CROSS-REFERENCE TO RELATED APPLICATION

[0001] This application claims benefit and priority from U.S. provisional patent application No. 60/619,960, filed on October 20, 2004, the contents of which are incorporated herein by reference.

FIELD OF THE INVENTION

[0002] The present invention relates generally to methods to resolve two stereoisomers, including two enantiomers, from a mixture of same, and particularly to methods using a resolving agent to resolve stereoisomers, particularly enantiomers.

BACKGROUND OF THE INVENTION

[0003] A racemic compound or racemic mixture contains equal amounts of each of a pair of enantiomers. A mixture of two enantiomers of a compound results when chemical reactions are conducted in an achiral environment, since there is no selection for the formation of one enantiomer in favour of the other. Physically and chemically, the two enantiomers in a pair are identical, but each of the pair often reacts selectively and specifically when it comes to interactions with other molecules, including chiral molecules, including enzymes and other biomolecules. In the pharmaceutical industry, such selectivity and specificity can be crucial, since each of the pair of enantiomers can have vastly different physiological effects from the other. Even if one enantiomer of a pair produces the desired therapeutic effects, the other may be inactive or produce undesirable side effects.

[0004] Consequently, there is an important need for methods for isolating a single enantiomer when new pharmaceutical drugs are developed, so that the pharmacological effects of each enantiomer can be tested separately. Other areas in which separation of two enantiomers may be required include the development of agrichemicals, flavors in foods and beverages, as well as compounds for use in fragrances.

[0005] Separation of a pair of enantiomers is generally conducted by resolution of

a mixture of the two enantiomers using enzymes. Existing methods of such enzyme- based enantioselective separation include use of enzymes as biocatalysts for enantioselective biotransformations; immobilisation of enzymes on solid supports as a chiral stationary phase for chromatography (Mano, N., et al. Chromatography 24: 19- 34 (2003)); entrapment of enzymes in membrane materials for enantioselective membrane separation (Lakshmi, B. B., and Martin, C. R. Nature 388: 758-760 (1997)); and use of enzymes in aqueous solution as a complexing agent for enantioselective ultrafiltration (Carnier, F., et al. Separation & Purification Technology 16: 243-250 (1999)).

[0006] The use of enzymes as biocatalysts for enantioselective resolution requires that the enzyme possesses at least a minimum level of activity or stability in order to be able to catalyse biotransformation of the substrates. Accordingly, enzymes that have almost no or low activity, stability or ability to transform the substrates would not be useful' in such enzymatic resolution processes. Since enzymes tend to be very sensitive molecules that denature or lose activity fairly readily, it would be advantageous to be able to deVelop methods of enantiomeric resolution that did not depend on an enzyme retaining its biological activity or ability to bind enantio- specifically to a substrate or other ligand.

SUMMARY OF THE INVENTION

[0007] In one aspect, the present invention provides a method for resolving a first stereoisomer and a second stereoisomer of a compound from a mixture of said first stereoisomer and said second stereoisomer, said method comprising: dissolving said compound and a resolving agent in an organic solvent, said resolving agent forming a first complex with said first stereoisomer and a second complex with said second stereoisomer, said first complex having a faster dissociation rate than said second complex; crystallizing said compound from said organic solvent, whereby as a result of the faster dissociation rate of the first complex relative to the second complex, more of the first stereoisomer crystallizes than said second stereoisomer and separating crystals of said compound from said organic solvent, to obtain a crystalline fraction enriched in said first stereoisomer and an organic solvent fraction enriched in said second stereoisomer. In one embodiment, the first stereoisomer is a first enantiomer and the second stereoisomer is a second enantiomer. In another

embodiment, the first stereoisomer is a first diastereomer and the second stereoisomer is a second diastereomer. In one embodiment, the resolving agent used is an enzyme.

[0008] In another aspect, the present invention provides a method for resolving a first enantiomer and a second enantiomer from a racemic compound, comprising: dissolving in an organic solvent the racemic compound and an enzyme that is capable of specifically binding the racemic compound to form a first complex with said first enantiomer and a second complex with said second enantiomer; crystallizing the compound to form crystals enriched for the first enantiomer over the second enantiomer; and separating the remaining organic solvent from the crystals.

