Synthesizing Chiral Drug Intermediates by Biocatalysis

Abstract

The chiral feature is a critical factor for the efficacy and safety of many therapeutic agents. At present, about 57% of marketed drugs are chiral drugs and about 99% of purified natural products are chiral compounds. There has been a tremendous potential of functional microorganisms and biocatalysts derived from them for the bioconversion of synthetic chemicals into drugs with high enantio-, chemo-, and regio-selectivities. Biocatalysis is becoming a key subassembly in the medicinal chemist’s toolbox. In fact, the intermediates of many important therapeutic agents such as sitagliptin, pregabalin, ragaglitazar, paclitaxel, epothilone, abacavir, atorvastatin, rosuvastatin, and omapatrilat have been successfully synthesized via biocatalysis. In this review, various biocatalytic systems that enable to synthesize these chiral drug intermediates are updated and discussed regarding their potential application in the pharmaceutical industry. Further development and increased utilization of biocatalysis for production of drugs with emphasis on green chemistry can be expected.

Introduction

The production of single enantiomers, which can be obtained by either biocatalytic or chemical synthesis of chiral intermediates, has become increasingly important in the development and use of pharmaceuticals and agrochemicals [56, 99, 100, 163, 209]. Biocatalysis is now keeping to maintain good growth momentum and is becoming a key constituent part in the toolbox for the chemists, and it has a place alongside chromatographic separations and chemocatalysis [183]. Biocatalysis systems (both whole-cell and isolated enzyme systems) are increasingly used as a synthetic route to produce complex molecules in industrial fields, such as synthesis of high value-added chemicals and drug intermediates [14, 18, 22, 28, 53, 55, 132, 176, 183, 232]. Further, the bioconversion field has gained more interest for the synthesis of chiral building blocks same as biologically active compounds [41], providing the development and employment of more economically and environmentally attractive processes. In contrast to the chemical synthesis, the biocatalysis shows the advantages of the reduction of environmental pollution and the high regio-, chemo-, and enantio-selectivity. And these biocatalyst-catalyzed reactions can be conducted at atmospheric pressure and ambient temperature to avoid the use of extreme conditions, which often could cause problems with racemization, isomerization, rearrangement of the compound, and epimerization. Therefore, biocatalysis possesses many appealing characteristics in the field of green chemistry particularly for the synthesis of chiral intermediates, which are subsequently used to produce food additives or enantiopure drugs [152]. And further development and employment of biocatalysis for high-productivity processes and green chemistry can also be expected [183].

Enzymes and functional microbes can be modified, evolved, redesigned, immobilized, and reused for many times. Overexpressed target enzymes and directed evolution of biocatalysts with ameliorative activity, stability, and selectivity can be tailor-made. Thus, biocatalysis makes the biotransformation process economically feasible and highly efficient. And the ways of designing biocatalysts, enzymes, and functional microbes will change the face of synthesis. Recent review articles about organic synthesis using various enzymes and biosystems [21, 48, 73, 101, 103, 115, 142, 154, 161, 164, 183, 187, 195, 201, 213, 214, 219, 229, 231, 272, 70, 218]have been reported, which summarized the main routes of producing single enantiomers of key drug intermediates by biocatalysis (Fig. 1), and some case studies and examples on the use of biocatalysis to synthesize biologically active compounds and pure building blocks in the development and utilization of pharmaceuticals.

Fig. 1.

The diagrammatic sketch of main methods for the synthesis of chiral drug intermediate by biocatalysis

Key Enzymes

Hydrolytic enzymes such as esterases, lipases, dehalogenases, proteases, nitrilases, acylases, epoxide hydrolases, amidases and decarboxylases, aminotransferases, dehydrogenases, aldolases, dioxygenases, and monooxygenases are the biocatalysts generally chosen for the biosynthesis of compounds with stereogenic centers [163, 215]. Some approaches have been executed, such as oxidative deaminization, enzymatic deracemization, reductive amination, and dynamic resolution, to achieve high e.e. and more than 50% yield through combination of chemo- and/or biocatalysts by a single biocatalysis or in sequential reactions. Dioxygenases have been successfully used in the production of chiral dihydrodiols through chemoenzymatic routes while monooxygenases were used in epoxidation, enantioselective hydroxylation, sulfoxidation, and Baeyer-Villiger reactions, and regioselective hydroxylation is utilized for synthesizing chiral compounds. Unlike the dioxygenases and monooxygenases, the decarboxylases and aldolases have been effectively used by acyloin condensation and aldol condensation reactions in the asymmetric synthesis progress. Cofactor regeneration is very economical and important for the cofactor-dependent progress due to the high cost of adding cofactors, such as NADH and NADPH [109, 234]. As the key enzymes, aminotransferases and dehydrogenases have been effectively used in biosynthesis of amino acids, amino alcohols, amines, and chiral alcohols with cofactors or cofactor regeneration. The hydrolytic enzymes have been used for synthesis of high value-added chiral compounds through the asymmetric synthesis of prochiral compounds or the kinetic resolution of racemic compounds.

Abundant access to various microbial cultures and enzymes applied in organic synthesis of key chiral intermediates and target drug products has been opened up through substrate engineering; medium engineering; protein engineering; biocatalyst engineering like directed evolution, random and site-directed mutagenesis of biocatalysts, biocatalytic cascade processes, cell-free synthetic biology; and reactor engineering and downstream processing like fermentation technology with high cell density for a long time in the past decade and future. The biocatalysis for biosynthesis or with chemosynthesis of chiral compounds looks promising in the future.

Functional Bacterial Strains

Under some circumstances, using whole microbial cells, natural functional bacteria or engineering bacteria, as biocatalysts is smarter rather than using isolated enzymes for cofactor regeneration due to instability of free enzyme outside the cell [29], and the multi-enzyme-catalyzed composite reactions and the whole-cell processes are cheaper to implement upstream for there is no need to isolate the enzyme. And whole cells should show prominent advantages in multistep conversion processes like side-chain cleavage of the β-sitosterol by using the whole cells of Mycobacterium sp. NRRL B-3805 [217]. It is just like redox enzymes which are frequently used, since the internal cellular metabolism is contributing to the regeneration of cofactors by this method. Whole cells (natural functional bacteria or engineering bacteria as “bags of enzymes”) are also employed when multi-enzyme-catalyzed step reactions are expected and designed for the formation of a multi-enzyme cluster. Compared with using isolated enzymes, using whole cells as the biocatalysts is economical and more attractive. Nowadays, 75% of routes for industrially executed redox bioconversions are using whole cells [92, 93]. Herein, we will describe some cases of the use of natural functional bacteria or engineering bacteria whole cells in the synthesis of chiral precursors of drugs.

Natural Functional Bacteria

Oxidative biocatalysts (oxidases, dehydrogenases, oxygenases, and peroxidases) possess many advantages in contrast to their chemical counterparts as the problems, such as uncontrollable and unpredictable final product structures, can be overcome for oxidation routes in traditional chemical catalyzing and the use of oxidative biocatalysts is economical [215]. Some representative cases implemented by oxidative biocatalysts are listed in Table S1. The 4-oxo-4-[3-(trifluoromethyl)-5,6-dihydro-[1, 2, 4]triazolo[4,3-a]pyrazin-7(8H)-yl]-1-(2,4,5-trifluorophenyl)butan-2-one (OTPP) was converted into the (S)-3-hydroxy-1-(3-(trifluoromethyl)-5,6-dihydro-[1, 2, 4]triazolo[4,3-a]pyrazine-7(8H)-yl)-4-(2,4,5-trifluorophenyl)butan-1-one ((S)-HTPP). A key intermediate for the preparation of sitagliptin employed to cure type 2 diabetes [128] was produced by using the Pseudomonas pseudoalcaligenes XW-40 (P. pseudoalcaligenes XW-40) with > 99.9% enantiomeric excess and > 99.5% yield [245] (entry 1, Table S1). And the Rhizopus microsporus var. rhizopodiformis ZJPH1308, a new fungus which showed the ability of catalysis OTPP to (S)-HTPP with > 99.9% e.e. value and 93.2% yield, was screened and identified [222] (entry 1, Table S1). The herbicide precursor, (R)-2-(4-hydroxyphenoxy)propanoic acid, was obtained by parahydroxylating the raw material (R)-2-phenoxypropionic acid as a representative case of the bio-oxidation-mediated hydroxylations designed by the BASF AG company [49] (entry 3, Table S1). A route by using whole cells was developed for the synthesis of the (S)-2-amino-6-hydroxyhexanoic acid, a key chiral intermediate for the preparation of an antihypertensive metalloendopeptidase inhibitor and a vasopeptidase inhibitor [159] (entry 4, Table S1). The (3S,4R,5S)-1,3,4,5-tetrahydroxy-6-(alkyl-amino)hexan-2-one, an important intermediate for preparing the oral α-glucosidase inhibitors which is used against retroviral infections or cure carbohydrate metabolism disorders, was obtained through catalyzed N-protected 1-amino-d-sorbitol by using the Gluconobacter oxydans’ whole cells in the Bayer AG company [126] (entry 5, Table S1). Dioxygenases and monooxygenases, which can introduce oxygen atom into the substrate, as the ideal biocatalysts to synthesize 2-quinoxalinecarboxylic acid which was used in the production of various biologically active compounds [253] (entry 6, Table S1)), sodium (3R,5R)-3,5-dihydroxy-7-((1S,2S,6S,8S,8aR)-6-hydroxy-2-methyl-8-((S)-2-methylbutanoyloxy)-1,2,6,7,8,8a-hexahydronaphthalen-1-yl)heptanoate) [156] (entry 7, Table S1), and cis,cis-muconic acid are used as important intermediates to synthesize a variety of agrochemicals and pharmaceuticals [264] (entry 8, Table S1).

