Increased racemate resolution of propranolol esters by lipase immobilized catalysis

R. Ávila*; R. Ruiz; D. Amaro-González*; O. Díaz; J. A. González and A. J. Núñez

Chemical Department, Center of Pharmaceutical Chemistry, PO Box 16042, Havana City, Cuba.
rizette@cqf.co.cu; daniel.amaro@cqf.co.cu

Abstract — A screening of six different immobilized lipases (C. rugosa, C. antartica, P fluorescens, M. miehie, R. niveus) was done to determinate the best for the stereospecific hydrolysis of RS-O-BP. An experimental design (2³) where the enzymatic activity was a function of temperature, buffer concentration and solvent was selected. P fluores-cens showed the best results for S(-) isomer, with an increase of the stereoespecificity ratio of 1.5, using tetrahydrofuran as solvent, and 15mM buffer phosphate concentration at 37 ^oC. C. rugosa show-ed the best results for R(+) isomer, with an increase of the stereoespecificity ratio of 1.7, using acetone as solvent, and 25 mM buffer phosphate concentra-tion at 15 ^oC. The enantiomeric excess of S(-) propranolol with P fluorescens was 87%, higher than those reported elsewhere.

Keyword — Candida Rugosa. Pseudomona Fluorescens. Enantioselectivity. Biotransformation. Lipase. Propranolol.

I. INTRODUCTION

Propranolol (1-isopropilamino-3-(1-naphtoxy)-2-pro-panol) is a known beta-adrenergic blocking agent commonly used for the treatment of arterial hyper-tension (AHT) and some cardiovascular disorders (Barret, 1985). However, its use has been overpassed by other beta-blocking agents, Barret (1985) found the side-effects, mainly in asthma patients. Muscle contraction effects have been the cause of bronchoconstriction observed in those patients by Goodman and Gillman (1985). As propranolol is a mixture of two stereoisomers, S(-) and separately on AHT and bronchoconstriction. Barret (1985) demons-trated that such side effect could be attributed to the R(+) propranolol isomer, and there-fore proposed to develop enriched S(-) propranolol isomer galenic formulation of the highest possible "optical" purity to avoid the observed adverse effect of commercially available racemic mixtures of propranolol.

Yost and Holtzman (1979) proposed a highly attractive method of propranolol racemate resolution at preparative scale. Propranolol stereoisomers were separated by multiple-step recrystallization of its di-(p-toluyl) tartaric acid salts.

Silber and Riegelman, 1980; Matsuo and Ohno, 1985; Terao et al., 1988 and Bevinakatti and Banerji, 1991, have proposed the preparation of S(-)-proprano-lol through immobilized catalysis using mainly several lipases as stereoespecific catalysts. The uses of immo-bilized enzymes versus other conventional chemical catalysis are being increasingly preferred by the phar-maceutical industry because of its higher selectivity and yield, with a less environmental impact.

Bevinakatti and Banerji (1991) have published examples about stereoselective esterification or hydro-lysis of several drug racemates for its resolution using lipases. Chiral discrimination is a long-standing pro-blem in the development, use and action of pharma-ceutical agents. Since many drugs are chiral com-pounds and interact with a chiral receptor in the body, only one of the enantiomers will show the optimal therapeutic action. To avoid side effects of the unwanted enantiomer, the strongly regulated pharma-ceutical industry increasingly demands drugs contai-ning only the biologically active enantiomer. How-ever, some synthetically produced chiral substances are still being sold in a racemic mixture. This is the case of Propranolol. Currently the commercially available preparation is a racemic mixture, in which only the S(-)-enantiomer has ß-adrenergic blocking activity (Barrett, 1985; Howe and Shanks, 1966; Potter, 1967; Rahn et al., 1974; Walle et al., 1984). Barrett (1985) indicated that its side effects on asthma patiens have been attributed to the R(+) isomer, which produces bronchoconstricction. Therefore, the resolu-tion of commercial propranolol racemate in order to obtain the bioactive S(-) isomer at its highest purity is a matter of concern for a better treatment of AHT and asthma patients.

Several research groups (Silber and Riegelman, 1980; Hermansson and von Bahr, 1980; Hermansson, 1982; Thompson, et al., 1982; Matsuo and Ohno, 1985; Terao, et al., 1988; Bevinakatti and Banerji, 1991) have reported the preparation of (S)-propranolol by lipase catalysed reactions. However, the length of the procedure and the low overall yields precludes the industrial application of these methods.

