Generation of novel family of reductases from PCR based library for the synthesis of chiral alcohols and amines

Article abstract of DOI:10.1016/j.enzmictec.2018.07.006

Biocatalysis has shown tremendous potential in the synthesis of drugs and drug intermediates in the last decade. Screening of novel biocatalysts from the natural genome space is the growing trend to replenish the harsh chemical synthetic routes, commonly used in the pharmaceutical and chemical industry. Here, we report a novel ketoreductase (KERD) and a nitrile reductase isolated from the PCR based library generated from the genome of Rhodococcus ruber and Bacillus subtilis, respectively. Both the proteins are hypothetical in nature as there is no putative homology found in the database, although both the enzymes have significant activity towards the synthesis of chiral alcohols and amines. Enzyme activity over a wide range of substrates (aromatic and aliphatic) for both the novel catalysts was observed. From the unique gene sequence to activity over a broad range of substrate and >99% conversion at higher concentrations (100 mM and above) entitles both the hypothetical enzymes as novel. The novel KERD has shown >99% selectivity for the synthesis of (S)-phenylethanol which makes it a potential candidate for industrial catalysis. The novel nitrile reductase has also shown promising activity for the synthesis of (R)-2-phenylethanolamine, which is a difficult moiety to synthesize chemically. In this report, starting from a homology based library, two highly potent whole cell biocatalysts are obtained.

Full text of DOI:10.1016/j.enzmictec.2018.07.006

Accepted Manuscript
Title: Generation of novel family of reductases from PCR
based library for the synthesis of chiral alcohols and amines
Authors: Pallvi Sehajpal, Seema Kirar, Saptarshi Ghosh,
Uttam Chand Banerjee
PII:
S0141-0229(18)30194-7
DOI:
Reference:
https://doi.org/10.1016/j.enzmictec.2018.07.006
EMT 9246
To appear in:
Enzyme and Microbial Technology
Received date:
Revised date:
Accepted date:
24-12-2017
26-6-2018
27-7-2018
Please cite this article as: Sehajpal P, Kirar S, Ghosh S, Banerjee UC,
Generation of novel family of reductases from PCR based library for the
synthesis of chiral alcohols and amines, Enzyme and Microbial Technology (2018),
https://doi.org/10.1016/j.enzmictec.2018.07.006
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Generation of novel family of reductases from PCR based library for the
synthesis of chiral alcohols and amines
Pallvi Sehajpal, Seema Kirar, Saptarshi Ghosh, Uttam Chand Banerjee*
Department of Pharmaceutical Technology (Biotechnology)
National Institute of Pharmaceutical Education and Research,
Sector 67, S. A. S. Nagar-160062, Punjab, India.
* Corresponding Author
ucbanerjee@niper.ac.in
Phone No: 91-172-2214682-85 Extn 2142 Fax: 91-172-2214692
Graphical abstract
Highlights

 Two highly potent biocatalysts, novel ketoreductase (KERD) and nitrile reductase
were obtained from the PCR based library generated from the genome of
Rhodococcus ruber and Bacillus subtilis, respectively.
 Both the proteins are found to be hypothetical as there is no putative homology found
in the database. Both the novel enzymes have shown activity over wide range of
substrates (aromatic and aliphatic).
 The novel KERD has shown > 99% selectivity for the synthesis of (S)-phenylethanol
which makes it a potential candidate for industrial catalysis.
 Novel nitrile reductase has also shown high stereo-specificity with > 99% conversion
for the synthesis of (R)-phenylethanolamine, which is difficult to synthesize
chemically. Phenylethanolamine has tremendous potential as pharmaceutical
intermediate.
Abstract
Biocatalysis has shown tremendous potential in the synthesis of drugs and drug intermediates
in the last decade. Screening of novel biocatalysts from the natural genome space is the
growing trend to replenish the harsh chemical synthetic routes, commonly used in the
pharmaceutical and chemical industry. Here, we report a novel ketoreductase (KERD) and a
nitrile reductase isolated from the PCR based library generated from the genome of
Rhodococcus ruber and Bacillus subtilis, respectively. Both the proteins are hypothetical in
nature as there is no putative homology found in the database, although both the enzymes
have significant activity towards the synthesis of chiral alcohols and amines. Enzyme activity
over a wide range of substrates (aromatic and aliphatic) for both the novel catalysts was
observed. From the unique gene sequence to activity over a broad range of substrate and >
99% conversion at higher concentrations (100 mM and above) entitles both the hypothetical
enzymes as novel. The novel KERD has shown > 99% selectivity for the synthesis of (S)-
phenylethanol which makes it a potential candidate for industrial catalysis. The novel nitrile
reductase has also shown promising activity for the synthesis of (R)-2-phenylethanolamine,
which is a difficult moiety to synthesize chemically. In this report, starting from a homology
based library, two highly potent whole cell biocatalysts are obtained.
Key words: Ketoreductase, Nitrile reductase, Hypothetical protein, Chiral alcohols and
amines, Drug intermediates, Biocatalysis

Introduction
Biocatalysis has shown tremendous potential in the synthesis of drugs and drug
intermediates in the last decade [1]. It has emerged as the green alternative over the
harsh chemical synthesis reactions used in pharmaceutical and chemical industries [2].
Stereoselective reduction of ketones to chiral alcohols and nitriles to amines are the
most useful reactions in organic synthesis [3,4]. Generation of side products, poor
substrate specificity and lengthy processes renders chemical catalysis inefficient for
industry [5,6]. Although, biocatalysis has emerged as environment friendly alternative
over the harsh chemical processes, however, poor substrate specificity, narrow
substrate range and low substrate turn over as the major obstacles for biocatalytic
reduction are also reported [7,8]. Ketoreductases are the group of enzymes involved in
catalyzing the asymmetric reduction (NADH/NADPH dependent) of prochiral ketones
to corresponding chiral alcohols [9-11]. Use of ketoreductase (KERDs) as catalyst is
advantageous over chemical method for the reduction of prochiral ketones due to its
high selectivity and environmentally benign nature [12-14]. Although there is a lack of
new ketoreductases emerging as biocatalysts, repurposing of existing KERDs have led
to novel biocatalytic processes. The catalyst repurposing is exemplified by the recent
work of enantioselective reduction of 2-chloro-1-(2,4-dichlorophenyl)-ethanone using
a KERD from Schefferomyces stipitis for the synthesis of (R)-2-chloro-1-(2,4-
dichlorophenyl)ethanol, an intermediate for the synthesis of several anti-fungal drugs
[15]. In recent years, generation of improved variants of KERDs through directed
evolution for better selectivity and improved yield have been reported [16,17]. Nitrile
reductase is a new family of protein involved in nitrile metabolism which converts
nitriles to amines [18]. The chemical reduction of nitriles to amines requires harsh
reaction conditions such as high pressure and temperature [19,20]. Nitrile reductase is
a new family of protein and an important bio-catalyst to develop green methodology
for nitrile reduction. The amines produced from the reduction of nitriles are valuable
drug intermediate for the pharmaceutical industry. There are few reports on this
unique biocatalyst and very less has been explored on nitrile reductase, in spite of
having tremendous potential for industrial application [18,21]. There is a need to
explore the microbial resources for screening of new biocatalysts with higher activity.
Metagenomic screening is one of most widely used approaches to retrieve novel

