
C.2 Isolation of Plasmids. Cleavage with Restriction Enz, Plasmids were isolated from the identified cultures using the cleared lysate method of Clewell, D.I3. and Helinski, Biochemistry 9: 4428 (1970), and purified by column chromatography on Biorad* A-50 Agarose. Smaller amounts (mini-preps) were prepared using the procedure of Birnboim, H.C. Nucleic Acids Research, 7:1513(1979).
* Trade mark .~ . _ .- -19_ 1 341 ~8 5 Fragments of the cloned plasmids were prepared for sequencing by treating about 20 i~g of plasmids with 10 to 50 units of the appropriate restriction enzyme or sequence of restriction enzymes in approximately 600 i~l solution containing the appropriate buffer for the restriction enzyme used for sequence of buffers); each enzyme incubation was at 37°C for one hour. After incubation with each enzyme, protein was removed and nucleic acids recovered by phenol-chloroform extraction and ethanol precipitation.
Alternatively, plasmids were fragmented by DNAase I digestion in the Presence of MnCl2 (Anderson, S. (1981). Nucleic Acids Res. 9, 3015) or.by sonication (D~eininger, P.L. (1983). Analyt. Biochem.
129, 216).
After cleavage, the preparation was treated for one hour at 37°C
~~ith 10 units Klenow DNA polymerase or T4 DNA polymerase in 100u1 of K1 enow buffer ( 50mt~1 KPi , ~pH 7.5, 7mt~ MgC12, lmhl BME ) , contai ni ng 50 nmol dNTP. Protein was removed and nucleic acids recovered as above, and the nucleic acids suspended in 40u1 of loading buffer for loading onto 6 percent polyacrylamide gel, as described above, for sizing. (Alternatively, fragments may be cloned direc-tly into an ~i13 vector. ) DtJA sequencing was performed by the dideoxynucleotide chain termination method (Sanger, F. et al (1977). Proc. Natl. Acad. Sci.
USA 74, 5463) after cloning the fragments in an M13-derived vector (Messing et al (19.81). Nucleic Acids Res. 9, 309).
C.3 Liaation Procedures DNA fragments, including cleaved expression plasmids were ligated by mixing the desired components (e.g. vector fragment cut from 0.25 ug plasmid is mixed with insert cut from 1 ug of plasmid in 20 ul reaction .mixture), which components were suitably end tailored to provide correct matching, with T4 DNA ligase.
APProximately 10 units ligase were required for ug quantities of '~' .' -20- 1 3 4 1 ~ 8 5 vector and insert components. The resulting plasmid vectors were then cloned by transforming E. coli K12 strain 294 (ATCC 31446) or DH-1 (ATCC 33849). The transformation and cloning, and selection of transformed colonies, were carried out: as described belovr. Presence of the desired sequence was confirmed by isolation of the plasmids from selected colonies, and DNA sequencing as described above.
Preparation A
Isolation, and Characterization of Sodium (Calcium) 2,5 Diketo-D-gluconate_ Because 2,5-DKG is not readily commercially available, it was prepared by isolation from Erwinia he rbicola, (ATTC 21998) or from Acetobacter cerinus, IFO 3263. Sodium 2;5-diketo-D-gluconate was isolated from the Erwinia fermentation by passage of the broth through AG1-X8, 100-200 mesh anion exchange column, washing with water, and eluting with Ci.05N HC1. Fractions containing 2,5-diketo-D-gluconic acid vrere pooled, neutralized v~ith sodium bicarbonate to pH 5.5, and lyophilized to dryness. (See for example, Bernaerts, M. et. al., AntoniE~ van Leeuwenhoek 37: 185 (1971)). Characterization performed by methods useful for organic acids involving HPLC and TLC (Gossele, F. et al., Zbl. Bokt. Hyg., 26 1: Abt. Orig. C, 178 (1980)), 13C Nt~iR and the formation of the bis 2,4-dinitrophenylhydrazone as described by Wakisika, Y. fir. Biol.
Chem. 28: 819 (19E~4) confirmed the identity of the compound.
(Alternatively the calcium salt may be prepared by passage of the concentrated broth throucth a column o'' a cation exchange resin (Dowex 50, Ca2+ form) followed by elution with water. Fractions containing 2,5-DKG by HPLC analysis are pooled and lyophilized to give a pale yellow calcium salt.) The following examples serve to illustrate but not to limit the invention:

Example 1 -- Isolation and Purification of 2.5-DKG Reductase from Corynebacterium A. Cell Lysis and :Extraction CorSrnebacterium sp. ATCC 31090 was l;rown in a 10-liter fermenter and harvested during log-phase growth. Cell paste was recovered by centrifugation of the fermentation broth, and stoned at -20 degree; C. 450 grams of cell past was thawed, resuspended in 650 ml 20mM Tris buffer pH 8.0, 0.5M NaCI to wash the cells, and the cells re-harvested by centrifugation, followed by resuspension in 650 ml Tris buffer containing 2mg/ml lysozyme to release intracellular proteins.
Cell debris was separated from soluble material by centrifugation, and the resulting pellet reextracted with Tris buffer containing 0.1 percent (w/w) Tween 80*, a non-ionic detergent. The extracts were assayed for 2,5-DKG reductase activity and pooled.
B . Ion Exchange C:hromatogranhv The crude cell extract (1264m1) was adsorbed batchwise onto diethylaminoethylcellulose (Whatman DE-52*, 250 ml wet settled volume), and stirred for 0.5 hrs at room temperature, followed by recovery of the DE-52 resin in a glass filter funnel. The DE-52 resin was packed into a 5x30 cm column and washed with Tris buffer until baseline was established (A280=0.7, A260/A280-1.7, indicating that nucleic acids were slowly washing off the column). The column was then eluted with a 0-1M linear NaCI gradient (1200m1) at a flow rate of 110m1/hr., and fractions were collected and assayed for DKG reductase activity.
Two distinct peaks of activity catalyzing DKG reduction were found: a peak eluting at approximately 0.4M NaCI contained the desired 2,5-DKG reductase (which 3 0 converts 2,5-DKG into 2-KLG):; another eluting at approximately 0.25M NaCI
which did not convert :~,5-DKG into 2-KLG.
* Trade mark C . Affinity ChromatogranhX
The 0.4M eluting peak from the DE-52 column was dialysed overnight vs.
20mM Tris pH 8.0 and. applied 'to a 2.5 x 4.5 column of Amicon Matrex Gel Blue A; (This resin consists of agaros;e beads with a covalently linked dye (Cibacron blue F3GA*) which has an ;affinity for enzymes utilizing NADH or NADPH as a cofactor.) The column was washed with Tris buffer and eluted with l.OmM
NADP. Fractions were; collected and assayed for DKG reductase activity, and the activity pool concentrated 16-fold by ultrafiltration (Amicon stirred cell YM-membrane).
D. High Pressure Li uid Chromato~ranhy (HPLC) The concentrated materiel from the Blue A column was dialysed overnight vs. 20mM Tris pH 8.0 and applied to an Altex TSK column (0.5 x 60cm) buffered with 200 mM ammonium bicarbonate. (The TSI~ column separates proteins according to molecular weight.) The 2,5-DKG reductase activity eluted in a single peak that corresponded to a molecular weight of 45,000 daltons and showed >99 percent purity -- i.e., was homol;eneous according to this criterion.
Example 2 -- Characterization of CorSrnebacterium 2.5-DKG Reductase 2 5 A. Electrophoresis The enzyme was electrophoresed in an acrylamide gel in the presence of sodium dodecyl sulfate (SDS). Under both non-reducing and reducing conditions, a single protein band with a molecular weight of approximately 34,000 daltons was found. No protein was found in the 45,000 dalton range (as was found with HPLC).
* Trade mark B. N-terminal Amino Acid Seguence Amino acid sequence data show that the purified enzyme contains a single ~J-terminal sequence. (u = undetermined) thr val pro ser ile val leu asn asp gly asn ser ile pro gln leu gly tyr gly val phe lys val pro pro ala asp ala gln arg ala val glu glu ala leu glu val gly t,yr a his ile asp a ala a a tyr gly C~ wino Acid Comcosition Amino acid hydrolysis data gives the folloE~ring composition:
viol a AA Fercent asx 11.52 thr 4.75 ser 3.85 glx 10.47 pro 6.15 gly 8.22 ai a 14.69 cys 0.00 val 7.47 met 1.36 ile 4.82 leu 7.90 tyr 2.40 phe 2.37 his 3.67 lys 3.10 arg 5.21 trp ~ 2.05 D. t;inetic Parameters,Substrate Specificity and Cofactor Requirements - -. -2d- 1 3 4 1 0 8 5 The following assay conditions were used to determine kinetic parameters:
Sodium phosphate buffer 150mi°i, pH 6.4, 25°C
NADPH 11-300uM
Enzyme l0ug 2,5-DKG 0.43-43mM
Assay volume 1.0 ml Michaelis constants (l~.m) 2,5-DhG: 15.5mP~1 NADPEi: 33.7uh1 Waximum velocity (Vmax) 9.8 units/mg (1 unit =1.0 uMole/min) The enzyme is specific fo r NADPH. No activity was observed with 2-KDG, 5-KDG, D-gluconic acid, 2-KLG, NADH.