Lipase catalysed reactions are superior to conven-tional chemical methods owing to mild reaction conditions, high catalytic efficiency and the inherent selectivity of natural catalysts which results in much purer products, according to Dordick (1991).

The ability of enzymes to catalyse reactions in a selective manner has been used for many years by organic chemists to obtain enantiomerically pure compounds. Lipases (triacylglycerol hydrolases (EC 3.1.1.3) form a large group of enzymes frequently used in fine chemistry catalysis. They are stable in nonpolar solvents and they have the remarkable ability of assuming a variety of conformations to accommo-date substrates of varying sizes and stereochemical complexities as Arroyo and Sinisterra (1994) reported. In organic media, there are examples related to enantioselective esterification or hydrolysis of racemic mixtures of drugs and agrochemical esters catalyzed by native or immobilized (Bevinakatti and Banerji, 1991) lipase but studies deal with the influence of structural variables on the enantioselectivity of immobilized enzymes are usually not reported in the organic chemistry literature.

The use of immobilized enzymes for such objective seems to be attractive in terms of the cost-benefit relationship and therefore a study conducted using several enzymes at different experimental conditions in order to establish a stereospecific procedure which may lead to obtain the propranolol S(-) isomer at the highest possible purity. The aim of this work has been the development of an enzymatic procedure to obtain the pure isomer from a propranolol racemic mixture and to establish the best conditions to achieve the biggest activity and enzymatic specificity.

II. METHODS

The RS-propranolol (RS-P) and the pure isomer were acquired from SIGMA. The screening were done with commercial lipases immobilized from Candida rugosa (CR), Candida antartica (CA), Pseudomonas fluores-cens (PF), Mucor mihei (MM), Rhizopus niveus (RN). All reagents for synthesis were from Fluka and all the solvents were of analytical grade. ¹H and ¹³C NMR spectra: Bruker Avance 400 instrument with SiMe₄ as internal standard, in CDCL₃; δ in ppm. HPLC-Mass Spectroscopy (LC-Mass): Agilent Technologies 1100 instrument LC/MSD Trap, equipped with ZORBAX Rapid resolution SB-C₁₈, (2.1 x 30 mm, 3.5μ) column; samples were dissolved in chloroform. HPLC (analytical and preparative): Agilent Technologies 1100 instrument equipped with UV detector, quaternary pump, thermostatized automatic injector and column compartment and Agilent ChemStation for LC systems; reverse phase column ChromSpher C18 (250 x 4,6 mm); flow 1 mL/min; mobile phase: Acetonitrile:H₂O:TEA; 60:40:0,14 (v/v/v).

A. Synthesis of RS-O-butyryl propranolol (RS-O-BP).

The RS-O-BP was synthetized follows the classical methodology of Gatterman (1961), in brief: RS-propranolol-HCl (1g; 3.8 mmol), was refluxed with dichloromethane (30 ml) and butyryl chloride (0.39 ml; 3.8 mmol) was added drop by drop. After two hours, the reaction mixture was washed successively with equal volumes of saturated aqueous sodium bicarbonate and brine. The organic layer was dried over anhydrous sodium sulphate and evaporated to dryness under reduced pressure to afford RS-O-BP. Yield: 85%; ¹H NMR (CDCl₃) δ (ppm): 0.97(3H,t), 1.22(3H,d), 1.31(3H,d), 1.67(2H,m), 2.33(2H,m), 3.40(1H,q), 3.83(1H,q), 4.12(1H,m), 4.29(2H,m), 5.59(1H,m), 6.87-8.21(aromatics). ¹³C NMR (CDCl₃) δ (ppm): 13.59, 18.33, 21.07, 21.80, 36.25, 41.33, 48.42, 68.51, 71.56, 104.82, 120.48, 121.89, 125.14, 125.52, 125.82, 126.32, 127.39, 134.42, 154.29, 173.80. ESI-MS m / z 352.4 [(M + Na)⁺].

B. Procedure for lipase immobilization.

2ml of octil sepharose (volume of the packed support) was added to 8 U of lipase suspended in 2 mL of sodium phosphate tampon 25mM pH=7 at 4^oC. A blanc suspension was also prepared adding 1ml of sepharose 4BCL. After 24h the immobilization process concluded filtering the suspension and washing the gel with abundant distilled water (Bastida et al., 1998).