enzyme catalysts from the microbial genome diversity [22]. Homology-driven
screening is one of the approaches used very frequently which is based on the search
for genes coding for conserved protein sequences or motifs. In this case, the library is
screened for the genes of interest using an oligonucleotide probe or (in cases of low
homology) through the use of degenerate primers in conjunction with the polymerase
chain reaction, followed by sub-cloning, expression and characterization of the gene
products [23,24]. Here, a PCR based method was used for the generation of random
gene library on the basis of gene sequences available in the gene database followed by
sub-cloning, expression and characterization of the gene products. Strains with
oxidoreductase activity reported previously from our group such as Rhodococcus
ruber [25] and Candida parapsilosis [26,27] were selected for this study to generate
library and screening for novel KERDs. Similarly, a PCR based gene library was
generated from Bacillus subtilis genome for the screening of a potent nitrile reductase
as one of the strains being reported to have similar activity [28]. At the end of this
study, two novel reductases were obtained from the library with prochiral ketone and
nitriles as substrate. Both the biocatalysts showed activity over a broad range of
substrate. Nitriles being very tough substrate to catalyze, the reduction of nitriles to
amino alcohols by the novel nitrile reductase makes it an attractive biocatalyst for
industry. This study showed how mining of gene libraries could led to the generation
of novel biocatalysts.
Materials and Methods
Bacterial strains, plasmids and reagents
The wild type strains Rhodococcus ruber (MTCC-1827) and Bacillus subtilis (MTCC 441),
were procured from Microbial Type Culture Collection Center, Institute of Microbial
Technology, Chandigarh, India. For cloning and expression, E. coli Top10 and BL-21 strains
were used, respectively. Plasmid vector pET-15b was used for generating the library and
pET-30a was used for generating the GDH based cofactor recycling system. All the
restriction enzymes (XhoI and BamHI), T4 DNA ligase and other PCR related reagents were
purchased from Fermentas (Thermo Scientific, Germany). Primers were procured from IDT
(USA). All the molecular biology related kits were purchased from standard companies.
Growth conditions
Rhodococcus ruber (MTCC 1827) was grown at 30 ºC (200 rpm) for 48 h in a medium
having the following composition (g/L): malt extract 10, yeast extract 4, glucose 4 and
calcium carbonate 2. Bacillus subtilis and E. coli were grown in Luria-Bertani (LB) medium

containing tryptone 10 g/L, yeast extract 5 g/L and sodium chloride 10 g/L at 37 ºC (200
rpm) for 24 and 12 h, respectively.
Chemicals
Acetophenone (purity 99%) was procured from S.D Fine Chemicals Ltd. (Mumbai India).
(RS)-1-phenylethanol (purity 99%) and (R)-mandelonitrile (purity 99%) were purchased from
Sigma-Aldrich (USA). Reduced form of nicotinamide adenine dinucleotide (NADH) was
procured from HiMedia Laboratories Pvt. Ltd. (Mumbai India). Different buffer components
were obtained from Qualigens Chemicals, Mumbai, India. The reagent grade solvents such as
hexane, ethyl acetate, and DMSO etc. were procured from various commercial sources.
Solvents for HPLC (analytical grade) such as hexane, 2-propanol, and methanol were
obtained from J. T. Baker, USA.
Analytical methods
Reversed phase HPLC (Waters Alliance e2695 series) based analysis of (R)-mandelonitrile
and (R)-2-phenylethanolamine was performed using C18 column (250 x 4.80 mm, 5 μm,
Merck, Germany) with methanol:water::6:4 (0.5 mL/min) as the mobile phase and detected at
254 nm (PDA detector). (R)-2-phenylethanolamine and (R)-mandelonitrile were eluted at
8.077 and 11.1 min, respectively. The optical purity of the products was confirmed by
reversed phase chiral HPLC using amylose-2 column (250 x 4.60 mm, Lux 5 μm) with
methanol:water::6:4 (0.5 mL/min) as the mobile phase and detected at 254 nm (PDA
detector). (R)-2-phenylethanolamine and (R)-mandelonitrile were eluted at 6.55 and 13.39
min, respectively. The bioreduction of acetophenone to (S)-phenylethanol was analyzed by
normal phase chiral HPLC using ODH column (250 x 4.6 mm, 5 μm) with
hexane:isopropanol::9:1 at a flow rate of 1 mL/min as the mobile phase and detected at 254
nm. Standard acetophenone, (S)-phenylethanol and (R)-phenylethanol were eluted at 10.8,
13.1, 15.1 min, respectively. The products formed in both the synthesis were also
characterized by LTQ-MS and NMR.
Generation of PCR-based library
The library was generated by the homology based PCR using the genomic DNA isolated
(using Genomic DNA isolation Kit) from the strains of Rhodococcus ruber and B. subtilis.
Since, homology driven approach was used for generating the library, the primers were
designed based on the sequence available in the database for both the reductases. For
ketoreductase, the available sequences of oxidoreductase from Rhodococcus ruber and for
nitrile reductase, the available gene sequence from B. subtilis (see SI, Figure S8) was used.

