WO2002064579A1 - Method for producing optically active, propargylic terminal epoxides - Google Patents

Abstract

The invention relates to a method for enantioselective representation of optically active, propargylic, terminal epoxides. According to the invention, in a first step of the inventive method, compounds of formula I are reacted with an alcoholdehydrogenase which preferably originates from horse liver or thermophilic microorganisms (e.g. thermoanaerobium brockii or thermoanaerobacter ethanolicus</i>) or from Lactobacillus and is recombina ntly overexpressed in Escherichia coli , in the presence of NAD (P) H. During said reaction, the oxogroup is enantioselectively reduced. Since ( S ) selective and ( R ) selective oxido reductases can be used, the synthesis of both enantiomers is possible. Representation of both enantiomers using the same enzyme is achieved with alkoholdehydrogenase from Lactobacillus brevis , since said compounds react with R<1> = H and compounds with R<1> = H to form alcohols having an opposite configuration. In the second step of the method, the alpha -halogenated alcohol is transformed into an enantiomer pure epoxide by means of a base.

Description

 Description

Process for the preparation of optically active, propargylic, terminal epoxides

The invention relates to a process for the preparation of optically active, propargylisehen, terminal epoxides according to the preamble of claim 1.

Optically active, propargylic, terminal epoxides are important intermediates in the synthesis of natural and active substances (e.g. HMG-CoA reductase inhibitors in K. Takahashi, T. Minami, Y. Ohara, T. Hiyama, Bull. Chem Soc. 1995, 68, 2649). Optical purity is one of the most important target values in the representation of these chiral compounds.

A number of processes are already known in the prior art for the production of chiral, non-racemic propargylic, terminal epoxides.

In M. opp, T. Kanger, A. Müraus, T. Pehk, Ü. Lille, Tetrahedron: Asymmetry 1991, 2, 943 describes the synthesis of (R) -1-t-butyldimethylsilyl-3, 4-epoxy-but-l-in. It is a process that starts from (S, S) - (+) -2, 3-O-isopropylidene-L-threitol (a tartaric acid derivative), which in 6 independent and sometimes very complex synthetic steps into the Epoxy is transferred. The product has an enantiomeric excess (ee> 99% and the overall yield is 20%. As the chirality already on the first synthesis level is introduced, a total of 80% of the optical activity used is lost again in the course of the display. The synthesis is described both in the original publication listed above and in the subsequent publication T. Kanger, P. Niidas, A.-M. Müürisepp, T. Pehk, M. Lopp, Tetrahedron: Asymetry 1998, 9, 2499 only with the t-butyldimethylsilyl group as a substituent.

Disadvantages here are the high number of synthesis steps (6), the moderate overall yield (20%) and the early introduction of the optical activity, which leave much to be desired. The applicability has so far only been shown for a single example, so that no universal applicability of the method can be seen.

R. Sanchez-Obregon, B. Ortiz, F. Walls, F. Yuste, JL Garcia Ruano, Tetrahedron: Asymmetry 1999, 10, 947 describes another process in which (commercially not available) propargyl esters and (R) - (+) -Methyl p-tolyl sulfoxide are converted to β-ketosulfoxides and the resulting keto function is reduced diastereoselectively syn or anti. Finally the sulfoxide is reduced and the compound is converted into the epoxide by splitting off the sulfide group. The most serious disadvantage of this process is the low ee of the epoxides shown, since the enantiomeric excesses achieved are only 66-78%. Even with the best example given, only 88% ee could be achieved, which is far from sufficient for use in drug synthesis are correct. The overall yield is also to be assessed rather moderately at 20 - 30%, not including the representation of the propargylic esters. Furthermore, it is a process in which it is not possible to recover the chiral auxiliary group which has split off.

Only the residues phenyl, methyl and n-propyl, all of which contain no functional groups, have been tested as substituents on the triple bond.

The low optical purity (ee's of 66 - 88%) and the moderate overall yield of 20 - 30% (which does not include the synthesis of the starting compound) are the most serious disadvantages of this method. The reaction procedure also has disadvantages, since it is not possible, for example, to recover the chiral auxiliary. The applicability has only been shown to a very limited extent so that transferability to a wide range of functional groups has not been tested.

The applicant is only aware of one optically active α-chloro or bromine-substituted propargyl alcohol in the literature. In CJH Helal, PA Magriotis, EJ Corey, J. Am. Chem. Soc. 1996, 118, 10938 l-chloro-4-tri-i-propylsilyl-3-butyn-2-one is reduced to the corresponding alcohol with 0.05 equivalents of an oxazaborolidine catalyst and 1.2 equivalents of catecholborane in dichloromethane. The synthesis is not absolutely enantioselective (ee = 95%) with a yield of 97%. The sterically demanding tri-i-propylsilyl group is necessary in order to achieve the ee of 95%, since the same publication was shown that z. For example, in the case of the corresponding non-chlorinated ketone, the replacement of the tri-i-propylsilyl group with the more common trimethylsilyl group results in a marked deterioration in the enantiomer ratio from 39: 1 (ee = 95%) to 14: 1 (ee = 87%) Has. In two more of Corey et al. This trend is confirmed in the examples given.

Unfortunately, the alcohol is not enantiomerically pure and a sterically demanding substituent on the triple bond is necessary in order to achieve an ee of 95%. Furthermore, the reaction was only shown using a single example.

In CW Bradshaw, W. Hummel, C.-H. Wong, J ^" . Org. Chem. 1992, 57, 1532 and in United States Patent 5,342,767 by C.-H. Wong, CW Bradshaw Aug. 30, 1994 describe the activity of the alcohol dehydrogenase from Lactobacillus kefir in the reduction of ketones. The activity of the enzyme on the substrate 1-chloro-4-trimethylsilyl-3-butyn-2-one is given as 52% based on the standard 1-phenoxy-2-propanone, which was determined in a UV assay This is an indirect detection, in which only the decrease in NADPH is tracked, and a side reaction which may be observed is described in DD Tanner, AR Stein, J \ Org. Chem. 1988, 53, 1642 and in LM Aleixo , M. de Carvalho, PJS Moran, JAR Rodrigues, Bioorg. Med. Chem. Let t. 1993, 3, 1637. In both studies it could be shown that in the reaction of α-halogen ketones with Oxidoreductases via a radical mechanism that can form dehalogenated ketones. In this reaction, too, NAD (P) H is consumed and activity is measured in the UV assay, but without the desired alcohol being produced. The actual formation of the halogenated alcohol is thus by Bradshaw et al. not been proven. Accordingly, the optical purity could not be determined or the product could be characterized in terms of other stabilities, etc.

