Asymmetric Routes to Chiral Secondary Alcohols

Chiral secondary alcohols are important intermediates and starting materials for the pharmaceutical industry. A wide variety of methodologies have been developed for preparing these synthons in high enantiomeric excess, including resolution of racemates using enzymes such as lipases and asymmetric reduction or hydrogenation of prochiral ketones. A decision on which technology to use must be made on a case-by-case basis and should take into account several factors, including the efficiency of the technology for a particular substrate, access to the catalysts, catalyst cost, and the sensitivity of the substrates to the proposed reaction conditions. Access to a broad portfolio of technologies is beneficial to allow the best solution for a particular product. This article describes several applications of these technologies for synthesizing secondary alcohol products at laboratory scale to multikilogram manufacture. The work was performed by Chirotech Technology, Custom Pharmaceutical Services (CPS–Chirotech), a wholly owned subsidiary of Dr. Reddy's Laboratories (Hyderabad, Andhra Pradesh, India).

Alcohol dehydrogenase reduction of ketones

The use of alcohol dehydrogenase (ADH) enzymes for synthesizing single-isomer secondary alcohols has significantly increased during the past five years. These enzymes catalyze the asymmetric reduction of a prochiral ketone and require a nicotinamide cofactor to provide the hydrogen that allows the reduction to occur. These cofactors are costly and therefore must be recycled. Several methods for recycling the cofactors are available, which include using glucose dehydrogenase (GDH), formate dehydrogenase (FDH), and isopropyl alcohol (IPA) (see Figure 1). Stable variants of GDH are available at large scale and reasonable cost although IPA is the most convenient and lowest-cost approach if IPA is a substrate for the ADH of interest.

Figure 1: Methods of cofactor regeneration in alcohol dehydrogenase (ADH)-catalyzed ketone reduction. NADPH is the reduced form of NADP+. NADP+ is nicotinamide adenine dinucleotide phosphate. NADH is the reduced form of NAD+. NAD+ is nicotinamide adenine dinucleotide. GDH is glucose dehydrogenase. FDH is formate dehydrogenase. (FIGURE IS COURTESY OF THE AUTHOR)

CPS–Chirotech has developed considerable expertise in using ADH enzymes to make secondary alcohols. It has a collection of more than 250 isolated in-house and commercially available enzymes, both (S)- and (R)-selective, arrayed in 96 well plates, to allow for efficient screening. Hits with these enzymes facilitate a fast response and allow scale-up to kilogram quantities of product in weeks. For products where a more efficient enzyme is needed, CPS–Chirotech has a broad collection of thousands of wild type (WT) organisms, which originate from diverse sources and include a large number of marine organisms. There are two choices with hits from this collection—either scale-up using the WT organism as whole cells or clone and overexpress the ADH of interest. Cloning the enzyme generally leads to a lower cost catalyst, especially when expressed using an efficient expression system. CPS–Chirotech uses a platform for protein production based on Pseudomonas fluorescens (1). This expression system uses a well characterized and safe strain of P. fluorescens (i.e., MB101) that is capable of high cell-density fermentations (in excess of 100 g/L of dry cells) without oxygen enrichment. P. fluorescens can be cultivated using a medium of simple and defined mineral salts, supplemented with an inorganic nitrogen source such as ammonia and a carbon source such as glucose. Annotated published genome sequences also provide a useful source of new ADHs. Synthetic gene synthesis is routinely offered at reasonable prices (i.e., $0.50/base pair) and can be codon-optimized for a nominated expression system.

Styrene oxide synthons

Chiral amino alcohol moieties (ArCH(O)CH₂N) account for a large proportion of secondary alcohol synthons found in clinical drug candidates. For example, this category accommodates the adrenoreceptor modulators used for treating asthma and related respiratory disorders, and these drug compounds are typically made by coupling a chlorohydrin (ArCH(OH)CH₂Cl) or styrene oxide with an amine component. In one example, the authors needed to develop a synthesis of (R)-3-chlorostyrene oxide on a multikilogram scale. (R)-3-Chlorostyrene oxide is a potential synthon for several clinical candidates, including Solabegron (3'-[2-[2(R)-(3-chlorophenyl)-2-hydroxyethylamino]ethyl amino]biphenyl-3-carboxylic acid hydrochloride) and SR-58611 (N-[7-(ethoxycarbonylmethoxy)-1,2,3,4-tetrahydronaphth-2(S)-yl]-2(R)-(3-chlorophenyl)-2-hydroxyethanamine hydrochloride).

