PROCESS FOR THE STEREOSELECTIVE ENZYMATIC REDUCTION OF KETO COMPOUNDS

Information

  • Patent Application
  • 20120040425
  • Publication Number
    20120040425
  • Date Filed
    March 03, 2010
    14 years ago
  • Date Published
    February 16, 2012
    12 years ago
Abstract
In a process for the stereoselective, in particular enantioselective enzymatic reduction of keto compounds to the corresponding chiral hydroxy compounds, wherein the keto compounds are reduced with an enantioselective, NADH-specific oxidoreductase, a polypeptide is used for reducing the keto compounds, which polypeptide exhibits an R-ADH-signature H-[P; A]-[I; A; Q; V; L]-[G; K]-R at positions 204-208 and the following further structural features in their entirety: (i) an N-terminal Rossmann-Fold GxxxGxG,(ii) an NAG-motif at position 87,(iii) a catalytic triad consisting of S 139, Y 152 and K 156,(iv) a negatively charged amino acid moiety at position 37,(v) two C-terminal motifs in the dimerization domain [A; S]-S-F and [V; I]-DG-[G; A]-Y-[T; C; L]-[A; T; S]-[Q; V; R; L; P],(vi) Val or Leu at position 159 (4 positions downstream of K 156),(vii) Asn at position 178, and(viii) a proline moiety at position 188.
Description

The present invention relates to a process for the stereoselective, in particular enantioselective enzymatic reduction of keto compounds to the corresponding chiral hydroxy compounds, wherein the keto compounds are reduced with an enantioselective, NADH-specific oxidoreductase. The invention relates in particular to the use of selected oxidoreductases in this process.


Optically active hydroxy compounds are valuable chirons with broad applicability for the synthesis of pharmacologically active compounds, aromatic substances, pheromones, agricultural chemicals and enzyme inhibitors. Thereby, an increasing demand for chiral compounds and thus chiral synthesis technologies can be noted particularly in the pharmaceutical industry, since, in the future, racemic compounds will hardly be used as pharmaceutical preparations.


The asymmetric reduction of prochiral keto compounds is a sector of stereoselective catalysis, wherein biocatalysis constitutes a powerful competitive technology versus chemical catalysis. The chemical asymmetric hydrogenation requires the use of highly toxic and environmentally harmful heavy metal catalysts, of extreme and thus energy-intensive reaction conditions as well as large amounts of organic solvents. Furthermore, these methods are often characterized by side reactions and insufficient enantiomeric excesses.


In nature, reductions of prochiral keto compounds to hydroxy compounds and oxidations proceeding inversely occur in numerous biochemical pathways, both in the primary metabolism and in the secondary metabolism, in every organism and are catalyzed by different types of secondary alcohol dehydrogenases (ADH) and oxidoreductases. Normally, these enzymes are cofactor-dependent.


In various sequencing projects, numerous amino acid sequences have, in each case, been identified for different organisms, which are presumably oxidoreductases of an unknown activity, function and enantio- and chemoselectivity, respectively. The basic suitability of oxidoreductases for use in industrial processes could recently be demonstrated on the basis of numerous examples.


Recently, it has been possible to show that the use of isolated oxidoreductases in aqueous/organic two-phase systems with organic solvents is extremely efficient, economical and feasible also at high concentrations (>5%). In the described systems, the keto compound to be reduced, which usually is poorly soluble in water, thereby forms the organic phase together with the organic solvent. Also, the organic solvent itself can partly be dispensed with. In that case, the organic phase is formed from the keto compound to be reduced (EP 1 383 899). Coenzyme regeneration is realized by the concurrent oxidation of secondary alcohols, for which, in most cases, the inexpensive water-miscible 2-propanol is used (WO 2006/087235). The use of 2-methyl-4-pentanol or 2-octanol for the regeneration of the cofactors NADH or NADPH is likewise advantageous (WO 2007/036257).


Numerous oxidoreductases of different specificities and selectivities can be found in the patent literature of recent years. Various examples can be found for S-specific oxidoreductases and dehydrogenases of high enantioselectivity, which, almost without exception, use NADH as a cofactor and thus can be used very economically under the above-described process conditions. Examples of such S-specific oxidoreductases are carbonyl reductases from Candida parapsilosis (CPCR) (U.S. Pat. No. 5,523,223 and U.S. Pat. No. 5,763,236; Enzyme Microb. Technol. 1993 November; 15(11):950-8) and Pichia capsulata (DE 10327454.4), carbonyl reductases from Rhodococcus erythropolis (RECR) (U.S. Pat. No. 5,523,223), Norcardia fusca (Biosci. Biotechnol. Biochem., 63(10) (1999), p. 1721-1729; Appl. Microbiol. Biotechnol. 2003 September; 62(4):380-6; Epub 2003 Apr. 26) and Rhodococcus ruber (J. Org. Chem. 2003 Jan. 24; 68(2):402-6). Furthermore, S-specific carbonyl reductases from Rhodotorula mucillaginosa, Microbacterium spec., Gordonia rubripertincta, Pichia stipitis and Pichia trehalophila (WO 2007/012428) are described.


Examples of R-specific secondary alcohol dehydrogenases are identified from organisms of the genus Lactobacillus: Lactobacillus kefir (U.S. Pat. No. 5,200,335), Lactobacillus brevis (DE 19610984 A1; Acta Crystallogr. D Biol. Crystallogr. 2000 December; 56 Pt 12:1696-8), Lactobacillus minor (DE 10119274) and Leuconostoc carnosum (WO 2007/012428). All these enzymes have the disadvantage that NADPH is required as a cofactor, which is substantially more expensive than NADH and thus involves relatively high process costs.


“R-specific” oxidoreductases are understood as those which reduce unsubstituted carbonyl compounds, such as, for example, 2-butanone, 2-octanone or acetophenone, to the corresponding R-hydroxy compounds, i.e., R-2-butanol, R-2-octanol and R-2-phenyl ethanol. 4-Halo-3-oxobutyric acid ester, for example, is reduced to S-4-halo-3-hydroxybutyric acid ester by R-specific oxidoreductases.


“S-specific oxidoreductases” are ADH which reduce unsubstituted carbonyl compounds, such as, for example, 2-butanone, 2-octanone or acetophenone, to the corresponding S-hydroxy compounds, i.e., S-2-butanol, S-2-octanol and S-2-phenyl ethanol. 4-Halo-3-oxobutyric acid ester, for example, is reduced to R-4-halo-3-hydroxybutyric acid ester by S-specific oxidoreductases.


In the present application, by the term “specific” is understood that the corresponding enantiomer is formed by >95%.


Furthermore, a number of unspecific NADPH-dependent enzymes from yeasts of the genus Candida magnoliae are described in the most recent patent literature (EP 1152054 A1, WO 2007/033928, WO 2006/046455, WO 2005/033094). Besides their NADPH-dependence, they have the major disadvantage that they cannot be used in processes with a substrate-coupled coenzyme regeneration, but always require a second enzyme system for the coenzyme regeneration. Glucose dehydrogenase with glucose as a cosubstrate is preferably used for this purpose. In addition, these enzymes have the disadvantage that they are normally unspecific, i.e., the reduction of unsubstituted carbonyl compounds, such as, for example, 2-butanone, 2-octanone or acetophenone, normally leads to racemic alcohols; only in exceptional cases, these enzymes operate selectively. Therefore, the enatioselective reduction of substrates such as 4-halo-3-oxobutyric acid ester, 3-oxoesters or aliphatic ketones, e.g., 2-octanone, is normally not possible with these enzymes.


Only a very limited number of robust, NADH-dependent and clearly R-specific oxidoreductases which are usable in processes with a substrate-coupled coenzyme regeneration with secondary alcohols, preferably 2-propanol, as cosubstrates are available.


A NADH-dependent R-specific oxidoreductase is identified, for example, from Pichia finlandica (EP 1179 595 A1). However, it exhibits a very small activity with 2-propanol and hence is unsuitable for processes with a substrate-coupled coenzyme regeneration.


Two further NADH-dependent R-specific oxidoreductases from Pichia farinosa and Candida nemodendra are described in WO 2007/012428. It has been possible to show therein for the enzyme from Pichia farinosa that a coenzyme regeneration with 2-propanol is also possible, but only at concentrations of up to 15% (v/v) isopropanol.