In C. Heiss, R. S. Phillips, J. Chem. Soc. , Perkin Trans. 1 2000, 2821 reported for the first time a formal configuration reversal in the reduction of propargyl ketones with isolated oxidoreductases. The alcohol dehydrogenase from Thermoanaerobacter ethanolicus TJS-ADH reduces ketones of the general structure:

The compounds with the residues R = methyl (ee of the alcohol formed = 60%), ethyl (ee = 80%), isopropyl (ee> 98%), t-butyl (ee = 85%) and propyl (ee = 51% ) are converted to products with (S) configuration, while ketones with the residues R = isobutyl (ee = 50%), neopentyl (ee = 66%), n-butyl (ee = 42%), isopentyl

(ee = 80%) and various keto-esters (ee = 82 -> 98%) to (R) -configured alcohols. The disadvantage here is that the formal configuration reversal cannot be precisely defined. It is only known that an (S) -configured product is still formed with R = propyl, while the larger isobutyl radical already leads to the (R) -alcohol. Thus, for other possible substrates, e.g. B. for ketones with halogenated alkyl radicals, do not clearly predict how the substitution pattern affects the stereoselectivity of the enzyme. With a few exceptions, the ee values of the alcohols obtained are low to moderate (42 - 80%), which is why the chiral alcohols are not suitable for use in natural and active ingredient synthesis. Furthermore, no example of the reduction of a ketone which is substituted on the triple bond is described for TE-ADH. The predominantly moderate ee values and the poorly predictable configuration of the products make the reduction using TE-ADH unsuitable for targeted natural product synthesis, especially since the literature does not describe TE-ADH reduction of ketones with substituted triple bonds.

In M.H. Ansari, T. Kusumoto, T. Hiyama, Tetrahedron Lett. 1993, 34, 8271 describes the reduction of methyl 3-keto-4-pentynate with baker's yeast.

While the derivative with the terminal triple bond is reduced to the (S) -enantiomer (ee = 80%) in 46% yield, the reduction of the corresponding ketone with a triethylsilyl substituent on the triple bond leads to the (R ) Product (ee = 82%). However, this is a whole-cell biotransformation in which the above-mentioned Reductions are presumably catalyzed by different enzymes. It is therefore not a formal reversal of the configuration of alcohols, the synthesis of which is catalyzed by the same oxidoreductase, but rather products which are formed by reducing a substrate by two or more enzymes. The moderate yields and optical purities achieved also reflect that different biocatalysts are likely to be competing in the reduction.

The formation of a propargylic, terminal epoxide from an α-halogen alcohol has been described three times in the literature. However, racemic compounds are used as starting materials in all three representations. In MR Lespieau, Bull. Soc. Chim. Fr. 1928, 4, 203 l-chloro-3-butyn-2-ol is reacted with KOH in ether without information on the yield to give the epoxide. The same reaction is described in H. Kleijn, J. Meijer, GC Overbeek, P. Vermeer, J. R. Neth. Chem. Soc. 1982, 101, 97. The reaction is carried out at 70 ° C and the yield is 45%. In D. Bernard, A. Doutheau, J. Gore, J. Moulinoux, V. Quemener, Tetra-hedron 1989, 45, 1429, 5-amino-1-chloro-5-methyl-3-hexin-2-ol is also used t-BuOK converted into the epoxide in only 5% yield. Due to the strongly basic conditions, substrates with more sensitive functional groups may not be suitable for the reactions described, which severely limits their general applicability. Another disadvantage is the fact that all examples have been carried out starting from racemic starting materials. This means that no say whether optically active alcohols undergo racemization due to the action of the strong bases used. With the exception of one amine substituent, the three methods have not been tested for substrates with functional groups, and the yields are low at 5% and 45%, respectively. Furthermore, the syntheses were carried out exclusively on racemic starting materials and products.

For non-propargyl-like terminal systems, a large number of reactions are described in the literature in which α-halogen alcohols are converted into epoxides. A wide range of reagents are available for this, but they have no general applicability. For each new epoxy formation conditions must be worked out that take into account the corresponding degree of activation, the functionalization of the alcohol and possible reactivities of the substituents.

The processes known from the prior art for the production of optically active, propargyl-terminal epoxides have in common that they lead to a moderate to low overall yield with a partially low enantiomeric excess.

It is therefore the object of the invention to provide a production process for optically active, propargyl-like, terminal epoxides which delivers a high total enantiomeric excess in high overall yield. The

The process should therefore be carried out in a few combined with a simple work-up of the intermediate products lead to the epoxides and thus do not have the disadvantages mentioned in the prior art.

Starting from the preamble of claim 1, the object is achieved according to the invention by the features specified in the characterizing part of claim 1.

With the method according to the invention, it is now possible to produce optically active, propargylic, terminal epoxides in high overall yield and high ee.

Advantageous developments of the invention are specified in the subclaims.

The invention is explained below and further advantages are described.

The tables and the figure show results and a reaction scheme for the formal configuration reversal according to the invention.

It shows:

Table 1: Substrates according to the invention with the activity of LB-ADH and the enantiomeric excess and the configuration of the products (alcohol) in% for -halogenated substrates. Table 2: Comparison of α-halogen-substituted substrates with structurally identical, non-α-halogen-substituted substrates.

Table 3: Activities of LB-ADH with other alkynones.

Table 4: Substrates of the LB-ADH with which a formal reversal of the configuration occurs.

Table 5: Activities of other oxidoreductases on

Example of the substrate 4-t-butyldimethylsilyl-l-chloro-3-butin-2-one.

Figure 1: A reaction scheme for the application of the formal configuration reversal for the representation of both enantiomers of propargyli- see terminal epoxides.

Tables 1 to 5 list the activities of Lactobacillus brevis alcohol dehydrogenase and further oxidoreductases in% based on the standard specified in each case.

The designation Nb means "not determined".

