**Biocatalytic Synthesis of Chiral Alcohols and Amino Acids for Development of Pharmaceuticals**

#### **Ramesh N. Patel**

**Abstract:** Chirality is a key factor in the safety and efficacy of many drug products and thus the production of single enantiomers of drug intermediates and drugs has become increasingly important in the pharmaceutical industry. There has been an increasing awareness of the enormous potential of microorganisms and enzymes derived there from for the transformation of synthetic chemicals with high chemo-, regio- and enatioselectivities. In this article, biocatalytic processes are described for the synthesis of chiral alcohols and unntural aminoacids for pharmaceuticals.

Reprinted from *Biomolecules*. Cite as: Patel, R.N. Biocatalytic Synthesis of Chiral Alcohols and Amino Acids for Development of Pharmaceuticals. *Biomolecules* **2013**, *3*, 741-777.

#### **1. Introduction**

For preparation of drugs and their intermediates, the synthesis of single enantiomers has become increasingly important in the pharmaceutical industry [1]. Single enantiomers can be produced by either by chemical or biocatalytic routes. The advantages of biocatalysis over chemical synthesis are that enzyme-catalyzed reactions are often highly enantioselective and regioselective. They can be carried out under mild conditions at ambient temperature and atmospheric pressure, thus avoiding the use of more extreme reaction conditions which could cause problems with isomerization, racemization, epimerization, and rearrangement of compound. Microbial cells and wide variety and class of enzymes derived there from can be used for chiral synthesis. Enzymes can be immobilized and reused for many cycles. In addition, enzymes can be over expressed to make biocatalytic processes economically efficient, and enzymes with modified activity can be tailor-made. Directed evolution of biocatalysts can lead to increased enzyme activity, selectivity and stability [2–15]. A number of review articles [16–31] have been published on the use of enzymes in organic synthesis. This chapter provides some examples of the use of enzymes for the synthesis chiral alcohols, unnatural amino acids, and amines for synthesis of phamaceuticals.

#### **2. Enzymatic Preparation of Chiral Alcohols**

#### *2.1. Hydroxy Buspirone (Antianxiety Drug): Enzymatic Preparation of 6-Hydroxybuspirone*

Buspirone (Buspar, **1**, Figure 1) is a drug used for treatment of anxiety and depression that is thought to produce its effects by binding to the serotonin 5HT1A receptor [32–34]. Mainly as a result of hydroxylation reactions, it is extensively converted to various metabolites and blood concentrations return to low levels a few hours after dosing [35]. A major metabolite, 6-hydroxybuspirone **2**, produced by the action of liver cytochrome P450 CYP3A4, is present at much higher concentrations in human blood than buspirone itself. For development of 6-hydroxybuspirone as a potential antianxiety drug, preparation and testing of the two enantiomers as well as the racemate was of interest. An enantioselective microbial reduction process was developed for reduction of 6-oxobuspirone **3**, to either (*R*)- and (*S*)-6-hydroxybuspirone **2**. About 150 microbial cultures were screened for the enantioselective reduction of **3**. *Rhizopus stolonifer* SC 13898, *Neurospora crassa*  SC 13816, *Mucor racemosus* SC 16198, and *Pseudomonas putida* SC 13817 gave >50% reaction yields and >95% e.e.s of (*S*)-6-hydroxybuspirone. The yeast strains *Hansenula polymorpha* SC 13845 and *Candida maltosa* SC 16112 gave (*R*)-6-hydroxybuspirone **2** in >60% reaction yield and >97% e.e. [36]. The NADP-dependent (*R*)-reductase (RHBR) from *Hansenula polymorpha* SC 13845 was purified to homogeneity, its *N*-terminal and internal sequences were determined and cloned and expressed in *Escherichia coli*. To regenerate the cofactor NADPH required for reduction we have also cloned and expressed the glucose-6-phosphate dehydrogenase gene from *Saccharomyces cerevisiae* in *Escherichia coli*. Recombinant cultures expressing (*R*)-reductase (RHBR) catalyzed the reduction of 6-ketobuspirone to (*R*)-6-hydroxybuspirone in 99% yield and 99.9% e.e. at 50 g/L substrate input [37].

