Azidolysis of epoxides catalysed by the halohydrin dehalogenase from Arthrobacter sp. AD2 and a mutant with enhanced enantioselectivity: an (S)-selective HHDH

Article abstract of DOI:10.1016/j.tetasy.2016.08.003

Halohydrin dehalogenase from Arthrobacter sp. AD2 catalysed azidolysis of epoxides with high regioselectivity and low to moderate (S)-enantioselectivity (E?=?1–16). Mutation of the asparagine 178 to alanine (N178A) showed increased enantioselectivity towards styrene oxide derivatives and glycidyl ethers. Conversion of aromatic epoxides was catalysed by HheA-N178A with complete enantioselectivity, however the regioselectivity was reduced. As a result of the enzyme-catalysed reaction, enantiomerically pure (S)-β-azido alcohols and (R)-α-azido alcohols (ee???99%) were obtained.

Full text of DOI:10.1016/j.tetasy.2016.08.003

Tetrahedron: Asymmetry 27 (2016) 930–935
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Azidolysis of epoxides catalysed by the halohydrin dehalogenase
from Arthrobacter sp. AD2 and a mutant with enhanced
enantioselectivity: an (S)-selective HHDH
a
b
a
c
´
ˇ ˇ b
Ana Mikleuševic , Ines Primozic , Tomica Hrenar , Branka Salopek-Sondi , Lixia Tang ,
a,
⇑
´
Maja Majeric Elenkov
a
ˇ
´
Ruder Boškovic Institute, Bijenicka c. 54, Zagreb 10000, Croatia
b Faculty of Science, University of Zagreb, Department of Chemistry, Zagreb, Croatia
c University of Electronic Science and Technology, No. 4, Section 2, North Jianshe Road, Chengdu, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Halohydrin dehalogenase from Arthrobacter sp. AD2 catalysed azidolysis of epoxides with high regiose-
lectivity and low to moderate (S)-enantioselectivity (E = 1–16). Mutation of the asparagine 178 to alanine
(N178A) showed increased enantioselectivity towards styrene oxide derivatives and glycidyl ethers.
Conversion of aromatic epoxides was catalysed by HheA-N178A with complete enantioselectivity, how-
ever the regioselectivity was reduced. As a result of the enzyme-catalysed reaction, enantiomerically pure
Received 1 June 2016
Accepted 2 August 2016
Available online 28 August 2016
(S)-b-azido alcohols and (R)-
a
-azido alcohols (ee P 99%) were obtained.
Ó 2016 Published by Elsevier Ltd.
1. Introduction
series of structurally different aliphatic epoxides was catalysed
by low to moderate (S)-enantioselectivity and high
regioselectivity.8 HheA was found to catalyse highly regioselective
azidolysis of spiroepoxides again with low enantioselectivity.9
Based on the aforementioned research, HheA can be considered
as a regioselective enzyme, displaying a broad substrate range
and low enantioselectivity, slightly preferring (S)-enantiomer.
Since the only known enantioselective wild-type (WT) HHDH is
the HheC of (R)-stereopreference, an (S)-selective enzyme would
be of great interest for application in asymmetric synthesis.
Recently, several HheA mutants have been successfully evolved
by semirational design.10 Among them, the N178A variant dis-
played an outstanding improvement in the kinetic resolution of
rac-2-chloro-1-phenylethanol (E > 200) compared to HheA-WT
(E = 1.7).
Halohydrin dehalogenases (HHDHs) are attractive enzymes due
to their catalytic promiscuity in catalysis of a number of non-
natural reactions.1 Their natural role is metabolism of vicinal
halohydrines to form epoxides. In the reverse reaction, epoxide
ring-opening, several non-natural nucleophiles are accepted.2
HHDH-catalysed conversions have been studied since the late
1960s, however the number and variety of known enzymes is still
limited.3 To date, only bacterial HHDHs have been isolated and
characterised. Among them, the best studied is the enzyme from
Agrobacterium radiobacter AD1 (HheC).4 It is a homotetrameric pro-
tein with 28 kDa subunits and displays high (R)-enantioselectivity
towards 2,2-disubstituted and aromatic epoxides.5 HHDH pro-
duced by Arthrobacter sp. AD2 (HheA) has a 33% amino acid
sequence similarity and a similar tertiary and quaternary structure
but a much more open active site compared to HheC.6 The low
sequence identities and different active site conformations suggest
that enzymes HheA and HheC are considerably different. Although
isolated and characterized in 1991,7 the biocatalytic potential of
HheA has not been broadly studied, mostly due to its apparently
low enantioselectivity. A low E-value was found for the conversion
of para-nitro-2-bromo-1-phenylethanol to epoxide (E < 3), with
the slight preference towards (S)-enantiomer.4b Cyanolysis of a
In order to gain more insight into the enantioselectivity of
HheA-WT and the mutant N178A in the azidolysis reaction, a vari-
ety of epoxides was investigated. Molecular docking analyses were
performed to understand the structural basis of the enantioselec-
tivity and regioselectivity of mutant N178A.
