A genetic selection system for evolving enantioselectivity of enzymes

Article abstract of DOI:10.1039/b814538e

As an alternative to screening in the directed evolution of enantioselective enzymes, a selection system has been implemented for a lipase-catalyzed hydrolytic kinetic resolution of a chiral ester. The Royal Society of Chemistry.

Full text of DOI:10.1039/b814538e

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
A genetic selection system for evolving enantioselectivity of enzymesw
¨ ´
Manfred T. Reetz,* Horst Hobenreich, Pankaj Soni and Layla Fernandez
Received (in Cambridge, UK) 20th August 2008, Accepted 3rd September 2008
First published as an Advance Article on the web 22nd September 2008
DOI: 10.1039/b814538e
As an alternative to screening in the directed evolution of
enantioselective enzymes, a selection system has been implemen-
ted for a lipase-catalyzed hydrolytic kinetic resolution of a chiral
ester.
pound as a poison for the organism. In such a system isosteric
pseudo-enantiomers10 are required, and the respective com-
pounds have to be used in a mixture. Depending upon the
specific system used, the choice of the ratio of the two starting
substrates employed in the mixtures constitutes a convenient
and efficient means for optimizing the selection pressure.
Directed evolution1 of enantioselective enzymes2 as catalysts
in synthetic organic chemistry has emerged as an alternative to
the traditional methods in asymmetric catalysis.3 It is based on
the combination of gene mutagenesis, expression and high-
throughput ee-screening of libraries of mutant enzymes. An
alternative to screening is selection,1d,4,5 in which cell survival
needs to be linked to the enantioselectivity of an enzyme-
catalyzed reaction. Ideally, only those colonies which harbor
variants displaying enhanced enantioselectivity appear on the
agar plates. The vast number of ‘‘junk’’ mutants that usually
need to be assayed would never be formed. Some selection
systems have been devised for evolving enzyme activity,1d,4,5
but the development of systems capable of selecting for
enantioselectivity is a more difficult problem.
Scheme 1 Genetic selection system for laboratory evolution of
enantioselectivity in a kinetic resolution.
In order to test our concept, we chose the hydrolytic kinetic
resolution of appropriately designed esters of chiral isopropyl-
idene glycerol 2 (IPG) as the model reaction, and the lipase
from Candida antarctica B (CALB)11 as the enzyme (Scheme 2).
In the hydrolytic kinetic resolution of the acetate rac-1, the
WT CALB displays a selectivity factor E of 1.9 in favor of the
reaction of (R)-1 with slightly preferential formation of (S)-29
(note that the assignment of the absolute configuration has
changed due to a switch in priority within the CIP nomencla-
ture). Since the reactive function of the substrate is not directly
bonded to the stereogenic center, obtaining acceptable enan-
tioselectivities in the hydrolytic kinetic resolution of this
substrate is difficult.12 We synthesized the acetate (S)-1 and
the fluoroacetate (R)-4 separately as pseudo-enantiomers. The
acid parts of the esters are nearly isosteric, yet hydrolytic
cleavage can be expected to lead to two very different scenar-
ios, namely the generation of acetic acid as a carbon source
and fluoroacetic acid as a poison for the organism, respectively
(Scheme 2). Thus, when using a mixture of the two compounds
under selection pressure, the system can be expected to provide
Previously, we devised a screening-system for assaying the
kinetic resolution of chiral esters catalyzed by an esterase,
which is based on differential growth properties,6 but it proved
to be less efficient than other screens utilizing spectroscopic
assays.2,4,7 Recently, a selection system for enantioselective
lipase variants of Bacillus subtilis was reported by Quax and
co-workers.8 They used a mutant library in an aspartate
auxotroph Escherichia coli which was supplemented with an
aspartate ester of the chiral alcohol isopropylidene glycerol
(IPG) in the (S)-form, assisted by an inhibitory compound,
namely a covalently binding phosphonate ester incorporating
the enantiomeric (R)-alcohol.