Furthermore, NADH-dependent enzymes from Leifsonia (U.S. Pat. No. 7,172,894 B2, Biosci. Biotechnol. Biochem., 70 (2), 2006, pages 418-426) and Devosia (WO 2004/027055) are known, wherein at least the enzyme from Leifsonia should be R-specific and capable of a substrate-coupled coenzyme regeneration.


However, the oxidoreductases known today, in particular the clearly R-specific enzymes, are not nearly sufficient for exploiting the entire market potential of stereoselective reductions of keto compounds. On the one hand, this can be explained by the fact that the individual enzymes have very different properties with respect to substrate spectrum, pH optimum as well as temperature and solvent stabilities, which often supplement each other. Therefore, even relatively similar homologous enzymes may exhibit a completely different conversion behaviour with regard to one particular substrate. On the other hand, not nearly all of the enzymes described are cloned and overexpressible to a sufficient extent, which means that these enzymes are not available for industrial use.


For exploiting the synthetic potential of the enzymatic asymmetric hydrogenation as extensively as possible, it is therefore necessary to be in possession of a portfolio of different industrially accessible oxidoreductases which is as broad as possible. Meanwhile, the requirements for enzymes which advantageously are usable industrially can be clearly defined.


Therefore, it is the object of the present invention to identify and provide stable, NADH-dependent and R-specific oxidoreductases which are suitable for processes for the reduction of keto compounds to chiral alcohols, using a substrate-coupled coenzyme regeneration.


For this purpose, it is necessary to provide oxidoreductases with a stability as high as possible against secondary alcohols, in particular against isopropanol. Moreover, the enzymes to be provided should be characterized by good expressibility in Escherichia coli (>500 units/g E. coli wet biomass).


Due to the high demand made on stability and expressibility in E. coli, it has been assumed in the present invention that bacterial enzymes are to be preferred over enzymes from yeast.


Furthermore, it is the object of the present invention to identify amino acid sequences and gene sequences, respectively, which, in a recombinantly expressed state, constitute stable NADH-dependent, (R)-specific oxidoreductases and are suitable in process engineering if an enzyme-coupled coenzyme regeneration is used.


The inventors have found that such sequences generally exhibit the known features of a “short-chain” dehydrogenase, such as, e.g., an N-terminal Rossmann-Fold GxxxGxG for cofactor binding, an NAG-motif at position 87 which provides for a stabilization of the central β-pleated sheet structure and a highly conserved catalytic triad consisting of S 139, Y 152 and K 156 in the active centre. For the purposes of the present application, the numerical allocation of individual homologous amino acid moieties in the protein sequences is carried out by the “multi-way protein alignment algorithm” with Score Matrix BLOSUM62 and the 3alpha,20beta-hydroxysteroid dehydrogenase from Streptomyces hydrogenans, Accession Code 2bhd_Strex as a reference.


With the plurality of amino acid sequences available on the internet which exhibit this signature, a definition of further features was required in order to restrict this sequence pool. The inventors therefore claimed that (R)-specific oxidoreductases having the described properties would have to exhibit, in addition, the following sequence features and patterns:

    • a negatively charged amino acid moiety at position 37
    • two C-terminal motifs in the dimerization domain: [A; S]-S-F and [V; I]-DG-[G;A]-Y-[T; C; L]-[A; T; S]-[Q; V; R; L; P]
    • an R-ADH-signature at positions 204 to 208: H-[P; A]-[I; A; Q; V; L]-[G; K]-R
    • Val or Leu at position 159 (4 positions downstream of K 156)
    • Asn at position 178
    • a proline moiety at position 188.


According to this pattern, the following sequences without an existing functional classification were selected from the sequence pool accessible from sequencing projects, were cloned and characterized with regard to their functions.













TABLE 1





SEQ ID NO
Origin
Length
Accession no.
Closest homologues







SEQ ID NO:1
Bacillus cereus
247 aa
NP-833194
Glucose-1-dehydrogenase



ATCC 14579


from Lysinibacillus






59% identity


SEQ ID NO:2
Methylibium
257 aa
YP-001020081
Short-chain dehydrogenase



petroleiphilum


Flavobacterium



Pm1


59% identity


SEQ ID NO:3
Mesorhizobium
251 aa
YP-676781
Putative oxidoreductase



sp.


Sinorhizobium






62% identity


SEQ ID NO:4
Streptomyces
263 aa
NP-631416
short-chain dehydrogenase



coelicolor


Comamonas testosteroni






57% identity


SEQ ID NO:5
Ralstonia
253 aa
YP-299653
short-chain dehydrogenase





(epimerase/dehydratase/
Bacillus





reductase SDR)
54% identity


SEQ ID NO:6
Herpentosiphon
252 aa
YP-001544828
short-chain dehydrogenase



aurantiacus


Comamonas testosteroni



ATCC 23779


54% indentity


SEQ ID NO:7
Paracoccus
249 aa
ZP-00631061
short-chain dehydrogenase,



denitrificans


Pedobacter






57% identity











2bhd_Strex Seq ID No 1 Seq ID No 2 Seq ID No 3 Seq ID No 4 Seq ID No 5 Seq ID No 6 Seq ID No 7


embedded image







2bhd_Strex Seq ID No 1 Seq ID No 2 Seq ID No 3 Seq ID No 4 Seq ID No 5 Seq ID No 6 Seq ID No 7


embedded image







2bhd_Strex Seq ID No 1 Seq ID No 2 Seq ID No 3 Seq ID No 4 Seq ID No 5 Seq ID No 6 Seq ID No 7


embedded image







2bhd_Strex Seq ID No 1 Seq ID No 2 Seq ID No 3 Seq ID No 4 Seq ID No 5 Seq ID No 6 Seq ID No 7


embedded image







2bhd_Strex Seq ID No 1 Seq ID No 2 Seq ID No 3 Seq ID No 4 Seq ID No 5 Seq ID No 6 Seq ID No 7


embedded image







2bhd_Strex Seq ID No 1 Seq ID No 2 Seq ID No 3 Seq ID No 4 Seq ID No 5 Seq ID No 6 Seq ID No 7


embedded image







Multi-way amino acid sequence comparison of selected proteins. Scoring matrix BLOSUM 62. Illustrated in shaded form: positionally specific identical amino acid moieties, compared to a reference sequence of 3alpha,20beta-hydroxysteroid dehydrogenase from Streptomyces hydrogenans, Accession Code 2bhd_Strex. The sequence motifs claimed by the inventors are framed.






Surprisingly, it has been possible to show for all seven sequences that they are NADH-dependent R-ADHs which possess sufficient stability, in particular against 2-propanol. Furthermore, it has been possible to show for all seven enzymes that they are suitable for enzymatic reduction processes with a cosubstrate-coupled coenzyme regeneration.


This is a surprising and unexpected result, particularly since the sequences exhibit a distinctly low degree of homology among each other (Table 2) and, even in sequence comparisons (Pubmed/Blast), no related sequences could be identified which would have given a functional hint in this direction (Table 1).


On the other hand, it has in turn been possible to show that the present enzymes differ sufficiently in their enzymatic properties and complement each other with regard to their process suitability, in particular with specific targets.


Thus, the use of the aforesaid oxidoreductases in the process according to the invention represents no equivalent alternative to the established prior art, but a broadening of the existing spectrum.









TABLE 2







Two-by-two amino acid sequence comparison of selected oxidoreductases

















SEQ ID
SEQ ID









NO: 1
NO: 2
SEQ ID


SEO ID





Bacillus


Methylibium

NO: 3
SEQ ID

NO: 6
SEQ ID





cereus


petro-


Meso-

NO: 4
SEQ ID

Herpento-

NO: 7


SEQ ID

ATCC

leiphilum


rhizobium


Streptomyces

NO: 5

siphon


Paracoccus



NO
Organism
14579
Pm1
sp.

coelicolor


Ralstonia


aurantiacus


denitrificans






SEQ ID

Bacillus

100
37/5
41/2
40/6
54/2
41/3
38/2


NO: 1

cereus




ATCC



14579


SEQ ID

Methylibium

37/5
100
37/6
33/8
41/5
39/4
43/4


NO: 2

petro-





leiphilum




Pm1


SEQ ID

Meso-

41/2
37/6
100
56/6
41/2
56/0
64/0


NO: 3

rhizobium




sp.