According to the invention, an α-halogen-substituted propargyl ketone is reduced to an alcohol by means of an alcohol dehydrogenase, after which a ring closure to give a propargylic epoxide takes place by reaction with a base with elimination of the corresponding hydrohalic acid. The alcohol dehydrogenase can, for example, from thermophilic microorganisms, such as. B. Thermoanaerobium brockii or Thermoanaerobacter ethanolicus, from horse liver or from Lactobacillus, preferably Lactojbacillus brevis or Lactobacillus kefir, preferably recombinantly. The corresponding muteins, which can be expressed by an allele, homolog or derivative of the associated nucleotide sequence, are also intended to be included. Of the alcohol dehydrogenases of Lactobacillus brevis and Lactobacillus kefir, it is also possible to use biologically active parts, a modified form (for example to increase the temperature stability), isoenzymes or mixtures thereof. Among modified enzymes, alcohol dehydrogenases, preferably from Lactobacillus, are preferred

To understand Lactobacillus brevis or Lactobacillus kefir, in which there are changes in the sequence, for example at the N- and / or C-terminus of the polypeptide or in the range of conserved amino acids, without however impairing the function of the enzyme. These changes can be made by exchanging one or more amino acids according to known methods.

Isoenzymes are understood to mean enzymes with the same or comparable substrate and activity specificity, but which have a different primary structure.

The alcohol dehydrogenase, which is preferably obtained from horse liver or thermophilic microorganisms (such as thermo- anaerobium brockii or Thermoanaerobacter ethanolicus) or from Lactobacillus, and in a particularly preferred embodiment is recompressively overexpressed in Escherichia coli, converts the compounds of general formula 1. If the reaction is not carried out in a whole cell transformation, a cofactor such as NAD (P) H is added.

Formula 1 shows the course of the reaction for the process according to the invention, in which R ^{1 is} a radical R, R ²

Halogen, preferably chlorine or bromine, ADH is an alcohol dehydrogenase, preferably from Lactobacillus, preferably Lactobacillus brevis or Lactobacillus kefir, and can be used as an isolated enzyme, as a cell-free crude extract or in a whole cell transformation, and the base is preferably an amine base or amidine -Base, preferably DBU.

ee> 99%

Formula picture 1

The alcohols and epoxides are advantageously formed in enantiomerically pure form (ee> 99%). The epoxides are obtained from commercially available starting materials, starting from a step upstream of formula 1, shown in only three synthesis steps. The epoxides are formed in good overall yield (70-90% based on the reduction and epoxidation), so that the process according to the invention is particularly economical. Chirality is introduced by means of a catalytic reaction, and the work-up can be carried out simply and efficiently by extraction. A large number of alkynones can be used for the synthesis according to the invention, which differ in a variety of substituents on the triple bond. These substituents can also include reactive functional groups. The results are shown in Table 3.

In the first synthesis step of the process, an α-

Halogen, preferably α-chloro or α-bromo-substituted propargyl ketone, as shown in formula 2, reduced. The reduction is catalyzed by an alcohol dehydrogenase, preferably from Lactobacillus, preferably Lactobacillus brevis or Lactobacillus kefir. The conversion is highly enantioselective, which reduces the prochiral, propargyl ketones to alcohols with an ee> 99%. A wide variety of substituents and functional groups can be used on the triple bond of the ketone, and compounds with reactive functional groups without side reactions are also reduced in 100% conversion to the desired products. This large tolerance to substituents on the triple bond opens up a general approach to

Synthesis of optically active, propargylic, terminal epoxides. Since the enzymatic reduction through If the extraction is worked up continuously or discontinuously, the alcohols can be isolated in high yields.

R^z = Halogen

R ^z = halogen

ee> 99%

Formula picture 2

Surprisingly, the reduction of propargyl ketones with a terminal triple bond (R ¹ = H, R ² ≠ H), especially when using ADH from Lactobacillus brevis, leads to a formal configuration reversal in the products. These alcohols also have an ee> 99%. The triple bond can then be deprotonated and linked to a variety of substituents. This additional synthesis step provides easy access to both enantiomers of propargylic alcohols in high optical purities, with the introduction of chirality being catalyzed by the same enzyme.

α-Halogen-substituted propargyl alcohols form the intermediate stage of the present process. Various α-chloro or α-bromo-substituted propargyl ketones (formula 2) with the alcohol dehydrogenase from Lactobacillus brevis have been implemented (Table 1). In all of the alkynones we tested, this enzyme showed a high reduction activity. In order to confirm the measured activities, the ketones were then preparatively reduced to the corresponding alcohols in 100% conversion in batches. The cofactor regeneration can take place via a coenzyme (e.g. formate dehydrogenase, formate). Alternatively, the cofactor NADPH can be regenerated by the LB-ADH itself in the presence of a secondary alcohol (e.g. i-propanol). This results in a further advantage, since many enzymes of the halogen ketones used here can be alkylated and thus inhibited or deactivated, which causes problems when using coenzymes. The previously determined high levels of activity could be confirmed in all batch approaches. The characterization of the propargylic alcohols showed an ee> 99%. A comparison of the activities with the chlorinated and the corresponding non-chlorinated ketones showed that the activity is surprisingly reduced by only a maximum of 50% by an α-chloro substituent (see Table 2). This is an unexpectedly good result, since less activities are expected for the enzymatic conversion of α-halogen-substituted ketone substrates, or a deactivation of the alcohol dehydrogenase could be assumed due to the high alkylation potential of the substrates. Further substrate screenings (see Table 3) were therefore carried out with the non-chlorinated propargylic ketones, since these are usually less expensive to access. The high activities [with a uniform ee of> 99%] of the different non-chlorinated substrates show that even with these diverse substituents on the triple bond, the corresponding α-chlorinated or α-brominated ketones could be used for the present process.

The radical R ¹ is a component from the group consisting of saturated or unsaturated aliphatic, aromatic or heteroaromatic hydrocarbons, which can be mono- or polysubstituted, where the substituents are alkyl (for example straight-chain, branched, saturated, unsaturated, cyclic alkyl), aryl ( for example phenyl, tosyl, naphthyl, condensed aromatic, heteroaromatic) or aralkyl groups or heteroatoms, such as. B. F, Cl, Br, I or 0, S, P, N or combinations thereof. Furthermore, the radical R ^{1 can be} a component from the group of the silyl compounds (for example trialkylsilyl, dialkyl arylsilyl, diarylalkylsilyl), the carbonyls (for example esters), the amines (for example tert. Amine, sec amine, prim. A in, cyclic amine) or a proton.

