Directed evolution of enantioselective hybrid catalysts: a novel concept in asymmetric catalysis

Article abstract of DOI:10.1016/j.tet.2007.03.177

The concept of directed evolution of enantioselective hybrid catalysts was proposed in 2001/2002 and implemented experimentally for the first time in a proof-of-concept study in 2006. The idea is based on directed evolution, which comprises repeating cycl

Full text of DOI:10.1016/j.tet.2007.03.177

Tetrahedron 63 (2007) 6404–6414
Directed evolution of enantioselective hybrid catalysts:
a novel concept in asymmetric catalysis
!
Manfred T. Reetz, Martin Rentzsch, Andreas Pletsch, Matthias Maywald, Peter Maiwald,
*
ˆ
Jerome J.-P. Peyralans, Andrea Maichele, Yu Fu, Ning Jiao, Frank Hollmann,
ꢀ
ꢀ
Regis Mondiere and Andreas Taglieber
ꢁ
Max-Planck-Institut fu€r Kohlenforschung, Kaiser-Wilhelm-Platz 1, 45470 Mu€lheim/Ruhr, Germany
Received 26 January 2007; revised 28 March 2007; accepted 30 March 2007
Available online 6 April 2007
Dedicated to Karl Wieghardt on the occasion of his 65th birthday
Abstract—The concept of directed evolution of enantioselective hybrid catalysts was proposed in 2001/2002 and implemented experimen-
tally for the first time in a proof-of-concept study in 2006. The idea is based on directed evolution, which comprises repeating cycles of random
gene mutagenesis/expression/screening in a Darwinistic sense for the purpose of improving the catalytic profile of enzymes. In the case of
hybrid catalysts, mutagenesis/expression of a protein is first performed with formation of a library of mutants, which are then modified chemi-
cally en masse with the introduction of an appropriate achiral ligand system harboring a transition metal. Screening these mutant hybrid
catalysts in a given transition metal-catalyzed reaction then leads to an improved catalyst, so that the corresponding gene can be used to start
another evolutionary cycle. This process can be repeated as often as needed until the desired catalytic profile has been reached, e.g., enhanced
enantioselectivity.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
enantioselectivity (ee>95%) is fairly broad. Nevertheless,
no single chiral catalyst is truly general. Indeed, this is why
research in this challenging area needs to be intensified,
especially with respect to the development of new concepts.
Asymmetric catalysis continues to be an active field of re-
search in synthetic organic chemistry. When applying this
form of catalysis, chemists have several main options,
namely transition metal catalysis,1 organocatalysis2 or bio-
catalysis.3 The optimal choice will depend on a number of
factors, including the nature of the immediate goal (small
or large scale), cost of catalysts, catalyst activity, enantio-
selectivity, and stability under operating conditions as well
as recyclability. When engaging in the challenging research
directed toward developing new synthetic chiral catalysts,
success depends upon the quality of design, but also on
experience, trial-and-error, and serendipity. Along a different
line, combinatorial asymmetric catalysis, based on the
design and use of modular ligands, has been implemented
experimentally with some degree of success4 including the
use of mixtures of monodentate P-ligands.5
The exploitation of enzymes in synthetic organic chemistry,
both in academic and industrial laboratories, has increased
dramatically during the past 20 years.3 The realization that
many of them can be used in organic solvents is one of the
important advances.7 Moreover, rapid progress in biotech-
nological engineering and microbiology has contributed
heavily to success in making biocatalysis practical, and
more progress can be expected in the near future.8 Neverthe-
less, the traditional problem in applied enzymology was
not solved until the 1990s, namely the limited substrate
scope and poor enantioselectivity that the enzymes often
display. Rational design based on site-specific mutagenesis
had been shown to be successful in some cases, but the
process is far from general due to the structural complexity
of proteins.9
The power of privileged synthetic catalyst systems6 lies in
the observation that the range of substrates showing high
In 1997 we proposed and implemented experimentally a fun-
damentally new and fairly general approach to asymmetric
catalysis, namely the directed evolution of enantioselective
enzymes for use in synthetic organic chemistry.10 It is based
on the appropriate combination of random gene mutagene-
sis, expression, and high-throughput screening, previously
Keywords: Asymmetric catalysis; Enzymes; Hydrogenation; Directed
evolution.
* Corresponding author. Tel.: +49 208 3062000; fax: +49 208 3062985;
e-mail: reetz@mpi-muelheim.mpg.de
Deceased on 19.5.2004.
!
0040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.03.177

M. T. Reetz et al. / Tetrahedron 63 (2007) 6404–6414
6405
etc.
A C D
A C D A B D
A B C
B C D A B D
B C
B C D
A
C D
B C
A
B C D
A B D
gene (DNA)
random
wild-type enzyme
A
mutagenesis
B
C D
C
A
B C
A
D
A B D
...
...
1
1
2
3
4
5
etc.
etc.
library of mutated genes
expression
A
B
C
D
repeat
WT
Figure 2. Schematic illustration of iterative saturation mutagenesis (ISM)
involving (as an example) four randomization sites A, B, C, and D of an
enzyme.14
2
3
4
5
library of mutated enzymes
screening (or selection)
for enantioselectivity
acids.11 The principle of ISM is illustrated schematically
in Figure 2, in which four sites A, B, C, and D are shown
(arbitrarily). Each site may be composed of one, two or three
(or more) amino acid positions. Following the generation of
the original saturation mutagenesis libraries, the upward
climb in the fitness landscape can proceed in several path-
ways by using the gene of a given hit from one library as
a template to perform another cycle of saturation mutagene-
sis at the other sites, and so on. When choosing a given up-
ward pathway so that each site is ‘visited’ only once (which
need not to be the case), then the scheme in Figure 2 results
in convergence and a total of only 64 experiments. Of
course, not all pathways need to be explored, but it is of con-
siderable theoretical interest to see if the order of choosing
the sites, i.e., the particular upward pathway, results in
different mutants. The nature of the to-be-improved catalytic
property determines the criterion for choosing the appropriate
sites A, B, C, D, etc.
2
2
positive mutant (hit)
Figure 1. Strategy for directed evolution of an enantioselective enzyme.10
used to enhance thermostability of proteins and stability
toward hostile solvents.11 In our case we had to develop,
inter alia, high-throughput ee assays,4e,12 which turned out
to be crucial for success. Moreover, it was not clear which
strategies for scanning protein sequence space should be
employed, and indeed, new approaches had to be developed.
Using the known molecular biological methods such as
error-prone polymerase chain reaction (epPCR), saturation
mutagenesis, and/or DNA shuffling the hits of the initial
mutant libraries (typically composed of 1000–10000 mem-
bers, i.e., clones) are used as templates for another round
of mutagenesis/expression/screening. The process can be
repeated as many times as necessary until the desired degree
of enantioselectivity has been reached (Fig. 1).10 It is the
evolutionary pressure exerted by such repeating cycles that
makes directed evolution logical. In principle, knowledge
of the structure or mechanism of the biocatalyst is not
necessary, yet much can be learned from a theoretical anal-
ysis of enantioselective mutant enzymes evolved by directed
evolution.13
The criterion for choosing the saturation mutagenesis sites
in the case of enantioselectivity or substrate scope is the
so-called combinatorial active-site saturation test (CAST).15
Accordingly, all sites having amino acids with side chains
next to the binding pocket are considered, not just one or
two sites as in previous studies regarding focused libraries.11
Thus, CASTing is the systematic generation of focused
libraries around the complete binding site. Consequently, it-
erative CASTing is an embodiment of ISM, which is useful
for enhancing enantioselectivity and enlarging the substrate
scope of enzymes.14a In the case of increasing thermostability
of a given enzyme by ISM, the criterion for choosing sensi-
tive sites is different.14b,c Accordingly, the B-factors, which
correlate with the smearing of electron density in X-ray
structures and thus flexibility are used to make a decision.
Only the positions at which the amino acids have high average
B-factors are chosen for ISM. We have shown that ISM is a
fast and efficient way to perform directed evolution in the
quest to enhance enantioselectivity14a and thermostability.14b,c
Although our original study, in which we successfully used
four rounds of epPCR in the quest to increase the enantio-
selectivity of a lipase, provided proof-of-principle,10 a new
challenge soon emerged. It became apparent that the
efficiency in probing protein sequence space had to be
improved, as in the directed evolution of other catalyst prop-
erties such as stability.11 This required new strategies such as
the appropriate application of saturation mutagenesis and
DNA shuffling.13 A more recent development is iterative
saturation mutagenesis (ISM),14 which is a symbiosis of
rational design and combinatorial saturation mutagenesis.
