Potent inhibition of mandelate racemase by a fluorinated substrate-product analogue with a novel binding mode

Article abstract of DOI:10.1021/bi401703h

Mandelate racemase (MR) from Pseudomonas putida catalyzes the Mg ²⁺-dependent 1,1-proton transfer that interconverts the enantiomers of mandelate. Because trifluorolactate is also a substrate of MR, we anticipated that replacing the phenyl rings of the competitive, substrate-product analogue inhibitor benzilate (K_i = 0.7 mM) with trifluoromethyl groups might furnish an inhibitor. Surprisingly, the substrate-product analogue 3,3,3-trifluoro-2-hydroxy-2-(trifluoromethyl)propanoate (TFHTP) was a potent competitive inhibitor [K_i = 27 ± 4 μM; cf. K_m = 1.2 mM for both (R)-mandelate and (R)-trifluorolactate]. To understand the origins of this high binding affinity, we determined the X-ray crystal structure of the MR-TFHTP complex to 1.68 A resolution. Rather than chelating the active site Mg²⁺ with its glycolate moiety, like other ground state analogues, TFHTP exhibited a novel binding mode with the two trifluoromethyl groups closely packed against the 20s loop and the carboxylate bridging the two active site Bronsted acid-base catalysts Lys 166 and His 297. Recognizing that positioning a carboxylate between the Bronsted acid-base catalysts could yield an inhibitor, we showed that tartronate was a competitive inhibitor of MR (K_i = 1.8 ± 0.1 mM). The X-ray crystal structure of the MR-tartronate complex (1.80 A resolution) revealed that the glycolate moiety of tartronate chelated the Mg²⁺ and that the carboxylate bridged Lys 166 and His 297. Models of tartronate in monomers A and B of the crystal structure mimicked the binding orientations of (S)-mandelate and that anticipated for (R)-mandelate, respectively. For the latter monomer, the 20s loop appeared to be disordered, as it also did in the X-ray structure of the MR triple mutant (C92S/C264S/K166C) complexed with benzilate, which was determined to 1.89 A resolution. These observations indicate that the 20s loop likely undergoes a significant conformational change upon binding (R)-mandelate. In general, our observations suggest that inhibitors of other enolase superfamily enzymes may be designed to capitalize on the recognition of the active site Bronsted acid-base catalysts as binding determinants.

Full text of DOI:10.1021/bi401703h

Article
pubs.acs.org/biochemistry
Potent Inhibition of Mandelate Racemase by a Fluorinated
Substrate-Product Analogue with a Novel Binding Mode
Mitesh Nagar,† Adam D. Lietzan,‡ Martin St. Maurice,‡ and Stephen L. Bearne*,†,§
†Department of Biochemistry and Molecular Biology, Dalhousie University, Halifax, Nova Scotia B3H 4R2, Canada
‡Department of Biological Sciences, Marquette University, Milwaukee, Wisconsin 53201-1881, United States
§Department of Chemistry, Dalhousie University, Halifax, Nova Scotia B3H 4R2, Canada
ABSTRACT: Mandelate racemase (MR) from Pseudomonas
putida catalyzes the Mg2+-dependent 1,1-proton transfer that
interconverts the enantiomers of mandelate. Because trifluor-
olactate is also a substrate of MR, we anticipated that replacing
the phenyl rings of the competitive, substrate-product analogue
inhibitor benzilate (Ki = 0.7 mM) with trifluoromethyl groups
might furnish an inhibitor. Surprisingly, the substrate-product
analogue 3,3,3-trifluoro-2-hydroxy-2-(trifluoromethyl)-
propanoate (TFHTP) was a potent competitive inhibitor [Ki
= 27 4 μM; cf. Km = 1.2 mM for both (R)-mandelate and
(R)-trifluorolactate]. To understand the origins of this high binding affinity, we determined the X-ray crystal structure of the
MR−TFHTP complex to 1.68 Å resolution. Rather than chelating the active site Mg2+ with its glycolate moiety, like other
ground state analogues, TFHTP exhibited a novel binding mode with the two trifluoromethyl groups closely packed against the
20s loop and the carboxylate bridging the two active site Brønsted acid−base catalysts Lys 166 and His 297. Recognizing that
positioning a carboxylate between the Brønsted acid−base catalysts could yield an inhibitor, we showed that tartronate was a
competitive inhibitor of MR (Ki = 1.8 0.1 mM). The X-ray crystal structure of the MR−tartronate complex (1.80 Å resolution)
revealed that the glycolate moiety of tartronate chelated the Mg2+ and that the carboxylate bridged Lys 166 and His 297. Models
of tartronate in monomers A and B of the crystal structure mimicked the binding orientations of (S)-mandelate and that
anticipated for (R)-mandelate, respectively. For the latter monomer, the 20s loop appeared to be disordered, as it also did in the
X-ray structure of the MR triple mutant (C92S/C264S/K166C) complexed with benzilate, which was determined to 1.89 Å
resolution. These observations indicate that the 20s loop likely undergoes a significant conformational change upon binding (R)-
mandelate. In general, our observations suggest that inhibitors of other enolase superfamily enzymes may be designed to
capitalize on the recognition of the active site Brønsted acid−base catalysts as binding determinants.
