Directed evolution of enantioselective enzymes: Iterative cycles of CASTing for probing protein-sequence space

Article abstract of DOI:10.1002/anie.200502746

(Chemical Equation Presented) ReCASTing for success: The recently introduced method of combinatorial active-site saturation test (CAST) has been applied in iterative cycles in the directed evolution of enantioselective wild-type epoxide hydrolases (WT-EH).

Full text of DOI:10.1002/anie.200502746

Communications
mutations that cover the whole gene (and thus enzyme) or a
[
10]
recombinant method such as DNA shuffling is applied.
More recently, the utility of generating focused libraries of
mutants that result from amino acid randomization at defined
[
2,5c,11]
positions in the enzyme, has started to be explored.
For
example, in the case of the directed evolution of enantiose-
lective lipases, we generated a focused library through the
simultaneous randomization of four amino acids located
[
5c]
adjacent to the binding pocket. In the quest to enhance
enzyme activity, other groups have created focused libraries
through the randomization of two, three, four, or more amino
[
11]
acids near the binding site, as in the case of chorismate
[
11a]
mutases.
In another study, systematic saturation mutagenesis at
every-single position of the Bacillus subtilis lipase, which is
composed of 181 amino acids, was considered, thereby
generating 181 focused libraries for the purpose of finding
[
12]
Directed Evolution
enantioselective mutants. This idea was developed inde-
pendently by scientists at Diversa (USA), who were able to
[
13]
DOI: 10.1002/anie.200502746
produce highly enantioselective nitrilases by this method.
Such saturation mutagenesis can be restricted to selected
positions near the active site, thereby lowering the screen-
Directed Evolution of Enantioselective Enzymes:
Iterative Cycles of CASTing for Probing Protein-
Sequence Space**
[14]
ing effort (although restricting catalyst diversity). Other
[2,11]
[11a,b]
focused libraries,
such as those of binary patterning,
are also useful tools. Although these strategies lead to
improved enzymes, they are not evolutionary in nature
unless at least one further round of mutagenesis/screening
follows. In many, if not most, practical cases this may be
necessary. Efficiency in scanning larger parts of the protein-
sequence space with the generation of high-quality libraries is
Manfred T. Reetz,* Li-Wen Wang, and in part
Marco Bocola
Directed evolution of enantioselective enzymes as catalysts in
synthetic organic chemistry has emerged as a fundamentally
[
15]
therefore mandatory. It minimizes the undesired formation
of inferior (e.g., inactive) enzyme mutants and thus reduces
the overall screening effort, which can be substantial in the
[
1]
new and useful approach to asymmetric catalysis. It is based
on earlier contributions by molecular biologists, biochemists,
and biotechnologists who developed efficient random-gene
mutagenesis methods, expression systems, and enzyme assays
in the quest to enhance the stability and/or activity of
[
4]
case of enantioselectivity.
Recently, we introduced a practical method in directed
evolution called a combinatorial active site saturation test
(CAST). CAST was originally developed with the purpose of
[
2]
enzymes. By passing through at least two cycles of muta-
[
3]
[16]
genesis/screening, Darwinistic character is introduced which
forms the general basis of directed evolution. Therein, we first
expanding the range of substrate acceptance of an enzyme.
Based on the 3D structure of the enzyme, two or three amino
acids, whose side chains reside next to the binding pocket, are
identified and the respective positions are then randomized
simultaneously with the creation of relatively small libraries
of mutants. Through the use of this method, it was possible to
obtain mutant lipases that catalyze the hydrolysis of esters
[
4]
developed high-throughput ee assays and applied them in
the evolution of active and enantioselective mutants of
[
5]
[6]
[7]
lipases, monooxygenases, and epoxide hydolases. Other
[
8]
researchers have also contributed to this new area of
asymmetric catalysis. Most directed evolution studies either
begin with one or more cycles of error-prone polymerase
derived from sterically demanding acids that are normally not
[
9]
[16]
chain reactions (epPCRs) to introduce more or less random
at all accepted as substrates.
CASTing thus means the
systematic design and screening of focused libraries around
the complete binding pocket. Owing to the fact that two or
[
12–14]
three amino acids (rather than just one
) are simulta-
[
*] Prof. Dr. M. T. Reetz, L.-W. Wang, Dr. M. Bocola
Max-Planck-Institut für Kohlenforschung
Kaiser-Wilhelm-Platz 1, 45470 Mülheim/Ruhr (Germany)
Fax: (+49)208-306-2985
neously randomized at each site, high catalyst diversity is
possible, therefore theoretically allowing the possibility of
cooperative effects. The method is a useful alternative to
epPCR as the starting point of directed evolution studies.