**Figure 1.** Hydroxy buspirone (antianxiety drug): Enzymatic preparation of 6-hydroxybuspirone.

The NAD-dependent (*S*)-reductase (SHBR) from *Pseudomonas putida* SC 16269 was also purified to homogeneity, its *N*-terminal and internal sequences were determined and cloned and expressed in *Escherichia coli*. To regenerate the cofactor NADH required for reduction we have also cloned and expressed the NAD+ dependent formate dehydrogenase gene from *Pichia pastoris* in *Escherichia coli*. Recombinant *Escherichia coli* expressing (*S*)-reductase was used to catalyze the reduction of 6-ketobuspirone to (*S*)-6-hydroxybuspirone, in >98% yield and >99.8% e.e. at 50 g/L substrate input [37].

*2.2. Cholesterol Lowering Agents: Enzymatic Preparation of (3S,5R)-Dihydroxy-6-(Benzyloxy) Hexanoic Acid, Ethyl Ester 4*

Compound **4** (Figure 2) is a key chiral intermediate required for the chemical synthesis of compound **5**, Arotvastatin **6**, and Rosuvastatin all are anticholesterol drugs which acts by inhibition of HMG CoA reductase [38–42].

**Figure 2.** Cholesterol lowering agents: Enzymatic preparation of (3*S*,5*R*)-dihydroxy-6- (benzyloxy) hexanoic acid, ethyl ester.

The enantioselective reduction of a diketone 3,5-dioxo-6-(benzyloxy) hexanoic acid, ethyl ester **7** to (3*R*,5*S*)-dihydroxy-6-(benzyloxy) hexanoic acid, ethyl ester **4** (Figure 2) was demonstrated by cell suspensions of *Acinetobacter calcoaceticus* SC 13876 [39,43]. On reduction of **7** by cell suspensions, the *syn*-**4** and *anti*-**8** dihydroxy esters were formed in the ratio of about 87:13, 83:17, 76:24 after 24 h at 2, 5 and 10 g/L of substrate input, respectively. There was no significant peak due to a monohydroxy ester. Chiral HPLC determined that the desired (3*R*,5*S*)-**4** was the major product with 99.4% e.e. Almost complete (>95%) conversion of the ethyl diketoester **7** to dihydroxy ester **4** in 24 h was seen up to a substrate concentration of 10 g/L and cell concentration of 200 g/L [39,43].

A mixture of ethyl 3-keto-5-hydroxy **9** (major) and 5-keto-3-hydroxy **10** (minor) was obtained from partial microbial reduction of ketoester **7**. These two mixtures were subjected to microbial reduction by *Acinetobacter* sp. SC13874 cells for 6 h (incomplete reduction). The reduction provided the dihydroxy esters with the isomeric composition. The results indicated that the second reduction of the monohydroxy compound by SC13874 cells was quite enantiospecific. Reduction of the 3-keto-5-hydroxy **9** provided predominantly the (3*R)-*hydroxy, while reduction of the 3-hydroxy-5-keto ester **10** provided predominantly the (5*S)*-hydroxy compound [43].