However, the list is only intended to be exemplary and not restrictive.

R ² can be a component from the group of the halogens, preferably Cl and Br.

In addition to the high enantioselectivity, the reduction is characterized by a very wide range of substrates because all of the propargylic ketones we use, which carry a wide variety of substituents and functional groups R ¹ on the triple bond, have been converted.

All alcohols which are substituted on the acetylene radical (R ¹ ≠ H) have (R) configuration if they are not α-halogenated (R ² = H) and (S) configuration in the case of α-halogenated products (R ² = halogen).

The reduction can be carried out in a temperature range between 0 ° C. and 90 ° C., preferably 5-70 ° C., particularly preferably 15-40 ° C. The pH can be between 4 and 10, ideally it is 5-8, preferably 6.5.

The enzymatic reaction can be carried out in aqueous solution, but organic solvents, preferably water-soluble solvents, preferably low-chain alcohols, preferably i-propanol or ethanol, can also be added. The admixture can be up to about 30% of the organic solvent and has the consequence that the organic reaction components are better dissolved. Furthermore, the enzymatic reaction can also be carried out without a solvent, e.g. B. in the gas phase.

The mild conditions of the enzymatic reduction of (pH = 5 - 8, room temperature) in connection with the simple work-up (the products can be obtained purely by extraction or adsorption-elution, e.g. on resins) provide gentle product access which also enables the representation of alcohols with reactive substituents.

The particular advantages of this reaction step are the enantiomeric purity of the products (ee> 99%) and the extremely broad substrate spectrum of the ADH used. Furthermore, a coenzyme is not absolutely necessary for the cofactor regeneration (the ADH used can regenerate the cofactor itself when adding cosubstrate) and there are mild reaction conditions, such as pH = 5 - 8 and room temperature. In addition, isolating the products is particularly easy.

The reduction of propargyl ketones with a terminal triple bond (R ¹ = H, R ² ≠ H) surprisingly leads to a formal reversal of the configuration of the products, especially when reduced with LB-ADH. These alcohols are also formed highly enantioselectively (exception: R ² = CH ₃ ; ee = 34%). The terminal triple bond can then be deprotonated and substituted in many ways. The corresponding synthesis scheme is shown in Figure 1. This additional synthesis step provides easy access to both enantiomers of propargylic alcohols in high optical purities, with the introduction of chirality being catalyzed by the same enzyme.

The main disadvantage of reducing ketones with an oxidoreductase is that only one of the two desired enantiomers is accessible. Due to the reduction with the alcohol dehydrogenase from Lactobacillus brevis or Lactobacillus kefir, a large number of propargylic alcohols with different residues at the triple bond accessible (R ¹ ≠ H, R ² = R). These alcohols have the (R) configuration (or (S) in the case of the α-halogenated substrates). Surprisingly, however, the use of ketones with a terminal triple bond (R ¹ = H), especially in the reduction with ADH from Lactobacillus brevis, leads to a formal configuration reversal in the products as soon as R ² ≠ H (Table 4). The formal configuration reversal already takes place with l-pentin-3-one

(R ¹ = H, R ² = CH ₃ ) instead. Here, however, the difference in size of the substituents is not yet sufficient to achieve an ee> 99%. Surprisingly, all other ketones with a triple bond (R ¹ = H, R ² ≠ H) we tested are reduced to the enantiomerically pure (S) -alcohols (or (R) in the case of the α-halogenated substrates) ( see table 4).

This is the first description of an oxidoreductase that can produce alcohols with opposite configurations in an enantiomerically pure manner (ee> 99%). Through an additional chemical step, this ability is used to optically synthesize both enantiomers of the same propargyl alcohol using only one alcohol dehydrogenase (see FIG. 1): after the enzymatic reduction of l-chloro-3-butyne-2- on to (R) -l-chloro-3-butyn-2-ol (or the corresponding bromine compounds), the terminal triple bond can be deprotonated with a base and substituted with a wide range of functionalities. However, this order is not mandatory, rather the ring closure can According to reaction step 2 of formula 1 for the epoxide, the proton of the terminal alkyne group is replaced first and then.

Suitable bases are, for example, metal organyls, preferably lithium organyls (for example n-butyl lithium), alcoholates (for example potassium butylate), amides (for example LDA). The deprotonation is carried out in a preferably inert, non-protic organic solvent, for example ethers (for example diethyl ether, THF) or alkanes (for example pentane, hexane) at a temperature in a range from preferably - 150 ° C. to 50 ° C, particularly preferably -80 ° C to 10 ° C, performed. Instead of the proton, any substituent can be introduced at position R ^{1 '} .

Practically all substituents which can be introduced by the procedure outlined can be considered as R ^{1 '} , but the same substituents which have been mentioned for R ^{1 can be} mentioned by way of example. This general approach enables the synthesis of a large number of propargylic alcohols. In connection with the enzymatic reduction of ketones which are substituted on the triple bond, it is now possible for the first time to visualize both enantiomers of a large number of α-halogeno-propargyl alcohols.

For the first time it is possible to generate alcohols with opposite configuration enantiomerically pure (ee> 99%) using only one oxidoreductase. A variety of substituents can be introduced to the triple bond and it is a precise prediction of the absolute configuration of the alcohols possible.

As an alternative to ADHs from lactobacilus, the enzymatic reduction of the propargylic α-halogen ketones can also be carried out with other oxidoreductases. Both horse liver ADH (HL-ADH), T ermoanaerojbium jbrocJii-ADH (TB-ADH), Candi da boidinii-ADH (CB-ADH) and Candida parapsilosis carbonyl reductase (CPCR) reduce, for example, 4-phenyl-l-chlorine -3-butin-2-one and 4- t-butyldimethylsilyl-l-chloro-3-butin-2-one to the corresponding alcohol (see Table 5). With all of the enzymes listed, the activity could be confirmed in batch batches by demonstrating the formation of the chiral propargylic alcohols by chromatography and spectroscopy.

The second and final step of the present process is the conversion of the enantiomerically pure alcohol into the epoxide. The step is carried out by eliminating HX (X = halogen) using a base while maintaining the enantiomeric purity.

Bases which can be used are, for example, hydroxides, alcoholates, cyanide ions, amine bases, such as primary, secondary or tertiary amines, cyclic amines, amidine bases such as DBN or preferably DBU, or generally preferably mild bases.