However, it is not, in itself, an evolutionary process because
only the initial libraries are considered. In our original
E-mail: reetz@mpi-muelheim.mpg.de
[
**] Support by the Deutsche Forschungsgemeinschaft (Schwerpunkt-
programm 1170 “Directed Evolution to Optimize and Understand
Molecular Biocatalysis”; Project RE 359/13-1) and the Fonds der
Chemischen Industrie is gratefully acknowledged. L.-W. W. thanks
the DAAD for a doctoral stipend. We thank H. Hinrichs for HPLC
analyses and J. D. Carballeira, M. Maywald and N. Otte for fruitful
discussions.
[
16]
paper,
we suggested that, for future work, Darwinistic
[
1,2,10,11,15]
character, which is central to directed evolution,
can
be introduced through the performance of further rounds of
mutagenesis/screening. This can be performed by using any
1
236
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 1236 –1241

Angewandte
Chemie
[
25]
one of the conventional methods, such as epPCR or DNA
shuffling after CASTing, which may lead to remote
substrate binding site.
The catalytically active triad, the
tyrosines 251 and 314 (bind and activate the epoxide through
H-bonds), and aspartate 192 (acts as the nucleophile in the
rate-determining step) are located in and around the
tunnel. Inspection of the 3D structure suggests the creation
of six CAST libraries
randomization at sites 193/195/196 (A), 215/217/219 (B),
329/330 (C), 349/350 (D), 317/318 (E), and 244/245/249 (F)
(Figure 2).
We have previously pointed out that for 95% coverage of
a CAST library resulting from the simultaneous random-
ization at two (or three) amino acid positions, an excess of
about 3000 (or 98000, respectively) clones needs to be
[
1,17]
effects.
As an alternative, we herein propose iterative
cycles of CASTing. Consider, for example, an enzyme for
which the CASTanalysis requires four randomization sites A,
B, C, and D. At each site, two or three amino acid positions
are randomized simultaneously depending on the CAST
analysis. Following the generation and screening of the four
respective initial libraries, the gene of each hit is then used as
a template for the second round of CASTing. In each case,
only two libraries are necessary, for example, the gene of the
best mutant arising from A is used as a template for
randomization at sites B and C, etc. (Figure 1). In practice,
not all branches of this confined protein-sequence space need
to be considered, nor is complete over-sampling absolutely
mandatory.
[
24]
[
26]
to be produced separately by
[
16,27]
screened.
However, one can settle for a lower degree of
[
28]
over-sampling and still be successful; we have made these
and other compromises in the present study. Experimentally,
we began by subjecting the WT-ANEH to CASTing with
generation of CAST libraries at sites A, B, and C, that is, in
three separate experiments, positions 193/195/196, 215/217/
2
19, and 329/330, respectively, were randomized. As shown in
Table 1, no enhancement in enantioselectivity was observed
in any of the mutants screened in the libraries that originated
from the randomization of sites A or C. In contrast, several
markedly improved mutants were discovered in the CAST
library produced by randomization at positions 215/217/219
(site B). Therefore, the decision was made to continue the
evolutionary process on this branch and not to consider, at
this point, CAST libraries that potentially originate by
randomization at sites D, E, and F, separately. One of the
best mutants created by randomizing at site B (215/217/219)
by using the WT gene, variant LW081, characterized by the
three mutations L215F, A217N, and R219S, leads to an
Figure 1. Schematic illustration of iterative CASTing involving (as an
example) four randomization sites A, B, C, and D: Confined protein-
sequence space for evolutionary enzyme optimization (redundancy in
some cases is expected).
[
29]
E value of 14. This is similar to the structurally related and
equally selective mutant LW080 (Table 1). Three other
mutants appear to be more active because they showed
The concept of CASTing is illustrated herein with the
directed evolution of enantioselective mutants of the epoxide
[
30]
higher conversions after shorter reaction times, however,
they were seen to be less stereoselective.