similar to that exhibited for mandelate was unexpected
andelate racemase (MR, EC 5.1.2.2) from Pseudomonas
2+
Mputida catalyzes the Mg -dependent 1,1-proton transfer considering that MR binds lactate with low affinity (Ki ≈ 30
mM).5 Indeed, the binding affinity that MR exhibits for TFL
that interconverts the enantiomers of mandelate via a two-base
exceeds that expected on the basis of hydrophobic effects
mechanism with His 297 and Lys 166 abstracting the α-proton
from (R)-mandelate and (S)-mandelate, respectively, as shown
in Scheme 1.1 MR is very proficient at discriminating between
the substrate in the ground state and the altered substrate in the
transition state, binding the latter species with an association
constant equal to 5 × 1018 M−1 and stabilizing the transition
state of the reaction by 26 kcal/mol.2,3 Consequently, MR has
been studied as a paradigm for enzymes that catalyze rapid
carbon−hydrogen bond cleavage of carbon acids with relatively
high pKa values.1 Recently, we demonstrated that β,γ-
unsaturation is not a necessary requirement for catalysis and
that MR is capable of catalyzing the racemization of the
enantiomers of 3,3,3-trifluorolactate (TFL). Although kcat for
this substrate was reduced by ∼100-fold relative to that of
mandelate as a substrate, the values of Km were 1.7 and 1.2 mM
for (S)- and (R)-TFL, respectively, similar to the corresponding
values of 1.2 and 1.0 mM observed for (R)- and (S)-mandelate,
respectively.4 The fact that MR bound TFL with an affinity
alone.4
Previously, we observed that the substrate-product analogue
benzilate (Ki = 0.67 mM) is a competitive inhibitor of MR,
binding with an affinity that is similar to that exhibited for
mandelate.6 Consequently, we rationalized that an inhibitor of
MR might be generated by replacing the phenyl rings of
benzilate with trifluoromethyl groups. Herein, we describe the
potent inhibition of MR by the substrate-product analogue
3,3,3-trifluoro-2-hydroxy-2-(trifluoromethyl)propanoate
(TFHTP). Surprisingly, the X-ray crystal structure of MR with
bound TFHTP reveals that the inhibitor is bound with an
orientation that differs markedly from the binding orientation
observed previously for the substrate,7 analogues of the
Received: December 23, 2013
Revised: January 27, 2014
Published: January 28, 2014
© 2014 American Chemical Society
1169
dx.doi.org/10.1021/bi401703h | Biochemistry 2014, 53, 1169−1178

Biochemistry
Article
mL), water (3.4 mL), and cyclohexylamine (1.6 mL) were
added, and the solution was stirred for 1 h. After removal of the
solvent using rotary evaporation, a white powder remained.