C. Enzymatic hydrolysis and effect of different parameters on its initial rate.

The starting conditions were: To a solution of RS-O-BP 0.5mM was added 20 % of organic solvent. For all the experiments the final reaction volume was 5 mL. The temperature of the mixture was adjusted to the selected value and 6.5U of immobilized lipase were added (the same procedure was used for pure isomers). The effect of different parameters on the initial rate was studied following the progress of the reaction with automatic titration (pH-stat) using a 25mM dissolution of NaOH and withdrawing a 100 µl aliquot each 5 min. and diluting it with 100 µl of the elution buffer. Samples of 20 µl were injected in the chromatograph using as mobile phase acetonitrile: H₂O: trietylamine (60: 40: 0,14; v/v/v) (Ávila-González et al., 1997). The area peak corresponding to RS-O-BP was measured and converted into concentration by using a calibration curve that was constructed injecting known amounts of RS-O-BP in the same conditions. RS-O-BP concentration was plotted against time to obtain a curve of substrate consumption.

The initial rate was obtained calculating the slope of the linear part of this curve and was expressed as the decrease in the substrate concentration (µM) per minute.

C. Experimental Design

To optimize the hydrolysis a central composed design 2³ was carried out; with three experiments in the central plane (Himmelblau, 1970)

As independent variables were studied: partition coefficient octanol/H₂O (Log P) ((-1.1)-(-0.49)); Temperature (T) (4-37 ⁰C); Ionic Strength (IS) (10-50 mM). The enzymatic activity (EA) in PF and CR enzymes was chosen as dependent variable.

III. RESULTS AND DISCUSSIONS

As it was said previously lipases catalyze the enantio- and regioselective hydrolysis and synthesis of a broad range of natural and synthetic esters, (Santaniello et al., 1993). Khmelnitsky and Rich. (1999) signed that some of the lipase-catalysis applications are carried out in organic solvents, which due to their toxicity must be eliminated from the final product. Thus, enzymatic procedures for the synthesis of chiral drugs in semi aqueous systems would suppose a valuable alternative to the traditional methods.

A screening with several immobilized lipases was done following the starting conditions of hydrolysis above mentioned.

Table 1 show that the enzymes with greater activity and stereospecific ratio were those from CR and PF.

A later study showed the influence of pH on the activity and stereospecificity of both these enzymes. Two pH values were tested (7 and 9) for further work, because the EA at pH values, outside this range, decreased as Chin-Shih (1994) reported (Table 2).

A neutral pH value for the optimization studies of reaction parameters was chosen because greater stereo-specificity was obtained. The results show that the enzymatic hydrolysis occurs faster at pH 9 than that at pH 7, but the enantiomeric ratio is lower. Probably, the lipase recognize both isomers at the same time however when pH 7 is used the enantiomeric ratio is higher for both of them. This behaviour could be explained by the capability of the lipase to recognize the compounds by its ionization grade. When the substrate is no-ionized it is better recognized for the enzyme than when it is ionized, this increases the enzyme affinity and also the stereospecificity. Similar results were obtained by Fernandez-Lorente et al. (2004).

Table 1: Hydrolysis of RS-O-BP with different lipases.

^a Enzymatic activity (EA) of RS-O-BP (U/mL gel).
^b Stereospecificity calculated as the relationship of R(+)/S(-).
* High values for S(-)

Table 2: Influence of pH on the EA.

^b Stereospecificity calculated as the relationship of R(+)/S(-).

Considering the results obtained at the starting conditions, the behavior of the EA was modeled as a function of the independent variables using a statistical polynomial model of third order of magnitude.

. (1)

The Student t-test and the adjustment of the model evaluated the significance of the coefficients by the approach of Fisher for a 95% of confidence.

The results indicated that the influence of parame-ters on the EA was Temp>Log P>IS. In case of the PF enzyme the model did not show the expected adjust-ment. A total of 7 models were evaluated. The next equation showed the best result for the lipase from PF.

(2)

The analysis of significance of the parameters offered similar results to those obtained with the other enzyme: Temp>Log P>IS.

For both enzymes it was proven that the levels of significance of the Log P and the ionic strength were similar (8.5-6.0 and 30.7-26.05), while the temperature was superior. It was also observed that the temperature influenced in a different way the two enzymes.