There are four main methods for preparing chiral styrene oxides: asymmetric expoxidation, two approaches based on kinetic resolution, and asymmetric reduction (see Figure 2). The kinetic resolution approaches (see Figures 2b and 2c) are likely to be the least favored because of the 50% theoretical yield of the resolution step and the extra step that is required beyond the corresponding asymmetric approach. Asymmetric epoxidation of a styrene (see Figure 2a) is the most direct, but biological approaches to this reaction suffer from both low yields and volumetric productivity, and the chemocatalytic approaches that may be applied are not presently feasible on commerical scale. The most attractive option is asymmetric reduction of a phenacyl halide followed by ring closure (see Figure 2d). Several chemocatalysts and biocatalysts are available for the reduction, and the final step appears facile. For (R)-3-chlorostyrene oxide, the authors decided to focus on the biocatalytic approach due to a simpler patent landscape as well as past work suggesting that catalytic asymmetric hydrogenation of the chloroketones is difficult to accomplish with both high selectivity and catalyst productivity.

Figure 2: Different approaches to the preparation of (R)-styrene oxides. (FIGURE IS COURTESY OF THE AUTHOR)

Screening of 3-chlorophenacyl chloride against CPS–Chirotech's collection of ADH enzymes revealed several hits, the best of which were investigated further. One enzyme that was available at scale gave a highly enantioselective bioreduction of 3-chlorophenacyl chloride with a conversion of > 90%. This reaction was optimized to proceed at 100 g/L substrate concentration to give 2-chloro-1-(3-chlorophenyl)ethanol of > 99.5% enantiomeric excess (ee) and > 90% yield. Cofactor recycling was achieved using glucose dehydrogenase, thereby allowing for the use of a concentration of 0.25 mM NADP⁺ in the reaction. The halohydrin product was used directly in the next step and cyclized using a base to give (R)-3-chlorostyrene oxide of > 99% ee. This material was purified by distillation to give a final product of > 99% purity. This procedure is highly amenable to large-scale manufacture and was readily scaled up using standard equipment to several 100-kg scale.

Asymmetric hydrogenation of ketones

Catalytic asymmetric hydrogenation of prochiral ketones is another efficient method to prepare secondary alcohols. Two efficient technologies have been developed in recent years. The first is pressure hydrogenation by which diphosphine ruthenium diamine complexes catalyze the hydrogenation of simple ketones with gaseous hydrogen (2). The precatalysts require base activation, with the resulting catalysts being among the most highly active systems developed. The second approach is catalytic transfer hydrogenation, by which monosulfonated diamino ruthenium arene complexes use hydrogen donors such as propan-2-ol and formic acid as hydrogen sources (3). Reaction conditions are mildly acidic, thus allowing the hydrogenation of base-sensitive ketones. Transfer hydrogenation does not require pressure, but catalyst loadings are higher. In both cases, either isomer of the alcohol can be prepared depending on the isomer of the catalyst used. CPS–Chirotech has access to both technologies via a license from the Japan Science and Technology Corporation (Tokyo). This relationship allows CPS–ChiroTech to develop processes for complementary substrates.

1-Aryl alcohols

1-Aryl alcohol synthons are often used as intermediates in the production of active pharmaceutical ingredients. Access to the prochiral ketone starting material is straightforward, and many are commercially available at low cost, thereby making asymmetric reduction of the ketone an attractive method to synthesize these intermediates. Asymmetric hydrogenation using catalysts derived from PhanePhos (i.e., (R) or (S)-4,12-bis(diphenylphosphino)-[2.2]-paracyclophane) is particularly efficient for these substrates in terms of enantioselectivity, conversion, and the amount of catalyst required (low to modest hydrogen pressure is required, and a ratio of molar substrate to catalyst (S/C ratio) as high as 2.4 million has been demonstrated). CPS–Chirotech has ready access to these catalysts at commerical scale, and therefore was the authors' method of choice when preparing 1-aryl alcohols.

(S)-1-(4-Fluorophenyl)ethanol was required at multikilogram scale (4). [((S)-xylyl-PhanePhos)Ru((R,R)-DPEN)Cl₂] (where DPEN is 1,2-diphenylethane-1,2-diamine) precatalyst was shown to proceed at a S/C ratio of 100,000 giving alcohol of 99% ee (see Figure 3). Wiped-film distillation of the ketone prior to reaction is required to achieve these low catalyst loadings. The chiral alcohol is obtained as a colorless oil of 99% purity and 99% ee in 93% yield with no metal contamination from the catalyst.

