CHIRALITY 10:321–324 (1998)
Enantiomeric Separation of rac-Indoprofen by a Biocatalytic Procedure RAFFAELE MORRONE, GIOVANNI NICOLOSI,* AND MARIO PIATTELLI Istituto CNR per lo Studio delle Sostanze Naturali di Interesse Alimentare e Chimico-Farmaceutico, Valverde, Italy
ABSTRACT Lipase from Candida antarctica, commercially available immobilised on acrylic resin as Novozymt 435, allows for enantioselective esterification of racindoprofen (±)-1, with methanol in a dioxane-toluene solvent system. A double esterification process affords methyl ester (−)-(R)-2 in 85% e.e. and enantiopure (+)-(S)-1, both in good chemical yield. Chirality 10:321–324, 1998. © 1998 Wiley-Liss, Inc. KEY WORDS: NSAID; esterification; resolution; Candida antarctica lipase; racIndoprofen A considerable body of research has shown that two enantiomers often possess quite different pharmacological and biological activity.1 Therefore, the production of chiral drugs as single enantiomers with the desired activity, the eutomers, is at the moment a strategic goal for the pharmaceutical industry. Among the chiral drugs, an outstanding position is occupied by the 2-arylpropionic acids (the ‘‘profen’’ family) widely used for their non-steroidal antiinflammatory activity that, however, resides essentially in the S-form,2 while the antipode has usually different pharmacological properties (for instance, (R)-flurbiprofen is an analgesic agent).3 Resolutions of racemic profens exploiting the stereoselectivity of lipases have been reported,4 among them the most favourable being those operating in non-aqueous media.5 More particularly, methodologies based on enantioselective esterification appear more advisable, since they do not require the preliminary conversion into the corresponding ester, as is compulsory in the application of procedures based on ester hydrolysis. Following this line, four racemic profens have been resolved, namely ibuprofen,6 naproxen,7 flurbiprofen,8 and suprofen.9 The aim of this work has been the resolution of (±)indoprofen (2-[4-(1-oxo-2-isoindolinyl)phenyl]propionic acid), presently marketed as racemate notwithstanding that it is known that its enantiomers are characterised by different degrees of interaction with serum proteins10 and different pharmacokinetic parameters.11 MATERIALS AND METHODS General 1 H NMR spectra were recorded in CDCl3 solution on an AC 250 Bruker instrument at 250 MHz (Karlsruhe, Germany). Optical rotations were measured on a DIP 370 Jasco instrument (Tokyo, Japan). TLC was performed on Si gel 60 F254 plates (Merck Darmstadt, Germany). Hplc analyses were performed on a reverse phase C18 column eluting with MeOH/citrate buffer 5 mM, pH 4, using a linear gradient from 40 to 80% MeOH in 10 min. Solvents
© 1998 Wiley-Liss, Inc.
(analytical grade) were used without further purification. rac-Indoprofen (±)−1 was purchased from Sigma (St. Louis, MO). Lipases from Aspergillus niger, Mucor javanicus, and Rhizopus javanicus were obtained from Amano International Co., Ltd. (Nagoya, Japan). Lipase from Candida cylindracea and porcine pancreatic lipase were purchased from Sigma. Immobilised lipases from Candida antarctica (Novozymt 435) and Mucor miehei (Lipozymet IM) were obtained from Novo Nordisk (Bagsvaerd, Danmark). Determination of the Enantiomeric Excesses of Optically Active Compounds The enantiomeric excess (e.e.) of ester (R)-2 was measured by 1H NMR in the presence of europium tris[3(heptafluoropropylhydroxymethylene)camphorate], (Eu(hfc)3). The -OMe signals for the two enantiomers were significantly different in their chemical shift values to allow a reliable determination of the peaks integration. The enantiomeric purity of acid (S)-1 was determined after conversion into its methyl ester by a conventional procedure (MeI/K2CO3 in acetone). E.e.’s of ethyl-, propyl-, butyl-, and hexyl esters were calculated from c = ees/(ees + eep) according to reference.12 Preliminary Enzymatic Esterification of rac-1 Esterification with different lipases. Lipase (100 mg)
of choice was added to CH3CN (2 ml) containing rac-1 (20 mg, 0.07 mmol) and methanol (12 µl, 4 eqv). The suspension was incubated at 45°C under shaking (300 rpm) for 2 days and then analysed by tlc. Esterification in different solvents. Rac-1 (50 mg, 0.17 mmol) was dissolved in the solvent of choice (5 ml) containing methanol (28 µl, 4 eqv) and of Novozymt 435 (500
*Correspondence to: Dr. Giovanni Nicolosi, Istituto CNR Studio Sostanze Naturali di Interesse Alimentare e Chimico-Farmaceutico, Via del Santuario 110, I-95028 Valverde CT, Italy. E-mail:
[email protected] Received for publication 4 August 1997; Accepted 6 November 1997
322
MORRONE ET AL.