No alteration of activity was observed in the presence of tdg++, Mn++, Ca++, Zn++, EDTA, cystein~e, ADP, ATP.
E. pH Optimum htaximal activity was observed at pH 6.4. The enzyme is active over a broad pH range of 5.0-7.6.
F. Stereosoecificitv and Quantitative Conversion In order to ciuantitate the conversion of 2,5-DKG into 2-KLG, a reaction was carried out; containing 2,5-DKG, 1.33 mM NADPH, 0.3mg 2,5-DKG reductase in 1 ml of O.1M Tr-is:Cl pH 7.5. After 5 h at 25°C
the reaction had stopped and was analyzed for NADPH oxidation and 30~ 2-KLG production.. A change in the absorbance at 380nm corresponding to 0.40 mM NADPH oxidized was observed. HPLC analysis on an organic acids column showed a single peak corresponding to 0.42 mM 2-KLG.
No 2-KDG (2-keto~-D-gluconic acid) or 5-KDG (5-keto-D-gluconic acid) was observed. HPLC analyses were verified by analysis of the GCMS
of per-trimethylailylated derivatives prepared by addition of -...
-25- 1 3 41 0 8 5 .
trimethylsilylimid.azole/ pyridine: 50;50 to a lyophilized reaction mixture for 30 min at 90°C, on a 25 m«ter 5 percent crosslinked phenylmethylsilicone fused silica bonded capillary column. Thin layer chromatography is also consistent ~~~ith the above results.
Example 3 -- Recombinant 2,5-DKG Reductase A. Probe Design The Tm of the,Corynebacterium sue. (ATCC 31090) DNA was measured and fourib to be 81.5°C in 7.5 mM sodium phosphate, 1 mh1 EDTA (pH
6.8). This corresponds to a G+C content of 71 percent, using Pseudomonas aeru inosa DNA (Tm = 79.7°C, G+C = 67 percent) as a standard. Hence, in the constructions, those codons known to be prevalent in bacterial DNAs of high G+C content (Goug, M., et al Nucleic Acids Res. 10, 7055 (1982)) were employed: phe, TTC; lys, AAG; val, GTG; pro, CCG; ala, GCC; asp, GAC; gln, CAG; arg, CGC;
glu, GAG; asn, AAC; ser, TCC; ile, ATC; leu, CTG; gly, GGC; tyr, TAC. The partial amino acid sequence of the 2,5-DKG reductase f rpm Corynebacterium sp. (ATCC 31090) is shown in Example 2; this sequence and the above codons were used to construct suitable probes, using the method of Anderson and Kingston (Proc. Natl. Acad.
Sci. USA 80, 6838 [1983]). Two 43 mess were synthesized by the phosphotriester method of Crea, R. _et _al., Nucleic Acids Res., 8:
2331 (1980) 5'GGCCTCCTCCACGGCGCGCTGGGCGTCGGCC GGCGGCACCT?G3' and 5'CTCCATCCCGCAGCTGGGCTACGGCGTGTTCAAGGTGCCGCCG3'. The oligonucleotide probes are phoshorylated with 100 uCi [Y-32P] ATP
(5,000 Ci/mmole, Amersham) and polynucleotide kinase (P-L
Biochemicals).
B. Construction of a Plasmid Genomic Library Genomic DNA was isolated from Corynebacterium sp_. ATCC 31090 by the method of Schiller et al. Antimicro. Agents Chemotherapy _18, 814 (lgg0). Large fragments (>100kb) were purified by CsCI density gradient centrifugation and partially digested with Sau 3A. The digest was size fractionated by agaros~e gel electrophoresis into size classes of 1-2kb, 2-3kb, 3-4kb, and 4-6kb. A genomic; library was prepared from each size class of DNA
fragments using the vectors pBR:322 and pACYCl84 (Bolivar, F. et al Gene 2, 95 (1977);
Chang, A.C.Y., and Cohen, S.IV. J. Bacteriol. 1.34, 1141 (1978)) by cleavage of the BamHI site and insertion of the Sau3A fragnnents using T4 DNA ligase. The resulting plasmids were used to transform a recA- derivative of E. coli strain MM294. (ATCC 31446) or DFf-1 (ATCC 33849) employing the transformation protocol of D. Hanahan (J. Mol.. Biol., 166, 557 1983). Each genomic library contained 104 - 105 independent recombinants. (In an alternative procedure, additional plasmid libr;~ries may be prepared in pBR322 using BamHI fragments (size range 2.0-2.5 Kb) and Pst fragments (size range 0.5-1.5 Kb) of Corynebacterium~. (ATCC 31.090) DNA.) C. Screening of the Plasmi<i Library in E. coli The colonies were picked into microtiter dishes, incubated overnight, and 2 0 stamped onto nitrocellulose filters (BA85) placed on LB plates containing ampicillin or chloramphenicol. The remaining portions of the colonies in the microtiter dishes were preserved by adding 25p,1 of 42 percent DMSO and storing at -20°C.
The transferred portions of the colonies were incubated for 8 to 9 hours at 2 5 37°C and amplified by 'transferring the filters to plates containing 12.5p,g/ml chloramphenicol or spectinomyc.in and incubating at 37°C overnight.
The plasmid library in E. coli is screened by filter hybridization. The DNA
is denatured and bound to duplicate filters as described by Itakura, K. et al.
Nucleic 3 0 Acids Res. 9:879-894 ( 1981 ). T'he filters for hybridization were wetted in about 10 ml per filter of 5X SET, 5X Denhardt's solution .and 50 pg/ml denatured ~34~88.5 salmon sperm DNA, a.nd 0.1 percent sodium pyrophosphate + 20 percent formamide (5X SET =: 50 mM Tris-HCl (pH 8.0), 5mM EDTA, 500 mM NaCI; 5X
Denhardt's solution = 0.1 percent bovine serum albumin, 0.1 percent polyvinylpyrolidone, 0.1 percent Ficoll; see Denhardt, Biochem. Biophys. Res.
Comm., 23:641 (1966)). The fiilters are prehybridized with continuous agitation at 42° for 14-16 hrs, and probed with ~1 x 108 cpm of probe as prepared in subparagraph A of this Example: at 42°C. The filters that were hybridized with the probes of subparagraph A are washed with .2 x SSC, .l percent SDS, .l percent sodium pyrophosphate at 42°C for 3 x 30 rains. Each of the duplicate filters is blotted dry, mounted on cardboard and exposed to Kodak* XR5 X-Ray film with Dupont* Cronex 11R :~tra life Lightning-plus intensifying screens at -70°C for 4-24 hrs.
Cells from positive colonies were grown up and plasmid DNA was isolated by the method of Clewell and Helinski (Proc. Natl. Acad. Sci. USA 62, 1159 [1969]). DNA was fragmented with Alul and PstI and subcloned into the vectors M13mp8 and Ml3mp!) (Messing, J. and Viera, J. [1982] Gene 19, 269).
2 0 Subclones that hybridized to the; probes were sequenced using the dideoxy chain termination procedure (Sanger, F. et al., Proc. l'datl. Acad. Sci. (USA) 74:

(1977)) in order to verify that the DNA coded for the 2,5-DKG reductase.
D. Construction of Expression Vectors for 2.5-DKG Reductase Gene The 2,5-DKG reductase gene, accompanied by either its own or a synthetic ribosome binding site, is inserted 'downstream' of the E. Coli tro (or tae) promoter or the pACYC184 CA'r promoter on expression plasmids, which also contain a tetracycline resistance l;ene or other selectable matker, and an origin of replication derived from plasmids ColEl. 1.5A, or RSF1010. Some constructs may contain, in addition, an active gene coding for E. Coli tro repressor or the E.coli lac repressor which allows the expression of the 2,5-DKG reductase gene to be regulated by exogenously added indole acryliic acid or IPTG. Various mutated versions of the above plasmids are also employed for the expre;~sion of 2,5-DKG reductase.
* Trade mark .v -28- ~ 3 41 0 8 5 Construction of Expression Vector for 2.5-Df;G Reductase A cloned 2.2 Kb BamHI fragment of Corynebacterium ~ (ATCC
31090) DNA, containing a portion of the 2,5-DKG reductase gene, was isolated with the 43-mer probes as described in Example 3. An 0.12 Kb PstI/BamHI fragment of this plasmid was further used as a probe to isolate an overlapping 0.88 Kb PstI fragment of Corynebacterium sp DNA, which contained the rest of the gene. As described in Figure 2, pDKGR2 (containing the 2.2 Kb BamHI fragment) was digested with NcoI, treated with E. coli DNA polymerase I Klenow fragment and dNTPs to~create flush-ended DNA, then further digested with BamHI to release an 0.87 Kb fragment; this fragment was purified by electrophoresis on low-melting-point agarose. The plasmid pDKGR9 (containing the 0.88 Kb PstI fragment) was digested with PstI and BamHI, and the resultant 0.76 Kb fragment similarly isolated on low-melting-point agarose. The 0.87 kb NcoI/BamHI fragment and the 0.76 Kb BamHI/PstI fragment were then combined with SmaI/PstI -digested M13mp9 and ligated to yield an M13 recombinant ("mitl2") with a 1.6 Kb insert of Corynebacterium sp DNA containing the entire 2p 2,5-DKG reductase gene (Figure 2).
To mitl2 single-stranded DNA a "deletion primer" (sequence:
ACGGCCAGTGAATTCTATGACAGTTCCCAGC) and AIuI fragments of M13mp9 DNA
were annealed. This template-primer combination was then treated with E. coli DNA polymerase Klenow fragment in the presence of dNTPs and T4 DNA ligase 'to create in vitro heteroduplex mitl2 RF
molecules, as described by Adelman et al., (DNA 2, 183 (1983).
These molecules were used to transform the host tM101, and recombinant phage incorporating the desired deletion were detected 30~ bY Plaque hybridization using the deletion primer as a probe (Adelman et al., (!)NA 2, 183 (1983)). This construction was designated mitl2o (Fig. 3).
The thitl2e RF DNA was digested with EcoRI and HindIII to yield a 1.5 Kb fragment containing the 2,5-DKG reductase gene. The human -2g- 1 341 08 5 growth hormone expression plasmid, pHGH207-lptrpeRl5', (pHGH207-lptrpaRIS' is a derivative of pHGH207-1 (de Boer et al., (1983), Proc. Natl. Acad. Sci.,USA _80, 21) in which the EcoRI site between the ampicillin resistance gene and the trp promoter has been deleted), was digested with EcoRI and PstI to yield a 1.0 Kb fragment containing the E. coli trp promoter and pBR~22 was digested with PstI and HindIII to yield a 3.6 Kb fragment. These three fragments were ligated together to form an expression plasmid for 2,5-DKG reductase, designated pmitl2e/trpl (Fig. 3). This plasmid eras unable to confe r tetracycline resistance on host cells. A