This was dissolved in water (16 mL), and the solution was
filtered. Acetone (150 mL) was added, and a white precipitate,
the cyclohexylamine salt, formed upon sitting.14 After being
cooled for 1 h at 4 °C, the precipitate was collected using
suction filtration. The cyclohexylamine salt was then converted
to the sodium salt by passing it through a column containing
AG 50W-X8 (Na+ form). The eluent was collected, and
lyophilization yielded a white powder (0.85 g, 90%): mp 238−
Scheme 1. Reaction Catalyzed by Mandelate Racemase
(MR) and Structures of Inhibitors Discussed in the Text
1
241 °C (crystal became opaque at 90 °C); H NMR (500
1
MHz, D2O) δ 4.12 (m); 13C NMR (63 MHz, D2O, H-
decoupled) δ 67.60 (dq, 1JC,P = 151.3 Hz, 2JC,F = 30.6 Hz, CH),
1
2
124.61 (qd, JC,F = 280.9 Hz, JC,P = 5.6 Hz, CF3); 31P NMR
1
3
(101 MHz, D2O, H-decoupled) δ 11.88 (q, JF,P = 7.43 Hz);
19F NMR (235 MHz, D2O) δ −71.08 (dd, JF,H = 15.30 Hz,
3
3JF,P = 7.43 Hz); HRMS (ESI−) m/z calcd for C2H3F3O4P [M
− H]− 178.9721, found 178.9735.
(R,S)-1-Hydroxyethylphosphonate (1-HEP). Synthesis of
dimethyl 1-hydroxyethylphosphonate was conducted as de-
scribed by Chen et al.15 Dimethyl phosphate (1.10 g, 10 mmol)
and acetaldehyde (0.44 g, 10 mmol) were stirred together for
10 min at room temperature, followed by addition of potassium
fluoride and basic alumina (1:1 mass ratio, 3 g each). After the
mixture had been stirred for an additional 30 min, methylene
chloride (15 mL) was added and the resulting suspension was
filtered. From the filtrate, methylene chloride was evaporated
under reduced pressure, leaving 0.68 g of dimethyl 1-
hydroxyethylphosphonate as a viscous yellow liquid. The
methyl groups were removed from dimethyl 1-hydroxyethyl-
phosphonate by treatment with trimethylsilylbromide using a
procedure similar to that described by Freeman et al.13
Dimethyl 1-hydroxyethylphosphonate (0.49 g, 3.2 mmol) was
dissolved in dry acetonitrile (10 mL). Trimethylsilylbromide
(2.26 mL, 17.5 mmol, 5.0 equiv) was added dropwise to the
solution. The solution was refluxed for 1 h under an
atmosphere of argon, and the solvent was removed using
rotary evaporation under reduced pressure to yield a slightly
yellow liquid. Dioxane (3.4 mL), water (3.4 mL), and
cyclohexylamine (1.6 mL) were added, and the solution was
stirred for 1 h at room temperature. After removal of the
solvent using rotary evaporation, a slightly yellow liquid was
obtained. This was dissolved in water (16 mL), and then
acetone (150−250 mL) was added until the cyclohexylamine
salt formed as a white precipitate. After being cooled overnight
at 4 °C, the precipitate was collected using suction filtration.
The cyclohexylamine salt was then converted to the sodium salt
by passing it through a column containing AG 50W-X8 (Na+
form). Fractions containing phosphonate were pooled and
lyophilized to obtain the sodium salt of 1-hydroxyethylphosph-
substrate,8−11 and intermediate analogues.12 Interestingly, the
higher binding affinity of TFHTP arises, in part, because the
carboxylate group forms a salt bridge between the Brønsted
acid−base catalysts His 297 and Lys 166. Recognition of this
novel binding mode led to our subsequent identification of
tartronic acid as an inhibitor of MR. We show that the glycolate
moiety of tartronic acid chelates the Mg2+ ion within the active
site in a manner similar to that of mandelate, while the second
carboxylate bridges the two Brønsted acid−base catalysts in a
manner similar to that of TFHTP.
MATERIALS AND METHODS
■
General. Tartronic acid was purchased from Alfa Aesar
(Ward Hill, MA). (R)-Mandelic acid, α-hydroxyisobutyric acid,
3,3,3-trifluoro-2-hydroxy-2-(trifluoromethyl)propanoic acid,
and all other reagents, unless mentioned otherwise, were
purchased from Sigma-Aldrich Canada Ltd. (Oakville, ON).