To generate the library, gradient PCR reactions were performed in Eppendorf Mastercycler
using the following PCR protocol comprising initial denaturation step (95 °C, 5 min),
followed by 30 cycles consisting of a denaturation step at 95 °C for 1 min, an annealing step
at 55 °C with a gradient of 10 °C for 45 sec and an elongation step at 72 °C for 1 min. From
the gradient PCR, annealing temperature of 48 °C was found to be ideal for R. ruber and 55
°C for B. subtilis. Further PCR reactions were carried out at optimized temperature. Being a
homology based approach, a gene sequence library was obtained in the range of 500-750 bp
(library A) from R. ruber and 750-1000 bp (library B) from B. subtilis.
Cloning of library into expression vector
To get functional protein from the library, the amplified sequences were cloned into an
expression vector. The gene libraries and vector pET-15b were digested with XhoI and
BamHI (Fast digest, Fermentas, 37 °C, 4 h). To prevent self-ligation, double digested vector
was given CIAP (Calf intestinal alkaline phosphatase, 37 °C, 1 h) treatment. Double digested
vector and insert (vector:insert::1:5) were ligated using ATP-dependent T4 DNA ligase
(16 °C, 16 h). The ligation mixture was transformed into chemically-competent E. coli Top10
cells and the transformants were screened on LB-ampicillin plate [29]. To confirm successful
cloning, the constructs were double digested with XhoI and BamHI (Fast digest, 37 °C, 4 h).
Further, the vector-construct was transformed into E. coli BL-21. After transformation, E.
coli BL-21 cells were placed on Amp-LB-agar plate and incubated at 37 °C to check for
positive clones. Further, the cloning was confirmed by carrying out PCR amplification using
the construct, isolated from the positive clones.
Expression of the cloned libraries
For the initial experiments, the positive clones were grown in Amp-LB medium (37 °C, 200
rpm) till 0.6 OD600, followed by induction with 1 mM IPTG. After 6 h post induction, the
cells were harvested and the expression profile was analyzed. To achieve the optimum
expression in the soluble fraction, the effect of IPTG concentrations (0.5, 1 mM), temperature
(20, 30 and 37 °C) and induction time on overexpression of the hypothetical proteins were
studied. The samples were analyzed by SDS-PAGE and NADH assay.
SDS-PAGE analysis
Protein sample of the recombinant strains were checked for expression profile. Wild type
strains were used as control for comparison of protein profile. SDS-PAGE 15% (w/v) was
used for the analysis following the standard protocol [30]. The cells were lysed to get the
cytosolic fraction. The lysis buffer contains 50 mM Tris buffer, 150 mM NaCl, 1 mM PMSF,

17.5 mM SDS and 1 mM EDTA. Broad range protein ladder (6.5-200 kDa) was used as the
marker. The gel was stained with Coomassie blue and destained following the standard
protocol to visualize the bands.
Western blotting
The bands from SDS-PAGE were transferred on to nitrocellulose membrane for 1 h in semi-
wet transfer unit. Ponceau S staining was used to confirm the transfer of protein bands into
the membrane. To prevent non-specific binding, blocking was done with 5% BSA for 3 h at
room temperature followed by 3-times washing. The membrane was then incubated (12 h, at
4 °C) with primary antibody [SIGMA monoclonal anti-polyhistidine antibody (1:5000)] in
TBS buffer. After overnight incubation, the membrane was washed and treated with
secondary antibody for 2 h. The alkaline phosphatase bound secondary antibody [SIGMA
anti-mouse IgG (Fc specific)] catalyzes colorimetric reactions using BCIP/NBT as substrate
resulting in dark- purple precipitate. The presence of desired His-tagged protein in the sample
will show dark-purple band over the membrane.
NADH assay
The recombinant strains were grown in 50 mL Amp-LB media at 37 °C for 6 h with
induction at optimized condition. Cells were harvested by centrifugation at 7000 x g for 10
min and resuspended in lysis buffer (50 mM Tris buffer of pH 8.0, 150 mM NaCl, 0.4 mM
dithiothreitol, 1.0 mM PMSF, 1 mM EDTA and 17.5 mM SDS). Further, the cells were
disrupted using a probe sonicator and the homogenate was centrifuged (15 min, 4 °C, 10000
x g). The supernatant was collected and used for the assay. The reaction mixture (0.250 mL)
contained 10 mM substrate, 1 mM NADH, 50 mM phosphate buffer (pH 6.5) and cell extract
with 0.1 mg protein. Substrate blank, enzyme blank and NADH blank were used as negative
control in the assay. Reductase activity was determined by measuring the decrease in
absorbance at 340 nm [31,32]. One unit of reductase activity was defined as µmol of NADH
disappeared per minute per mg protein. All activity measurements were carried out in
triplicate along with the controls. The absorbance of the samples was normalized against the
control in each run.
Computational modeling of the novel catalysts
The cloned fragments were sequenced through IDT sequencing service (See SI, Figure S7).
The positive sequences were analyzed by BLAST and they were found to be hypothetical
proteins with no putative sequence homology. No putative homology of the two hypothetical
proteins in the database renders them as novel biocatalysts. In order to obtain the protein
sequence, the gene sequence is converted to reading frames using ExPASY tool. Being a