[0009] Other aspects and features of the present invention will become apparent to those of ordinary skill in the art upon review of the following . description of specific embodiments of the invention in conjunction with the accompanying figures.

BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In the figures, which illustrate, by way of example only, embodiments of the present invention,

[0011] Figure 1 contains the chemical structure for ketoprofen; and

[0012] Figures 2A-C are HPLC chromatograms of (A) the beginning solution of racemic ketoprofen, (B) filtrate I and (C) filtrate II showing the absorbance peaks for acetone, and the R and S enantiomers of ketoprofen, respectively, for (A) and the absorbance peaks for the R and S enantiomers of ketoprofen, respectively, for both (B) and (C).

DETAILED DESCRIPTION

[0013] This invention provides a novel method for stereoisomeric resolution, including enantiomeric resolution, for example by using enzymes to form enzyme- enantiomer complexes for enantioselective separation by crystallisation. The enzyme- mediated enantioselective separation is carried out by making use of the difference in dissociation of the enzyme-enantiomer complexes regardless of the enzymes' activity and stability.

[0014] Due to their topography and arrangement of functional groups, enzymes

are able to distinguish between the two enantiomers of a substrate and to catalyze reactions involving each enantiomer at different rates. This difference in catalytic rate has been used to resolve racemates kinetically (Klibanov, A. M. Nature 409: 241-246 (2001); Yang, H., et al. Journal of Organic Chemistry 64: 1709-1712 (1999)).

[0015] Kinetic analyses indicate that, in many cases, the enantio-recognition of an enzyme for its substrate is determined, at least in part, by the dissociation rate of the enzyme-substrate complex rather than by its formation rate, as evidenced by similar values for the Michaelis constant K _m (which reflects the affinity of substrate to enzyme to form the enzyme-substrate complex) and the markedly different maximum velocity V _max values (which indicates the dissociation of the enzyme-substrate complex) for the two enantiomers of an enzyme substrate (Kamiya, N., et al. Biotechnology and Bioengineering 65: 227-232 (1999); Okahata Y, et al. Tetrahedron Asymmetry 6: 1311-1322 (1995); Watanabe, K., et al. Bioorganic Chemistry, 29: 65- 76 (2001)).-'^'

[0016] An enzyme's stability and biospecificity for its natural ligand tends to be dependent on an aqueous reaction environment, and enzyme reactions are therefore typically conducted in aqueous buffer systems. However, many pharmaceutically active small molecules have low or no solubility in aqueous solutions and tend to be much more soluble in organic solvents. This presents a difficulty in using enzymes as biocatalysts to resolve two enantiomers kinetically, since in many cases enzymes show very low activity and/or stability in organic solvents (Kamiya, N., et al. Biotechnology and Bioengineering 65: 227-232 (1999); Okahata Y, et al. Tetrahedron Asymmetry 6: 1311-1322 (1995)).

[0017] The inventors have surprisingly discovered that a chiral resolving agent such as an enzyme that binds to a chiral compound in a specific manner can be used to separate the enantiomers of that compound, even when the enzyme is not recognizing the compound as a substrate for catalysis. The enzyme will interact differently with each enantiomer, thus leading to greater sequestering of one enantiomer by the enzyme than the other enantiomer due to the different dissociation constants of the two enantiomer-enzyme complexes, leaving the less sequestered enantiomer available for crystallization. Since the catalytic activity of the enzyme is not required for resolution, this novel resolution method can be conveniently

performed in organic solvents.

[0018] As will be understood, the term "stereoisomer" refers to one of a set of isomers whose molecules have the same number and kind of atoms bonded to each other, but which differ in the way these atoms are arranged in space.

[0019] In one embodiment of the present method, a starting mixture of two enantiomers of a compound is used, the compound being a compound which is desired to be resolved into separate solutions of each enantiomer, or which is desired to be further enantiomerically enriched. The mixture may already be enantiomerically enriched for one enantiomer, or the mixture may be a racemic mixture of the two enantiomers.