SEQ ID

Streptomyces

40/6
33/8
56/6
100
43/3
53/5
54/5


NO: 4

coelicolor



SEQ ID

Ralstonia

54/2
41/5
41/2
43/3
100
45/1
46/4


NO: 5


SEO ID

Herpento-

41/3
39/4
56/0
53/5
45/1
100
53/1


NO: 6

siphon





aurantiacus



SEQ ID

Paracoccus

38/2
43/4
64/0
54/5
46/4
53/1
100


NO: 7

denitrificans










The degree of homology in Table 2 is indicated as percent of identical amino acid moieties and as the number of shifts in an optimum pairwise alignment. The optimum alignment is thereby determined by means of the BLAST-algorithm (Basic Local Alignement Search Tool) (Altschul et al. 1990, Proc. Natl. Acd. Sci. USA. 87: 2264-2268).


As a basis, the PAM30 matrix is used as a scoring matrix for evaluating the sequence similarity (Dayhoff; M. O., Schwarz, R. M., Orcutt, B. C. 1978. “A model of evolutionary change in Proteins” in “Atlas of Protein Sequence and structure” 5(3) M. O. Dayhoff (ed) 345-352, National Biomedical Research foundation).


It is the object of the invention to provide a process for the enantioselective enzymatic reduction of keto compounds to the corresponding chiral hydroxy compounds which covers a broad substrate spectrum and yields high turnovers under process technological conditions.


According to the invention, said object is achieved by a process of the initially mentioned kind which is characterized in that, for reducing the keto compounds, a polypeptide is used which exhibits an R-ADH-signature H-[P; A]-[I; A; Q; V; L]-[G; K]-R at positions 204-208 and the following further structural features in their entirety:

  • (i) an N-terminal Rossmann-Fold GxxxGxG,
  • (ii) an NAG-motif at position 87,
  • (iii) a catalytic triad consisting of S 139, Y 152 and K 156,
  • (iv) a negatively charged amino acid moiety at position 37,
  • (v) two C-terminal motifs in the dimerization domain [A; S]-S-F and [V; I]-DG-[G; A]-Y-[T; C; L]-[A; T; S]-[Q; V; R; L; P],
  • (vi) Val or Leu at position 159 (4 positions downstream of K 156),
  • (vii) Asn at position 178, and
  • (viii) a proline moiety at position 188.


According to a preferred embodiment of the process according to the invention, a polypeptide is used for reducing the keto compounds

  • (a) which comprises one of the amino acid sequences SEQ ID NO:1 to SEQ ID NO:7, or
  • (b) in which at least 70% of the amino acids are identical to the amino acids of one of the amino acid sequences SEQ ID NO:1 to SEQ ID NO:7, or
  • (c) for which a nucleic acid sequence from the group consisting of SEQ ID NO:8 to SEQ ID NO:14 encodes, or
  • (d) for which a nucleic acid sequence encodes which hybridizes to one of the nucleic acid sequences mentioned in (c) under stringent conditions.


By a nucleic acid sequence which hybridizes, for example, to SEQ ID NO:14 under stringent conditions, a polynucleotide is understood which can be identified via the colony hybridization method, the plaque hybridization method, the Southern hybridization method or comparable methods, using SEQ ID NO:14 or partial sequences of SEQ ID NO:14 as a DNA probe. For this purpose, the polynucleotide immobilized on a filter is hybridized, for example, to SEQ ID NO:14 in a 0.7-1 M NaCl solution at 65° C. Hybridization is carried out as described, for instance, in Protocol 32 of Molecular Cloning, A Laboratory Manual, Second Edition (Cold Spring Harbor Laboratory Press, 1989) or in similar publications. Subsequently, the filter is washed with a 0.1 to 2-fold SSC solution at 65° C., wherein a 1-fold SSC solution is understood to be a mixture consisting of 150 mM NaCl and 15 mM sodium citrate.


A polynucleotide which hybridizes to the polynucleotide SEQ ID NO:8, SEQ ID NO:9, SEQ ID NO:10, SEQ ID NO:11, SEQ ID NO:12, SEQ ID NO:13 or SEQ ID NO:14 under the above-mentioned stringent conditions should exhibit at least 68% sequence identity to the polynucleotide sequence SEQ ID NO:8, SEQ ID NO:9, SEQ ID NO:10, SEQ ID NO:11, SEQ ID NO:12, SEQ ID NO:13 or SEQ ID NO:14, respectively, better an identity of at least 80%, even better an identity of 95%.


According to a further preferred embodiment, keto compounds of general formula I are used,




embedded image


wherein R1, R2 and R3 independently of each other are selected from the group consisting of


1) —H, provided that R1 is not H,


2) —(C1-C20)-alkyl, wherein alkyl is linear-chain or branched,


3) —(C2-C20)-alkenyl, wherein alkenyl is linear-chain or branched and optionally contains up to four double bonds,


4) —(C2-C20)-alkynyl, wherein alkynyl is linear-chain or branched and optionally contains up to four triple bonds,


5) —(C6-C24)-aryl,


6) —(C1-C8)-alkyl-(C6-C14)-Aryl,


7) —(C5-C14)-heterocycle,


8) —(C3-C7)-cycloalkyl,


wherein the moieties mentioned above under 2) to 8) are unsubstituted or substituted one, two or three times, independently of each other, by —OH, halogen, —NO2, —NH2, —NHP and/or -M, wherein P stands for —(C1-C7)-alkyl, —(C2-C7)-alkenyl, —(C2-C7)-alkynyl, —(C6-C14)-aryl or a protective group selected from benzyloxy carbonyl, triphenyl methyl and t-butyl carbonyl, and M stands for —(C1-C7)-alkyl, —(C2-C7)-alkenyl, —(C2-C7)-alkynyl, —(C6-C14)-aryl or —(C1-C8)-alkyl-(C6-C14)-aryl, wherein —(C6-C14)-aryl in —(C1-C8)-alkyl-(C6-C14)-aryl is unsubstituted or substituted one, two or three times by halogen, and


9) —CH2—X—R, wherein X stands for O or S and R is selected from —(C1-C7)-alkyl, phenyl and benzyl, wherein phenyl and benzyl are substituted one, two or three times by —(C1-C7)-alkyl, —S(C1-C3)-alkyl, —O(C1-C3)-alkyl, —NO2, —SCF3, halogen, —C(O)(C1-C3)-alkyl and/or —CN,


or R1 forms with R2, R1 with R3 or R2 with R3, a ring comprising 3-6 C-atoms which is unsubstituted or substituted one, two or three times, independently of each other, by —OH, halogen, —NO2, —NH2, —NHP and/or -M, and the remaining moiety is selected from the moieties mentioned above under 1) to 9).


According to another preferred embodiment, diketones of general formula II




embedded image


are used, wherein


R5 stands for (CH2)n with n=0-20, —(C6-C14)-aryl or —(C5-C14)-heterocycle,


R4 and R6, independently of each other, stand for —(C1-C20)-alkyl or an ester group,


wherein R4, R5 and R6 are unsubstituted or substituted one, two or three times, independently of each other, by —(C1-C4)-alkyl, —OH, halogen, —NO2 and/or —NH2.


In the process according to the invention, it is furthermore preferred that

    • the oxidized cofactor NAD formed by the oxidoreductase/dehydrogenase is regenerated continuously, and
    • a secondary alcohol having the general formula RXRYCHOH is used for cofactor regeneration, wherein RX and RY, independently of each other, are a branched or unbranched C1-C8-alkyl group and Ctotal≧3.


The term “NAD” means nicotinamide adenine dinucleotide, the term “NADH” stands for reduced nicotinamide adenine dinucleotide.


The secondary alcohols (=cosubstrates) are converted to the corresponding ketones and NADH with the aid of the oxidoreductases used and NAD, whereby a regeneration of NADH occurs. In doing so, the portion of the cosubstrate for the regeneration ranges from 5 to 95% by volume, based on the total volume, preferably from 10 to 70%, particularly preferably from 20 to 50%.


Particularly preferably, 2-propanol or 2-butanol is used as a secondary alcohol for cofactor regeneration. Preferably, in case of 2-propanol, 10-70% (v/v), particularly preferably 20-50% (v/v), is used.


By the term “aryl”, aromatic carbon moieties comprising 6 to 24 carbon atoms within the ring/in the (anellated and completely conjugated) rings are understood. —(C6-C24)-Aryl moieties are, for instance, phenyl, naphthyl, e.g., 1-naphthyl, 2-naphthyl, biphenylyl, e.g., 2-biphenylyl, 3-biphenylyl and 4-biphenylyl, anthracyl, fluorenyl, phenantryl or cyclopentanoperhydrophenantryl. Biphenylyl moieties, naphthyl moieties and in particular phenyl moieties are preferred aryl moieties.


