HheG, a halohydrin dehalogenase with activity on cyclic epoxides

Article abstract of DOI:10.1021/acscatal.7b01854

Halohydrin dehalogenases (HHDHs) are of biotechnological interest due to their promiscuous epoxide ring-opening activity with a set of negatively charged nucleophiles, enabling the formation of C-C, C-N, or C-O bonds. The recent discovery of HHDH-specific sequence motifs aided the identification of a large number of halohydrin dehalogenases from public sequence databases, enlarging the biocatalytic toolbox substantially. During the characterization of 17 representatives of these phylogenetically diverse enzymes, one HHDH, namely HheG from Ilumatobacter coccineus, was identified to convert cyclic epoxide substrates. The enzyme exhibits significant activity in the azidolysis of cyclohexene oxide and limonene oxide with turnover numbers of 7.8 and 44 s^-1, respectively. As observed for other HHDHs, the cyanide-mediated epoxide ring-opening proceeded with lower rates. Wild-type HheG displays modest enantioselectivity, as the resulting azido- and cyanoalcohols of cyclohexene oxide ring-opening were obtained in 40% enantiomeric excess. These biocatalytic findings were further complemented by the crystal structure of the enzyme refined to 2.3 ?. Analysis of HheG's structure revealed a large open cleft harboring the active site. This is in sharp contrast to other known HHDH structures and aids in explaining the special substrate scope of HheG.

Full text of DOI:10.1021/acscatal.7b01854

Research Article
pubs.acs.org/acscatalysis
HheG, a Halohydrin Dehalogenase with Activity on Cyclic Epoxides
Julia Koopmeiners,†,§ Christina Diederich,‡,∥ Jennifer Solarczek,† Hauke Voß,† Janine Mayer,†
Wulf Blankenfeldt,†,‡ and Anett Schallmey*,†
†Institute for Biochemistry, Biotechnology and Bioinformatics, Technische Universitat Braunschweig, Spielmannstraße 7, 38106
̈
Braunschweig, Germany
‡Structure and Function of Proteins, Helmholtz Centre for Infection Research, Inhoffenstraße 7, 38124 Braunschweig, Germany
S
* Supporting Information
ABSTRACT: Halohydrin dehalogenases (HHDHs) are of
biotechnological interest due to their promiscuous epoxide
ring-opening activity with a set of negatively charged
nucleophiles, enabling the formation of C−C, C−N, or C−O
bonds. The recent discovery of HHDH-specific sequence motifs
aided the identification of a large number of halohydrin
dehalogenases from public sequence databases, enlarging the
biocatalytic toolbox substantially. During the characterization of
17 representatives of these phylogenetically diverse enzymes,
one HHDH, namely HheG from Ilumatobacter coccineus, was
identified to convert cyclic epoxide substrates. The enzyme
exhibits significant activity in the azidolysis of cyclohexene oxide
and limonene oxide with turnover numbers of 7.8 and 44 s−1,
respectively. As observed for other HHDHs, the cyanide-mediated epoxide ring-opening proceeded with lower rates. Wild-type
HheG displays modest enantioselectivity, as the resulting azido- and cyanoalcohols of cyclohexene oxide ring-opening were
obtained in 40% enantiomeric excess. These biocatalytic findings were further complemented by the crystal structure of the
enzyme refined to 2.3 Å. Analysis of HheG’s structure revealed a large open cleft harboring the active site. This is in sharp
contrast to other known HHDH structures and aids in explaining the special substrate scope of HheG.
KEYWORDS: halohydrin dehalogenase, epoxide ring-opening, cyclohexene oxide, limonene oxide, biotransformation, crystal structure
the synthesis of highly enantioenriched oxazolidinones6,7 and
INTRODUCTION
■
tertiary alcohols.8,9
Halohydrin dehalogenases (HHDHs; also called haloalcohol
HHDHs belong to the superfamily of short-chain dehydro-
dehalogenases or epoxidases) are biocatalytically attractive
genases/reductases (SDR), with which they share considerable
enzymes for the synthesis of various β-substituted alcohols due
sequence homology as well as structural and mechanistic
to their promiscuous epoxide ring-opening activity with a range
features.10 Recently, HHDH-specific sequence motifs have
of anionic nucleophiles such as azide, cyanide, nitrite, cyanate,
been described that enable the unambiguous discrimination of
and formate.1 As their natural reaction, they catalyze the
HHDH sequences from the overwhelming majority of SDR
enzymes.11 Using a database mining approach based on these
reversible dehalogenation of vicinal haloalcohols with formation
of the corresponding epoxides. Since their discovery in 1968,2
motifs, the number of HHDH sequences could be increased by
many different biotechnological applications of HHDHs have
at least 10-fold. Until 2016, 69 diverse HHDH enzyme
been described on the basis of merely a handful of known
HHDH enzymes. One of the most prominent examples is their
use in the synthesis of ethyl (R)-4-cyano-3-hydroxybutyrate, a
chiral synthon for the production of statin side chains.3 This
process has been applied on the industrial scale, employing a
stereoselective ketoreductase in combination with a highly
engineered variant of HheC from Agrobacterium tumefaciens
AD1.4 The putative process variant of HheC carries a minimum
of 35 amino acid exchanges which had been generated by
ProSAR-driven enzyme evolution to significantly improve the
volumetric productivity of the HHDH-catalyzed step.4 Other
examples for biocatalytic applications of HHDHs include the
synthesis of chiral C3 units, such as epihalohydrins,5 as well as
sequences have been reported that can be subdivided into six
phylogenetic subtypes with sequence identities below 30%.1 A
selection of 17 new HHDHs, representing all phylogenetic
subtypes, has been characterized in more detail.12 Interestingly,
five of the tested enzymes displayed higher thermostabilities
and temperature optima in comparison to the well-known
HheC.
Until now, HHDHs have been described to convert only
terminal epoxides, resulting in a highly regioselective attack of
Received: June 7, 2017
Revised: August 24, 2017
© XXXX American Chemical Society
6877
DOI: 10.1021/acscatal.7b01854
ACS Catal. 2017, 7, 6877−6886

ACS Catalysis
Research Article
the nucleophile at the sterically less hindered β-carbon atom. In
a study by Elenkov et al., vicinally disubstituted as well as cyclic
epoxides have been tested in reactions with HheC in addition
to HheA2 from Arthrobacter sp. AD2 and HheB2 from
Mycobacterium sp. GP1.13 None of the enzymes used, however,
were able to convert these sterically more demanding epoxide
substrates. With our large set of highly diverse new HHDHs at
hand, we anticipated that it should be possible to identify
enzymes with novel catalytic features. In particular, the selective
epoxide ring-opening of cyclic epoxides using HHDHs would
be of high interest, as this would enable access to chiral
synthons for the production of pharmaceuticals such as
(1R,2S)-4-((2-cyanocyclohexyl)oxy)-2-(trifluoromethyl)-
benzonitrile, a potent androgen receptor antagonist.14
azidolysis reaction. Similarly, HheG also achieved the highest
product yields (98% after 24 h) in the azidolysis of 5 in
comparison to other HHDHs (Figure 1). Only HheE5 (62%
conversion) and HheF (79% conversion) also exhibited
significant activity on 5 in comparison to the chemical
background reaction giving 15% of 6. To the best of our
knowledge, HheG therefore is the first HHDH reported to
accept cyclic epoxide substrates in epoxide ring-opening
reactions.
Kinetic Characterization of HheG. In order to investigate
HheG’s ability to convert cyclic substrates in more detail,
kinetic parameters of HheG in the conversion of epoxides 2
and 5 with azide and cyanide were determined. In addition,
HheG kinetics in the dehalogenation of chlorocyclohexanol (1)
were measured (Table 1). Since fitting of the obtained
experimental data to the Michaelis−Menten equation was
poor, the Hill equation was used instead. Comparison of the
obtained kinetic constants for the three tested substrates (1, 2,
and 5) revealed significant differences. The highest K50 value
(30 mM) was observed for haloalcohol 1, whereas the K50 value
for the corresponding epoxide 2 was lower. Furthermore,
HheG’s turnover number (kcat) in the azidolysis of 2 was more
than 20 times higher than the dehalogenation of the
corresponding chloroalcohol 1. Moreover, azidolysis of 5 with
a kcat value of 44 s−1 was about 5.5 times faster than the
azidolysis of epoxide 2 (kcat value of 7.8 s−1). Interestingly, the
Hill coefficient for the azidolysis of 5 (nH = 1.7) by HheG is
lower than that for azidolysis of 2 (nH = 3). This suggests a
stronger cooperativity for binding of epoxide 2 to HheG than
for binding of epoxide 5. A comparison of the obtained K50
values for the nucleophiles azide (8.6 mM) and cyanide (20
mM) indicates a higher affinity of HheG toward azide.
