Erratum: Converting an esterase into an epoxide hydrolase

Full text of DOI:10.1002/anie.200806276

Communications
DOI: 10.1002/anie.200806276
Divergent Evolution
Converting an Esterase into an Epoxide Hydrolase**
Helge Jochens, Konstanze Stiba, Christopher Savile, Ryota Fujii, Juin-Guo Yu,
Tatsiana Gerassenkov, Romas J. Kazlauskas,* and Uwe T. Bornscheuer*
Dedicated to Professor Kalle Hult on the occasion of his 65th birthday
Divergent evolution has created superfamilies of enzymes
with the same protein fold, but different catalytic abilities. For
example, the a/b-hydrolase superfamily[1] enzymes all cata-
also introduced enzymatic activity into non-catalytic proteins,
but these experiments all required extensive substitutions.[12]
Herein, we test the hypothesis that a few amino acid
substitutions are sufficient to interconvert the catalytic
abilities of different enzymes within the a/b-hydrolase
family. Our test case is to convert an esterase from Pseudo-
monas fluorescens (PFE) into an epoxide hydrolase. Mech-
anistic considerations suggest that two or three amino acid
substitutions could convert an esterase mechanism into an
epoxide hydrolase mechanism (Scheme 1). Esterases have a
Ser-His-Asp catalytic triad,[13] whereas epoxide hydrolases
use an Asp-His-Asp triad,[14] so one substitution is needed in
the triad. In addition, epoxide hydrolases contain two
tyrosines to protonate the epoxide oxygen during catalysis,
which are two more substitutions. In one case, only one of
these tyrosines was essential to catalysis,[15] so perhaps only
one is needed.
lyze reactions involving
a
nucleophilic attack: ester,[2]
amide,[3] epoxide,[4] and alkyl halide hydrolysis,[5] cyanide
addition to aldehydes forming a carbon–carbon bond[6] as well
as several others. Many of these enzymes, and especially
lipases and esterases, accept a wide range of substrates, each
with high stereoselectivity, making them versatile catalysts for
organic synthesis.[2,7]
The different catalytic abilities require distinct mechanis-
tic steps, but some of these mechanistic steps may be shared.
X-ray crystallography and biochemical studies suggest that
the new mechanistic steps require only a few amino acid
substitutions, but the amino acid sequences of enzymes within
a superfamily differ by hundreds of substitutions and possible
insertions and deletions.
Previous reports that involved changing the catalytic
activity of an enzyme required different approaches. Chang-
ing from hydrolysis of a thioester to hydrolysis of a b-lactam
required insertion, deletion, and substitution of loops as well
as amino acid substitutions.[8] Changing 4-chlorobenzoyl-CoA
dehalogenase to a crotonase activity required eight amino
acid substitutions.[9] However, in a few cases, a single amino
acid substitution could introduce new catalytic activity: from
a racemase to an aldolase, from an esterase to a perhydrolase,
from an epimerase to an o-succinylbenzoate synthase, and
from a decarboxylase to a racemase.[10,11] Researchers have
To identify the positions of these residues within the PFE
we compared the structures and amino acid sequences of six
epoxide hydrolases (PDB entries: 2E3J, 2CJP, 1S8O, 1CQZ,
1EHY, and 1Q07) with three esterases (PDB entries: 1VA4,
1P0, and 1ZOI) using clustalw.[17] This comparison (Figure 1)
identified the position of the catalytic nucleophile (D94), four
possible positions for the two mechanistically important
tyrosines (Y125, Y139, Y143, Y195) and three further
amino acids that are conserved in epoxide hydrolases but
are missing from PFE (P29, H93, K188). The position of one
of these tyrosines was identified as being at amino acid 195,
but as it was not definitely clear where to introduce the
second tyrosine, it was placed at all three alternative positions.
Consequently, the following mutants were created by Quik-
Change site-directed mutagenesis and expressed recombi-
nantly in E. coli: M1: S94D; M2: S94D, F125Y, V195Y; M3:
S94D, F143Y, V195Y; M4: S94D, F125Y, F143Y, V195Y; M5:
L29P, S94D, F125Y, K188M; M6: L29P, S94D, F125Y, V195Y;
M7: L29P, F93H, S94D, F125Y, V139Y, V195Y. As expected,
all mutants showed no esterase activity against p-nitrophenyl
acetate above background levels (< 10 mUmgÀ1), as the
catalytic nucleophile was replaced by an aspartate. Unfortu-
nately, all these mutants also showed no detectable epoxide
hydrolase activity towards p-nitrostyrene oxide (see Support-
ing Information).
