PLP-independent racemization: Mechanistic and mutational studies of: O -ureidoserine racemase (DcsC)

Article abstract of DOI:10.1039/c7ob03013d

O-Ureidoserine racemase (DcsC) is a PLP-independent enzyme in the biosynthetic route to the antibiotic d-cycloserine. Here we present the recombinant expression and characterization of a significantly more active DcsC variant featuring an N-terminal SUMO-tag. Synthesis of enantiomeric pure inhibitors in combination with site-specific mutation of active site cysteines to serines of this enzyme offers closer insights into the mechanism of this transformation. Homology modelling with a close relative (diaminopimelate epimerase, DapF) inspired C- and N-terminal truncation of DcsC to produce a more compact yet still active enzyme variant.

Full text of DOI:10.1039/c7ob03013d

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
PLP-independent racemization: mechanistic and
mutational studies of O-ureidoserine racemase
(DcsC)†
Cite this: DOI: 10.1039/c7ob03013d
Yeong-Chan Ahn, Conrad Fischer,
Marco J. van Belkum and John C. Vederas
*
O-Ureidoserine racemase (DcsC) is a PLP-independent enzyme in the biosynthetic route to the antibiotic
D-cycloserine. Here we present the recombinant expression and characterization of a significantly more
active DcsC variant featuring an N-terminal SUMO-tag. Synthesis of enantiomeric pure inhibitors in com-
bination with site-specific mutation of active site cysteines to serines of this enzyme offers closer insights
into the mechanism of this transformation. Homology modelling with a close relative (diaminopimelate
epimerase, DapF) inspired C- and N-terminal truncation of DcsC to produce a more compact yet still
active enzyme variant.
Received 5th December 2017,
Accepted 16th January 2018
DOI: 10.1039/c7ob03013d
rsc.li/obc
diaminopimelate epimerase (DapF). DapF catalyzes the inter-
conversion of LL-diaminopimelic acid and meso-diaminopime-
Introduction
Living cells generate predominantly L-amino acids. However, lic acid, both of which are intermediates in the biosynthesis of
bacteria make several D-amino acids for incorporation into the L-lysine in bacteria and plants.17 Recent crystal structures of
peptidoglycan cell wall layer.1,2
A number of enzymes DapF reveal how this enzyme catalyzes the epimerization of
that produce D-amino acids are racemases or epimerases substrate involving a thiol–thiolate pair18,19 or a redox-active
based on their overall mechanistic transformation.3 Two disulfide bridge20,21 within the active site. In DapF, initially
types of D-amino acid racemases can be found in nature: pyri- the active site cysteines are located near termini of α-helices in
doxal 5′-phosphate (PLP)-dependent and PLP-independent the enzyme whose dipoles reduce the thiol pKa to generate a
racemases.4,5
thiol–thiolate pair in the active site at neutral pH. Closure of
PLP-dependent racemases employ the pyridoxal cofactor to the enzyme upon substrate binding expels water from the
make an imine of the substrate amino acid to increase the active site and generates a hydrophobic interior that enhances
acidity of the α-proton.6,7 Since the pKa of the α-position of the basicity of the thiolate. The cysteine thiolate then acts as base
parent zwitterionic amino acid is about 29,8 deprotonation at to deprotonate the α-carbon of the rigidly held substrate22 and
that site is unfavorable. The PLP moiety acts as an electron the other cysteine functions as an acid to re-protonate this site
sink to decrease the pKa of the α-hydrogen by delocalization of from the opposite side. DapF also decreases the pKa of the
negative charge throughout the extended conjugated α-proton by stereoelectronic alignment of the bond between
π-system.7 In contrast, PLP-independent racemases generally the α-carbon and its hydrogen with the carboxylate π system.22
do not require any cofactor or metal.9–11 These enzymes can In addition, extensive hydrogen bonding of the carboxylate of
induce racemization of the α-carbon by using a thiolate–thiol the substrate with amide hydrogens of the enzyme in the
pair within the active site.12 Several PLP-independent race- active site helps to increase the acidity of the α-proton.18,23
mases have been characterized in the literature, including glu- This type of process can not be accomplished without enzy-
tamate racemase,13 proline racemase,14 aspartate racemase,15 matic machinery at present.
and isoleucine racemase.16 One intensively studied example is
O-Ureidoserine racemase (DcsC) is the enzyme that converts
(S)-L-ureidoserine to (R)-D-ureidoserine for D-cycloserine biosyn-
thesis in various Streptomyces species. It has been recently
identified as PLP-independent enzyme that acts similarly to
DapF (71% identity, accession number WP_078593587).24,25
Our recent work suggested that, like DapF, DcsC also has a
thiol–thiolate pair in the active site of the enzyme (Fig. 1).25
Exploring this relationship, we decided to test the inhibition
of DcsC and active site DcsC mutants with optically pure
Department of Chemistry, University of Alberta, Edmonton, AB, Canada, T6G 2G2.
E-mail: john.vederas@ualberta.ca; Fax: +1-780 492 2134; Tel: +1-780 492 5475
†Electronic supplementary information (ESI) available: Compound characteriz-
ation (MS sequencing, 1H-NMR substrate and inhibition assays), X-ray crystal
structure and analytical data (1H-NMR, 13C-NMR, IR, HRMS). CCDC 1588579
and 1588578. For ESI and crystallographic data in CIF or other electronic format
see DOI: 10.1039/c7ob03013d
This journal is © The Royal Society of Chemistry 2018
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 1 Substrate enantiomers 1a and 1b and racemic inhibitor 2.