plasmid that would encode a complete tetracycline resistance function was constructed as follows. pmitl2e/trpl DIJA was digested with HindIII, treated with E. coli DNA polymerase I Klenow fragment and dRITPs to produce flush ended DNA, then digested with SphI; the resultant 5.6 Y,b fragment was purified by electrophoresis on low melting agarose. :Similarly, pBR322 DNA was digested with EcoRI, treated with _E. col_i DNA polymerase I Klenow fragment and dNTPs, digested with S~hI, and the resultant 0.56 Kb fragment purified on low melting agaros~e. The 5.6 Kb and C.56 Kb fragments were then ligated together to yield a tetracycline-resistant 2,5-DKG reductase expression plasmid, designated ptrpl-:.5 (Figure 3). The DNA
sequence of the trp promoter, synthetic ribosome binding site, and 2,5-DKG reductase gene contained on this plasmid is shown in Figure 4.
E. Production of Recombinant 2,5-DKG Reductase Cells are prepared for transformation by the method of Lacy and Sparks, Phytopatholoaical Society, 69:. 1293-1297 (1979). Briefly a 30- loop of suitable host cells, Erwinia herbicola (ATCC 21998), or _E. coli MM294 (ATCC 314645), is inocu'ated into 5 ml of LB medium and incubated at 3.0°C for 14-16 hrs. A 1:100 dilution of this overnight growth is inoculated into LE>, the culture grown to OD5g0 of 0.4, and the cells recovered by centrifugation at 4°C. The pellet was resuspended in 1/3 volume of 10 mM NaCI, again -3c- 1 3 4 1 0 $ 5 centrifuged at 4°C, the pellet resuspended in an equal volume of 30 mM CaCl2, and after 60 minutes at 0°C, the cells again centrifuged at 4°C. The pellet is resuspended in 1/12 volume of 30 rr~ CaCl2 and stored at 4°C overnight. (Alternatively, cells may' , be resuspended in 30 mho CaCl2, 15 percent glycerol, and stored at -70°C.) Transformation is effected by incubation of 1 ug plasmid in 0.2 ml of competent cells at 0°C for 2 hr followed by heating to 42°C
for 1 min. 3 ml of LB broth is added and cells are allowed to recover for 4 hrs at 37°C, then cells are plated on selective medium as described by Lacy and Sparks (supra). Successful transformants are grown on LB broth to a density of O.D.550 = 1.0, then centrifuged and resuspended in rr,inimal medium in the presence of 0.2 percent glucose. IAA or IPTG is then added to the medium, and after 0.5-1.0 hrs, the cells recovered by centrifugation and lysed by treatment with lysozyme and a detergent (Tween 80).
The supernatant is aasayed for the presence of 2,5-DKG reductase 2p as outlined in Example 21D. (Table I), and the proteins analyzed by SDS polyacrylamide gel electrophoresis (Fig. 5). The protein band representing the recombiinant 2,5-DKG reductase was identified immunologically using the Western blotting procedure (Tobin, H, et al., Proc. Natl. ~4cad. Sci. USA 76, 4350 (1979)).
TABLE I
2,5-DKG reductase activity 30- Extract . (e Absorbance (340 nm) min-1/100 uL) pBR322 transformed - 0.087 Erwinia herbicola (ATCC .?1998) ptrpl-35 transforrned - 1,925 Erwinia herbicola (ATCC 21998) Example 4 -- Production of 2-KL~ntactin9 Recombinant Organism with 2,5-DKG
Cells of ptrpl-35 transformed Erw. _her_bicola (ATCC 21998) were 5streaked from a frozen glycerol stock onto LB solid medium (10 9/L