TABLE 1. Esterification of rac-1 with CH3OH catalysed by Candida antarctica lipase in different solvents*
Entry 1 2 3 4 5 6 7
Solvent
Time (d)
Conv. (%)a
e.e. Ester (%)b
Ec
e.e. Acid (%)d
CH3CN tert-Amyl alcohol THF Dioxane Dioxane-Toluene 1:1 Dioxane-t-BME 1:1 Dioxane-CH2Cl2 1:1
12 12 12 4 2 2 2
33 20 14 19 33 22 20
82 72 85 84 89 87 91
15 7 14 14 26 18 26
40 18 14 20 44 25 23
*Substrate, 10 mg/ml; enzyme, 100 mg/ml; T = 45°C, 300 rpm. a Determined by HPLC. b Determined by 1H NMR analysis in the presence of Eu(hfc)3. c Enantiomeric ratio, calculated according Chen et al.12 d Determined by 1H NMR analysis in the presence of Eu(hfc)3 after conversion into its methyl ester.
mg). The suspension was incubated at 45°C and shaken (300 rpm) until an appropriate conversion of the substrate was reached. The reaction was then stopped by filtering off the enzyme, the solvent removed in vacuo, and the residue partitioned between sat. aq. NaHCO3 and tert-buthyl methyl ether (t-BME). The organic phase was taken to dryness to afford ester (−)-(R)-2. Esterification with different alcohols. Esterification of rac-1 with ethanol, propanol, butanol, and hexanol was carried out as above. The reaction mixture was partitioned between sat. aq. NaHCO3 and t-BME and the aqueous phase acidified with H2SO4 and extracted with t-BME to yield (+)-(S)-1. Esterification in the presence of different water amounts. Four samples of a solution of rac-1 (50 mg,
0.17 mmol) in dioxane-toluene 1:1 v/v (5 ml), previously dried by shaking overnight with molecular sieves (4 Å), were each added with methanol (28 µl, 4 eqv) and water (0, 5, 10, or 20 µl, respectively). Novozymt 435 (500 mg), dried 12 h in vacuo, was then added to each sample and the mixtures were incubated at 45°C under shaking (300 rpm). After 60 h the reaction conversion and enantiomeric purity of products were determined as above. Preparative Resolution of rac-1
Novozymt 435 (10 g) and methanol (0.58 ml, 4 eqv) were added to a solution of (±)-1 (1 g, 3.6 mmol) in dioxane-toluene 1:1 vol/vol (100 ml). The suspension was shaken (300 rpm) at 45°C until the conversion reached 40% (60 h), then the reaction was stopped, filtering off the enzyme. Solvents were removed in vacuo and the residue was partitioned between sat. aq. NaHCO3 and t-BME. The organic phase was dried over Na2SO4 and then evaporated in vacuo to afford methyl ester (−)-(R)-2 (400 mg, 38%, e.e. 85%); [a]D = −27.1 (c 1.5, CHCl3). 1 H NMR d 1.52 (3H, d, J = 7.15 Hz), 3.67 (3H, s), 3.74
(1H, q, J = 7.15 1Hz), 4.85 (2H, s), 7.36 (2H, d, J = 8.60 Hz), 7.47–7.65 (3H, m), 7.82 (2H, d, J = 8.60 Hz), 7.92 (1H, d, J = 8.20 Hz). Anal.: Calc. for C18H17NO3: C, 73.20; H, 5.80. Found: C, 73.32; H, 5.87%. The aqueous phase was acidified with H2SO4 to yield a precipitate of the unreacted acid, (+)-(S)-1 (590 mg, yield 59%, e.e. 62%). The enantiomerically enriched acid (500 mg) was again subjected to enzyme-catalysed esterification until the conversion reached 25% (27 h). Workup of the reaction mixture afforded methyl ester (−)-(R)-2 (120 mg, 23%, e.e. 61%) and acid (+)-(S)-1 (350 mg, 70%, e.e. >95%); [a]D = + 48.1 (c = 0.05, DMSO) [reference13 + 48.0 (c = 0.05, DMSO)]. RESULTS AND DISCUSSION