Fig. 1 Proposed epimerization of O-ureidoserine (1) catalyzed by DcsC.
about 20 min (Fig. S4†). However, neither of the Cys → Ser
mutants (M1-DcsC or M2-DcsC) showed any deuterium incor-
poration with either 1a or 1b after 24 hours. According to the
proposed mechanism of DcsC, in the native enzyme both
active site thiols/thiolates have important roles in deprotona-
tion and re-protonation of the α-position of 1. Hence, in both
enzyme mutants, deprotonation of the α-carbon of the sub-
strate by the single active site thiolate, if it occurs at all, does
not result in exchange with solvent. The racemization process
does not occur, presumably because of the significantly higher
pKa of an active site serine as a proton donor on the opposite
side.
inhibitors to obtain new insights into the enzyme’s unusual
mechanism. Within this paper we describe the preparation
and activity of site specific mutants of DcsC as well as the syn-
thesis of optically pure inhibitors and enzyme interaction with
these inhibitors. Furthermore, the activity and kinetic perform-
ance of a truncated version of DcsC (tDcsC) are studied, pre-
senting a more compact enzyme version. Such structural
investigations would assist understanding of the distal
binding site necessary for substrate side chain recognition and
very strict enzyme specificity.25
Results and discussion
Enzyme purification and mutation of DcsC
Inhibition of SUMO-DcsC and mutants
To show that the active site of DcsC consists of a thiol–thiolate
pair, the irreversibly-binding oxirane inhibitor 2 was syn-
thesized in racemic form (Scheme 1).25 Similar to the aziri-
dine-based inhibition of glutamate racemase30 and diaminopi-
melate epimerase,23,31–34 it was expected that the epoxide ring
would react with an active site cysteine, giving rise to a
covalent attachment (Fig. 2). Each cysteine in the active site is
expected to be inhibited by one enantiomer of the inhibitor at
a time (Fig. 2) resulting in an m/z-peak of increased mass
(43.825 kDa for SUMO-DcsC). The enzyme mutants (M1-DcsC,
M2-DcsC) each were also successfully inhibited by racemic
inhibitor 2, indicated by additional mass peaks in the
DcsC fused to a His-tagged small ubiquitin-like modifier
(SUMO) was overexpressed in Escherichia coli BL21(DE3). The
SUMO-fusion protein was chosen to increase protein over-
expression and stability, and achieve a His-tagged free DcsC
protein after cleavage of the fusion tag.26 After purification by
Ni2+-affinity chromatography the SUMO tag was removed by
treatment with SUMO protease. The mass of the DcsC was
detected by ESI-TOF (30.383 kDa, Fig. S1†). However, removal
of the SUMO-tag resulted in an unstable protein that quickly
precipitated even at 0 °C. Since the SUMO-tag has been shown
to be beneficial for the crystallization of proteins,27–29 we
studied further properties including the tag (SUMO-DcsC, m/z
43.649 kDa, Fig. S2†).
A previous study by our group suggested locations of the
active site cysteine residues responsible for the activity of a
C-terminal His-tagged DcsC (Fig. 1).25 Single-site mutation of
these cysteines to serines should allow confirmation of the
suggested bilateral epimerization mechanism. Therefore, site-
directed mutagenesis was used to independently change Cys81
to Ser81 and Cys227 to Ser227 of SUMO-DcsC (M1-DcsC and
M2-DcsC, respectively), and both enzyme mutants were iso-
lated as described above. The loss of 16 m/z units in the
respective ESI-TOF-MS (m/z 43.633 kDa, Fig. S2†) and trypsin-
based MS–MS sequencing confirms the desired mutation of
the M1-and M2-DcsC proteins (Fig. S3†).
To confirm the activity of the enzymes, the natural substrate
isomers, L-O-ureidoserine (1a) and D-O-ureidoserine (1b) were
synthesized (Scheme 1).25 The activity of SUMO-DcsC, M1-
DcsC and M2-DcsC were determined by NMR (20 mM Tris,
D2O, pD 7.6). Non-mutated SUMO-DcsC successfully
exchanges the α-hydrogen of 1a or 1b to deuterium, as
Fig. 2 a. Putative inhibition of Cys81 by (S)-Inhibitor. b. Putative inhi-
observed by disappearance of the α-H peak at 4.05 ppm after bition of Cys227 by (R)-Inhibitor.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Organic & Biomolecular Chemistry
Paper
ESI-TOF-MS spectra (Fig. S2†). The difference between two
main peaks in the mass spectra is 176 Da (m/z ratio), corres-
ponding to the molecular weight of 2. MS–MS sequencing data
confirmed the desired sites of inhibition (Fig. S3†). These
results suggest that SUMO-DcsC and the mutants can be inhi-
bited by racemic oxirane inhibitor 2. However, SUMO-DcsC
and both mutants showed low inhibition (less than 40%,
Fig. S2†), and we were not able to obtain crystals from these
proteins to solve the protein structure.
Fig. 3 Electrospray mass spectra of inhibited M1-DcsC mutants (A1: 2a,
A2: 2b) and M2-DcsC mutants (B1: 2a, B2: 2b). The nominal mass of the
inhibitor is 176 Da.
Synthesis of chiral inhibitors
To investigate the mechanism of DcsC (Fig. 2), we targeted the
synthesis of the optically pure inhibitors 2a and 2b. Since
racemic oxirane inhibitor 2 was somewhat unstable under
aqueous conditions at room temperature due to epoxide ring
opening, we chose to synthesize stable diastereomeric ester
precursor mixtures 7a/8a and 7b/8b and to separate pure enan-
tiomers from them (Scheme 2). Hydrolysis of each stereo-
chemically pure ester would then give the optically pure inhibi-
tors. (S)-1-Phenylethanol (4a) and (R)-1-phenylethanol (4b)
were used as chiral entries for the synthesis. Briefly, bromo-
acrylic acid was esterified separately with either alcohol 4a or
4b in a Mitsunobu reaction. The bromoesters (5a and 5b) were
treated individually with tert-butyl peroxide, using NaH as
base in ether. Peroxides 6a and 6b were then reacted with
hydroxyurea to form the diastereomeric precursor inhibitor
mixtures 7a/8a and 7b/8b, respectively. The diastereomeric
inhibitors in each mixture were separated by recrystallization.