2. Results and discussion
2.1. Enantioselectivity of HheA-WT and N178A mutant
Initially, a set of structurally different epoxides 1a–1k was cho-
sen to test the enantioselectivity of HheA in the azide mediated
⇑
Corresponding author.
http://dx.doi.org/10.1016/j.tetasy.2016.08.003
0957-4166/Ó 2016 Published by Elsevier Ltd.

´
931
A. Mikleuševic et al. / Tetrahedron: Asymmetry 27 (2016) 930–935
ring-opening reaction (Table 1). The kinetic resolution experiments
were performed by using crude extract containing recombinant
enzyme prepared from Escherichia coli cells. Enzymatic reactions
were performed in Tris–SO4 buffer (pH 7.0) containing 0.5–2%
DMSO, NaN3 (5 mM) and epoxide (5 mM), except 1k was used in
2 mM concentration due to low aqueous solubility. Reactions were
monitored by periodically taking samples from the reaction mix-
ture, followed by extraction and GC or HPLC analysis. HheA
showed a broad substrate range by converting all tested epoxides
into azido alcohols. The nucleophile attacks at the less hindered
carbon of the epoxide moiety resulting in regiospecific reaction.
All reactions occurred with low to moderate enantioselectivity
(E = 1–16) and a slight (S)-preference for the majority of the sub-
strates (Table 1). These results confirmed previous observations
of HheA as a biocatalyst displaying poor enantioselectivity and
high regioselectivity.4b,8,9
Recently, a saturation mutagenesis library was constructed
based on the knowledge of the X-ray structure of HheA and
rational considerations.10 Residues L141, V136 and N178 located
in the two active-site loops were targeted, and resulted in several
active mutants with improved properties. The N178A variant,
Table 1
Kinetic resolution of epoxides 1a–1k catalysed by HheA
O
OH
R1
R2
HheA
Buffer
R1
R2
+
NaN3
N3
which displayed
a remarkable improvement in the kinetic
1a 1k
-
2a 2k
-
resolution of rac-2-chloro-1-phenylethanol (E > 200) with (S)-
stereopreference, was selected for a more detailed study. For this,
a small set of substrates 1b and 1g–1k that represent different
classes of epoxide was chosen to investigate the enantioselectivity
of HheA-N178A in the ring-opening reaction with azide (Table 2).
Reaction of aliphatic 1b and carboxymethyl 1g substituted
epoxides occurred with no difference in enantioselectivity. A sig-
nificant increase was observed with glycidyl ethers 1h and 1i, as
well as for previously reported styrene oxide 1k.10 For 1h the E-
value increased from 1 to 48, for 1i from 1.5 to 20 and for 1k from
5 to 108. A drop of enantioselectivity was observed only with 1j.
Styrene oxide derivatives and glycidyl ethers were azidolysed with
up to 48-fold higher enantioselectivity, which exposed this class of
epoxides as favourable substrates for N178A mutant. To further
explore this enzyme, a series of aromatic epoxides was tested
and results are represented in Table 3. Racemic substrates used
for this screening were commercially available or synthesised
according to literature procedure (see Experimental section).
The mutant enzyme showed high enantioselectivity towards
substituted styrene oxides 1k–1o and the same enantiopreference
for (S)-epoxides. Conversion proceeded rapidly to 50%, but after
that it drastically slowed down, resulting in efficient kinetic reso-
lution and formation of enantiomerically pure secondary alcohols
2 (ee P 99%). The only exception was pyridine derivative 1p that
was converted with moderate enantioselectivity and lower rate
compared to phenyl-substituted epoxides (Table 3, entry 8). These
results clearly confirm that replacement of the active site residue
Asn178 by alanine is crucial in controlling the enantioselectivity
of HheA. Besides the fact that this mutation beneficially influences
the enantioselectivity and activity,10 it has however a detrimental
effect on the regioselectivity. While HheA-WT shows high prefer-
ential attack at the terminal position, mutant N178A catalyses
Substrate
R1
R2
E-Valuea
Config.b
1a
1b
1c
1d
1e
1f
1g
1h
1i
Et
n-Bu
i-Pr
t-Bu
c–Hex
n-Pen
CH2CO2Me
CH2OCH2CH@CH2
CH2OPh
CH2Ph
Ph
H
H
H
H
H
Me
H
H
H
H
H
1.5
9
7
10
2
2
1
1
2
16
5
(S)
(S)
(S)
(S)
(S)
(R)
(R)
(R)c
(S)c
(S)
(S)
1j
1k
a
b
c
E-Values were calculated from eep and ees.
Absolute configuration of the faster reacting enantiomer.
Reversed absolute configuration due to the Cahn–Ingold–Prelog rule.