Similarly, HheG displayed a stronger cooperativity for binding
of CN− (nH = 3.3) than for N3− (nH = 1.7). This might explain
why the determined kcat value in the cyanolysis of 2 was more
than 1000-fold lower than that in the respective azidolysis
reaction. For azide a fixed nucleophile concentration of 50 mM
was used, while the nucleophile concentration was set to 75
mM for cyanide. Especially in the case of cyanide, this
nucleophile concentration might not have been high enough to
exclude any major influence on the kinetic experiments. For
cyanide, however, higher nucleophile concentrations lead to
substantial nonenzymatic product formation observed as high
chemical background. The dependence of HHDH activity in
the epoxide ring-opening reactions on the type of nucleophile is
well in line with previous observations.15 When nine different
nucleophiles were tested in the epoxide ring-opening of
epoxybutane using HheC, a 64-fold higher kcat value was
observed when azide was used as the nucleophile in comparison
to cyanide. Moreover, of all nine tested nucleophiles HheC
exhibited the highest activity with azide.
RESULTS AND DISCUSSION
■
Conversion of Cyclic Epoxides by HHDHs. To
investigate the activity of new HHDHs toward cyclic epoxides,
our 17 previously characterized HHDHs12 together with the
well-known HheC from A. tumefaciens10 were applied in
conversions of cyclohexene oxide (2) and (+)-cis/(+)-trans-
limonene oxide (5). Reactions were carried out using sodium
azide as the nucleophile to yield azido alcohols 4 and 6,
respectively, as products (Scheme 1). One enzyme, HheG from
Scheme 1. Substrates and Products of Cascade and Epoxide
a
Ring-Opening Reactions
a
Substrates: 2-chlorocyclohexanol (1), cyclohexene oxide (2),
limonene oxide (5); products: 2-cyano-1-cyclohexanol (3); 2-azido-
1-cyclohexanol (4); 2-azido-1-methyl-4-(prop-1-en-2-yl)cyclohexan-1-
ol (6a), and 2-azido-2-methyl-5-(prop-1-en-2-yl)cyclohexan-1-ol (6b).
Ilumatobacter coccineus, catalyzed the complete conversion of
cyclic epoxide 2, yielding 2-azido-1-cyclohexanol (4) as the
product (Figure 1). All other tested HHDHs, including the
previously described HheC, showed only low conversion (up to
40%). Considering that the chemical background azidolysis
already yielded up to 30% of 4, only HheG showed a significant
enzymatic activity on 2. Interestingly, in previous studies on the
substrate scope of HheG, only low specific activities were
obtained toward aliphatic haloalcohols such as 1,3-dichloro-2-
propanol (<0.1 U mg−1) and ethyl 4-chloro-3-hydroxybutyrate
(0.2 U mg−1) as well as the arylaliphatic haloalcohol 2-chloro-1-
phenylalcohol (<0.1 U mg−1).12 The corresponding epoxides,
epichlorohydrin and styrene oxide, were also hardly converted
with azide as nucleophile. The only HheG substrate for which a
high activity has been reported so far is epoxide glycidylphenyl
ether, for which a product yield of 100% was achieved in the
Interestingly, this is the first report for cooperative substrate
binding to a HHDH. Previous kinetic analysis of several other
HHDHs has not provided evidence for cooperative behavior of
these enzymes.7,10,16 The Hill coefficients reported here may
therefore indicate that HheG varies from other HHDHs by
requiring structural changes across the homotetramer in the
course of the catalytic cycle.
In a previous report, it was observed that HheG is quite
10
50
thermolabile, exhibiting a T value (half inactivation temper-
ature after incubation for 10 min) of 40 °C.12 In these initial
stability experiments, however, residual activity was determined
on the basis of substrate dehalogenation. Given that HheG is
6878
DOI: 10.1021/acscatal.7b01854
ACS Catal. 2017, 7, 6877−6886

ACS Catalysis
Research Article
Figure 1. Conversions obtained in reactions of cyclohexene oxide (2) and (+)-cis/(+)-trans-limonene oxide (5) using purified HHDH and azide as
nucleophile. Control reactions were carried out in buffer without enzymes.
Table 1. Kinetic Parameters of HheG with Substrates Chlorocyclohexanol (1), Cyclohexene Oxide (2), and (+)-cis-/(+)-trans-
Limonene Oxide (5) Using Cyanide (NaCN) or Azide (NaN3) as Nucleophile, as well as K50 Values of Nucleophiles Azide and
Cyanide
K50 (mM)
kcat (s−1
)
kcat/K50 (mM−1 s−1
)
nH
chlorocyclohexanol (1)
30.0 3.0
22.2 2.0
23.7 2.6
19.4 2.4
8.6 0.5
0.3 0.03
0.01 0.01
2.9 0.6
2.6 0.5
3.0 0.9
1.7 0.3
1.7 0.2
3.3 0.3
cyclohexene oxide (2) [NaCN]
cyclohexene oxide (2) [NaN3]
(+)-cis-/(+)-trans-limonene oxide (5) [NaN3]
NaN3 [cyclohexene oxide]
(2 0.00) × 10−4
7.8 0.35
(7.7 0.00) × 10−6
0.3 0.13
44 0.01
2.3 0.00
NaCN [cyclohexene oxide]
19.8 0.5
more active in the epoxide ring-opening of cyclic epoxides,
temperature and pH optima of HheG were determined in the
epoxide ring-opening of 2 using N3 as nucleophile here
biocatalytic route toward the androgen receptor antagonist
(1R,2S)-4-((2-cyanocyclohexyl)oxy)-2-(trifluoromethyl)-
benzonitrile.14
−
(Figure S1 in the Supporting Information). This gave a
temperature optimum of 30 °C and a complete enzyme
inactivation at 40 °C, confirming that HheG is quite unstable.
The obtained pH optimum between pH 6 and 7 is in the range
of previously studied HHDHs with HheA, HheB, and HheC,
showing an optimum in the epoxide ring-opening between pH
4 and 7.10,17−20
In the azidolysis of 2, HheG exhibited an ee value of 40% for
the formation of one enantiomer of azidoalcohol 4. To
determine the stereoconfiguration of this enantiomer, azidoal-
cohol 4 was chemically reduced to the corresponding
aminoalcohol and further analyzed by chiral GC after
derivatization. By comparison with enantiopure authentic
standards of (1R,2R)- and (1S,2S)-2-aminocyclohexan-1-ol, it
was observed that HheG favors the formation of (1S,2S)-4
(Scheme 1). Interestingly, this is the opposite enantiomer in
comparison to the preferentially formed 1S,2R enantiomer of
cyanoalcohol 3.
Whole-Cell Biotransformation of Cyclohexene Oxide
and Cascade Reaction. Production of cyanoalcohol 3 and
azidoalcohol 4 was also tested using HheG-overexpressing E.
coli cells at OD600 = 40. In addition, we extended our
investigation to the cascade-type reaction starting from
chloroalcohol 1 (Scheme 1). This strategy of combining
dehalogenation and epoxide ring-opening in a cascade-type
reaction using HHDHs was previously reported for the
conversion of ethyl 4-chloro-3-hydroxybutyrate into ethyl 4-
cyano-3-hydroxybutyrate.4 With 20 mM chloroalcohol 1 as the
starting material, the whole-cell system efficiently catalyzed the
formation of epoxide 2 as intermediate as well as subsequent
epoxide ring-opening due to the presence of 40 mM NaCN in
the reaction mixture. After 24 h, cyanoalcohol 3 was obtained
Stereoselectivity of HheG. Cyanolysis of 2 leads to the
production of trans-2-hydroxycyclohexane-1-carbonitrile (3).