Cell extracts of A. *calcoaceticus* SC 13876 in the presence of NAD+ , glucose, and glucose dehydrogenase reduced **7** to the corresponding monohydroxy compounds [3-hydroxy-5-oxo-6-(benzyloxy) hexanoic acid ethyl ester **9** and 5-hydroxy-3-oxo-6-(benzyloxy) hexanoic acid ethyl ester **10**]. Both **9** and **10** were further reduced to the (3*R*,5*S*)-dihydroxy compound **4** in 92% yield and 99% e.e. by cell extracts. (3*R*,5*S*)-**4** was converted to **11**, a key chiral intermediate for the synthesis of compound **5** and Atorvastatin **6**. Three different ketoreductases were purified to homogeneity from cell extracts of *A. calcoaceticus* SC 13876 and their biochemical properties were compared. Reductase I only catalyzes the reduction of ethyl diketoester **7** to its monohydroxy products whereas reductase II catalyzes the formation of dihydroxy products from monohydroxy substrates. A third reductase (III) was identified which catalyzes the reduction of diketoester **7** to *syn*-(3*R*,5*S*)-dihydroxy ester **4** [44], which now has been cloned and expressed in *E. coli* [44] and the reduction of diketoester 7 to *syn*-(3*R*,5*S*)-dihydroxy ester **4** was demonstrated by recombinant enzyme at 50 g/L substrate input with 10 g/L cell suspensions.

#### *2.3. Atorvastatin: Enzymatic Preparation of (R)-4-Cyano-3-Hydroxybutyrate*

An enzymatic process for the preparation of ethyl (*R*)-4-cyano-3-hydroxybutyric acid **12** (Figure 3), a key intermediate for the synthesis of Atorvastatin **6** was developed by Codexis [45]. In this process, first the enzymatic synthesis of ethyl (*S*)-4-chloro-3-hydroxybutyric acid derivatives **13** was carried out by ketoreductase-catalyzed conversion of 4-chloro-3-ketobutyric acid derivatives **14** [46]. The genes encoding halohydrin dehydrogenase from *Agrobacterium tumefaciens*, ketoreductase from *Candida magnoliae*, glucose dehydrogenase from *Bacillus subtilis* and formate dehydrogenase from *Candida boidinii* were separately cloned into *Escherichia coli* BL21. Each enzyme was then produced by fermentation, isolated and characterized. Then ethyl (*R*)-4-cyano-3-hydroxybutyrate **12** (Figure 3) was prepared from ethyl 4-chloroacetoacetate **14** by the following procedure: Ethyl 4-chloroacetoacetate 14 was incubated at pH 7.0 with ketoreductase, glucose dehydrogenase, and NADP<sup>+</sup> for 40 h to produce ethyl (*S*)-4-chloro-3-hydroxybutyrate **13** which was extracted with ethyl acetate, dried, filtered and concentrated to yield ~97% pure ester. The dried ethyl (*S*)-4-chloro-3-hydroxybutyrate **13** was dissolved in phosphate buffer and mixed with halohydrin dehalogenase and sodium cyanide at pH 8.0. After 57 h, essentially pure ethyl (*R*)-4-cyano-3-hydroxybutyrate **12**, an intermediate used in HMG-CoA reductase inhibitors syntheses, was recovered [45].

#### **Figure 3.** Atorvastatin: Enzymatic preparation of (*R*)-4-cyano-3-hydroxybutyrate.

#### *2.4. Preparation of (S)-4-Chloro-3-Hydroxybutanoic Acid Methyl Ester*

**(***S*)-4-chloro-3-hydroxybutanoic acid methyl ester **15** (Figure 4) is a key chiral intermediate in the total chemical synthesis of **16**, an inhibitor of HMG CoA reductase [46,47]. The reduction of 4-chloro-3-oxobutanoic acid methyl ester **17** to (*S*)-4-chloro-3-hydroxybutanoic acid methyl ester **15** (Figure 4) by cell suspensions of *Geotrichum candidum* SC 5469. In the biotransformation process, a reaction yield of 95% and e.e. of 96% were obtained for (*S*)-**15** by glucose-, acetate- or glycerol-grown cells (10% w/v) of *G. candidum* SC 5469 at 10 g/L substrate input. The e.e. of (*S*)-**15** was increased to 98% by heat-treatment of cell-suspensions (55 °C for 30 min) prior to conducting the bioreduction of **17** [48].