Table 2
Enantioselectivity of HheA-WT and HheA-N178A variant in the ring-opening reaction
with azide
Substrate
HheA-WT
HheA-N178A
E-Valuea,b
E-Valuea,b
1b
1g
1h
1i
1j
1k
10 (S)
1 (R)
1 (R)
10 (S)
1 (R)
48 (R)c
20 (R)c
6 (S)
1.5 (S)c
16 (S)c
5 (S)d
108 (S)d
a
E-Values were calculated from eep and ees.
Absolute configuration of the faster reacting epoxide is indicated between
brackets.
b
c
Reversed absolute configuration due to the CPI rule.
Data from Ref. 10.
d
Table 3
Conversion of epoxides 1k–1p catalysed by HheA-N187Aa
O
OH
N3
HheA N178A
buffer
+
NaN3
N3
OH
+
(R)
Ar
(S)
Ar
Ar
1k-1p
2k-2p
3k-3p
Entry
Substrate
Ar
t (h)
Conv. (%)
ee 1 (%)
ee 2 (%)
ee 3 (%)
Ratio (2:3)
1
2
3
4
5
6
7
8
rac-1k
rac-1l
Ph
2
45
50
50
100
0
51
52
35
72 (R)
>99 (R)
>99 (R)
/
99 (S)
>99 (S)
>99 (S)
100 (S)
/
>99 (S)
>99 (S)
63 (R)c
91 (R)
>99 (R)
>99 (R)
100 (R)
/
57:43
69:31
70:30
70:30
/
81:19
94:6
100:0
4-Cl-Ph
4-Br-Ph
4-Br-Ph
4-Br-Ph
4-CN-Ph
4-NO2-Ph
2-Py
0.75
0.75
0.75
1
2
1.5
rac-1m
(S)-1m
(R)-1m
rac-1n
rac-1o
rac-1p
/
85 (R)
98 (R)
43 (R)
n.d.b
n.d.b
/
20
a
b
c
General conditions: 2 mM epoxide in Tris–SO4, 2 mM NaN3, cell-free extract.
Not determined.
Apparent inversion of configuration of epoxides (S)-1p to alcohols (R)-2p is due to a different substituent priority according to the CIP rule.

´
A. Mikleuševic et al. / Tetrahedron: Asymmetry 27 (2016) 930–935
932
formation of both regioisomers 2 and 3. The less hindered carbon
atom is predominantly attacked, however, the /b ratio 2:3
structures with the lowest energy. A model of the enzyme active
site was built by selecting the most important amino acids within
10 Å from the catalytic triad (positions of their atoms fixed). The
position of amino acids and the azide ion were frozen during the
optimizations. Optimized geometries obtained for enantiomers of
1k and 1m are presented at the Figures 1 and 2, respectively.
The epoxide oxygen atom of (R)-1k is positioned in such a way
that can form two strong hydrogen bonds (distances below 3.0 Å)
with Tyr147 and Ser134 whereas (S)-enantiomer forms weaker
hydrogen bonds (Fig. 1).
a
depends on the para-substituent. The highest regioisomeric ratio
was observed with epoxides bearing electron-withdrawing sub-
stituents 1n–1p, while the lowest was for styrene oxide 1k
(57:43). Nevertheless, the formation of primary alcohols 3 obvi-
ously is enzyme catalysed, since they were isolated in highly enan-
tioenriched form (up to >99% ee). Not surprisingly, the
configuration on the stereogenic centre was found to be (R). In
order to confirm the stereochemical course of the nucleophilic
a-
attack and formation of (R)-3, enzymatic reactions were performed
with pure enantiomers. For these experiments epoxide 1m was
chosen as substrate. Reaction of racemic 1m showed formation
of the mixture of secondary (S)-2m and primary (R)-3m alcohols
in the ratio of 70:30 (Table 3, entry 3). Enantiomerically pure (R)-
1m and (S)-1m were obtained from racemic 1m by preparative
HPLC and subjected to enzymatic reaction. (S)-1m was completely
converted with identical stereochemical pattern as racemic com-
pound (RS)-1m and both products were formed in the ratio
2m:3m = 70:30, equal to reaction with racemic substrate (Table 3,
entry 4). From the other hand, no reaction occurred with (R)-enan-
tiomer (Table 3, entry 5). These results confirm that the enzyme-
catalysed reaction is highly enantioselective, and nucleophilic b-
attack occurs at the sterically less hindered C3 atom of (S)-enan-
tiomer with the retention of the absolute configuration, while
Although the epoxide carbon atoms are in the vicinity of the
nucleophile, those positions are geometrically unfavourable for
the nucleophilic attack to occur. On the other hand, for (S)-1k,
additional complex was found (Fig. 1, thinner grey model). The aro-
matic ring is positioned in a similar location as the previous one,
but the epoxide ring is rotated and its oxygen atom pointed
towards Ser135, making two strong H-bonds with Ser134 and
Ser135. This binding is geometrically in a correct position for the
nucleophilic attack to occur since azide ion is located ca. 3.3 Å from
Cb. Cb hydrogen atom is positioned the vicinity of the Ca hydrogen
atom of Ala178 (H–H 1.7 Å). However, this is a non-productive
complex, suggesting that a conformational change at the active site
of HheA might occur.