The 1S,2R enantiomer of this compound serves as an
intermediate in the synthesis of the androgen receptor
antagonist (1R,2S)-4-((2-cyanocyclohexyl)oxy)-2-
(trifluoromethyl)benzonitrile21 and is therefore of industrial
interest. To analyze HheG’s selectivity in the cyanolysis of 2,
reactions with purified HheG, 5 mM 2, and 20 mM NaCN
were carried out at room temperature and analyzed by chiral
GC after extraction. A moderate enantiomeric excess (ee) of
cyanoalcohol 3 with 37% ee after 30 min reaction time was
obtained. The enantiomeric excess slightly decreased upon
prolonged incubation due to a significant chemical background
cyanolysis reaction. Using chiral GC and a lipase-catalyzed
enantiomeric separation strategy,22,23 the stereoconfiguration of
the preferentially formed enantiomer was determined and
confirmed that HheG favors the formation of (1S,2R)-3
(Scheme 1). Thus, HheG’s enantiopreference opens a new
6879
DOI: 10.1021/acscatal.7b01854
ACS Catal. 2017, 7, 6877−6886

ACS Catalysis
Research Article
with a yield of 50%. After that, the product yield increased only
slightly, indicating an inactivation of the whole-cell catalyst
(Figure 2A). Interestingly, the chemical background cyanolysis
Figure 3. Azidolysis of (+)-cis-/(+)-trans-limonene oxide (5) to
azidoalcohol 6 using 20 mM 5, 40 mM NaN3, and whole cells of E. coli
BL21(DE3) overexpressing HheG with an OD600 value of 40. Control
reactions contained E. coli BL21(DE3) cells carrying empty vector
pET-28a. Error bars represent standard deviations of duplicate
measurements.
Figure 2. Product yields of whole-cell biotransformations: comparison
of (A) cascade reaction starting from haloalcohol 1 (20 mM) to
cyanoalcohol 3 and (B) epoxide ring-opening of 2 (20 mM) to
cyclohexan-1-ol (6b), depending on the carbon atom that is
attacked by the nucleophile (Scheme 1). Due to the lack of
commercial standards for azidoalcohols 6a,b, the azidolysis
reaction of (+)-trans-limonene oxide ((+)-trans-5) using whole
cells was scaled up for subsequent product isolation. On the
basis of NMR analysis of purified product 6, HheG converts
(+)-trans-5 exclusively into (+)-trans-2-azido-1-methyl-4-(prop-
1-en-2-yl)cylohexan-1-ol ((+)-trans-6a). Hence, the nucleophile
exclusively attacked the sterically less hindered carbon atom of
the epoxide ring, resulting in formation of a tertiary alcohol
with absolute regioselectivity. This regioselectivity is in line
with the reported selectivity of previously characterized
HHDHs in the epoxide ring-opening of 2,2-disubstituted
terminal epoxides also yielding the corresponding tertiary
alcohols exclusively.13
Overall Structure and Active Site of HheG. To
investigate possible structural determinants of the unique
substrate scope of HheG, we determined the crystal structure of
HheG at 2.3 Å resolution (see the Supporting Information). As
found previously for other HHDHs, HheG forms a stable
homotetramer (Figure 4A). The monomer exhibits a
Rossmann-fold-like architecture characterized by a central
seven-stranded parallel β-sheet (β1−β7) flanked by eight α-
helices (Figure 4B). The active site of HheG is located in a long
and deep cleft (Figure 5A) formed between the helices α4 and
α5 and the helix−turn−helix motif involving helices α6 and α7.
The catalytic triad comprised of Ser152, Tyr165, and Arg169
can be found at the bottom of this partially electropositive cleft
(Figure 6). Ser152 is located in a flexible loop region between
the 310-helix η1 and the fifth β-strand, while Tyr165 and Arg169
are at the C-terminal pole of α5. In the structure of HheG, the
strongly electropositive halide binding site is occupied by a
water molecule, which is coordinated by the side chain of
Thr195 (Figure 4C).
product 3 using HheG expressing E. coli BL21(DE3) cells (OD600
=
40) and 40 mM sodium cyanide (NaCN). Control reactions were
performed using the same reaction conditions with whole cells of E.
coli BL21(DE3) lacking HheG. Error bars represent standard
deviations of duplicate measurements.
of 2 in the cascade reaction was very low (<5% after 120 h)
despite the high nucleophile load. This reduced chemical
background had a beneficial effect on the ee value of 3, as in the
cascade reaction (1S,2R)-3 was obtained with 40% ee. When
the cascade reaction was compared to the epoxide ring-opening
of 2 using whole cells, it was striking that in both cases similar
yields (around 60%) of cyanoalcohol 3 were obtained (Figure
2B). The chemical background cyanolysis of 2, however, was
significantly higher when epoxide 2 was used as the starting
substrate and caused a slight drop in ee value of 3 to 36%. In
further tests with increased substrate and nucleophile loads (50
and 100 mM, respectively), the product yield of 3 could be
increased to roughly 80%. A higher nucleophile concentration,
on the other hand, led to a higher unselective chemical
background, resulting in a drop in ee value for 3 of below 30%.
In future applications, cyanide could be titrated to the reaction
mixture as reported previously for the synthesis of ethyl (R)-3-
hydroxyglutarate by combination of an HHDH and a
nitrilase.24 This way, the nucleophile concentration can be
kept at a value that enables sufficient HHDH cyanolysis activity
and at the same time limits the chemical background reaction
to ensure maximum product enantiomeric excess.
The whole-cell cascade setup was also successfully applied in
reactions producing azidoalcohol 4 from chloroalcohol 1. Here,
100% yield with low chemical background (<5%) and an ee
value for product 4 of 40% were obtained after 24 h.
Biotransformation of Limonene Oxide and Product
Analysis. Azidolysis of (+)-cis-/(+)-trans-5 to produce
azidoalcohol 6 was also carried out using E. coli whole cells
overexpressing HheG (Figure 3). Production of 6 using whole
cells was very fast and yielded 80% conversion already after 1 h
incubation at room temperature. Due to the fast conversion, the
chemical background remained low. Azide-mediated epoxide
ring-opening of 5 could lead to the formation of two possible
regioisomers, namely 2-azido-1-methyl-4-(prop-1-en-2-yl)-
cyclohexan-1-ol (6a) and 2-azido-2-methyl-5-(prop-1-en-2-yl)-
Residual electron density was observed in proximity of the
catalytic residue Ser152 (Figure S2 in the Supporting
Information). This unattributed density, which superposes
with the binding site of ligands from other HHDH structures,
was also observed in the presence of reducing agent and in
several structures obtained from different crystallization
conditions of varying pH from 4.5 to 8.5. The quality of the
electron density maps does not allow unambiguous identi-
fication of the bound compounds, which have likely been
6880
DOI: 10.1021/acscatal.7b01854
ACS Catal. 2017, 7, 6877−6886

ACS Catalysis
Research Article
Figure 4. (A) Tetrameric biological assembly of the halohydrin dehalogenase HheG from Ilumatobacter coccineus shown as cartoon and surface
representations. The individual chains are depicted in different colors. (B) Secondary structure elements within the Rossmann-fold-like architecture
of a HheG monomer. The residues constituting the catalytic triad, Ser152, Tyr165, and Arg169, are highlighted as balls and sticks. (C) Close-up on
the active site of HheG comprised of the catalytic triad shown with its hydrogen-bonding network (- - -), as well as the additional residues Thr195
and Phe203 with the former coordinating the water molecule (Wat516) occupying the halide binding site.
Figure 5. Surface representation of HheG from I. coccineus and HheC (mutant) from A. tumefaciens (PDB entry: 3ZN2),25 respectively. (A) In HheG
the active site harboring the catalytic triad (highlighted in red) is located in a large open cleft shown in teal. (B) In HheC, a narrow tunnel (shown in
yellow) is leading to the active site with the catalytic triad highlighted in red.
copurified with the recombinant protein. Their presence,
however, may explain why all attempts to cocrystallize HheG
with substrates 1, 2, and 5 were unsuccessful.
in HheC (Figure S3A in the Supporting Information), where it
reaches into the active site of an adjacent monomer via a
conserved tryptophan residue.27
Structural Comparison to Related HHDHs. A compar-
ison of the structure of HheG with 14 previously published
HHDH structures yields rmsd values on Cα positions ranging
from 1.47 (PDB entry: 3ZN2) to 1.71 Å (PDB entry: 4Z9F).