hypothetical protein, online server LOMETS (Local Meta Threading Server) has been used to
generate homology model [33]. It is an online web service for prediction of protein structure.
It generates 3D-models by collecting high scoring target to template alignments from 9
locally installed threading programmes (FFAS-3D, HHsearch, MUSTER, pGenTHREADER,
PPAS, PRC, PROSPECT 2, SP3 and SPARKS-X). The pros of using LOMETS are that it
gives the best 10 threading models selected from 160 models by the confidence score and top
10 target-template alignments of individual threading servers. It predicts spatial C-alpha and
side chain contact, distance map and generates full length models by MODELLER guided by
consensus restraints. The homology model developed for both the novel catalyst has been
used for in-silico screening using Swiss Dock Server with different carbonyl and nitrile
compounds [34,35]. The docking study was utilized to depict the binding interaction of the
novel enzymes with different substrates based on H-bond length and lowest binding energy.
The amino acid residues involved in enzyme-ligand interaction were analyzed using
Chimera. Similarly, the binding of NADH with both the enzymes was analyzed by docking.
At the end of the study, both the novel sequences were submitted to the NCBI database with
the gene ID, MH121608 (500-1) and MH135302 (Bac-2) respectively.
General reaction protocol for biocatalytic synthesis of chiral amines
For the whole cell biocatalysis, cell suspension (200 mg/mL) of recombinant strains prepared
in phosphate buffer (50 mM, pH 7.8) was used. The bioreduction reaction mixture was
consisted of cell suspension (1 mL), substrate (70 mM) dissolved in DMSO and 5% glucose
as cofactor recycling substrate [36]. The reaction was performed at 37 °C in an orbital shaker
(200 rpm) for 24 h. To recover the product, solvent extraction using ethyl acetate was used.
The product was further characterized by reversed phase HPLC, Mass spectrometry and
NMR.
Generation of cofactor recycling system
Reductase requires NADH as cofactor in the biotransformation reaction. In an attempt to
improve the biotransformation by incorporating a cofactor recycling machinery, glucose
dehydrogenase gene (GDH) from B. subtilis was co-expressed in the same cells of E. coli
strain. GDH gene was PCR amplified from B. subtilis using respective primers (see
supporting information S6) and cloned into E. coli Top10 cells using pET30a vector. After
confirmation of successful cloning of GDH, the construct was transformed into chemically
competent E. coli BL21 cells with overexpression of 500-1 and Bac-2. The resulting dual
system, whole cell biocatalyst showed improved substrate utilization at an increased
concentration.

Optimization of process parameters for the biocatalytic reaction
Various reaction parameters of both the recombinant biocatalysts were optimized to get
maximum yield of the chiral alcohols. To study the interrelation among four important
parameters (temperature, pH, substrate concentration and time) a series of reactions were
designed. The aim was to find out the best combination of pH, temperature, substrate
concentration and reaction time. The physico-chemical parameters have profound effect on
the reaction kinetics and affect the biocatalyst activity and stability. The biocatalytic reaction
was optimized by selecting a pH range from 5 to 9, temperature from 25 to 50 °C. The
substrate concentration was varied from 30 to 150 mM. Reaction time was varied from 6 to
72 h. Samples were analyzed by HPLC.
Synthesis of (R)-2-phenylethanolamine from (R)-mandelonitrile using the novel
biocatalyst
Whole cell mediated biotransformation of (R)-mandelonitrile was carried out using the novel
nitrile reductase (Bac-2 over expressed in E. coli BL21). The reaction mixture contained cell
suspension (200 mg/mL) in 50 mM phosphate buffer (pH 7.8) and 100 mM substrate [(R)-
mandelonitrile)] dissolved in DMSO. The biocatalysis reaction was carried out in an
incubator shaker (37 °C, 200 rpm) for 24 h. The cells were removed by centrifugation (7000
x g) and the supernatant was extracted with ethyl acetate to recover the product. The
recovered product was characterized as (R)-2-phenylethanolamine by IR, NMR, mass
spectrometry and reversed phase HPLC. White solid, 1H NMR (400 MHz, CDCl3): 7.38 (d, J
= 4 Hz, 4H), 7.33 (m, 1H), 4.68 (m, 1H), 3.05 (dd, J = 8 Hz, 1H), 2.86 (dd, J = 4 Hz, 1H); IR
(KBr): tmax (cm-1) 3418, 2918, 1652, 1019, 1406, 953, 709; calculated mass: 137.08, observed
mass: 137.07.
Synthesis of phenylethanol from acetophenone using novel KERD
Whole cell mediated bioconversion of acetophenone to (S)-phenylethanol was carried out
using the novel KERD (500-1 over expressed in E. coli BL21). The reaction mixture was
consisted of cell suspension (200 mg/mL) in 50 mM phosphate buffer (pH 7.8) and 120 mM
substrate (acetophenone) dissolved in DMSO. The biocatalytic reaction was carried out at
37 °C (200 rpm) for 24 h. After completion of the reaction, the cells were removed by
centrifugation (7000 x g) and the supernatant was extracted with ethyl acetate to isolate the
product. The recovered product was characterized as (S)-phenylethanol by NMR, mass
spectrometry and reversed phase chiral HPLC. Oily liquid,1H NMR (400 MHz, MeOD):

7.36-7.28 (m, 4H), 7.23-7.19 (m, 1H), 4.83-4.78 (m, 1H), 3.30-3.29 (t, J = 1.5 Hz, 1H), 1.43-
1.41 (d, J = 6.52 Hz, 3H); calculated mass: 122.16, observed mass: 122.35.
Results and Discussion
Generation of novel KERD and Nitrile reductase from PCR based library
Random homology based gene library was generated from the genome of Rhodococcus ruber
and Bacillus subtilis using PCR with primers designed on the basis of available sequence in
the database of these enzymes from different strains. From the gel picture, it is evident that
there was a formation of amplified sequences ranging from 500 bp to 1500 bp (See SI, Figure
S1). Sequences from 500-750 bp, 750-1000 bp and 1000-1500 range were cloned into
pET15b vector and transformed into E. coli Top 10 cells. The positive transformants were
selected on Amp-LB plate. Post confirmation, the construct was transformed into E. coli
BL21 cells for the expression of novel proteins from the library. For the optimum expression
of the proteins, induction conditions (IPTG concentration, temperature and induction time)
were optimized. Though, under increased IPTG concentration (from 0.25 to 1.0 mM) no
significant difference in the activities was found, however, 1 mM IPTG has the highest
expression over SDS-PAGE. Higher incubation temperature during over-expression of the
desired protein may result into inclusion body formation (pellet fraction) which causes loss of
enzyme activity. Hence, the temperature dependent expression profile was studied between
20 to 37 °C with 1 mM IPTG. It was seen that 30 °C was ideal for 500-1 gene expression
whereas 37 °C was the best for Bac-2 gene expression. Effect of induction time was studied
at 1 mM IPTG by incubating the cultures at 30 °C (500-1) and 37 °C (Bac-2) for 2 to 6 h.
Significant difference in the activities was found under different set of conditions. Incubation
time of 6 h was found to be ideal for the expression of both the catalysts under the optimized
condition (See SI, Figure S2).
Screening of novel biocatalysts for reductase activity
The substrate specificity of the novel proteins was determined for reductase activity against
various carbonyl and nitrile substrates. A broad substrate range consisting of both alkyl and
aromatic moieties was selected for this purpose. Crude protein samples were used to screen
the substrates with NADH as the cofactor. The crude cell lysate from wild type E. coli BL-21
cells was used as the negative control. It is evident that the novel KERD 500-1 has the
highest specificity for acetophenone, while it is showing minimal activity with various other
ketones (Table 1).
< Table 1 >