[
18–20]
hydrolase from Aspergillus niger (ANEH)
as catalysts in
the hydrolytic kinetic resolution of the glycidyl ether 1 with
formation of diols (R)- and (S)-2. The wild-type enzyme (WT-
ANEH) catalyzes this reaction with a low selectivity factor
The mutant gene that corresponds to enzyme LW081 was
then used as a template for a second cycle of CASTing,
specifically by (arbitrarily) focusing on positions 329/330 (site
C). Substantial improvements in enantioselectivity were
observed with mutant LW086 (E = 21) being the best
enzyme (Table 1). Mutant LW086 is characterized by two
additional mutational changes of M329P and L330Y. Before
continuing the ascent in enantioselectivity, it was of interest to
perform CASTing with one of the mutant genes that encodes
(E = 4.6) in favor of (S)-2 [Eq. (1)]. In previous work, the
[
30]
an enzyme displaying enhanced activity but not improved
enantioselectivity. For this purpose, gene LW037 (Table 1)
was subjected to randomization at positions 329/330 with the
creation of a new library. However, no substantially improved
mutants were found (Table 1).
We then refocused our attention on mutant LW086 and
used the corresponding gene in another cycle of CASTing.
This time, positions 349/350 (site D) were chosen for
simultaneous randomization. The best mutant was LW123
with one new mutational change C350V (E = 24). At this
point, two different decisions for further evolutionary opti-
mization appeared plausible: use of the same gene (LW086)
for randomization at the other sites of the enzyme (A, B, C, E,
enantioselectivity in this reaction was increased to E = 10.8 by
screening several libraries produced by epPCR at different
mutation rates (the number of total clones amounted to
[
7a]
2
0000).
The ee screening was performed through the
[
21]
application of our previously described MS-based system
following a pre-screen to identify the active clones. In the
present study, we employed an improved expression system in
E. coli, adapted the MS screen, and used an efficient pre-
selection system for the elimination of inactive clones.
[
22]
[
7b]
[
23]
[
24]
The X-ray crystal structure of the WT-ANEH reveals,
amongst other things, a narrow hydrophobic tunnel as the
Angew. Chem. Int. Ed. 2006, 45, 1236 –1241
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1237

Communications
focusing on sites E and F. This led to mutants LW126 (E = 49)
and LW144 (E = 35) as shown in Table 1 and Figure 3.
Upon subjection of the gene LW126 to two separate and
independent rounds of CASTing at positions 193/195/196 (site
A) and 244/245/249 (site F), no substantial improvements in
enantioselectivity were observed. Parallel to these efforts and
before examining other positions, such as those defined by B,
C, or D, gene LW144 was used as the template for CASTing in
a different branch of the protein-sequence space. Random-
ization at positions 317/318 (site E) resulted in the identifi-
cation of the, to date, most enantioselective mutant LW202,
which shows a selectivity factor of E = 115 Æ 10 (Table 1 and
Figure 3). This corresponds to a 25-fold increase in enantio-
selectivity relative to the WT, which therefore shows that
iterative CASTing is an efficient way to engineer the ANEH
tunnel. Sequencing revealed the presence of two new
mutations, T317W and T318V. This means that the best
enzyme variant (LW202) is characterized by a total of nine
mutational changes (L215F, A217N, R219S, L249Y, T317W,
T318V, M329P, L330Y, and C350V). At this stage we refrain
from speculating about the origin of enhanced enantioselec-
tivity. Before any sound conclusions can be made, isolation of
mutant LW202 in its pure form is necessary, as are kinetic and
detailed theoretical studies. Moreover, the question whether
all nine mutational changes in the best mutant LW202 are
actually necessary for high enantioselectivity needs to be
addressed systematically. An initial experiment points to a
Figure 2. CASTing of the epoxide hydrolase from Aspergillus niger
[
24]
(
ANEH) based on the X-ray crystal structure of the WT. Top: Defined
randomization sites A–E; Bottom: Top view of tunnel-like binding
pocket showing sites A–E (blue) and the catalytically active Asp192
(
red).
and/or F), or continuation of optimization with LW123.
Herein we chose the latter, and created two new CAST
libraries by using the LW123 gene as the template and
Figure 3. Iterative CASTing in the evolution of enantioselective epoxide
hydrolases as catalysts in the hydrolytic kinetic resolution of rac-1.
1
238
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 1236 –1241

Angewandte
Chemie
[
a]
Table 1: Results of CASTing (I) and iterative CASTing (II).
[
c]
[d]
[e]
Random- Template
ization
sites
Mutant tr
[min]
Conversion [%] ee [%]
ee [%]
s
Newly introduced
mutations
E
No.