The overall folds of all HHDHs, including HheG, are rather
similar and resemble the characteristic Rossmann fold that is
found in all members of the SDR family. However, marked
differences exist between HheG and the other members of this
enzyme class. In general, HheG features a seventh β-strand
which can be found in HheA and HheC as well but which is
missing in HheB. Similar to the case for HheA and HheB,
HheG lacks an extended C-terminus which can be found only
The most distinct differences between HheG and other
HHDHs, however, are the shape and size of the active site
pocket. While HheC, HheB, and HheA have rather narrow
substrate tunnels leading to the active site,28 HheG features a
large open cleft (Figure 5). Using the program KVFinder29 the
volume of this active site cleft was calculated to be ∼520 Å3. As
a comparison, the active site pocket in HheC (PDB entry:
3ZN2) is only 190 Å3 (Figure S4 in the Supporting
Information). The loop spanning residues His37 to Phe52,
which is not present in the other structures and connects the β2
strand to helix α2, delimits the active site cleft of HheG on one
side. Consequently, the α2 helix, not present in HheB
6881
DOI: 10.1021/acscatal.7b01854
ACS Catal. 2017, 7, 6877−6886

ACS Catalysis
Research Article
displays 74% sequence identity to HheG and was therefore also
studied for its activity on cyclic epoxides 2 and 5. Using 0.4 mg
mL−1 purified enzyme in reactions with 20 mM substrate and
40 mM nucleophile (NaN3 and NaCN), products 3, 4, and 6
were obtained in 39, 99, and 79% conversion, respectively, after
overnight incubation at room temperature. HheG2 displayed
also some enantioselectivity in the conversion of 2, giving
cyanoalcohol 3 with 22% ee and azidoalcohol 4 with 41% ee,
which are comparable to the data obtained for HheG. Hence,
the ability to convert cyclic epoxides is not exclusive to HheG
but might likely represent a general feature of G-type HHDH
subfamily members. To verify this hypothesis, a PHI-BLAST
database search for potential HheG homologues in the most
recent GenBank release 220 was performed using our
previously published HHDH sequence motifs11 with HheG
as the query sequence. However, no additional member of the
G-type subfamily could be identified. Thus, it remains to be
examined whether further HheG homologues will exhibit the
same substrate preference for cyclic epoxides.
Figure 6. Close-up of the electropositive active site pocket of HheG.
The electrostatic potential was calculated using the ABPS plugin26 of
PyMOL and is represented on the molecular surface with a scale
ranging from −15 (red) to 15 (blue) kbT/e. The residues of the
catalytic triad Ser152, Tyr165, and Arg169 and the conserved residues
Thr13 and Gly20, as well as Phe203 and the water molecule occupying
the halide binding site (Wat516), are shown as balls and sticks.
CONCLUSION
structures, adopts an atypical position almost parallel to α1
(Figure S3A in the Supporting Information). Another
important structural feature that determines the shape of the
active site pocket of HheG is helix α6. This helix is longer in
HheG, while in the other HHDH structures it corresponds to a
loop that closes the active site. In the latter structures, an
aromatic residue (Tyr185 in HheA2, Tyr169 in HheB from
Corynebacterium sp. N-1074, or Phe186 in HheC) is mainly
responsible for delimiting the active site tunnel within this
region (loop) together with the residues of the catalytic triad
and an additional conserved aromatic residue (Phe12 in HheC
and HheA2 or Tyr19 in HheB). Not only is the corresponding
residue Phe203 in the α6 helix of HheG displaced by at least 2.0
Å (regarding Cα positions) in comparison to HheA2, HheB, or
HheC but also its side chain is flipped away from the active site
by ∼180°. As such, it occupies the position which in related
halohydrin dehalogenases is occupied by an adjacent aromatic
residue (Phe170 in HheB, Phe186 in HheA2, or Tyr187 in
HheC), inducing an opening of the active site. Further, the
conserved aromatic residue Tyr18 is displaced by at least 2.4 Å
(regarding Cα positions) in comparison to the corresponding
residues in structures of HheA2/HheC or HheB. This
displacement leads to additional widening of the active site
cleft of HheG (Figure S3B). While it cannot be excluded that
substrate binding induces a movement of α6 or at least a flip of
Phe203 to close the active site, such a conformational change
has not been observed in other HHDHs. It might be speculated
that such substrate-induced conformational changes, on the
other hand, could account for the observed cooperative
behavior of HheG. Further biochemical and structural
investigations, however, will be necessary to explain the
observed cooperativity in the future.
■
HheG from Ilumatobacter coccineus is special among other
halohydrin dehalogenases, as it can be applied in the conversion
of sterically demanding cyclic epoxides while its activity toward
standard HHDH substrates is comparably low. Using this
enzyme, different cyclic β-substituted alcohols such as 2-azido-
and 2-cyano-1-cyclohexanols can be accessed, which are even
formed with moderate enantiomeric excess. With this, HheG
broadens the versatility of the HHDH enzyme toolbox for
biocatalytic routes toward fine chemicals and pharmaceutical
synthons. For future practical applications, however, enzyme
optimization is still required. Especially the enzyme’s activity
with cyanide, its enantioselectivity, and its stability have to be
enhanced by protein engineering, which is subject to current
investigation in our group. The crystal structure of HheG
reported here aids in explaining the substrate preference of this
enzyme, as its active site is positioned in a wide-open cleft,
which is in contrast to rather buried active sites found in other
HHDH structures. Additionally, this structure will substantially
facilitate further protein engineering campaigns of HheG and
its close homologue HheG2, which also exhibits activity toward
cyclic epoxides.
EXPERIMENTAL SECTION
■
Chemicals. All chemicals were of analytical or the highest
available grade. Substrates 2-chlorocyclohexanol (1), (+)-cis-/
(+)-trans-limonene oxide (5) and (+)-trans-limonene oxide
((+)-trans-5) were purchased from Sigma-Aldrich; cyclohexene
oxide (2) and trans-2-hydroxycyclohexane-1-carbonitrile (3)
were obtained from Acros Organics.
Bacterial Strains, Plasmids, and Enzymes. E. coli
BL21(DE3) Gold (Life Technologies, Darmstadt, Germany),
E. coli C43(DE3) (Lucigen Corporation, Middleton, WI, USA),
and E. coli Top10 (Life Technologies) were used as hosts for
heterologous protein production. pET-28a-based vectors
harboring a T7 promoter and resulting in N-terminal
hexahistidine (His6) tag fusion were used for expression of
HHDH genes.11 Vector pBAD-hheC harboring an ara
promotor was used for the expression of the HheC gene.30
Enzyme Expression and Purification. Heterologous
enzyme production of HHDHs was carried out as reported
previously.12 For the production of recombinant HheG, 500
Overall, we propose that the large active site cleft, in
comparison to the narrow active site pockets in related
HHDHs, explains why HheG preferentially converts cyclic
epoxides over smaller aliphatic substrates. At the same time, this
observation may indicate that the enzyme is able to
accommodate even larger or sterically more demanding
substrate molecules. The large epoxide trans-stilbene oxide,
however, was not converted by HheG in previous tests.12
HheG Homologues. Using our motif-driven sequence
database search, we recently identified a close homologue of
HheG.1 The enzyme from Ilumatobacter nonamiensis (HheG2)
6882
DOI: 10.1021/acscatal.7b01854
ACS Catal. 2017, 7, 6877−6886

ACS Catalysis
Research Article
mL TB medium (4 mL L−1 glycerol, 12 g L−1 peptone, 24 g L−1
yeast extract, 0.17 M KH2PO4, 0.74 M K2HPO4) supplemented
with 50 mg L−1 kanamycin and 0.2 mM IPTG was inoculated
with 10% (v/v) overnight culture of E. coli BL21(DE3) pET-
28a-hheG and incubated for 24 h at 22 °C. Cells were harvested
by centrifugation (4400g, 20 min at 4 °C), and the cell pellet
was stored at −20 °C until further use.
Temperature and pH Profiles of HheG. To determine
HheG’s temperature and pH optima, reactions were carried out
in 0.5 mL using 20 mM 2, 40 mM NaN3, and 0.4 mg mL−1
HheG. Temperature profile reactions were performed in 50
mM Tris·SO4, pH 8.0, buffer at temperatures ranging from 4 to
90 °C. After 30 min, 500 μL of TBME containing 0.1% (v/v)
dodecane was added to each of the reaction mixtures for
extraction. In case of pH profiles, buffers of varying pH were
used (50 mM citrate buffer at pH 4.0−6.5, 50 mM sodium
phosphate buffer at pH 6.0−8.0, 50 mM Tris·SO4 buffer at pH
7−9, and glycine NaOH buffer at pH 8−11) and reaction
mixtures were extracted after incubation at 30 °C for 40 min
using 500 μL of TBME containing 0.1% (v/v) dodecane.