Here we present an alternative approach based on pseudo-
enantiomeric mixtures which does not require such surrogate
substrates nor enzyme inhibitors.9 The basic idea is to mimic
kinetic resolution in such a way that ‘‘positive’’ and ‘‘negative’’
components are placed in a single system according to the
absolute configuration of the chiral compound of interest,
thereby ensuring simultaneous selection for activity and
enantioselectivity (Scheme 1). For example, in order to evolve
(R)-selectivity, the (R)-substrate needs to contain a positive
component which serves as a potential energy source for the
host organism, thereby promoting growth following the de-
sired enantioselective cleavage reaction. At the same time the
(S)-substrate is designed so as to contain a negative compo-
nent, the respective cleavage reaction generating a toxic com-
Max-Planck-Institut fur Kohlenforschung, Kaiser-Wilhelm-Platz 1,
¨
45470 Mulheim, Germany. E-mail: reetz@mpi-muelheim.mpg.de
¨
w Electronic supplementary information (ESI) available: Experimental
details. See DOI: 10.1039/b814538e
Scheme 2 Model system for genetic selection based on a mixture
comprising an enantiomer (S)-1 which provides an energy source (3)
for the host organism and a pseudo-enantiomer (R)-4 which generates
a poison (5).
ꢀc
This journal is The Royal Society of Chemistry 2008
5502 | Chem. Commun., 2008, 5502–5504

View Article Online
primarily mutants of CALB which are selective for the (S)-
substrate in the model reaction. In the present case this would
mean reversal of the sense of enantioselectivity, which we
chose to pursue.
has been achieved, this being with a difficult substrate. The
activities of all mutants are in the same range as the wild type,
as similar conversions were obtained in test experiments with
the same amount of supernatant. We then compared these
observations with the results of screening the ‘‘normal’’ en-
zyme library not under selection pressure, GC analysis being
used as the screening assay. This representative sampling was
done for statistical reasons. Complete sampling was not
necessary for this purpose. Of 192 active mutants analyzed
for enantioselectivity, only 25 were found to be (S)-selective,
among them six (3.1%) with an E-value higher than 4 (highest
observed E-value: 10). This comparison likewise shows that
the selection pressure strongly favors the appearance of (S)-
selective mutants.
Exploratory experiments indicated that the yeast Pichia
pastoris, which allows for acceptable expression of CALB,11c
constitutes an adequate organismic host for the model study.
The desired effect of growth inhibition induced by fluoro-
acetate with acetate as the carbon source was shown to be
operating at various pH values. At pH 6, growth is slightly
retarded, but we expected this to be insufficient in selection
experiments. Fortunately, complete inhibition was achieved at
lower pH, and after optimization pH 4.6 was chosen for all
subsequent selection studies. Due to problems with catabolite
repression and background growth in the case of the methanol-
inducible pPICZa system, the constitutive pGAPZa was
chosen as expression vector (extracellular).
Table 1 Sequence analysis of the eight mutants obtained from the
selection system and their subsequent use as catalysts in the hydrolytic
kinetic resolution of rac-1
As a mutagenesis method we chose the Combinatorial
Active-Site Saturation Test (CAST), which had previously
been developed for controlling substrate scope13 and/or en-
hancing the enantioselectivity14,15 of enzymes. It involves
systematic saturation mutagenesis16 at relevant sites around
the complete binding pocket with formation of focused li-
braries,17 a given site being composed of one or more amino
acid positions. In the case of CALB, several CAST sites
around the binding pocket appeared logical as judged by the
published X-ray structure,11b but we restricted the present
study to the site comprising amino acid positions 278 and
281, given the relatively low transformation efficiency of P.
pastoris. These were subjected to simultaneous randomization
using NNK and NDT codon degeneracy, respectively (N:
adenine/cytosine/guanine/thymine; K: guanine/thymine; D:
adenine/guanine/thymine; T: thymine). These encode all 20
proteinogenic amino acids or only 12 (Phe, Leu, Ile, Val, Tyr,
His, Asn, Asp, Cys, Arg, Ser, Gly), respectively. Following
optimization experiments in liquid culture and on solid plates,
selection plates with 0.3% (17 mM) of the acetate (S)-1 and
0.003% (0.16 mM) of the fluoroacetate (R)-4 were found to be
best. The degree of selective pressure was adjusted in such a
way that the parental variant was eliminated and sufficiently
enantioselective mutants were able to grow. For the second
generation of directed evolution, the hurdle would be set
higher by simply increasing the concentration of (R)-4. Under
constant conditions a CAST library was spread out on an
agarose plate followed by incubation, which led to the appear-
ance of about 70–80 colonies on the plate. At this point we
decided not to harvest all of them, which means missing some,
but restricted further investigation to ten of the larger ones.