Furthermore, to evaluate the impact of a para-substituent on
the aromatic ring, docking and subsequent ONIOM calculations
of 1m enantiomers to the mutant HheA-N178A were performed.
Overlay of the obtained optimized geometries for (R)- and (S)-
enantiomers of 1m in the active site are presented in Figure 2.
Similar binding patterns to that of 1k can be observed. Again,
the most productive binding (Fig. 2, thinner grey model) for the
reaction to occur is that in which (S)-epoxide ring is rotated mak-
ing two stronger H-bonds with Ser134 and Ser135. Azide ion is
close and geometrically in the best position for the nucleophilic
the
a-attack (C2 atom) is accompanied by inversion of configura-
tion. As a result, (R)-configured azido alcohol 3 is formed from
the (S)-epoxide.
2.2. Docking studies and quantum chemical calculations
To gain insight into how the single mutation at the residue 178
results in such a drastic change of enantioselectivity and regiose-
lectivity for the azidolysis of para-substituted styrene oxides, in sil-
ico mutagenesis of N178 to A and docking of 1k and 1m
enantiomers to the mutant HheA-N178A were performed. To
investigate the electronic effects of the important interactions
within the active site, a hybrid approach combining quantum
chemical (B3LYP/6-311G(d)) and semiempirical (PM6) calculations
within the ONIOM scheme were carried out for the docked
attack. The distance from nucleophile to the epoxide C
a atom is
ca. 0.6 Å longer than that to Cb resulting in a higher regioselectivity
of the attack. Due to the bigger size of the aromatic substituent
(para-bromine), a slight shift in the positions of the aromatic group
and epoxide ring compared to binding of equivalent 1k-enan-
tiomers is observed. Exchange of Asn178 to Ala increased the
conformational flexibility of Ser135 hydroxyl group. The strong
Figure 2. Overlay of 1m enantiomers ((R)-white, (S)-grey, thicker and thinner
model, bromine atom represented with dark red colour) and the azide ion in the
HheA-N178A active site model represented by some structurally important amino
acids obtained by ONIOM calculations. Hydrogen bonds are indicated as green lines,
distances given in Å.
Figure 1. Overlay of 1k enantiomers ((R)-white, (S)-grey thicker and thinner
model) and the azide ion in the HheA-N178A active site model represented by some
structurally important amino acids obtained by ONIOM calculations. Hydrogen
bonds of (S)-enantiomers are indicated as green lines, distances given in Å.

´
933
A. Mikleuševic et al. / Tetrahedron: Asymmetry 27 (2016) 930–935
H-bond (ca. 3 Å) between Asn amide nitrogen atom and Ser OH
group, present in the HheA-WT, vanished in the mutant. Loss of
that H-bond at the same time increased potency of Ser OH group
to act as an H-bond donor towards the (S)-enantiomer of the sub-
strate. Since the mutation dramatically changed the reaction out-
come, it can be assumed that the observed enantio- and
regioselectivity of reactions are also affected by the possible
greater conformational adjustment of the protein not included in
present quantum chemical study.10
After stirring for 20 min at room temperature, a solution of 2-hep-
tanone (2.0 g, 17.8 mmol) in DMSO (10 mL) was added dropwise.
Reaction mixture was stirred for 22 h. Water was added (20 mL)
and the mixture extracted with diethyl ether (3 ꢀ 30 mL). The
combined organic layers were washed with water (2 ꢀ 30 mL)
and brine (20 mL), dried over Na2SO4, filtered and evaporated
under reduced pressure. Column chromatography (SiO2; hexane–
ethyl acetate, 9.5:0.5) furnished the pure epoxide 1f (1.1 g, 48%)
as a colourless liquid. dH (300 MHz, CDCl3) 0.87–0.91 (3H, m),
1.27–1.34 (11H, m), 2.57 (1H, d, J 5.0 Hz), 2.60 (1H, d, J 5.0 Hz);
dC (150 MHz, CDCl3) 14.0, 20.9, 22.6, 24.9, 31.8, 36.7, 53.9, 57.1.
3. Conclusions
In conclusion, it was shown that halohydrin dehalogenase HheA
catalyzed ring-opening reaction of various structurally diverse
epoxides with azide as the nucleophile. This reaction proceeded
with high regioselectivity and low to moderate enantioselectivity.