Product formation was followed by gas chromatography.
Negative control reactions without enzyme were performed
in parallel to determine the chemical background of epoxide
ring-opening. Negative controls and reactions containing
enzyme were carried out in duplicate, and the chemical
background was subtracted from enzymatic conversions. In
temperature and pH profile tests, the highest observed
conversion in the sample series was set to 100%.
Steady-State Kinetics of HheG. Enzyme kinetics of HheG
toward chloroalcohol 1 were determined by monitoring halide
release12 in microtiter-plate format using a CLARIOstar
spectrophotometer (BMG Labtech, Ortenberg, Germany).
Dehalogenation reactions were performed in 25 mM Tris·
SO4 buffer at pH 7 containing 0.5−150 mM substrate 1 in a
total volume of 1.5 mL. The reaction mixtures were incubated
at 30 °C after 100 μg mL−1 HheG was added. To calculate
initial activities, 100 μL of each sample was withdrawn after 0.5,
1.5, 2.5, 3, 4, 5, 6, 10, 15, 20, 30, and 60 min and mixed with
100 μL of assay reagent (equal volumes of solution I (0.25 M
NH4Fe(SO4)2 in 9 M HNO3) and solution II (saturated
solution of Hg(SCN)2 in absolute ethanol)). Absorbance at
460 nm was measured immediately, and the resulting
dehalogenase activities were calculated using a standard curve
for Cl− in the range of 0−2 mM. Activities were calculated in
μmol min−1, plotted against the substrate concentration, and
fitted according to the Hill equation for cooperative binding
(eq 1)
Purification of HheG was carried out by affinity chromatog-
raphy. A frozen cell pellet was resuspended in buffer A (50 mM
Tris·SO4, 500 mM Na2SO4, 25 mM imidazole, pH 7.9),
containing 1 mg mL−1 lysozyme and 100 μM phenyl-
methylsulfonyl fluoride as protease inhibitor. Cells were
disrupted by sonication (six cycles of 1 min at an amplitude
of 60% and 1 min pause, on ice), and the resulting crude extract
was centrifuged (16600g, 30 min, 4 °C) to obtain HheG-
containing cell free extract (CFE). This CFE was filtered
through a 0.45 μm cellulose acetate membrane filter (Sarstedt,
Numbrecht, Germany) and loaded (flow rate 2 mL min−1) on a
̈
5 mL HisTrap FF column (GE Healthcare, Freiburg, Germany)
̈
pre-equilibrated with buffer A using an Akta FPLC system (GE
Healthcare). To remove nonspecifically bound proteins, the
column was washed with five column volumes of buffer A at a
flow rate of 5 mL min−1. Elution of target protein was achieved
using a 100 mL gradient of 25−500 mM imidazole in buffer A
(flow rate 2 mL min−1). Fractions containing target protein
HheG were identified by SDS-PAGE. Fractions of highest
purity were pooled and concentrated by ultrafiltration using
Amicon Ultra-15 centrifugal filter units with 10 kDa NMWL
(Merck Millipore, Darmstadt, Germany). Afterward, the
concentrated purified protein was desalted using a PD-10
column (GE Healthcare) and TE buffer (10 mM Tris·SO4, pH
7.9, 4 mM EDTA). Protein concentration was measured on a
NanoDrop 1000 spectrophotometer (Thermo Scientific,
Waltham, MA, USA) at 280 nm using a molar extinction
coefficient for His-tagged HheG of 15470 M−1 cm−1 and a
molecular weight of 29881 Da.
For crystallization studies, HheG was expressed and purified
as described above using an additional size-exclusion
chromatography step. After affinity chromatography, the
collected fractions were loaded on a HiLoad 26/600 Superdex
200 pg column (GE Healthcare) and HheG was eluted with
crystallization buffer (see below), applying a flow of 2 mL
min−1.
V
max[S]nH
(K50)nH + [S]nH
V =
(1)
Selenomethionine-labeled HheG (SeMet His6-HheG) was
produced as described above but using minimal media (M9)
containing 60 mg L−1 selenomethionine. Protein purification
was carried out using a protocol similar to that described above.
Bioconversion of Cyclic Epoxides using 17 HHDHs.
Conversion of cyclic epoxides 2 and 5 by 17 different HHDHs
was tested on a small scale and analyzed by gas chromatography
(GC) and GC coupled with mass spectrometry (GC-MS).
Reactions were performed in a total volume of 1 mL in 50 mM
Tris·SO4 buffer, pH 8.0, at room temperature with 20 mM
sodium azide (NaN3) using 150 μg of each HHDH and 5 mM
of substrate 2 or 5. Respective negative control reactions
without enzyme but with only substrate with NaN3 were carried
out in parallel. After 24 h, 0.5 mL samples of each reaction
mixture were withdrawn and extracted using an equivalent
volume of tert-butyl methyl ether (TBME) supplemented with
0.1% (v/v) dodecane as internal standard. After the organic
extracts were dried over anhydrous MgSO4, samples were
injected on GC for quantification and on GC-MS for product
identification.
where V is the reaction velocity, Vmax the maximum velocity of
the reaction, [S] the substrate concentration, K50 the substrate
concentration at half-maximal reaction velocity, and nH the Hill
coefficient. Chemical background activities in control reactions
containing no enzyme were subtracted before fitting.
In the case of epoxide ring-opening of substrates 2 and 5,
reactions were analyzed by GC. Kinetic constants of HheG for
ring-opening of 2 were determined using either NaCN or NaN3
as nucleophile at fixed concentrations of 75 and 50 mM,
respectively. Kinetic constants of HheG for ring opening of 5
were determined using 50 mM NaN3 as nucleophile. Reactions
were carried out in 1.6 mL of 50 mM Tris·SO4 buffer at a pH of
8 containing 5−150 mM 2 and 300 μg mL−1 (in the case of
NaCN) or 156.25 μg mL−1 (in the case of NaN3) HheG at 22
°C. After 10, 30, 60, and 90 min, each 400 μL sample was taken
and extracted with 400 μL of TBME containing 0.1% (v/v)
dodecane. Product formation was quantified by GC on the
basis of a standard curve for product 3 or via substrate
depletion in the case of azidoalcohol 4. Activities were
6883
DOI: 10.1021/acscatal.7b01854
ACS Catal. 2017, 7, 6877−6886

ACS Catalysis
Research Article
calculated in μmol min−1, plotted, and fitted as described above.
Chemical background activities in control reactions without
enzyme were subtracted before fitting.
Whole-Cell Biotransformations and Cascade Reac-
tions. For whole-cell conversions, HheG was heterologously
expressed in E. coli BL21(DE3) pET-28a-hheG as mentioned
above. Cells containing empty pET-28a vector were used as a
negative control. For cyanolysis of epoxide 2, frozen cells were
resuspended in 50 mM Tris·SO4 buffer, pH 8, containing 20
mM 2 and 40 mM NaCN to reach a final OD600 value of 40.
For cascade reactions from chloroalcohol 1 to cyanoalcohol 3
via dehalogenation of 1 and cyanolysis of intermediate 2, frozen
cells were resuspended in 50 mM Tris·SO4 buffer, pH 8,
containing 20 mM 1 and 40 mM NaCN to reach a final OD600
value of 40. All reaction mixtures were incubated at room
temperature with constant stirring (800 rpm). A 500 μL sample
was taken after 1, 3, 24, 48, and 120 h, extracted, and analyzed
using achiral and chiral GC. Product yields were calculated
using a standard curve for cyanoalcohol 3.
For the conversion of (+)-cis-/(+)-trans-5 using whole cells,
reactions were carried out as described but using 20 mM 5 and
40 mM NaN3 as substrate and nucleophile, respectively.
Samples were taken after 1, 2, and 24 h incubation at room
temperature with stirring (800 rpm) and analyzed on achiral
GC after extraction. Conversion (percent) was calculated on
the basis of relative peak areas.
