Novel hybrid esterase-haloacid dehalogenase enzyme

Article abstract of DOI:10.1002/cbic.201000258

(Figure Presented) Promiscuous proteins: The identification and characterization of a wild-type promiscuous hydrolase that efficiently catalyzes the hydrolysis of ester bonds and haloacids in a single protein has been reported. This study has demonstrated the utility of metagenomics to access novel catalytic activities.

Full text of DOI:10.1002/cbic.201000258

DOI: 10.1002/cbic.201000258
Novel Hybrid Esterase-Haloacid Dehalogenase Enzyme
Ana Beloqui,[a] Julio Polaina,[b] Josꢀ Marꢁa Vieites,[a] Dolores Reyes-Duarte,[c] Rodrigo Torres,[d]
Olga V. Golyshina,[e, f] Tatyana N. Chernikova,[f] Agnes Waliczek,[f] Amir Aharoni,[g] Michail M. Yakimov,[h]
Kenneth N. Timmis,[f] Peter N. Golyshin,[e, f] and Manuel Ferrer*[a]
The existence of different catalytic mechanisms (or reaction
types) in the same active site is an example of catalytic promis-
cuity.[1,2] The promiscuity can result from natural evolution of
an enzyme, which enhances organism metabolic flexibility and
environmental fitness, or from laboratory evolution and can be
exploited in numerous synthetic applications.[2] a/b-Hydrolase-
fold proteins belong to one of the largest protein superfamilies
within the a/b class of folds and exhibit enormous sequence
diversity,[3,4] fold plasticity, and activities.[2e,f] As they also exhibit
high conservation of tertiary structures and catalytic triads,
they have been suggested to have evolved from a common
protein ancestor,[5] from which divergent evolution led to the
emergence of a large number of promiscuous enzymes.[5b,6]
Serine esterases and haloacid dehalogenases[7] are a/b-type
hydrolases that differ in their topological features, the nature/
position of the nucleophile, and the geometry of their catalytic
scaffolds, while esterases (EC 3.1.1.1) preferentially hydrolyze
water-soluble simple esters and contain a Ser/Asp(Glu)/His cat-
alytic triad, haloacid dehalogenases (HAD; EC 3.8.1.2) catalyze
the conversion of haloacid compounds into the corresponding
alcohols and hydrogen halides by means of an Asp/Asp/His
catalytic triad. Although, it has been suggested that dehaloge-
nases are evolutionarily related to esterases,[8] so far no protein,
either naturally occurring or laboratory generated, has been
reported to possess both activities.
We describe here
a multifunctional a/b-hydrolase-fold
enzyme,[9] designated REBr, mined from a metagenome library
established from the DNA of a microbial community from sea-
water contaminated with crude oil. The protein showed a
novel hydrolytic phenotype, namely the cleavage of both
common p-nitrophenyl (pNP) and short aliphatic esters, and or-
ganic haloalkanoates. The existence of these two activities in a
single protein is remarkable as they involve distinct catalytic
mechanisms (for details see Figure S1 in the Supporting Infor-
mation).
[a] Dr. A. Beloqui, Dr. J. M. Vieites, Dr. M. Ferrer+
CSIC, Institute of Catalysis
The rEBr gene (933 bp),[9] encodes a protein (310 AA, Mr =
33852 Da) that exhibits high homology (up to 63% identity,
75% similarity) with a number of a/b-fold hydrolases (see Fig-
ure S2). Not only does this protein hydrolyze a series of com-
mercially available common pNP and nonactivated short fatty
acid esters as propyl propionate and ethyl butyrate, but also
haloacids: [(kcat/Km)]ester/[(kcat/Km)]haloacid factor of ~4:1 for the
best substrates, optimally at 408C and pH 8.0–8.5 (Figure S3;
Tables 1 and S2). Weak though measurable activity with halo-
alkanes was detected and no epoxide tested was hydrolyzed.