The reaction is successful using DBU as a gentle base at room temperature and proceeds with 100% conversion (see formula 3). Depending on the functionalization, the isolated yield is 60-80%. All tested substituents and functional groups on the triple bond are stable under the very mild conditions. Thus, the α-halogenated, propargylic alcohols (accessible through the enzymatic reduction in a large variety) can be used for the formation of epoxides, so that the reaction has general applicability. The reaction proceeds without racemesis, whereby the ee of> 99% is retained.

R = halogen ee> 99%

Formula picture 3

Advantageously, this reaction step is generally applicable, gives good to high yields under mild conditions and there is no racemization.

The radical R ¹ is a component from the group consisting of saturated or unsaturated aliphatic, aromatic or heteroaromatic hydrocarbons, which can be mono- or polysubstituted, where the substituents are alkyl (for example straight-chain, branched, saturated, unsaturated, cyclic alkyl), aryl ( for example phenyl, tosyl, naphthyl, condensed aromatic, heteroaromatic) or aralkyl groups or heteroatoms, such as. B. F, Cl, Br, I or 0, S, P, N or combinations thereof. Furthermore can the radical R ^{1 is} a component from the group of the silyl compounds (for example trialkylsilyl, dialkyl arylsilyl, diarylalkylsilyl), the carbonyls (for example esters), the amines (for example tert. amine, sec amine, primary amine, cyclic Amine) or a proton.

However, the list is only intended to be exemplary and not restrictive.

R ² can be a component from the group of the halogens, preferably Cl and Br.

The reaction can be carried out in numerous solvents. For example, water and aqueous buffer systems or a protic or aprotic solvent or a solvent mixture are suitable, but the reaction can also take place without a solvent (for example in the gas phase).

The methods known from the prior art for the preparation of propargylic, terminal epoxides can only be used to a limited extent with the substrates specified here. The yields are usually less than 10% and some of the functionalized alcohols are not stable under these conditions.

Since this method represents the first general access to optically active, propargylic, terminal epoxides in high optical purities, a multitude of new application possibilities arise e.g. B. for natural product and drug synthesis. The Nucleophilic opening of the epoxides gives access to a number of interesting compounds. For example, when 1.2 equivalents of KCN are added in ethanol / water (4/1) at RT, the corresponding nitrile is formed after 100 hours with 100% conversion. After extraction, the pure product is obtained without further working up. It is not necessary to isolate the epoxide beforehand to synthesize the nitrile, since under the conditions described it can also be formed directly from the halogenated alcohol (formula 4). The intermediate stage of the epoxide was detected by GC-MS, but is immediately implemented in situ. The nitrile obtained can e.g. B. in the synthesis of HMG-CoA reductase inhibitors. This synthesis is described in K. Takahashi, T. Minami, Y. Ohara, T. Hiyama, Bull. Chem. Soc. Jpn. 1995, 68, 2649, the nitrile used (R ¹ = TBDMS; see formula 4) being shown in a 7-step synthesis.

R = halogen ee> 99%

Formula picture 4 Experimental part

Methods: All solvents were used in pa quality and dried if necessary using standard methods. The chemicals were purchased from Aldrich, Lancaster, TCI and Fluka. HL-ADH, TB-ADH and CB-ADH were purchased from Sigma; LB-ADH and CPCR were isolated at the Institute of Enzyme Technology at Heinrich-Heine University Düsseldorf (for LB-ADH see: B. Riebel, dissertation, Heinrich-Heine University Düsseldorf, 1996; for CPCR see: J. Peters, dissertation , Heinrich Heine University Düsseldorf, 1993). TLC: silica gel 60 F254 (Merck). Column chromatography: silica gel 60 (Merck). NMR spectra: AMX-300 ( ¹ H: 300 MHz, ¹³ C: 75.5 MHz) from Bruker Physik AG. Chemical shifts are reported in ppm relative to CHC1 ₃ as the internal standard. GC: Chrompack CP9002 with an FS-Cyclodex-ß-I / P (CS GmbH) and a Lipodex E column (Macherey-Nagel). HPLC: Hewlett Packard Series 1100 with a Chiracel OB and a Chiralpak AD column (Daicel Chem. Ind.) At 20 ° C, 0.5 ml / min. GCMS: HP 6890 series GC system equipped with an HP 5973 mass detector (EI, 70 eV) and an HP-5MS column (T _GC (injector) = 250 ° C, time program (oven): T _{0 mιn} = 60 ° C, T _{3 min} = 60 ° C, T _{14 min} = 280 ° C (20 ° / min), T _{19 min} = 280 ° C. Rotation values: Polarimeter 241 (Perkin Elmer).

substrate representation

The propargyl ketones were prepared using methods known from the literature. The synthesis can be carried out via deprotonation of a terminal alkyne with subsequent coupling of a Weinreb amide (S. Nahm, SM Weinreb, Tetrahedron Lett. 1981, 22, 3815), via the coupling of a trimethylsilyl-substituted alkyne with a Acid chloride (L. Birkofer, A. Ritter, H. Uhlenbrauck, Chem. Ber. 1963, 96, 3280) and the oxidation of the corresponding alcohol (K. Bowden, IM Heilbron, ERH Jones, BCL Weedon, J. Chem. Soc 1946, 39).

reduction

Enzyme Assays: For the photometric determination of the enzyme activity, the following solutions were combined and the decrease in absorbance at 340 nm was followed (ε _N AD (P) H 6.22 L mol ^"1 cm ^{" 1} ), 20 ° C: 970 μL solution of the Ketones in TEA / NaOH buffer (5 M ketone [2 mM for aromatically substituted alkinones due to the lower solubility], 100 mM TEA, ImM MgCl ₂ ) pH 7.0, 20 μL

NAD (P) H (12.5 mM) and 10 μL enzyme solution. The initial decrease in absorbance was recorded and set relative to the values of the standard substrate 5-oxo-hexanoic acid ethyl ester.

LB alcohol dehydrogenase-catalyzed reductions in propargylic ketones (preparative scale):

The ketone (1 mmol) was at RT with 50 mg NADP ⁺ , 1.5 mL i-propanol and 50 U LB-ADH (1 U enzyme reduced 1 μmol substrate per minute) in 100 mL TEA / NaOH buffer (100 mM TEA, 1 mM MgCl ₂ ; pH 6.5) stirred. After 16 h the reaction mixture was extracted with CH ₂ C1 ₂ (3 x 40 mL). The combined organic phases were dried over sodium sulfate, concentrated in vacuo and chromatographed on silica gel (hexane / ethyl acetate).