Repeated recrystallization from a benzene–EtOAc mixture
(2 : 1) gave enantiomerically pure precursors 7a and 7b, whose
structures were confirmed by X-ray diffraction analysis (Fig. S5,
Table S1†). Saponification of each by LiOH furnishes the pure
chiral inhibitors 2a and 2b, respectively. Both inhibitors (10
eq.) were separately added to M1-DcsC and M2-DcsC mutants
and samples were submitted to ESI-TOF-MS analysis to ident-
ify and confirm the sites of inhibition. Whereas M1-DcsC was
only inhibited by 2a, the M2-DcsC mutant was inhibited by 2b
(Fig. 3 and Fig. S6†).
In addition, MS–MS sequencing data confirmed the inhi-
bition sites (Fig. S7†). These data indicate that the free active
site cysteine in both mutants (C227 in M1-DcsC and C81 in
M2-DcsC) can each only react with one inhibitor enantiomer,
thus supporting the bilateral reaction mechanism (Fig. 2).
Truncation of DcsC
To obtain additional structural information about DcsC, we
compared the modeled structure of the protein to the reported
coordinates of DapF from Haemophilus influenzae (PDB: 1bwz,
Fig. 4).18 Closer examination reveals
a flexible N- and
C-terminus of DcsC consisting of 9 and 3 amino acids, respect-
ively (Fig. S8†). Truncation of these less-structured segments
could result in an enzyme (tDcsC) that maintains catalytic
activity, but has improved stability. In contrast to the
SUMO-DcsC, we were not able to cleave the SUMO-tag from
tDcsC after purification. Hence for activity testing, SUMO-
Fig. 4 Ribbon-model structure of DcsC (lightblue) based on the mod-
eling of DapF from Haemophilus influenzae (gray, PDB ID: 1bwz, Seq.
identity: 42%). The truncated areas (C- and N-terminal) are highlighted
in green.
Scheme 2 Synthesis of chiral oxirane inhibitors.
This journal is © The Royal Society of Chemistry 2018
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 Kinetic parameters of DcsC variants
Species
Direction
vmax (mM s−1
)
Km (mM)
Kcat (1 s−1
)
Kcat/Km (1 M−1 s−1
)
Ref.
DcsC-His6
DcsC-His6
1a → 1b
1b → 1a
1a → 1b
1b → 1a
1a → 1b
1b → 1a
1a → 1b
1b → 1a
0.084
0.015
0.022
0.047
<0.001
<0.001
0.042
0.046
110
17
12
158
29
1436
1706
41 130
45 302
n.d.
n.d.
21 137
8642
25
25
—
—
—
—
—
—
SUMO-DcsC
SUMO-DcsC
DcsC (tag-free)
DcsC (tag-free)
SUMO-tDcsC
SUMO-tDcsC
475
1450
n.d.
n.d.
315
194
32
n.d.
n.d.
15
22
tDcsC was incubated with the natural substrates 1a and 1b protein that quickly precipitates even at 0 °C, preventing deter-
(20 mM Tris, D2O, pD 7.6). The activity profiles of SUMO- mination of kinetic parameters. Unlike for the His-tagged
tDcsC are somewhat lower compared to the native enzyme (see DcsC version, the substrate affinity towards the O-ureidoserine
below).
enantiomers (direction of transformation, 1a → 1b or 1b → 1a)
In order to characterize the inhibition properties of SUMO- is comparable for the SUMO-tagged variants.
tDcsC, mutants M1-tDcsC and M2-tDcsC were generated that
incorporate a serine for cysteine mutations in the active site in
analogous fashion to those generated in the native protein
(Cys72 → Ser in M1-tDcsC and Cys218 → Ser in M2-tDcsC).
ESI-TOF-mass spectra reveal the proteins of desired mass
(tDcsC: 42 317 Da, M1-tDcsC and M2-tDcsC: 42 301 Da,
Fig. S9†). The sequence was confirmed to be correct by apply-
ing MS–MS digest analysis (Fig. S10†). Incubation of both
mutants with the natural substrates 1a and 1b as before
showed no activity, underlining the proposed bilateral mecha-
nism. Both mutants were almost completely inhibited by the
expected compatible enantiomer 2a or 2b (Fig. S9†). However,
a second site of alkylation becomes available in the truncated
versions (Cys41), indicated by a small additional mass peak
(Fig. S9†). This suggests a somewhat looser structure with
greater access for inhibitors to the enzyme interior.
Conclusions
Herein we report the recombinant expression of
O-ureidoserine racemase (DcsC) featuring an N-terminal
SUMO-tag. Purified SUMO-DcsC racemizes the natural sub-
strates L- and D-O-ureidoserine 1a and 1b with a 30-fold
increased catalytic activity in comparison to a C-terminally
His-tagged enzyme version. Individual mutation of both active
site cysteines and incubation with inhibitor enantiomers 2a
and 2b reveals selective inactivation depending on the chirality
of the inhibitor, thus supporting the bilateral reaction mecha-
nism of the enzyme. Truncation of the C- and N-terminal ends
of the protein as well as removal of the SUMO-tag results in a
de-stabilization of the protein structure and reduction of the
enzyme activity. Crystallization studies are underway ultimately
aiming to obtain a single X-ray structure of the enzyme, which
will help us to rationalize the interior dimensions and distal
site recognition of the active site.