For the PF enzyme the increase of the temperature caused an increase of the activity, however, in the case of CR it was inverse, the biggest activities were observed at low temperatures.

The denaturalization processes caused in some lipases by the increment of temperature could explain such results (Dordick, 1989).

With the purpose of determining the best conditions for RS-O-BP hydrolysis, it was used a software (MATLAB V 4.2) to search the global maximum of the system. The results show the conditions that guarantee the maximum stereospecificity in the reaction system (Table 3).

The experiments were carried out and they were compared with the calculated data. The following results were obtained (Table 4). The EA of the immobilized PF enzyme in the reaction of hydrolysis of the RS-O-BP was of 15.1 U/mL gel. With pure isomers, the activity with the R(+) isomer was 1.7 U/mL gel and for the S(-) isomer was 13.9 U/mL gel. The stereospecificity was 0.12.

For CR enzyme the experiment was not carried out to the exact conditions that were described due to the Log P maximum was obtained at –0.1 (Activity = 9.4), value that does not correspond to any solvent, so the nearest data-number was chosen, in this case acetone with –0.23. Under these new conditions the theoretical model was repeated and the value of 9.2 U/mL gel was obtained which do not have significative differences from the initial value.

Table 3: Theoretical results of the EA obtained from the optimization model.

The experiment under these news conditions show-ed (Table 4) that the CR enzyme has an activity of 8.9 U/mL gel. With pure isomers, the activity with the R(+) isomer was 10.7 U/mL gel and for the S(-) isomer was 1.5 U/mL gel, the stereospecificity was 7.1.

Table 4: Comparation of the experimental results with the calculated values from optimization model.

(*) Theoretical results.
^b Stereospecificity calculated as the relationship of R(+)/S(-).

Fig. 1. Progress curve of the reaction with PF in the optimal conditions.

As it can be appreciated, CR enzyme model describes with more reproducibility the theoretical and practical results that PF enzyme. The results showed that the stereospecificity increased for both enzymes compared to the initial values. The enantiomeric excess of (S)-propranolol (ee_s) was determined at the end point of the reaction (0.058mM; 40% of substrate conversion) (Fig. 1). Under these conditions values of ee_s of 87% and E= 47 (Chen et al., 1982) were obtained which are higher than that reported by Matsuo and Ohno (1985) for the enzymatic preparation of (S)-propranolol.

IV. CONCLUSIONS

The best results for the enzymatic hydrolysis of RS-O-BP were obtained with PF and CR. The better conditions were 37 ^oC, tetrahydrofuran and buffer phosphate 15mM for PF and 15 ^oC, acetone and buffer phosphate 25 mM for CR.

The immobilized enzyme CR is stereospecific for the hydrolysis of the R-O-BP and the PF for the S-O-BP. The values of stereospecificity increased for both enzymes compared to the values in the initial conditions. Under these conditions an ee 87% and E=47 were obtained.

ACKNOWLEDGEMENTS
This work was supported in part by a research grant from Institute of Catalysis and Petroleochemistry of Madrid, Spain. Professor J. M. Guisán for his recommendation and help.