The respective enzyme variants were sequenced and evaluated
for enantioselectivity in the kinetic resolution of rac-1.
Transformant
(enzyme variant)
Enantio-
selectivity E-value
Sequence
Sel1
Sel2
Sel3
Sel4
Sel5
Sel6
Sel7
Sel8
(S)-1
(S)-1
(S)-1
(S)-1
(S)-1
(S)-1
(S)-1
(S)-1
6
8
3
6
8
3
6
3
Leu278Trp/Ala281Asp
Leu278Asp/Ala281Leu
Leu278Asn/Ala281Asn
Leu278Trp/Ala281Asp
Leu278Asp/Ala281Leu
Leu278Ile/Ala281His
Leu278Ser/Ala281Leu
Leu278Ile/Ala281His
Finally, in order to provide additional support to our
conclusions, we tested the growth behavior in liquid cultures
containing either the WT or the mutants obtained indepen-
dently from the screening process. Accordingly, the mutants
were incubated with a mixture comprising 0.3% (S)-1 (17 mM)
and 0.004% (R)-4 (0.21 mM) under otherwise identical con-
ditions, growth behavior being monitored by OD-measure-
ments (Fig. 1). It can be seen that the (S)-selective variants E2-
G2 (E = 8; Leu278Pro/Ala281Leu) and E2-A12 (E = 10;
Leu278Ala/Ala281Leu) maintain the growth rate of the host
P. pastoris in liquid culture fairly well, whereas the (R)-
selective WT leads to complete growth inhibition. In the case
The results of these experiments are noteworthy in several
respects (Table 1). Eight of the ten mutants proved to be (S)-
selective, only one favoring slightly (R)-1, and one being
inactive. Thus, the expectations are borne out very well. The
selection system provides primarily the desired (S)-selective
variants, the percentage of false positives being gratifyingly
low (20%). The enantioselectivity, as measured by the selec-
tivity factor E, ranges between 3 and 8, which are respectable
values considering the fact that reversal of enantioselectivity
Fig. 1 Growth behavior of CALB variants with P. pastoris as the
host organism in liquid cultures containing 0.3% (S)-1 and 0.004%
(R)-4.
ꢀc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 5502–5504 | 5503

View Article Online
of the moderately (S)-selective variant E1-D6 (E = 4; Leu278-
His/Ala281Cys), greatly reduced but not completely inhibited
growth is observed. Upon performing the experiment with (S)-
1 in the absence of (R)-4, no inhibition was observed. Instead,
growth correlated only with the activity of the corresponding
mutant regarding the (S)-substrate. Thus, these experiments
likewise demonstrate that the concept of applying simulta-
neously positive and negative genetic selection pressure is
sufficient and necessary for differentiating enzyme variants
according to their respective degrees of enantioselectivity.
Due to the extracellular enzyme expression of the present
system, we cannot expect an enrichment of (S)-mutants in a
mixture of variants in liquid culture. The situation in an
intracellular system might be different.
(j) T. Matsuura and T. Yomo, J. Biosci. Bioeng., 2006, 101,
449–456; (k) S. Bershtein and D. S. Tawfik, Curr. Opin. Chem.
Biol., 2008, 12, 151–158.
2 Reviews of directed evolution of enantioselective enzymes: (a) M.
T. Reetz, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 5716–5722; (b)
M. T. Reetz, Directed evolution of enantioselective enzymes as
catalysts for organic synthesis, in Advances in Catalysis, ed. B. C.
Gates and H. Knozinger, Elsevier, San Diego, 2006, vol. 49,
¨
pp. 1–69; (c) M. T. Reetz, Directed evolution as a means to engineer
enantioselective enzymes, in Asymmetric Organic Synthesis with
´
Enzymes, ed. V. Gotor, I. Alfonso and E. Garcıa-Urdiales,
Wiley-VCH, Weinheim, 2008, pp. 21–63.
3 (a) E. N. Jacobsen, in Comprehensive Asymmetric Catalysis, ed. E.