Kinetic parameters of various DcsC variants
The catalytic parameters for various DcsC variants were deter-
1
mined by time-dependent H-NMR spectroscopy and are sum-
marized in Table 1. The catalytic parameters for the
C-terminal His-tagged enzyme version of DcsC published pre-
viously25 are also included in this table for reference. The data
were extracted from the signal ratio between the vanishing
Experimental
Fusion protein expression and purification
α-hydrogen (at
δ ∼4.05 ppm) and the β-hydrogen (at DNA sequences encoding DcsC and tDcsC were purchased
δ ∼4.2 ppm) (Fig. S11†). The initial velocities (3 min) were used from BioBasic Inc. (Ontario, Canada) and cloned separately
to calculate the Michaelis–Menten parameters. In general, into the pET SUMO (small ubiquitin-like modifier) expression
these calculations reveal two trends: the catalytic efficiency vector (Invitrogen). Clones were sequenced to confirm that
(Kcat/Km) in both directions is almost 30-times higher for the DcsC and tDcsC were in frame with the His-tagged SUMO
SUMO-tagged DcsC version in comparison to the His-tagged fusion protein. The resulting pET SUMO-DcsC and pET SUMO-
enzyme version. One possible explanation for this enhanced tDcsC plasmids were transformed into Escherichia coli BL21
catalytic performance could be the predominantly dimeric (DE3). For protein expression, the E. coli transformant was
state of SUMO-DcsC compared to a higher oligomeric state of grown in 1 L LB Broth, Miller (Luria–Bertani) (10 g of tryptone,
the His-tagged version (compare Fig. S12†), which has been 5 g of yeast extract, 10 g of NaCl) at 37 °C with shaking (225
described as catalytic accelerator for other racemases.35 rpm) to an optical density (OD600) of 0.6–0.8 using kanamycin
Truncation of the enzyme (SUMO-tDcsC, Table 1) renders as selective pressure. Next, the cells were chilled in an ice bath
these variants 50–80% less active than the native version but for 10 minutes before IPTG (0.2 mM) was added to the culture
still 5–14-times more active than the C-terminally His-tagged to induce protein expression. The culture was then incubated
variant. Removal of the SUMO-tag results in an instable at 23 °C for 16 hours with shaking (225 rpm). The cells were
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Organic & Biomolecular Chemistry
Paper
harvested by centrifugation (6300 rpm, 4 °C, 30 min) and sus- EtO2 was distilled over Na; MeOH, EtOH, acetonitrile as well as
1
pended in lysis buffer [100 mM sodium-phosphate (pH 7.8), DMF were dried over activated 4 Å molecular sieves. H-NMR
300 mM NaCl, 20 mM imidazole, 1 mM EDTA] with 2 mg of (600 MHz/400 MHz) and 13C-NMR (151 MHz/125 MHz) data
lysozyme and 10 unit of DNase I (Thermo Scientific). The cells were obtained on
a Agilent/Varian VNMRS four-channel
were lysed using a Constant Systems Cell Disruptor, model TS spectrometer at room temperature with TMS or the solvent
(Constant Systems, Ltd) operated at 20 kpsi. The lysate was signal as internal standard. High-resolution mass spectra were
centrifuged (15 000 rpm, 30 min), and the supernatant con- recorded on an Agilent 6220 oaTOF with electron ionization.
taining the fusion protein was isolated. The lysate was mixed IR-spectra were measured using a Nicolet 8700 spectrometer.
with Ni-NTA resin (Qiagen), and loaded on a fritted column. Specific rotation was measured on a Sodium D line (589 nm)
The resin was washed with 20 mL lysis buffer, and the fusion with PerkinElmer 241 Polarimeter. Circular dichroism spectra
protein was eluted with 10 mL lysis buffer containing 300 mM were recorded on an OLIS globalworks CD spectrophotometer.
imidazole. The fusion protein solution was dialyzed against Staring compound 3 and (R)-1-phenylethanol, (S)-1-phenyl-
3 L of buffer [20 mM Tris-HCl (pH 7.8), 5 mM DTT] for 20 h at ethanol are commercially available from Sigma-Aldrich.
4 °C and concentrated by centrifugation using an Amicon
Ultra centrifugal filter (30 kDa, 4000g, 20 min). Average recov-
ery was 10–30 mg L−1 culture.
1-Phenylethyl-2-bromomethyl acrylate [5a, 5b]
To a solution of bromomethylacrylic acid 3 (1.50 g, 9.09 mmol)
and diisopropyl azodicarboxylate (DIAD) (1.80 mL, 9.14 mmol)
in diethyl ether (15 mL), was added a solution of (S)-1-phenyl-
ethanol 4a (1.20 mL, 9.90 mmol) and triphenylphosphine
(2.36 g, 9.09 mmol) in diethyl ether (15 mL), dropwise at 0 °C.
The mixture was stirred at the same temperature for 30 min,
and then at room temperature for 48 h. The mixture was fil-
tered and washed with Et2O (20 mL). After concentration of
the filtrate, the crude product was purified by flash column
chromatography (5% EtOAc in hexanes) to furnish the acrylate
as a yellow liquid (1.47 g, 60%).
Cleavage of fusion protein
To cleave the SUMO tag, approximately 20 mg of fusion
protein was incubated with 25 units of His-tagged SUMO pro-
tease (McLab, South San Francisco, CA) in 50 mM Tris-HCl
(pH 8.0), 1 mM DTT, 0.2% Igepal CA-360 (Sigma) and 150 mM
NaCl in a total volume of 5 mL. Digestion of the fusion protein
was performed for 20 h at 4 °C. Cleavage of the fusion protein
checked by sodium dodecyl sulfate-polyacrylamide gel electro-
phoresis (SDS-PAGE). The cleavage mixture was added to 1 mL
of Ni-NTA agarose (Qiagen) resin and loaded on a fritted
column to remove the His-tagged SUMO and His-tagged
SUMO protease. The flow through containing the enzyme was
collected and checked for purity by SDS-PAGE. The enzyme
solution was then dialyzed against 3 L of buffer [20 mM Tris-
HCl (pH 7.8), 5 mM DTT] for 20 h at 4 °C and concentrated by
centrifugation using an Amicon Ultra centrifugal filter
(30 kDa, 4000g, 20 min). Average yield was around 7 mg of
enzyme.
5a (R). 1H-NMR (600 MHz, CDCl3) δ 7.40 (m, 5H, Ph–H),
6.39 (d, 1H, J = 1.0 Hz, vCH), 6.01 (d, 1H, J = 1.0 Hz, vCH),
5.95 (q, 1H, J = 6.9 Hz, Ph–CH), 4.21 (dd, 2H, J = 2.5, 1.0 MHz,
Br–CH2), 1.58 (d, 3H, J = 8.5 Hz, CH–CH3). 13C-NMR (150 MHz,
CDCl3) δ 20.3, 21.2, 28.5, 73.4, 125.6, 127.5, 128.1, 128.3, 138.3,
141.5, 164.1. [α]2D5 = +23.24 (c = 0.500, CH2Cl2). IR (thin film)
ν 3087, 3063, 2960, 2930, 2872, 1722, 1632, 1453, 1186 cm−1
.