Similarly, the other novel catalyst, Bac-2 was screened against various nitriles and found to
have comparatively higher specificity for mandelonitrile as the substrate (Table 2). Similar
pattern of activity with other nitrile substrates indicates its broad substrate range.
< Table 2 >
Computational modeling of hypothetical KERD and Nitrile reductase
To depict the structure activity relationship, homology model for both the hypothetical
proteins was generated based on the sequence (Figure 1). The Ramachandran plot generated
against the homology model of the novel KERD (500-1) showed that 80.2% residues are in
the most favored region (A, B, L). G-factor score of -0.23 (not below-0.5) from the
PROCHECK analysis indicates that the structural parameters of the model are in the usual
range. The NADH dependence of the novel KERD for activity was validated through its
binding to NADH (Arg124, His277, His278, and Gln279) (Figure 1). Similarly, the
Ramachandran plot of the novel nitrile reductase (Bac-2) depicts that 87.2% amino acids are
in the most favored regions (A, B, L). Validation of the model was performed by
PROCHECK analysis with G-factor score of -0.22 indicating that the structural properties are
in usual range. The role of NADH in novel NR activity was confirmed through its interaction
(Ser 245 and Arg 244) with the enzyme (Figure 2).
To validate the assay based screening with computational data, an in-silico screening with
eleven different carbonyl substrates (both aromatic and aliphatic) was performed for the
novel KERD (Table 3) (See SI, Figure S3). Among them, acetophenone showed the ideal
binding (Thr39, 2.052 Å) with shortest H-bond length (Figure 1), which also supports the
earlier screening data. Hence, acetophenone was selected as model substrate for further
studies. The role of threonine residue in the catalytic site of ketoreductase is previously being
mentioned in literature [37].
< Figure 1 >
< Table 3 >
Likewise, an in-silico screening for nitrile reductase activity was done with ten different
nitriles (both aromatic and aliphatic) (See SI, Figure S4). It is evident from the docking study
that the novel nitrile reductase has a broad substrate range (Table 4). Among them,
mandelonitrile was selected as the model substrate for further studies. Docking of
mandelonitrile reveals that Arg181 is the key residue involved in catalysis (Figure 2). Role of
arginine residue for nitrile reductase activity is also mentioned in previous literature [38].
< Figure 2 >

< Table 4 >
Generation of co-factor recycling system
Reductase activity is solely dependent on cofactors (NAD+/NADH, NADP+/NADPH) and
the continuous supply of cofactors is very essential for the conversion. A dual catalyst with
co-factor recycling system was developed by overexpressing glucose dehydrogenase (GDH,
from B. subtilis) in E. coli, along with KERD and nitrile reductase. The over expression of
both the proteins was confirmed by western blotting (Figure 3). Further, the dual catalyst was
used for the synthesis of enantiopure alcohols and amino alcohols.
< Figure 3 >
Biocatalytic synthesis of (S)-phenylethanol and (R)-2-phenylethanolamine using the
novel dual enzyme catalyst
The novel catalyst with cofactor recycling system was used for the synthesis of (S)-phenyl
ethanol and (R)-2-phenylethanolamine, two important pharmaceutical intermediates. (S)-
selective reduction of acetophenone was performed using the whole cells of the dual catalysts
in the presence of glucose, as cofactor recycling substrate with 70 mM acetophenone
(substrate for the novel KERD) (Scheme 1A).
After 24 h of reaction, ~70% conversion with enantiomeric excess of > 99% was observed. It
signifies that the novel KERD has higher enantio-selectivity and the cofactor recycling
system has worked efficiently to replenish adequate amount of NADH into the catalytic
system. Similarly, (R)-mandelonitrile (70 mM) was reduced to (R)-2-phenylethanolamine
using the novel nitrile reductase with 80% conversion using the GDH dependent cofactor
recycling system (Scheme 1B).
Determination of various reaction parameters
To achieve the maximum yield, effect of various physico-chemical parameters was
studied. For optimization experiments, one parameter at a time approach was used.
The effect of reaction time on the biocatalytic conversion was studied to obtain the
best yield in the synthesis of chiral alcohols. The course of reduction using 500-1 as
catalyst revealed that at 36 h, the highest conversion of 93.33% for acetophenone was
obtained which led to the formation of the product, (S)-1-phenyl ethanol with 99.5%

enantiomeric excess (Figure 4a). Hence, for further experiments, reaction time of 36 h
was selected.
< Figure 4 >
The Bac-2 mediated reduction of mandelonitrile showed the highest conversion (>
95%) at 48 h towards the formation of (R)-phenylethanolamine (Figure 5a). With
further increase in reaction time, there was no change on the percentage conversion.
Hence for all the subsequent experiments, bioreduction was carried out for 48 h using
the novel nitrile reductase Bac-2. The ideal reaction pH for the novel KERD (500-1)
and nitrile reductase (Bac-2) was found to be pH 7.0 and pH 8.0, respectively, with the
highest conversion (Figures; 4d, 5d).
< Figure 5 >
The optimum pH of nitrile reductase has also been reported in literature to be above
7.5. The reaction temperature of 30 °C was found to be optimum for (S)-phenylethanol
production using the novel KERD (500-1) with 94% conversion, whereas the highest
conversion of 98% with (R)-mandelonitrile as substrate using Bac-2 was observed at
48 °C (Figures 4c, 5c). From the data, it is evident that acetophenone concentration of
125 mM was optimum for the whole cell catalysis using the novel KERD 500-1 and
further the yield gradually decreased with the increasing substrate concentration
(Figure 4b). Hence, all the subsequent reactions were carried out with 125 mM
acetophenone. Similarly for the novel nitrile reductase (Bac-2), it was observed that
the bioconversion of (R)-mandelonitrile to (R)-2-phenylethanolamine was maximum
with 100 mM substrate.
The novel enzymes (500-1 and Bac-2) generated from PCR based library, demonstrated their
catalytic potential for the bioreduction of ketone and nitriles. One of the novel catalysts, Bac-
2 showed bioconversion of mandelonitrile to (R)-2-phenylethanolamine, which is the
precursor for several drugs and having great pharmaceutical importance. It is used as the
precursor to the various β-blocker cardiovascular and catecholamine nervous system drugs.
The (R)-enantiomer has an anti-asthma activity and it is an important building block for the
synthesis of salbutamol, clenbuterol (anti-asthmatic drugs) etc. The other novel catalyst, 500-
1 also showed improved reduction of acetophenone to (S)-phenylethanol, another very useful
drug intermediate. Both the higher substrate conversion and selectivity by both the novel
catalysts adds to their industrial potential.
Synthesis of (S)-phenylethanol using novel KERD