[0020] As will be understood, the term enantiomerically enriched describes a mixture of two enantiomers in which one enantiomer is present in greater quantity than the other ^' enantiomer.

[0021] As will be understood, the term racemic compound is used to refer to a mixture of equal amounts of two compounds that are mirror images of each other, or enantiomers. The term racemate refers to a homogeneous phase containing equimolar amounts of enantiomers. Where a racemic compound is for a chemical having one or more than one chiral center, each chiral center in one of the two compounds forming the racemic compound will be of opposite chirality to the corresponding chiral center in the other compound in the racemic compound. The term enantiomer, used interchangeably with enantiomeric compound, refers to one of the two compounds that are mirror images of each other, and which form a racemic compound when present in equal quantities.

[0022] Thus, a resolving agent is dissolved in an organic solvent. The resolving agent is a chiral compound that can interact with and bind specifically to the compound that is to be resolved into enantiomers, to form a resolving agent/enantiomer complex for each of the two enantiomers. Each resolving agent/enantiomer complex will have different dissociation rates. The resolving agent may be a chiral small molecule, a protein, or another chiral biomolecule. "Binding in a specific manner" or "specific binding" refers to a binding interaction between an enantiomer of the compound and the resolving agent that occurs at one or more

particular sites, or between particular functional groups, on both the compound and the resolving agent, and which binding is saturable at a given concentration of either resolving agent or compound.

[0023] The solvent is any organic solvent in which the compound that is to be resolved is soluble and in which the resolving agent is soluble or in which the resolving agent may be solubilized through use of a suitable surfactant. Preferably, the organic solvent is not miscible with water, although a water-miscible solvent may be used. In certain embodiments the solvent is apolar. The organic solvent may be, for example, a hydrophobic solvent, for example, isooctane, hexane, heptane, cyclohexane, octane or toluene.

[0024] In one embodiment, the resolving agent is a protein or a peptide. In a specific embodiment, the resolving agent is an enzyme. The protein, peptide or enzyme may, be any protein, peptide or enzyme that will bind the compound in a specific manner when dissolved in the organic solvent in which the compound is to be enantiomerically resolved. _Λ

[0025] When the resolving agent is an enzyme, although the racemic compound to be resolved may be a substrate of the enzyme, and may interact with the enzyme by binding at the active site, the enzyme need not bind the substrate at the enzyme active site, nor be active in the organic solvent. Furthermore, the compound does not need to be a substrate of the enzyme, since the enzyme is not used as a catalyst in the present method. However, if the racemic compound is a substrate of the enzyme, the enzyme need not be able to transform or catalyze the transformation of the substrate in the organic solvent. For example, when dissolved in the organic solvent, the enzyme may lack or be deficient in one or more biological activities normally associated with the enzyme. In certain embodiments, the enzyme may be an apoenzyme or a specifically mutated inactive enzyme.

[0026] In certain embodiments the enzyme is a lipase or a chymotrypsin, including a lipase from Candida rugosa, Mucor jananicus or Rhizopus oryzae or α- chymotrypsin or it may be penicillin G acylase enzyme.

[0027] If the resolving agent is not soluble, or not fully soluble in the organic solvent that is to be used, for example when the resolving agent is an enzyme, the

enzyme may be solubilized with the help of a surfactant followed by addition of the enzyme/surfactant complex to an organic solvent. The surfactant may be added to the organic solvent, for example an organic solvent in which the enzyme is to be solubilized. An aqueous buffer containing the enzyme may then be mixed with the organic solvent containing the surfactant, which may then be separated from the aqueous solution, for example by settling of the mixture or by centrifugation to separate the aqueous and non-aqueous phases. If a water-miscible solvent is used, it is not necessary to remove the aqueous phase. However, the resulting mixture of aqueous solution and organic solvent should be suitable to dissolve and crystallize the compound that is to be resolved.