The enzyme cleaved a full set of halides at both terminal and
subterminal positions, with catalytic efficiencies increasing in
the order bromide (1-fold), fluoride (1.1-fold), chloride (12-fold),
Marie Curie 2, 28049 Madrid (Spain)
Fax: (+34)91-585-4872
E-mail: mferrer@icp.csic.es
[b] Dr. J. Polaina
CSIC, Instituto de Agroquꢀmica y Tecnologꢀa de Alimentos
Paterna, 46980 Valencia (Spain)
Fax: (+34)96-3636301
[c] Dr. D. Reyes-Duarte
Universidad Autꢁnoma Metropolitana-Cuajimalpa (UAM-C)
Artificios 40. Col. Hidalgo, Deleg. ꢂlvaro Obregꢁn
01120 Mꢃxico D.F. (Mexico)
[d] Dr. R. Torres
School of Chemistry, Universidad Industrial de Santander
678 Bucaramanga (Colombia)
Fax: (+57)7634-9069
[e] Dr. O. V. Golyshina, Prof. P. N. Golyshin+
School of Biological Sciences, Bangor University
Gwynedd, LL57 2UW (UK)
Table 1. Kinetic parameters for the wild-type REBr enzyme.[a]
Fax: (+44)1248-38-3629
[f] Dr. O. V. Golyshina, Dr. T. N. Chernikova, A. Waliczek, Prof. K. N. Timmis,
Prof. P. N. Golyshin+
Substrate
Km [mm]
kcat [sÀ1
]
kcat/Km [sÀ1 mÀ1
]
bromoacetate
0.42Æ0.04
0.52Æ0.07
0.34Æ0.02
0.31Æ0.02
12.4Æ3.10
22.5Æ4.20
1.01Æ0.24
3.25Æ0.32
0.21Æ0.04
1614.0Æ23.3
1183.0Æ82.2
624.0Æ41.5
298.0Æ21.5
341Æ14.3
3.8ꢂ106
2.3ꢂ106
1.8ꢂ106
9.6ꢂ105
2.8ꢂ104
1.2ꢂ104
3.0ꢂ105
8.0ꢂ102
3.5ꢂ106
16.2ꢂ106
1.3ꢂ102
HZI-Helmholtz Centre for Infection Research
Inhoffenstrasse 7, 38124 Braunschweig (Germany)
Fax: (+49)531-6181-4199
2,3-dibromopropionate
2-bromopropionate
3-bromopropionate
2-bromobutyrate
2-bromocaproate
monochloroacetate
monoiodoacetate
monofluoroacetate
p-nitrophenyl butyrate
ethyl butyrate
[g] Dr. A. Aharoni
University of Ben-Gurion of the Negev
Beer-Sheva 84105 (Israel)
Fax: (+972)647-921
261.8Æ20.0
307.0Æ18.3
2.7Æ0.3
[h] Prof. M. M. Yakimov
739.0Æ26.8
Institute for Coastal Marine Environment
Spianata S. Raineri, 86, Messina 98122 (Italy)
Fax: (+39)0906-69007
1.64Æ0.35 26500Æ830
90Æ5.0
11.7Æ0.6
[+] These authors contributed equally to this work.
[a] For reaction conditions (pH 8.5, T=408C) see ref. [9]. Activities to-
wards other common esters and haloacids are shown in detail in
Table S4.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201000258.
ChemBioChem 2010, 11, 1975 – 1978
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1975

and iodide (4800-fold), within the same substrate series. We
further estimated the apparent enantioselectivity (Eapp) of the
REBr enzyme by separate measurements of the kinetic parame-
ters for the hydrolysis of two model enantiomers, namely (S)-
and (R)-3-bromo-2-methylpropionate. It should be mentioned
that the ratios obtained by these measurements were not true
enantiomeric ratios (Etrue), because the rates of hydrolysis of
the enantiomers were measured separately. As shown in
Table S2, the enzyme showed a higher preference for the S-
configured halogen over the R-configured substrate (Eapp ~41).