% active
p
[
b]
mutants mutants
screened
I
A
No substantial improvements
4000
2200
500
38
25
44
9
>
14=
LW081 60
LW080 60
LW037 10
LW041 10
LW044 10
34
34
42
34
44
80
81
53
61
49
44
45
40
32
39
L215F A217N R219S
L215F A217N R219T
R219T
14>
>
>
WT-ANEH
B
5
>
>
A217P R219H
R219M
6>
>
;
4
C
No substantial improvements
9
II
LW086 60
LW102 60
LW094 60
LW096 60
LW123 60
LW122 60
LW125 60
LW126 60
LW127 60
LW142 30
LW144 60
40
31
71
78
50
27
32
47
56
61
47
84
86
37
27
84
85
82
89
74
59
87
60
41
97
100
72
27
37
86
99
100
81
M329P L330Y
M32K L330T
M329T L330Q
M329S L330P
C350V
L349V C350V
L349V
T318N
T317F T318N
21>
>
=
LW081
19
2
000
87
C
8>
>
LW037
;
8
9
24=
D
E
LW086
LW123
16 2200
6
;
16
9
49=
34 1600
66
;
T317I
T318L
L249Y
35
F
A
F
E
35 1300
3100
9
4
27
63
No substantial improvements
No substantial improvements
LW126
LW144
1600
LW202 60
48
95
94
T317W T318V
>100 2000
[
a] CASTing performed as a means to evolve enantioselective epoxide hydrolases as catalysts in the hydrolytic kinetic resolution of rac-1. [b] t =
r
reaction time. [c] ee =enantioselectivity of the product. [d] ee =enantiopurity of the substrate. [e] E=enantioselectivity of the reaction.
p
s
substantial cooperative effect in at least one case. This
concerns mutation C350V, introduced in mutant LW086,
which raises the E value from 21 to (only) 24 (Figure 3;
Table 1). We discovered that upon reintroduction of cysteine
at position 350 in the final mutant LW202, enantioselectivity
drops sharply from E = 115 to E = 60. This means that
mutation C350V introduced in the early phase of the evolu-
tionary process plays a significant role in the final mutant
LW202. The mechanism of this cooperative effect needs to be
studied in the future.
Keywords: asymmetric catalysis · directed evolution ·
enantioselectivity · enzymes · epoxides
.
[1] For reviews of directed evolution of enantioselective enzymes,
see: a) M. T. Reetz, Proc. Natl. Acad. Sci. USA 2004, 101, 5716 –
5722; b) M. T. Reetz in Methods in Enzymology, Vol. 388 (Eds.:
D. E. Robertson, J. P. Noel), Elsevier, San Diego, CA, 2004,
pp. 238 – 256; c) M. T. Reetz, Tetrahedron 2002, 58, 6595 – 6602;
d) See also N. J. Turner, Trends Biotechnol. 2003, 21, 474 – 478.
2] For reviews of directed evolution, see: a) Methods in Molecular
Biology, Vol. 230 (Eds.: F. H. Arnold, G. Georgiou), Humana,
Totowa, NJ, 2003; b) Directed Molecular Evolution of Proteins
(or How to Improve Enzymes for Biocatalysis), (Eds.: S.
Brakmann, K. Johnsson), Wiley-VCH, Weinheim, 2002;
c) Enzyme Functionality—Design, Engineering, and Screening,
[
The present results were obtained by screening a total of
only 20000 clones. This is about the same number of clones
previously screened when using epPCR (at different mutation
[
7a]
rates) as the method for library formation, yet the results of
the two approaches are dramatically different (E = 115 versus
E = 10.8). In iterative CASTing, the lower numbers of
screened mutants are linked with higher numbers of libraries
which need to be created. However, the latter is not the
bottleneck, because a library of ANEH mutants can be
generated in one day. It is likely that if iterative CASTing
were to be performed more extensively to cover more of the
focused branches of the confined protein-sequence space
(Eds.: A. Svendsen), Marcel Dekker, New York, 2004; d) Evolu-
tionary Methods in Biotechnology (Clever Tricks for Directed
Evolution), (Ed.: S. Brakmann, A. Schwienhorst), Wiley-VCH,
Weinheim, 2004; For a historical account of directed evolution
which includes the early contributions of Manfred Eigen, see:
e) A. Koltermann, U. Kettling, Biophys. Chem. 1997, 66, 159 –
177.
[
3] One of the first reports regarding more than one cycle of
mutagenesis/selection leading to evolutionary pressure was
published in the mid 1980s, although this crucial idea was not
generalized until the 1990s: see reference [2] and H. Liao, T.
McKenzie, R. Hagemann, Proc. Natl. Acad. Sci. USA 1986, 83,
576 – 580.