By the term “halogen”, an element from the family of fluorine, chlorine, bromine or iodine is understood.


By the term “—(C1-C20)-alkyl”, a hydrocarbon moiety is understood the carbon chain of which is linear-chain or branched and comprises 1 to 20 carbon atoms, for example, methyl, ethyl, propyl, isopropyl, butyl, tertiary butyl, pentyl, hexyl, heptyl, octyl, nonyl, decanyl etc.


By the term “—(C3-C7)-cycloalkyl”, cyclic hydrocarbon moieties such as cyclopropyl, cyclobutyl, cyclopentyl, cyclohexyl or cycloheptyl are understood.


The term “—(C5-C14)-heterocycle” stands for a monocyclic or bicyclic 5-membered to 14-membered heterocyclic ring which is partially or completely saturated. N, O and S are examples of heteroatoms. Moieties derived from pyrrole, furan, thiophene, imidazole, pyrazole, oxazole, isoxazole, thiazole, isothiazole, tetrazole, 1,2,3,5-oxathiadiazole-2-oxide, triazolone, oxadiazolone, isoxazolone, oxadiazolidinedione, triazole, which are substituted, e.g., by F, —CN, —CF3 or —C(O)—O—(C1-C4) alkyl, 3-hydroxypyrro-2,4-dione, 5-oxo-1,2,4-thiadiazole, pyridine, pyrazine, pyrimidine, indole, isoindole, indazole, phthalazine, quinoline, isoquinoline, quinoxaline, quinazoline, cinnoline, carboline and benz-anellated, cyclopenta-, cyclohexa- or cyclohepta-anellated derivatives of said heterocycles are examples for the term “—(C5-C14)-heterocycle”. The moieties 2- or 3-pyrrolyl, phenyl pyrrolyl such as 4- or 5-phenyl-2-pyrrolyl, 2-furyl, 2-thienyl, 4-imidazolyl, methylimidazolyl, e.g., 1-methyl-2-, -4- or -5-imidazolyl, 1,3-thiazole-2-yl, 2-pyridyl, 3-pyridyl, 4-pyridyl, 2-, 3- or 4-pyridyl-N-oxide, 2-pyrazinyl, 2-, 4- or 5-pyrimidinyl, 2-, 3- or 5-indolyl, substituted 2-indolyl, e.g., 1-methyl-, 5-methyl-, 5-methoxy-, 5-benzyloxy-, 5-chloro- or 4,5-dimethyl-2-indolyl, 1-benzyl-2- or -3-indolyl, 4,5,6,7-tetrahydro-2-indolyl, cyclohepta[b]-5-pyrrolyl, 2-, 3- or 4-quinolyl, 1-, 3- or 4-isoquinolyl, 1-oxo-1,2-dihydro-3-isoquinolyl, 2-quinoxalinyl, 2-benzofuranyl, 2-benzothienyl, 2-benzoxazolyl or benzothiazolyl or dihydropyridinyl, pyrrolidinyl, e.g., 2- or 3-(N-methylpyrrolidinyl), piperazinyl, morpholinyl, thiomorpholinyl, tetrahydrothienyl or benzodioxolanyl are particularly preferred.


Preferred compounds of Formula I are, for example, ethyl-4-chloroacetoacetate, methyl acetoacetate, ethyl-8-chloro-6-oxooctanoic acid, ethyl-3-oxovalerate, 4-hydroxy-2-butanone, ethyl-2-oxovalerate, ethyl-2-oxo-4-phenylbutyric acid, ethyl pyruvate, ethyl phenyl glyoxylate, 1-phenyl-2-propanone, 2-chloro-1-(3-chlorophenyl)ethane-1-one, acetophenone, 2-octanone, 3-octanone, 2-butanone, 1-[3,5-bis(trifluoromethyl)phenyl]ethane-1-one, 1,4-dichloro-2-butanone, acetoxyacetone, phenacyl chloride, ethyl-4-bromoacetoacetate, 1,1-dichloroacetone, 1,1,3-trichloroacetone or 1-chloroacetone.


Aromatic or heteroaromatic compounds of general formula R—CO—CH2-X are particularly preferred compounds, wherein X=halogen and R represents a moiety according to the above definition of Formula I.


Further preferred compounds are diketones such as 2,5-hexanedione, 2,4-pentanedione, 2,7-octanedione, 2,7-dimethyl-3,6-octanedione or diacetyl.


In the process according to the invention, the oxidoreductases can be used either in a completely purified or in a partially purified state or the process can be performed with cells containing the described oxidoreductases. In doing so, the cells used can be provided in a native, permeabilized or lysed state. Preferably, the cloned oxidoreductases according to SEQ ID NO:1, SEQ ID NO:2, SEQ ID NO:3, SEQ ID NO:4, SEQ ID NO:5, SEQ ID NO:6, SEQ ID NO:7 and homologues thereof, respectively, are used.


5.000 to 10 Mio U of oxidoreductase are used per kg of keto compound to be converted (no upper limit) The enzyme unit 1 U corresponds to the enzyme amount which is required for converting 1 mmol of the keto compound per minute (min)


In the process according to the invention, the keto compound is used in an amount ranging from 3% to 50%, based on the total volume, preferably ranging from 5% to 40%, in particular from 10% to 30%.


The TTN (total turn over number=mol of reduced keto compound/mol of cofactor used) achieved in the processes according to the invention normally ranges from 102 to 105, preferably, however, it is ≧103.


For the regeneration of the cofactor, an alcohol dehydrogenase can additionally be added, preferably, however, the process is carried out only with an oxidoreductase (substrate-coupled coenzyme regeneration).


The aqueous portion of the reaction mixture in which the enzymatic reduction proceeds preferably contains a buffer, e.g., a potassium phosphate, tris/HCl or triethanolamine buffer, having a pH value of from 5 to 10, preferably a pH value of from 6 to 9. In addition, the buffer can contain ions for stabilizing or activating the enzymes, for example, zinc ions or magnesium ions.


While the process according to the invention is being carried out, the temperature suitably ranges from about 10° C. to 70° C., preferably from 20° C. to 45° C.


In a further possible embodiment of the process according to the invention, the enzymatic conversion is carried out in the presence of organic solvents which are not miscible with water or are miscible with water only to a limited degree and which are not involved in the cofactor regeneration, i.e., do not contain any oxidizable hydroxy groups. Said solvent can, for example, be a symmetric or unsymmetric di(C1-C6)alkyl ether, a linear-chain or branched alkane or cycloalkane. Preferably, diethyl ether, tertiary butyl methyl ether, diisopropyl ether, dibutyl ether, ethyl acetate, butyl acetate, heptane, hexane, toluene, dichloromethane, cyclohexane or mixtures thereof is/are used as additional organic solvents.


The concentration of the cofactor NAD(H) in the aqueous phase generally ranges from 0.001 mM to 1 mM, in particular from 0.01 mM to 0.1 mM.


In the process according to the invention, a stabilizer of oxidoreductase/dehydrogenase can also be used. Suitable stabilizers are, for example, glycerol, sorbitol, 1,4-DL-dithiothreitol (DTT) or dimethyl sulfoxide (DMSO).


The enzymatic reduction itself proceeds under mild conditions so that the alcohols produced do not react any further. The process according to the invention has a long service life, an enantiomeric purity of normally more than 95% of the chiral alcohols produced and a high yield, based on the amount of keto compounds used.


The process according to the invention is carried out, for example, in a closed reaction vessel made of glass or metal. For this purpose, the components are transferred individually into the reaction vessel and stirred under an atmosphere of, e.g., nitrogen or air. The reaction time ranges from 1 hour to 48 hours, in particular from 2 hours to 24 hours.


Subsequently, the reaction mixture is reprocessed. For this purpose, the aqueous phase is separated, the organic phase is filtered. Optionally, the aqueous phase may be extracted once again and reprocessed further like the organic phase. Thereupon, the solvent is optionally evaporated from the filtered organic phase.


By way of the following examples, the invention is illustrated in further detail and the exceptional quality of the identified oxidoreductases is demonstrated in comparison to the prior art.







EXAMPLE 1
Cloning and Expression of Oxidoreductase SEQ ID NO:5

A) Cultivation of Ralstonia eutropha JMP134


Cells of Ralstonia eutropha DSM 4048 were cultivated in the following medium (pH 7.0) at 30° C. in a bacterial incubator: 0.5% peptone, 0.3% meat extract. On day 3 of the cultivation, cells were separated from the culture medium by centrifugation and stored at −80° C.