tryptone, 5 g/L yeast extract, 10 g/L tdaCl, 15 g/L agar) containing mg/L tetracycline, 'then incubated 48 hrs at 30°C. A single colony was picked and used to inoculate 5 ml of LB liquid medium containing 5 mg/L tetracycline, and this was shaken in a test tube for 16 hrs 10at 30°C. 1.0 ml of this culture was used to inoculate 100 ml LB
liquid medium containing 5 mg/L tetracyc,iine, and this was then shaken at 200 rpm for 16 hrs>. at 30°C. Cells were harvested by centrifugation, washe d once in an equal volume of fresh LB medium, then resuspended in 50 ml of LB medium containing 5 mg/L
tetracycline and 20 g/L 2,5-DY,G. A 10 ml aliquot of this was shaken in a 125 ml flask at 200 rpm for 16 hrs at 30°C. The resultant broth was analyzed by HPLC and found to contain 2.0 g/L of 2-KLG. A
control culture containing pBR322-transi'ormed Erw. herb (ATCC
21998), treated in a similar fashion, contained no 2-KLG.
Example 5 -- Production of 2-KLG f rpm Glucose by Recombinant Organism Cells of ptrpl-35-transformed Erw. herbicola (ATCC 21998) were streaked from a frozen glycerol stock onto LB solid medium (10 g/L
tryptone, 5 g/L yeast extract, 10 g/L NaCI, 15 g/L agar) containing 5 mg/L tetracycline, then incubated for 48 hrs at 30°C. A single colony was picked and used to inoculate 5 ml of LB liquid medium containing 5 mglL tetracycline, and this was shaken in a test tube for 16 hrs at 30°C. 1.0 ml of this culture was used to inoculate 100 ml of ATCC medium 1038 (3.0 g/L glucose, 5.0 g/L yeast extract, 5.0 g/L peptone, 7.5 g/L CaC03, pH 7.0) containing 5 mg/L
tetracycline, and this was shaken at 200 rpm in a 500 ml flask at 30°C for 16 hrs. Cells from 75 ml of this culture were then harvested by centrifugation, resuspended in 50 ml of fresh ATCC
medium 1038 containing 5 mg/L tetracycline and 20 g/L glycerol, and _~ ~J~__ 1341085 shaken at 200 rpm in a 500 ml flask at; 30°C for 48 hrs. The resultant broth was analyzed by HPLC and GC/MS and found to contain 1.0 g/L 2-KLG. A control culture containing pBR322-transformed Erw.
herbicola (ATCC 21998), treated in a similar fashion, contained no 2-KLG.
Example 6 -- Production of 2-KLG by the Organism of the Present Invention To demonstrate the production of 2-KLG from readily available carbon sources, the following experiment was performed.
Production of 2-KL.G from Glucose The production of 2-I;LG by Erwinia herbicola (ATCC
21998)/ptrpl-35 vras accomplished by growing the cells in a medium containing the fo'ilowing:
Yeast Extract (INestles) lOg/1 Calcium Carbonate 20 g/1 2p Corn Steep Liquor 10 g/1 Glucose 20 g/1 Tetracycline 5 mg/1 The glucose and the tetracycline were sterilized separately and added prior to inoculation.
An inoculum was prepared by adding 1.0 rnl of a fro w stock culture to 50 ml of Luria broth containing 5 g/1 glucose and 5 mg/1 tetracycline. The 250 rrnl baffled flask containing the inoculum was ~ incubated at 30°C for lf.hours with shaking.
A 250 ml baffled flask was filled with 50 ml of the production medium above and inoculated with 1 ml of the inoculum. The flask was incubated with shaking at 30°C. The pH of the medium at the time of inoculation was 5.1 due to the acidity of corn steep -33- 1 3 41 ~ 8 5 liquor. After 57 hours the pH had risen to 8.71 and 2-KLG was shown to be present at a concentration of O.fi mg/ml by HPLC. The presence of 2-KLG was confirmed by HPLC and GC-t~1ass Spectrometry.