Accordingly, a Cartesian view of the protein is considered
in which predetermined sites composed of one, two or three
amino acid positions are chosen for saturation mutagenesis,
based on rational considerations using structural information
(X-ray or homology model).14 Saturation mutagenesis is a
molecular biological method with which amino acid ran-
domization at one, two, three or more positions is induced,
meaning the introduction of all 20 proteinogenic amino
Directed evolution of enantioselective enzymes is now well
established as a method to create biocatalysts for asymmet-
ric organic transformations.10,13,14a,16 Numerous academic
and industrial studies have appeared. Nevertheless, this
novel approach has clear limitations, the most important
one being the fact that enzymes cannot catalyze numerous
synthetically important transformations known to be possi-
ble by transition metal catalysis.1,17 Some years ago we rea-
soned that it would be intriguing to apply the principle of

6406
M. T. Reetz et al. / Tetrahedron 63 (2007) 6404–6414
Figure 3. Concept of directed evolution of hybrid catalysts showing the flow of genetic information from the gene to transition metal hybrid catalysts.18
directed evolution to synthetic transition metal catalysis. We
therefore proposed the notion of directed evolution of hybrid
catalysts.18 It had been known for decades that achiral
ligand/metal moieties can be anchored covalently or non-
covalently to proteins acting as a host.19 Accordingly, a given
system provides a single chiral catalyst, the wild-type (WT)
protein providing a defined local environment around the
synthetic catalytically active transition metal center. One
of the milestones in this type of catalysis is the Whitesides
system, in which a biotinylated diphosphine/rhodium moiety
is allowed to interact non-covalently with avidin.20 Low ee
values in olefin-hydrogenation were observed (see below),
but a new concept had been introduced by this seminal study.
Nevertheless, in this and other systems,19 there is no reason
to believe that the particular protein environment is optimal.
Therefore, the directed evolution of such hybrid catalysts
provides a fascinating perspective: the possibility to tune a
synthetic transition metal catalyst by an evolutionary ap-
proach (Darwinism in the test tube!).18 The concept is shown
schematically in Figure 3. The first study providing proof-of-
principle was published recently21 (see Section 2).
One of the first examples that we proposed18a for the directed
evolution of hybrid catalysts was the Whitesides system.20
Accordingly, biotin is first attached covalently to an achiral
diphosphine–Rh complex via a spacer to form a complex of
the type 1, which then binds with high affinity to avidin.
Whitesides actually used the analogous complex in which
norbornadiene takes the place of cod. The respective
complex was used as a single catalyst in the Rh-catalyzed
hydrogenation of a-acetamido-acrylic acid, the observed
ee ranging between 33% and 44% depending upon the con-
ditions used.20 Ward has used chemical tuning (variation of
spacer length in 1) and rational protein design to improve the
enantioselectivity significantly in the same reaction using
streptavidin,22 which is fundamentally different from our
Darwinistic approach. We likewise employed the biotiny-
lated diphosphine–Rh complex 1, but chose to use the ester-
ified substrate 2 because this facilitates medium-throughput
analysis by gas chromatography.21 The reaction mixtures
can be extracted with ethyl acetate in a parallel manner,
whereas the acid of the Whitesides–Ward system is accessible
efficiently only by continuous extraction.
O
Ph2
2. Results
HN NH
P
Rh
BF4
H
H
PPh2
N
S
As in any fundamentally new venture, proof-of-principle
was the immediate goal.21 The challenge in putting the
scheme shown in Figure 3 into practice is formidable for
several reasons. Firstly, unlike the process in normal directed
evolution of enzymes (Fig. 1), the (over)expressed mutant
protein present in each well of the microtiter plates needs
to be separated from the rest of the proteins that are of no
interest, but that are nevertheless present. Normally, such
proteins do not affect the process of directed evolution of
enzymes adversely, but in the present system chemical
modification with introduction of metal/ligand entities
would occur indiscriminately with formation of numerous
catalysts in each well. Secondly, in classical directed evolu-
tion, only a very small amount of mutant is needed in each
well due to the high activity of enzymes. In contrast, syn-
thetic catalysts have much lower activities, which means
that considerably larger amounts of a mutant protein need
to be present in each well. Therefore, the expression system
of the protein host has to be very efficient. Thirdly, chemical
modification has to be essentially quantitative. If part of the
achiral synthetic catalyst remains unbound in the presence of
the desired hybrid catalyst, false information will be gener-
ated by the screening system. We originally underestimated
these technical problems,18 which is the reason why it has
taken us so long to provide the proof-of-principle.21
O
1
H2 (6 bar)
catalyst 1 (0.2%)
protein (1.3 free binding
site per catalyst)
H
N
O
H
N
O
O
O
OMe
OMe
*
H2O, 10% DMF
(1 mL)
2
3
0.1 M AcOH, pH 4
r.t., 8 h; [2] = 4·10-2 M
Initially we considered avidin as the host protein,18a but
production of eukaryotic proteins is both time consuming
and low yielding. We therefore turned to streptavidin, a
genetically unrelated bacterial protein, which also binds
biotin with high affinity.21 Several expression systems for
streptavidin have been described,23 and some of them were
compared. Unfortunately, we ran into problems with the in-
sufficient expression level and purification in parallel form.
The best solution for our purpose turned out to be based
on pET11b-sav,24 which encodes 12 residues of T7-tag
followed by Asp and Gln and residues 15–159 of the mature
streptavidin. Escherichia coli strain BL21(DE3) transformed
with this plasmid and grown in Studier’s auto-induction
media,25 ZYP5052, requires less monitoring than conven-
tional induction with IPTG at mid-log phase and thus allows

M. T. Reetz et al. / Tetrahedron 63 (2007) 6404–6414
6407
multiple unattended overnight cultures.21 Before reaction
with 1, streptavidin mutants were purified using standard
agarose/iminobiotin affinity column chromatography. Strep-
tavidin or streptavidin/biotin themselves do not catalyze the
hydrogenation reaction. Unbound complex 1 does not occur
because an excess of free binding site of streptavidin was
present. Per tetramer of streptavidin, 3.8–3.9 free binding
sites were determined by standard titration using fluores-
cence quenching of biotin/4-fluorescein. About 150 ee deter-
minations were made.
Our optimized adaptation was crucial in simplifying the
screening task.21 However, the system is still not fully suited
for screening thousands of mutants since it requires a
150 mL culture scale to provide a sufficient amount of strep-
tavidin. Typically we used 1.04ꢀ10ꢁ7 mol of binding site,
which is equivalent tow1.7 mg protein (based on a MW of
16.5 kDa per monomer and expecting 3.8–3.9 free binding
sites per tetramer as obtained for the WT).21
Figure 4. Selected close (purple) and distal (blue) sites from Rh(I) centers
(red) of two important calculated conformers of the complex streptavidin/
1
21 based on the X-ray structure of streptavidin/biotin.27
When lower amounts of streptavidin were obtained for
a given mutant, the culture scale had to be increased by up
to fivefold and/or the amount of hybrid catalyst used in the re-
action was decreased from 0.2% to 0.1%. Following titration
of the mutant streptavidins for the purpose of determining the
amount present, they were transferred into glass vessels of
an in-house adapted reactor block,26a which was used in
the Chemspeed AcceleratorÔ SLT 100 Synthesizer.21 It is
important to point out that traces of O2 need to be removed
from the reactors, which otherwise destroy the Rh-catalyst.
Enantioselectivity was determined by conventional gas
chromatographic analysis of the reaction mixtures.
Leu110, Ser112, and Leu124. These are located about 4–
˚
6 A away from the Rh(I) of the two calculated major con-
formers. We speculated that they could influence directly
the conformation of the catalyst or catalyst/substrate com-
plex.21 The second type of sites is ‘distal’: Glu51, Tyr54,
Trp79, Asn81, Arg84, Asn85, and His87, which are located
further away from Rh(I). These second sphere CAST-
positions could influence the structure of the enzyme as
a whole because they are involved in hydrogen bonding
between secondary elements.
Saturation mutagenesis was initiated at positions 110, 112,
and 124 using the QuikChange method (Stratagene) and
pET11b-sav.24 In each saturation experiment about 200–
300 clones were harvested and screened, which correspond
to an oversampling of >95% coverage.28
It is obvious that this system really does not fulfill the tech-
nological requirements for implementing the concept of
directed evolution of hybrid catalyst efficiently. Thus, we had
two options: (1) invest more time in improving the (over)
expression system for streptavidin, which would subse-
quently allow the practical formation of large libraries of
mutant protein hosts (each mutant in sufficiently large quan-
tities); or (2) use the partially optimized expression system
described above, and then produce fairly small libraries.