[0028] Alternatively, the surfactant may first be added to the aqueous solution containing the enzyme, which is then mixed with the organic solvent. The aqueous buffer solution then is separated from the organic solvent, if immiscible with water, resulting in the enzyme solubilized by the help of the surfactant in the organic solvent.

[0029] If a surfactant is used, the surfactant should be one that will help solubilize the resolving agent (for example an enzyme) and which is soluble in the organic solvent that is to be used. Preferably, for ease of separation of the phases, the surfactant is not one which will create a microemulsion between the aqueous buffer and the organic solvent. In certain embodiments the surfactant is an ionic surfactant, which allows for the surfactant to form an ion pair with the enzyme resolving agent if the enzyme has charged surface groups. In certain embodiments, the surfactant is able to form reverse micelles encapsulating the enzyme. In a particular embodiment the surfactant is sodium bis (2-ethylhexyl) sulfosuccinate.

[0030] The amount of surfactant used will depend on the particular surfactant, the amount and type of resolving agent used, and the amount and type of organic solvent used. Enough surfactant is used to solubilize a sufficient amount of resolving agent in the desired volume of organic solvent, which can be readily determined by a skilled person. For example, sufficient amounts of surfactant are used so as to form an ion- paired complex with the enzyme or to form reverse micelles containing the enzyme. The concentration of the resolving agent in the organic solvent can be determined using standard methods, including protein staining or UV absorbance if the resolving agent is a protein, peptide or an enzyme. A protein detection method should be

chosen based on the organic solvent used so as to prevent interference of the solvent with the protein concentration determination.

[0031] The mixture of enantiomers of the compound is then dissolved in the organic solvent containing the resolving agent. It will be understood that the order of addition of the resolving agent and the mixture of enantiomers is not critical and that alternatively, the mixture of enantiomers may be first dissolved in the organic compound, followed by addition of the resolving agent.

[0032] The volume of organic solvent used and the amount of the mixture of enantiomers of the compound can be varied, depending on the nature of both solvent and the compound, as well as the desired concentration of the compound and the total amount of enantiomer that is desired to be resolved. The most efficient ratio of enantiomer mixture:resolving agent is 1 :1, although a ratio of >1 or <1 may also be used. At yatios below 1 :1, the capacity of the resolving agent to resolve the enantiomers will not be fully utilized, and at ratios greater than 1 :1, each resolution step will be less efficient than at a ratio of 1 :1, which may increase the number of resolution steps required to reach a desired level of resolution of enantiomers.

[0033] Once the mixture of enantiomers of the compound is dissolved in the solvent containing the solubilized resolving agent, the compound is allowed to crystallize.

[0034] In one embodiment, the crystallization is achieved by reducing the volume of solvent, thereby concentrating the compound and enzyme in the remaining solvent until the compound is allowed to crystallize.

[0035] A convenient way to reduce the volume of the solvent is to allow the solvent to evaporate. Such a method is slow enough to allow for ordering of the enantiomer molecules within the solvent, leading to the formation of crystals. However, it will be appreciated that other methods of reducing the volume and/or crystallizing of the enantiomers can be used. For example, for a molecule that crystallizes readily in a particular organic solvent, filtration using a filter having pores small enough to retain the resolving agent and the compound may be used.

[0036] Once crystals of the compound have formed, the remaining solvent

containing un-crystallized enantiomers and the resolving agent is separated from the crystals. The un-crystallized enantiomer solution will be enriched for one enantiomer over the other (i.e. the enantiomer for which the resolving agent/enantiomer complex has a slower dissociation rate), due to the differential dissociation of the resolving agent with each of the enantiomers, and the resultant preferential sequestering of one enantiomer by the resolving agent, reducing the amount of the preferentially sequestered enantiomer available for crystallization. Conversely, the crystalline fraction will be enriched in other enantiomer (i.e. the enantiomer for which the resolving agent/enantiomer complex has a faster dissociation rate).