The genes of three other a/b-hydrolases[10] homologous to
REBr were cloned from bacterial genomes and hyperexpressed,
then their products were purified. All enzymes showed slightly
lower (ca. sixfold) kcat/Km values for the cleavage of pNP esters
than REBr, and none of them hydrolyzed haloacids or haloal-
kanes. These REBr homologues thus behaved as normal ester
hydrolases (Table S3; Figure S3).
Figure 2. Proposed catalytic and stabilizing residues of the REBr protein. The
hydrolases CumD (EC 3.7.1.9; PDB: 2D0D; 20% identity) of Pseudomonas flu-
orescens IPO1 and HasD (EC 3.7.1.8; PDB: 2VF2; 25% identity) from Mycobac-
terium tuberculosis were identified as the best templates and used for ho-
mology modeling.
Halo-organics inhibited the catalytic activity of REBr towards
pNP butyrate sevenfold, and pNP butyrate strongly inhibited
the HAD reaction (75-fold; Figure 1); this demonstrates the
promiscuous behavior of the REBr enzyme and its ability to
concomitantly hydrolyze both esters and carbon–halogen
bonds. This unusual biochemical feature pointed to the possi-
bility of the coexistence of two different catalytic abilities in
the same protein.
almost complete loss of activity (catalytic efficiencies of the
mutants could not be determined because of the apparently
high value of Km), and S139G exhibited significant reduction
(ca. 20000-fold) in catalytic efficiency for 2-bromopropionate
(Table S4). This last result can be explained by the up to 75-
fold greater Km value coupled with a significant (328-fold) re-
duction in the kcat value.
In addition, amino acids E258, R260, V261, and L288 were
identified from the 3D model as being potential halide-stabiliz-
ing residues (Figure 2), therefore E258Q, R260L, V261A, and
L288A variants were generated.[9] None of the mutations had a
dramatic effect on the esterase activity of REBr (33–99% rel.
Km; 52–116% rel. kcat; 62–198% rel. kcat/Km; Table S4), but the
E258Q, R260L, and V261A substitutions dramatically reduced
activity towards halogenated compounds due to an up to 480-
fold lower kcat and 117-fold higher Km towards 2-bromopropio-
nate.
The kinetic constants of 2-bromopropionate binding to REBr
were further measured by monitoring intrinsic protein fluores-
cence by stopped-flow methods.[9] Addition of 2-bromopropio-
nate (which was used as model halo-organic) at different con-
centrations to REBr resulted in a clear fluorescence enhance-
ment (not shown) similar to that reported for the halohydrin
dehalogenase of Agrobacterium radiobacter.[11] The kinetics
could be fitted to a single exponential equation,[11] thus yield-
Figure 1. Reciprocal substrate inhibition of REBr-mediated hydrolysis of pNP
butyrate and 2-bromopropionate. The two left-hand bars show the kcat
values for hydrolysis of the individual substrates alone, whereas the right-
hand bars show the values obtained with an equimolar mixture of sub-
strates. Reaction conditions: [E]o =0–12 nm, 3 mm p-nitrophenyl butyrate or
2-bromopropionate, 100 mm Tris-sulfate, pH 8.5, T=408C. Kinetic parameter
(kcat) was calculated as in ref. [9]. Results shown are the average of three in-
dependent assays Æthe standard deviation.
ing an observed rate constant (kobs) from which k2 and k
À2
values could be calculated together with K1 and Kd values
(Table 2).