The stereoselective reduction of ketones (also well-known as stereoselective reduction of a C=O bond into its corresponding alcohol) to form a stereogenic center (^*CH–OH) is an important process in the asymmetric synthesis because of its wonderful atom economy in the preparation of bioactive molecules and pharmaceuticals [93, 208]. And this route, the asymmetric hydrogenation of ketones by biocatalysis, can be regarded as an ideal greener strategy in contrast to the chemical catalysts. Here, some preparative examples and cases of bio-reductions catalyzed by whole cells are listed in Table S2. An important chiral precursor, (S)-1-(benzo[d][1, 3]dioxol-5-yl)propan-2-ol [2] (entry 1, Table S2), in the synthesis of benzodiazepines like Talampanel™ and a key intermediate, (1R,4S,6R)-6-hydroxybicyclo[2.2.2]octan-2-one [109] (entry 2, Table S2), for the production of antitumoral taxane derivatives were synthesized by using whole cells. A key alcohol intermediate for the synthesis of the atazanavir (an HIV endopeptidase inhibitor and an acyclic aza-peptidomimetic), the tert-butyl ((2S,3R)-4-chloro-3-hydroxy-1-phenylbutan-2-yl)carbamate, was obtained by using Rhodococcus erythropolis SC 13854 (R. erythropolis SC 13854) with 95% yield and 99% e.e. value [173] (entry 3, Table S2). The atazanavir was approved by the Food and Drug Administration to treat autoimmune diseases (AIDS) [13, 191]. The (1S,2R)-[3-chloro-2-hydroxy-1-(phenylmethyl)propyl]carbamic acid, 1,1-dimethylethyl ester, an important chiral intermediate for the preparation of the HIV protease inhibitor atazanavir, was synthesized by using Hansenula, Rhodococcus, and Brevibacterium strains with e.e. of 99.4% [173] (entry 4, Table S2). And this route was further improved by mutation and screening of desired mutant for the synthesis of the important intermediate for the preparation of the atazanavir [20]. The (1S)-[3-chloro-2-oxo-1-(phenylmethyl)propyl]carbamic acid, 1,1-dimethylethyl ester was catalyzed to its corresponding chiral alcohol, a key intermediate for chemical synthesis of a class of HIV protease inhibitors [7], by using the Mortierella ramanniana SC 13850 and Streptomyces nodosus SC 13149 (S. nodosus SC 13149) [170]. And 67% reaction yield and 99.9% e.e. value were gained by employing the cells of S. nodosus SC 13149 while 54% reaction yield and 99.9% e.e. value were obtained by employing the cells of M. ramanniana SC 13850 [170] (entry 5, Table S2). Chemical reduction of the chloroketone (1S,2R)-[3-chloro-2-hydroxy-1-(phenylmethyl)propyl]carbamic acid, 1,1-dimethylethyl ester often generates the undesired chlorohydrin diastereomer [257]. The key intermediate (S,E)-methyl 2-(2-(3-(3-(2-(7-chloroquinolin-2-yl)vinyl)phenyl)-3-hydroxypropyl)phenyl)-2-methyl-propanoate for the production of anti-asthma montelukast was implemented by using Microbacterium campoquemadoensis [205] (entry 6, Table S2). The Aureobasidium pullulans was used in stereoselective reduction of the a-keto ester ethyl 2-oxo-2-(1′,2′,3′,4′-tetrahydro-1′,1′,4′,4′-tetramethyl-6′-naphthalenyl)acetate, which was a key intermediate for product retinoic acid receptor modulators used as anticancer and dermatological drugs with 94% yield and 97% enantioselectivity [172] (entry 7, Table S2). The production of the (S)-(3-(cyclopentyloxy)-4-methoxyphenyl)(phenyl)methanol, an ideal precursor to synthesize the phosphodiesterase inhibitor which is selected to treat asthma, was implemented by the Merck company [30] (entry 8, Table S2). The cell suspensions of the Acinetobacter calcoaceticus SC 13876 (A. calcoaceticus SC 13876) were used for the synthesis of the (3S,5R)-dihydroxy-6-(benzyloxy) hexanoic acid ethyl ester, a key important intermediate for the production of cholesterol-lowering agents, such as rosuvastatin and atorvastatin, with e.e. value of 97% and reaction yield of 85% [145, 169] (entry 9, Table S2).

Chiral alcohols are widely used as the key drug intermediate for the synthesis of some important pharmaceuticals. The Aspergillus niger (A. niger) was employed to produce the (R,S)-3-(2-methoxyphenoxy)propane-1,2-diol, a β-blocker used for the treatment of certain arrhythmias and hypertension, with 90% e.e. value and 50% conversion rate [134] (entry 1, Table S3). Then, the (R,S)-3-(2-methoxyphenoxy)propane-1,2-diol was converted to (S)-moprolol through chemical synthesis with two steps [134]. The Geotrichum candidum SC 5469 (G. candidum SC 5469) was used for the synthesis of the (S)-4-chloro-3-hydroxybutanoic acid methyl ester, an important chiral intermediate for the preparation of a HMG-CoA reductase inhibitor [106, 244], with 95% reaction yield and 96% e.e. value (entry 2, Table S3). And the e.e. value of the (S)-4-chloro-3-hydroxybutanoic acid methyl ester increased to 98% by using cell suspensions with heat treatment [166, 174]. The Mortierella ramanniana ATCC 38191 (M. ramanniana ATCC 38191) and Pullularia pullulans ATCC 16623 (P. pullulans ATCC 16623) were screened from various microorganisms to introduce the enantioselective reduction of 1-(4-fluorophenyl)4-[4-(5-fluoro-2-pyrimidinyl)1-piperazinyl]-1butanone to (R)-1-(4-fluorophenyl)4-[4-(5-fluoro-2-pyrimidinyl)1-piperazinyl]-1-butanone, which was used for the development of a new class of antipsychotic pharmaceuticals [58, 59, 111, 110], with a 100% reaction yield and a 98.9% e.e. value [168] (entry 3, Table S3). The (S)-5-hydroxyhexanenitrile and ethyl-(S)-5-hydroxyhexanoate are key and ideal chiral intermediates in the preparation of anti-Alzheimer’s drugs [184]. And Pichia methanolica SC 16116 was employed to synthesize both of these chiral compounds by enantioselective reduction of 5-oxohexanenitrile and ethyl-5-oxohexanoate [151] (entry 4, Table S3). Furthermore, the Candida antarctica (C. antarctica) lipase was also employed in the preparation of anti-Alzheimer’s drugs, and a 42% reaction yield and a > 99% e.e. value were achieved [167] (entry 5, Table S3). Organisms from genus Baker’s yeast, Pichia, Candida, Rhodotorula, Sphingomonas, Hansenula, and Saccharomyces were used for the preparation of the key chiral intermediates, (S)-methyl 4-(2′-acetyl-5′-fluorophenyl)-butanol and (S)-1-(2′-bromo-4′-fluorophenyl)ethanol, for the production of several potential anti-Alzheimer’s drugs [199, 173] (entry 6, Table S3). It was approved by the FDA that paclitaxel is used for the treatment of metastatic breast cancer and ovarian cancer, and paclitaxel is also used for curing other various kinds of cancers [94, 117, 158]. The cell suspensions of Hansenula fabianii SC 13894 (H. fabianii SC 13894) and H. polymorpha SC 13865 were selected to synthesize the (2R,3S)-N-benzoyl-3-phenylisoserine ethyl ester, an important precursor for the semi-synthetic process of the paclitaxel with over 94% of e.e. value and over 80% of reaction yield [167] (entry 7, Table S3). The Rhodococcus globerulus SC 16305 (71.8% enantiomeric excess) and H. polymorpha SC 13824 (73.8% enantiomeric excess) were identified to synthesize the (S)-2-chloro-1-(3-chlorophenyl)ethanol which is required in the production of a IGF-1 receptor inhibitor (an anticancer program) [252] (entry 8, Table S3). The Rhodotorula piliminae ATCC 32762 (R. piliminae ATCC 32762) and P. delftensis MY 1569 were introduced in the synthesis of the key chiral alcohol for the chloroketone and keto ester, and the 97% yield of the (S)-alcohols was reached [125] (entry 10, Table S3). The Candida parapsilosis IFO 1396 was screened for the preparation of the (R)-4-chloro-3-hydroxybutanoate and (R)-1, 3-butanediol, which is of great significance in the synthesis of azetidinone derivatives utilized for the production of carbapenem antibiotics and penem [105] (entry 11, Table S3). The 4,5-dihydro-4-(4-methoxyphenyl)-6-(trifluoromethyl)-1H-1-benzazepin-2,3-dione was used for the generation of the 1,3,4,5-tetrahydro-3-hydroxy-4-(4-methoxyphenyl)-6-(trifluromethyl)-2H-1-benzazepin-2-one, a key intermediate for the synthesis of the diltiazem commonly used for the treatment of angina and hypertension [181], with 96% reaction yield and 99.8% e.e. value by using Nocardia salmonicolor SC 6310 (N. salmonicolor SC 6310) [165] (entry 12, Table S3).