(R) -4-phenyl-3-butyn-2-ol: 85% yield; ee> 99% determined by means of HPLC on Chiracel OB (i-hexanes: i-propanol = 95: 5),

R _t = 24.3 min (R); ^X H-NMR (CDCl ₃ ): δ = 1.55 (d, J = 6.6 Hz, 3H, CH ₃ ), 2.13 (s, 1H, OH), 4.75 (q, -7 = 6.6 Hz, 1H), 7.31- 7.42 (m, 5H, Ar-H). - ¹³ C-NMR (CDC1 ₃ ): δ = 24.3 (CH ₃ ), 58.7 (CH), 83.9, 90.9 (C _q ), 122.5, 128.2, 128.3, 131.6 (C _ar ). GCMS: R _t = 8.0 min; m / z (%) = 146 (50) [M ⁺ ], 131 (100)

[M ⁺ - CH ₃ ], 103 (61) [M ⁺ - CH ₃ - CO], 77 (35) [C ₆ H ₅ ⁺ ]. [α] ²² _D +43.8 (c = 0.6, Et ₂ 0) (literature value of (S) -2a: [α] ²⁵ _D - 44.8 (c = 1.0, Et ₂ 0, ee = 97%); K. Nakamura , K. Takena-ka, A. Ohno, Tetrahedron: Asymmetry 1998, 9, 4429.)

(R) -4- (4-methoxy-phenyl-3-butyn-2-ol: 64% yield; ee> 99% determined by means of HPLC on Chiracel OB (i-hexane: i-propanol = 85:15), R _t = 29.0 min (R); ^ -NMR (CDC1 ₃ ): δ = 1.55 (d, J = 6.5 Hz, 3H, CH ₃ ), 1.90 (s, 1H, OH), 3.80 (s, 3H, OCH ₃ ), 4.75 (q, -7 = 6.5 Hz, 1H), 6.82 (d, J = 9.5 Hz, 2H), 7.37 (d, J = 9.5 Hz, 2H,

Ar-H). - ¹³ C-NMR (CDC1 ₃ ): δ = 24.7 (CH ₃ ), 55.5 (OCH ₃ ), 59.1 (CH), 84.1, 89.8 (C _q ), 114.1, 114.8, 133.3, 159.8 (C _ar ). GCMS: R _t = 10.3 min; m / z (%) = 176 (40) [M ⁺ ], 161 (100) [M ⁺ - CH ₃ ], 145 (5) [M ⁺ - 2CH ₃ ], 133 (26) [M ⁺ - CH ₃ - CO], 118 (14) [M ⁺ - 2CH ₃ - CO]. [α] ²² _D +34.8

(c = 1.0, Et ₂ 0) (literature value of (S) -2b: [α] ²⁵ _D - 35.1 (c = 1.0, Et ₂ 0, ee = 97%); K. Nakamura, K. Takenaka, A. Ohno, Tetrahedron: Asymmetry 1998, 9, 4429.) (S) -l-pentin-3-ol: 81% yield; ee = 34% determined by GC on Lipodex E (T = 35 ° C), R _t = 30.5 min (S) + 32.0 min (R); ^X H NMR (CDCl ₃ ): δ = 1.01 (t, J = 7.2 Hz, 3H, CH ₃ ), 1.72 (m, 2H, CH ₂ ), 2.12 (s, 1H, OH), 2.40 (s, 1H , Cl), 3.98 (t, J = 6.6 Hz, 1H, C3). - ¹³ C NMR

(CDC1 ₃ ): δ = 9.4 (CH ₃ ), 30.6 (CH ₂ ), 63.3 (CHOH), 72.9 (CH), 84.9 (C _q ). [α] ²² _D -8.4 (c = 0.5, Et ₂ 0) (literature value of (12) -6a: [α] ²² _D +34.0 (c = 2.2, Et ₂ 0, ee>97%); O. Steiner , C. Tamm, Tetrahedron Lett. 1993, 34, 6729.)

(S) -l-hexin-3-ol: 86% yield; ee> 99% determined by GC on HP FS-Cyclodex-ß-l / P (T = 60 ° C), R _t = 26.3 min (S); ^X H-NMR (CDC1 ₃ ): δ = 0.94 (t, J = 7.3 Hz, 3H, CH ₃ ), 1.42 - 1.77 (m, 4H, CH ₂ ), 2.10 (s, 1H, Cl),

2.12 (s, 1H, OH), 4.27 (t, J = 6.5 Hz, 1H, C3). - ¹³ C-

NMR (CDC1 ₃ ): δ = 18.4 (CH ₃ ), 25.5 (CH ₂ ), 39.8 (CH ₂ ), 64.5 (CHOH), 66.0 (CH), 72.9 (C _q ). GCMS (T = 60 ° C): R _t = 4.2 min; / z (%) = 97 (5) [M ⁺ - H], 83 (33) [M ⁺ - CH ₃ ], 70 (17) [M ⁺ - H ₂ 0], 55 (100) [M ⁺ - CH ₃ - CO].

[α] ²² _D -35.1 (c = 1.5, Et ₂ 0) (literature value: [α] ²² _D -24.0 (c = 2.4, Et ₂ 0; ee = 68%); K. Mori, H. Akao, tetrahedron 1980, 36, 91.)

(S) -4-t-butyldimethylsilyl-l-chloro-3-butyn-2-ol: 55%

Yield; ^ -NMR (CDC1 ₃ ): δ = 0.16 (s, 6H, Si-CH ₃ ), 0.98 (s, 9H, t-bu), 2.51 (s, 1H, OH), 3.68 (d, J ^" = 4.0 Hz, 2H, CH ₂ C1), 4.51 (t, J = 4.0 Hz, 1H, CH). - ¹³ C-NMR (CDC1 ₃ ): δ = - 1.3 (SiCH ₃ ), 12.1 (t-bu-C _q ), 23.4 (t-bu-CH ₃ ), 48.6 (CH ₂ ), 63.1 (CHOH), 92.0, 102.4 (C _q ).