In order to identify important residues in REBr for substrate
binding and catalysis, we developed a 3D model (Figures S4
and 2) to guide site-directed mutagenesis. It revealed a catalyt-
ic pocket conserved in esterases (S139/D259/H287), which was
confirmed by establishing and analyzing S139G, D259N, and
H287Q variants of REBr. The Km, kcat and kcat/Km values for pNP
esters were all significantly reduced (Table S4). A radiochemical
assay with the substrate methyl [1-14C]butyrate[9] has confirmed
its binding and formation of the acyl-enzyme intermediate of
S139 (not shown), which is consistent with its role as a nucleo-
phile. Surprisingly, the D259N and H287Q variants showed an
The rate of substrate binding (k2: ~153 sÀ1) was about two-
fold higher than the rate of release of alkyl halide from the
active site (kÀ2: ~43–78 sÀ1); this suggests that binding is not a
limiting step for substrate conversion. This observation is con-
sistent with the higher carbon–halogen cleavage rates of REBr
(see Table 1), as compared with those of known dehalogenases
(kcat up to 41 sÀ1).[12] Like the parent protein, REBr mutants
D259N, H287Q, and E258Q exhibited fluorescence enhance-
ment upon addition of halo-organics at different concentra-
tions, and displayed similar Kd values (Figure 3). However, sub-
1976
www.chembiochem.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 1975 – 1978

Table 2. K1, k2, k and Kd constants for the REBr enzyme. 2-Bromopropio-
À2
nate in 100 mm Tris-sulfate (pH 8.5) was used as substrate for binding
experiments[a]
K1 [mm]
k2 [sÀ1
]
k
À2
[sÀ1
]
Kd [mm]
1.21Æ0.44
152.5Æ13.5
77.9Æ9.6
0.41Æ0.09
[a] Constants were calculated according to refs. [9] and [11].
Figure 3. Apparent dissociation constant (Kd) of wild-type REBr and its mu-
tants with 2-bromopropionate. Experiments were carried out in 100 mm Tris-
sulfate, pH 8.5, at 408C. Results shown are the average of three independent
assays Æthe standard deviation. Values were calculated according to Equa-
tion (5) in refs. [9] and [11].
strate addition to REBr mutants S139G, R260L, V261A, and
L288A did not induce any marked fluorescence increase (so ki-
netic parameters could not be determined); this suggests that
the mutated residues in these mutants affected the binding
sites of the enzyme.
Figure 4. A) Suggested esterase and B) haloacid dehalogenase catalytic
mechanisms for the REBr enzyme. Catalytic and binding residues are shown
in each model (for details see Figure S6). Substrates are shown in red. Resi-
dues marked in gray in panel A are not thought to play a direct role in ester-
ase activity, whereas residues in gray in panel B are suggested to play a role
in substrate binding. Although, the model might not represent the exact
locations of residues in the protein structure, it summarizes in a graphical
view the interactions that should exist around the catalytic core in the REBr
protein.
Time courses of substrate consumption and covalent
enzyme–intermediate complex formation were followed in a
reaction of 43 mm of [14C]2-bromopropionate and 78 mm of
wild-type protein. After filtration of the acid-quenched reaction
samples, enzyme, substrate-bound enzyme, and unconsumed
substrate were separated by HPLC.[9] Covalent attachment of a
radiolabel to the enzyme at position E258 was demonstrated
by separating the quenched reaction mixtures of parental and
mutant enzyme by SDS-PAGE. There was a coincidence of the
14C label and the parental enzyme band, and an absence of
labeled mutant enzyme (not shown), thereby confirming the
role of E258 as the nucleophile in carbon–halogen cleavage.
These biochemical analyses on mutant enzymes confirmed
residues S139, E258, D259, R260, V261, H287, and L288 in the
REBr parental enzyme as multifunctional amino acids partici-
pating in substrate binding and cleavage of a number of differ-
ent halogenated and unhalogenated substrates (Figure 4), with
S139, D259, and H287 serving as nucleophile, acid, and base,
respectively, for the cleavage of common esters. Furthermore,
these amino acid residues, together with E258, R260, V261,
and L288, play significant roles in the transformation of halo-
organics (see the putative catalytic mechanism in Figure 4).
Indeed, E258 could act as the nucleophile, whereas H287
might activate the water molecule needed for carbon–halogen
cleavage, with D259 stabilizing the H287 during the reaction. It
seems likely that the putative primary halide-stabilizing resi-
dues are R260 and V261, in combination with S139, which is
probably involved in stabilization of the COOH group, and
L288. Further determination of the structure of REBr by X-ray
crystallography is needed to confirm these notions.