Depression and anxiety as psychiatric disorders are major health problems worldwide. Though lots of treatments have been developed for these two orders, the reduction of side effect and the increment of efficacy are still required [104, 155]. The (R)-1-cyclopropylethylamine and (R)-sec-butylamine are important chiral intermediates used to synthesize the CRF-1 receptor antagonists which are taken for the novel pharmacological treatments for stress disorders, depression, and anxiety [66, 129, 155, 224]. And the deacylation or acylation reactions for the preparation of racemic amines were executed by using oxidases or lipases and proteases [71, 72, 183]. The enantioselective reduction of 6-oxobuspirone to either (S)- or (R)-6-hydroxybuspirone was established by screening about 150 microorganisms [67, 175]. And the 6-oxobuspironeto is a key material for the preparation of the buspirone which is used as a drug for curing depression and anxiety through binding the serotonin 5HT1A receptor [107, 144, 262]. Bacillus megaterium (B. megaterium SC 6394) isolated from several soil samples was used to synthesize the (R)-1-cyclopropylethylamine and (R)-sec-butylamine [85] (entry 1, Table S4). The epothilones, a novel kind of natural product cytotoxic compounds, were obtained by using myxobacterium Sorangium cellulosum; it showed ideal result in the trigger apoptosis progress [62, 11] (entry 3, Table S4). Using Amycolatopsis orientalis SC 15847 (A. orientalis SC 15847), a microbial hydroxylation route was executed for the preparation of epothilone F with epothilone B as the precursor (entry 4, Table S4). After that, the key gene, ferredoxin gene from A. orientalis SC 15847, was expressed in the Streptomyces rimosus for the synthesis of the epothilone F from epothilone B with 80% yield [8] (entry 5, Table S4).

Engineered Bacteria

Using genetic engineering to produce target engineering bacteria with specific function is an ideal route to equip more active biocatalysts. Using genetic engineering protocols, the E. coli cells containing the NADPH-dependent carbonyl reductase, which was discovered through genome mining, from Yarrowia lipolytica ACA-DC 50109 (Y. lipolytica ACA-DC 50109), were employed to catalyze the ethyl 4-chloro-3-oxobutanoate to ethyl (S)-4-chloro-3-hydroxybutanoate, a key intermediate for the preparation of statins (blockbuster drugs) with 90% yield and 99% e.e. value [259] (entry 1, Table S5). And the Escherichia coli cells co-expressing the NADPH-dependent carbonyl reductase, which was also found through genome mining, from Rhodococcus pyridinivorans and glucose dehydrogenase (GDH), were employed to produce ethyl (S)-4-chloro-3-hydroxybutanoate with 91% yield and 99% e.e. value [258] (entry 2, Table S5). The recombinant E. coli which contains a formate dehydrogenase (FDH) from Mycobacterium sp. and a carbonyl reductase from Pichia finlandica was used as a biocatalyst to synthesize a key chiral intermediate for anticholesterol drugs, (2R,3S)-ethyl 2-chloro-3-hydroxybutanoate, with 85% yield and 99% e.e. value [113] (entry 3, Table S5), and the FDH (a dehydrogenase for cofactor regeneration) and carbonyl reductase (a ketoreductase to reduce the target substrate) were inserted in the E. coli system in this case. The recombinant E. coli which contains an enoate reductase (also known as old yellow enzyme) from Candida macedoniensis was used for stereoselective reduction C=C bond to produce 6R-levodione, an important intermediate in the preparation of some flavors and carotenoids [114] (entry 4, Table S5). The (S)-methyl 4-chloro-3-hydroxybutanoate (entry 5, Table S5) and (S)-ethyl 4-chloro-3-hydroxybutanoate (entry 6, Table S5) were synthesized by bio-reduction to generate the lateral chain of atorvastatin, a blockbuster drug for the treatment of lowering blood cholesterol [159, 118]. Using whole cells as biocatalysts, the two enzymes, a modified form of phenylalanine dehydrogenase (PheDH) from Thermoactinomyces intermedius and an engineered FDH expressed in single recombinant E. coli, were used for the synthesis of the (S)-3-hydroxyadamantylglycine, a key chiral intermediate for the synthesis of saxagliptin, an inhibitor of dipeptidyl peptidase-4 and an antidiabetic drug (an antihyperglycemic agent) with 100% e.e. and 98% yield [196, 239, 83] (entry 7, Table S5). The glucose-6-phosphate dehydrogenase from Saccharomyces cerevisiae, NADPH-dependent R reductase from H. polymorpha SC 13845, NADH-dependent S reductase from Pseudomonas putida SC 16269, and FDH from Pichia pastoris were selected and expressed in E. coli, respectively. Then, these four recombinant E. coli strains were used for the conversion of the 6-ketobuspirone to (R)- or (S)-6-hydroxybuspirone, a key material for the preparation of the buspirone which is used as a drug for resistant anxiety and depression through binding the serotonin 5HT1A receptor [107, 144, 262], with > 99.9% e.e. value and > 98% yields [67] (entry 8, Table S5). A ketoreductase III from Acinetobacter sp. 13874 was identified, cloned in the E. coli expression system for the synthesis of (3S,5R)-dihydroxy-6-(benzyloxy) hexanoic acid, ethyl ester. And this reaction was executed by using the recombinant E. coli’s cells or extracts with 99% yield [68, 75] (entry 10, Table S2).

Using the E. coli system to overexpress the phenylalanine dehydrogenase (PheDH) from Sporosarcina sp., the (S)-2-amino-5-(1,3-dioxolan-2-yl)-pentanoic acid (one of the three building blocks for the synthesis of the Vanlev) was synthesized through reductive amination of the corresponding a-keto acid with 98% enantioselectivity and 91% yield. By using E. coli cells overexpressing PheDH from Sporosarcina sp., 197 kg of the products was obtained in three batches of 1600 L with 91% yield and 98% enantioselectivity [159] (entry 9, Table S5). The recombinant E. coli which contained a transaminase from B. megaterium SC 6394 was employed for the synthesis of the (R)-1-cyclopropylethylamine and (R)-sec-butylamine by the recombinant cells with 42% isolated yield and 99% e.e. value [85] (entry 2, Table S4). A ketoreductase III from Acinetobacter sp. 13874 was identified, cloned in the E. coli expression system for the synthesis of (3S,5R)-dihydroxy-6-(benzyloxy) hexanoic acid ethyl ester. And this reaction was executed by using the recombinant E. coli’s cells or extracts with 99% yield [68, 75] (entry 10, Table S2). And an extract of E. coli which has two recombinant enzymes, a glucose-6-phosphate dehydrogenase from Saccharomyces cerevisiae (S. cerevisiae) and a ketoreductase from H. polymorpha SC 13824, was used to produce (S)-2-chloro-1-(3-chloro-4-fluorophenyl)ethanol with 100% e.e. value and 89% yield [83] (entry 9, Table S3). The stereoselective reduction of C=N bond by amino acid dehydrogenases is a great exciting bio-reduction to prepare high value-added amino acids and their derivatives [159]. The leucine dehydrogenase (LeuDH) from Bacillus cereus and the FDH from Candida boidinii were expressed in one genetic circuit in the E. coli system to synthesize l-tert-leucine (L-Tle), a building block to prepare many drugs with different activities (antitumoral agents or HIV endopeptidase inhibitors), with high enantiopure form (> 99% e.e.) and yield (approximately 100%) [109] (entry 10, Table S5). Using recombinant E. coli to express LeuDH from Thermoactinomyces intermedius and FDH from P. pastoris, the L-Tle was synthesized with 99.5% stereoselectivity and 95% yield [124] (entry 11, Table S5). The ethyl (R)-4-cyano-3-hydroxybutyric acid, an ideal intermediate for the production of carnitine, was synthesized by using the glucose dehydrogenase from Bacillus subtilis, halohydrin dehydrogenase from Agrobacterium tumefaciens (A. tumefaciens), FDH from C. boidinii, and ketoreductase from Candida magnoliae. These four enzymes were separately cloned and expressed into E. coli BL21, and were subsequently isolated and characterized to synthesize the ethyl (R)-4-cyano-3-hydroxybutyric acid [38, 39] (entry 12, Table S5).