HRMS (ESI) [M + Na]+ m/z calcd for C12H13BrNaO2: 290.9991,
found 290.9991.
5b (S). Starting with (R)-1-phenylethanol 4b. 1H-NMR
(600 MHz, CDCl3) δ 7.40 (m, 5H, Ph–H), 6.39 (d, 1H, J = 1.0 Hz,
vCH), 6.01 (d, 1H, J = 1.0 Hz, vCH), 5.95 (q, 1H, J = 6.9 Hz,
Ph–CH), 4.25 (dd, 2H, J = 2.5, 1.0 MHz, Br–CH2), 1.58 (d, 3H,
J = 6.4 Hz, CH–CH3). 13C-NMR (150 MHz, CDCl3) δ 20.1, 21.3,
Site-directed mutagenesis of DcsC and tDcsC enzymes
QuikChange II Site-Directed Mutagenesis Kit (Agilent
Technologies) was used to construct plasmids encoding DcsC
and tDcsC with cysteine to serine residues mutations in the
active site according to the manufacturer’s instructions.
Primer pair MVB206 (5′-CCGTTCCGCGCAGTCTGGCAACGG
TGCACG-3′) and MVB207 (5′-CGTGCACCGTTGCCAGACTGCG
CGGAACGG-3′) was used to create the C81S mutation, whereas
primer pair MVB208 (5′-GTGAAACCCTGGCGTCCGGTTCC
GGTGCATG-3′) and MVB209 (5′-CATGCACCGGAACCGG
ACGCCAGGGTTTCAC-3′) was used to create the C227S
28.4, 73.5, 125.7, 127.5, 128.0, 128.3, 138.1, 141.5, 164.1. [α]D25
=
−26.71 (c = 0.500, CH2Cl2). IR (thin film) ν 3087, 3064, 2980,
2931, 2872, 1722, 1632, 1453, 1187 cm−1. HRMS (ESI) [M + Na]+
m/z calcd for C12H13BrNaO2: 290.9991, found 290.9991.
1-Phenylethyl-2-tert-butylperoxomethyl acrylate [6a, 6b]
mutation. The presence of these mutations was verified by A solution of tert-butylhydroperoxide (1.50 mL, 8.26 mmol) in
nucleotide sequencing of the whole coding region. The result- dry CH2Cl2 (20 mL) was stirred at −5 °C under a stream of
ing plasmids were transformed into E. coli BL21(DE3) to over- argon. To this was added sodium hydride (0.28 g, 7.0 mmol,
express and isolate the mutant enzymes.
60% dispersion in mineral oil). After 15 minutes, 5a or 5b
(1.47 g, 5.48 mmol) was added by dropwise at −5 °C. After 4 h,
the mixture was filtered and washed with dry CH2Cl2. After
General information
All non-aqueous reactions were done with oven-dried glassware concentration of the filtrate, the crude product was purified by
under argon. Solvents were dried by follows: CH2Cl2 was dis- flash column chromatography (4% EtOAc in hexanes) to
tilled over CaH2; tetrahydrofuran (THF) was distilled over Na; obtain 6a or 6b as a yellow liquid (0.72 g, 46%).
This journal is © The Royal Society of Chemistry 2018
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
6a (R). 1H-NMR (600 MHz, CDCl3) δ 7.38 (m, 5H, Ph–H), mixture was cooled to 0 °C on an ice-water bath. To this
6.35 (d, 1H, J = 1.0 Hz, vCH), 5.91 (d, 1H, J = 1.0 Hz, vCH), mixture was added potassium tert-butoxide solution
a
5.94 (q, 1H, J = 6.6 Hz, Ph–CH), 4.63 (dd, 2H, J = 3.0, 1.0 MHz, (0.80 mL, 0.80 mmol, 1 M in THF) dropwise, and the solution
O–CH2), 1.58 (d, 3H, J = 7.0 Hz, CH–CH3), 1.18 (s, 9H, O–C– turned light yellow. The mixture was stirred at 0 °C for 30 min,
CH3); 13C-NMR (150 MHz, CDCl3) δ 21.1, 25.1, 72.7, 79.8, and another 4 h at room temperature. Solvent was then
125.8, 127.5, 128.2, 136.7, 141.8, 165.0. [α]2D5 = +20.73 (c = removed by a rotary evaporator attached to high-vacuum. The
0.500, CH2Cl2). IR (thin film) ν 3089, 3065, 3034, 2980, crude yellow mixture was applied to silica gel chromatography
2932, 2874, 1723, 1637, 1454, 1363, 1162 cm−1. HRMS (ESI) (9 : 1 EtOAc : hexane, then pure EtOAc), to yield the diastereo-
[M + Na]+ m/z calcd for C16H22NaO4 301.1410, found 301.1409.
mers 7b (1S,2R) and 8b (1S,2S) as a yellowish solid–liquid
6b (S). 1H-NMR (600 MHz, CDCl3) δ 7.38 (m, 5H, Ph–H), 6.35 mixture (0.17 g, 23%). Repeated re-crystallization from
(d, 1H, J = 1.0 Hz, vCH), 5.91 (d, 1H, J = 1.0 Hz, vCH), 5.94 benzene/EtOAc yielded 7b (1S,2R) as crystalline material.