HheA-N178A mutant showed enhanced enantioselectivity and
reduced regioselectivity towards the azidolysis of styrene oxide
derivatives. All para-substituted styrene oxides could be resolved
in excellent enantioselectivities and enantiomerically pure (S)-b-
4.2.2. 4-Bromostyrene oxide 1m
Epoxide 1m was prepared according to a literature procedure.13
4-Bromobenzaldehyde (1.0 g, 5.4 mmol) and trimethylsulfonium
iodide (1.76 g, 8.6 mmol) were dissolved in DMSO (8 mL). Potas-
sium tert-butoxide (0.92 g, 8.2 mmol) was dissolved in 8 mL of
DMSO and added to the reaction mixture. Reaction mixture was
stirred over night at room temperature under argon. Reaction
was quenched by addition of water (16 mL). Reaction mixture
was extracted with dichloromethane (3 ꢀ 30 mL), combined
organic layers were dried over Na2SO4, filtered and evaporated
under reduced pressure. Column chromatography (SiO2; hexane–
ethyl acetate, 9:1) furnished the pure epoxide 1m (0.79 g, 74%)
as a colourless liquid. dH (300 MHz, CDCl3) 2.74 (1H, dd, J1 5.5 Hz,
J2 2.5 Hz), 3.13 (1H, dd, J1 5.5 Hz, J2 4.0 Hz), 3.82 (1H, dd, J1
4.0 Hz, J2 2.5 Hz), 7.14 (2H, d, J 8.5 Hz), 7.46 (2H, d, J 8.5 Hz); dC
(75 MHz, CDCl3) 51.2, 51.8, 122.0, 127.2, 131.7, 136.8.
azido alcohols and (R)-a-azido alcohols could be obtained. It is
remarkable that a single mutation results in such a dramatic
change in the enantioselectivity. It is clear that residue 178 in
the halide binding pocket plays a critical role in determining the
enantioselectivity of these epoxide azidolysis. HheA-N178A repre-
sents a first highly (S)-enantioselective HHDH. With the molecular
modelling studies, binding modes of enantiomers and the impor-
tant interactions within the enzyme active site were determined.
4.2.3. 4-Cyanostyrene oxide 1n
4. Experimental
4.1. General
Epoxide 1n was prepared according to a modified literature
procedure.14 Sodium borohydride (106 mg, 2.7 mmol) was added
to a solution of 2-bromo-1-(4-cyanophenyl)-ethanone (568 mg,
2.54 mmol) in methanol (7 mL) and stirred at 0 °C. After 1.5 h,
water (10 mL) was added and reaction mixture was stirred for
1 h at room temperature. Reaction mixture was extracted with
dichloromethane (3 ꢀ 10 mL), combined organic layers were dried
over Na2SO4, filtered and evaporated under reduced pressure. The
dry residue was dissolved in a small amount of diethyl ether, fol-
lowing by addition of potassium hydroxide (15 mL, 1 M) and the
reaction mixture was stirred over night at room temperature.
Reaction mixture was extracted with dichloromethane
(4 ꢀ 10 mL), combined organic layers were washed with brine,
dried over Na2SO4, filtered and evaporated under reduced pressure.
Column chromatography (SiO2; hexane–ethyl acetate, 7:3) fur-
nished the pure epoxide 1n (289 mg, 78%) as a colourless liquid.
dH (300 MHz, CDCl3) 2.74 (1H, dd, J1 5.5 Hz, J2 2.5 Hz), 3.18 (1H,
dd, J1 5.5 Hz, J2 4.0 Hz), 3.88–3.90 (1H, m), 7.37 (2H, d, J 8.5 Hz),
8.62 (2H, d, J 8.5 Hz); dC (75 MHz, CDCl3) 51.5, 51.6, 112.0, 118.6,
126.1, 132.3, 143.3.