Single Expression or/and Isolated Enzymes

Using oxidoreductases to synthesize high-value chiral building blocks is a powerful and highly efficient tool to aid chemists for designing more productive and sustainable synthesis of drugs. Actually, reductions or oxidations as ideal biocatalysts involved about 30% of the industrial processes as redox is a greatly efficient and attractive choice to the point of being environmentally friendly compared with the use of chemical catalysts [208]. Even though whole-cell catalysis (natural functional bacteria or engineering bacteria as “bags of enzymes”) has many advantages, it also showed some drawbacks [55]: (1) limited substrates or product permeability; (2) substrate or product inhibition, such as inhibition of the cell growth and biocatalyst activities; (3) low total output, low enzymatic activities, purification and separation of the target product are time-consuming and complicated, and the values of the environmental factor E would increase [208]; (4) the multiple reductases are often present in the whole cell which lead to the insufficiency of enantiomeric excess in this route [29]; (5) degradation and consumption of the product or substrate through cellular metabolism resulting in low enantiomeric excesses and yields; and (6) generation of unwanted by-products for the presence of multiple enzymes. And these shortcomings of the whole cells can be overcome by using isolated enzymes, which is an ideal method to generate a highly efficient coenzyme recycling system for industrial and lab scale, especially important for the former progress [12]. Compared with whole cells, one-step reactions catalyzed by isolated enzymes provide significant benefits of direct access of substrates and avoiding side reactions. A large number of new enzymes and biosynthetic pathways have been reported by genome sequencing, increasing considerably the field of enzymology and permitting the development of drug design [10]. In this review, we will show some cases of the use of single expression or/and isolated enzymes in the synthesis of chiral precursors of pharmaceuticals.

Oxidoreductase

Pharmaceutical production will benefit greatly from the selectivity earnings associated with the enzymatic catalysis by using biocatalysts such as oxidoreductase. Bio-oxidation reactions and bio-reduction reactions have been widely used in the pharmaceutical field. For example, boceprevir and telaprevir (both hepatitis C virus protease inhibitors) [120, 130] and the esomeprazole (the proton-pump inhibitor) [15] were manufactured industrially by using oxidases while sitagliptin was synthesized by using reductase for combating the type 2 diabetes [128]. The mutant D11 obtained by using rational design, directed evolution, and high-throughput screening with the monoamine oxidase from A. niger as the wild type was employed to synthesize the important pharmaceutical intermediates for the preparation of solifenacin and levocetirizine with > 97% e.e. value [63, 64] (entry 1, Table S6). And the monoamine oxidase mouton D5 was employed to synthesize the TRPV1 inhibitor and serotonin receptor inhibitor with 99% e.e. value [42] (entry 2, Table S6). The cyclohexanone monooxygenase was applied in the preparation of the 2-(benzhydrylsulfinyl) acetic acid, an important material for the preparation of the armodafinil. And it was approved by FDA that the armodafinil can be used as a stimulant-like pharmaceutical to cure shift work disorder, obstructive sleep apnea, narcolepsy, and other sleep disorders [3] (entry 3, Table S6). Bacterial laccases and the vanillyl alcohol oxidase from Penicillium simplicissimum (P. simplicissimum) were used for the synthesis of pinoresinol, which shows a protective effect of combating multiple health disorders, by a two-step one-pot synthesis in vitro [188] (entry 4, Table S6). Then, the pinoresinol reductase from Arabidopsis thaliana (A. thaliana) and pinoresinol lariciresinol reductase from Forsythia intermedia (F. intermedia) were employed for the establishment of a one-pot “two-cell” system to produce the enantiopure pinoresinol with more than 97% e.e. value [189] (entry 5, Table S6). The aromatic amino alcohols and N-carbobenzyloxy (Cbz)-protected aliphatic compound were oxidized to their corresponding α-hydroxyaldehydes by the galactose oxidase mutants [90]. Two engineered carbonyl reductases, P170H/L174Y and P170R/L174Y, were used to catalyze the 3-(dimethylamino)-1-(2-thienyl)-propan-1-one and 3-(dimethylamino)-1-phenylpropan-1-one to 3-(dimethylamino)-1-(2-thienyl)-1-ol and 3-(dimethylamino)-1-phenylpropan-1-ol, respectively. And the 3-(dimethylamino)-1-(2-thienyl)-1-ol and 3-(dimethylamino)-1-phenylpropan-1-ol are key intermediates for the production of some antidepressant drugs [270] (entries 6 and 7, Table S6). A new carbonyl reductase (KlAKR) from Kluyveromyces lactis (K. lactis) was discovered and employed to asymmetrically reduce t-butyl 6-cyano-(5R)-hydroxy-3-oxohexanoate to t-butyl 6-cyano-(3R,5R)-dihydroxyhexanoate, the crucial chiral precursor for the synthesis of the Lipitor^® (atorvastatin calcium) [135] (entry 8, Table S6). Then, the KlAKR was engineered by using a semi-rational design method based on molecular docking and homology modeling, and a mutant Y295W/W296L which exhibited 11.25-fold catalytic efficiency (k_cat/K_m) than that of a wild type was obtained [136]. The (R)-alpha-hydroxyketal, a great chiral precursor of an important pharmaceutical intermediate used to synthesize a chemokine receptor inhibitor, was synthesized by using a recombinant ketoreductase with > 99% e.e. value [121] (entry 9, Table S6). Coupling with a GDH for NADPH cofactor regeneration, an economical route was developed to produce the (R)-α-hydroxyketal with a 96–98% yield [121] (entry 10, Table S6). The (R)-tetrahydrothiophene-3-ol (entry 11 Table S6) or (S)-2-butanol (entry 11, Table S6), an important component for a potent antibacterial, sulopenem, was synthesized by using the native ketoreductase from Lactobacillus kefir and the engineered enzyme with directed evolution [76, 131]. The (S)-3, 5-bistrifluo-romethylphenyl ethanol, a raw material for synthesis of antagonists of receptors K-1, was produced by using the alcohol dehydrogenase (ADH) from R. erythropolis by Merck & Co. Inc. (entry 12, Table S6). The genetically modified ketone reductases, which were generated by directed evolution and were 3000 times than that of the native enzyme from Microbacterium campoquemadoensis MB5614, were used in the synthesis of (S,E)-methyl 2-(3-(3-(2-(7-chloroquinolin-2-yl)-vinyl)phenyl)-3-hydroxypropyl)benzoate, montelukast (an anti-asthma drug), with 90% yield and 99% e.e. value [131] (entry 13, Table S6). The ketoreductase KRED1001 was employed in the preparation of the (R)-2-hydroxy-3, 3-dimethylbutanoic acid, a key chiral intermediate for the synthesis of a thrombin inhibitor [131]. The GDH was introduced for cofactor NADPH regeneration, and the (R)-2-hydroxy-3,3-dimethylbutanoic acid was saponified without epimerization with > 99.5% e.e. value and 82% isolated yield [197] (entry 14, Table S6). Multifarious protein engineering technologies, such as mutation approach, substrate walking, and modeling, were employed in modifying a transaminase to enhance the productivity of the (R)-3-amino-1-(3-(trifluoromethyl)-5,6-dihydro-[1, 2, 4]triazolo[4,3-a]pyrazin-7(8H)-yl)-4-(2,4,5-trifluorophenyl)butan-1-one-itagliptin (sitagliptin, used as an oral antihyperglycemic agent) [122, 212] (entry 15, Table S6) or some other chiral amines [198]. An engineered transaminase was used in the preparation of sitagliptin with 99% enantiomeric excess and 92% yield (entry 16, Table S6), and this route was of more productivity and efficiency of synthesis of the target product compared with the corresponding chemical route [198]. And by using a chiral biocatalyst, sitagliptin was synthesized by enantioselective hydrogenation of an enamine [80, 89, 88]. A FDH for cofactor regeneration and a d-lactate dehydrogenase (D-LDH) for reduction reaction were selected to synthesize the (R)-3-(4-fluorophenyl)-2-hydroxy propionic acid, a key building block for the production of the rhinovirus protease inhibitor AG7088 [52, 265] (entry 17, Table S6), through the use of a membrane reactor [228]. The l-lysine ε-aminotransferase from Sphingomonas paucimobilis SC 16113 (S. paucimobilis SC 16113) was selected and overexpressed in E. coli for the synthesis of the [(4S)-(4a,7a,10ab)]-1-octahydro-5-oxo-4-[[(phenylmethoxy)-carbonyl]amino]-7H-pyrido-[2,1-b][1, 3]thiazepine-7-carboxylic acid, an important intermediate in the production of the omapatrilat which is used as an antihypertensive drug by inhibiting neutral endopeptidase and angiotensin-converting enzyme [171, 192] (entry 18, Table S6). (R)-Amino acid, an important material for the preparation of antimigraine drugs (also known as calcitonin gene–related peptide receptors) or other pharmaceuticals such as anticoagulants and fertility drugs [53, 87, 147, 230, 261], can be synthesized by combining with (R)-hydantoinases with (R)-acylases or (R)-carbamoylases, (R)-amidases, (R)-transaminases, and HNO₂. And an (S)-amino acid was synthesized by using an (S)-amino acid dehydrogenase and an (R)-amino acid oxidase with a precursor, racemic amino acid [82]. The (R)-2-amino-3-(7-methyl-1H-indazol-5-yl)-propanoic acid, an important intermediate for the production of antagonists of a calcitonin gene–related peptide receptors (for the preparation of antimigraine drugs and other pharmaceuticals) [31, 32], was prepared by using a (S)-amino acid oxidase, which was obtained from Proteus mirabilis (P. mirabilis) and expressed in E. coli, and a commercial (R)-transaminase with 68% isolated yield and > 99% e.e. value [86] (entry 19, Table S6). Then, the (R)-transaminase from B. thuringiensis was expressed in E. coli and employed in this progress, leading to a nearly complete bioconversion without requirement of additional biocatalysts (enzymes) for removing the pyruvate [86]. Two enzymes, (S)-aminotransferase from Sporosarcina ureae and (R)-amino acid oxidase from Thamniochloris variabilis expressed in E. coli, respectively, were put together for the synthesis of the (S)-amino-3-[3-{6-(2-methylphenyl)}ridyl]-propionic acid (a key material for glucagon-like peptide). Meanwhile, a (S)-aminotransferase from Burkholderia sp. was expressed in the E. coli system and applied in this route, and the isolated yield reached 99.4% and with e.e. value of more than 99.4% [160] (entry 20, Table S6). The (R)-cyclohexylalanine, another important (R)-amino acid which is an important chiral material for the production of thrombin inhibitor inogatran [77], was obtained by using the modified meso-diaminopimelate (R)-dehydrogenase from Corynebacterium glutamicum with > 99% e.e. value and 98% yield [149, 235] (entry 21, Table S6). And the modified enzyme form with nicotinamide cofactor dependence, high stereoselectivity, and broad substrate range was obtained through random and rational mutagenesis [235]. The LeuDH from Bacillus sphaericus ATCC 4525 was used in the synthesis of the α-keto-β-hydroxyisovalerate, a key material for the synthesis of the (S)-β-hydroxyvaline which is an important chiral intermediate for the production of tigemonam (orally active monobactam) [69], with a 98% reaction yield and a 99.8% enantiomeric excess [81] (entry 22, Table S6). And either GDH from B. megaterium or FDH from C. boidinii was employed in this progress for NADH regeneration, which is necessary in this reaction [81].