(
Figure 1), even more hits would be discovered. For practical
purposes this is not necessary, although we plan to do so for
[
31]
theoretical reasons. We anticipate that iterative CASTing
may also be useful when attempting to evolve other enzyme
properties such as enhanced activity and/or broadened
[
4] For reviews of high-throughput ee assays, see: a) M. T. Reetz,
Angew. Chem. 2001, 113, 292 – 320; Angew. Chem. Int. Ed. 2001,
[
16,32]
substrate acceptance.
40, 284 – 310; b) M. T. Reetz, in Enzyme Functionality—Design,
Engineering, and Screening (Ed.: A. Svendsen), Marcel Dekker,
New York, 2004, pp. 559 – 598; c) M. T. Reetz, in Evolutionary
Methods in Biotechnology (Eds.: S. Brakmann, A. Schwien-
horst), Wiley-VCH, Weinheim, 2004, pp. 113 – 141; d) M. T.
Received: August 4, 2005
Revised: October 28, 2005
Published online: January 13, 2006
Angew. Chem. Int. Ed. 2006, 45, 1236 –1241
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1239

Communications
Reetz, in Methods in Molecular Biology, Vol. 230 (Eds.: F. H.
Kato, H. Nakano, T. Yamane, J. Mol. Biol. 2003, 331, 585 – 592;
e) J. A. Wells, M. Vasser, D. B. Powers, Gene 1985, 34, 315 – 323;
f) J. D. Stevenson, S. J. Benkovic, J. Chem. Soc. Perkin Trans. 2
2002, 1483 – 1493; g) M. Wilming, A. Iffland, P. Tafelmeyer, C.
Arrivoli, C. Saudan, K. Johnsson, ChemBioChem 2002, 3, 1097 –
1104; h) A. Juillerat, T. Gronemeyer, A. Keppler, S. Gendreizig,
H. Pick, H. Vogel, K. Johnsson, Chem. Biol. 2003, 10, 313 – 317;
i) L. B. Evnin, J. R. Vµsquez, C. S. Craik, Proc. Natl. Acad. Sci.
USA 1990, 87, 6659 – 6663; j) A. R. Schmitzer, F. LØpine, J. N.
Pelletier, Protein Eng. Des. Sel. 2004, 17, 809 – 819; k) J. Yang, Y.
Koga, H. Nakano, T. Yamane, Protein Eng. 2002, 15, 147 – 152;
l) M. S. Warren, S. J. Benkovic, Protein Eng. 1997, 10, 63 – 68;
m) T. Yano, S. Oue, H. Kagamiyana, Proc. Natl. Acad. Sci. USA
1998, 95, 5511 – 5515; n) E. M. Gabor, D. B. Janssen, Protein
Eng. Des. Sel. 2004, 17, 571 – 579; o) D. K. Dube, L. A. Loeb,
Biochemistry 1989, 28, 5703 – 5707.
Arnold, D. Georgiou), Humana Press, Totowa, NJ, 2003,
pp. 259 – 282; e) M. T. Reetz, in Methods in Molecular Biology,
Vol. 230 (Eds.: F. H. Arnold, G. Georgiou), Humana Press,
Totowa, NJ, 2003, pp. 283 – 290; f) D. Wahler, J.-L. Reymond,
Curr. Opin. Biotechnol. 2001, 12, 535 – 544.
[
5] a) M. T. Reetz, A. Zonta, K. Schimossek, K. Liebeton, K.-E.
Jaeger, Angew. Chem. 1997, 109, 2961 – 2963; Angew. Chem. Int.
Ed. Engl. 1997, 36, 2830 – 2832; b) K. Liebeton, A. Zonta, K.
Schimossek, M. Nardini, D. Lang, B. W. Dijkstra, M. T. Reetz,
K.-E. Jaeger, Chem. Biol. 2000, 7, 709 – 718; c) M. T. Reetz, S.
Wilensek, D. Zha, K.-E. Jaeger, Angew. Chem. 2001, 113, 3701 –
3703; Angew. Chem. Int. Ed. 2001, 40, 3589 – 3591; d) D. Zha, S.
Wilensek, M. Hermes, K.-E. Jaeger, M. T. Reetz, Chem.
Commun. 2001, 2664 – 2665.
[
[
6] a) M. T. Reetz, B. Brunner, T. Schneider, F. Schulz, C. M.