[*] Dr. C. Savile, Dr. R. Fujii, J.-G. Yu, T. Gerassenkov,
Prof. R. J. Kazlauskas
University of Minnesota, Department of Biochemistry, Molecular
Biology & Biophysics
1479 Gortner Avenue, 174 Gortner Lab, St. Paul, MN 55108 (USA)
Fax: (+1)612-625-5780
E-mail: rjk@umn.edu
Dipl.-Biochem. H. Jochens, Dr. K. Stiba, Prof. Dr. U. T. Bornscheuer
Institute of Biochemistry, Dept. of Biotechnology & Enzyme
Catalysis, University of Greifswald
Felix-Hausdorff-Strasse 4, 17487 Greifswald (Germany)
Fax: (+49)3834-86-80066
E-mail: uwe.bornscheuer@uni-greifswald.de
Homepage: http://www.chemie.uni-greifswald.de/~biotech
[**] We thank Prof. Dick Janssen for providing the gene of Agrobacterium
radiobacter epoxide hydrolase and Dr. Santosh Padhi for preparation
of several site-directed mutants. U.T.B. thanks the German Research
Foundation (DFG, grant Bo1862/4-1), and R.J.K. the US National
Science Foundation (CHE-0616560) for financial support.
Because it was not obvious why the created mutants failed
to catalyze the reaction, we applied directed evolution to
search for missing key amino acids that correct the imperfect
geometry. Starting from mutants M4–M6, we created a
mutant library by a error-prone polymerase chain reaction
(epPCR) and selected active variants using a growth assay.[18]
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200806276.
3532
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 3532 –3535

Angewandte
Chemie
Scheme 1. Mechanistically essential amino acid residues in Agrobacterium radiobacter epoxide hydrolase (EchA).[15] A) Formation and liberation of
the alkyl enzyme intermediate derived from styrene oxide as substrate and Pseudomonas fluorescens esterase (PFE).[16] B) Formation and liberation
of the acetyl enzyme intermediate derived from phenyl acetate as substrate.
To test the role of the loop,
the whole loop in the PFE
mutant M7 (containing muta-
tions L29P, F93H, S94D, F125Y,
V139Y, V195Y) was replaced
by the corresponding element
of the EchA by a PCR-based
method based (EchA number-
ing: P132 to Y152). The result-
ing chimera (M8) could be
expressed in E. coli as a soluble
enzyme by coexpression of
chaperones (see the Supporting
Figure 1. Part of the comparison of the structure and amino acid sequences of three esterases
(PFE=1VA4) with six different epoxide hydrolases (EH; for PDB codes, see text). The catalytic aspartate
(D) and one of the mechanistically important tyrosines (Y) of the epoxide hydrolases and the catalytic
serine (S) of the esterases are highlighted.
Information).
Protein
(ca.
Although we identified eight clones out of 25000 that grew
under these conditions, in the subsequent HPLC validation,
no activity against p-nitrostyrene oxide was detected.
A further structural comparison of the six epoxide
hydrolases showed that a loop with 20 amino acids at the
supposed entrance to the active site[16] differs in PFE
(Figure 2). In the epoxide hydrolases, this loop lies to one
side of the active site, but in PFE, this loop blocks the
entrance. The loop may also position the mechanistically
important Y152 (position 139 in PFE) as it is the last residue
of this loop.
10 mg) was isolated after His-tag purification and desalting.
This chimera catalyzed the slow hydrolysis of p-nitrostyrene
oxide (Figure 3) with an initial activity of 9 mUmgÀ1 and a
turnover number of 0.01 sÀ1 at a 50 mm substrate concentra-
tion. The catalytic activity differed from batch to batch and
some samples showed no activity. In spite of extensive
experimentation (see the Supporting Information), we have
not found a satisfactory explanation. We hypothesize that the
mutations hinder protein folding, and subtle changes in
experimental conditions change the amount of properly
folded enzyme. We could not measure KM and Vmax values
Angew. Chem. Int. Ed. 2009, 48, 3532 –3535
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3533

Communications
The experimental results indicate that the epoxide hydro-
lase loop is essential for the newly created activity, but at the
molecular level, the reason is uncertain. One possibility is that
the esterase loop favors a nonproductive substrate orienta-
tion. The esterase substrate, p-nitrophenyl acetate, and
epoxide hydrolase substrate, p-nitrostyrene oxide, most
likely orient themselves slightly differently in the active site.