The commercial grade reagents and solvents were used without
further purification. The commercially available racemic substrates
1,2-epoxybutane 1a, 1,2-epoxyhexane 1b, allyl glycidyl ether 1h,
2,3-epoxypropyl-benzene 1j and styrene oxide 1k, as well as azi-
dotrimethylsilane, (R,R)-N,N0-bis(3,5-di-tertbutylsalicylidene)-1,2-
cyclohexanediaminochromium(III) chloride, trimethylsulfoxonium
iodide, absolute DMSO, L-arabinose, ampicillin sodium salt, b-mer-
captoethanol, sorbitol, and EDTA were purchased from Sigma–
Aldrich. Racemic 1,2-epoxy-3-methylbutane 1c, 3,3-dimethyl-1,2-
epoxybutane 1d, 1,2-epoxy-3-phenoxypropane 1i and 4-chloros-
tyrene oxide 1l were purchased from Alfa Aesar. Bacto-tryptone,
yeast extract and bacto-agar were purchased from Difco. Complete
Protease Inhibitor Cocktail Tablets were supplied by Roche, while
glycerol was obtained from Kemika. Rac-2-cyclohexyl-oxirane 1e8
and rac-methyl-3,4-epoxybutyrate 1g11 were prepared as previ-
ously described. Enantiomerically pure (R)-1m and (S)-1m were
obtained from racemic 1m by preparative HPLC using
a
semipreparative Chiralpak AS column (250 ꢀ 46 mm, Daicel) with
0.5% 2-PrOH/Hexane as eluent (4 mL/min). Racemic azido alcohols
were prepared by ring-opening reactions of the corresponding
epoxide with sodium azide in water.12 Enzymes, HheA-WT and
mutant HheA-N178A were prepared by overexpression in E. coli
strain MC1061 according to a previously described protocol and
used as cell-free extract.9,10
4.2.4. 4-Nitrostyrene oxide 1o
Epoxide 1o was prepared according to a literature procedure.14
Sodium borohydride (85 mg, 2.2 mmol) was added to a solution of
2-bromo-1-(4-nitrophenyl)-ethanone (500 mg, 2 mmol) in metha-
nol/THF (1:1.2, 5 mL) at 0 °C. After that, solution of sodium hydrox-
ide (3 mL, 2 M) was added and reaction mixture was stirred for 1 h
at room temperature. The solvent was evaporated, acetic acid was
added (to pH 4) and reaction mixture was extracted with dichlor-
omethane (4 ꢀ 10 mL), combined organic layers were washed with
solution of sodium hydrogen carbonate, water, brine, dried over
Na2SO4, filtered and evaporated under reduced pressure. The dry
residue was dissolved in a small amount of diethyl ether, potas-
sium hydroxide (15 mL, 1 M) was added and reaction mixture
was stirred over night at room temperature. Reaction mixture
4.2. Synthesis of racemic substrates
4.2.1. 1,2-Epoxy-2-methylheptane 1f
Epoxide 1f was prepared according to a literature procedure.13
DMSO (20 mL) was added dropwise to a mixture of trimethylsul-
foxonium iodide (5.1 g, 23.1 mmol) and sodium hydride (a 60%
disp. in mineral oil; 0.94 g, 23.1 mmol) cooled to 0 °C, under argon.

´
A. Mikleuševic et al. / Tetrahedron: Asymmetry 27 (2016) 930–935
934
was extracted with dichloromethane (4 ꢀ 10 mL), combined
organic layers were washed with brine, dried over Na2SO4, filtered
and evaporated under reduced pressure. Column chromatography
(SiO2; hexane–ethyl acetate, 7:3) furnished the pure epoxide 1o
(0.30 g, 90%) as a yellow powder. dH (300 MHz, CDCl3) 2.78 (1H,
dd, J1 5.5 Hz, J2 2.5 Hz), 3.19 (1H, dd, J1 5.5 Hz, J2 4.0 Hz), 3.96
(1H, dd, J1 4.0 Hz, J2 2.5 Hz), 7.42 (2H, d, J 8.5 Hz), 8.18 (2H, d, J
8.5 Hz); dC (150 MHz, CDCl3) 51.4, 51.6, 123.8, 126.2, 145.3, 147.8.
from Gaussian 09 program package.20 All energy values from single
point calculations were arranged in the 7-way array and search for
all local minima was performed by using combinatorial algorithm
built in program for multivariate analysis moonee.21
All these local minima were subjected to geometry optimization
procedure using the semiempirical PM6 method and subsequent
clustering of geometries was performed and classified on the basis
of the energy values. Results were inspected visually and on the
basis of the energy values some structure were selected and reop-
timized using the QM/QM 2-layer ONIOM approach with semiem-
pirical PM6 method for the outer layer, and density functional
theory B3LYP/6-31G(d) method, for the inner layer of system.22,23
4.2.5. 2-Pyridyl-oxirane 1p
Epoxide 1p was prepared according to a literature procedure.15
Vinyl pyridine (1.0 g, 9.5 mmol) was dissolved in tetrahydrofurane
(100 mL) at 0 °C in an inert atmosphere. Water (150 mL) was
added dropwise until a saturated solution was formed. After that,
N-bromosuccinimide (1.86 g, 10.5 mmol) was added and reaction
mixture was stirred for 2 h at room temperature. Reaction mixture
was concentrated at rotary evaporator, THF was removed and
aqueous layer was extracted with dichloromethane (3 ꢀ 50 mL).