B) Amplification of the Gene Encoding for Selective Oxidoreductase

Genomic DNA was extracted according to the method described in “Molecular Cloning” by Manniatis & Sambrook (see above). The resulting nucleic acid served as a template for the polymerase chain reaction (PCR) involving specific primers which were derived from the gene sequence published under number 53760931 in the NCBI database. In doing so, the primers were provided in a 5′-terminal position with restriction sites for the endonucleases Nde I and Xho I, for subsequent cloning into an expression vector.


Amplification was carried out in a PCR buffer [10 mM tris-HCl, (pH 8.0); 50 mM KCl; 10 mM MgSO4; 1 mM dNTP Mix; 20 pMol of each primer and 2.5 U of Platinum Pfx DNA polymerase (Invitrogen)] with 500 ng of genomic DNA and the following temperature cycles:

  • Cycle 1:
    • 94° C., 2 min
  • Cycle 2×30:
    • 94° C., 30 sec
    • 55° C., 30 sec
    • 68° C., 60 sec
  • Cycle 3:
    • 68° C., 7 min
    • 4° C., ∞


The resulting PCR product having a size of about 750 bp was restricted after purification via a 1% agarose gel with the aid of the endonucleases Nde I and Xho I and was ligated into the backbone of the pET21a vector (Novagen), which backbone had been treated with the same endonucleases. After transforming 2 μl of the ligation batch into E. coli Top 10 F′ cells (Invitrogen), plasmid DNAs of ampicillin-resistant colonies were tested for the presence of an insert having a size of 750 bp by using a restriction analysis with the endonucleases Nde I and Xho I. Plasmid preparations from the clones which were positive for the fragment were subjected to a sequence analysis and subsequently transformed into Escherichia coli BL21 Star (Invitrogen).


EXAMPLE 2
Expression of the Recombinant Oxidoreductases in E. coli

The Escherichia coli strains BL21 Star (Invitrogen, Karlsruhe, Germany) and RB791 (E. coli genetic stock, Yale, USA), respectively, which had been transformed with the expression construct, were cultivated in 200 ml LB medium (1% tryptone, 0.5% yeast extract, 1% NaCl) with ampicillin (50 μg/ml) or carbenicillin (50 μg/ml), respectively, until an optical density of 0.5, measured at 550 nm, was reached. The expression of recombinant protein was induced by adding isopropylthiogalactoside (IPTG) at a concentration of 0.1 mM. After 8 and 16 hours of induction, respectively, at 25° C. and 220 rpm, the cells were harvested and frozen at −20° C. For the activity test, 10 mg of cells were mixed with 500 μl of 100 mM TEA buffer pH 7.0 and 500 μl of glass beads and were digested for 10 mM using a globe mill. The lysate obtained was then used for the respective measurements in a diluted state. The activity test was composed as follows: 870 μl of 100 mM TEA buffer pH 7.0, 160 μg NAD(P)H, 10 μl diluted cell lysate. The reaction was started by adding 100 μl of a 100 mM substrate solution to the reaction mixture.


The oxidoreductases of the present invention could be expressed very efficiently in Escherichia coli. Table 3 shows the activities of the individual enzymes which were achieved in the expression.














TABLE 3







Expression
Expression

Activity



vector
strain
Reference
U/g




















SEQ ID NO: 1
pET21a
BL21 Star
2-octanone
2650 U/g


SEQ ID NO: 2
pET21a
BL21 Star
methyl
1570 U/g





acetoacetate


SEQ ID NO: 3
pQE70
BL21 Star
CLAEE
3130 U/g


SEQ ID NO: 4
pET21a
BL21 Star
ethyl-2-oxo-4-
 400 U/g





phenyl





butanoate


SEQ ID NO: 5
pET21a
BL21 Star
CLAEE
2890 U/g


SEQ ID NO: 6
pQE70
BL21 Star
methyl
 840 U/g





acetoacetate


SEQ ID NO: 7
pET21a
BL21 Star
methyl
3295 U/g





acetoacetate









EXAMPLE 3
Characterization of the Recombinant Oxidoreductases SEQ ID NO:1-SEQ ID NO:7

3a: pH-Optimum


The buffers listed in Table 4 were produced. The concentration of the respective buffer components was in each case 50 mM.










TABLE 4





pH-value
Buffer system
















4
Na-acetate/acetic acid


4.5
Na-acetate/acetic acid


5
Na-acetate/acetic acid


5.5
KH2PO4/K2PO4


6
KH2PO4/K2PO4


6.5
KH2PO4/K2PO4


7
KH2PO4/K2PO4


7.5
KH2PO4/K2PO4


8
KH2PO4/K2PO4


8.5
KH2PO4/K2PO4


9
glycine/NaOH


9.5
glycine/NaOH


10
glycine/NaOH


11
glycine/NaOH









Measuring Batch (30° C.)-pH Optimum Reduction:

870 μl of each of the buffer systems mentioned in Table 3


20 μl of NADH 10 mM


10 μl of a diluted enzyme


After about 2 to 3 min of incubation,


100 μl of a substrate solution (100 mM) were added.


The respective reference substrate (Table 3) was used as a substrate for each oxidoreductase. The reaction was pursued for 1 min at 340 nm In order to determine the pH-optimum, the enzymatic reaction in the respective buffer listed in Table 4 was determined In order to determine the pH-optimum for the oxidation reaction, NAD was used as the cofactor and 2-propanol or 2-octanol was used as the substrate.


The results for the oxidoreductases SEQ ID NO:1-SEQ ID NO:7 are compiled in Table 5.













TABLE 5







SEQ ID NO
pH-Opt. Red.
pH-Opt. Ox.









SEQ ID NO: 1
5.5-6.5
9.0-9.5



SEQ ID NO: 2
5.0-5.5
 9-10



SEQ ID NO: 3
5.5
10-11



SEQ ID NO: 4
5.5
 9-10



SEQ ID NO: 5
7.0-7.5
9.5-10 



SEQ ID NO: 6
5.5-6.5
7.5



SEQ ID NO: 7
5.5
10-11










3b: pH Stability

The stability of the recombinant oxidoreductases was examined by storing them in the buffer systems mentioned in Table 4. For this purpose, the different buffers (50 mM) were prepared in the range from pH 4 to 11, and the oxidoreductases produced according to Example 4 were diluted therewith. After 30, 60 and 120 minutes of incubation, 10 μl were taken from the batch and used in the activity test according to Example 3a.


The initial value was thereby the measured value which was obtained immediately after the dilution (1:20) of the enzyme in a potassium phosphate buffer 50 mM pH=7.0. Under the given conditions, said value corresponded to an extinction change of approx. 0.70/min and was set as a 100% value. All subsequent measured values were put in relation to this value.


In Table 6, the pH ranges in which the enzymes exhibited no less than 50% of the initial activity with an incubation of 120 mM are compiled for the named oxidoreductases.












TABLE 6







SEQ ID NO
pH-range stability









SEQ ID NO :1
 4.5-9.5



SEQ ID NO: 2
7.0-10



SEQ ID NO: 3
5.5-11



SEQ ID NO: 4
 5.5-9.5



SEQ ID NO: 5
7.5-11



SEQ ID NO: 6
  6-11



SEQ ID NO: 7
  5-11










3c: Temperature Optimum

In order to determine the optimum test temperature, the enzyme activity for the oxidoreductases used according to the invention was measured in the standard measuring batch in a temperature range from 15° C. to 70° C.


The temperature optima determined are compiled in Table 7:












TABLE 7







SEQ ID NO
Topt









SEQ ID NO: 1
55° C.



SEQ ID NO: 2
65° C.



SEQ ID NO: 3
n.d



SEQ ID NO: 4
n.d



SEQ ID NO: 5
45° C.



SEQ ID NO: 6
50° C.



SEQ ID NO: 7
45° C.










3d: Temperature Stability

In an analogous manner as described under Example 3c, the temperature stability was determined for the range from 15° C. to 70° C. For this purpose, a dilution of the oxidoreductases used according to the invention was in each case incubated at the respective temperature for 60 min and 180 min and was subsequently measured at 30° C. with the above-mentioned test batch. In Table 8, the temperature ranges in which the enzymes exhibited no less than 50% of the initial activity with an incubation of 120 min are compiled for the oxidoreductases.












TABLE 8







SEQ ID NO
Temperature range









SEQ ID NO: 1
15-45° C.



SEQ ID NO: 2
15-35° C.



