the potential of the enzyme. In recent years, more research-
ers have been paying attention to the bioreduction approach
of producing (R)-3-quinuclidinol, since biocatalyzed asym-
metric reductions can offer highly selective reactions, environ-
Table 1. Asymmetric Reduction of 3-Quinuclidinone HCl with
Coexpressed Cells of E. coli (pET28a-ArQR-BmGDH)
a
7
c
mentally benign processes, and energy-effective operations.
8
RrQR from Rhodotorula rubra, QNR and BacC from
substrate
(g/L)
cell
NAD
þ
time
(h)
ee
c
entry
(g/L)
(mM)
conv (%)
(%)
9
Microbacterium luteolum JCM 9174, and Nocardia sp.
1
2
3
4
5
6
81
162
162
242
242
242
10
10
10
10
10
5
0.1
0.1
0
0.5
1.5
4.0
20
100
100
100
54
>99
>99
>99
>99
>99
>99
1
0
WY1202 can convert 3-quinuclidinone stereospecifically
to (R)-3-quinuclidinol, but the common disadvantages of
all these reductases were their poor substrate tolerance and
low volumetric productivity, which limits their industrial
application.
0
b
0.1
0.1
4.5
10
100 (90)
b
>99 (90)
a
Herein, a new reductase ArQR (GenBank accession no.:
YP_002542435.1) was identified from Agrobacterium
radiobacter ECU2556 through the screening of various
microorganisms stocked in our labratory. After heter-
ologous expression in E. coli, it can efficiently convert
Reaction conditions: 3-quinuclidinone HCl (0.8ꢀ2.4 g), dry cells of
E. coli BL21 harboring pET28a-ArQR-BmGDH (0.05ꢀ0.1 g), D-glucose
þ
(
1.5 equivalent), NAD (0ꢀ1 μmol), 10 mL phosphate buffer (200 mM,
b
pH 7.0), 30 °C. pH was kept at 7.0 with 2 M NaOH. Isolated yield of
(R)-3-quinuclidinol. Determined by GC analysis.
c
3
and excellent enantioselectivity.
-quinuclidinone to (R)-3-quinuclidinol with high activity
internal cofactor of the E. coli cells could probably not meet
the need of high substrate loading.
To construct the cofactor regeneration system, ArQR and
BmGDH (glucose dehydrogenase from Bacillus megaterium)
were coexpressed in E. coli in a tandem mode. Considering
that the reduction of 3-quinuclidinone was a rate-limiting
step comparing with the oxidation step of glucose, E. coli
BL21 (DE3) (pET28a-ArQR-BmGDH), rather than E. coli
BL21 (DE3) (pET28a-BmGDH-ArQR), was chosen for
further research. The lyophilized cells were employed
as the biocatalyst to perform the reductive reaction of
ꢀ1
Further reaction was tried using 242 g L substrate and
þ
0.1 mM of external NAD . As expected, the substrate
conversion reached 99% in 4.5 h (Table 1, entry 5), and the
ꢀ1 ꢀ1
spaceꢀtime yield of (R)-3-quinuclidinol reached 916 g L d .
Considering the internal content (ca. 9.7 μmol/g dry cell) of
þ
11
NAD(H/ ) inE. coli cells, the total turnover number (TTN)
was estimated to be around 7500, which represents the highest
among all the 3-quinuclidinone reductases reported to
date. The time course of bioreduction showed that increasing
3
-quinuclidinone.
To optimze the reaction conditions, the biocatalyst
þ
the amount of NAD gave faster bioreduction (Figure 1).
þ
dosage, substrate loading, and external NAD concentra-
tion have to be assessed. Because both the substrate and
product are water-miscible, all the reactions were per-
formed in aqueous solution. Initial experiments were
performed in 10 mL of potassium phosphate buffer (200 mM.
þ
pH 7.0) with 5 mmol of substrate (81 g/L), 1 μmol of NAD ,
7.5 mmol of glucose, and 0.1 g of lyophilized cells of E. coli
(pET28a-ArQR-BmGDH), and the pH of the reaction mix-
ture was controlled at 7.0 by titrating 2 M NaOH. Surpris-
ingly, the substrate was completely transformed to the desired
product within merely 0.5 h (Table 1, entry 1). Clearly it was
possible to transform more substrate, so the substrate loading
ꢀ
1
was raised to 162 g L without changing the other condi-
tions, and the reaction was completed easily within 1.5 h
Figure 1. Reduction of 3-quinuclidinone-HCl by E. coli trans-
formats harboring pET28a-ArQR-BmGDH: (() 162 g/L sub-
strate, 0.1 mM NAD was added; (9) 162 g/L substrate without
addition of NAD ; (2) 242 g/L substrate, 0.1 mM NAD
added.
(Table 1, entry 2). In a stepwise manner, a further attempt was
made to convert 162 g L substrate without external supple-
þ
ꢀ1
þ
þ
þ
ment of NAD . It was found that the conversion also reached
100% in less than 4 h (Table 1, entry 3). When the substrate
was further increased up to 242 g L , without any external
ꢀ
1
Otherwise, to be a good catalyst to render a chemical munu-
facturing process feasible, it needs to meet the requirement of
þ
NAD , only 44% substrate was converted to product even if
the reaction time was extended to 20 h (Table 1, entry 4). The
ꢀ1
ꢀ1
g100 g L substrate loading, e5 g L biocatalyst loading,
13
e24 h reaction time, g98% conversion, and g99% ee.
(
7) (a) Ni, Y.; Xu, J. H. Biotechnol. Adv. 2012, 30, 1279–1288. (b)
Huisman, G. W.; Liang, J.; Krebber, A. Curr. Opin. Chem. Biol. 2010,
4, 122–129.
8) Uzura, A.; Nomoto, F.; Sakoda, A.; Nishimoto, Y.; Kataoka,
M.; Shimizu, S. Appl. Microbiol. Biotechnol. 2009, 83, 617–626.
9) Isotani, K.; Kurokawa, J.; Suzuki, F.; Nomoto, S.; Negishi, T.;
Matsuda, M.; Itoh, N. Appl. Environ. Microbiol. 2013, 79, 1378–1384.
10) Wang, Y.; Li, J. J.; Wu, Q. Q.; Zhu, D. M. J. Mol. Catal. B:
Enzym. 2013, 88, 14–19.
1
(11) San, K. Y.; Bennett, G. N.; Berrıos-Rivera, S. J.; Vadali, R. V.;
´
Yang, Y. T.; Horton, E.; Rudolph, F. B.; Sariyar, B.; Blackwood, K.
Metab. Eng. 2002, 4, 182–192.
(12) Chen, W.; Cheng, X. F.; Zhou, Z. H.; Liu, J. J.; Wang, H. Z. Mol.
Biol. Rep. 2013, 40, 1145–1154.
(13) Luetz, S.; Giver, L.; Lalonde, J. Biotechnol. Bioeng. 2008, 101,
647–53.
(
(
(
4
918
Org. Lett., Vol. 15, No. 19, 2013