To determine the regioselectivity of HheG in the conversion
of (+)-trans-5, a whole-cell biotransformation was performed
on a preparative scale and the resulting product 6 was purified
to homogeneity. In total, 40 mM (+)-trans-5 (600 mg) was
converted using 80 mM NaN3 in 10 mL of 50 mM Tris·SO4
buffer, pH 8, containing HheG-overexpressing E. coli cells at
OD600 = 40. The reaction mixture was incubated for 24 h at 22
°C. A 500 μL sample was extracted using TBME (0.1% (v/v)
dodecane) and analyzed by GC to determine conversion. For
purification of product 6, the whole reaction mixture was
extracted three times with each 10 mL of dichloromethane
(DCM) and the resulting crude product was purified by silica
gel chromatography with a gradient from 50% to 100% DCM
in pentane. The purified product was obtained as a colorless oil
in 34% yield (212.1 mg).
K50 values for the nucleophiles cyanide (NaCN) and azide
(NaN3) were determined in reactions with a total volume of 1.6
mL in 50 mM Tris·SO4 buffer at a pH of 8 containing 5−150
mM nucleophile, 10 mM 2, and 300 or 156.25 μg mL−1 HheG,
respectively. After 30 min, 90 min, 4 h, and 24 h (NaCN) or 10,
30, 60, and 90 min (NaN3), each 400 μL sample was taken,
extracted as described above, and analyzed by GC. Activities (in
μmol min−1) were calculated, plotted, and fitted as described
above. Chemical background activities in control reactions
without enzyme were subtracted before fitting.
For preparation of a NaCN stock solution, NaCN salt was
dissolved in 10 mM NaOH solution and stored in a closed vial.
High pH of the stock solution is required to prevent formation
of toxic HCN gas. Additionally, all reactions using CN− as
nucleophile were handled in a ventilated fume hood with HCN
gas monitoring present.
Stereoselectivity Determination. Conversion of cyclo-
hexene oxide (2) to the chiral products trans-2-hydroxycyclo-
hexane-1-carbonitrile (3) and 2-azidocyclohexan-1-ol (4) was
followed using achiral and chiral GC. Reactions were performed
using 0.25 mg mL−1 HheG, 20 mM NaCN or NaN3,
respectively, and 5 mM substrate 2 in a total volume of 2
mL of 50 mM Tris·SO4, pH 8. Each 0.5 mL sample was taken
after 0.5, 2, and 24 h and extracted using 500 μL of TBME
containing 0.1% (v/v) dodecane. Conversion (percent) was
calculated on the basis of peak areas from achiral GC analysis,
and product enantiomeric excess (eeP in percent) was
determined by chiral GC analysis.
To distinguish the enantiomers of 3 on chiral GC, an
enzymatic kinetic resolution of 3 was carried out using lipase PS
from Pseudomonas cepacia (Sigma-Aldrich) and vinyl acetate as
an acyl donor as described elsewhere.22,23 A reaction in 3 mL
volume was performed using 100 mg of lipase PS, 100 mM
commercial trans-2-hydroxycyclohexane-1-carbonitrile (3), 200
mM vinyl acetate in TBME at 22 °C, and shaking at 800 rpm.
Samples were taken after 1.5 h and after overnight incubation
and injected on chiral GC.
For determination of the preferentially formed enantiomer of
4, the product obtained in a HheG-catalyzed azidolysis of 2 was
reduced chemically to 2-aminocyclohexan-1-ol, for which
enantiopure 1R,2R and 1S,2S product standards are available
(Sigma-Aldrich). Production of 4 was carried out using 30 mL
of HheG overexpressing E. coli cells with an OD600 value of 40
using 50 mM 2 and 100 mM NaN3 in 50 mM Tris·SO4 buffer,
pH 8. The reaction mixture was incubated at 22 °C for 24 h
and stirred at 800 rpm. Afterward, the reaction mixture was
extracted three times using 30 mL of ethyl acetate. The
combined organic extracts were dried over Na2SO4, and solvent
was removed by evaporation. Approximately 0.75 mmol of 4
were dissolved in a 1/1 solution of methanol and ethyl acetate
for subsequent chemical reduction of azidoalcohol 4. For in situ
production of hydrogen, 1.7 mL of triethylsilane was added to
the mixture. As catalyst, 0.03 g of palladium on carbon was
added and the reduction reaction was carried out for 90 min at
room temperature. Afterward, the catalyst was removed by
centrifugation and the solvent was evaporated under reduced
pressure. The resulting aminoalcohol was dissolved in pyridine,
derivatized using trifluoroacetic anhydride (incubation for 1 h
at 65 °C), and analyzed by chiral GC.
(+)-trans-2-Azido-1-methyl-4-(prop-1-en-2-yl)cyclohexan-
1
1-ol ((+)-trans-6a). [α]21 = +124.8. H NMR (600 MHz,
D
DMSO-d6): δ 4.70−4.68 (m, CH2; 2H), 4.65−4.64 (s, OH;
1H), 3.56−3.55 (m, CH, 1H), 2.05−1.99 (tt, 3J = 11.9, 3.2 Hz,
CH, 1H), 1.94−1.88 (ddd, J = 13.6, 12.0, 3.0 Hz, CH2 1 H),
1.68 (s, CH3, 3 H), 1.68−1.63 (m, CH2, 1H), 1.57−1.49 (ddd,
J = 23.6, 12.0, 4.5 Hz, CH2, 1H) 1.47−1.41 (m, CH2, 2H)
1.41−1.35 (m, CH2, 1H), 1.12 (s, CH3, 3H). 13C NMR (151
MHz, DMSO-d6): δ 149.0 (Cq), 109.1 (CH2), 69.0 (Cq), 66.2
(CH), 37.9 (CH), 33.6 (CH2), 30.2 (CH2), 27.4 (CH3), 25.6
(CH2), 20.8 (CH3). ESI-HRMS: [M + Na+] m/z 218.12648
(calculated [M + Na+] m/z 218.12638).
Analytical Methods. Achiral and chiral GC analyses were
performed on a GC2010 plus gas chromatograph (Shimadzu,
Duisburg, Germany) with FID detection. Achiral separation
was carried out using a OPTIMA 5 MS column (Macherey
Nagel, Duren, Germany) with a length of 30 m, an inner
̈
diameter of 0.25 mm, and a film thickness of 0.25 μm.
Separation of substrates and products was achieved using two
different temperature programs: program 1 (100 °C, 3 min//50
°C min−1//200 °C//20 °C min−1//300 °C, 2.5 min) for
compounds 1−4 and program 2 (80 °C, 1 min//10 °C
min−1//160 °C//20 °C min−1//300 °C) for compounds 5 and
6 with a total flow of 1 mL min−1 and hydrogen as carrier gas.
The substrates eluted at retention times of 4.6 min (1), 3.3 min
6884
DOI: 10.1021/acscatal.7b01854
ACS Catal. 2017, 7, 6877−6886

ACS Catalysis
Research Article
(2), 7.3 min (cis-5), and 7.4 min (trans-5), while the products
eluted at 5.6 min (3), 5.5 min (4), 10.9 min (cis-6), and 11.3
min (trans-6). Chiral separation was carried out using two
different chiral columns, LIPODEX E and HYDRODEX γ-
DIMOM (Macherey-Nagel), each with a length of 25 m, an
inner diameter of 0.25 mm, and a film thickness of 0.25 μm.
Enantiomers were separated on the Lipodex E column using a
temperature program of 80 °C, 5 min//10 °C min−1//110 °C,
5 min//2 °C min−1//180 °C, 5 min with a total flow of 0.8 mL
min−1 and hydrogen as carrier gas. Retention times were as
follows: rac-1 (9.5 and 9.7 min), 2 (2.6 min), (1R,2S)-2-
cyanocyclohexyl acetate (21.6 min), (1S,2S)-4 (17.8 min),
(1R,2R)-4 (18.1 min), (1R,2S)-3 (30.6 min), and (1S,2R)-3
(31.0 min). Enantiomers on the Hydrodex γ-DIMOM column
were separated using a temperature program of 60 °C, 30
min//10 °C min−1//195 °C, 5 min with a total flow of 3.27 mL
min−1 and hydrogen as carrier gas. Retention times were as
follows: (1R,2R)-2-aminocyclohexan-1-ol (26.6 min) and
(1S,2S)-2-aminocyclohexan-1-ol (27.4 min).