Claims (37)

1. A recombinant plasmid comprising a DNA segment encoding the enzyme 2,5-diketogluconic acid reductase, the DNA segment having the nucleotide sequence as set out in Figure 4 or variations thereof which encode for

2,5-diketogluconic acid reductase activity.
2. A process for producing a recombinant microorganism capable of converting glucose to 2-keto-L-gulconic acid which comprises transforming a host cell capable of converting glucose to 2,5-diketogluconic acid with an expression vector which encodes for 2,5-diketogluconic acid reductase activity.

3. The process of Claim 2 wherein said host cell is of the genus Erwinia.

4. The process of Claim 2 wherein said host cell is Erwinia herbicolau (ATCC 21998).

5. A process according to any one of Claims 2 to 4 wherein said expression vector contains a DNA sequence as set out in Figure 4 or variations thereof which encode for 2,5-diketogluconic acid reductase activity.

6. A process for constructing a replicable expression vehicle capable of effecting the expression of a DNA
sequence encoding 2,5-diketogluconic acid reductase, which process comprises linking said DNA sequence operably to a promoter compatible with a bacterial host.

7. A process according to Claim 6 wherein said DNA
sequence is as set out in Figure 4 or variations thereof which encode for 2,5-diketogluconic acid reductase activity.

8. A process for producing 2,5-diketogluconic acid reductase which comprises culturing recombinant host cells containing a recombinant vector capable of effecting the expression of 2,5-diketogluconic acid reductase.

9. A process according to Claim 8 wherein said recombinant vector contains a DNA sequence as set out in Figure 4 or variations thereof which encode for 2,5-diketogluconic acid reductase activity.