(q, 1H, J = 6.6 Hz, Ph–CH), 4.63 (dd, 2H, J = 3.0, 1.0 MHz,
1H-NMR (600 MHz, CDCl3) δ 7.38 (m, 5H, Ph–H), 5.94 (q,
O–CH2), 1.58 (d, 3H, J = 7.0 Hz, CH–CH3), 1.18 (s, 9H, O–C– 1H, J = 6.5 Hz, Ph–CH), 4.41(d, 1H, J = 10.1 Hz, CH2ONH), 4.13
CH3); 13C-NMR (150 MHz, CDCl3) δ 21.4, 25.2, 72.9, 79.9, (d, 1H, J = 10.1 Hz, CH2ONH), 3.16 (d, 1H, J = 6.0 Hz, oxirane-
125.6, 127.5, 128.1, 136.6, 141.7, 165.1. [α]2D5 = −19.23 (c = CH2), 3.02 (d, 1H, J = 6.0 Hz), 1.61 (d, 3H, J = 6.5 Hz, CH–CH3);
0.500, CH2Cl2). IR (thin film) ν 3088, 3065, 3034, 2980, 13C-NMR (150 MHz, CDCl3) δ 21.6, 50.3, 74.7, 124.8, 128.2,
2932, 2873, 1724, 1638, 1454, 1363, 1162 cm−1. HRMS (ESI) 128.8, 165.3. IR (thin film) ν 3415, 3392, 3171, 3041, 2989,
[M + Na]+ m/z calcd for C16H22NaO4 301.1410, found 301.1410.
2939, 2877, 1719, 1683, 1457, 1373, 1177 cm−1. HRMS (ESI)
[M + Na]+ m/z calcd for C13H16N2O5Na 303.0951, found
303.0950. Evaporation of the filtrate and re-crystallization gave
pure 8b (1S,2S) showing identical analytical data as the 1R,2R-
(R)-1-Carboxy-(1-phenylethyl)-1-O-ureido-methoxy oxiranes 7a
(1R,2S) and 8a (1R,2R)
Compound 6a (R) (0.720 g, 2.58 mmol) and hydroxyurea compound obtained from the (R)-precursor.
(0.250 g, 3.28 mmol) were dissolved in dry DMF (15 mL), and
the mixture was cooled to 0 °C on an ice-water bath. To this
1-Carboxy-1-O-ureido-methoxy-(S)-oxirane 2a (S)
mixture was added
a potassium tert-butoxide solution Ester 7a (1R,2S) (1.0 mg, 0.004 mmol) was saponified in D2O
(0.80 mL, 0.80 mmol, 1 M in THF) dropwise, and the solution (0.7 mL) containing LiOH solution (0.010 mL, 0.010 mmol, 1
turned light yellow. The mixture was stirred at 0 °C for 30 min, M solution in D2O). After 10 min, the reaction was deemed to
and another 4 h at room temperature. Solvent was then be complete by 1H-NMR, and the solvent was removed by
removed by a rotary evaporator attached to high-vacuum. The lyophilization.
crude yellow mixture was applied to silica gel chromatography
1H-NMR (600 MHz, CDCl3) δ (600 MHz, D2O) 4.43 (d, 1H,
(9 : 1 EtOAc : hexane, then pure EtOAc), to yield the diastereo- J = 10.2 Hz, CH2ONH), 3.72 (d, 1H, J = 10.1 Hz, CH2ONH), 2.95
mers 7a (1R,2S) and 8a (1R,2R) as a yellowish solid–liquid (s, 2H, oxirane-CH2); 13C-NMR (150 MHz, CDCl3) δ 49.2, 57.1,
mixture (0.17 g, 23%). Repeated re-crystallization from 78.5, 163.1, 174.5 IR (thin film) ν 3301 (broad), 2938, 1671,
benzene/EtOAc yielded 7a (1R,2S) as crystalline material.
1619, 1425, 1107 cm−1. HRMS (ESI) [M − H]− m/z calcd for
1H-NMR (600 MHz, CDCl3) δ 7.38 (m, 5H, Ph–H), 5.94 (q, C5H7N2O5 175.0360, found 175.0357.
1H, J = 6.5 Hz, Ph–CH), 4.41(d, 1H, J = 10.1 Hz, CH2ONH), 4.13
1-Carboxy-1-O-ureido-methoxy-(R)-oxirane 2b (R)
(d, 1H, J = 10.1 Hz, CH2ONH), 3.16 (d, 1H, J = 6.0 Hz, oxirane-
CH2), 3.02 (d, 1H, J = 6.0 Hz), 1.61 (d, 3H, J = 6.5 Hz, CH–CH3); Ester 7b (1S,2R) (1.0 mg, 0.04 mmol) was saponified in D2O
13C-NMR (150 MHz, CDCl3) δ 21.8, 50.1, 74.2, 75.1, 124.8, (0.7 mL) containing LiOH solution (0.010 mL, 0.010 mmol,
128.2, 128.8, 165.4 IR (thin film) ν 3415, 3392, 3174, 3040, 1 M solution in D2O). After 10 min, the reaction was deemed
2989, 2940, 2878, 1718, 1682, 1457, 1373, 1176 cm−1. HRMS to be complete by 1H-NMR, and the solvent was removed by
(ESI) [M + Na]+ m/z calcd for C13H16N2O5Na 303.0951, found lyophilization.
303.0954.
1H-NMR (600 MHz, CDCl3) δ (600 MHz, D2O) 4.43 (d, 1H,
Evaporation of the filtrate after re-crystallization gave pure J = 10.2 Hz, CH2ONH), 3.72 (d, 1H, J = 10.1 Hz, CH2ONH), 2.95
1
8a (1R,2R). H-NMR (600 MHz, CDCl3) δ 7.38 (m, 5H, Ph–H), (s, 2H, oxirane-CH2); 13C-NMR (150 MHz, CDCl3) δ 49.2, 57.2,
5.94 (q, 1H, J = 6.5 Hz, Ph–CH), 4.37 (d, 1H, J = 10.1 Hz, 78.5, 163.3, 174.7 IR (thin film) ν 3301 (broad), 2938, 1671,
CH2ONH), 4.05 (d, 1H, J = 10.1 Hz, CH2ONH), 3.23 (d, 1H, J = 1619, 1425, 1107 cm−1. HRMS (ESI) [M − H]− m/z calcd for
6.0 Hz, oxirane-CH2), 2.96 (d, 0.6H, J = 6.0 Hz), 1.61 (d, 3H, J = C5H7N2O5 175.0360, found 175.0358.
6.5 Hz, CH–CH3); 13C-NMR (150 MHz, CDCl3) δ 21.8, 50.1,
General protocol for methanolysis of cycloserine25
74.2, 75.1, 124.8, 128.2, 128.8, 165.2. HRMS (ESI) [M + Na]+ m/z
calcd C13H16N2O5Na 303.0951, found 303.0952.