GC-MS analysis of (+)-trans-6a was performed on a GCMS-
QP2010SE (Shimadzu) instrument equipped with a ZB-5MS
GUARDIAN column (30 m × 0.25 mm, 0.25 μm film
thickness). The product (+)-trans-6a was dissolved in diethyl
ether. The injector temperature was set to 250 °C and interface
temperature to 280 °C. Helium was used as carrier gas with a
flow of 1.5 mL min−1, and separation was performed with 50
°C, 3 min//12 °C min−1//300 °C, 8 min as the temperature
program. The product was detected at a retention time of 12.6
min with m/z 195 (calculated [M]+ m/z 195).
diameter were cryoprotected by addition of 25% (v/v) glycerol
to their respective mother liquors, mounted on nylon loops,
and flash-frozen in liquid nitrogen. Diffraction data were
collected at 100 K at beamline X06DA (PXIII) at the Swiss
Light Source of the Paul Scherrer Institute on a PILATUS 2M-
F hybrid-pixel detector (DECTRIS Ltd., Baden-Daettwil,
Switzerland).
Indexing and integration of all diffraction data was performed
using XDS.31 For scaling the programs AIMLESS32 of the
CCP4 suite33 and Xscale were used for the native and
anomalous data sets, respectively. Statistics of the data sets are
presented in Table S1 in the Supporting Information.
Structure Determination and Refinement. Initial phases
for HheG could be obtained using SAD and the anomalous
signal of selenium extending to 2.7 Å. Overall, 108 selenium
sites were located using SHELXD34 and, after density
modification and phase extension to 2.3 Å using the native
data set and SHELXE,34 the initial phases were provided to
phenix.autobuild,35 which built approximately 65% (1838
residues) of the 10 HheG copies constituting the asymmetric
unit of the crystal. Further model building and refinement were
carried out by alternating rounds of manual adjustment in
COOT36 and maximum likelihood refinement in phenix.re-
fine37 of the PHENIX software suite.38 Structural flexibility was
modeled using Translation/Libration/Screw refinement.39 In
the last step of refinement, water molecules were attributed to
their nearest TLS group using a script developed by Reichelt
and Blankenfeldt (unpublished). Refinement statistics of the
final model are reported in Table S1 in the Supporting
Information. Final structure validation was done with
MolProbity.40 Diffraction data and coordinates have been
deposited in the Protein Data Bank (PDB entry: 5O30).41 All
structural illustrations were generated using the PyMOL
Mass spectrometric analysis of (+)-trans-6a was carried out
in diethyl ether using ESI-HRMS on a LTQ-Orbitrap Velos
spectrometer (ThermoFisher Scientific, Bremen, Germany).
Localization of the carbon−nitrogen bond (C−N3) in
(+)-trans-6a was determined by NMR (AV ll-600, Bruker,
Billerica, MA, USA) for which different 2D correlation spectra
(1H−1H COSY, 1H−1H HMBC, NOESY, and 1H−15N
HMBC) were determined using deuterated dimethyl sulfoxide
as solvent.
Molecular Graphics System version 1.8.4 (Schrodinger, LLC,
̈
New York City, NY, USA).
Database Search for HheG Homologues. A PHI-
BLAST42 search in the nr and env_nr databases of GenBank
(release 220) was performed using the combination of two
previously reported HHDH sequence motifs: T-X4-(F/Y)-X-G-
A solution of (+)-trans-6a in DCM (c = 10 mg mL−1) was
analyzed with a polarimeter (Propol Digital Automatic
Polarimeter, Dr. Kernchen, Seelze, Germany) to obtain the
specific rotation value.
X
50−150-S-X12-Y-X3-R and the HheG protein sequence (Gen-
Bank accession number AMQ13576.1) as query. A distance
tree of the obtained hits was generated on the basis of the
neighbor joining method and inspected for new members of
the HHDH G-type subfamily.
Crystallization and Data Collection. Crystallization
experiments of N-terminally hexahistidine-tagged HheG wild
type at 14 mg mL−1 and selenomethionine derivatized (SeMet)
His6-HheG at 24 mg mL−1, both in 10 mM Tris·SO4 buffer, pH
8.0, containing 2 mM EDTA and 5 mM β-mercaptoethanol
were carried out via the sitting-drop vapor diffusion method at
room temperature in 96-well INTELLI-PLATES (ARI - Art
Robbins Instruments, Sunnyvale, CA, USA) using a nano-
dispensing robot (Honeybee 963, Genomic Solutions,
Huntingdon, U.K.) and commercial sparse matrix screening
suites. Drops consisting of 200 nL of protein solution and 200
nL of precipitant were equilibrated against a 60 μL reservoir.
Crystals of native HheG were obtained within a few days
from a condition of the WIZARDS I+II screening suite
(Emerald Biosystems Inc., Bainbridge Island, WA, USA),
comprised of 0.1 M sodium acetate pH 4.5 and 1 M
ammonium phosphate dibasic. SeMet His6-HheG crystallized
under conditions comprising 0.1 M imidazole pH 8.0, 0.2 M
lithium sulfate, and 10% (w/v) PEG 3000.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acscatal.7b01854.
■
S
pH and temperature profiles and additional structural
information (PDF)
AUTHOR INFORMATION
Corresponding Author
*A.S.: tel, +49 531 391-55400; fax, +49 531 391-55401; e-mail,
a.schallmey@tu-braunschweig.de.
■
ORCID
Anett Schallmey: 0000-0002-6670-0574
Present Addresses
§Institute of Chemistry and Biotechnology, Competence
Center for Biocatalysis, Zurich University of Applied Sciences,
For both native and derivatized protein, hexagonal prism
shaped crystals of approximately 90 μm length and 150 μm
Einsiedlerstrasse 31, 8820 Wadenswil, Switzerland.
̈
6885
DOI: 10.1021/acscatal.7b01854
ACS Catal. 2017, 7, 6877−6886

ACS Catalysis
Research Article
∥Department of Molecular Biology, Max-Planck-Institute for
Biophysical Chemistry, Max Planck Society, Am Fassberg 11,
(20) Nagasawa, T.; Nakamura, T.; Yu, F.; Watanabe, I.; Yamada, H.
Appl. Microbiol. Biotechnol. 1992, 36, 478−482.
(21) Vaidyanathan, R.; Hesmondhalgh, L.; Hu, S. Org. Process Res.
Dev. 2007, 11, 903−906.
Gottingen 37077, Germany.
̈
Notes
(22) Forro,
Asymmetry 1997, 8, 3095−3099.
(23) Utczas, M.; Szekely, E.; Forro,
Siman
́
E.; Lundell, K.; Fulop, F.; Kanerva, L. T. Tetrahedron:
̈
̈
The authors declare no competing financial interest.
́
́ ̋ ̋
E.; Szollosy, A.; Fulop, F.;
̈
̈
́
́
ACKNOWLEDGMENTS
́
di, B. Tetrahedron Lett. 2011, 52, 3916−3918.
■
(24) Yao, P.; Wang, L.; Yuan, J.; Cheng, L.; Jia, R.; Xie, M.; Feng, J.;
We thank Dr. Marcus Schallmey for technical assistance with
chiral GC analysis and suggestions on the manuscript.
Furthermore, we thank Dr. Rainer Wardenga from Enzymicals
AG (Greifswald, Germany) for scientific collaboration and
discussion of the biocatalytic data. The authors also thank the
staff at beamline PXIII (X06DA) of the Swiss Light Source
(PSI, Villingen, Switzerland), in particular Dr. Meitian Wang,
for beamline access and support. This work was financially
supported by the German Federal Ministry for Economic
Affairs and Energy (BMWi) within the AIF-ZIM funding
scheme (award numbers KF3041401 and KF2584302).
Wang, M.; Wu, Q.; Zhu, D. ChemCatChem 2015, 7, 1438−1444.
(25) Schallmey, M.; Jekel, P.; Tang, L.; Majeric
́
Elenkov, M.;
Hoffken, H. W.; Hauer, B.; Janssen, D. B. Enzyme Microb. Technol.
̈
2015, 70, 50−57.
(26) Baker, N. A.; Sept, D.; Joseph, S.; Holst, M. J.; McCammon, J.
A. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 10037−10041.
