Angewandte
Chemie
nucleotide primers (see the Supporting Information). Sixty
clones were analyzed, of which seven were R-selective and six
carried mutation W185R together with additional mutations
(Table 1). The variant showing the lowest ERapp = 1.3 carried
the mutation W185T. These results indicate that replacement
of tryptophan 185 by arginine plays a major role to shift the
enantioselectivity of EstA towards the R substrate. All the
clones we identified within the microlibrary covering amino
acid position 185 carried additional mutations that most likely
accumulated during PCR assembly. Variant 2-R-43 showed
the highest enantioselectivity, with an ERtrue (determined by
GC analysis) of 15.5 representing 81% ee at 38% conversion
(see the Supporting Information). The viability of our method
was convincingly demonstrated by the results shown in
Figure 1c,d: EstA variant 2-R-43 was clearly distinguishable
from EstA wild type in the FACS analysis using the same
labeling conditions as in the initial enantioselectivity screen.
We then addressed the notion that it is mutation W185R
which mainly contributes to enhanced enantioselectivity of
EstA, and we additionally investigated whether expression on
the cell surface had an impact on its catalytic properties. Thus,
EstAwild type and variant W185R with no further amino acid
substitutions were constructed by targeted mutagenesis,
purified, and refolded from inclusion bodies (see the Sup-
porting Information). Whereas isolated EstA protein again
proved to be nonselective (Etrue ꢀ 1), variant W185R dis-
played an ERtrue of 10, with only a slight concomitant
reduction of specific activity (5.2 UmgÀ1 vs. 4.1 UmgÀ1 for
octanoic acid p-nitrophenyl ester and 50mUmg À1 vs.
18 mUmgÀ1 for (R)-2-MDA p-nitrophenyl ester), indicating
that evolution of enantioselectivity does not necessarily
largely compromise catalytic activity.
No previous attempts have been made to engineer EstA
from Pseudomonas aeruginosa. However, lipase A from the
same bacterial strain was extensively studied by directed
evolution using epPCR, saturation mutagenesis, and DNA
shuffling, medium-throughput screening being performed by
a UV/Vis-based assay. This approach with iterative muta-
genesis and selection of several hundred to several thousand
clones per cycle led to improved enantioselectivity of the
kinetic resolution of 2-methyl-decanoic acid p-nitrophenyl
ester (E = 1.1 for the WT and E = 51 for the best muta-
nt).[1j,2a,b] Our strategy has two major advantages: first, only
those clones that display substantial enzymatic activity which
is comparable to the wild-type enzyme are selected by FACS
(see the Supporting Information). Second, very large libraries
can be screened rapidly for enantioselective enzymes, thereby
increasing the number of accessible screening events by
several orders of magnitude as compared to previous
ee screens. Therefore, the bottleneck of directed evolution
of enantioselective hydrolases is no longer the screening step
but the actual generation of large libraries.
reported.[6b] At present, this technique is restricted to the
enantioselectivity screen for hydrolysis of tyramide esters of
chiral carboxylic acids by esterases, or for oxidation of chiral
phenol derivatives by peroxidases, which has been reported
previously.[7] It will be interesting to see how this technique
compares to in vitro approaches that allow similar through-
put,[8] and whether it can be extended to enzymes other than
esterases or peroxidases by direct or indirect coupling of
enzyme activity/enantioselectivity to the formation of hydro-
gen peroxide or tyramide.
Our results are also relevant with regard to the recent
claim that massive mutagenesis might be the method of
choice to introduce high functional diversity in a given
library.[9] This requires high-throughput screening of large
numbers of clones, as only a tiny fraction of variants that may
contain over 20mutations per gene can be expected to display
a functional enzyme, a requirement that can be fulfilled by
our method. The method can also be expected to be useful in
iterative saturation mutagenesis when considering sites in the
enzyme comprising more than four amino acid positions.[10]
Received: November 14, 2007
Revised: February 7, 2008
Published online: June 2, 2008
Keywords: asymmetric catalysis · directed evolution ·
.
enzyme catalysis · hydrolases · kinetic resolution
[1] Reviews of directed evolution of functional enzymes:
a) Directed Enzyme Evolution: Screening and Selection Meth-
ods, Vol. 230 (Eds.: F. H. Arnold, G. Georgiou), Humana,
Totowa, 2003; b) Directed Molecular Evolution of Proteins
(Eds.: S. Brakmann, K. Johnsson), Wiley-VCH, Weinheim,
2002; c) Evolutionary Methods in Biotechnology (Eds.: S.
Brakmann, A. Schwienhorst), Wiley-VCH, Weinheim, 2004;
e) K. A. Powell, S. W. Ramer, S. B. del CardayrØ, W. P. C.
Stemmer, M. B. Tobin, P. F. Longchamp, G. W. Huisman,
Hibbert, F. Baganz, H. C. Hailes, J. M. Ward, G. J. Lye, J. M.
Engineering Protocols, Vol. 352 (Eds.: K. M. Arndt, K. M.
Müller), Humana Press, Totowa, 2007; j) review of directed
evolution of enantioselective enzymes: M. T. Reetz, Proc. Natl.
[2] a) M. T. Reetz, K. Zonta, N. Schimossek, K. Liebeton, K. E.
Genetic Selection and Fingerprinting (Ed.: J.-L. Reymond),
Wiley-VCH, Weinheim, 2005.
In conclusion, we have developed a high-throughput
screening assay for the identification and isolation of enan-
tioselective esterase mutants from large combinatorial libra-
ries generated by the techniques of directed evolution.
Extension to other enzyme types, such as lipases, should be
straightforward, as the bacterial surface display of lipases with
biotechnological relevance using the same approach has been
[3] Typical examples of directed evolution of enantioselective
enzymes:[1j,2a,b] a) M. Wada, C. C. Hsu, D. Franke, M. Mitchell,
2091 – 2098; b) G. J. Williams, S. Domann, A. Nelson, A. Berry,
Angew. Chem. Int. Ed. 2008, 47, 5085 –5088
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5087