The literature procedure was adapted.25 Acetyl chloride
(7.14 mL, 100 mmol) was added to dry methanol (100 mL)
cooled in an ice bath and stirred under argon for 15 min. Next,
1.00 g of enantiomerically pure cycloserine (9.80 mmol) was
(S)-1-Carboxy-(1-phenylethyl)-1-O-ureido-methoxy oxiranes 7b
(1S,2R) and 8b (1S,2S)
Compound 6b (S) (0.72 g, 2.58 mmol) and hydroxyurea (0.25 g, dissolved in the acidic methanol solution and warmed to room
3.28 mmol) were dissolved in dry DMF (15 mL), and the temperature. After 15 min, a reflux condenser was connected
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Organic & Biomolecular Chemistry
Paper
to flask, and the mixture was stirred for 15 h at 70 °C. the α-CH to α-CD, the progress of the reaction was monitored
Insoluble materials were removed by filtration, and solvent by the disappearance of the α-CH peak.
was removed by a rotary evaporator. Ethyl acetate (700 mL) was
Enzyme kinetics monitoring by CD spectroscopy
added to recrystallize the product, and white solid was precipi-
tated after an hour of stirring. The O-aminoserine methyl ester
(1.25 g) was filtered and analyzed by NMR and CD, which dis-
played expected properties.25
The ureidoserine samples (300 µL, 2–15 mM, phosphate buffer
(pH = 7.8), 1 mM DTT) were placed in 1 mm quartz cell, and
reactions were monitored at 30 °C, at 206 nm. The CD signal
(in mdeg) was recorded every two seconds, for five minutes.
The reaction was initiated by adding truncated DcsC (2–5 μg),
and the quartz cell was directly placed in CD spectrophoto-
meter. The data were processed to Excel and GraphPad prism.
The enzyme initial velocity (Vo) was calculated within the
linear range of the CD-signal decay (initial 2 minutes), and
kinetics were extracted using Michaelis–Menten analysis.
General ureidoylation and saponification protocol
The literature procedure was adapted.25 O-Aminoserine methyl
ester (300 mg, 2.20 mmol) was dissolved in 10 mL water, and
potassium cyanate (180 mg, 2.22 mmol) was added to the solu-
tion. pH of solution was adjusted to 4–5 to get maximal yield,
and the solution was stirred for 2 h at room temperature.
When the reaction was finished as determined by TLC, the
solution was adjusted to pH ∼1 with 1 M HCl to terminate the
reaction. Solvent was removed by rotatory evaporator, and the
oily ester product was used for next reaction without purifi-
cation. The ester was dissolved in 20 mL of water and 2 M pot-
assium hydroxide (2 ml, 4.0 mmol) was added. The mixture
was stirred at room temperature until the reaction was finished
as determined by TLC (NH4OH : i-PrOH = 3 : 7). The mixture
was adjusted to pH 4 with 1 M HCl to terminate the reaction,
and excess EtOH (20 mL) was added to precipitate the product.
Solvent was evaporated by rotatory evaporator, and the product
was purified by Varian BondElut C18 SPE cartridge. The car-
tridge was washed with water first (20 mL), and a water solu-
tion of crude compound was loaded to cartridge. The product
was eluted with water (10 mL) and then MeCN–H2O (1 : 1) solu-
tion. The fractions containing product were collected and lyo-
philized to yield the pure isomer of O-ureidoserine 1 as a solid.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
The authors gratefully acknowledge Dr Robert McDonald and
Dr Michael Ferguson (University of Alberta) for X-ray crystallo-
graphic analyses, Mark Miskolzie (University of Alberta
Nuclear Magnetic Resonance Laboratory) for NMR analyses,
Wayne Moffat and Jennifer Jones (University of Alberta
Analytical and Instrumentation Laboratory) for IR and CD ana-
lyses, Gareth Lambkin (University of Alberta) for biological ser-
vices, and Jing Zheng and Bela Reiz (University of Alberta Mass
Spectrometry Laboratory). This work was supported by the
Natural Sciences and Engineering Research Council of Canada
(NSERC) and Alberta Innovates Health Solutions (AIHS
Fellowship to CF) for financial support.
X-ray diffraction
Crystals of 7a (1R,2S) and 7b (1S,2R), suitable for analysis were
obtained by slow cooling of a hot solution of each pure isomer
in benzene/EtOAc (2 : 1). The intensity data were collected on a
Bruker APEX II diffractometer with CuKα radiation (λ =
1.5418 Å) using ω- and φ-scans. Reflections were corrected for
background and Lorentz polarization effects. Preliminary
structure models were derived by application of direct
methods and were refined by fullmatrix least squares calcu-
lation based on F2 for all reflections. All hydrogen atoms were
included in the models in calculated positions and were
refined as constrained to bonding atoms. The crystal data and
experimental parameters are summarized in the ESI and the
structures are deposited under CCDC 1588579 (7a) and CCDC
1588578 (7b).†
Notes and references
1 R. J. Thompson, G. H. Bouwer, D. A. Portnoy and
F. R. Frankel, Infect. Immun., 1998, 66, 3552–3561.
2 N. E. Caceres, N. B. Harris, J. F. Wellehan, Z. Feng, V. Kapur
and R. G. Barletta, J. Bacteriol., 1997, 179, 5046–5055.
3 K. Soda and N. Esaki, Pure Appl. Chem., 1994, 66, 709–714.
4 S. Y. Choi, N. Esaki, M. Ashiuchi, T. Yoshimura and
K. Soda, Proc. Natl. Acad. Sci. U. S. A., 1994, 91, 10144–
10147.
5 C. T. Walsh, J. Biol. Chem., 1989, 264, 2393–2396.
6 S. Sun and M. D. Toney, Biochemistry, 1999, 38, 4058–4065.
7 J. C. Vederas and H. G. Floss, Acc. Chem. Res., 1980, 13,
455–463.