We chose the latter strategy, i.e., we settled for only a few
hundred mutants in each mutagenesis experiment and pro-
ceeded with directed evolution on a ‘small scale’, which in
fact was still labor intensive.21
Unfortunately, we observed that in some cases protein
variants were formed in amounts too small for fulfilling
the experimental prerequisites regarding reproducible hy-
drogenation. Thus, these were not considered for further
study. Nevertheless, a few mutants showing enantioselectiv-
ity different from the WT were observed in a reproducible
manner, the best variant I leading to 35% ee (R).21 It is char-
acterized by mutation Ser112Gly.
It is clear that a single round of saturation mutagenesis does
not yet constitute an evolutionary process. Therefore, itera-
tive CASTing,14a which is an embodiment of iterative satu-
ration mutagenesis (ISM), was performed using the gene that
encodes mutant I and saturating at position 49. This led to
a double mutant II having mutations Asn49His/Ser112Gly
and showing an ee value of 54% (R) in the model reaction
(Fig. 5).21 Finally, a third-generation saturation experiment
was performed using the gene, which encodes mutant II
and focusing once more on position 112. This experiment
was designed to test whether glycine at position 112 is really
the best choice when combining with histidine at position
49. This provided the improved mutant III leading to an
ee of 65% (R).21 Surprisingly, the original mutation
Ser112Gly was reverted back to serine (Fig. 5). This means
that the best variant is characterized by a single mutation
(Asn49Val). We then proceeded to positions 51, 54, 79, 81,
In an initial experiment we observed that WT-streptavidin/1
is a poor catalyst in the hydrogenation of 2, leading to an ee
of only 23% in favor of (R)-3. Rather than targeting the
whole protein for amino acid substitution by error-prone
PCR,11 we applied CASTing15 in the present hybrid catalyst
system.21 Unfortunately, an X-ray structure of the conjugate
was not available. Therefore, the CAST sites for amino acid
randomization were chosen on the basis of modeling the
biotinylated Rh-complex 1 into the X-ray structure27 of
streptavidin/biotin using Moloc and Accelerys DS visual-
izer. Figure 4 shows an excerpt of the modeled structure
and the amino acid sites that appeared to be appropriate
for CAST experiments.21
Two types of sites for saturation mutagenesis were consid-
ered: the first is ‘close’ positions, specifically Asn49,

6408
R (% ee)
M. T. Reetz et al. / Tetrahedron 63 (2007) 6404–6414
+65%
III (Asn49Val)
L
L
60
50
chemical
M
+54%
FG
FG
protein
modification
II (Asn49Val/Ser112Gly)
protein
40
+35%
I (Ser112Gly)
Figure 6. Implantation of ligand/metal moieties in proteins (FG¼functional
30
group).18,26
WT
20
+23%
IV (Asn49His)
+8%
appropriate functional group anywhere in the protein as
desired (or remove it).
10
Racemic
10
V (Asn49His/Leu124Phe)
-7%
Indeed, these ideal characteristics contribute to the reason
why the WT papain has been modified chemically in previ-
ous studies at position Cys25.19 Either SN2-reactions or
Michael additions have been used in such endeavors, and this
is exactly what we strived for in exploratory experiments.18,26
For example, a manganese–salen complex 7 was introduced
by a Michael process with formation of 8. Dipyridyl moie-
ties complexing copper, palladium or rhodium were also
introduced in papain by appropriate reactions at Cys25.18,26
Preliminary experiments regarding epoxidation using the
hybrid catalyst containing the Mn/salen moiety and hydro-
genation employing the Rh-catalyst led to ee values of up
to only 10%, but this is not surprising.18,26 Again, there is
no reason to believe that an achiral synthetic catalyst in a
given chiral protein environment should necessarily lead to
high enantioselectivity.
S (% ee)
Figure 5. Directed evolution of hybrid catalysts comprising streptavidin/1,
the Rh-catalyzed hydrogenation 2/3 serving as the model reaction
(40–90% yield).21
84, 85, and 87. Unfortunately, saturation mutagenesis exper-
iments were not successful because no soluble protein was
obtained. In contrast, saturation mutagenesis at position 49
on WT template led directly to mutants III and IV. The latter
is characterized by Asn49His. This variant has lower enan-
tioselectivity than the WT (ee¼8% (R)), suggesting the pos-
sibility of inverting stereoselectivity. Therefore, the plasmid
encoding variant IV was utilized as a template for saturation
mutagenesis in hope of inverting the sense of enantioselec-
tivity. Indeed, upon focusing on position 124 mutant V
(Asn49His/Leu124Phe) was identified, which is (S)-selec-
tive, although not by a great degree (ee¼7%).21 It would
be interesting to expand the present study by applying itera-
tive saturation mutagenesis to residues in the second shell
around the close amino acid positions already considered
(49, 110, 112, and 124).
N
O
N
O
N
OH
HO
HO
OH
+
5
O
4
DIC, NaHCO3
DMAP (cat.),
THF
(53%)
This work demonstrates for the first time that it is possible to
apply the methods of directed evolution to increase and/or to
invert enantioselectivity of a hybrid catalyst composed of
a synthetic achiral transition metal catalyst anchored to
a host protein.21 Due to the technical problems associated
with the expression system, only very small mutant libraries
could be generated, which in itself was labor intensive.
Nevertheless, proof-of-principle has been provided for the
first time.21
N
N
O
O
N
O
HO
OH
O
6
Mn(OAc)2,
LiCl, O2
CH2Cl2/MeOH
(90%)
Parallel to our efforts regarding non-covalent binding based
on the biotin/streptavidin system,18,21 we also considered
covalent binding (Fig. 6).18,26
N
O
N
O
O
MnCl
N
O
O
The functional group (FG) in an appropriate binding pocket
or cavity of a protein can be, for example, a thiol moiety
belonging to cysteine or hydroxy originating from serine.
It is conceivable that the WT of a potential protein host
harbors only a single cysteine in a cavity large enough to
accommodate the synthetic catalyst (and also the substrate
in subsequent catalysis). This is the case with papain, a
well-known cysteine protease having the only free cysteine
in a relatively large cavity (Fig. 7).29 Of course, site-specific
mutagenesis can be used to place an amino acid with an
O
7
papain
N
N
O
O
MnCl
N
O
O
O
papain
S
O
8

M. T. Reetz et al. / Tetrahedron 63 (2007) 6404–6414
6409
In the quest to find yet another alternative protein for hosting
synthetic transition metal catalysts, we turned to serum albu-
mins.30 These robust and easy to handle proteins are present
at high concentrations in blood plasma, functioning as trans-
port carriers for a variety of compounds such as fatty acids,
bile acids, bilirubin, and hemin.31 It was known from the X-ray
work of Curry that iron–protoporphyrin dimethyl ester binds
in the subdomain IB of human serum albumin (HSA) and
that weak axial coordination by Tyr161 contributes to the
binding.32 Moreover, Gross had previously anchored water-
soluble sulfonylated FeIII– and MnIII–corroles to various
serum albumins and used these conjugates as catalysts in
the H2O2-based asymmetric sulfoxidation of prochiral thio-
ethers (up to 74% ee using WT).33 In addition, it had been
reported that the sodium salts of di-, tri-, and tetrasulfonic
acid derivatives of porphyrins, phthalocyanines, and corroles
bind strongly to serum albumins,34 analogously to iron–
protoporphyrin dimethyl ester.32
We therefore contemplated the use of the commercially
available CuII–phthalocyanine complex 13, expecting it to
bind strongly to the IB subdomain of HSA or to analogous
regions of other serum albumins such as bovine serum
albumin (BSA), which is also a cheap and robust protein.
Figure 7. Globular structure of papain featuring a single free cysteine
(Cys).26b,29
NaO3S
SO3Na
We also proposed the idea of introducing bidentate ligand
systems in lipases by designing appropriate phosphonate
inhibitors.18a,26 Itwaswellknownthatphosphonateinhibitors
reactcovalentlyatthecatalyticallyactiveserinesiteofalipase
selectively in the presence of other serine moieties in the
enzyme. We were able to prove the incorporation of diphos-
phine 11in thelipase fromBacillussubtilis(Lip A). However,
after 24 h in water the inhibitor was hydrolytically cleaved,
which is not surprising due to the presence of the second
p-nitrophenol leaving group. It became clear that phospho-
nate 12 has to be replaced by an analog, which bears only
one leaving group, i.e., by exchanging one of the p-nitrophe-
nol groups for an alkoxy or amino moiety.18a,26 Such work is
in progress, but we also concentrated on other approaches.