Combined organic layers were dried over Na2SO4, filtered and
evaporated under reduced pressure. Column chromatography
(SiO2; dichloromethane–methanol, 98:2) furnished the pure bro-
moalcohol as a colourless liquid in 64% yield (1.2 g). To a solution
of 2-bromo-1-pyridin-2-yl-ethanol (500 mg, 2.5 mmol) in metha-
nol (5 mL) potassium carbonate (400 mg, 2.9 mmol) was added
and reaction mixture stirred during 3 h at room temperature. Then
water (10 mL) was added and reaction mixture was extracted with
dichloromethane (3 ꢀ 15 mL), combined organic layers were dried
over Na2SO4, filtered and evaporated under reduced pressure. Col-
umn chromatography (SiO2; dichloromethane–methanol, 98:2)
furnished the pure epoxide 1p (167 mg, 55%) as a colourless liquid.
dH (300 MHz, CDCl3) 2.94 (1H, dd, J1 6.0 Hz, J2 2.5 Hz), 3.18 (1H, dd,
J1 6.0 Hz, J2 4.0 Hz), 4.02 (1H, dd, J1 4.0 Hz, J2 2.5 Hz), 7.20–7.29 (2H,
m), 7.68 (1H, dt, J1 7.5 Hz, J2 1.5 Hz), 8.55–8.56 (1H, m); dC
(75 MHz, CDCl3) 50.4, 52.8, 119.7, 123.1, 136.8, 149.4, 157.2.
4.5. Determination of enantiomeric purity and absolute
configuration
The enantiomeric excesses (ee’s) of the formed azido alcohols
and remaining epoxides were determined by chiral GC or HPLC
analyses under conditions described in Supplementary
information.
Absolute configurations were assigned by chiral GC or HPLC
analysis using reference compounds. In the case of epoxides 1a–
1e,8 1g,11 1j5a
,
1k16 and 1o5b assignment was based on
previously reported data. The enantiomerically enriched epoxides
(R)-1f, (S)-1h, (S)-1i, (R)-1l, (R)-1m, (R)-1n and (R)-1p were
prepared by (R,R)-(salen)CrCl catalysed ring-opening with
TMSiN3 according to Lebel and Jacobsen.17 Absolute configuration
of primary azido alcohol 3m was assigned by comparison of elu-
tion order on HPLC column with published data.18 Absolute config-
urations of 3k and 3l were assigned by analogy.
Acknowledgement
Financial support by the Unity through Knowledge Fund (UKF
grant no. 51) is gratefully acknowledged.
4.3. Kinetic resolution of epoxides
Supplementary data
To Tris–SO4 buffer (50 mM, pH 7.0) a stock solution of epoxide
in DMSO was added at room temperature (final concentration
5 mM for 1a–1j and 2 mM for 1k–1p), followed by addition of a
stock solution of NaN3 in water (0.5 mL, final conc. 5 mM or
2 mM). Reactions were initiated by addition of 1 mL of cell-free
extract in TEMG buffer (final volume 10 mL). The progress of the
reaction was followed by periodically taking samples (0.5 mL) from
reaction mixture. Samples were extracted with MTBE (1.0 mL) con-
taining mesitylene as internal standard and analysed by GC for
conversion and regioselectivity ratio. In parallel, chiral GC and/or
HPLC analyses were performed to determine the enantiomeric pur-
ity of the product and remaining substrate. The non-catalysed
reactions of epoxides 1k–1p with NaN3 were followed by monitor-
ing epoxide consumption in the absence of enzyme.
Supplementary data (chiral GC and HPLC analyses and NMR
spectra) associated with this article can be found, in the online ver-
sion, at http://dx.doi.org/10.1016/j.tetasy.2016.08.003.
References
1. a) Lutje Spelberg, J. H.; de Vries, E. J. In Enzyme Catalysis in Organic Synthesis;
Drauz, K., Gröger, H., May, O., Eds.; Wiley-VCH: Weinheim, 2012; Vol. 1, pp
363–411; b) Schallmey, M.; Floor, R. J.; Szyman´ ski, W.; Janssen, D. B. In
Comprehensive Chirality; Carreira, E. M., Yamamoto, H., Eds.; Elsevier:
Amsterdam, 2012; Vol. 7, pp 143–155; c) Majeric´ Elenkov, M.; Szyman´ ski,
W.; Janssen, D. B. In Science of Synthesis, Biocatalysis in Organic Synthesis; Faber,
K., Fessner, W.-D., Turner, N. J., Eds.; Thieme: Stuttgart, 2014; Vol. 2, pp 507–
527.
2. Hasnaoui, G.; Majeric´ Elenkov, M.; Lutje Spelberg, J. H.; Hauer, B.; Janssen, D. B.
ChemBioChem 2008, 9, 1048–1051.
3. a) Nagasawa, T.; Nakamura, T.; Yu, F.; Watanabe, I.; Yamada, H. Appl. Microbiol.
Biotechnol. 1992, 36, 478–484; b) Assis, H. M. S.; Sallis, P. J.; Bull, A. T.;
Hardman, D. J. Enzyme Microb. Technol. 1998, 22, 568–574; c) Nakamura, T.;
Nagasawa, T.; Yu, F.; Watanabe, I.; Yamada, H. Appl. Environ. Microbiol. 1994, 60,
1297–1301; d) van Hylckama Vlieg, J. E. T.; Tang, L.; Lutje Spelberg, J. H.;
Smilda, T.; Poelarends, G. J.; Bosma, T.; van Merode, A. E. J.; Fraaije, M. W.;
Janssen, D. B. J. Bacteriol. 2001, 183, 5058–5066; e) Schallmey, M.;
Koopmeiners, J.; Wardenga, R.; Schallmey, A. Appl. Environ. Microbiol. 2014,
80, 7303–7315.