[0037] Depending on the dissociation rates between the resolving agent and each enantiomer and the crystallization rates of the two enantiomers, as well as the crystal form and the intermolecular interactions within the crystal form, the crystals formed may be wholly of one enantiomer or the other, and both types of crystals may form or only one typ^'of crystal may form, or crystals containing both enantiomers may also form. However, given that one enantiomer is sequestered to a greater extent by the resolving agent, the crystal phase should be enriched for one enantiomer over the other.

[0038] At this point, methods may be used to determine the amount of enantiomer excess (e.e.) of the compound remaining in the solvent. As will be understood, for a mixture of (+)- and (-)-enantiomers, with composition given as the mole or weight fractions F ₍₊₎ and F _H (where F ₍₊₎ 4- F _(-) = 1), the enantiomer excess is defined as |F ₍₊₎ - F(.)| (and the percent enantiomer excess by 100|F ₍₊₎ - F _(-) | ). The enantiomer excess can be determined by measurement of the amount of each enantiomer remaining, for example, by analysing an amount of the remaining solvent containing the enantiomers using a chiral HPLC column.

[0039] The above steps of crystallizing the compound and separating the crystals from the solvent/resolving agent/enantiomers can be repeated as necessary until the desired degree of resolution between the two enantiomers is reached, for example, repeated from 1 to 10 times.

[0040] The present method can result in a percent enantiomeric enrichment of one enantiomer in the remaining solvent of, for example, 10, 20, 25, 30, 40, 50, 60, 70, 75,

80, 90, 95, 96, 97, 98 or 99%, depending on the dissociation rate of the resolving agent with each enantiomer and the number of times that the method is repeated.

[0041] Furthermore, since the crystals of the compound will also be enantiomerically enriched, albeit for the enantiomer that is not enriched in the remaining solvent, once the desired degree of enantiomeric resolution has been obtained, the enantiomer contained in the crystals can be harvested to obtain the resolved, or partially resolved enantiomer contained in the crystals. If desired, and if the crystals contain an enriched enantiomer mixture, the crystals can be redissolved in the organic solvent and the method can be repeated in order to further resolve the enantiomer mixture.

[0042] This above method can be readily scaled up in order to provide large quantities of resolved enantiomers.

[0043] A's ⁷ well, the resolving agent can be readily recovered for repeated usage, for example by ultrafiltration using a filter with an appropriate molecular weight cut¬ off, particularly where the resolving agent is an enzyme, due to the relatively weak interaction and the large molecular weight difference between the enzyme and enantiomers.

[0044] The present method makes it possible to resolve a mixture of enantiomers of a compound using enzymes that cannot be used as biocatalysts due to low activity and/or stability. This method makes it possible to obtain a high enantiomer excess using enzymes that would not be useful for enantiomer selection by biotransformation methods. Furthermore, since the method is performed in an organic solvent in which the racemic compound is soluble, it is not necessary to form a diastereoisomeric salt with the racemic compound prior to crystallization, negated the requirement to subsequently isolate the free base or free acid of the enantiomer once resolved.

[0045] Although the above description is given in respect of separation of two enantiomers from a mixture, it will be understood that the present method is also useful for separation of two diastereomers of a compound from a mixture of the diastereomers. That is, a chiral resolving agent can be used that interacts with both diastereomers in a mixture, but which has different dissociation rates for each diastereomer. A diastereomer is a chiral compound that has more than one chiral

center, and which differs from another stereoisomer of the compound (the other diastereomer in the mixture) with respect to stereochemistry at one or more chiral centers, but not at every chiral center and therefore is not an enantiomer of the other stereoisomer.

[0046] The invention is further illustrated by the following non-limiting examples.

EXAMPLES

[0047] This example illustrates the resolution of ketoprofen (shown in Figure 1), a member of the profen family (ibuprofen, ketoprofen, flurbiprofen, naproxen etc), as an example to show how this method works, using three lipases from Candida rugosa (Sigma, Type VII), Mucor jananicus (Lipase M "Amano" 10) and Rhizopus oryzae (Fluka). These enzymes were first made soluble in organic solvents by forming an ion-paired complex with an ionic surfactant, sodium bis (2-ethylhexyl) sulfosuccinate (AOT) ₅ as follows.

