CHEMCATCHEM
COMMUNICATIONS
To further optimize the biotransformation of 11, reactions
were performed at a higher substrate loading. At a substrate
concentration of 20 mm, the reactions proceeded to comple-
tion, although higher concentrations up to 100 mm led to in-
complete conversion. A preparative-scale biotransformation
tive scale, and this points the way towards future applications
of these IREDs. Despite the unexpected broad specificity of the
wild-type enzyme, there is undoubtedly further scope for engi-
neering these enzymes to enhance their properties for future
applications in the synthesis of chiral amines.
(200 mg) was successfully performed at 20 mm by using the
whole-cell biocatalyst. Conversion was observed to proceed to
Experimental Section
>
98% for amine (S)-12, which was isolated in 87% yield.
We also explored the effect of substitutions at the 2-position
The (S)-IRED biocatalyst was produced by using E. coli BL21 (DE3)
bearing a pET-28a (+) plasmid with an inserted codon-optimized
of pyrrolines (i.e., 1a–d) and also the effect of ring size on the
level of conversion and ee (i.e., 3 and 5; Table 3). For the series
(
S)-IRED gene cloned by using NdeI and XhoI restriction sites. Culti-
vation was performed in lysogeny broth (LB) medium (500 mL) at
78C and 250 rpm to OD600 nm =0.6–0.8 (OD=optical density) and
3
recombinant protein expression was induced by using 0.2 mm iso-
propyl b-d-1-thiogalactopyranoside (IPTG). Cells were harvested
after a further cultivation at 208C and 250 rpm for 18–22 h. Cell
disruption was performed by ultrasonication of a suspension of
E. coli BL21 (DE3) expressing the (S)-IRED in sodium phosphate
buffer (100 mm, pH 7.0).
Table 3. Whole-cell biocatalysis with cyclic imine reductase substrates of
different ring sizes. Conversions after 6 and 18 h as well as enantioselec-
tivities after 18 h are given.
Substrate
% Conversion
18 h
% ee
Product
6
h
18 h
2
+
1
1
1
1
3
5
a
b
c
57
55
36
17
42
>98
>98
>95
87
>98
98
>98
>98
(S)-2a
The (S)-IRED was purified by Ni -chelating affinity chromatogra-
phy by using a HisTrap column whereby elution was performed by
using sodium phosphate buffer (100 mm, pH 7.0) containing
[
a]
28
14
25
>98
>98
(R)-2b
[
a]
(R)-2c
[
a]
d
(R)-2d
(S)-4
(S)-6
3
00 mm NaCl and 300 mm imidazole.
Kinetic constants were determined from a liquid-phase spectro-
photometric assay by monitoring the decrease in NADPH at 340
e=6.22 mm cm ) or 370 nm (e=2.216 mm cm ). Reaction
[a] The formation of the (R)-amine is due to a change in the Cahn–
À1
À1
À1
À1
(
Ingold–Prelog assignment.
mixtures contained sodium phosphate buffer (100 mm, pH 7.0),
NADPH to an absorbance of 0.8–1 at the respective wavelength,
1
% (v/v) dimethyl sulfoxide, and the substrate at the desired con-
of pyrrolines, conversions were found to vary with the sub-
stituents (17–55%), although high ee values were generally ob-
tained (87–98%). Comparison of 1a, 3, and 5 revealed that the
centration. The reaction was started by adding the purified
enzyme to the mixture. One unit of the (S)-IRED is defined as the
amount of protein that oxidizes 1 mmol NADPH per minute.
(
S)-IRED is highly tolerant of ring size and indeed favors six-
Biotransformations were performed at 308C in sodium phosphate
buffer (100 mm, pH 7.0) by using resting cells of E. coli BL21 (DE3)
expressing the (S)-IRED at a final OD600 nm of 30. The reaction mix-
tures typically contained 5 mm of the substrate imine, 50 mm of
glucose, and 2% (v/v) dimethylformamide unless otherwise stated.
Samples were typically taken after 0, 1, 3, 6, 18, and 30 h, extracted
by using dichloromethane, and used directly for analysis. Negative-
control biotransformations were performed in the same way by
using cells harboring empty pET-28a (+) vector cultivated and in-
duced like the cells expressing the biocatalyst. A preparative-scale
biotransformation for the enantioselective reduction of 11 was per-
formed on a 200 mg scale at a substrate concentration of 20 mm
with 100 mM of glucose and cells added at a final OD600nm of 30
for 24 h. Conversion was >98% and (S)-12 was isolated in 87%
yield.
and seven-membered rings.
The data shown in Tables 2 and 3 may provide insight into
the exact role of imine reductases in the synthesis of cyclic sec-
ondary and tertiary amines. The biosynthesis of the natural
product eleagnine (16a) in pea seedlings from l-tryptophan
has been shown to proceed via imine 15a, which was itself
generated in situ by internal cyclization of an amine onto an
[
19]
aldehyde. Conversion of 15a into 16a was suggested to be
due to a putative imine reductase. In separate studies, it was
shown that the reduction of a series of dihydro-b-carbolines
could be effected by using the biocatalytic capabilities of dif-
[20a]
ferent eukaryotic organisms such as Saccharomyces bayanus
[
20b]
and Eisenia foetida.
The use of S. bayanus appears to be the
only report of imine reductase activity in a yeast, which are
well documented to exhibit high ketoreductase and ene
reductase activities.
Acknowledgements
In conclusion we have shown that the (S)-imine reductase
from Streptomyces sp. GF3546 is able to catalyze the highly
enantioselective reduction of five-, six-, and seven-membered
imines, including substituted dihydro-b-carbolines, which high-
lights the preference for six-membered imines as well as the
broad substrate scope of this enzyme. In addition, we demon-
strated that this (S)-IRED also catalyzes the reduction of imini-
um ions. Reactions performed with the use of E. coli whole-cell
biocatalyst supplemented with glucose gave excellent conver-
sions (>95%) and yields of the isolated products on a prepara-
The research leading to these results received funding from the
European Union’s Seventh Framework Programme FP7/2007–
2013 under grant agreement no. 266025 (BIONEXGEN, to F.L.). We
are also grateful for funding from the Marie Curie Initial Training
Network (Biotrains FP7-ITN-238531, to D.G.). We also acknowl-
edge support from EPSRC and AstraZeneca (to S.H.) and the
Royal Society for a Wolfson Research Merit Award (NJT).
ꢀ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 4
&
3
&
ÞÞ
These are not the final page numbers!