**Figure 4.** Chloesterol lowering agents: Preparation of (*S*)-4-chloro-3-hydroxybutanoic acid methyl ester.

In an alternate approach, the asymmetric reduction of ethyl 4-chloroacetoacetate to (*S*)-4-chloro-3-hydroxybutonoate was demonstrated by a secondary alcohol dehydrogenase (PfODH) from *Pichia finlandica*. The gene encoding PfODH was cloned from *P. finlandica* and over expressed in *Escherichia coli*. Formate dehydrogenase was used to regenerate the cofactor NADH required for this reaction. Using recombinant *E. coli* coexpressing both PfODH and formate dehydrogenase from *Mycobacetrium* sp. produced to (*S*)-4-chloro-3-hydroxybutonoate in 98.5% yield and 99% e.e. at 32 g/L substrate input [49].

## *2.5. Rhinovirus Protease Inhibitor: Enzymatic Process for the Preparation of (R)-3-(4-Fluorophenyl)-2-Hydroxy Propionic Acid*

(*R*)-3-(4-fluorophenyl)-2-hydroxy propionic acid **18** (Figure 5) is a building block for the synthesis of AG7088, a rhinovirus protease inhibitor **19** [50,51]. The preparation of **18** using a biocatalytic reduction of **20** in a membrane reactor [52]. A continuous enzymatic process for an efficient synthesis of (*R*)-3-(4-fluorophenyl)-2-hydroxy propionic acid at multikilogram scale with a high space-time yield (560 g/L/day) using a membrane reactor. The product was generated in excellent enantiomeric excess (e.e. >99.9%) and good overall yield (68%–72%).

Using this method, an overall quantity of 23 kg of **18** was prepared. The key step was an aqueous enzymatic reduction using D-lactate dehydrogenase (D-LDH) and formate dehydrogenase (FDH). Mechanistically, the keto acid salt **20** is stereoselectively reduced to the corresponding *R*-hydroxy acid **18** in the presence of D-lactate dehydrogenase by NADH. The cofactor itself is oxidized to NAD<sup>+</sup> in the process. Subsequently, in the presence of formate dehydrogenase, NAD<sup>+</sup> is reduced back to NADH by ammonium formate, which was oxidized to CO2 and NH3. In this fashion the expensive cofactor NAD<sup>+</sup> is regenerated by FDH, and only a catalytic amount of NAD<sup>+</sup> was required [52].

#### *2.6. Enzymatic Preparation of Chiral Intermediates for Atazanavir*

Atazanavir **21** (Figure 6) is an acyclic aza-peptidomimetic, a potent HIV protease inhibitor [53,54] approved by the Food and Drug Administration for treatment of Auto Immune Diseases (AIDS). An enzymatic process was developed for the preparation of (1*S*,2*R*)-[3-chloro-2-hydroxy-1-(phenylmethyl) propyl]carbamic acid, 1,1-dimethylethyl ester **22**, a key chiral intermediate required for the total synthesis of the HIV protease inhibitor atazanavir. The diastereoselective reduction of (1*S*)-[3-chloro-2-oxo-1-(phenylmethyl)propyl] carbamic acid, 1,1-dimethylethyl ester **23** was carried out using *Rhodococcus*, *Brevibacterium*, and *Hansenula* strains to provide **22**. Three strains of *Rhodococcus* gave >90% yield with a diastereomeric purity of >98% and an e.e. of 99.4% [55]. An efficient single-stage fermentation-biotransformation process was developed for the reduction of ketone **23** with cells of *Rhodococcus erythropolis* SC 13845 to yield **22** in 95% with a diasteromeric purity of 98.2% and an e.e. of 99.4% at substrate input of 10 g/L. The reduction process was further improved by generating mutants and selection of desired mutant for conversion of **23** to (1*S*,2*R*)-**22** at substrate input of 60 g/L [56]. (1*S*,2*R*)-22 was converted to epoxide **24** and used in the synthesis of atazanavir. Chemical reduction of chloroketone **23** using NaBH4 produces the undesired chlorohydrin diastereomer [57].

**Figure 6.** Atazanavir (antiviral agent): Enzymatic reparation of (1*S*,2*R*)-[3-chloro-2-hydroxy-1-(phenylmethyl) propyl]-carbamic acid,1,1-dimethyl-ethyl ester.