(27) de Jong, R. M.; Tiesinga, J. J. W.; Rozeboom, H. J.; Kalk, K. H.;
Tang, L.; Janssen, D. B.; Dijkstra, B. W. EMBO J. 2003, 22, 4933−
4944.
(28) Watanabe, F.; Yu, F.; Ohtaki, A.; Yamanaka, Y.; Noguchi, K.;
Yohda, M.; Odaka, M. Proteins: Struct., Funct., Genet. 2015, 83, 2230−
2239.
(29) Oliveira, S. H. P.; Ferraz, F. A. N.; Honorato, R. V.; Xavier-Neto,
J.; Sobreira, T. J. P.; de Oliveira, P. S. L. BMC Bioinf. 2014, 15, 197−
204.
(30) Haak, R. M.; Tarabiono, C.; Janssen, D. B.; Minnaard, A. J.; de
Vries, J. G.; Feringa, B. L. Org. Biomol. Chem. 2007, 5, 318−323.
(31) Kabsch, W. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2010, 66,
133−144.
(32) Evans, P. R.; Murshudov, G. N. Acta Crystallogr., Sect. D: Biol.
Crystallogr. 2013, 69, 1204−1214.
(33) Winn, M. D.; Ballard, C. C.; Cowtan, K. D.; Dodson, E. J.;
Emsley, P.; Evans, P. R.; Keegan, R. M.; Krissinel, E. B.; Leslie, A. G.
W.; McCoy, A.; McNicholas, S. J.; Murshudov, G. N.; Pannu, N. S.;
Potterton, E. A.; Powell, H. R.; Read, R. J.; Vagin, A.; Wilson, K. S.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 2011, 67, 235−242.
(34) Sheldrick, G. M. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2010,
66, 479−485.
REFERENCES
■
(1) Schallmey, A.; Schallmey, M. Appl. Microbiol. Biotechnol. 2016,
100, 7827−7839.
(2) Castro, C. E.; Bartnicki, E. W. Biochemistry 1968, 7, 3213−3218.
(3) Ma, S. K.; Gruber, J.; Davis, C.; Newman, L.; Gray, D.; Wang, A.;
Grate, J.; Huisman, G. W.; Sheldon, R. A. Green Chem. 2010, 12, 81−
86.
(4) Fox, R. J.; Davis, S. C.; Mundorff, E. C.; Newman, L. M.;
Gavrilovic, V.; Ma, S. K.; Chung, L. M.; Ching, C.; Tam, S.; Muley, S.;
Grate, J.; Gruber, J.; Whitman, J. C.; Sheldon, R. A.; Huisman, G. W.
Nat. Biotechnol. 2007, 25, 338−344.
(5) Lutje Spelberg, J. H.; Tang, L.; Kellogg, R. M.; Janssen, D. B.
Tetrahedron: Asymmetry 2004, 15, 1095−1102.
(6) Elenkov, M. M.; Tang, L.; Meetsma, A.; Hauer, B.; Janssen, D. B.
Org. Lett. 2008, 10, 2417−2420.
(7) Mikleuse
Janssen, D. B.; Majeric
̌
vic,
́
A.; Hamersa
̌
k, Z.; Salopek-Sondi, B.; Tang, L.;
Elenkov, M. Adv. Synth. Catal. 2015, 357,
(35) Terwilliger, T. C.; Grosse-Kunstleve, R. W.; Afonine, P. V.;
Moriarty, N. W.; Zwart, P. H.; Hung, L. W.; Read, R. J.; Adams, P. D.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 2008, 64, 61−69.
(36) Emsley, P.; Lohkamp, B.; Scott, W. G.; Cowtan, K. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 2010, 66, 486−501.
(37) Afonine, P. V.; Grosse-Kunstleve, R. W.; Echols, N.; Headd, J. J.;
Moriarty, N. W.; Mustyakimov, M.; Terwilliger, T. C.; Urzhumtsev, A.;
Zwart, P. H.; Adams, P. D. Acta Crystallogr., Sect. D: Biol. Crystallogr.
2012, 68, 352−367.
́
1709−1714.
(8) Elenkov, M. M.; Hoeffken, H. W.; Tang, L.; Hauer, B.; Janssen,
D. B. Adv. Synth. Catal. 2007, 349, 2279−2285.
(9) Fuchs, M.; Simeo, Y.; Ueberbacher, B. T.; Mautner, B.; Netscher,
T.; Faber, K. Eur. J. Org. Chem. 2009, 2009, 833−840.
(10) van Hylckama Vlieg, J. E. T.; Tang, L.; Spelberg, J. H. L.; Smilda,
T.; Poelarends, G. J.; Bosma, T.; van Merode, A. E. J.; Fraaije, M. W.;
Janssen, D. B. J. Bacteriol. 2001, 183, 5058−5066.
́
(38) Adams, P. D.; Afonine, P. V.; Bunkoczi, G.; Chen, V. B.; Davis,
(11) Schallmey, M.; Koopmeiners, J.; Wells, E.; Wardenga, R.;
Schallmey, A. Appl. Environ. Microbiol. 2014, 80, 7303−7315.
(12) Koopmeiners, J.; Halmschlag, B.; Schallmey, M.; Schallmey, A.
Appl. Microbiol. Biotechnol. 2016, 100, 7517−7527.
(13) Elenkov, M. M.; Hauer, B.; Janssen, D. B. Adv. Synth. Catal.
2006, 348, 579−585.
I. W.; Echols, N.; Headd, J. J.; Hung, L.-W.; Kapral, G. J.; Grosse-
Kunstleve, R. W.; McCoy, A. J.; Moriarty, N. W.; Oeffner, R.; Read, R.
J.; Richardson, D. C.; Richardson, J. S.; Terwilliger, T. C.; Zwart, P. H.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 2010, 66, 213−221.
(39) Schomaker, V.; Trueblood, K. N. Acta Crystallogr., Sect. B: Struct.
Crystallogr. Cryst. Chem. 1968, 24, 63−76.
(14) Hu, L. Y.; Du, D.; Hoffman, J.; Smith, Y.; Fedij, V.; Kostlan, C.;
Johnson, T. R.; Huang, Y.; Kesten, S.; Harter, W.; Yue, W. S.; Li, J. J.;
Barvian, N.; Mitchell, L.; Lei, H. J.; Lefker, B.; Carroll, M.; Dettling,
D.; Krieger-Burke, T.; Samas, B.; Yalamanchili, R.; Lapham, K.;
Pocalyko, D.; Sliskovic, D.; Ciotti, S.; Stoller, B.; Hena, M. A.; Ding,
Q.; Maiti, S. N.; Stier, M.; Welgus, H. Bioorg. Med. Chem. Lett. 2007,
17, 5983−5988.
(40) Chen, V. B.; Arendall, W. B.; Headd, J. J.; Keedy, D. A.;
Immormino, R. M.; Kapral, G. J.; Murray, L. W.; Richardson, J. S.;
Richardson, D. C. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2010, 66,
12−21.
(41) Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T.
N.; Weissig, H.; Shindyalov, I. N.; Bourne, P. E. Nucleic Acids Res.
2000, 28, 235−242.
́
(15) Hasnaoui-Dijoux, G.; Majeric Elenkov, M.; Lutje Spelberg, J. H.;
(42) Zhang, Z.; Schaffer, A. A.; Miller, W.; Madden, T. L.; Lipman,
̈
Hauer, B.; Janssen, D. B. ChemBioChem 2008, 9, 1048−1051.
(16) Tang, L.; Lutje Spelberg, J. H.; Fraaije, M. W.; Janssen, D. B.
Biochemistry 2003, 42, 5378−5386.
D. J.; Koonin, E. V.; Altschul, S. F. Nucleic Acids Res. 1998, 26, 3986−
3990.
(17) You, Z.-Y.; Liu, Z.-Q.; Zheng, Y.-G. Appl. Microbiol. Biotechnol.
2013, 97, 9−21.
(18) Yu, F.; Nakamura, T.; Mizunashi, W.; Watanabe, I. Biosci.,
Biotechnol., Biochem. 1994, 58, 1451−1457.
(19) Nakamura, T.; Nagasawa, T.; Yu, F.; Watanabe, I.; Yamada, H.
Appl. Environ. Microbiol. 1994, 60, 1297−1301.
6886
DOI: 10.1021/acscatal.7b01854
ACS Catal. 2017, 7, 6877−6886