Once the catalytic conditions were optimized, a preparative scale synthesis of (S)-1-phenyl
ethanol was carried out under the optimized condition using the novel KERD. The new
KERD with co-factor recycling system showed 90% conversion and > 99% enantiomeric
excess for 120 mM substrate. The product was characterized by chiral HPLC, mass and NMR
(See SI, Figure S5). The novel KERD reported here also showed better activity (conversion
and ee) over the recently reported ketoreductase (with 89% conversion and enantiomeric
excess of 98% for 40 mM substrate) [39].
Synthesis of (R)-2-phenylethanolamine using novel nitrile reductase
Preparative scale enzymatic synthesis of (R)-2-phenylethanolamine was carried out from (R)-
mandelonitrile using the novel biocatalyst under the optimized condition. A conversion of >
99% with > 99% enantiomeric excess was achieved with 100 mM substrate. The product was
recovered and characterized by chiral HPLC, mass and NMR (See SI, Figure S6). In one of
the recent studies on biocatalytic synthesis of chiral amino alcohols using novel catalyst, 70%
conversion and enantiomeric excess of 98% was reported with 60 mM substrate which is
much lower in comparison to the novel catalyst reported here [30].
Conclusion
From the homology based PCR dependent gene library developed using R. ruber and B.
subtilis genome, two potent hypothetical proteins with significant reductase activity towards
different substrates were screened out. The unique sequence of both the hypothetical neoteric
enzymes speaks for their novelty. Both the in-vitro and in-silico screening of enzyme activity
were performed against various substrates using both the novel catalysts. The novel KERD
showed broad substrate range and higher selectivity. Asymmetric reduction of acetophenone
to (S)-phenylethanol was optimized for 120 mM substrate, with > 90% conversion and > 99%
enantiomeric excess. Likewise, asymmetric reduction by the novel nitrile reductase was
shown using (R)-mandelonitrile as the model substrate. The novel reductases also showed
activity over a broad substrate range of both alkyl and aromatic moieties. Under optimized
condition, > 99% conversion with 100 mM (R)-mandelonitrile was achieved and (R)-2-
phenylethanolamine was synthesized as the product. The novelty of this nitrile reductase is
further resonated through the conversion of such higher concentration of substrate which is
rare in this group of catalysts. Both the novel catalysts have shown significant activities
which make them potential candidate for industrial catalysis.

Acknowledgements
PS acknowledges the funding support by Department of Biotechnology (DBT) project no.
396 (BT/PR4807/PID/6/645/2012), Govt. of India to carry out this work. SK is supported by
Department of Biotechnology (DBT, Govt. of India) for this project. SG acknowledges
Indian Council of Medical Research (ICMR, Govt. of India) for funding to work under this
project.

References
[1] A. Schmid, J. S. Dordick, B. Hauer, A. Kiener, M. Wubbolts and B. Witholt, Industrial
biocatalysis today and tomorrow, Nature. 409 (2001) 258-268.
[2] L. Marx, N. R. Lombardia, J. Farnberger, W. Kroutil, A. I. B. Mateos, F. L. Gallego, F.
Moris, J. G. Sabin, and P. Berglund, Chemoenzymatic approaches to the synthesis of the
calcimimetic agent cinacalcet employing transaminases and ketoreductases, Adv. Synth.
Catal. (2017) doi: 10.1002/adsc.201701485.
[3] M. M. Sheeba, M. M. Tamizh, L. J. Farrugia, R. Karvembu, Asymmetric hydrogenation
of pro-chiral ketones catalyzed by chiral Ru(II)-benzene organometallic compounds
containing amino acid based aroylthiourea ligands, J. Organomet. Chem. (2016)
doi:10.1016/j.jorganchem.2016.12.016.
[4] D. Milstein, A. Mukherjee, D. Srimani, and Y. Ben-David, Low-pressure hydrogenation
of nitriles to primary amines catalyzed by ruthenium pincer complexes scope and mechanism,
ChemCatChem. (2016) doi: 10.1002/cctc.201601416.
[5] S. Enthaler, K. Junge, D. Addis, G.Erre, and M. Beller, A practical and benign synthesis
of primary amines through ruthenium-catalyzed reduction of nitriles, ChemSusChem 1
(2008) 1006-1010.
[6] M. Yu Jin, J. Li, R. Huang, Y. Zhou, L. Wa Chung and J. (Joelle) Wang, Catalytic
asymmetric trifluoromethylthiolation of carbonyl compounds via a diastereo-and
enantioselective cu-catalyzed tandem reaction, Chem. Commun. (2018) DOI:
10.1039/C8CC02097C.
[7] a) M. Wills, J. Hannedouche, New methodology for the asymmetric reduction of ketones,
Curr. Opin. Drug Discov. Devel. 5 (2002) 881-891. b) R. Noyori, Asymmetric catalysis in
organic synthesis, first ed., John Wiley & Sons, New York, 1994.
[8] a) R. Noyori, M. Kitamura, T. Ohkuma, Toward efficient asymmetric hydrogenation:
Architectural and functional engineering of chiral molecular catalysts, Proc. Natl. Acad. Sci.
U.S.A. 101 (2004) 5356-5362. b) R. Noyori, Chiral metal complexes as discriminating
molecular catalysts, Science. 248 (1990) 1194-1199. c) C.A. Sandoval, T. Ohkuma, K.
Muniz, R. Noyori, Mechanism of asymmetric hydrogenation of ketones catalyzed by