N
Cu
N
N
N
N
N
N
N
SO3Na
NaO3S
13
In the case of HSA we employed the X-ray data of Curry to
model the desired 13/HSA complex (Fig. 8).30 Absolute
proof that binding occurs in this manner was not obtained,
i.e., Figure 8 simply presents a useful model.
As the model reaction to be mediated by hybrid catalysts
comprised of 13/serum albumins, we chose the Diels–Alder
reaction of the H2O-soluble azachalcone 14 with cyclopen-
tadiene 15 to form adduct 16. This reaction had been origi-
nally devised by Engberts,35 who used CuII-complexes of
amino acids in aqueous medium (ee up to 74%). Later it
O
P
O
O
O
O
Ph2P
NH x HCl
+
O2N
NO2
PPh2
10
NO2
9
O
O
P
Ph2P
DMAP (1.0 eq.),
NEt3 (1.0 eq.),
r.t., 18 h
N
O
O
NO2
PPh2
(60%)
11
NO2
O
P
O
O
Ph2P
N
O lipase
lipase
PPh2
12
Figure 8. Model of 13/HSA30 based on the X-ray structure of hemin/HSA.32
NO2

6410
M. T. Reetz et al. / Tetrahedron 63 (2007) 6404–6414
was employed by Feringa in the study of CuII-conjugates of
DNA as catalysts.36
from ee¼23% to ee¼65%.21 If in the future a highly
improved expression system for streptavidin could be devel-
oped, then this system should be re-considered. It is very
likely that much better enantioselectivities can then be
achieved by iterative CASTing using considerably larger
mutant libraries or by other mutagenesis methods such as
DNA shuffling.
O
N
R
+
R
Until that goal has been reached, other systems need to be
considered, such as appropriate lipase/inhibitor complexes,
HSA conjugates, or alternative approaches. Another possi-
bility is to employ thermophilic proteins as hosts, which
can be purified by simple heating.26b,c,38 Once the optimal
system has been established (or systems!), a platform (or
platforms) would exist from which a number of intriguing
goals can be strived for, especially in view of the amazingly
large number of catalytic aqueous organometallic reactions
already known.17b No doubt, the use of molecular biology
to tune a synthetic transition metal catalyst in a Darwinistic
sense (Fig. 3) offers exciting perspectives for the future,
including the theoretical interpretation of enhanced enantio-
selectivity and/or activity.39
O
N
14 a R = H
b R = CH3
15
16
c R = OCH3
d R = NO2
e R = Cl
Surprisingly, we observed high enantioselectivities when
using BSA as the host for 13 (85–93% ee) in favor of the
expected endo-products 16.30 In the case of 13/HSA (Fig. 8),
the reaction of 14a with 15 led to an ee value of 85%. BSA or
HSA alone do not catalyze the reaction. Thus, since a good
expression system for HSA has been reported,37 this system
could be employed in a future study regarding the directed
evolution of enantioselective hybrid catalysts. Iterative
CASTing based on the model in Figure 8 could then be
used to increase the enantioselectivity of the present
Diels–Alder cycloaddition or of other CuII-catalyzed pro-
cesses. However, a procedure for en masse purification of
HSA mutants needs to be developed first.
5. Experimental
5.1. Directed evolution of the hybrid catalyst comprising
streptavidin/1
5.1.1. Synthesis of the Rh-complex 1. To a solution of
[Rh(cod)2]BF4 (80.3 mg; 197 mmol) in dry dichloromethane
(7 mL) was added at 0 ꢂC (3aS,4S,6aR)-5-(2,3,3a,4,6,6a-
hexahydro-2-oxo-1H-thieno[3,4-d]imida-zol-4-yl]-pentanoic
acid-N,N-bis-(2-diphenyl-phosphanyl-ethyl)amide20,40 (165 mg;
247 mmol). Dichloromethane (7 mL) was added, the cooling
bath was removed, and the mixture was stirred for 3 h. After
cooling to 0 ꢂC, n-pentane (40 mL) was added and the
mixture stirred for 0.5 h. The precipitated yellow solid was
filtered, washed three times with n-pentane (60 mL), and
dried in vacuo for 16 h. This provided compound 1
3. Discussion
In order to put the idea of directed evolution of hybrid
catalysts into practice, we have considered three different
approaches, namely the biotinylated Rh–diphosphine com-
plex 1 bound non-covalently to streptavidin, transition metal
complexes such as Mn/salen entities bound covalently to
papain, and a Cu(II)–phthalocyanine complex anchored
non-covalently to serum albumins. Proof-of-principle was
achieved only in the first case, which provokes the question
regarding the source of enhanced or reversed enantioselec-
tivity. Presently, it is essentially impossible to provide a
sound answer on a molecular level. It is not even clear
whether the Whitesides system adheres to the Halpern (or
anti-Halpern) rule. Thus, more work is necessary to clarify
these points.
(169 mg; 89%) as a yellow solid. IR (KBr): n¼3402 cmꢁ1
,
3249 (br, NH), 3054 (w, Ar-H), 2922 (m, CH2), 1706 (s,
(HN)2C]O), 1643 (s, CH2C]O), 1586 (w, CAr]CAr),
1572, 1480, 1435 (s, P-Ph), 1370, 1333, 1311, 1267, 1182,
1160, 1055 (br, H–C]), 998, 940, 859, 814, 746 (m, Ar-
1
H), 698 (s, Ar-H), 518, 492. H NMR (400 MHz, CD2Cl2)
d 1.19–1.32 (m, 3H, 3-H, 4-H), 1.34–1.51 (m, 3H, 3-H, 5-
H), 1.92–1.94 (m, 2H, 1-H), 2.11–2.28 (m, 12H, COD–
4. Conclusions
2
CH2, NCH2CH2PPh2), 2.59 (d, J¼12.8 Hz, 1H, 60-Hb),
3
Following our original proposal regarding the directed
evolution of hybrid catalysts (Fig. 3) and some exploratory
experiments,18 we were able to provide proof-of-principle
in 2006.21 This was based on the use of the Whitesides sys-
tem comprising a biotinylated Rh–diphosphine complex and
streptavidin as the protein host.21 However, due to the inef-
ficiency of the current expression systems of streptavidin,
only very small mutant libraries even after intensive labora-
tory effort could be generated (less than 300 clones per
library). Nevertheless, we applied iterative CASTing suc-
cessfully in the ‘mini’ directed evolution study.21 In three
cycles of iterative saturation mutagenesis at selected sites
around the modeled Rh-center in the hybrid catalyst the ee
of Rh-catalyzed olefin-hydrogenation increased stepwise
2.82 (dd, 2J¼12.8 Hz, J¼5.0 Hz, 1H, 60-Ha), 3.08 (q,
2
3
3J¼7.7 Hz, 40-H), 3.34 (dd, J(H,P)¼19.9 Hz, J¼4.9 Hz,
2H, E-NCH2CH2PPh2), 3.72 (dd, 2J(H,P)¼21.4 Hz,
3J¼5.2 Hz, 2H, Z-NCH2CH2PPh2), 4.18–4.20 (m, 1H, 3a0-
H), 4.37–4.40 (m, 5H, COD–CH, 6a0-H), 5.01 (br s, 1H,
30-H), 5.61 (br s, 1H, 10-H), 7.03–7.46 (m, 20 H, Ar-
H) ppm. 13C NMR (101 MHz, CD2Cl2) d 21.57 (ꢁ, CH2),
24.04 (ꢁ, CH2), 27.58 (ꢁ, CH2), 29.52 (ꢁ, CH2), 30.33
(ꢁ, CH2), 32.61 (ꢁ, CH2), 39.84 (ꢁ, C-40), 45.88 (ꢁ, E-
NCH2CH2PPh2), 46.99 (ꢁ, Z-NCH2CH2PPh2), 56.29 (+,
C-30), 60.80 (+, C-6a0), 62.55 (+, C-3a0), 99.27 (+, COD–
CH), 101.58 (COD–CH), 128.26 (+, Ar-C), 128.36 (+, Ar-
C), 128.64 (+, Ar-C), 130.59 (+, Ar-C), 130.72 (+, Ar-C),
131.77 (+, Ar-C), 131.86 (+, Ar-C), 132.15 (+, Ar-C),

M. T. Reetz et al. / Tetrahedron 63 (2007) 6404–6414
6411
162.58 (Cquart, C-20), 173.70 (Cquart, C-1) ppm. 31P NMR
(162 MHz, CD2Cl2) d 13.91 (dd, 1J(P,Rh)¼143 Hz, 2J(P,P)¼
34.7 Hz, PPh2), 22.69 (dd, 1J(P,Rh)¼142 Hz, 2J(P,P)¼
34.3 Hz, PPh2) ppm. MS (ESI/pos., CH2Cl2) m/z (%): 878
(7) [M+ꢁBF4], 770 (100) [M+ꢁCOD–BF4]. Elemental anal-
ysis for C46BF4H55N3O2P2SRh (965.68): C 57.28 (57.24),
H 5.81 (5.74), N 4.30 (4.35), P 6.46 (6.41), Rh 10.60
(10.66) found (calculated).