Sequence alignments (Figure S5) to detect whether the
amino acids identified here as critical for activity are located in
conserved regions of a/b-hydrolases showed that three out of
seven—S139, D259 and H287 (REBr numbering)—are absolute-
ly conserved in all proteins. The quality of the alignment in
this region is very good, with no gaps and with an almost
exact superimposition of secondary-structure predictions. No
variants of either REBr, or any of the three other a/b-hydrolase
homologues analyzed in this study that bears a mutation in
these positions exhibited any catalytic activity. However, the
core motif E258·R260·V261·L288 determined in this study to be
critical for dehalogenase activity is not present in any other
homologue protein. These differences might explain the sub-
strate specificity of the protein under investigation, which,
together with previous studies,[6] suggests that active-site ar-
ChemBioChem 2010, 11, 1975 – 1978
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1977

Askew, A. C. O’Donoghue, F. Hollfelder, J. Am. Chem. Soc. 2008, 130,
16547–16555; d) M. Nardini, B. W. Dijkstra, Curr. Opin. Struct. Biol. 1999,
9, 732–737; e) M. Nardini, I. S. Ridder, H. J. Rozeboom, K. H. Kalk, R.
Rink, D. B. Janssen, B. W. Dijkstra, J. Biol. Chem. 1999, 274, 14579–
14586; f) M. Holmquist, Curr. Protein Pept. Sci. 2000, 1, 209–235.
[3] D. L. Ollis, E. Cheah, M. Cygler, B. Dijkstra, F. Frolow, S. M. Franken, M.
Harel, S. J. Remington, I. Silman, J. Schrag, Protein Eng. 1992, 5, 197–
211.
[4] M. E. Glasner, J. A. Gerlt, P. C. Babbitt, Curr. Opin. Chem. Biol. 2006, 10,
492–497.
[5] a) M. Skjøt, L. De Maria, R. Chatterjee, A. Svendsen, S. A. Patkar, P. R. Øs-
tergaard, J. Brask, ChemBioChem 2009, 10, 520–527; b) C. Li, M. Hassler,
T. D. Bugg, ChemBioChem 2008, 9, 71–76.
[6] H. Jochens, K. Stilba, C. Savile, R. Fujii, J.-G. Yu, T. Gerassenkov, R. J.
Kazlauskas, U. T. Bornscheuer, Angew. Chem. 2009, 121, 3584–3587;
Angew. Chem. Int. Ed. 2009, 48, 3532–3535.
[7] For reviews, see: a) U. T. Bornscheuer, FEMS Microbiol. Rev. 2002, 26, 73–
81; b) D. B. Janssen, I. J. Dinkla, G. J. Poelarends, P. Terpstra, Environ. Mi-
crobiol. 2005, 7, 1868–1882; c) A. M. Burroughs, K. N. Allen, D. Dun-
away-Mariano, L. Aravind, J. Mol. Biol. 2006, 361, 1003–1034.
[8] R. M. De Jong, J. J. M. Tisesinga, H. J. Rozeboom, K. H. Kalk, L. Tang, D. B.
Janssen, B. W. Dijkstra, EMBO J. 2003, 22, 4933–4944.
[9] a) Detailed experimental procedures are available in the Supporting
Information. b) Tentative phylogenetic affiliation is shown in Figure S2.
c) A graphical representation of optimal pH (~8.5) and temperature
(~408C) is shown in Figure S3.
chitecture relatedness is a major determining factor for the
evolution of new catalytic activities.