SEQ ID NO: 3
n.d



SEQ ID NO: 4
n.d



SEQ ID NO: 5
15-35° C.



SEQ ID NO: 6
15-35° C.



SEQ ID NO: 7
15-45° C.











3e: Stability Against 2-propanol


The stability of the oxidoreductases to be examined against 2-propanol was tested by diluting the lysates obtained in Example 2 (from a recombinant expression) in a buffer (KPP 100 mM, pH=7.0) containing the corresponding percentage of 2-propanol (approx. 10 units/ml buffer). The batch was incubated at room temperature with constant thorough mixing (thermomixer at 170 rpm). After 24 h of incubation, 10 μl each were taken from the aqueous phase and used for the determination of the enzyme activity in the standard test batch (potassium phosphate buffer (KPP) 100 mM, pH=7.0, 0.2 mM NADH, 10 mM substrate). Also in this case, the initial value immediately after the dilution in the buffer was set to 100%, and all further values were put in relation thereto.









TABLE 9







Enzyme activity after 24 h of incubation


with mixtures containing 2-propanol














10% 2-
20% 2-
30% 2-
40% 2-


SEQ ID NO
Buffer
Propanol
Propanol
Propanol
Propanol
















SEQ ID NO: 1
100%
  100%
  100%
  100%
30%



SEQ ID NO: 2
100%
  100%
  100%
  100%
30%


SEQ ID NO: 3
100%
80-100%
50-75%
25-30%
0%


SEQ ID NO: 4
100%
80-100%
50-75%
   0%
0%


SEQ ID NO: 5
100%
  100%
  100%
  100%
0%


SEQ ID NO: 6
100%
80-100%
50-75%
50-75%
50-75%


SEQ ID NO: 7
100%
80-100%
50-75%
25-30%
25-30%









As can be seen from Table 9, all examined sequences exhibit an astounding stability in 2-propanol, i.e., all oxidoreductases are stable for at least 24 h in 20% isopropanol, but most enzymes still display substantial residual activities of up to 75% even after 24 h at 40% (v/v).


This shows that the sequences SEQ ID NO:1-SEQ ID NO:7 are particularly suitable for use in a substrate-coupled process with 2-propanol.


3f: Comparison of the Substrate Spectra of Oxidoreductases SEQ ID NO:1-SEQ ID NO:7

The substrate spectrum of the sequences to be characterized was determined by measuring the enzyme activity for reduction and oxidation with a number of ketones and alcohols. For this purpose, the standard measuring batch according to Example 2 was used with different substrates.


The activity with methyl acetoacetate was set to 100% for all enzymes, and all other substrates were put in relation thereto.









TABLE 10







Substrate spectra/reduction
















SEQ ID
SEQ ID
SEQ ID
SEQ ID
SEQ ID
SEQ ID


Substrate
SEQ ID NO: 1
NO: 2
NO: 3
NO: 4
NO: 5
NO: 6
NO: 7





1-Phenyl-2-
 15%
 7%
70%
37%
<1%
<1%
16%


propanone


Phenacyl
3.4%
 1%
50%
<1%
<1%
15%
10%


chloride


Acetophenone
3.5%
 6%
30%
<1%
<1%
3.5% 
32%


Acetonaphthone
  3%
 9%
30%
 7%
<1%
25%
17%


Butyrophenone
  3%
 1%
 <%
<1%
<1%
<1%
<1%


2-Octanone
 92%
 2%
33%
70%
26%
23%
49%


3-Octanone
 22%
 2%
15%
 5%
13%
 2%
15%


2-Butanone
8.5%
4.5% 
30%
 5%
<1%
 7%
 6%


Ethyl-2-
  3%
60%
70%
17%
<1%
15%
<1%


oxovalerate


Ethyl-2-oxo-
  3%
50%
118% 
35%
45%
18%
10%


4-phenyl


butyric acid


Ethyl pyruvate
140% 
70%
100% 
64%
13%
116% 
160% 


Ethyl phenyl
4.8%
53%
7.5% 
 3%
<1%
<1%
<1%


glyoxylate


Ethyl-4-
12.5% 
20%
21%
100% 
50%
10%
<1%


chloro-


acetoacetate


Methyl
100% 
100% 
100% 
100% 
100% 
100% 
100% 


acetoacetate


Ethyl-3-
3.5%
30%
76%
11%
<1%
10%
12%


oxovalerate


Acetone
4.5%
16%
38%
 5%
<1%
3.5% 
 8%









The enantiomeric excesses were determined for selected substrates. The following reaction batch was used for this:

    • 160 μl buffer
    • 100 μl NAD (0.4 mg/ml)
    • 20-50 μl 2-propanol
    • 50 μl enzyme solution
    • 2 mg substrate
    • Reaction conditions: 28° C., 24 h


Upon completion of the incubation period, the samples were extracted with solvents depending on the substrate and were centrifuged. One aliquot was withdrawn, optionally, the extractant was removed completely, subsequently, the sample was dissolved in a suitable solvent, and the enantiomeric excess was determined by GC.









TABLE 11







Enantioselectivity for selected substrates















SEQ ID
SEQ ID
SEQ ID
SEQ ID
SEQ ID
SEQ ID
SEQ ID


Substrate
NO: 1
NO: 2
NO: 3
NO: 4
NO: 5
NO: 6
NO: 7





Phenacyl
≧99.9% S
≧99.9% S
≧99.9% S
97.5% S
60-80% S
≧99.9% S
≧99.9% S


chloride


Acetophenone
   99.0% R
   98.5% R
   99.0% R


   99.0% R
   98.5% R


2-Pentanone
≧99.9% R
     98% R
     98% R
     95% R
   99.0% R
     97% R
   99.5% R


2-Octanone
≧99.9% R
≧99.9% R
≧99.9% R
≧99.9% R
≧99.9% R
≧99.9% R
≧99.9% R


2-Butanone
≧96.0% R
n.d
n.d
n.d
n.d
rac
     50% R


Ethyl
   98.0% R
≧99.9% R
≧99.9% R
≧99.9% R
≧99.9% R
≧99.9% R
≧99.9% R


pyruvate


Ethyl-4-
   99.0% S
≧99.9% S
≧99.9% S
≧99.9% S
≧99.9% S
≧99.9% S
≧99.9% S


chloro-


acetoacetate


Methyl
≧99.9% R
≧99.9% R
≧99.9% R
98.5% R
≧99.9% R
≧99.9% R
≧99.9% R


acetoacetate





“rac”: unselective reduction, i.e., both enantiomers form at a roughly identical ratio


n.d = not determined, no conversion


The enantiomeric excess is calculated as follows: ee(%) = ((R-alcohol − S-alcohol)/(R-alcohol + S-alcohol)) × 100.






The potential of the above-described oxidoreductases toward oxidation and hence the suitability for regeneration were compared by employing the following activity test: 870 μl of 100 mM TEA buffer, pH 8.0, 160 μg NAD, 10 μl diluted cell lysate. The reaction was started by adding 100 μl of a 100 mM substrate solution to the reaction mixture.


In this case, the substrate with the highest activity was set to 100%.









TABLE 12







Substrate spectra/oxidation
















SEQ
SEQ
SEQ


SEQ




ID
ID
ID


ID



SEQ ID
NO:
NO:
NO:
SEQ ID
SEQ ID
NO:


Substrate
NO: 1
2
3
4
NO: 5
NO: 6*
7

















S-2-Octanol
 0%
0%
0%
0%
 0%
0%
0%


R-2-Octanol
100% 
31%
100%
100%
100% 
100%
100%


S-2-Butanol
 1%
20%
30%
2%
2.5% 
<2%
0%


R-2-Butanol
 3%
100%
58%
18%
14%
60%
50%


S-Phenyl-2-
 0%
0%
0%
0%
 0%
0%
0%


propanol


R-Phenyl-2-
<1%
20%
50%
2%
25%
0%
50%


propanol


2-Methyl-4-
<1%
20%
10%
1%
2.5% 
0%
0%


pentanol


2-Propanol
1.6% 
20%
100%
7%
 7%
60%
20%


Cyclohexanol
 0%
20%
0%
0%
 0%
0%
0%





*Oxidation hardly detectable






As can be seen from Tables 10-12, all seven sequences are clearly R-specific oxidoreductases. However, the examined enzymes differ clearly in the spectrum of preferred substrates as well as in the oxidation behaviour. Differences arise with regard to the preference for short-chain, rather low-molecular substrates over long-chain carbonyl compounds, similarly, different aromatic systems or also different substituents (such as halogen) are tolerated to a different degree.