Clouthier, M. M. Kayser, Angew. Chem. 2004, 116, 4167 – 4170;
Angew. Chem. Int. Ed. 2004, 43, 4075 – 4078; b) M. T. Reetz, F.
Daligault, B. Brunner, H. Hinrichs, A. Deege, Angew. Chem.
[12] S. A. Funke, A. Eipper, M. T. Reetz, N. Otte, W. Thiel, G.
van Pouderoyen, B. W. Dijkstra, K.-E. Jaeger, T. Eggert, Bio-
catal. Biotransform. 2003, 21, 67 – 73.
2
004, 116, 4170 – 4173; Angew. Chem. Int. Ed. 2004, 43, 4078 –
[13] G. DeSantis, K. Wong, B. Farwell, K. Chatman, Z. Zhu, G.
Tomlinson, H. Huang, X. Tan, L. Bibbs, P. Chen, K. Kretz, M. J.
Burk, J. Am. Chem. Soc. 2003, 125, 11476 – 11477.
[14] a) G. P. Horsman, A. M. F. Liu, E. Henke, U. T. Bornscheuer,
R. J. Kazlauskas, Chem. Eur. J. 2003, 9, 1933 – 1939; b) S. Park,
K. L. Morley, G. P. Horsman, M. Holmquist, K. Hult, R. J.
Kazlauskas, Chem. Biol. 2005, 12, 45 – 54; c) R. Kazlauskas,
Chem. Listy 2003, 67, 329 – 337.
4081.
7] a) M. T. Reetz, C. Torre, A. Eipper, R. Lohmer, M. Hermes, B.
Brunner, A. Maichele, M. Bocola, M. Arand, A. Cronin, Y.
Genzel, A. Archelas, R. Furstoss, Org. Lett. 2004, 6, 177 – 180;
For an improved expression system of ANEH in E. coli, see:
b) F. Cedrone, S. Niel, S. Roca, T. Bhatnagar, N. Ait-Abdelkader,
C. Torre, H. Krumm, A. Maichele, M. T. Reetz, J. C. Baratti,
Biocatal. Biotransform. 2003, 21, 357 – 364.
8] For further examples of enhanced enantioselectivity (evolu-
tionary process was not involved in all cases), see: a) Amino-
transferases: G. W. Matcham, A. R. S. Bowen, Chim. Oggi 1996,
[15] S. Lutz, W. M. Patrick, Curr. Opin. Biotechnol. 2004, 15, 291 –
297.
[
[16] M. T. Reetz, M. Bocola, J. D. Carballeira, D. Zha, A. Vogel,
Angew. Chem. 2005, 117, 4264 – 4268; Angew. Chem. Int. Ed.
2005, 44, 4192 – 4196.
1
4, 20 – 24; b) Esterases: U. T. Bornscheuer, J. Altenbuchner,
[
14]
H. H. Meyer, Biotechnol. Bioeng. 1998, 58, 554 – 559; c) Hydan-
toinases: O. May, P. T. Nguyen, F. H. Arnold, Nat. Biotechnol.
2
speed, M. A. Sogorb, F. Wu, S.-B. Hong, F. M. Raushel,
Biochemistry 2001, 40, 1325 – 1331; e) M. Chen-Goodspeed,
M. A. Sogorb, F. Wu, F. M. Raushel, Biochemistry 2001, 40,
[17] For examples of remote effects, see: a) M. Bocola, N. Otte, K.-
E. Jaeger, M. T. Reetz, W. Thiel, ChemBioChem 2004, 5, 214 –
223; b) S. Oue, A. Okamoto, T. Yano, H. Kagamiyama, J. Biol.
Chem. 1999, 274, 2344 – 2349; c) A. Iffland, P. Tafelmeyer, C.
Saudan, J. Johnsson, Biochemistry 2000, 39, 10790 – 10798; d) M.
Kumar, K. K. Kannan, M. V. Hosur, N. S. Bhavesh, A. Chatter-
jee, R. Mittal, R. V. Hosur, Biochem. Biophys. Res. Commun.
2002, 294, 395 – 401; e) P. K. Agarwal, S. R. Billeter, P. T. R.
Rajagopalan, S. J. Benkovic, S. Hammes-Schiffer, Proc. Natl.
Acad. Sci. USA 2002, 99, 2794 – 2799; f) H. Zhao, F. H. Arnold,
Protein Eng. 1999, 12, 47 – 53; g) L. K. Jackson, J. Baldwin, R.
Akella, E. J. Goldsmith, M. A. Phillips, Biochemistry 2004, 43,
12990 – 12999; h) B. Lingen, D. Kolter-Jung, P. Dünkelmann, R.