GCMS: R _t = 8.1 min; / z (%) = 218 (5) [M ⁺ ], 203 (1) [M ⁺ - CH ₃ ], 161 (21) [M ⁺ - t-Bu], 95 (100) [M ⁺ - t-bu - 2CH ₃ - Cl]. (The rotation value is determined according to the following synthesis step; see (S) - 1 - t-butyldimethylsilyl - 3, 4 - epoxy-1-butyne)

Synthesis of race alcohols and separation of enantiomers (analytical scale):

The propargyl ketone (10 μmol) and 0.2 mg NaBH ₄ (5 μmol) were stirred in 1 ml EtOH at 0 ° C. for 1 h. The reaction mixture is hydrolyzed with 1 M HC1 (1 ml), saturated NaCl solution (4 ml) is added and the reaction mixture is extracted with CH ₂ C1 ₂ (1 ml). All alcohols were obtained in 100% conversion.

4-phenyl-3-butyn-2-ol: The enantiomers were separated by HPLC on Chiracel OB (i-hexane: i-propanol = 95: 5), R _t = 24.3 min (R) + 31.9 min (S).

4- (4-methoxy-phenyl) -3-butyn-2-ol: The enantiomers were separated by HPLC on Chiracel OB (i-hexane: i-propanol = 85:15), R _t = 29.0 min (R) + 38.2 min (S).

l-Butyn-3-ol: The enantiomers were separated by GC on HP FS-Cyclodex-β-I / P (T = 35 ° C), R _t = 13.9 min (R) + 15.5 min (S).

l-Pentin-3-ol: The enantiomers were separated by GC on HP Lipodex E (T = 35 ° C), R _t = 30.5 min (S) + 32.0 min (R). l-Hexin-3-ol: The enantiomers were separated by GC on HP FS-Cyclodex-β-I / P (T = 60 ° C), R _t = 26.3 min (S) + 28.4 min {R).

l-Octin-3-ol: The enantiomers were determined by GC on HP

FS-Cyclodex-ß-I / P (T = 80 ° C) separated, R _t = 41.8 min (S) + 44.9 min (fl) .GCMS: R _t = 5.0 min; m / z (%) = 125 (3) [M ⁺ - H], 111 (3) [M ⁺ - CH ₃ ], 107 (7) [M ⁺ - H ₂ 0 - H], 97 (19) [ M ⁺ - C ₂ H ₅ ], 83 (42) [M ⁺ - CH ₃ - CO], 55 (100) [C ₄ H ₇ ].

l-chloro-4-trimethylsilyl-3-butyn-2-ol: GCMS: R _t = 7.0 min; m / z (%) = 176 (1) [M ⁺ ], 161 (1) [M ⁺ - CH ₃ ], 127 (100) [M ⁺ - CH ₃ - Cl], 95 (93) [M ⁺ - 3CH ₃ - Cl].

l-bromo-4-trimethylsilyl-3-butyn-2-ol: GCMS: R _t = 7.7 min; m / z (%) = 221 (1) [M ⁺ ], 206 (1) [M ⁺ - CH ₃ ], 139 (64) [M ⁺ - Br], 127 (100) [M ⁺ - CH ₃ - Br].

General procedure for analysis of 4-silyl-3-butyne

2-ole: The silyl group of the alcohols was split off by adding (Bu) ₄ NF (0.2 mL; 1 M in THF) in THF (1 mL). After 1 h, saturated NaCl solution (2 mL) was added and the mixture was extracted with CH ₂ C1 ₂ . The products were obtained in 100% conversion and analyzed as listed below.

l-Chloro-3-butin-2-ol: The enantiomers were separated by GC on Lipodex E (T = 60 ° C), R _t = 59.1 min (S) + 65.4 min (1?). l-Bromo-3-butin-2-ol: The enantiomers were separated by means of GC on Lipodex E (T = 70 ° C.), R _t = 67.2 min (S) + 70.7 min (R).

Enzyme-catalyzed reduction of propargylic alcohols (analytical scale): The ketone (10 μmol) was at RT with 0.5 mg NAD (P) ⁺ , 15 μL 2-propanol (with HL-ADH: 12 μL ethanol) and 0.5 U enzyme in 1 mL TEA / NaOH buffer (100 M TEA, 1 mM MgCl ₂ ; pH 6.5) carefully shaken. After 16 h the reaction mixture was extracted with 200 μL CH ₂ C1 ₂ . The analysis of the approaches was carried out as described above; the activities, configurations and ee values of the alcohols are given in Tables 1-5.

epoxidation

(S) -lt-Butyldimethylsilyl-3,4-epoxy-but-l-in: To a solution of 100 mg (460 μmol) (S) - 4- butyldimethylsilyl-l-chloro-3-butyne-2 -ol in 50 ml CH ₂ C1 ₂ 216 mg (1.4 mmol) DBU were added at RT and stirred. After 22 h the solution is washed with water, dried over sodium sulfate and concentrated in vacuo. Column chromatography (pentane: CH ₂ C1 ₂ = 3: 1) gave 58 mg (317 μmol, 69% yield) of the epoxide.

^ -NR (CDC1 ₃ ): δ = 0.12 (s, 6H, Si-CH ₃ ), 0.94 (s, 9H, t-bu), 2.92 (d, J = 3.4 Hz, 2H), 3.38 (t, J = 3.4 Hz, 1H). - ¹³ C-NMR (CDCI3): δ = - 4.6 (SiCH ₃ ), 16.7 (t-bu-C _q ), 26.2 (t-bu-CH ₃ ), 40.1 (CH), 49.2 (CH ₂ ), 87.8 . 102.7 (C _q ). GCMS: R _t = 7.5 min; / z (%) = 182 (8) [M ⁺ ], 167 (4) [M ⁺ - CH ₃ ], 125 (100) [M ⁺ - t-Bu], 109 (14) [M ⁺ - t- bu - CH ₃ ], 97 (76) [M ⁺ -t-bu - CO]. [α] ²² _D +71.6 (c = 1.02, CH ₂ C1 ₂ ) (literature value of (R) -1-t-butyldimethylsilyl-3,4-epoxy-but-l-in: [α] ²⁰ _D - 72.3 (c =

5.63, CH ₂ C1 ₂ ); T. Kanger, P. Niidas, A.-M. Müürisepp,

T. Pehk, M. Lopp, Tetrahedron: Asymmetry 1998, 9,

2499.)

subsequent chemistry

(R) - 4-t-Butyldimethylsilyl-l-cyano-3-butyn-2-ol: To a solution of 100 mg (460 μmol) (S) - 4-t-butyldimethylsilyl-l-chloro-3- butin-2-ol in 100 ml ETOH / H ₂ 0 (4/1), 38 mg (575 mmol) KCN were added and the mixture was stirred at 50 ° C. After 16 h the ethanol is removed in vacuo. The mixture is then extracted with ethyl acetate, the organic phases are dried over sodium sulfate and concentrated in vacuo. Without further purification, 73 mg (350 μmol, 76% yield) of the pure nitrile were obtained.

