Communications
DOI: 10.1002/anie.200704606
Esterases
Complete Inversion of Enantioselectivity towards Acetylated Tertiary
Alcohols by a Double Mutant of a Bacillus Subtilis Esterase**
Sebastian Bartsch, Robert Kourist, and Uwe T. Bornscheuer*
Enantiomerically pure compounds are playing a rapidly
growing role as building blocks in organic chemistry. Biocat-
alysts are frequently used for their synthesis owing to their
high chemo-, regio-, and stereoselectivity, making them often
superior to chemical catalysis.[1] However, the wild-type (WT)
enzyme often does not show broad substrate specificity in
combination with sufficiently high selectivity. These limita-
tions can be overcome by using rational protein design or
directed evolution. Whereas several examples have been
reported in which the enantioselectivity could be increased by
these methods, a switch in enantiopreference could be
achieved only rarely.[2] Most importantly, only E values > 50
are of synthetic importance, and this threshold is difficult to
surpass starting from an enzyme with opposite natural
selectivity. For instance, May et al. converted a d-selective
hydantoinase into an l-selective variant, but this gave only
20% ee,[2d] and Zha et al.[2e] inversed the enantioselectivity of
a lipase from Ps. aeruginosa to E = 30. In addition, the
variants might lose considerable activity as reported for an
arylmalonate decarboxylase with inverted selectivity.[3] Thus,
changing the enantioselectivity of an enzyme is still a major
challenge in protein engineering.
inversion of the enantioselectivity of BS2 towards acetylated
tertiary alcohols (Scheme 1).
Scheme 1. Kinetic resolution of rac-1 using the wild-type esterase and
E188D of Bacillus subtilis (BS2) yields the R enantiomer (top) while the
double-mutant E188W/M193C yields the S enantiomer (bottom).
An esterase from Bacillus subtilis (BS2) was identified as
a GGG(A)X-hydrolase[4] showing activity towards the ace-
tates of tertiary alcohols. Esterases containing the more
common GX motif in the oxyanion pocket do not show
activity towards the esters of sterically more demanding
tertiary alcohols because of the smaller active site. With the
mutants G105A and E188D two variants with excellent R
enantioselectivity (E > 100) towards 1,1,1-trifluoro-2-phenyl-
but-3-yn-1-yl acetate (1) were created, while the wild-type
gave ER = 42.[5] Optically active tertiary alcohols are impor-
tant building blocks for organic synthesis[6] and have been
recently applied for the synthesis of an orally active A2A
receptor antagonist in a mouse catalepsy model.[7] Further-
more, fluorine-containing optically active alcohols are also of
great interest owing to their potential use as ferroelectric
liquid crystals and drugs.[8]
Recently,[5a] we identified, based on computer modelling,
residue E188 as having a strong influence on the enantiose-
lectivity of esterase BS2 towards acetylated tertiary alcohols:
While mutant E188D had an E value of ER > 100 towards 1,
mutant E188F had an inversed enantiopreference (ES = 3).
Encouraged by this observation, we decided to apply focused
directed evolution based on saturation mutagenesis inspired
by the recently proposed CASTing[9] method developed by
the Reetz group. From the crystal structure of the highly
homologous BsubpNBE esterase (pdb entry: 1QE3[10]), three
adjacent amino acids (E188, A190, and M193) were selected
as target residues as they all point into the active site.
Furthermore these residues are located around the catalytic
triad. Hence, a focused library with simultaneous saturation
mutagenesis at these three positions was created using
degenerate primers bearing NNK codons, which theoretically
should lead to 203 = 8000 variants.
Here we report a focused directed-evolution approach
based on a previous rational-protein-design study[5a] for the
The resulting library of 2800 mutants was first submitted
to a prescreening on agar plates in which only ꢀ 40% of all
clones still showed activity. To facilitate high-throughput
screening of the remaining variants, we applied a previously
developed spectrophotometric assay based on the quantifi-
cation of the released acetic acid, which is converted to
NADHin an enzyme cascade. [11] Using enantiomerically pure
(S)-1 and (R)-1 in separate wells of microtiterplates, we
determined the apparent enantioselectivity Eapp for approx-
imately 1100 clones. The best variant turned out to be the
[*] S. Bartsch, R. Kourist, Prof. U. T. Bornscheuer
Institute of Biochemistry
Deparment of Biotechnology & Enzyme Catalysis
Greifswald University
Felix-Hausdorff-Strasse 4, 17487 Greifswald (Germany)
Fax: (+49)3834-86-80066
E-mail: uwe.bornscheuer@uni-greifswald.de
[**] We are very grateful to Prof. Karl Hult, Linda Fransson, and Martin
Veld (KTH, Stockholm, Sweden) for fruitful discussions and
introduction to the docking program.
1508
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 1508 –1511