At the inception of this work rac-indoprofen, (±)-1, was subjected to esterification in CH3CN with one of the following lipases: lipase from Aspergillus niger, Mucor javanicus, Rhizopus javanicus, Candida cylindracea, Candida antarctica (immobilised, Novozymt 435), Mucor miehi (immobilised, Lipozymet IM), and from porcine pancreas. Methanol was used as nucleophile, since in previous studies it worked successfully in the enzyme-catalysed esterification of flurbiprofen8 and suprofen.9 Among the lipases screened, only Novozymt 435 was active and was used in further experiments. Due to the sparing solubility of (±)-1 in most common solvents, only dioxane, tert-amyl alcohol, and THF were considered in the study of the effect of the solvent variation on the reaction rate. The data summarised in Table 1 (entries 1–4) show that esterification generally proceeds very slowly, the best results being obtained with dioxane (20% conversion after 4 days). Since these results were inadequate, we resorted to mixtures (1:1 vol/vol) of dioxane and a solvent of high lipophilicity (in which indoprofen is insoluble), since it is known that hydrophobic solvents increase the activity of lipase.14 Actually, the three co-solvents all gave good re-
323
BIOCATALYTIC RESOLUTION OF RAC-INDOPROFEN
TABLE 2. Influence of the nucleophile in the esterification of (±)-indoprofen catalysed by Candida antarctica lipase* Alcohol
Time (h)
Conv. (%)a
e.e. Ester (%)b
Ec
e.e. Acid (%)d
32 48 48 48 48
25 22 14 9 9
90 85 90 89 90
26 16 22 19 21
30 24 15 9 9
Methanol Ethanol n-Propanol n-Butanol n-Hexanol
*Substrate, 10 mg/ml; enzyme, 100 mg/ml; T = 45°C, 300 rpm. a Determined by HPLC. b Calculated from c=ees/(ees+eep). c Enantiomeric ratio calculated according to Chen et al.12 d Determined by 1H NMR analysis in the presence of Eu(hfc)3 after conversion into its methyl ester.
sults (Table 1, entries 5–7), the best one, in terms of both reaction rate and enantioselectivity, being obtained with toluene. After 48 h in dioxane-toluene, ester (−)-2 is obtained with 33% chemical yield and 89% enantiomeric excess (enantiomeric ratio,12 E = 26). The 1:1 dioxane-toluene mixture was then used in the examination of the effect of the chain length of the alcohol. As is apparent from the data in Table 2, the reaction rate decreases with the increasing length of the alcohol chain, whereas the enantioselectivity is not significantly affected. Moreover, when the nucleophile is methanol, variation of its concentration (methanol/indoprofen ratio from 1:1 to 20:1) has no appreciable influence on the reaction rate. Finally, we investigated the influence of water, which is known to affect both conversion and enantiomeric purity of the product in lipase-mediated esterfication.14,15 The data in Table 3 indicate that in the case at hand addition of water has an adverse effect on both reaction rate and enantioselectivity. The E value, which is 22 for the ‘‘dry’’ system, drops to 11 after addition of 0.4% water to the solvent system. In principle, since Novozymt 435 has R stereopreference, it seems possible to obtain the eutomer S with high enantiomeric excess, carrying the reaction to conversion levels beyond 50%. However, the reversibility of esterification causes a worsening of the enantiomeric purity with increased incubation time. In fact, the gradual accumulation of water produced in the forward reaction will give rise to reverse catalysis and the enantiomer esterefied faster will be the one suffering hydrolysis more easily, thus resulting in a decreased enantiomeric excess of the product. TABLE 3. Effect of added water on the lipase-catalysed esterification of (±)-indoprofen with methanol* % Added water 0.0 0.1 0.2 0.4
Conv. (%)a
e.e. Ester (%)b
Ec
e.e. Acid (%)d
41 32 28 20
85 80 82 80
22 13 14 11
60 38 31 20
*Substrate, 10 mg/ml; enzyme, 100 mg/ml; T = 45°C, 300 rpm. a Determined by HPLC. b Determined by 1H NMR analysis in the presence of Eu(hfc)3. c Enantiomeric ratio, calculated according Chen et al.12 d Determined by 1H NMR analysis in the presence of Eu(hfc)3 after conversion into its methyl ester.