Analysentechnik, Meerbusch, Germany), carrier (N2), flow
1.7 mL/min, temperature profile 100 ꢂC for 12.5 min. The
uncertainty in the ee values is ꢃ2%.
5.2. Synthesis of papain-conjugate 8
5.2.1. Preparation of salen 4. The procedure of Jacobsen42
was used: to the mixture of 2,5-dihydroxy-benzaldehyde
(61 mg; 0.44 mmol), 3,5-di-tert-butylsalicylaldehyde (310 mg;
1.32 mmol), and ethylene diamine (53 mg; 0.88 mmol) in
20 mL of dichloromethane was added 2 g of silica gel. After
stirring for 18 h at room temperature, the solvent was re-
moved in vacuo. The solid (SiO2 and compounds) was
brought onto a chromatography column (25 mm diameter)
and filled with 20 g of silica gel. Using a 3:1 mixture of
hexane/diethyl ether and an argon pressure, elution of the
undesired product (non-polar yellow symmetrical salen)
was performed, followed by elution of 4 using a 1:1 solvent
mixture. This has to occur rapidly, avoiding high tempera-
tures, so that undesired new formation of symmetrical salen
products is avoided. The solvent is removed in vacuo, yield-
5.1.2. Expression system of streptavidin and mutant
library formation. The best expression system for the pro-
ject, although not ideal, is based on pET11b-sav described
by Gallizia,24 which was employed in all experiments.
This construct encodes 12 residues of T7-tag followed by
Asp and Gln and residues 15–159 of the mature streptavidin.
E. coli strain BL21(DE3) transformed with this plasmid and
grown in Studier’s auto-induction media,25 ZYP5052, re-
quires less monitoring than conventional induction with
IPTG at mid-log phase and thus allows multiple unattended
overnight cultures. This adaptation was crucial in simplify-
ing the screening task. It requires a 150 mL culture scale
to provide a sufficient amount of streptavidin. Typically
1.04ꢀ10ꢁ7 mol of binding site was used, which is equivalent
tow1.7 mg protein (based on a MW of 16.5 kDa per mono-
mer and expecting 3.8–3.9 free binding sites per tetramer as
obtained for the WT). However, when lower amounts of
streptavidin were obtained for a given mutant, the culture
scale was increased by up to fivefold and/or the amount of
hybrid catalyst used in the reaction was decreased to 0.1%.
Titrated mutant streptavidins were transferred into glass
vessels of an in-house adapted reactor block21,26a for the
Chemspeed AcceleratorÔ SLT 100 Synthesizer. Saturation
mutagenesis at the defined sites in streptavidin (Fig. 1)
was performed using the standard QuikChange method of
Stratagene41 and pET11b-sav.24
1
ing 4 as an orange-yellow solid (134 mg; 77%). H NMR
(400.1 MHz, d6-DMSO) d 1.26 (s, 9H, CH3), 1.37 (s,
9H, CH3), 3.90 (m, 4H, CH2 diiminoethane), 6.77 (d,
3JH,H¼8.6 Hz, 1H, CH), 6.79 (s, 1H, CH), 6.69 (d,
3JH,H¼8.6 Hz, 1H, CH), 7.24 (s, 1H, CH), 7.30 (s, 1H,
CH), 8.49 (s, 1H, HC]N), 8.58 (s, 1H, HC]N), 8.93 (s,
1H, OH), 12.48 (s, 1H, OH), 13.95 (s, 1H, OH). 13C NMR
(100.6 MHz, d6-DMSO) d 29.6 (CH3), 31.6 (CH3), 34.2
(quart. C), 34.9 (quart. C), 59.1 (CH2 diiminoethane), 59.4
(CH2 diiminoethane), 116.8 (CH), 117.2 (CH), 118.0 (C im-
ine), 119.2 (C imine), 120.3 (CH), 126.5 (CH), 126.6 (CH),
136.0 (C-t-butyl), 139.8 (C-t-butyl), 150.0 (C–OH), 153.1
(C–OH), 158.2 (C–OH), 167.1 (C]N), 168.4 (C]N). MS
(EI) m/z (%): 396 (100) [M+], 381 (30) [M+ꢁCH3], 259
(14), 234 (52), 218 (19), 163 (16).
5.1.3. Enantioselective hydrogenation. All reactions were
performed using the Chemspeed AcceleratorÔ SLT 100
Synthesizer21 equipped with in-house adapted reactor
blocks.26a All solvents were thoroughly degassed three times
as well as redistilled in the case of dimethylformamide. Such
cautions must be taken due to the high sensitivity of the
reaction toward oxygen. The reaction vessels of the reactor
block26a were evacuated at high vacuum and flushed with
argon, a process that was repeated three times. The catalyst
solution was prepared at room temperature by the addition
of a solution of [Rh(cod)2]BF4 in dimethylformamide to a
stirred solution of the ligand in dimethylformamide as
well. The catalyst thus prepared has to be used immediately
as it degrades readily after a few hours. The preformed Rh-
catalyst (Section 5.1.1) can also be used. First, the substrate
in solution in the acetate buffer was added to the protein,
lyophilized in the reaction flask, and then the catalyst solu-
tion was added to the shaken solution. All conditions (H2
(6 bar); 0.2% Rh-catalyst 1; H2O/10% DMF; 0.1 M
AcOH/pH 4; 22 ꢂC; 8 h) are optimal, although the catalyst
amount can be lowered with some confidence to 0.01%
whenever needed.
5.2.2. Preparation of ester 6. The mixture of 4 (134 mg;
0.338 mmol), acid 543 (122 mg; 0.721 mmol), diisopropyl-
carbodiimide (170 mmol), and NaHCO3 (75 mg;
0.89 mmol) in a 0.1 M solution of N,N-4-dimethylaminopyr-
idin in THF (20 mL) was stirred for 2 h at room temperature.
The solvent was removed in vacuo and the mixture extracted
with ethyl acetate. Chromatography over silica gel (20 g)
using a 1:1 mixture of ethyl acetate and hexane afforded 6
as a yellow solid (115 mg; 62%). IR (KBr, cmꢁ1): 2960
(s, n[C–H], t-butyl), 1750 (m, n[C]O], aryl ester), 1710
(s, n[C]O], maleinimide), 1640 (s, n[C]N]), 1170 (m,
n[C–O], aromatic), 830 (w, d[]CH], aryl). 1H NMR
(400.1 MHz, d4-methanol) d 1.32 (s, 9H, CH3), 1.44 (s,
3
9H, CH3), 2.88 (t, JH,H¼6.8 Hz, 2H, CH2 propionic acid),
3
3.94 (t, JH,H¼6.8 Hz, 2H, CH2 propionic acid), 3.99 (m,
4H, CH2 diiminoethane), 6.87 (s, 2H, CH maleinimide),
6.88 (dd, JH,H¼4.9 und 8.9 Hz, 2H, CH aromatic), 7.12
3
3
(s, 1H, CH aromatic), 7.09 (d, JH,H¼8.9 Hz, 1H, CH aro-
matic), 7.19 (s, 1H, CH aromatic), 7.41 (s, 1H, CH aro-
matic), 8.47 (s, 1H, CH imine), 8.48 (s, 1H, CH Imin). 13C
NMR (100.6 MHz, d4-methanol) d 30.0 (CH3), 32.0
(CH3), 34.1 (CH2 propionic acid), 34.7 (CH2 propionic
acid), 35.0 (quart. C), 36.0 (quart. C), 60.2 (CH2 diimino-
ethane), 60.5 (CH2 diiminoethane), 118.7 (CH aromatic),
119.5 (C imino group), 119.7 (C imino group), 125.0 (CH ar-
omatic), 127.3 (CH aromatic), 127.5 (CH aromatic), 129.9
The reaction mixtures were extracted with ethyl acetate and
analyzed by chiral gas chromatography using GC Hewlett
Packard 6890N chromatograph equipped with DiMePe-
BETA-ivadex-1 chiral column (25 m, 0.25 mm, 0.15 mm,

6412
M. T. Reetz et al. / Tetrahedron 63 (2007) 6404–6414
(CH aromatic), 135.6 (CH maleinimide), 137.5 (C-t-butyl),
138.7 (C-t-butyl), 141.4 (C–O aromatic), 143.4 (C–O aro-
matic), 167.6 (C]N), 169.4 (C]N), 172.2 (C]O), 180.8
(C]O). MS (EI) m/z (%): 547 (100) [M+], 532 (21)
[M+ꢁCH3], 396 (21), 259 (19), 234 (29) [C4H9].