Feldmann, J. Grötzinger, M. Pohl, M. Müller, ChemBioChem
2003, 4, 721 – 726; i) S. McQ. Gould, D. S. Tawfik, Biochemistry
2005, 44, 5444 – 5452; j) V. G. H. Eijsink, S. Gåseidnes, T. V.
Borchert, B. van den Burg, Biomol. Eng. 2005, 22, 21 – 30.
[18] a) C. Morisseau, A. Archelas, C. Guitton, D. Faucher, R.
Furstoss, J. C. Baratti, Eur. J. Biochem. 1999, 263, 386 – 395;
b) K. M. Manoj, A. Archelas, J. C. Baratti, R. Furstoss, Tetrahe-
dron 2001, 57, 695 – 701; c) M. Arand, H. Hemmer, H. Dürk, J.
Baratti, A. Archelas, R. Furstoss, F. Oesch, Biochem. J. 1999,
344, 273 – 280.
000, 18, 317 – 320; d) Phosphotriesterases: M. Chen-Good-
1
332 – 1339; f) Monoamine oxidases: M. Alexeeva, A. Enright,
M. J. Dawson, M. Mahmoudian, N. J. Turner, Angew. Chem.
2
3
Mol. Biol. 2003, 331, 585 – 592; h) Aldolases: M. Wada, C.-C.
Hsu, D. Franke, M. Mitchell, A. Heine, I. Wilson, C.-H. Wong,
Bioorg. Med. Chem. 2003, 11, 2091 – 2098; i) Epoxide hydro-
lases: B. van Loo, J. H. L. Spelberg, J. Kingma, T. Sonke, M. G.
Wubbolts, D. B. Janssen, Chem. Biol. 2004, 11, 981 – 990; j) L.
Rui, L. Cao, W. Chen, K. F. Reardon, T. K. Wood, Appl.
Environ. Microbiol 2005, 71, 3995 – 4003.
002, 114, 3309 – 3312; Angew. Chem. Int. Ed. 2002, 41, 3177 –
180; g) Lipases: Y. Koga, K. Kato, H. Nakano, T. Yamane, J.
[
9] a) D. W. Leung, E. Chen, D. V. Goeddel, Technique 1989, 1, 11 –
1
2
5; b) R. C. Cadwell, G. F. Joyce, PCR Methods Appl. 1992, 2,
8 – 33; c) T. Eggert, M. T. Reetz, K.-E. Jaeger, in Enzyme
Functionality—Design, Engineering, and Screening (Ed.: A.
Svendsen), Marcel Dekker, New York, 2004, pp. 375 – 390;
d) For an illuminating study concerning high-error-rate
epPCR, see: D. A. Drummond, B. L. Iverson, G. Georgiou,
F. H. Arnold, J. Mol. Biol. 2005, 350, 806 – 816.
[19] For reviews of epoxide hydrolases, see: a) K. Faber, R. V. A.
Orru in Enzyme Catalysis in Organic Synthesis (Eds.: K. Drauz,
H. Waldmann), Wiley-VCH, Weinheim, 2002, pp. 579 – 608;
b) C. Morisseau, B. D. Hammock, Annu. Rev. Pharmacol.
Toxicol. 2005, 45, 311 – 333; c) A. Archelas, R. Furstoss, Curr.
Opin. Chem. Biol. 2001, 5, 112 – 119.
[
10] For a review of DNA shuffling, see: K. A. Powell, S. W. Ramer,
S. B. del CardayrØ, W. P. C. Stemmer, M. B. Tobin, P. F. Long-
champ, G. W. Huisman, Angew. Chem. 2001, 113, 4068 – 4080;
Angew. Chem. Int. Ed. 2001, 40, 3948 – 3959.
[
11] For examples of focused libraries, see: references [2,5c,12–15]
and a) S. V. Taylor, P. Kast, D. Hilvert, Angew. Chem. 2001, 113,
[20] For other important work on the directed evolution of enantio-
selective epoxide hydrolases, see references [8i,j].
3
408 – 3436; Angew. Chem. Int. Ed. 2001, 40, 3310 – 3335; b) H.
[21] a) M. T. Reetz, M. H. Becker, H.-W. Klein, D. Stöckigt, Angew.
Chem. 1999, 111, 1872 – 1875; Angew. Chem. Int. Ed. 1999, 38,
1758 – 1761; W. Schrader, A. Eipper, D. J. Pugh, M. T. Reetz,
Can. J. Chem. 2002, 80, 626 – 632.