Aldolases

Aldolases are often used as the ideal choice to catalyze aldol additions to the corresponding asymmetric compounds with high catalytic efficiency and selectivity [35]. The N-acetyl-d-neuraminic acid aldolase either from Clostridium perfringens (C. perfringens) or E. coli was applied to produce N-acetyl-d-neuraminic acid, an important intermediate to synthesize a selective and potent influenza virus sialidase inhibitor Relenza which is an important drug for curing the two types most responsible for flu epidemics [5, 44, 116, 139] (entry 1, Table S7). Then, this progress was improved by implementing the reaction in an excellent enzyme membrane reactor for enhancing the productivity of N-acetyl-d-neuraminic acid synthesis [123]. And the N-acetyl-d-neuraminic acid aldolase from E. coli was immobilized to synthesize quantities of N-acetyl-d-neuraminic acid [40, 140]. And immobilized enzyme technology would benefit from the enzyme reusability [27]. The 2-deoxyri-bose-5-phosphate aldolase was selected to synthesize the 2,4-dideoxyhexose derivative which acts as a key role in the production of some cholesterol-lowering agents and HMG-CoA reductase inhibitors with a 96% diastereomeric excess and more than a 99% e.e. value by Diversa Corporation, and the aldolase was expressed in the E. coli system [215] (entry 2, Table S7).

Lipases

Lipases are general candidates of biocatalyst for the biosynthesis of chiral compounds by the kinetic resolution of amides and esters or the enantioselective enzymatic desymmetrization of amines and secondary alcohols to synthesize the corresponding amides or chiral esters due to their good stability and high activity in organic solvents. Several labs or industrial processes have already used the lipase as catalysts in the synthesis of pharmaceuticals as described. The lipase from Thermomyces lanuginosus (T. lanuginosus) was immobilized to prepare (S)-indanol for the synthesis of the rasagiline mesylate, an active key ingredient of the medicine AZILECT^® which is used to combat Parkinson’s disease, with > 99% e.e. value [60, 223] (entry 3, Table S7). The lipase from Candida rugosa (also called C. cylindracea lipase) as the ideal candidate has been employed as serine hydrolases in the production of (S)-2-arylpropionic acids (active isomer) and non-steroidal anti-inflammatory drugs (NSAIDs), like ibuprofen ((S)-2-(4-isobutylphenyl)propanoic acid, (S)-ibuprofen) (entry 4, Table S7). The (S)-ibuprofen was synthesized in industrial scale by the Pfizer, and this route was achieved by C. rugosa lipase through enantioselective hydrolysis of the racemic methoxyethyl ester (entry 4, Table S7), and the C. rugosa lipase was immobilized in a membrane reactor [206, 207]. Two isoenzymes from C. antarctica, A (CALA) and B (CALB) which presented several differences, have been described. CALB was used as the biocatalyst to synthesize the antithrombotic compound lotrafiban S-16 by GlaxoSmithKline Pharmaceuticals (entry 5, Table S7), and the enzyme was immobilized in aqueous medium [241]. CALB has been also used in the synthesis of (S)-monoester, (S)-3-(3,4-dichlorophenyl)-5-ethoxy-5-oxopentanoic acid (entry 6, Table S7), by the Schering-Plough company, and (S)-3-(3,4-dichlorophenyl)-5-ethoxy-5-oxopentanoic acid is a key chiral intermediate for the production of antagonists of tachykinin receptors NK1 and NK2, which were used in the treatment of arthritis, asthma, and migraine in various biological processes like vasodilatation, secretion, inflammation, and pain transmission [95, 237]. By using DNA shuffling, a chimeric lipase B with high activity for the diethyl 3-[3′, 4′-dichlorophenyl]glutarate was created with three homologous lipases from Hyphozyma sp. CBS 648.91, C. antarctica ATCC 32657, and Cryptococcus tsukubaensis ATCC 24555 as the female parent. And the activity of the lipase was improved 20-fold than that of the parents [220] while the thermostability of the lipase B was enhanced by directed evolution [266] (entry 7, Table S7). The (S)-tert-butyl 2-carbamoyl-2,3-dihydro-1H-pyrrole-1-carboxylate (entry 8, Table S7), an important intermediate for the production of the oral hypoglycemic agent saxagliptin, was synthesized by CALB in Bristol-Myers Squibb [65]. And the (1S,2R)-2-(methoxycarbonyl)cyclohex-4-ene-1-carboxylic acid (entry 9, Table S7), an important intermediate for the generation of a chemokine receptor’s potential modulator, was synthesized by immobilized C. antarctica CALB (Novozyme 435^@) in Bristol-Myers Squibb (69). The Novozyme 435 (Chirazyme L-2 or C. antarctica lipase B) was selected for the preparation of the ribavirin, an antiviral agent for the treatment of hepatitis C [57, 180, 225] (entry 10, Table S7). The lipase from Pseudomonas cepacia or BMS lipase (extracellular enzyme obtained from Pseudomonas sp. SC 13856), which showed good selectivity of secondary alcohols’ transesterification or esters’ hydrolysis and was immobilized through polypropylene Accurel, was employed to synthesize the potent anticancer agent paclitaxel (Taxol) [6] (entry 11, Table S7), an antimitotic agent for the treatment of breast and ovarian cancer. Semi-synthetic pathways of paclitaxel were much more attractive alternatives as the 1 kg of paclitaxel was isolated consuming around 3000 tree barks. The porcine pancreatic lipase was immobilized to synthesize (R)-oxiran-2-yl-methanol, which was used as an important chiral intermediate for the synthesis of β-blockers, by the DSM company [207] (entry 12, Table S7). Pseudomonas fluorescens lipase as one of the most ideal enzymes in the resolution of racemic compounds was used in the synthesis of a chiral building block, (3S,3aR,6aR)-3-hydroxy-3,3a,4,6a-tetrahydro-2H-cyclopenta[b]furan-2-one, which is required for the preparation of the nucleoside analogue carbovir, an antiviral agent used for combating AIDS [54] (entry 13, Table S7). Lipase from Serratia marcescens, such as S. marcescens Sr41 8000, was applied in the synthesis of a key chiral intermediate for the calcium canal blocker diltiazem by Tanabe Pharmaceutical Co. [143] (entry 14, Table S7). Pregabalin [(S)-3-(aminomethyl)-5-methylhexanoic acid], a γ-aminobutyric acid (a lipophilic GABA analogue) was developed to cure some central nervous system disorders, such as social phobia, epilepsy, generalized anxiety disorder, neuropathic pain, and chronic pain like spinal cord injury and fibromyalgia [127, 204] (entry 15, Table S7). A new greater manufacturing route for synthesis of pregabalin was developed by using a commercially available lipase compared with the traditional methods of chemical synthesis.