REFERENCES
1. Arroyo, M. and J.V. Sinisterra, "High Enantioselec-tive Esterification of 2-Arylpropionic Acids Catalyzed by Immobilized Lipase from Candida antarctica: A Mechanistic Approach", J. Org. Chem., 59, 4410-4417 (1994).         [ Links ]
2. Ávila-González R., Díaz O., Ruiz R., González J.A. "Determinación de propranolol por cromatografía líquida de alta resolución", R. CENIC, Ciencias Químicas, 29, 147-149 (1997).         [ Links ]
3. Barrett, A.M., "Cardiobeta-adrenoceptor blokade: the guest for selectivity", J Pharmacol, 2, 95-108 (1985).         [ Links ]
4. Bastida, A., P. Sabuquillo, P. Armisen, R. Fernández-Lafuente, J. Huguet and J.M. Guisán, "A single step purification, inmobilization, and hyperactivation of lipases via interfacial adsorption on strongly hydrophobic supports", Biotechnology and bioengineering, 58, 486-493 (1998).         [ Links ]
5. Bevinakatti, H.S. and A.A. Banerji, "Practical Chemoenzymatic synthesis of both enantiomers of propranolol", J Org Chem, 56, 5372-5375 (1991).         [ Links ]
6. Chen, C.S., Y. Fujimoto, G. Girdaukas and C.J. Sih, "Quantitative analyses of biochemical kinetic resolutions of enantiomers", J Am Chem Soc 104, 7294-7299 (1982).         [ Links ]
7. Chin-Shih Ch., Wakefield, R.I. Patent US 5,334,534 (1994).         [ Links ]
8. Dordick, J.S., "Principles and applications of nonaqueous enzymology". In Blanch HW, Clark DS, editors. Applied biocatalysis, 1. New York: Marcel Dekker: 1-52 (1991).         [ Links ]
9. Dordick, J.S., "Enzymatic catalysis in monophasic organic solvents", Enzyme Microb. Technol., 11, 194-211 (1989).         [ Links ]
10. Fernández-Lorente, G., J.M. Palomo, C. Mateo, J.M. Guisán and R. Fernández-Lafuente, "Resolution of paroxetine precursor using different lipases influence of the reaction conditions on the enantioselectivity of lipases", Enzyme and Microbial Technology, 34, 264-269 (2004).         [ Links ]
11. Goodman, L.S. and A. Gillman, The Pharmacological basis of therapeutics, Fourth Edition, MacMillan Company, New York. (1985).         [ Links ]
12. Gatterman, L., Practices of Organic Chemistry, 2da Edición Española, 159-162 (1961).         [ Links ]
13. Hermansson, J. and C. von Bahr, "Simultaneous determination of d- and l-Propranolol in human plasma by high-performance liquid chromatography", J. Chromatogr., 221, 109-117 (1980).         [ Links ]
14. Hermansson, J., "Separation and quantification of (R)- and (S)-Propranolol as their diastereomeric derivatives in human plasma by reversed-phase ion-pair chromatography", Acta Pharm Suec, 19, 11-24 (1982).         [ Links ]
15. Himmelblau, D.M., Process Analysis by Statistical Methods, John Wiley & sons. New York (1970).         [ Links ]
16. Howe, R. and R.G. Shanks, "Optical isomers of propranolol", Nature, 210, 1336-1368 (1966).         [ Links ]
17. Khmelnitsky, Y.L. and J.O. Rich, "Biocatalysis in non organic solvents", Curr Opin Chem Biol, 3, 47-53 (1999).         [ Links ]
18. Matsuo, N. and N. Ohno, "Preparation of optically active 1-acetoxy-2-aryloxypropionitriles and its application to a facile synthesis of (S)-(-)-Propranolol", Tetrahedron Lett, 26, 5533-5534 (1985).         [ Links ]
19. Potter, L.T., "A radiometric microassay of acetylcholinesterase", J Pharmacol Exp Ther, 156, 500-506 (1967).         [ Links ]
20. Rahn, K.H., A. Hawlina, F. Kersting and G. Planz, "Studies on the antihypertensive action of the optical isomers of propranolol in man", Naunyn Schmiedebergs Arch Pharmacol, 286, 319-323 (1974).         [ Links ]
21. Santaniello, E., P. Ferraboschi and P. Grisenti, "Lipase-catalyzed transesterification in organic solvents: applications to the preparation of enantiomerically pure compounds", Enzyme Microb Technol, 15, 367-382 (1993).         [ Links ]
22. Silber, B. and S. Riegelman, "Stereospecific assay for (-)- and (+)- propranolol in human and dog plasma", J Pharmacol Exp Ther, 215, 643-648 (1980).         [ Links ]
23. Terao, Y., M. Murata, K. Achiwa, T. Nishino, M. Akamatsu and M. Kamimura, "Highly efficient lipase-catalysed asymmetric synthesis of chiral glicerol derivatives leading to practical synthesis of S-propranolol", Tetrahedron Lett, 29, 5173-5176 (1988).         [ Links ]
24. Thompson, J.A., J.L. Holtzman, M. Tsuru and C.L. Lerman, "Procedure for the chiral derivatization and chromatographyc resolution of R-(+)- and S-(-)- propranolol", J Chromatogr, 238, 470-475 (1982).         [ Links ]
25. Walle, T., U.K. Walle, M.J. Wilson, T.C. Fagan and T.E. Gaffney, "Stereoselective ring oxidation of propranolol in man", Br J Clin Pharmacol, 18, 741-748 (1984).         [ Links ]
26. Yost, Y. and J.L. Holtzman, "Resolution of (+/-) Propranolol", J Pharm Sci, 16, 1181-1182 (1979).         [ Links ]