Figure 3: Asymmetric reduction approach to (S)-1-(4-fluorophenyl)ethanol. S/C is the molar substrate to catalyst ratio. iPrOH is propan-2-ol, and t-BuOK is potassium tert-butoxide. GC is gas chromatography, and ee is enantiomeric excess. (FIGURE IS COURTESY OF THE AUTHOR)

1-Aryl-2-imidazol-1-yl ethanols

Two imidazolyl acetophenone substrates were examined for reduction via catalytic asymmetric hydrogenation (see Figure 4) (5). Pressure hydrogenation was initially examined using diphosphine ruthenium diamine complexes; however, no significant conversion was observed under various conditions. When standard conditions for catalytic asymmetric transfer hydrogenation using [(R,R)TsDPENRu (cymene)Cl] as the catalyst precursor were applied, both ketones were reduced in high conversions and ee. Reaction conditions were optimized so that (S)-1-phenyl-2-imidazol-1-yl ethanol could be produced in 97% ee in quantitative yield and (S)-1-(2,4-dichlorophenyl)-2-imidazol-1-yl-ethanol in 91% ee in 98% yield.

Figure 4: Asymmetric transfer hydrogenation of imidazolyl acetophenones. SC is the molar substrate to catalyst ratio, and ee is enantiomeric excess. (FIGURE IS COURTESY OF THE AUTHOR)

Lipase-catalyzed resolution

Lipase-catalyzed production of chiral secondary alcohols via transesterification of the racemic alcohol or hydrolysis of the ester are well-known methods frequently run at large scale. This route has the disadvantage that the yield of the desired isomer is a maximum of 50%; however, in many cases this approach is still the method of choice. Lipases are industrially suitable enzymes due to their stability, tolerance of organic solvents and high temperatures, low price, and availability. In addition, the off-isomer can sometimes be racemized and recycled to increase efficiency.

Propylene glycol alkyl or aryl ethers

Chiral 1,2-propanediols are useful synthons for preparing cardiovascular drugs, antiviral drugs, and enantiomerically pure crown ethers (6). Propylene oxide-derived racemic glycol ethers are readily available at large scale and low cost. A viable approach for synthesizing single isomers is through lipase-catalyzed resolution of these racemates (7). A screen of hydrolases for their ability to catalyze enantioselective hydrolysis of an aqueous solution of a racemic mixture of test glycol alkyl ether acetates revealed Candida antarctica lipase B (CAL-B) to be the most enantioselective enzyme for all substrates screened (all E > 50, where E is the enantioselectivity constant). Moreover, immobilized CAL-B was the best catalyst for transesterification-based resolution in organic solvent with various acyl donors, for example, ethyl acetate and vinyl acetate. For large-scale manufacture, the transesterification reaction using an immobilized lipase (see Figure 5) was favored for several reasons. No additional ester-formation step is necessary. The reaction can be run with much higher substrate loadings. The work-up is easier as it is possible to directly recover products by distillation following bioresolution, and the enzyme can be easily recycled without loss of activity either in batch mode or in a continuous process with enzyme packed in a column. As an example, the resolution of 1-n-propoxy-2-propanol was conducted in batch mode at a 1-L scale giving (S)-alcohol in 99% ee; more than 20 consistent recycles of the enzyme were obtained. These reactions have been scaled to several 100-kg scale in an economic process made possible via the short-reaction sequence, effective recycling of enzyme, the low-cost starting material, and straightforward isolation of the reaction products.

Figure 5: Lipase-catalyzed production of propylene glycol ethers. CAL-B is Candida antarctica lipase B, and ee is enantiomeric excess. (FIGURE IS COURTESY OF THE AUTHOR)

Conclusion

Chiral secondary alcohols are important intermediates and starting materials for pharmaceutical compounds. Examples cited in this article describe biocatalytic asymmetric reduction and catalytic hydrogenation of prochiral ketones as well as the resolution of racemates using enzymes such as lipases as efficient methods to make various aryl and alkyl secondary alcohols.

Christopher J. Cobley, PhD, is head of chemocatalysis, and Karen E. Holt-Tiffin,* PhD, is head of biocatalysis, both at Chirotech Technology, Dr. Reddy's Custom Pharmaceutical Services, 162 Cambridge Science Park, Cambridge, CB4 0GH, UK, tel. +44 (0) 1223 728010, keholt.tiffin@drreddys.com.

*To whom all correspondence should be addressed.

References

1. C. H. Squires et al., BioProcess Intl., 2 (11) 54–59, 2004.

2. R. Noyori and T. Ohkuma, Angew. Chem. Int. Ed. 40 (1), 40–73, 2001.

3. Noyori and S. Hashiguchi, Acc. Chem. Res. 30 (2), 97–102, 1997.

4. D. Chaplin et al., Org, Process Res. Dev. 7 (1), 89–94 (2003).

5. I.C. Lennon and J.A. Ramsden, Org. Process Res. Dev. 9 (1), 110–112 (2005).

6. B.H. Hoff et al., Tetrahedron: Asymmetry, 7 (11), 3181–3186 (1996).

7. S.M. Resnick et al., "Enzymatic Resolution of Propylene Glycol Alkyl (or Aryl) Ethers and Ether Acetates," WO 2003083126, March 2003.