Transaminases

Transaminases are widely used in the chiral amines or chiral amino synthesis and the chiral amines or chiral amino acids are key building blocks for lots of bioactive pharmaceuticals and natural products, such as sitagliptin and atazanavir [177]. Sitagliptin was developed by spiroindolone compounds and Codexis for malaria treatment [37, 198]. The transaminase mutant ATA-303 was evolved based on ATA-013 transaminase to produce the key chiral amine intermediate for the synthesis of vernakalant, which is used as an antiarrhythmic agent to treat the atrial fibrillation and most common cardiac arrhythmia, with > 99.5% e.e. value and 85% assay yield [133] (entry 16, Table S7). Monoamine oxidases and x-transaminases were employed to synthesize the 2,5-disubstituted pyrrolidines, key scaffolds for various natural products, and pharmaceutical compounds, with > 94% excellent enantioselectivity and > 98% diastereoselectivity [153] (entry 17, Table S7). Modular cascade biocatalysis was established to synthesize (S)-phenylethanolamine and its derivatives, key chiral precursors for lots of bioactive compounds like antitumor potential drugs BMS-577098 [236] and BMS-536924 [252], with high yields and high e.e. value [255] (entry 18, Table S7). And five enzymes containing x-transaminase were employed for the establishment of two enzyme modules and were overexpressed in E. coli to produce (S)-amino alcohols in one-pot conversion [255] (entry 19, Table S7). The transaminase ATA-036 was employed to synthesize the 1-((2R,4R)-2-(1H-benzo[d]imidazol-2-yl)-1-methylpiperidin-4-yl)-3-(4-cyanophenyl)urea, a smoothened receptor inhibitor which was confirmed to have a potential impact on a variety of blood-related cancers by Pfizer [150] (entry 20, Table S7), with 99% e.e. value and 85% yield [179]. An x-transaminase mouton from Arthrobacter sp. was obtained to produce 17-a-amino steroids with 99% e.e. value and > 83% yield [78] (entry 21, Table S7). Ruegeria sp. TM1040 transaminase was evolved for the preparation of some important bulky chiral amines and it showed high stereoselectivity and 8900-fold higher activity than that of the wild type [78, 178] (entry 22, Table S7). Rational design was used to broaden the substrate spectrum of the amine transaminase from Vibrio fluvialis (V. fluvialis), and the ideal mutant, W57F/R88H/V153S/K163F/I259M/R415A/V422A, was obtained with > 1716 times higher reaction rate and > 99% e.e. value of pure (S)-amine [51] (entry 23, Table S7).

Lyases and Hydrolase

Lyases are used for the preparation of chiral compounds as they can catalyze their substrate to generate new double bonds through removing the groups from the substrates by methods other than oxidation and hydrolysis. The (+)-γ-lactamase from Bradyrhizobium japonicum USDA 6 (B. japonicum USDA 6) was discovered and overexpressed in E. coli [273]. Then, the (+)-γ-lactamase was applied in the synthesis of the (+)-γ-lactam, an important chiral intermediate for abacavir and carbovir (the antiviral carbocyclic nucleoside drugs) which were used in curing the “swine flu” and HIV infection, respectively, with excellent enantioselectivity (> 99.8% e.e. value) and high yield (> 49.9%) [61, 62, 223] (entry 24, Table S7). The γ-lactam 2-azabicyclo[2.2.1]hept-5-en-3-one, an ideal intermediate for the synthesis of the abacavir (also known as Ziagen, a 2-aminopurine nucleoside analogue) which is selected as a reverse transcriptase inhibitor to treat hepatitis B viruses and human HIV, was synthesized using the Rhodococcus NCIMB40213 and Pseudomonas solanacearum NCIMB40249 as these two microorganisms have γ-lactamase [54, 141] (entry 25, Table S7). And some commercially available enzymes were involved for the preparation of the target products. The whole Erwinia herbicola (E. herbicola) cells containing a tyrosine phenol lyase were applied in the synthesis of the l-3,4-dihydroxyphenylalanine, an important drug predominantly applied in curing Parkinson’s disease [162] (entry 26, Table S7). Lyases were also used in the synthesis of other pharmaceutical intermediate and pharmaceuticals [223]. A tyrosine phenol lyase and a monooxygenase P450 BM3 were used for the establishment of a one-pot two-step cascade scheme to synthesize 3-methoxy-l-tyrosine, which is more stable than l-3,4-dihydroxyphenylalanine and has positive effects on the cure of Parkinson’s disease, with > 97% e.e. value and an output of 2 g/L [45] (entry 27, Table S7). The whole-cell catalyst, E. coli cells consisting of the cellobiose 2-epimerase from Caldicellulosiruptor saccharolyticus (C. saccharolyticus), was applied in the production of the lactulose, which is generally applied in nutraceuticals, pharmaceuticals, and food industries [243], with 65.1% conversion yield [243] (entry 28, Table S7). Then, the best mutant G4-C5, R5M/I52V/A12S/K328I/F231L which was obtained through high-throughput screening after four cycles of random mutagenesis, was used for the preparation of lactulose with the elevated conversion yield of 76% [211] (entry 29, Table S7). N-Succinylamino acid racemase and d-succinylase were applied in the synthesis of d-phenylalanine with 91.1% conversion yield and 86.7% e.e. value [221] (entry 30, Table S7). The N-succinylamino acid racemase from Geobacillus kaustophilus CECT4264 (G. kaustophilus CECT4264) and l-carbamoylase from Geobacillus stearothermophilus CECT43 were immobilized to produce a panel of unnatural and natural l-amino acids [216]. d-Phenylalanine and its derivatives are important building blocks for many drugs and natural products, such as d-phenylalanyl-l-prolyl-l-arginine chloromethyl ketone and nateglinide [156]. d-Substituted phenylalanines were synthesized by using an engineered phenylalanine ammonia lyase with 98% e.e. value and the phenylalanine ammonia lyase was obtained from cyanobacteria Anabaena variabilis (A. variabilis) [157] (entry 31, Table S7). Phenylalanine ammonia lyase was also used for the synthesis of unnatural l-amino acids [193]. The phenylalanine ammonia lyase from Rhodotorula graminis was evolved to synthesize l-Br-phenylalanine with a > 99% e.e. value and a 94% yield [193] (entry 32, Table S7). The (S)-b- or (S)-a-arylalanines was synthesized by using the ammonia lyase EncP from Scirpus maritimus and its rationally designed mutants [246] (entry 33, Table S7).

Endopeptidases

Endopeptidases, such as serine endopeptidases, are a kind of hydrolases, hydrolyze proteins, which bio-catalyze the hydrolysis of peptidic bonds. These enzymes are ideal biocatalysts because they not only unrequired consuming cofactors but also are very regio-, enantio-, and chemo-selective. And the reaction is often implemented under mild reaction conditions [185]. The alkaline endopeptidase (also called as subtilisin A or subtilisin Carlsberg) from Bacillus licheniformis was used in the preparation of (S)-2-benzyl-3-(2-methyl-1-morpholino-1-oxopropan-2-ylsulfonyl)propanoic acid, a chiral intermediate for the synthesis of ciprokiren and renin inhibitor remikiren, with 99% e.e. value and 50% yield [50] (entry 34, Table S7).

Isomerases

Isomerases have great potential in synthetic medicinal chemistry and pharmaceuticals as they can transform cheap, abundant sugars into rare, expensive sugars [9]. The mannose-6-phosphate isomerase and l-arabinose isomerase were used in the preparation of l-ribose, a key precursor for the preparation of pharmaceutically l-form sugar which was used in the production of those targeting human cytomegalovirus, HIV, and hepatitis B/C virus [260] (entry 35, Table S7). The l-arabinose isomerase from Paenibacillus polymyxa (P. polymyxa) was immobilized on nanomaterials to produce the l-ribulose with 61.8% conversion rate [260] (entry 36, Table S7). A high-efficiency three-step separation-integrated route was built by combining a d-xylose isomerase, a d-tagatose epimerase, and an invertase to synthesize d-psicose which shows the ability to enhance d-glucose tolerance and insulin sensitivity in type 2 diabetes [96], with 89% yield and 99.9% purity [240] (entry 37, Table S7). Two mutants, Var8 and IDF10-3, were obtained by the directed evolution of the d-tagatose epimerase and employed in the synthesis of the d-psicose in an enzyme membrane reactor [19] (entry 38, Table S7).