A correct p-nitrophenyl group orientation for the esterase
substrate orientation may be nonproductive for the epoxide
hydrolase substrate. The observed substrate inhibition by p-
nitrostyrene oxide supports the notion that nonproductive
orientations of the substrate at the active site are possible. A
second possible role for the epoxide hydrolase loop is some
required motion. The X-ray structure of EchA gave a non-
interpretable electron density for amino acids G138 to H148,
suggesting that this region is flexible.[19] The third possibility is
that the epoxide hydrolase loop corrects the orientation of
Y139. Mutant M7 contained the Y139 mutation, but did not
show epoxide hydrolase activity. The new loop most likely
changes the orientation of this amino acid.
Figure 2. Multiple structural alignment of six epoxide hydrolases
(green) with PFE (light orange) showing that a loop (A120–V139)
present in PFE (dark orange) possibly blocks the entrance to the active
site (red sticks).
Changing just the obvious residues (serine to aspartate,
introduction of two tyrosines) was not sufficient, but an
additional substitution of a loop conferred epoxide hydrolase
activity into the esterase scaffold. The specific activity we
measured at substrate concentrations near the detection limit
was only 800-fold less than the activity of the template, which
is a true epoxide hydrolase. This catalytic activity is too low
for practical use, but demonstrates that the principle of
interconversion of enzyme activities within the a/b-hydrolase
family is possible, and thus opens opportunities to extend
catalysis to new, non-natural reactions.
Figure 3. Hydrolysis of p-nitrostyrene oxide catalyzed by loop mutant
M8. The protein concentration was 0.5 mgmLÀ (16.7 mm), so complete
conversion corresponds to approximately three substrate molecules
hydrolyzed per enzyme molecule.
Received: December 23, 2008
Published online: April 6, 2009
owing to substrate inhibition (see the Supporting Informa-
tion). This inhibition was significant even at concentrations of
50 mm, and lower substrate concentrations were below the
detection limit. The loop mutant showed high enantioselec-
tivity (E > 100) for the (R)-enantiomer of p-nitrostyrene
oxide (Figure 4). The template epoxide hydrolase, EchA, also
favors the (R)-enantiomer. We also detected low background
activity in crude E. coli cell extract (0.06 mUmgÀ1), but this
reaction favored the (S)-enantiomer. The His-tag purification
used to prepare the loop mutant removed this contaminating
activity.
Keywords: directed evolution · enantioselectivity ·
.
enzyme catalysis · epoxide hydrolases · esterases
[1] D. L. Ollis, E. Cheah, M. Cygler, B. Dijkstra, Protein Eng. 1992,
5, 197.
[2] a) R. J. Kazlauskas, U. T. Bornscheuer in Biotechnology, Vol. 8a,
2nd ed (Eds.: H. J. Rehm, G. Reed, A. Pꢀhler, P. J. W. Stadler,
D. R. Kelly), Wiley-VCH, Weinheim, 1998, p. 37; b) M. Cygler,
J. D. Schrag, J. L. Sussman, M. Harel, I. Silman, M. K. Gentry,
B. P. Doctor, Protein Sci. 1993, 2, 366.
[3] P. Carter, J. A. Wells, Nature 1988, 332, 564.
[4] a) M. Arand, A. Cronin, M. Adamska, F. Oesch, Methods
Enzymol. 2005, 400, 569; b) A. Archelas, R. Furstoss, Trends
Biotechnol. 1998, 16, 108; c) K. Faber, M. Mischitz, W. Kroutil,
Acta Chem. Scand. 1996, 50, 249.
[5] a) S. Barth, M. Fischer, R. D. Schmid, J. Pleiss, Bioinformatics
2004, 20, 2845; b) D. B. Janssen, Curr. Opin. Chem. Biol. 2004, 8,
150; c) S. Fetzner, F. Lingens, Microbiol. Rev. 1994, 58, 641;
d) P. E. Swanson, Curr. Opin. Biotechnol. 1999, 10, 365.
[6] a) D. V. Johnson, A. A. Zabelinskaja-Mackova, H. Griengl,
Curr. Opin. Chem. Biol. 2000, 4, 103; b) F. Effenberger, S.
Fꢁrster, H. Wajant, Curr. Opin. Biotechnol. 2000, 11, 532; c) M.
Schmidt, H. Griengl, in Biocatalysis—from discovery to appli-
cation, Vol. 200 (Ed.: W.-D. Fessner), Springer, Berlin, 1999,
pp. 193.