Xiong, B. L. Buckwalter, H. M. Shieh, M. H. Hecht, Proc. Natl.
Acad. Sci. USA 1995, 92, 6349 – 6353; c) J. U. Bowie, R. T. Sauer,
Proc. Natl. Acad. Sci. USA 1989, 86, 2152 – 2156; d) Y. Koga, K.
1
240
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 1236 –1241

Angewandte
Chemie
[
[
22] For information on pre-screens for EH activity, see: a) F. Zocher,
M. M. Enzelberger, U. T. Bornscheuer, B. Hauer, R. D. Schmid,
Anal. Chim. Acta 1999, 391, 345 – 351; b) D. Wahler, J.-L.
Reymond, Angew. Chem. 2002, 114, 1277 – 1280; Angew.
Chem. Int. Ed. 2002, 41, 1229 – 1232; c) K. Doderer, R. D.
Schmid, Biotechnol. Lett. 2004, 26, 835 – 839.
23] Herein, we employed an efficient selection system based on
bacterial (E. coli) growth. The latter is directly related to the
presence of active epoxide hydrolases which catalyze the
detoxicating hydrolysis of epoxide substrates such as 1: M. T.
Reetz, L.-W. Wang, Comb. Chem. High Throughput Screening,
in press.
[
24] a) J. Y. Zou, B. M. Hallberg, T. Bergfors, F. Oesch, M. Arand,
S. L. Mowbray, T. A. Jones, Structure 2000, 8, 111 – 122; b) http://
www.pdb.ccdc.cam.ac.uk/pdb.
[
25] For theoretical analyses of other enzymes with narrow tunnel-
like binding pockets, some of which were re-engineered, see:
reference [1a]; a) J. Schmitt, S. Brocca, R. D. Schmid, J. Pleiss,
Protein Eng. 2002, 15, 595 – 601; b) H. Henriksson, J. Ståhlberg,
R. Isaksson, G. Pettersson, FEBS Lett. 1996, 390, 339 – 344; c) M.
Holmquist, F. Haeffner, T. Norin, K. Hult, Protein Sci. 1996, 5,
8
3 – 88.
26] Based on the X-ray crystal structure of ANEH (see reference
24]), the nonresolved loop from postions 320 to 328 over the
[
[
binding site was constructed by molecular modeling. The CAST
libraries were then defined by using the completed protein
model with substrate 1 in the binding pocket.
[
27] For the application of mathematical expressions useful in the
construction, description, and evaluation of protein libraries,
see: a) A. D. Bosley, M. Ostermeier, Biomol. Eng. 2005, 22, 57 –
61; b) W. M. Patrick, A. E. Firth, J. M. Blackburn, Protein Eng.
2003, 16, 451 – 457; c) W. M. Patrick, A. E. Firth, Biomol. Eng.
2005, 22, 105 – 112.
[
28] In our previous lipase project (reference [5c]), we created a
focused library of mutants through the simultaneous random-
ization of amino acid positions 160/161/162/163 by using
appropriate cassette mutagenesis. For a 95% sequence space
coverage, this would require the screening of about three million
clones (see references [16,27]). Although only 5000 were actually
evaluated, markedly improved mutants were in fact found: The
selectivity factor in the kinetic resolution of a chiral ester
increased from E = 1.2 (WT) to 30 (mutant) (reference [5c]).
29] All E values were derived through the application of the formula
of Sih et al: C.-S. Chen, Y. Fujimoto, G. Girdaukas, C. J. Sih, J.
Am. Chem. Soc. 1982, 104, 7294 – 7299.
[
[
30] The conclusion regarding relative activity is only qualitative and
is uncertain because it is based on the assumption of equal
expression levels for the various mutants. A sound conclusion
can only be made following a kinetic study of the isolated and
purified mutants (which we are currently performing).
[
[
31] A different type of evolutionary process can be envisioned
through the combination of the mutations of two (or more)
enzyme variants identified in the initial CAST libraries.
32] For a review of practical asymmetric catalysis emphasizing the
need to develop systems showing broad substrate scope, see: M.
Breuer, K. Ditrich, T. Habicher, B. Hauer, M. Keßeler, R.
Stürmer, T. Zelinski, Angew. Chem. 2004, 116, 806 – 843; Angew.
Chem. Int. Ed. 2004, 43, 788 – 824.
Angew. Chem. Int. Ed. 2006, 45, 1236 –1241
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1241