Anyway, the desired enantiomer can be obtained with satisfying e.e. quenching the reaction near 40% conversion and subjecting the enriched S-acid to a second esterification cycle. In a gram-scale run, rac-indoprofen was incubated with methanol in the presence of Novozymt 435 using dioxanetoluene as solvent system. The reaction was stopped after 60 h to give (R)-indoprofen methyl ester with 85% e.e. and an unreacted acid enriched in the S-enantiomer with 62% e.e. This was again subjected to enzyme-catalysed esterification in the same conditions to afford, after 27 h, a methyl ester with low e.e. (60%) and the pharmacologically important (S)-acid with high chemical (70%) and optical (>95%) yields. The methyl ester isolated in the two steps can be transformed into rac-indoprofen by hydrolysis using the basic condition. The recovered enzyme, dried in vacuo at room temperature for 12 h is reused with unchanged enantioselectivity and a very small loss of activity (10% after 12 cycles). In conclusion lipase from Candida antarctica catalyses the esterification of rac-indoprofen in organic medium to furnish enantiopure (S)-indoprofen of pharmaceutical interest. The R enantioform, recovered as ester with low e.e., can be subjected to alkaline hydrolysis, affording recyclable rac-indoprofen. ACKNOWLEDGMENTS
The authors thank Novo Nordisk for a generous gift of Novozymt 435 and Lipozymet IM. LITERATURE CITED 1. (a) Caldwell, J. ‘‘Chiral pharmacology’’ and the regulation of new drugs. Chem. Indust. (London) 176–179, 1995. (b) Cannarsa, M.J. Single enantiomer drugs: new strategies and directions. Chem. Indust. (London) 374–378, 1996. 2. Shen, T.Y. Perspectives in nonsteroidal anti-inflammatory agents. Angew. Chem., Int. Ed. Eng. 11:460–472, 1972. 3. Brune, K., Beck, W.S., Geisslinger, G., Menzel-Soglowek, S., Peskar, B.M., Peskar, B.A. Aspirin-like drugs may block pain independently of prostaglandin synthesis inhibition. Experientia 47:257–261, 1991. 4. (a) Gu, Q-M., Chen, C-S., Sih, C.J. A facile enzymatic resolution process for the preparation of (+)-S-2-(6-methoxy-2-naphthyl)propionic acid (naproxen). Tetrahedron Lett. 27:1763–1766, 1986. (b) Hernaiz, M.J., Sanchez-Montero, J., Sinisterra, V.J. Comparison of the enzymic activity of commercial and semipurified lipase of Candida cylindracea in the hydrolysis of the esters of (R,S)-2-arylpropionic acids. Tetrahedron 50:10749–10760, 1994.
324
MORRONE ET AL.
5. Palomer, A., Cabre, M., Ginesta, J., Mauleon, D., Carganico, G. Resolution of rac-ketoprofen esters by enzymatic reactions in organic media. Chirality 5:320–328, 1993. 6. Mustranta, M. Use of lipases in the resolution of racemic ibuprofen. Appl. Microbiol. Biotechnol. 38:61–66, 1992. 7. Tsai, S-W., Wei, H-J. Enantioselective esterification of racemic naproxen by lipases in organic solvent. Enzyme Microb. Technol. 16: 328–333, 1994. 8. Morrone, R., Nicolosi, G., Patti, A., Piattelli, M. Resolution of racemic flurbiprofen by lipase-mediated esterification in organic solvent. Tetrahedon: Asymmetry 6:1773–1778, 1995. 9. Mertoli, P., Nicolosi, G., Patti, A., Piattelli, M. Convenient lipaseassisted preparation of both enantiomers of suprofen, a non-steroidal anti-inflammatory drug. Chirality 8:377–380, 1996. 10. Lapicque, F., Muller, N., Payan, E., Dubois, N., Netter, P. Protein binding and stereoselectivity of nonsteroidal anti-inflammatory drugs. Clin. Pharmacokinet. 25:115–125, 1993. 11. Tosolini, G.P., Moro, E., Forgione, A., Ranghieri, M., Mandelli, V. GLC
determination of plasma levels of enantiomers of a-[4-(1-oxo-2isoindolinyl)phenyl]propionic acid. J. Pharm. Sci. 63:1072–1076, 1974. 12. Chen, C.-H., Fujimoto, Y., Girdauskas, G., Sih, C.J. Quantitative analyses of biochemical kinetic resolution of enantiomers. J. Am. Chem. Soc. 104:7294–7299, 1982. 13. Kametani, T., Kigasawa, K., Hiiragi, M., Ishimaru, H., Haga, S., Shirayama, K. Studies on the syntheses of analgesic. Part 50 (1). Synthesis of optically active 2-[4-(1-oxo-2-isoindolinyl)phenyl] propanoic Acid. [Studies on the syntheses of heterocyclic compounds. Part 742 (2)]. J. Heterocyclic Chem. 15:369–375, 1978. 14. (a) Parida, S., Dordick, J.S. Substrate structure and solvent hydrophobicity control lipase catalysis and enantioselectivity in organic media. J. Am. Chem. Soc. 113:2253–2259, 1991. (b) Nakamura, K., Kinoshita, M., Ohno, A. Effect of solvent on lipase-catalyzed transesterification in organic media. Tetrahedron 50:4681–4690, 1994. 15. Kirchner, G., Scollar, M.P., Klibanov, A.M. Resolution of racemic mixtures via lipase catalysis in organic solvent. J. Am. Chem. Soc. 107: 7072–7076, 1985.