Figure 4. HPLC chromatogram for the conversion of p-nitrostyrene
oxide by loop mutant M8. 50 mL of a prepared sample (c=50 mm) was
measured on a HPLC using Chiracel OD-H column (for sample
preparation, see the Supporting Information). (R,S)-p-Nitrostyrene
oxide (tR =10.577 min), (R)-p-nitrophenylethanediol (tR =21.590 min),
(S)-p-nitrophenylethanediol (tR =27.797 min).
3534
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 3532 –3535

Angewandte
Chemie
[7] a) U. T. Bornscheuer, R. J. Kazlauskas, Hydrolases in Organic
Synthesis—Regio- and Stereoselective Biotransformations, 2nd
ed., Wiley-VCH, Weinheim, 2006; b) R. D. Schmid, R. Verger,
Angew. Chem. 1998, 110, 1694; Angew. Chem. Int. Ed. 1998, 37,
1608.
[8] H. S. Park, S. H. Nam, J. K. Lee, C. N. Yoon, B. Mannervik, S. J.
Benkovic, H. S. Kim, Science 2006, 311, 535.
[9] H. Xiang, L. Luo, K. L. Taylor, D. Dunaway-Mariano, Biochem-
istry 1999, 38, 7638.
[13] V. K. Ward, B. C. Bonning, T. Huang, T. Shiotsuki, V. N.
Griffeth, B. D. Hammock, Int. J. Biochem. 1992, 12, 1933.
[14] R. Rink, M. Fennema, M. Smids, U. Dehmel, D. B. Janssen, J.
Biol. Chem. 1997, 272, 14650.
[15] a) B. van Loo, J. H. L. Spelberg, J. Kingma, T. Sonke, M. G.
Wubbolts, D. B. Janssen, Chem. Biol. 2004, 11, 981; b) R. Rink,
J. H. L. Spelberg. R. J. Pieters, J. Kingma, M. Nardini, R. M.
Kellogg, B. W. Dijkstra, D. B. Janssen, J. Am. Chem. Soc. 1999,
121, 7417.
[10] a) F. P. Seebeck, D. Hilvert, J. Am. Chem. Soc. 2003, 125, 10158;
b) D. M. Z. Schmidt, E. C. Mundorff, M. Dojka, E. Bermudez,
J. E. Ness, S. Govindarajan, P. C. Babbitt, J. Minshull, J. A. Gerlt,
Biochemistry 2003, 42, 8387.
[11] a) P. Bernhardt, K. Hult, R. J. Kazlauskas, Angew. Chem. 2005,
117, 2802; Angew. Chem. Int. Ed. 2005, 44, 2742; b) Y. Terao, K.
Miyamoto, H. Ohta, Chem. Commun. 2006, 3600.
[12] a) D. Rꢁthlisberger, O. Khersonsky, A. M. Wollacott, L. Jiang, J.
DeChancie, J. Betker, J. L. Gallaher, E. A. Althoff, A. Zanghel-
lini, O. Dym, S. Albeck, K. N. Houk, D. S. Tawfik, D. Baker,
Nature 2008, 453, 190; b) B. Seelig, J. W. Szostak, Nature 2007,
448, 828; c) L. Jiang, E. A. Althoff, F. R. Clemente, L. Doyle, D.
Rꢁthlisberger, A. Zanghellini, J. L. Gallaher, J. L. Betker, F.
Tanaka, C. F. Barbas III, D. Hilvert, K. N. Houk, B. L. Stoddard,
D. Baker, Science 2008, 319, 1387.
[16] a) J. D. Cheeseman, A. Tocilj, S. Park, J. D. Schrag, R. J.
Kazlauskas, Acta Crystallogr. Sect. D 2004, 60, 1237; b) N.
Krebsfꢂnger, F. Zocher, J. Altenbuchner, U. T. Bornscheuer,
Enzyme Microb. Technol. 1998, 22, 641; c) G. P. Horsman,
A. M. F. Liu, E. Henke, U. T. Bornscheuer, R. J. Kazlauskas,
Chem. Eur. J. 2003, 9, 1933.
[17] npsa-pbil.ibcp.fr/cgi-bin/npsa_automat.pl?page = npsa_clustalw.
html.
[18] Mutant libraries were copied on agar plates that contained
glycidol butyrate. Active clones should then generate glycerol
butyrate, which serves as a carbon source for E. coli on the
minimal media, thus allowing selection for active epoxide
hydrolase variants.
[19] M. Nardini, I. S. Ridder, H. J. Rozeboom, K. H. Kalk, R. Rink,
D. B. Janssen, B. Dijkstra, J. Biol. Chem. 1999, 274, 14579.
Angew. Chem. Int. Ed. 2009, 48, 3532 –3535
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3535