[0048] One hundred milliliters of lipases (10 mg/ml) (in 0.01 M phosphate buffer, pH 6.0) containing CaCl ₂ (10 mM for the lipase from Candida rugosa and 100 mM for the lipase from Rhizopus oryzae) or 300 mM NaCl (for the lipase from Mucor javanicus) was mixed with 100 ml isooctane containing AOT (2 mM) under magnetic stirring at 650 rpm for 10 min. The mixture was then placed at room temperature for 15 min followed by centrifugation at 6000 rpm for 5 min. The transparent isooctane layer was collected and the protein content determined based on the UV absorption at 280 nm. The lipase concentration was adjusted by evaporating or adding isooctane when necessary.

[0049] In the case of α-chymotrypsin, 100 ml of α-chymotrypsin (1 mg/ml) (in 0.01 M phosphate buffer, pH 7.8) containing CaCl ₂ (2 mM) was mixed with 100 ml isooctane containing AOT (2 mM). The other operations were the same with those for the lipases.

[0050] Ketoprofen was analyzed by HPLC with a Chiralcel™ OJ-H column (4.6 x 250 mm, Daicel Industrial, Japan). Samples (5 μl) were eluted by a mixture of n- hexane: 2-propanol: acetic acid (90:10:0.5, v/v/v) at 1.0 ml min ^"1 and detected at 254 nm.

[0051] The examples for the enzyme-mediated enantioselective crystallization of ketoprofen are as follows.

[0052] Example 1: Ketoprofen (180 mg) was dissolved in acetone (6 ml) followed by addition of 60 ml isooctane containing the ion-paired lipase from Candida rugosa (0.36 mg/ml). The transparent solution thus formed was allowed to evaporate naturally in a fume cupboard to a final volume of 12.5 ml, leading to a crystallization of 97.0% of ketoprofen. The crystals were filtered off and the R- and S- enantiomers in the filtrate I were analyzed (HPLC) to be 0.139 and 0.286 mg/ml, respectively, corresponding to an e.e. of 34.5%. The filtrate I was further evaporated to 1.8 ml, leading to 38.5% of ketoprofen crystallized from filtrate I. The crystals were removed by filtration and the R- and S-enantiomers in the filtrate II were respectively 0.306 and 1.51 mg/ml, corresponding to an e.e. of 66.7%. The chromatograms of the beginning solution of racemic ketoprofen, filtrate I and filtrate II are showπ'ih Figures 2A-C, respectively.

[0053] Detailed legend £ρr Figure 2: Chromatograms of ketoprofen in the beginning transparent solution (A), filtrate I (B) and filtrate II (C). Ketoprofen (180 mg) was dissolved in acetone (6 ml) followed by addition of 60 ml isooctane containing the ion-paired Candida rugosa lipase (0.36 mg/ml). The transparent solution thus formed was let evaporate naturally in a fume cupboard to 12.5 ml and the crystals thus formed were filtered off to give the filtrate I, which was further evaporated to 1.8 ml and the crystals thus formed were removed to give filtrate II. The first peak in (A) is acetone.

[0054] Example 2: Ketoprofen (180 mg) was dissolved in acetone (12 ml) followed by addition of 60 ml isooctane containing the ion-paired lipase from Mucor javanicus (0.36 mg/ml). The transparent solution thus formed was let evaporate naturally in a fume cupboard to 15.5 ml, leading to a crystallization of 96.5% of ketoprofen. The crystals were filtered off and the R- and S-enantiomers in the filtrate I were 0.173 and 0.228 mg/ml, respectively, corresponding to an e.e. of 13.7%. The filtrate I was further evaporated to 2.2 ml, leading to a crystallization of 40.6% of ketoprofen from filtrate I. The crystals were removed and the R- and S-enantiomers in the filtrate II were 0.588 and 1.091 mg/ml, respectively, corresponding to an e.e. of 29.9%.