4.4. Molecular modelling
Quantum mechanical docking calculation were performed by
systematic scanning of the HheA-N178A active site (PDB code:
1ZMO,6 N178 residue changed to alanine) using search algorithm
implemented in our program code qcc.19 Three molecular transla-
tional degrees of freedom were scanned with the step of 1.0 Å
whereas three molecular rotational degrees of freedom were
scanned with increments of 45°. Besides these 6 molecular degrees
of freedom, rotations around single bonds that correspond to the
torsional movement of epoxide ring in investigated compounds
were also investigated. In each case the model of the active site
was built. Single point calculation for the model of active site and
investigated compounds were performed using the PM6 method
´
4. a) Janssen, D. B.; Majeric Elenkov, M.; Hasnaoui, G.; Hauer, B.; Lutje Spelberg, J.
H. Biochem. Soc. Trans. 2006, 34, 291–295; b) Lutje Spelberg, J. H.; Tang, L.; van
Gelder, M.; Kellogg, R. M.; Janssen, D. B. Tetrahedron: Asymmetry 2002, 13,
1083–1089; c) Tang, L.; Lutje Spelberg, J. H.; Fraaije, M. W.; Janssen, D. B.
Biochemistry 2003, 42, 5378–5386.
´
5. a) Majeric Elenkov, M.; Tang, L.; Hauer, B.; Janssen, D. B. Adv. Synth. Catal. 2007,
349, 2279–2285; b) Lutje Spelberg, J. H.; van Hylckama Vlieg, J. E. T.; Tang, L.;
Janssen, D. B.; Kellogg, R. M. Org. Lett. 2000, 3, 41–43; c) Molinaro, C.; Guilbault,
A.-A.; Kosjek, B. Org. Lett. 2010, 12, 3772–3775.

´
935
A. Mikleuševic et al. / Tetrahedron: Asymmetry 27 (2016) 930–935
6. de Jong, R. M.; Kalk, K. H.; Tang, L.; Janssen, D. B.; Dijkstra, B. J. Bacteriol. 2006,
18. Cattoën, X.; Pericàs, M. A. Tetrahedron 2009, 65, 8199–8205.
19. Hrenar, T. qcc, Quantum Chemistry Code rev. 0.682, 2016.
188, 4051–4056.
7. van den Wijngaard, A. J.; Reuvekamp, P. T. W.; Janssen, D. B. J. Bacteriol. 1991,
20. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.;
Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.;
Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven,
T., ; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.;
Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Keith, T.; Kobayashi, R.; Normand,
J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.;
Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas,
O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09 Revision
D.01; Gaussian: Wallingford CT, 2013.
173, 124–129.
8. Majeric´ Elenkov, M.; Hauer, B.; Janssen, D. B. Adv. Synth. Catal. 2006, 348, 579–
585.
´
ˇ ˇ
9. Majeric Elenkov, M.; Primozic, I.; Hrenar, T.; Smolko, A.; Dokli, I.; Salopek-
Sondi, B.; Tang, L. Org. Biomol. Chem. 2012, 10, 5063–5072.
´
10. Tang, L.; Zhu, X.; Zheng, H.; Jiang, R.; Majeric Elenkov, M. Appl. Environ.
Microbiol. 2012, 78, 2631–2637.
11. Majeric´ Elenkov, M.; Tang, L.; Hauer, B.; Janssen, D. B. Org. Lett. 2006, 8, 4227–
4229.
12. Fringuelli, F.; Piermatti, O.; Pizzo, F.; Vaccaro, L. J. Org. Chem. 1999, 64, 6094–
6096.
13. Corey, E. J.; Chaykovsky, M. J. Am. Chem. Soc. 1965, 87, 1353–1364.
14. Shen, J.; Zhang, L.; Song, W.; Meng, T.; Wang, X.; Chen, L.; Feng, L.; Xu, Y.; Shen,
J. Acta Pharmacol. Sin. 2013, 34, 441–452.
15. Nadeau, R. G.; Hanzlik, R. P. Methods Enzymol. 1969, 15, 346–351.
16. Hasnaoui, G.; Lutje Spelberg, J. H.; de Vries, E.; Tang, L.; Hauer, B.; Janssen, D. B.
Tetrahedron: Asymmetry 2005, 16, 1685–1692.
21. Hrenar, T. moonee, Program for Manipulation and Analysis of Multi- and
Univariate Data rev. 0.6826, 2016.
22. Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
23. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1998, 37, 785–789.
17. Lebel, H.; Jacobsen, E. N. Tetrahedron Lett. 1999, 40, 7303–7306.