Although a variety of enzymes are thus available which, in principle, indeed catalyze similar reactions, it was found out that both enzymes which catalyze the target reaction extremely efficiently and enzymes which do not catalyze the target reaction at all can be found among those individual substrates.


Furthermore, differences in enantioselectivity are also observed which indeed appear minor at first sight, but can be absolutely decisive for the applicability of individual enzymes for specific reduction processes given a specification of ee>99.0-99.9%, which today is common for pharmaceutical applications.


Also, the oxidative properties and hence the applicability in different processes with a substrate-coupled coenzyme regeneration vary between the examined enzymes as well as in comparison to the prior art.


Thus, the use of the named oxidoreductases in the process according to the invention does not represent an equivalent alternative to the established prior art, but a broadening of the existing spectrum.


EXAMPLE 4
Comparison of Oxidoreductases SEQ ID NO:1-SEQ ID NO:7 in Specific Processes

On the basis of the examples listed below, it shall now be shown for a few specific processes that the enzymes SEQ ID NO:1 to 7 can provide special solutions to concrete questions which also clearly exceed the existing prior art.


4.a Reduction of 3,6-octanedione to R,R-3,6-octanediol


For the purpose of screening for suitable enzymes for the development of a process with a substrate-coupled coenzyme regeneration for reducing 3,6-octanedione to R,R-3,6-octanediol, all enzymes were examined in the following reaction batch:


450 μl buffer, pH=7.0


0.05 mg NAD

50 μl 2-propanol


10 mg 3,6-octanedione


After 24 h of incubation at 25° C., the samples were extracted by means of dichloromethane and analyzed by GC.


In Table 13, the conversions and reaction products obtained are indicated.














TABLE 13






3,6-
Product (S)
Product (R)
S,S-3,6-
R,R-3,6-


SEQ ID NO
Octanedione
reduced once
reduced once
Octanediol
Octanediol





















SEQ ID NO: 1
95%
0%
 5%
0%
0%



SEQ ID NO: 2
 3%
0%
17%
0%
80%


SEQ ID NO: 3
10%
0%
45%
0%
45%


SEQ ID NO: 4
76%
0%
24%
0%
0%


SEQ ID NO: 5
95%
0%
 5%
0%
0%


SEQ ID NO: 6
24%
0%
59%
0%
17%


SEQ ID NO: 7
43%
0  
47%
0  
10%


WO 2007/012428
85%
0%
15%
0%
0%



Pichia farinosa



WO 2007/012428
95%
0%
 5%
0%
0%



Candida nemodendra



EP 1179 595 A1
 3%
0%
17%
0%
80%



Pichia finlandica



WO 2004/027055
40%
0%
50%
0%
10%



Devosia










As can be seen from Table 13, among the available enzymes, such can be found which, under the above-mentioned reaction conditions, are very well capable of reducing 3,6-octanedione to the desired R,R-octanediol (SEQ ID NO:2, SEQ ID NO:3 and P. finlandica), such which practically do not accept the substrate at all (SEQ ID NO:1, SEQ ID NO:5 and C. nemodendra) as well as such which would have to be used preferably if it was intended to produce the compound reduced once (SEQ ID NO:4).


4.b Reduction of 2,7-dimethyl-3,6-octanedione to (S,S)-2,7-dimethyl octane-3,6-diol


For the purpose of screening for suitable enzymes for a process with a substrate-coupled coenzyme regeneration for reducing 2,7-dimethyl-3,6-octanedione to (S,S)-2,7-dimethyl octane-3,6-diol, the enzymes were examined in the following reaction batch:


450 μl buffer, pH=7.0


0.05 mg NAD

50 μl 2-propanol


10 mg 2,7-dimethyl-3,6-octanedione


After 24 h of incubation at 25° C., the samples were extracted by means of dichloromethane and analyzed by GC.


In Table 14, the conversions and reaction products obtained are indicated.













TABLE 14






2,7-Dimethyl-3,6-

(S,S)-2,7-Dimethyl
(R,R)-2,7-Dimethyl-


SEQ ID NO
octanedione
Meso
octane-3,6-diol
3,6-octanediol







SEQ ID NO: 1
100%
0%
0%
0%


SEQ ID NO: 2
 23%
0  
77% 
0%


SEQ ID NO: 3
100%
0%
0%
0%


SEQ ID NO: 4
100%
0%
0%
0%


SEQ ID NO: 5
100%
0%
0%
0%


SEQ ID NO: 6
100%
0%
0%
0%


SEQ ID NO: 7
100%
0%
0%
0%


WO 2007/012428
100%
0%
0%
0%



Pichia farinosa



WO 2007/012428
100%
0%
0%
0%



Candida nemodendra



EP 1179 595 A1
100%
0%
0%
0%



Pichia finlandica



WO 2004/027055
100%
0%
0%
0%



Devosia










As can be seen from Table 14, SEQ ID NO:2 provides a solution to this specific question, by contrast, no enzyme from the prior art is capable of catalyzing the desired reaction.


4.c Reduction of 2,4-pentanedione to R,R-2,4-pentanediol


For the purpose of screening for suitable enzymes for a process with a substrate-coupled coenzyme regeneration for reducing 2,4-pentanedione to R,R-2,4-pentanediol, the enzymes were examined in the following reaction batch:


450 μl buffer, pH=7.0


0.05 mg NAD

50 μl 2-propanol


10 mg 2,4-pentanedione


After 24 h of incubation at 25° C., the samples were extracted by means of dichloromethane and analyzed by GC.


In Table 15, the conversions and reaction products obtained are indicated.













TABLE 15






2,4-
Compound
(S,S)-2,4-
(R,R)-2,4-



Pentane-
(R) re-
Pentane-
Pentane-


SEQ ID NO
dione
duced once
diol
diol




















SEQ ID NO: 1
0%
70%
0%
30%



SEQ ID NO: 2
0%
19%
0%
81%


SEQ ID NO: 3
2%
95%
0%
3%


SEQ ID NO: 4
6%
93%
0%
1%


SEQ ID NO: 5
60% 
40%
0%
0%


SEQ ID NO: 6
100% 
 0%
0%
0%


SEQ ID NO: 7
7%
91%
0%
2%


WO 2007/012428
100% 
 0%
0%
0%



Pichia farinosa



WO 2007/012428
55% 
45%
0%
0%



Candida nemodendra



EP 1179 595 A1
30% 
70%
0%
0%



Pichia finlandica



WO 2004/027055
43% 
57%
0%
0%



Devosia










As can be seen from Table 15, the conversion of 2,4-pentanedione to R,R-2,4-pentanediol is possible with the oxidoreductases SEQ ID NO:1 and SEQ ID NO:2. Most enzymes convert this substrate only as far as to a mono-reduced compound. The prior art provides no solution to this, either.


On the other hand, for an efficient recovery of the mono-reduced compound, the oxidoreductases SEQ ID NO:3, SEQ ID NO:4 and SEQ ID NO:7 are even more recommendable than the prior art.


4.e Reduction of 1,1,1-trifluoroacetone to R-1,1,1-trifluoropropanol


For the purpose of screening for suitable enzymes for a process with a substrate-coupled coenzyme regeneration for reducing 1,1,1-trifluoroacetone to R-1,1,1-trifluoropropanol, the enzymes were examined in the reaction batch indicated below.


In doing so, the reaction batch accounts for the specific demand made on the reprocessing and conversion of this highly volatile substrate. In this case, a regeneration with isopropanol is not possible, in fact, the use of methyl pentanol is preferred in order to, on the one hand, prevent the volatile substrate (bp=22° C.) from escaping during the reaction and, on the other hand, enable the reprocessing of the reaction product R-1,1,1-trifluoropropanol (bp=80° C.). (The separation of 2-propanol would be impossible because of identical boiling points.)


450 μl buffer, pH=7.0


0.05 mg NAD

500 μl 2-methyl-4-pentanol


20 μl 1,1,1-trifluoroacetone


After 24 h of incubation at 25° C., the samples were analyzed by GC.


In Table 16, the conversions and reaction products obtained are indicated.