In conclusion, our study has further demonstrated the utility
of metagenomics to access novel catalytic activities, and has
delivered the first experimental evidence of a functional switch
from esterase to dehalogenase activity in a single protein. Our
results have several implications. First, they demonstrate the
global role of the highly conserved triad Ser/Asp/His in the
hydrolysis of both ester– and carbon–halogen bonds. Second,
the co-occurrence of both esterase and dehalogenase activities
in the REBr protein could provide a significant competitive
advantage in organisms thriving in environments containing
both classes of substrates.[13,14] In this context, mutations at
residues that have a significant role in stabilization, abstraction
or orientation of substrates in the active site (also having a
role in defining a certain level of physical/chemical freedom)
will provide variants from which new specificities can evolve
with or without an increase in catalytic promiscuity.
Acknowledgements
[10] a) Three predicted a/b-hydrolases from Alcanivorax borkumensis SK2
(Abo_1197; 32% similarity; Acc. Nr. YP_692917), Marinobacter aquaolei
VT-8 (Maqu_3452; 34% similarity; Acc. Nr. YP_960710) and Pseudomo-
nas aeruginosa PAO1 (Pao_2949; 36% similarity; Acc. Nr. NP_251639)
were selected as suitable candidates for comparative analysis. b) The
full nucleotide and amino acid sequences of those enzymes are report-
ed in the Supporting Information.
This research was supported by the Spanish MEC (projects
BIO2006-11738 and CSD2007-00005), CSIC-COLCIENCIAS (project
2006CO0016) and the European Commission within the FP7
(projects KBBE-226977 (MAMBA) and KBBE-245226 (MAGIC-PAH).
A.B. thanks the Spanish MEC for a FPU fellowship. O.V.G. was
supported by a grant from the Federal Ministry for Science and
Education (BMBF) within the GenoMikPlus initiative (0313751K).
[11] L. Tang, D. E. Torres PazmiÇo, M. W. Fraaije, R. M. de Jong, B. W. Dijkstra,
D. B. Janssen, Biochemistry 2005, 44, 6609–6618.
[12] For examples of highly active dehalogenases, see: a) R. J. Pieters, M.
Fennama, R. M. Kellogg, D. B. Janssen, Bioorg. Med. Chem. Lett. 1999, 9,
161–166; b) T. Bosma, J. Damborsky, G. Stucki, D. B. Janssen, Appl. Envi-
ron. Microbiol. 2002, 68, 3582–3587; c) T. Nagata, Z. Prokop, S. Marvano-
va, J. Sykorova, M. Monincova, M. Tsuda, J. Damborsky, Appl. Environ.
Microbiol. 2003, 69, 2349–2355.
Keywords: dehalogenases
·
esterases
·
hydrolysis
·
metagenomes · protein design
[1] For reviews, see: a) U. T. Bornscheuer, R. J. Kazlauskas, Angew. Chem.
2004, 116, 6156–6165; Angew. Chem. Int. Ed. 2004, 43, 6032–6040;
b) O. Khersonsky, C. Roodveldt, D. Tawfik, Curr. Opin. Chem. Biol. 2006,
10, 498–508; c) K. Hult, P. Berglund, Trends Biotechnol. 2007, 25, 231–
238.
[13] a) A. Aharoni, L. Gaidukov, O. Khersonsky, S. McQ Gould, C. Roodveldt,
D. S. Tawfik, Nat. Genet. 2005, 37, 73–76; b) K. Brenner, L. You, F. H.
Arnold, Trends Biotechnol. 2008, 26, 483–489.
[14] T. Futagami, Y. Morono, T. Terada, A. H. Kaksonen, F. Inagaki, Appl. Envi-
ron. Microbiol. 2009, 75, 6905–6909.
[2] For recent examples that illustrate unnatural reactions of synthetic im-
portance, see: a) G. Hasnaoui-Dijoux, M. M. Elenkov, J. H. L. Spelberg, B.
Hauer, D. B. Janssen, ChemBioChem 2008, 9, 1048–1051; b) J. K. Lassila,
D. Herslag, Biochemistry 2008, 47, 12853–12859; c) L. F. Olguin, S. E.
Received: April 26, 2010
Published online on August 16, 2010
1978
www.chembiochem.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 1975 – 1978