(15) and stirred for 3.5 days at 5 ꢂC, followed by extraction
with diethyl ether. Subsequently, 10 mL of methanol was
added to the aqueous phase to destroy the structure of pro-
tein. Following extraction with ethyl acetate (90 mL/three
times), the organic phases were combined and dried with
anhydrous Na2SO4. After evaporation, the residue was puri-
fied through silica gel chromatography (eluent: petroleum
hexane/ethyl acetate¼10:1) to afford 92 mg of 16a (85%,
endo/exo¼93:7, 89% ee for endo-16a). The stereochemical
assignments were made by comparison with authentic
samples.35,36
5.2.3. Preparation of Mn-complex 7. The solution of ester
6 (17 mg; 31 mmol), Mn(OAc)2$2H2O (11 mg; 61 mmol),
and LiCl (13 mg; 31 mmol) in methanol (4 mL) was stirred
under air for 16 h at room temperature. The volume of the
mixture was reduced in vacuo to 0.5 mL and then diluted
with CH2Cl2. After washing with an aqueous solution of
NaCl, the organic phase was separated and dried over
Na2SO4. Removal of the solvent in vacuo afforded 7
(18 mg; 90%) as a black solid. MS (methanol, ESI pos.)
m/z (%): 600 [MꢁCl]+, 632 [MꢁCl+CH3OH]+. High resolu-
tion MS (ESI, methanol): calcd for C31H35N3O6Mn:
600.190639, found: 600.190776, difference: 0.23 ppm. IR
(KBr, cmꢁ1): 2960 (s, n[C–H], t-butyl), 1750 (m, n[C]O],
aryl ester), 1710 (s, n[C]O], imide), 1620–1540 (s,
n[C]C], n[C]N]), 1170 (m, n[C–O], aromatic), 830 (w,
d[]CH], aryl).
Acknowledgements
This work was supported by the Fonds der Chemischen
Industrie, the Deutsche Forschungsgemeinschaft (SPP
1170, RE 359/13-1), and the EU (IBAAC MCRTN-CT-
2003-505020).
References and notes
5.2.4. Formation of papain conjugate 8. In an initial step
papain-lyophilisate (10 mg) obtained from Sigma–Aldrich
was activated by treatment with L-cysteine (10 mg) in H2O
(1 mL) for 30 min. The non-reacted cysteine was then sepa-
rated by gel permeation chromatography (Sephadex G-25;
50 mM acetate buffer at pH¼5.2). The protein fraction
was identified by its absorption at 280 nm and its proteolytic
activity tested. The fractions containing active papain (about
3 mL) were combined and kept at 4 ꢂC for further use. This
mixture (3 mL) was mixed with phosphate buffer (300 mL;
1 M) and brought to pH¼7.0. It was then treated with com-
plex 7 (0.648 mg; 1.08 mL of DMSO). After 6 h no protease
activity was observed, showing that conjugation with forma-
tion of 8 had occurred.
1. (a) Knowles, W. S. Angew. Chem. 2002, 114, 2096–2107;
Angew. Chem., Int. Ed. 2002, 41, 1998–2007; (b) Noyori, R.
Angew. Chem. 2002, 114, 2108–2123; Angew. Chem., Int. Ed.
2002, 41, 2008–2022; (c) Sharpless, K. B. Angew. Chem.
2002, 114, 2126–2135; Angew. Chem., Int. Ed. 2002, 41,
2024–2032; (d) Comprehensive Asymmetric Catalysis;
Jacobsen, E. N., Pfaltz, A., Yamamoto, H., Eds.; Springer:
Berlin, 1999; Vols. I–III.
2. Reviews of organocatalytic reactions: (a) Lelais, G.;
MacMillan, D. W. C. Aldrichimica Acta 2006, 39, 79–87; (b)
Seayad, J.; List, B. Org. Biomol. Chem. 2005, 3, 719–724; (c)
List, B. Tetrahedron 2002, 58, 5573–5590; (d) Asymmetric
€
Organocatalysis; Berkessel, A., Groger, H., Eds.; Wiley-
VCH: Weinheim, 2004; (e) Dalko, P. I.; Moisan, L. Angew.
Chem. 2001, 113, 3840–3864; Angew. Chem., Int. Ed. 2001,
40, 3726–3748.
5.3. Enantioselective Diels–Alder reactions catalyzed
by the hybrid catalyst 13/BSA
3. (a) Enzyme Catalysis in Organic Synthesis: A Comprehensive
Handbook, 2nd ed.; Drauz, K., Waldmann, H., Eds.; VCH:
Weinheim, 2002; Vols. I–III; (b) Biotransformations in
Organic Chemistry, 4th ed.; Faber, K., Ed.; Springer: Berlin,
2000; (c) Schmid, A.; Dordick, J. S.; Hauer, B.; Kiener, A.;
Wubbolts, M.; Witholt, B. Nature (London) 2001, 409, 258–
268; (d) Koeller, K. M.; Wong, C.-H. Nature (London) 2001,
409, 232–240.
5.3.1. Small scale reaction. An aqueous formate buffer
(2 mL; 30 mM; pH 4.0) was prepared and 39.6 mg of BSA
added. Subsequently a 12.5 mL solution of the commercially
available Cu-complex 13 (Aldrich) in H2O (40 mM) was
added. This solution was stirred at room temperature for
30 min. Then, 21.2 mL solution of dienophile 14a in
CH3CN (0.94 M) was added. The resulting mixture was
cooled to 3 ꢂC. The reaction was started with the addition
of 10 mL pure cyclopentadiene (15) and stirred for 3 days
at 3 ꢂC, followed by extraction with diethyl ether (6 mL),
removal of the ether, and measurement of the conversion
and ee value by chiral HPLC (Chiralcel-OD-H column, elu-
tion with n-heptane/i-propanol¼98:2; anisol as the internal
standard).
4. (a) Gennari, C.; Piarulli, U. Chem. Rev. 2003, 103, 3071–3100;
€
(b) Dahmen, S.; Brase, S. Synthesis 2001, 1431–1449; (c)
Hoveyda, A. H. Handbook of Combinatorial Chemistry;
Nicolaou, K. C., Hanko, R., Hartwig, W., Eds.; Wiley-VCH:
Weinheim, Germany, 2002; Vol. 2, pp 991–1016; (d)
de Vries, J. G.; de Vries, A. H. M. Eur. J. Org. Chem. 2003,
799–811; (e) Reetz, M. T. Angew. Chem. 2001, 113, 292–
320; Angew. Chem., Int. Ed. 2001, 40, 284–310; (f) Francis,
M. B.; Jacobsen, E. N. Angew. Chem. 1999, 111, 987–991;
Angew. Chem., Int. Ed. 1999, 38, 937–941; (g) Berkessel, A.;
Riedl, R. J. Comb. Chem. 2000, 2, 215–219; (h) Ding, K.;
Du, H.; Yuan, Y.; Long, J. Chem.—Eur. J. 2004, 10, 2872–
2884.
5.3.2. Larger scale reaction. The solution of 792 mg BSA
in 40 mL formate buffer (30 mM; pH 4.0) was treated with
a 250 mL solution of 13 in H2O (40 mM). This solution
was stirred at room temperature for 30 min. Then 424 mL
solution of dienophile 14a in CH3CN (0.94 M) was added.
The resulting mixture was cooled to 3 ꢂC. The reaction was
started with the addition of 200 mL pure cyclopentadiene
5. (a) Reetz, M. T.; Sell, T.; Meiswinkel, A.; Mehler, G. Angew.