10. A recombinant expression vector capable of effecting the expression of a DNA sequence encoding 2,5-diketogluconic acid reductase in prokaryotic host cells.

11. A recombinant expression vector according to Claim wherein said expression vector contains a DNA sequence as set out in Figure 4 or variations thereof which encode for 2,5-diketogluconic acid reductase activity.

12. Microbial host cells or cell culture transformed with the expression vector of Claim 10 or 11.

13. A recombinant prokaryotic cell or cell culture containing the DIVA sequence encoding 2,5-diketogluconic acid reductase and capable of effecting its expression.

14. A recombinant prokaryotic cell or cell culture according to Claim 13 wherein the DNA segment encoding the enzyme 2,5-diketogluconic acid reductase is as set out in Figure 4 or variations thereof which encode for 2,5-diketogluconic acid reductase activity.

15. A process for converting 2,5-diketogluconic acid into 2-keto-L-gluconic acid which comprises contacting a mixture comprising 2,5-diketogluconic acid with a culture containing recombinant host cells expressing the gene for 2,5-diketogluconic acid reductase.

16. A process according to Claim 15 wherein said expression vector contains a DNA sequence as set out in Figure 4 or variations thereof which encode for 2,5-diketogluconic acid reductase activity.

17. The process of Claims 15 or 16 comprising in addition converting the 2-keto-L-gluconic acid into ascorbic acid.

18. Plasmid pmit 12.

19. Plasmid pmit 12.DELTA..

20. Plasmid pmit 12.DELTA./trp.

21. Plasmid ptrp 1-35.

22. The DNA sequence as set out in Figure 4 or allelic variations thereof.

23. The DNA sequence of Claim 22 comprising the 296th through the 1854th nucleotide.

24. A recombinant microorganism selected from the group consisting of Erwinia, Acetomonas, Acetobacter and Gluconobacter species containing a DNA segment as set out in Figure 4 or variations thereof which encode for 2,5-diketogluconic acid reductase activity capable of converting glucose or an other carbon source commonly utilized by bacteria for growth, to 2-keto-L-gluconic acid.

25. The recombinant microorganism of Claim 24 which is of the genus Erwinia.

26. The recombinant microorganism of Claim 24 wherein the ordinary metabolite is glucose.

27. A process for converting glucose or other carbon source commonly utilized by bacteria for growth to 2-keto-L-gulconic acid which process comprises culturing in a medium containing said glucose or other carbon source under suitable metabolic conditions, recombinant bacterial cells containing an expression vector effective in expressing a DNA sequence encoding 2,5-diketogluconic acid reductase capable of effecting the conversion of said glucose or ordinary carbon source into 2-keto-L-gulconic acid.

28. The process of Claim 27 wherein the ordinary metabolite is glucose.

29. The process of Claim 27 wherein the 2-keto-L-gulconic acid is converted into ascorbic acid.

30. The process of Claim 27 wherein the recombinant cell is of the genus Erwinia.

31. A process according to any one of Claims 27 to 30 wherein said DNA sequence is as set out in Figure 4 or variations thereof which encode for 2,5-diketogluconic acid reductase activity.

32. The process of converting glucose into 2-keto-L-gulconic acid which comprises:
(a) transforming a prokaryotic host cell capable of effecting the conversion of said glucose to 2,5-diketogluconic acid with an expression vector encoding and capable of expressing the DNA sequence for 2,5-diketogluconic acid reductase; and (b) culturing said transformed prokaryotic host cell in a medium containing said glucose under suitable metabolic conditions.

33. The process of Claim 32 wherein said host cell is of the genus Erwinia.

34. The process of Claim 32 wherein said host cell is Erwinia herbicola ATCC 21998.

35. A process according to Claim 32, 33 or 34 wherein said DNA sequence is as set out in Figure 4 or variations thereof which encode for 2,5-diketogluconic acid reductase activity.

36. A process for production of 2-keto-L-gulconic acid which comprises cloning and expressing in a culture of Erwinia herbicola the 2,5-diketogluconic acid reductase gene.

37. A process according to Claim 36 wherein said DNA
sequence is as set out in Figure 4 or variations thereof which encode for 2,5-diketogluconic acid reductase activity.