binap/1,2-diamine-ruthenium(ii) complexes, J. Am. Chem. Soc. 125 (2003) 13490-13503. d)
R. V. Hoffman, N. Maslouh, F. Cervantes-Lee, Highly stereoselective syntheses of syn- and
anti-1,2-amino alcohols, J. Org. Chem. 67 (2002) 1045-1056. e) E. J. Corey, J. O. Link, A
new chiral catalyst for the enantioselective synthesis of secondary alcohols and deuterated
primary alcohols by carbonyl reduction, Tetrahedron Lett. 30 (1989) 6275-6278.
[9] M. Hall, A. S. Bommarius, Enantioenriched compounds via enzyme-catalyzed redox
reactions, Chem. Rev. 111 (2011) 4088-4110.
[10] D. Monti, G. Ottolina, G. Carrea, S. Riva, Redox reactions catalyzed by isolated
enzymes, Chem. Rev. 111 (2011) 4111-4140.
[11] S. M. D. Wildeman, T. Sonke, H. E. Schoemaker, O. May, Biocatalytic reductions: from
lab curiosity to “first choice”, Acc. Chem. Res. 40 (2007) 1260-1266.
[12] G. W. Huisman, J. Liang, A. Krebber, Practical chiral alcohol manufacture using
ketoreductases, Curr Opin Chem Biol. 14 (2010) 122-129.
[13] T. Matsuda, R. Yamanaka, K. Nakamura, Recent progress in biocatalysis for asymmetric
oxidation and reduction, Tetrahedron: Asymmetry. 20 (2009) 513-557.
[14] U. Bornscheuer, G. Huisman, R. Kazlauskas, S. Lutz, J. Moore, K. Robins, Engineering
the third wave of biocatalysis, Nature. 485 (2012) 185-194.
[15] Y. Shang, Q. Chen, X. Kong, Y. Zhang, J. Xu, H. Yu, Efficient synthesis of (R)-2-
Chloro-1-(2,4-dichlorophenyl)ethanol with a ketoreductase from Scheffersomyces stipitis
CBS 6045, Adv. Synth. Catal. 359 (2017) 426-431.
[16] F. Qin, B. Qin, T. Mori, Y. Wang, L. Meng, X. Zhang, X. Jia, I. Abe, S. You,
Engineering of Candida glabrata ketoreductase 1 for asymmetric reduction of α‑halo
ketones, ACS Catal. 6 (2016) 6135-6140.
[17] G. Zheng, Y. Liu, Q. Chen, L. Huang, H. Yu, W. Lou, C. Li, Y. Bai, A. Li, J. Xu,
Preparation of structurally diverse chiral alcohols by engineering ketoreductase CgKR1,
ACS Catal. 7 (2017) 7174-7181.
[18] P. D. de Maria, Nitrile reductase: a forthcoming wave in biocatalysis? ChemCatChem. 3
(2011) 1683-1685.

[19] S. Gowda, D. C. Gowda, Application of hydrazinium monoformate as new hydrogen
donor with raney nickel: a facile reduction of nitro and nitrile moieties, Tetrahedron. 58
(2002) 2211-2213.
[20] C. Gunanathan, M. Holscher, W. Leitner, Reduction of nitriles to amines with
h2 catalyzed by non-classical ruthenium hydrides-water-promoted selectivity for
primary amines and mechanistic investigations, Eur. J. Inorg. Chem. 22 (2011) 3381-
3386.
[21] L. Yang, S. L. Koh, P. W. Suttonb, Z. X. Liang, Nitrile reductase as a biocatalyst:
opportunities and challenges, Catal. Sci. Technol. 4 (2014) 2871-2876.
[22] U. T. Bornscheuer, G. W. Huisman, R. J. Kazlauskas, S. Lutz, J. C. Moore, K.
Robins, Engineering the third wave of biocatalysis, Nature. 485 (2012) 185-194.
[23] G. A. Behrens, A. Hummel, S. K. Padhi, S. Schtzle, U. T. Bornscheuer, Discovery and
protein engineering of biocatalysts for organic synthesis, Adv. Synth. Catal. 353 (2011)
2191-2215.
[24] P. J. Bell, A. Sunna, M. D. Gibbs, N.C. Curach, H. Nevalainen, P. L. Bergquist,
Prospecting for novel lipase genes using PCR, Microbiology. 148 (2002) 2283-2291.
[25] E. Hamnevik, C. Blikstad, S. Norrehed, M. Widersten, kinetic characterization of
Rhodococcus ruber DSM 44541 alcohol dehydrogenase A, J. Mol. Catal. B: Enzymatic.
99 (2014) 68-78.
[26] S. M. Amrutkar, L. Banoth, and U. C. Banerjee, One-pot synthesis of (R)-1-(1-naphthyl)
ethanol by stereoinversion using Candida parapsilosis, Tetrahedron Lett. 54 (2013)
3274-3277.
[27] S. Ghosh, L. Banoth, U. C. Banerjee, Biocatalytic deracemization: An efficient one-pot
synthesis of (R)-α-methyl-4-pyridinemethanol using whole cells of Candida
parapsilosis, Biocatalysis. 1 (2015) 59-66.
[28] W. K. Lee, S. G. Van Lanen, D. Iwata-Reuyl, Mechanistic studies of Bacillus subtilis
QueF, the nitrile oxidoreductase involved in queuosine biosynthesis, Biochemistry. 46
(2007) 12844-12854.