Chem. 2003, 115, 814–817; Angew. Chem., Int. Ed. 2003, 42,
790–793; (b) Reetz, M. T. Chim. Oggi 2003, 21, 5–8; (c)

M. T. Reetz et al. / Tetrahedron 63 (2007) 6404–6414
6413
Reetz, M. T.; Li, X. Tetrahedron 2004, 60, 9709–9714; (d)
Reetz, M. T.; Mehler, G.; Meiswinkel, A.; Sell, T.
Tetrahedron: Asymmetry 2004, 15, 2165–2167; (e) Reetz,
M. T.; Ma, J.-A.; Goddard, R. Angew. Chem. 2005, 117, 416–
419; Angew. Chem., Int. Ed. 2005, 44, 412–415; (f) Reetz,
M. T. Comprehensive Coordination Chemistry II; Ward,
M. D., Ed.; Elsevier: Amsterdam, 2004; Vol. 9, pp 509–548;
(g) Reetz, M. T.; Guo, H. Beilstein J. Org. Chem. 2005, 1, 3;
(h) Reetz, M. T.; Li, X. Angew. Chem. 2005, 117, 3019–
3021; Angew. Chem., Int. Ed. 2005, 44, 2959–2962; (i) Reetz,
M. T.; Li, X. Angew. Chem. 2005, 117, 3022–3024; Angew.
Chem., Int. Ed. 2005, 44, 2962–2964; (j) Reetz, M. T.;
Surowiec, M. Heterocycles 2006, 67, 567–574; (k) Pen˜a, D.;
Minnaard, A. J.; Boogers, J. A. F.; de Vries, A. H. M.; de
Vries, J. G.; Feringa, B. L. Org. Biomol. Chem. 2003, 1,
1087–1089; (l) Hoen, R.; Boogers, J. A. F.; Bernsmann, H.;
Minnaard, A. J.; Meetsma, A.; Tiemersma-Wegman, T. D.;
de Vries, A. H. M.; de Vries, J. G.; Feringa, B. L. Angew.
Chem. 2005, 117, 4281–4284; Angew. Chem., Int. Ed. 2005,
44, 4209–4212.
13. (a) Reetz, M. T.; Wilensek, S.; Zha, D.; Jaeger, K.-E. Angew.
Chem. 2001, 113, 3701–3703; Angew. Chem., Int. Ed. 2001,
40, 3589–3591; (b) Reetz, M. T. Proc. Natl. Acad. Sci. U.S.A.
2004, 101, 5716–5722; (c) Bocola, M.; Otte, N.; Jaeger, K.-
E.; Reetz, M. T.; Thiel, W. ChemBioChem 2004, 5, 214–223;
(d) Reetz, M. T.; Puls, M.; Carballeira, J. D.; Vogel, A.;
Jaeger, K.-E.; Eggert, T.; Thiel, W.; Bocola, M.; Otte, N.
ChemBioChem 2007, 8, 106–112.
14. (a) Reetz, M. T.; Wang, L.-W.; in part Bocola, M. Angew.
Chem. 2006, 118, 1258–1263; Erratum, 2556; Angew. Chem.,
Int. Ed. 2006, 45, 1236–1241; Erratum, 2494; (b) Reetz,
M. T.; Carballeira, J. D.; Vogel, A. Angew. Chem. 2006, 118,
7909–7915; Angew. Chem., Int. Ed. 2006, 45, 7745–7751; (c)
Reetz, M. T.; Carballeira, J. D. Nat. Protoc. 2007, 2, 891–903.
15. (a) Reetz, M. T.; Bocola, M.; Carballeira, J. D.; Zha, D.; Vogel,
A. Angew. Chem. 2005, 117, 4264–4268; Angew. Chem., Int.
Ed. 2005, 44, 4192–4196; (b) Reetz, M. T.; Carballeira, J. D.;
€
Peyralans, J. J.-P.; Hobenreich, H.; Maichele, A.; Vogel, A.
Chem.—Eur. J. 2006, 12, 6031–6038.
16. Reetz, M. T. Comprehensive Review of Directed Evolution of
€
6. Yoon, T. P.; Jacobsen, E. N. Science 2003, 299, 1691–1693.
7. Klibanov, A. M. Nature (London) 2001, 409, 241–246.
8. Industrial Biotransformations; Liese, A., Seelbach, K.,
Wandrey, C., Eds.; Wiley-VCH: Weinheim, 2006.
Enantioselective Enzymes; Gates, B. C., Knozinger, H., Eds.;
Advances in Catalysis; Elsevier: San Diego, CA, 2006; Vol.
49, pp 1–69.
17. (a) Parshall, G. W.; Ittel, S. D. Homogeneous Catalysis: The
Applications and Chemistry of Catalysis by Soluble
Transition Metal Complexes, 2nd ed.; Wiley: New York, NY,
1992; (b) Cornils, B.; Herrmann, W. A. Aqueous-Phase
Organometallic Catalysis, Concepts and Applications, 2nd
ed.; Wiley-VCH: Weinheim, 2004.
18. (a) Reetz, M. T.; Rentzsch, M.; Pletsch, A.; Maywald, M.
Chimia 2002, 56, 721–723; (b) Reetz, M. T. Tetrahedron 2002,
58, 6595–6602; (c) Reetz, M. T. Patent WO 92,18645, 2001;
9. (a) Imanaka, T.; Atomi, H. Enzyme Catalysis in Organic
Synthesis:
Waldmann, H., Eds.; Wiley-VCH: Weinheim, 2002; Vol. 1,
A
Comprehensive Handbook; Drauz, K.,
ꢀ
ꢀ ꢀ
pp 67–95; (b) Cedrone, F.; Menez, A.; Quemeneur, E. Curr.
Opin. Struct. Biol. 2000, 10, 405–410; (c) Harris, J. L.; Craik,
C. S. Curr. Opin. Chem. Biol. 1998, 2, 127–132; (d)
Kazlauskas, R. J. Curr. Opin. Chem. Biol. 2000, 4, 81–88; (e)
Li, Q.-S.; Schwaneberg, U.; Fischer, M.; Schmitt, J.; Pleiss,
J.; Lutz-Wahl, S.; Schmid, R. D. Biochim. Biophys. Acta
2001, 1545, 114–121.
€
(d) Maiwald, P. Dissertation, Ruhr-Universitat Bochum, 2001.
19. (a) Qi, D.; Tann, C.-M.; Haring, D.; Distefano, M. D. Chem.
Rev. 2001, 101, 3081–3111; (b) Polgar, L.; Bender, M. L.
J. Am. Chem. Soc. 1966, 88, 3153–3154; (c) Schultz, P. G.
Science (Washington, D.C.) 1988, 240, 426–433; (d)
Khumtaveeporn, K.; DeSantis, G.; Jones, J. B. Tetrahedron:
Asymmetry 1999, 10, 2563–2572; (e) Smith, H. B.; Hartman,
F. C. J. Biol. Chem. 1988, 263, 4921–4925; (f) Nicholas,
K. M.; Wentworth, P., Jr.; Harwig, C. W.; Wentworth, A. D.;
Shafton, A.; Janda, K. D. Proc. Natl. Acad. Sci. U.S.A. 2002,
99, 2648–2653; (g) Hamachi, I.; Shinkai, S. Eur. J. Org.
Chem. 1999, 539–549; (h) Lu, Y. Curr. Opin. Chem. Biol.
2005, 9, 118–126; (i) Lu, Y.; Valentine, J. S. Curr. Opin.
Struct. Biol. 1997, 7, 495–500; (j) Lu, Y.; Berry, S. M.;
Pfister, T. D. Chem. Rev. 2001, 101, 3047–3080; (k) Barker,
P. D. Curr. Opin. Struct. Biol. 2003, 13, 490–499; See also:
(l) Kaiser, E. T. Angew. Chem. 1988, 100, 945–955; Angew.
Chem., Int. Ed. Engl. 1988, 27, 913–922; (m) Hilvert, D.;
Kaiser, E. T. Biotechnol. Genet. Eng. Rev. 1987, 5, 297–318;
(n) de Vries, J. G.; Lefort, L. Chem.—Eur. J. 2006, 12, 4722–
4734; (o) Kruithof, C. A.; Casado, M. A.; Guillena, G.;
Egmond, M. R.; van der Kerk-van Hoof, A.; Heck, A. J. R.;
Klein Gebbink, R. J. M.; van Koten, G. Chem.—Eur. J. 2005,
11, 6869–6877; (p) Panella, L.; Broos, J.; Jin, J.; Fraaije,
M. W.; Janssen, D. B.; Jeronimus-Stratingh, M.; Feringa,
B. L.; Minnaard, A. J.; de Vries, J. G. Chem. Commun.
(Cambridge) 2005, 5656–5658; (q) Davis, B. G. Curr. Opin.