N. Jacobsen, A. Pfaltz and H. Yamamoto, Springer, Berlin, 1999,
vol. I–III; (b) R. Noyori, Angew. Chem., Int. Ed., 2002, 41,
2008–2022; (c) K. B. Sharpless, Angew. Chem., Int. Ed., 2002, 41,
2024–2032; (d) A. Berkessel, in Asymmetric Organocatalysis, ed. A.
Berkessel and H. Groger, Wiley-VCH, Weinheim, 2004; (e) M. T.
¨
In conclusion, we have provided proof-of-principle of a
novel selection system for potential application in the directed
evolution of the enantioselectivity of enzymes. The idea is to
utilize a mixture composed of one enantiomer as a potential
energy source for the host organism and the opposite pseudo-
enantiomer as a potential poison leading to growth inhibition,
these being released by the enantioselective cleavage reaction.
By mimicking kinetic resolution with non-surrogate substrates,
the acetate–fluoroacetate-system allows for the bias-minimal
selection for activity and enantioselectivity simultaneously.
Exploiting the properties of acetic acid and fluoroacetic
acid is not the only possibility for an appropriate energy
source–poison couple. Extension to the desymmetrization of
meso-type substrates in which the two enantiotopic groups are
appropriately labeled is, in principle, possible. We conclude
that the present approach is relevant in any enzyme-catalyzed
kinetic resolution or desymmetrization reaction, provided the
two chiral isosteric moieties functioning as a potential energy
source or poison, respectively, can be designed and incorpo-
rated in the system. The next step in applying the underlying
principle described herein is the establishment of a corre-
sponding E. coli system which would allow very large libraries
to be evaluated in order to obtain practical catalysts for use in
organic chemistry.
Reetz, Angew. Chem., Int. Ed., 2008, 47, 2556–2588.
4 (a) J.-L. Reymond, Enzyme Assays – High-throughput Screening,
Genetic Selection and Fingerprinting, Wiley-VCH, Weinheim, 2006;
(b) H. Lin and V. W. Cornish, Angew. Chem., Int. Ed., 2002, 41,
4402–4425.
5 (a) H. Leemhuis, V. Stein, A. D. Griffiths and F. Hollfelder, Curr.
Opin. Struct. Biol., 2005, 15, 472–478; (b) A. Aharoni, G. Amitai,
K. Bernath, S. Magdassi and D. S. Tawfik, Chem. Biol., 2005, 12,
1281–1289; (c) Y. L. Boersma, M. J. Droge and W. J. Quax, FEBS
¨
J., 2007, 274, 2181–2195; (d) C. H. Collins, J. R. Leadbetter and F.
H. Arnold, Nat. Biotechnol., 2006, 24, 708–712; (e) M. T. Reetz
and L.-W. Wang, Comb. Chem. High Throughput Screening, 2006,
9, 295–299; (f) U. T. Bornscheuer, J. Altenbuchner and H. H.
Meyer, Bioorg. Med. Chem., 1999, 7, 2169–2173.
6 M. T. Reetz and C. J. Ruggeberg, Chem. Commun., 2002,
1428–1429.
¨
7 (a) M. T. Reetz, A. Zonta, K. Schimossek, K. Liebeton and K.-E.
Jaeger, Angew. Chem., Int. Ed. Engl., 1997, 36, 2830–2832; (b) R. J.
Kazlauskas, in Enzyme Assays – High-throughput Screening, Ge-
netic Selection and Fingerprinting, ed. J.-L. Reymond, Wiley-VCH,
Weinheim, 2006, pp. 17–39; (c) M. T. Reetz, in Enzyme Assays –
High-throughput Screening, Genetic Selection and Fingerprinting,
ed. J.-L. Reymond, Wiley-VCH, Weinheim, 2006, pp. 41–76.
8 Y. L. Boersma, M. J. Droge, A. M. van der Sloot, T. Pijning, R. H.
¨
Cool, B. W. Dijkstra and W. J. Quax, ChemBioChem, 2008, 9,
1110–1115.
9 H. Hobenreich, Dissertation, Ruhr-Universitat Bochum, Bochum,
¨
¨
2008.