[29] J. Sambrook, E. F. Fritsch, T. Maniatis, Molecular cloning. A laboratory manual, cold
spring harbor laboratory press, New York, (1989) 1626.
[30] S. Wu, Y. Zhou, T. Wang, H. P. Too, D. I. C. Wang, Z. Li, Highly regio- and
enantioselective multiple oxy- and amino-functionalizations of alkenes by modular
cascade biocatalysis, Nat. Commun. 7 (2016) 11917-11930.
[31] P. Soni, U. C. Banerjee, Enantioselective reduction of acetophenone and its derivatives
with new yeast isolate Candida tropicalis PBR MTCC 5158, Biotechnol. J. 1 (2006) 80-
85.
[32] B. Vallee, F. Hoch, Zinc: A component of yeast alcohol dehydrogenase, Proc. Natl.
Acad. Sci. USA. 41 (1955) 327-338.
[33] S. Wu, Y. Zhang, LOMETS: A local meta-threading-server for protein structure
prediction, Nucleic Acids Res. 35 (2007) 3375-3382.
[34] Z. Bikadi, E. Hazai, Application of the PM6 semi-empirical method to modeling
proteins enhances docking accuracy of AutoDock, J. Cheminform. 1 (2009) 1-15.
[35] R. Huey, G. M. Morris, A. J. Olson, D. S. Goodsell, A semi-empirical free energy force
field with charge-based desolvation, J. Comput. Chem. 28 (2007) 1145-1152.
[36] M. Singh, S. Singh, S. Deshaboina, H. Krishnen, R. Lloyd, K. Holt-Tiffin, A.
Bhattacharya, R. Bandichhor, Asymmetric reduction of a key intermediate of
eslicarbazepine acetate using whole cell biotransformation in a biphasic medium, Catal.
Sci. Technol. 2 (2012) 1602-1605.
[37] C. Filling, K. D. Berndt, J. Benach, S. Knapp, T. Prozorovski, E. Nordling, R.
Ladenstein, H. Jornvall, U. Oppermann, Critical residues for structure and catalysis in
short-chain dehydrogenases/reductases, J. Biol. Chem. 277 (2002) 25677-25684.
[38] Y. Kim, M. Zhou, S. Moy, J. Morales, M.A. Cunningham, A. Joachimiak, High-
resolution structure of the nitrile reductase QueF combined with molecular simulations
provide insight into enzyme mechanism, J. Mol. Biol. 404 (2010) 127-137.

[39] F. D. Nasario, T. Cazetta, P. J. S. Moran, J. A. R. Rodrigues, Deracemization of 1-
phenylethanol via tandem biocatalytic oxidation and reduction, Tetrahedron:
Asymmetry. 27 (2016) 404-409.

List of Figures
Figure 1. (a) Homology model of novel KERD; (b) Ramachandran plot of the model;
(c) Binding interaction of NADH to the enzyme; (d) Docking study showing
interaction of the catalyst with acetophenone as substrate.
Figure 2. (a) Homology model of novel nitrile reductase; (b) Ramachandran plot of the
model; (c) Binding interaction of NADH to the enzyme; (d) Docking study
showing interaction of the catalyst with mandelonitrile as substrate.
Figure 3. Gel images for dual system protein expression. (a) SDSPAGE showing expression
of dual system catalyst. Lane 1: Uninduced dual system, Lane 2: Induced 500-1,
Lane 3: Induced Bac-2, Lane 4: Induced dual system for 500-1, Lane 5: Induced
dual system for Bac-2, Lane 6: Protein marker. (b) Western blot showing distinct
spots indicating overexpression of both the hypothetical proteins along with GDH
overexpression as dual catalyst system.
Figure 4. (a) Course of reaction for the bioconversion of acetophenone to (S)-phenylethanol
using the novel KRED, (b) Effect of substrate concentration on the bioconversion
of acetophenone to (S)-phenylethanol using the novel KRED, (c) Effect of
temperature on the bioconversion of acetophenone to (S)-phenylethanol using the
novel KRED, (d) Effect of pH on the bioconversion of acetophenone to (S)-
phenylethanol using the novel KRED.
Figure 5. (a) Course of reaction for the bioconversion of (R)-mandelonitrile to (R)-phenyl-
ethanolamine using the novel biocatalyst, (b) Effect of substrate concentration on
the bioconversion of (R)-mandelonitrile to (R)-phenyl-ethanolamine using the
novel biocatalyst, (c) Effect of temperature on the bioconversion of (R)-
mandelonitrile to (R)-phenylethanolamine using the novel biocatalyst, (d) Effect of
pH on the bioconversion of (R)-mandelonitrile to (R)-phenylethanolamine using
the novel biocatalyst.

(a)
(c)
(d)
(b)
Figure 1.

(c)
(a)
(b)
(d)
Figure 2.

(a)
(b)
Figure 3.

Figure 4.

Figure 5.

Scheme 1. (A) Biocatalytic synthesis of (S)-phenylethanol by the selective asymmetric
reduction of acetophenone using the novel KERD-GDH dual catalyst. (B) Bioconversion of
(R)-mandelonitrile to (R)-2-phenylethanolamine using the novel nitrile reductase-GDH dual
catalyst

List of Tables
Table. 1 Substrate screening for ketoreductase activity of 500-1 proteinc
Enzyme activityb
Substratea (ketones)
(µmol/min) (n=3)
Acetophenone
87900 ± 540
No activity
24300 ± 280
25200 ± 670
No activity
4-acetylphenyl acetate
1-(pyridin-3-yl)ethanone
1H-indene-1,3(2H)-dione
Hydroxytetralone
1-(3,5-bis(trifluoromethyl)phenyl)ethanone 24400 ± 740
6-methoxy 2-acetonaphthone
1-(4-chlorophenyl)ethanone
1-(4-bromophenyl)ethanone
Acetonapthone
9700 ± 420
25100 ± 1150
24900 ± 860
No activity
4-Hydroxyacetophenone
35600 ± 940
aSubstrate concentration was 10 mM for the activity screening
bUnit enzyme activity was defined as the amount of protein required to convert 1 µmol
substrate to product per unit time (U/mg enzyme).
cReactions were carried out in 96 well plate using protein isolated from the recombinant
strain

Table. 2 Substrate screening for nitrile reductase activity of Bac-2 proteinc
Enzyme activityb
Substratea (Nitriles)
(µmol/min) (n=3)
Mandelonitrile
29860 ± 437
26070 ± 1101
27090 ± 455
27030 ± 472
27800 ± 529
27890 ± 586
27100 ± 507
25000 ± 612
25800 ± 1409
27800 ± 1076
26470 ± 726
Benzonitrile
Isonicotinonitrile
1H-indole-5-carbonitrile
2-(thiophen-2-yl)acetonitrile
3-Phenylpropanenitrile
Piperonylonitrile
4-Methoxybenzonitrile
Valeronitrile
Isobutyronitrile
Pyrazinecarbonitrile
aSubstrate concentration was 10 mM for the activity screening
bUnit enzyme activity was defined as the amount of protein required to
convert 1 µmol substrate to product per unit time (U/mg enzyme).
cReactions were carried out in 96 well plate using protein isolated from
the recombinant strain