Other Enzymes

Although oxidoreductase, transaminases, aldolases, lactamase, and lyases are the most commonly used biocatalysts in the biocatalytic synthesis of chiral drug intermediates and chiral pharmaceuticals, some other enzyme cases also deserved to be described, as listed in Table S7. The cytidine deaminase from recombinant strain E. coli was immobilized to synthesize the Epivir [(2′R-cis)-2′-deoxy-3-thiacytidine], which is approved by the FDA to cure HIV, and it was used more than 15 cycles [139, 140] (entry 39, Table S7). The nitrilases, which can bioconvert a variety of nitriles, were also widely employed as the ideal biocatalyst in chemical transformations and one of the most representative cases was shown in the production of (R)-3-hydroxy-4-cya-nobutyric acid, an important intermediate in the synthesis of the Lipitor (a cholesterol-lowering agent) with 98% yield and 95% e.e. value [46] (entry 40, Table S7). And a nitrilase and its improved version, which was obtained by using site-saturation mutagenesis, were used for the synthesis of 3-hydroxyglutaronitrile [46, 47] and it was demonstrated that nitrilases can be used to synthesize various chiral carboxylic acid [48, 119]. The (S)-2-ethoxy-3-(4-hydroxyphenyl)propanoic acid, a key intermediate for the synthesis of ragaglitazar (a new antidiabetic drug), was synthesized using the esterase from Aspergillus oryzae with 99.6% enantiomeric purity and 48% yields [54] (entry 41, Table S7). The phenylpyruvate decarboxylase from Achromobacter eyrydice (A. eyrydice) was used for the synthesis of the PhCH₂COCH(OH)R, the α-hydroxyketones (acyloin) which is an important class of chiral intermediates in organic synthesis [54, 74] (entry 42, Table S7). Semi-synthetic pathways, as an ideal approach, were used in overcoming satisfactory chemical and pharmacological limitations of β-lactam antibiotics, one of the generally employed and the most important pharmaceuticals. 7-Aminocephalosporanic acid (7-ACA), 6-amin-openicillanic acid (6-APA), and 7-amine-3-desacetoxycephalosporanic acid (7-ADCA) have been prepared by hydrolysis of cephalosporin G and penicillin G. The 7-ADCA was obtained from cephalosporin G by using penicillin G acylase from Actinomyces viscosus or E. coli with 93% yields (entry 43, Table S7). And the Dr. Vig Medicaments has successfully prepared the 6-APA by using penicillin G acylase from immobilized B. megaterium, A. viscosus, or E. coli [132] (entry 44, Table S7).

Co-expression of the Key Enzymes

The most prominent routes for the synthesis of chiral alcohol, chiral amino acid, and chiral amine were executed and usually required cofactor NADH or NADPH. Coenzyme regeneration is a very ideal method for the coenzyme-dependent progress due to the addition of the costly cofactor [234] and it is an ideal selection to co-express the key enzyme for coenzyme regeneration to reduce costs and help achieve green production. In the last years, scientists spent a lot of energy on building cofactor regeneration systems and realizing the regeneration of cofactor by using different enzymatic, electrochemical, and chemical approaches [12], as less than 0.1 mol% cofactor used can be taken as economically acceptable [102]. However, these effects do not achieve the preparative relevance. Enzyme-coupled (one enzyme catalyzes the synthesis of the target product, and another enzyme (auxiliary enzyme) is used to regenerate the coenzyme) and substrate-coupled (an enzyme catalyzes oxidation and reduction reactions with cofactor regeneration, such as a ketone substrate and a cheap alcohol were used as the substrate-coupled materials) coenzyme regeneration, two enzymatic regeneration strategies, were established. FDH is the best-known ideal enzyme for regeneration of cofactor as the reaction is practically irreversible, very easy, and the carbon dioxide can be eliminated in a timely manner, which is beneficial to the separation of products and helpful reaction to the direction of generating coenzymes. In some cases, the phosphite dehydrogenase was also used as the auxiliary enzyme [238]. And the GDH is also used for generating coenzymes for the high activity of GDH and practical irreversibility of the reaction [102, 215]. Using cofactor coupling to synthesize valuable molecules is becoming a great attractive methodology [269]. And it could be used to synthesize the high value-added products, such as L-Tle and 1, 3-dihydroxiacetone, by using different enzymes to catalyze a reductive and an oxidative reaction at the same time with raw materials, trimethylpyruvic acid, and glycerol.

There are many enantiopure molecules synthesized by using enzyme-coupled or substrate-coupled route. The FDH from C. boidinii and ADH from R. erythropolis used the enzyme-coupled to produce (S)-1-phenylpropan-2-ol, a key intermediate to prepare the amphetamines (sympathomimetics), and this progress was developed by the Forschungszentrum Jülich GmbH Co. (entry 1, Table S8). The (S)-3-hydroxyadamantylglycine, a key chiral intermediate for the synthesis of saxagliptin, an inhibitor of dipeptidyl peptidase-4, and an antidiabetic drug, was synthesized by using a modified form of PheDH from T. intermedius and expressed in P. pastoris with an engineered FDH for cofactor recycling [197] (entry 8, Table S4). The reductive amination route was scaled up to synthesize saxagliptin by using an engineered PheDH and FDH which were expressed in a single E. coli cell with 98% yield and 100% e.e. value [84, 239]. And the modified PheDH has 12 amino acid extensions in the C-terminus area and two amino acid mutations at the same area [84, 239]. The (R)-4-chloro-3-hydroxybutanoate esters (methyl ester and ethyl ester), which were used to generate the lateral chain of atorvastatin, can be prepared by using a recombinant ketoreductases with GDH and glucose (entry 2, Table S8) or by using the whole cells of E. coli [138, 159] (entry 2, Table S8). An engineered ketoreductase was used for the synthesis of the (S)-3-(N,N,dimethylamino)-1-(thien-2-yl)propan-1-ol, which was used as a precursor for duloxetine (an antidepressant drug), and using GDH and glucose for cofactor regeneration [199] (entry 3, Table S8). The recombinant E. coli, which co-expresses both secondary alcohol dehydrogenase from P. finlandica and FDH from Mycobacterium sp., was employed to produce (S)-4-chloro-3-hydroxybutonoate, an important chiral intermediate for the preparation of a HMG-CoA reductase inhibitor [106, 244], with 99% e.e. value and 98.5% yield [208] (entry 4, Table S8).

Fusion expression forms a multifunctional biocatalyst that was also used in the synthesis of drug intermediate. The functional fusion with two or more enzymes is built to aim at offering several advantages than individual enzymes in reaction kinetics and enzymatic catalysis. Proper linker peptides can reduce folding interference from each other for fusion enzymes making it function as independently as possible by inserting between individual enzymes [36]. A FDH from C. boidinii and a PheDH from Bacillus halodurans were tethered by using a linker peptide for constructing a cofactor regeneration system to rapid analysis of phenylketonuria (PKU) and high-efficiency synthesis of l-phenylalanine, an important material for the preparation of aspartame (sweetener), with 99% conversion rate [108] (entry 5, Table S8). Peptide linkers were used for the establishment of bifunctional LeuDH and FDH to generate enzymatic complex for L-tert-leucine (L-Tle) biotransformation and efficient cofactor regeneration [271] (entry 6, Table S8).