10 (a) A. Horeau and A. Nouaille, Tetrahedron Lett., 1990, 31,
2707–2710; (b) M. T. Reetz, M. H. Becker, H.-W. Klein and D.
Stockigt, Angew. Chem., Int. Ed., 1999, 38, 1758–1761.
¨
Generous support by the German-Israeli Project Coopera-
tion (DIP) and the Fonds der Chemischen Industrie is grate-
fully acknowledged. We also thank A. Vogel for helpful
11 Reviews of CALB and other lipases:12c (a) R. D. Schmid and R.
Verger, Angew. Chem., Int. Ed., 1998, 37, 1608–1633; (b) J.
Uppenberg, M. T. Hansen, S. Patkar and T. A. Jones, Structure
(Cambridge, MA, U. S.), 1994, 293–308; (c) J. C. Rotticci-Mulder,
M. Gustavsson, M. Holmquist, K. Hult and M. Martinelle, Protein
Expression Purif., 2001, 21, 386–392.
discussions and D. Klutt for assistance in the synthesis of
¨
substrates.
12 (a) F. Molinari, K. S. Cavenago, A. Romano, D. Romano and R.
Gandolfi, Tetrahedron: Asymmetry, 2004, 15, 1945–1947; (b) A.
Mezzetti, C. Keith and R. J. Kazlauskas, Tetrahedron: Asymmetry,
2003, 14, 3917–3924; (c) U. T. Bornscheuer and R. J. Kazlauskas,
Hydrolases in Organic Synthesis, Wiley-VCH, Weinheim, 2005.
13 M. T. Reetz, M. Bocola, J. D. Carballeira, D. Zha and A. Vogel,
Angew. Chem., Int. Ed., 2005, 44, 4192–4196.
14 (a) M. T. Reetz, L.-W. Wang and M. Bocola, Angew. Chem., Int.
Ed., 2006, 45, 1236–1241; Erratum 2494; (b) C. M. Clouthier,
M. M. Kayser and M. T. Reetz, J. Org. Chem., 2006, 71,
8431–8437; (c) M. T. Reetz, J. J.-P. Peyralans, A. Maichele, Y.
Fu and M. Maywald, Chem. Commun., 2006, 4318–4320.
15 (a) S. Bartsch, R. Kourist and U. T. Bornscheuer, Angew. Chem.,
Int. Ed., 2008, 47, 1508–1511; (b) L. Liang, J. Zhang and Z. Lin,
Microb. Cell Fact., 2007, 6, 36.
Notes and references
1 General reviews of directed evolution: (a) K. M. Arndt and K. M.
Muller, Protein Engineering Protocols (Methods in Molecular
¨
Biology), Humana Press, Totowa, New Jersey, 2007, vol. 352; (b)
F. H. Arnold and G. Georgiou, Directed Enzyme Evolution:
Screening and Selection Methods, Humana Press, Totowa, New
Jersey, 2003, vol. 230; (c) N. J. Turner, Trends Biotechnol., 2003,
21, 474–478; (d) S. V. Taylor, P. Kast and D. Hilvert, Angew.
Chem., Int. Ed., 2001, 40, 3310–3335; (e) S. Brakmann and
A. Schwienhorst, Evolutionary Methods in Biotechnology – Clever
Tricks for Directed Evolution, Wiley-VCH, Weinheim, 2004; (f) S.
Lutz and W. M. Patrick, Curr. Opin. Biotechnol., 2004, 15,
291–297; (g) E. G. Hibbert, F. Baganz, H. C. Hailes, J. M. Ward,
G. J. Lye, J. M. Woodley and P. A. Dalby, Biomol. Eng., 2005
22, 11–19; (h) S. B. Rubin-Pitel and H. Zhao, Comb. Chem.
High Throughput Screening, 2006, 9, 247–257; (i) J. Kaur and
R. Sharma, Crit. Rev. Biotechnol., 2006, 26, 165–199;
16 Details of QuikChanget protocol of Stratagene: H. H. Hogrefe, J.
Cline, G. L. Youngblood and R. M. Allen, BioTechniques, 2002,
33, 1158–1165.
17 For examples of focused libraries,1 see list in Electronic Supple-
mentary Information.
ꢀc
This journal is The Royal Society of Chemistry 2008
5504 | Chem. Commun., 2008, 5502–5504