[0055] Example 3: Ketoprofen (180 rag) was dissolved in acetone (12 ml) followed by addition of 60 ml isooctane containing the ion-paired lipase from Rhizopus oryzae (0.36 mg/ml). The transparent solution thus formed was let evaporate naturally in a fume cupboard to 12.5 ml, leading to a crystallization of 98.2% of ketoprofen. The crystals were removed and , the R- and S-enantiomers in the filtrate were 0.116 and 0.142 mg/ml, respectively, corresponding to an e.e. of 10.1%. The filtrate was further evaporated to 2.0 ml, leading to 63.2% of ketoprofen crystallized from filtrate I. The crystals were removed and the R- and S-enantiomers in the filtrate II were 0.211 and 0.382 mg/ml, respectively, corresponding to an e.e. of 28.7%.

[0056] It should be mentioned that the three lipases showed almost no activity or very low activity and enantioselectivity (E<2) for ketoprofen esterification with ethanol in isooctane, which makes them not feasible to be used as biocatalysts for commercial resolution of ketoprofen through the esterification route, but they worked well in the'^dnzyme-mediated crystallization process, giving a high e.e. of the S- ketoprofen in the remaining filtrate after a two-step crystallization, although further improvement is still possible fey further evaporation of the solvents. The enzymes are expected to be easily recovered and reused by ultrafiltration considering the large weight difference and weak interaction between enzyme and enantiomer molecules.

[0057] Example 4: Ketoprofen (360 mg) was dissolved in acetone (30 ml) followed by addition of 120 ml isooctane containing the ion-paired α-chymotrypsin (0.70 mg/ml). The transparent solution thus formed was let evaporate naturally in a fume cupboard to 83 ml, leading to a crystallization of 89.5% of ketoprofen. The crystals were filtered off and the R- and S-enantiomers in the filtrate I were analyzed (HPLC) to be 0.180 and 0.227/ml, respectively, corresponding to an e.e. of 21.2%. The filtrate I was further evaporated to 8.4 ml, leading to 88.4% of ketoprofen crystallized from filtrate I. The crystals were removed by filtration and the R- and S- enantiomers in the filtrate II were respectively 0.085 and 0.440 mg/ml, corresponding to an e.e. of 67.6%.

[0058] Therefore, this invention provides a novel tool for resolving racemic compounds using enzymes especially those that cannot be used as biocatalysts due to low activity and/or stability, let alone low enantioselectivity. Table 1 shows the resolution of ketoprofen by lipases from Candida Rugosa, Mucorjavanicus, Rhizopus

oryzae and by α-chymotrypsin. For the lipases, ketoprofen (180 mg) was dissolved in acetone (6 or 12 ml) followed by addition of isooctane (60 ml) containing the ion- paired lipases (0.36 mg/ml) and the transparent solution was evaporated naturally. The remaining solvent was separated from the crystals by filtration and let evaporate to further improve the e.e. in the remaining solvent.

Table 1: Enantiomeric excess of ketoprofen in the remaining filtrate solutions after the first and second step enzyme-mediated crystallization

First-step crystallization Second-step crystallization

Enzyme % Crystallized ³ % e.e. ^b % Crystallized ⁰ % e.e. ^b

Candida rugosa

97.0 34.5 48.8 66.7 lipase

Mucor javanicus

96.5 13.7 50.5 29.9 lipase

Rhizopus oryzae

98.2 10.1 69.4 lipase 28.7 α-chymotrypsin 89.5 21.2 88.4 67.6

[0059] For Table 1: ^a' based on the initial ketoprofen; ^b' percent enantiomeric excess (e.e.) in the remaining filtrate solutions; ^c' based on the ketoprofen in the remaining solution after the first-step crystallization.

[0060] AU documents referred to herein are fully incorporated by reference.

[0061] Although various embodiments of the invention are disclosed herein, many adaptations and modifications may be made within the scope of the invention in accordance with the common general knowledge of those skilled in this art. Such modifications include the substitution of known equivalents for any aspect of the invention in order to achieve the same result in substantially the same way. All technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art of this invention, unless defined otherwise. The invention is intended to encompass all modification within its scope, as defined by the claims.