Regulatory Element and SD Sequences for Biocatalytic Synthesis

One of the most significant decisions when designing and planning a manufacturing process employing biocatalysis method is whether the conversion is executed using one or more biocatalysts (enzymes) contained within an isolated enzyme or whole-cell extract [183]. A regulatory expression system to coordinate enzymes by cascade connection of protein-protein expressions and regulation of strength of RBS was established for the preparation of L-Tle, an important chiral drug intermediate for the preparation of the drugs telaprevir (HCV) and Reyataz (HIV), with approximately 100% yield and > 99% enantiopure form [109] (entry 1, Table S9). And the schematic diagram and flow chart of this experiment was shown as Fig. 2. A trienzyme CLEA including a non-selective nitrilase from P. fluorescens EBC 191, a (S)-hydroxynitrile lyase from Manihot esculenta (M. esculenta), and an amidase from R. erythropolis was employed for the construction of a biocatalytic cascade system to produce the (S)-mandelic acid with > 99% e.e. value and 90% yield [108, 33] (entry 2, Table S9). And the rate of the biocatalytic cascades was higher than that of admixtures of the respective cross-linked enzyme aggregates [33]. Three different routes were developed to synthesize the chiral intermediates for the preparation of omapatrilat 1: (1) 2-keto-6-hydroxyhexanoic acid was conversed to l-6-hydroxynorleucine by using beef liver glutamate dehydrogenase, and d-amino acid oxidase from T. variabilis or porcine kidney was employed to synthesize (1) the 2-keto-6-hydroxyhexanoic acid for the chemical synthesis of the 2-keto-6-hydroxyhexanoic acid since it was lengthy; (2) the PDH from T. intermedius was selected to synthesize the (S)-2-amino-5-(1,3-dioxolan-2-yl)-pentanoic acid and the PDH was overexpressed in P. pastoris and E. coli; and (3) the l-lysine-aminotransferase from S. paucimobilis SC 16113 was employed to produce the [4S-(4a,7a,10ab)]1-octahydro-5-oxo-4[[(phenylmethoxy)carbonyl]amino]-7H-pyrido-[2,1-b] [1, 3]thiazepine-7-carboxylic acid, and a E. coli which contain the enzyme was also used in this synthetic process. Therefore, there are some biocatalysts that are applied to the synthesis of the chiral intermediates of omapatrilat 1. Chemoenzymatic cascade reactions and biological cascade reactions also continue to attract enhancing attention and make significant progress [43, 242]. The safety-based process development and smart design ensure the productivity and quality of the pharmaceuticals manufactured are also the increasing key challenges in pharmaceutical synthesis [248]. And the yield of the target product would significantly increase while the regulatory elements or SD sequences were involved in multi-enzyme-catalyzed composite reaction to balance the speed between the reactions and the cost. Taken together, it is envisaged that the regulatory element and SD sequences for catalytic synthesis will become more generally used in the production of pharmaceutical intermediates and pharmaceuticals in the future.

Fig. 2.

The schematic diagram and flow chart for the construction of the regulatory expression system. I: Design genetic circuit; II: construct genetic circuit; III: expression and regulation of the target protein; IV: a tunable multi-enzyme-coordinate expression coenzyme regeneration system was constructed for biosynthesis of L-tert-leucine

Metabolic Network (Predesign)

We foresee the enhanced application of synthetic biology and metabolic engineering as an ideal strategy for the fast implementation of biocatalysis routes into drug development, and this strategy will give an exciting future for biocatalytic routes, such as in multistep processes using one designed whole cell to express multiple enzymes. Lately, cell-free systems and redesigned recombinant whole cell were designed and developed to execute novel complex biocatalytic reactions in vitro [17, 25, 91, 146, 263]. Telescoping multistep synthesis to generate one-pot catalytic cascades is ultimate in green conversion methodologies and the soul of synthesis as this system could avoid the purification of intermediates and time-consuming isolation. Such catalytic cascade syntheses, which are built by truly emulating the enzymatic progresses in the metabolic pathways in living cell, show several advantages: drive equilibria go in the direction of product and avoiding the use for excess reagents, the incompatibility of the catalysts with each other can be overcome, less solvent, fewer unit operations, reactor volume, higher space-time yields and volumetric, shorter cycle times, and less waste, which mean substantial environmental and economic benefits [209]. The envisaged environmental and economic benefits of biocatalytic cascade processes (also known as the cell-free synthetic biology) make these progresses become a focus of attention in the last few years [210]. Free enzymatic bioconversion pathways are designed and engineered through assembling more than ten purified enzymes by using synthetic biology, and the coenzymes, which were used for the synthesis of the target products, could not be prepared by using a single enzyme. Cell-free synthetic biological systems (biocatalytic cascade processes) are more efficient, easier to build, and allow greater engineering flexibility than the modification of primitive functional cells as they are not influenced by complexity, cell viability, cellular membranes, physiology, and/or cell walls [263, 268], solving equilibrium issues [92] and substrate and/or product inhibition [93].

Construction of de novo metabolic pathways has placed great expectations to biosynthesize useful advanced intermediates. And it is predicted that novel biosynthetic pathways will be predesigned and assembled to generate an ideal metabolic pathway for a one-step synthesis of many semi-synthetic natural products in the future [24]. A metabolic pathway was constructed and optimized for the production of cis-(1S,2R)-indandiol and trans-(1S,2R)-indandiol by using metabolic engineering [23] (entry 3, Table S9). Indene was oxidized to mixtures of trans- and cis-indandiols and related metabolites by Rhodococcus sp. and P. putida isolates, and its metabolism was consistent with dioxygenase and monooxygenase activity. The P. putida was applied in the synthesis of the cis-(1S,2R)-indandiol, a key intermediate for the preparation of CRIXIVAN (indinavir sulfate) which was a protease inhibitor for the cure of AIDS. It was confirmed that the dihydrodiol dehydrogenase was required to resolve racemic mixtures of cis-indandiol while the toluene dioxygenase from P. putida was cloned in E. coli [23]. These synthetic biological systems show prominent use for the synthesis of biofuels, e.g., hydrogen, ethanol and butanol [263, 267, 268], organic acids and amino acids [249], and chiral pharmaceuticals [23, 24, 109, 233].

Conclusions and Future Perspectives

Due to the increasing strict environmental, quality, economical, and safety requirements of industrial synthetic procedures, as well as the growing demand of compounds which are enantiomerically pure, the development of more healthy, sustainable, and economically and environmentally attractive strategies for the product of chiral biologically active molecules, such as chiral drug intermediate and chiral drug, still remains a great challenge in the pharmaceutical industry. Biocatalysis is an ideal option to create or resolve one or more chiral centers [183]. Biocatalysis has been successfully used in almost all kinds of bioconversion reactions in the last few decades, and it showed benefits for the pharmaceutical and fine chemical research industries [223]. Many biocatalysis routes have been successfully built, which is more sustainable and environment friendly than that of chemical synthetic progress. It is no doubt that using biocatalysts to synthesize the key enantiopure molecules is an ideal and potential route from the cases and examples presented in this review. And it will enhance the usefulness of bioconversion progresses inside the frame of industrial biotechnology with the increasing demands of greener synthetic progresses; the great recent advances in protein chemistry, biochemistry, molecular cloning, random and site-directed mutagenesis, DNA shuffling, and direct evolution of biocatalyst; the achievement in stabilization and immobilization of enzymes or cells, bioprocess engineering, and modeling techniques [183]; and the successfully acquired new enzyme, such as silico enzymes [112], with the ability to perform a catalytic reaction that has never been done before. For example, the protein engineering can modify the biocatalyst to match a programmed synthesis or conquer process limitations like product or inhibition [16, 26]. The bioconversion field, which is executed by biocatalyst, has emerged as a good alternative for the traditional synthetic method, because of the exquisite regio-, enantio-, and chemo-selectivities commonly displayed by function cells or enzymes; thus, biocatalysis is becoming a universal methodology for the biosynthesis of chiral pure compounds in laboratory scale as well as in industrial scale. Additionally, the use of biocatalysts often circumvents the requirement for functional group activation, hence avoiding deprotection and protection steps usually required in conventional organic syntheses. And these properties afford courses which generate less or shorter waste, and are, hence, both economically and environmentally smarter than traditional routes.

It is still required to engineer function microorganisms and enzymes to enhance their stability, activity, selectivity, and productivity under the mild reaction conditions [227]. There are many routes to enhance robustness of biocatalysts and function microorganism, among them (1) immobilization may possibly be one of the most conventional methods used and studied [79]; (2) the extreme conditions have been used to screen the extremophiles for excavation of novel biochemical pathways and new enzymes [87, 250, 256, 4, 205, 60, 34, 179, 182, 96, 216]; (3) the use of enzyme molecular engineering techniques, substrate engineering techniques, like directed evolution [1, 84, 137, 142, 186, 187, 194, 202, 203, 226, 251, 266, 272, 200], and DNA shuffling [97, 98, 148, 220] has vastly contributed to the new function of gene cluster or biocatalysts, able to work effectively in experimental conditions greatly different from the “natural” ones, in terms of substrate spectrum, presence of organic solvents, pH, temperature, etc.; (4) the use of regulatory elements [109], such as promoter and ribosome bind site (RBS) [196], to regulate and coordinate the expression of key enzymes which are the key catalyst for the synthesis of target products, and regulation and optimization of metabolic network through gene engineering and protein engineering [247]; (5) using synthetic biology, computational systems biology, and metabolic engineering to design and optimize artificial metabolic network to synthesize high-value chemicals such as omega-3 LCPUFAs [254] and docosahexaenoic acid (DHA) [190] and chiral drug intermediates such as L-Tle [109], cis-(1S,2R)-indandiol [23], and trans-(1S,2R)-indandiol [23]. And we expect that the holistic approach of the methods described in this review will play an ideal and important enabling role in the development of chiral drug intermediates, high value-added chemicals, and chiral pharmaceuticals.