TABLE 16






1,1,1-





Trifluoro-
R-1,1,1-
S-1,1,1-


SEQ ID NO
acetone
Trifluoropropanol
Trifluoropropanol







SEQ ID NO: 1
 0%
99%
 1%


SEQ ID NO: 2
20%
60%
40%


SEQ ID NO: 3
 2%
70%
30%


SEQ ID NO: 4
15%
80%
20%


SEQ ID NO: 5
50%
85%
15%


SEQ ID NO: 6
n.d
n.d
n.d


SEQ ID NO: 7
n.d
n.d
n.d


WO 2007/012428
 0%
97%
 3%



Pichia farinosa



WO 2007/012428
50%
99%
 1%



Candida nemodendra



EP 1179 595 A1
50%
99%
 1%



Pichia finlandica



WO 2004/027055
n.d
n.d
n.d



Devosia










As can be seen from Table 16, in this case, oxidoreductase SEQ ID NO:1 offers the best compromise between conversion and selectivity. Amazingly, most enzymes are not capable at all of reducing this substrate with a satisfactory enantioselectivity, whereby the diversity of the described enzymes despite a basically identical functionality is again emphasized.


4.f Reduction of 2-chloro-1-(3-hydroxyphenyl)ethane-1-one to (1S)-2-chloro-1-(3-hydroxyphenyl)ethane-1-ol


For the purpose of screening for suitable enzymes for a process with a substrate-coupled coenzyme regeneration for reducing 2-chloro-1-(3-hydroxyphenyl)ethane-1-one to (1S)-2-chloro-1-(3-hydroxyphenyl)ethane-1-ol, the enzymes were examined in the following reaction batch:


400 μl buffer, pH=7.0


0.02 mg NAD

100 μl 2-propanol


25 mg 2-chloro-1-(3-hydroxyphenyl)ethane-1-one


After 24 h of incubation at 25° C., the samples were extracted and analyzed by GC.


In Table 17, the conversions and reaction products obtained are indicated.












TABLE 17






2-Chloro-1-(3-
(1R)-2-Chloro-1-(3-
(1S)-2-Chloro-1-(3-


SEQ ID NO
hydroxyphenyl)ethane1-one
hydroxyphenyl)ethane-1-ol
hydroxyphenyl)ethane-1-ol







SEQ ID NO: 1
100% 
n.d
n.d


SEQ ID NO: 2
55%
0%
45%


SEQ ID NO: 3
40%
0%
60%


SEQ ID NO: 4
99%
n.d
n.d


SEQ ID NO: 5
100% 
n.d
n.d


SEQ ID NO: 6
25%
0%
75%


SEQ ID NO: 7
83%
0%
17%


WO 2007/012428
100% 
n.d
n.d



Pichia farinosa



WO 2007/012428
100% 
n.d
n.d



Candida nemodendra



EP 1179 595 A1
98%
n.d
n.d



Pichia finlandica



WO 2004/027055
100% 
n.d
n.d



Devosia










As can be seen from Table 17, the conversion of 2-chloro-1-(3-hydroxyphenyl)ethane-1-one to (1S)-2-chloro-1-(3-hydroxyphenyl)ethane-1-ol is basically possible with the oxidoreductases SEQ ID NO:2, SEQ ID NO:3, SEQ ID NO:6 and SEQ ID NO:7, with those enzymes converting the substrate absolutely selectively. No enzyme from the prior art is capable of reducing this specific substrate.


4.g Reduction of ethyl-4-chloroacetoacetate to (S)-ethyl-4-chloro-3-hydroxybutyrate


In the following, the enzymes SEQ ID NO:1-7 to be examined are tested in the industrial process for the reduction of ethyl-4-chloroacetoacetate to (S)-ethyl-4-chloro-3-hydroxybutyrate and are compared to the prior art.


For this purpose, the substrate ethyl-4-chloroacetoacetate was used at a high concentration (200 g/l) according to the prior art (WO 2006/087235) and the efficiency of the enzymes in the process with a substrate-coupled regeneration with 2-propanol was compared.


600 μl buffer, pH=8.0


0.2 mg NAD

200 μl 2-propanol


200 μl ethyl-4-chloroacetoacetate


240 μl enzyme recombinant from E. coli (20-60 units)


After 24 h of incubation at 25° C., the samples were extracted and analyzed by GC.


In Table 18, the conversions and reaction products obtained are indicated.












TABLE 18






Ethyl-4-
(S)-Ethyl-4-chloro-3-
(R)-Ethyl-4-chloro-3-


SEQ ID NO
chloroacetoacetate
hydroxybutyrate
hydroxybutyrate







SEQ ID NO: 1
90%
10%
0%


SEQ ID NO: 2
 0%
100% 
0%


SEQ ID NO: 3
 0%
100% 
0%


SEQ ID NO: 4
70%
30%
0%


SEQ ID NO: 5
85%
15%
0%


SEQ ID NO: 6
60%
40%
0%


SEQ ID NO: 7
75%
25%
0%


WO 2007/012428
100% 
n.d
n.d



Pichia farinosa



WO 2007/012428
100% 
n.d
n.d



Candida nemodendra



EP 1179 595 A1
80%
20%
0%



Pichia finlandica



WO 2004/027055
75%
25%
0%



Devosia










As can be seen from Table 18, in particular the enzymes SEQ ID NO:2 and SEQ ID NO:3 provide NADH-dependent alternatives to the existing, usually NADPH-dependent processes, which alternatives are interesting for the reduction of ethyl-4-chloroacetoacetate to (S)-ethyl-4-chloro-3-hydroxybutyrate. In contrast, the enzymes from the prior art are unsuitable since they display only insufficient conversions or no conversions at all, respectively, under process conditions.

Claims
  • 1. A process for stereoselective enzymatic reduction of keto compounds to corresponding chiral hydroxy compounds, wherein the keto compounds are reduced with an enantioselective, NADH-specific oxidoreductase, the process comprising: reducing the keto compounds using a polypeptide which exhibits an R-ADH-signature H-[P; A]-[I; A; Q; V; L]-[G; K]-R at positions 204-208 and has the following further structural features in their entirety:(i) an N-terminal Rossmann-Fold GxxxGxG,(ii) an NAG-motif at position 87,(iii) a catalytic triad consisting of S 139, Y 152 and K 156,(iv) a negatively charged amino acid moiety at position 37,(v) two C-terminal motifs in the dimerization domain [A; S]-S-F and [V; I]-DG-[G; A]-Y-[T; C; L]-[A; T; S]-[Q; V; R; L; P],(vi) Val or Leu at position 159 (4 positions downstream of K 156),(vii) Asn at position 178, and(viii) a proline moiety at position 188.
  • 2. A process according to claim 1, wherein a polypeptide is used for reducing the keto compounds: (a) which comprises one of the amino acid sequences SEQ ID NO:1 to SEQ ID NO:7, or(b) in which at least 70% of the amino acids are identical to the amino acids of one of the amino acid sequences SEQ ID NO:1 to SEQ ID NO:7, or(c) for which a nucleic acid sequence from the group consisting of SEQ ID NO:8 to SEQ ID NO:14 encodes, or(d) for which a nucleic acid sequence encodes which hybridizes to one of the nucleic acid sequences mentioned in (c) under stringent conditions.
  • 3. A process according to claim 1, wherein keto compounds of general formula I are used,
  • 4. A process according to claim 1, wherein diketones of general formula II
  • 5. A process according to claim 1, wherein: an oxidized cofactor NAD formed by the oxidoreductase/dehydrogenase is regenerated continuously, anda secondary alcohol having the general formula RXRYCHOH is used for cofactor regeneration, wherein RX and RY, independently of each other, are a branched or unbranched C1-C8-alkyl group and Ctotal≧3.
  • 6. A process according to claim 5, wherein an alcohol from the group consisting of 2-propanol, 2-butanol, 2-pentanol, 4-methyl-2-pentanol, 2-heptanol and 2-octanol, preferably 2-propanol, is used as a cosubstrate or a secondary alcohol, respectively.
  • 7. A process according to claim 1, wherein the keto compound is used in an amount ranging from 3% to 50% based on the total volume.
  • 8. A process according to claim 1, wherein TTN (total turn over number=mol of keto compound/mol of cofactor used) is ≧103.
  • 9. A process according to claim 1, wherein the process is carried out in an aqueous organic two-phase system.
  • 10. A process according to claim 1, wherein, in addition, an organic solvent selected from diethyl ether, tertiary butyl methyl ether, diisopropyl ether, dibutyl ether, ethyl acetate, butyl acetate, heptane, hexane and cyclohexane is used.
Priority Claims (1)
Number Date Country Kind
09450050.1 Mar 2009 EP regional
PCT Information
Filing Document Filing Date Country Kind 371c Date
PCT/EP2010/052701 3/3/2010 WO 00 10/28/2011