8 J. P. Richard, T. L. Amyes, B. Goryanova and X. Zhai, Curr.
Opin. Chem. Biol., 2014, 21, 1–10.
Enzyme kinetics monitoring by 1H-NMR
To analyze the kinetics of truncated DcsC by NMR, enantio-
merically pure ureidoserine (2–15 mM) was dissolved to Tris
buffer (D2O, 20 mM, pD = 7.8). To initiate the reaction, trun-
cated DcsC (3–6 μg) was added and 1H NMR was recorded
every 3 min. To determine enzyme activity, the peak of β-CH2
9 M. E. Tanner, Acc. Chem. Res., 2002, 35, 237–246.
peak (mutiplet, 4.2–4.3 ppm) was compared with the α-CH 10 A. D. Radkov and L. A. Moe, Appl. Microbiol. Biotechnol.,
peak (triplet, 3.9–4.0 ppm). Since truncated DcsC exchanges
2014, 98, 5363–5374.
This journal is © The Royal Society of Chemistry 2018
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
11 N. Bernardo-García, P. Sánchez-Murcia, F. Gago, F. Cava 24 T. Kumagai, Y. Koyama, K. Oda, M. Noda, Y. Matoba and
and J. A. Hermoso, Curr. Org. Chem., 2016, 20, 1222–1231.
12 P. Conti, L. Tamborini, A. Pinto, A. Blondel, P. Minoprio,
M. Sugiyama, Antimicrob. Agents Chemother., 2010, 54,
1132–1139.
A. Mozzarelli and C. De Micheli, Chem. Rev., 2011, 111, 25 D. Dietrich, M. J. van Belkum and J. C. Vederas, Org.
6919–6946. Biomol. Chem., 2012, 10, 2248–2254.
13 T. Lundqvist, S. L. Fisher, G. Kern, R. H. A. Folmer, Y. Xue, 26 M. P. Malakhov, M. R. Mattern, O. A. Malakhova,
D. T. Newton, T. A. Keating, R. A. Alm and B. L. M. de
Jonge, Nature, 2007, 447, 817–822.
M. Drinker, S. D. Weeks and T. R. Butt, J. Struct. Funct.
Genomics, 2004, 5, 75–86.
14 A. Buschiazzo, M. Goytia, F. Schaeffer, W. Degrave, 27 R. Assenberg, O. Delmas, S. C. Graham, A. Verma,
W. Shepard, C. Grègoire, N. Chamond, A. Cosson,
A. Berneman, N. Coatnoan, P. M. Alzari and P. Minoprio,
Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 1705–1710.
N. Berrow, D. I. Stuart, R. J. Owens, H. Bourhy and
J. M. Grimes, Acta Crystallogr., Sect. F: Struct. Biol. Cryst.
Commun., 2008, 64, 258–262.
15 L. Liua, K. Iwatab, M. Yohdab and K. Mikia, FEBS Lett., 28 D. R. Smyth, M. K. Mrozkiewicz, W. J. Mcgrath,
2002, 528, 114–118.
P. Listwan and B. Kobe, Protein Sci., 2003, 12, 1313–
1322.
29 C. D. Lee, H. C. Sun, S. M. Hu, C. F. Chiu, A. Homhuan,
S. M. Liang, C. H. Leng and T. F. Wang, Protein Sci., 2008,
17, 1241–1248.
16 R. C. Strauch, E. Svedin, B. Dilkes, C. Chapple and X. Li,
Proc. Natl. Acad. Sci. U. S. A., 2015, 112, 11726–11731.
17 C. W. Koo and J. S. Blanchard, Biochemistry, 1999, 38,
4416–4422.
18 B. Pillai, M. M. Cherney, C. M. Diaper, A. Sutherland, 30 M. E. Tanner and S. Miao, Tetrahedron Lett., 1994, 35,
J. S. Blanchard, J. C. Vederas and M. N. G. James, Proc.
Natl. Acad. Sci. U. S. A., 2006, 103, 8668–8673.
19 L. Hor, R. C. J. Dobson, C. Dogovski, C. A. Hutton and
M. A. Perugini, Acta Crystallogr., Sect. F: Struct. Biol. Cryst.
Commun., 2010, 66, 37–40.
4073–4076.
31 B. Pillai, M. M. Cherney, C. M. Diaper, A. Sutherland,
J. S. Blanchard, J. C. Vederas and M. N. G. James, Biochem.
Biophys. Res. Commun., 2007, 363, 547–553.
32 C. M. Diaper, A. Sutherland, B. Pillai, M. N. G. James,
P. Semchuk, J. S. Blanchard and J. C. Vederas, Org. Biomol.
Chem., 2005, 3, 4402–4411.
20 H.-Y. Sagong and K.-J. Kim, Sci. Rep., 2017, 7, 42318.
21 J. S. Park, W. C. Lee, J. H. Song, S. I. Kim, J. C. Lee,
C. Cheong and H.-Y. Kim, Acta Crystallogr., Sect. F: Struct. 33 C. W. Koo, A. Sutherland, J. C. Vederas and J. S. Blanchard,
Biol. Cryst. Commun., 2013, 69, 42–44. J. Am. Chem. Soc., 2000, 122, 6122–6123.
22 M. H. Gelb, Y. Lin, M. A. Pickard, Y. Song and J. C. Vederas, 34 H. Liu, V. R. Pattabiraman and J. C. Vederas, Org. Lett.,
J. Am. Chem. Soc., 1990, 112, 4932–4942. 2007, 9, 4211–4214.
23 B. Pillai, V. A. Moorthie, M. J. van Belkum, S. L. Marcus, 35 L. Hor, R. C. J. Dobson, M. T. Downton, J. Wagner,
M. M. Cherney, C. M. Diaper, J. C. Vederas and
M. N. G. James, J. Mol. Biol., 2009, 385, 580–594.
C. A. Hutton and M. A. Perguini, J. Biol. Chem., 2013, 288,
9238–9248.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2018