#### *2.7. Enzymatic Reduction Process for Synthesis of Montelukast Intermediate*

The discovery of the biological activity of the slow reacting substance of anaphylaxis (SRS-A) and its relation to the leukotrienes (LTC4, LTD4, and LTE4) and asthma, the search for leukotriene antagonists has been intensive. As part of an ongoing program for the development of specific LTD4 antagonists for the treatment of asthma and other associated diseases at Merck have identified Montelukast **25** (Figure 7) as LTD4 antagonist [58–60].

Merck has described the synthetic route for the production of montelukast, using a stereoselective reduction of a ketone **26** to the (*S*)-alcohol **27** as the key step. The alcohol subsequently undergoes a Sn2 displacement with a thiol to give the *R*-configured final product [59,60]. The reduction of the ketone **26** to produce the chiral alcohol **27** requires stoichiometric amounts of the chiral reducing agent (í)-flchlorodiisopino campheylborane [(í)-DIP-chloride]. (í)-DIP-chloride is selective and avoids the side reactions but it is corrosive and moisture-sensitive, causing burns if it is allowed to contact the skin. The reaction must be carried out at í20 to í25 °C to achieve the best stereoselectivity. The quench and extractive work-up generate large volumes of waste solvent, due to the product's low solubility. The potential advantages of biocatalytic transformation of ketone to alcohol were recognized early on by researchers at Merck. However, only two microorganisms were identified as having activity on the bulky and hydrophobic substrate [61]. Due to several reasons, an enzyme-catalyzed process for reduction of the ketone **26** was developed by Codexis. A ketoreductase was developed by directed evolution by high throughput screens using a slurry of the ketone substrate and high isopropanol concentration. Beneficial mutations among the various improved mutants were recombined in each round, and new mutations were made guided by ProSAR. The productivity of the final enzyme was improved 2,000-fold and stability was also substantially increased [62].

**Figure 7.** Enzymatic reduction process for synthesis (*S*)-alcohol **27** for Montelukast intermediate.

The final process was carried out as a slurry-to-slurry reaction at 45 °C, with the sparingly soluble ketone **26** being converted to an almost equally insoluble alcohol **27** at a concentration of 100 g/L substrate in aqueous isopropanol and toluene. A reaction yield of 99.3% and enantiomeric excess of 99.9% was obtained for alcohol 27 [62].

#### *2.8. Anticancer Drug: Enzymatic Preparation of C-13 Paclitaxel Side-Chain Synthon*

Among the antimitotic agents, paclitaxel (taxol®) **28** (Figure 8), a complex, polycyclic diterpene, exhibits a unique mode of action on microtubule proteins responsible for the formation of the spindle during cell division. Various types of cancers have been treated with paclitaxel and it was approved for use by the FDA for treatment of ovarian cancer and metastatic breast cancer [63–65]. A key precursor for the paclitaxel semi-synthetic process is the chiral C-13 paclitaxel side-chain **29**. An enzymatic enantioselective microbial reduction of 2-keto-3-(N-benzoylamino)-3-phenyl propionic acid ethyl ester **30** to yield (2*R*,3*S*)-*N*-benzoyl-3-phenyl isoserine ethyl ester **29** was demonstrated using two strains of *Hansenula* [66]. Preparative-scale bioreduction of ketone **30** was demonstrated using cell suspensions of *Hansenula polymorpha* SC 13865 and *Hansenula fabianii* SC 13894 in independent experiments. In both batches, a reaction yield of >80% and e.e.s of >94% were obtained for (2*R*,3*S*)-**29**. In a single-stage process, cells of *H. fabianii* were grown in a 15-L fermentor for 48 h, then the bioreduction process was initiated by addition of 30 g of substrate and 250 g of glucose and continued for 72 h. A reaction yield of 88% with an e.e. of 95% was obtained for (2*R*,3*S*)-**29**.