^ -NMR (CDC1 ₃ ): δ = 0.16 (s, 6H, Si-CH ₃ ), 0.98 (s, 9H, t-bu), 1.50 - 2.00 (s, 1H, OH), 2.78 (d, J = 5.8 Hz, 2H, CH ₂ CN), 4.66 (t, J = 5.8 Hz, 1H, CH). - ¹³ C-NMR (CDCI ₃ ): δ = - 4.6 (SiCH ₃ ), 16.6 (t-bu-C _q ), 26.2 (t-bu-CH ₃ ), 27.6 (CH ₂ ), 59.0 (CHOH), 91.3, 103.3 (C _q ), 116.3 (CN). GCMS: R _t = 9.5 min; m / z (%) = 191 (3) [M ⁺ -H ₂ 0], 176 (2) [M ⁺ - H ₂ 0 - CH ₃ ], 152 (64) [M ⁺ - t-Bu], 134 (59) [M ⁺ - t-Bu - H ₂ 0], 111 (100) [M ⁺ - t-Bu - CH ₃ - CN]. Table 1

'based on 5-oxohexanoic acid ethyl ester as standard (in%) ^** configuration of the resulting alcohols

Table 2:

bezogen auf 5-Oxohexansäureethylester als Standard (in %) "Konfiguration der resultierenden Alkohole Tabelle 3

based on 5-oxohexanoic acid ethyl ester as standard (in%) "configuration of the resulting alcohols Table 3

^* based on ethyl 5-oxohexanoate as standard (in%) ^** not determined

^*** Configuration of the resulting alcohols (continued in Table 3)

Table 4

'based on 5-oxohexanoic acid ethyl ester as standard (in%) ^** configuration of the resulting alcohols

Table 5:

oxidoreductase

R ¹ = TBDMS

'Based on the respective standard: 5-oxohexanoic acid ethyl ester for Lactobacillus Jbrevis-ADH and Candida parapsilosis-CR; Cyclohexanone in horse liver ADH; 2-butanone in Thermoanaerobium brockii-ADH; Acetone in Candida boidinii -ADH) in%

"Configuration of the resulting alcohols ^* " very slow

Claims

Patent claims 1. Process for the preparation of optically active, propargylic, terminal epoxides of the formula 1,

formula 1 characterized in that an α-halogen-substituted propargylic ketone is reduced to an alcohol by means of an alcohol dehydrogenase and the resulting optically active, propargylic alcohol is reacted with a base to form an optically active, propargylic, terminal epoxide. 2. The method according to claim 1, characterized in that an alcohol dehydrogenase is used which is derived from Lactobacillus, Thermoanaerobium, Thermoethano- licus or horse liver. 3. The method according to claim 2, characterized in that the alcohol dehydrogenase is derived from Lactobacillus brevis or Lactobacillus kefir. 4. The method according to claim 2, characterized in that the alcohol dehydrogenase is derived from Thermoanaerobium brockii. 5. The method according to any one of claims 1 to 4, characterized in that the alcohol dehydrogenase is recombinant. 6. The method according to any one of claims 1 to 5, characterized in that biologically active parts, a modified form, isoenzymes or mixtures of alcohol dehydrogenase are used individually or in combination. 7. The method according to any one of claims 1 to 6, characterized in that the alcohol dehydrogenase is used in a full-line transformation or in isolation. 8. The method according to any one of claims 1 to 7, characterized in that the α-halogen substituted propargyl ketone is a compound of general formula 2:

9. The method according to claim 8, characterized in that Rl is a component from the group of saturated or unsaturated aliphatic, aromatic or heteroaromatic hydrocarbons, which can be mono- or polysubstituted, the substituents being alkyl, aryl or aralkyl groups or heteroatoms, such as B. F, Cl, Br, I or 0, S, P, N or combinations thereof or a component from the group of the silyl compounds, the carbonyls, the amines or a proton. 10. The method according to claim 8 or 9, characterized in that R2 is a halogen, preferably Cl or Br. 11. The method according to any one of claims 1 to 10, characterized in that at least one component from the group hydroxide, alcoholate, cyanide, amine base, amidine base is used as the base. 12. The method according to any one of claims 1 to 11, characterized in that a mild base is used as the base. 13. The method according to any one of claims 1 to 12, characterized in that an amidine base, preferably DBU or DBN, is used as the base. 14. The method according to any one of claims 1 to 13, characterized in that the reduction by means of the alcohol dehydrogenase is carried out in a temperature range between 0 ° C and 90 ° C. 15. The method according to claim 14, characterized in that the reduction is carried out in a temperature range between 5 ° C to 70 ° C. 16. The method according to any one of claims 1 to 15, characterized in that the reduction with the alcohol dehydrogenase is carried out at a pH of 4 to 10. 17. The method according to any one of claims 1 to 16, characterized in that the reduction with the alcohol dehydrogenase is carried out in aqueous solution or without a solvent. 18. The method according to any one of claims 1 to 17, characterized in that the reduction with the alcohol dehydrogenase is carried out in the gas phase. 19. The method according to claim 17, characterized in that an organic solvent is added to the aqueous solution. 20. The method according to any one of claims 1 to 19, characterized in that the reaction to the propargylic epoxide is carried out in a solvent. 21. The method according to any one of claims 1 to 19, characterized in that the reaction to the propargylic epoxide is carried out without a solvent. 22. The method according to any one of claims 1 to 21, characterized in that a ketone with a terminal triple bond is used as the α-halogen substituted propargyl ketone, the proton of the terminal triple bond being replaced by a substituent after the reaction with the alcohol dehydrogenase can. 3. The method according to any one of claims 1 to 22, characterized in that the resulting optically active, propargylic, terminal epoxy is used as the starting material for the production of a chiral product while opening the Epoxidrin-.