Biotechnol. 2003, 14, 379–386; (r) Eckermann, A. L.;
Barker, K. D.; Hartings, M. R.; Ratner, M. A.; Meade, T. J.
J. Am. Chem. Soc. 2005, 127, 11880–11881; (s) Ward, T. R.
Chem.—Eur. J. 2005, 11, 3798–3804.
10. (a) Reetz, M. T.; Zonta, A.; Schimossek, K.; Liebeton, K.;
Jaeger, K.-E. Angew. Chem. 1997, 109, 2961–2963; Angew.
Chem., Int. Ed. 1997, 36, 2830–2832; (b) Reetz, M. T.;
€
Zonta, A.; Schimossek, K.; Liebeton, K.; Jager, K.-E. Patent
EP 1,002,100 B1, 2003; (c) Schimossek, K. Dissertation,
€
Ruhr-Universitat Bochum, 1998; (d) Reetz, M. T. Pure Appl.
Chem. 1999, 71, 1503–1509.
11. (a) Directed Enzyme Evolution: Screening and Selection
Methods; Arnold, F. H., Georgiou, G., Eds.; Humana:
Totowa, New Jersey, NJ, 2003; Vol. 230; (b) Directed
Molecular Evolution of Proteins (or How to Improve
Enzymes for Biocatalysis); Brakmann, S., Johnsson, K., Eds.;
Wiley-VCH: Weinheim, 2002; (c) Evolutionary Methods in
Biotechnology (Clever Tricks for Directed Evolution);
Brakmann, S., Schwienhorst, A., Eds.; Wiley-VCH:
Weinheim, Germany, 2004; (d) Taylor, S. V.; Kast, P.;
Hilvert, D. Angew. Chem. 2001, 113, 3408–3436; Angew.
Chem., Int. Ed. 2001, 40, 3310–3335; (e) Powell, K. A.;
ꢀ
Ramer, S. W.; del Cardayre, S. B.; Stemmer, W. P. C.; Tobin,
M. B.; Longchamp, P. F.; Huisman, G. W. Angew. Chem.
2001, 113, 4068–4080; Angew. Chem., Int. Ed. 2001, 40,
3948–3959; (f) Rubin-Pitel, S. B.; Zhao, H. Comb. Chem.
High Throughput Screen. 2006, 9, 247–257; (g) Kaur, J.;
Sharma, R. Crit. Rev. Biotechnol. 2006, 26, 165–199.
12. (a) Reetz, M. T. Angew. Chem. 2002, 114, 1391–1394; Angew.
Chem., Int. Ed. 2002, 41, 1335–1338; (b) Reetz, M. T. Enzyme
Assays—High-throughput Screening, Genetic Selection and
Fingerprinting; Reymond, J.-L., Ed.; Wiley-VCH: Weinheim,
2006; pp 41–76.

6414
M. T. Reetz et al. / Tetrahedron 63 (2007) 6404–6414
20. (a) Wilson, M. E.; Whitesides, G. M. J. Am. Chem. Soc. 1978,
100, 306–307; See also: (b) Lin, C.-C.; Lin, C.-W.; Chan,
A. S. C. Tetrahedron: Asymmetry 1999, 10, 1887–1893.
21. Reetz, M. T.; Peyralans, J. J.-P.; Maichele, A.; Fu, Y.; Maywald,
M. Chem. Commun. (Cambridge) 2006, 4318–4320.
22. (a) Collot, J.; Gradinaru, J.; Humbert, N.; Skander, M.; Zocchi,
A.; Ward, T. R. J. Am. Chem. Soc. 2003, 125, 9030–9031; (b)
Thomas, C. M.; Ward, T. R. Chem. Soc. Rev. 2005, 34,
337–346.
32. Zunszain, P. A.; Ghuman, J.; Komatsu, T.; Tsuchida, E.; Curry,
S. BMC Struct. Biol. 2003, 3, 6.
33. Mahammed, A.; Gross, Z. J. Am. Chem. Soc. 2005, 127, 2883–
2887.
34. Mahammed, A.; Gray, H. B.; Weaver, J. J.; Sorasaenee, K.;
Gross, Z. Bioconjugate Chem. 2004, 15, 738–746.
35. (a) Otto, S.; Engberts, J. B. F. N. J. Am. Chem. Soc. 1999, 121,
6798–6806; (b) Otto, S.; Boccaletti, G.; Engberts, J. B. F. N.
J. Am. Chem. Soc. 1998, 120, 4238–4239.
23. (a) Bayer, E. A.; Ben-Hur, H.; Wilchek, M. Methods Enzymol.
1990, 184, 80–89; (b) Thompson, L. D.; Weber, P. C. Gene
1993, 136, 243–246; (c) Sano, T.; Cantor, C. R. Proc. Natl.
Acad. Sci. U.S.A. 1990, 87, 142–146; (d) Wu, S. C.; Qureshi,
M. H.; Wong, S. L. Protein Expr. Purif. 2002, 24, 348–356;
(e) Wu, S.-C.; Wong, S.-L. Appl. Environ. Microbiol. 2002,
68, 1102–1108.
36. Roelfes, G.; Feringa, B. L. Angew. Chem. 2005, 117, 3294–
3296; Angew. Chem., Int. Ed. 2005, 44, 3230–3232.
37. (a) Barr, K. A.; Hopkins, S. A.; Sreekrishna, K. Pharm. Eng.
1992, 12, 48–52; (b) Ohtani, W.; Nawa, Y.; Takeshima, K.;
Kamuro, H.; Kobayashi, K.; Ohmura, T. Anal. Biochem.
1998, 256, 56–62.
€
38. Taglieber, A. Projected Dissertation, Ruhr-Universitat
24. Gallizia, A.; de Lalla, C.; Nardone, E.; Santambrogio, P.;
Brandazza, A.; Sidoli, A.; Arosio, P. Protein Expr. Purif.
1998, 14, 192–196.
Bochum, 2007.
39. The same applies to the anchoring of organocatalysts2 to
proteins.26c
25. Studier, F. W. Protein Expr. Purif. 2005, 41, 207–234.
€
40. Nuzzo, R. G.; Haynie, S. L.; Wilson, M. E.; Whitesides, G. M.
J. Org. Chem. 1981, 46, 2861–2867.
26. (a) Maywald, M. Dissertation, Ruhr-Universitat Bochum,
€
2005; (b) Pletsch, A. Dissertation, Ruhr-Universitat Bochum,
2004; (c) Rentzsch, M. Dissertation, Ruhr-Universitat
41. The widely used QuikChange protocol (QuikChange site-
directed mutagenesis kit. Instruction Manual, 2003,
Stratagene, La Jolla, CA) is based on a number of previous
publications. These and other developments have been re-
viewed in: (a) Dominy, C. N.; Andrews, D. W. Methods in
Molecular Biology; Casali, N., Preston, A., Eds.; Humana:
Totowa, NJ, 2003; Vol. 235, pp 209–223; (b) Vandeyar,
M. A.; Weiner, M. P.; Hutton, C. J.; Batt, C. A. Gene 1988,
65, 129–133; (c) Kirsch, R. D.; Joly, E. Nucleic Acids Res.
1998, 26, 1848–1850.
€
Bochum, 2004.
27. Weber, P. C.; Ohlendorf, D. H.; Wendoloski, J. J.; Salemme,
F. R. Science 1989, 243, 85–88.
28. (a) Bosley, A. D.; Ostermeier, M. Biomol. Eng. 2005, 22, 57–
61; (b) Patrick, W. M.; Firth, A. E. Biomol. Eng. 2005, 22,
105–112.
29. Kamphuis, I. G.; Kalk, K. H.; Swarte, M. B. A.; Drenth, J.
J. Mol. Biol. 1984, 179, 233–256.
30. Reetz, M. T.; Jiao, N. Angew. Chem. 2006, 118, 2476–2479;
Angew. Chem., Int. Ed. 2006, 45, 2463–2466.
31. Peters, T. All about Albumin: Biochemistry, Genetics and
Medical Applications; Academic: San Diego, CA, 1996.
42. Annis, D. A.; Jacobsen, E. N. J. Am. Chem. Soc. 1999, 121,
4147–4154.
43. Rich, D. H.; Gesellchen, P. D.; Tong, A.; Cheung, A.; Buckner,
C. K. J. Med. Chem. 1975, 18, 1004–1010.