Nuclear magnetic resonance (NMR) analyses were conducted
at the Nuclear Magnetic Resonance Research Resource (NMR-
3) using either a Bruker/Tecmag AC-250 spectrometer, a
̈
Bruker AV-300 spectrometer, or a Bruker AV-500 spectrom-
̈
̈
eter. Chemical shifts (δ) for proton (1H), carbon (13C), and
phosphorous (31P) spectra are reported in parts per million
relative to the HOD signal for spectra obtained in D2O. High-
resolution (HR) electrospray ionization (ESI) mass spectra
(MS) were recorded using a Bruker microTOF Focus
̈
orthogonal ESI-TOF mass spectrometer instrument operating
in either positive or negative ion mode. Circular dichroism
(CD) assays were conducted using a JASCO (Easton, MD) J-
810 spectropolarimeter with a jacketed cell holder. Melting
points are uncorrected.
1
onate as a white powder (0.224 g, 41%): mp 223−225 °C; H
(R,S)-1-Hydroxy-2,2,2-trifluoroethylphosphonate
(TFHEP). The ethyl groups were removed from diethyl 1-
hydroxy-2,2,2-trifluoroethylphosphonate by treatment with
trimethylsilylbromide using a procedure similar to that
described by Freeman et al.13 Diethyl 1-hydroxy-2,2,2-
trifluoroethylphosphonate (1.0 g, 4.24 mmol) was dissolved
in 10 mL of dry acetonitrile. Trimethylsilylbromide (2.63 mL,
20.4 mmol, 4.8 equiv) was added dropwise to the solution. The
solution was refluxed for 1 h under an atmosphere of argon,
and the solvent was removed using rotary evaporation under
reduced pressure to yield a slightly yellow liquid. Dioxane (3.4
NMR (D2O, 500 MHz) δ 1.23 (dd, 3JH,H = 7.1 Hz, 3JH,P = 15.9
3
2
Hz, 3H, CH3), 3.74 (dq, JH,H = 7.1 Hz, JH,P = 5.9 Hz, 1H,
OCHP); 13C NMR (D2O, 126 MHz, H-decoupled) δ 17.10
1
(CH3), 64.72 (d, JC,P = 157.9 Hz, OCHP); 31P NMR (D2O,
1
202 MHz, H-decoupled) δ 20.94; 31P NMR (D2O, 202 MHz,
1
2
3
1H-coupled) δ 20.94 (dq, JH,P = 5.91 Hz, JH,P = 15.90 Hz);
HRMS (ESI+) m/z calcd for C2H6Na2O4P [M + H]+ 170.9799,
found 170.9800.
1
Concentration Correction for TFHEP and 1-HEP. H
NMR spectroscopy was used to correct for the presence of
1170
dx.doi.org/10.1021/bi401703h | Biochemistry 2014, 53, 1169−1178

Biochemistry
Article
waters of hydration and sodium ion stoichiometry and thereby
permit accurate estimation of the concentration of TFHEP and
1-HEP in solution. Equimolar solutions (∼20 mM based on the
theoretical molecular weight) of 1-HEP (170.01 g/mol) or
TFHEP (223.98 g/mol) and sodium acetate were prepared in 5
mL of D2O. 1H NMR spectra for these solutions were obtained,
and integrated signal intensities corresponding to the α-H and
α-CH3 of 1-HEP and the α-H of TFHEP were calculated with
reference to the signal intensities of the methyl protons of the
sodium acetate internal standard. The actual concentrations of
1-HEP and TFHEP were then calculated from the ratio of
signal intensities and the known concentration of sodium
acetate.
the enzyme preparation (≥99%) was checked using 12%
sodium dodecyl sulfate−polyacrylamide gel electrophoresis.
The tag was not removed from the enzyme.
Enzyme Assays. MR activity was assayed using a CD-based
assay by following the change in ellipticity of mandelate at 262
nm with a 1 cm light path (unless otherwise indicated) as
described by Sharp et al.17 All kinetic assays, including
inhibition experiments with TFHEP, 1-HEP, TFHTP,
tartronate, and α-hydroxyisobutyrate, were conducted at 25
°C in Na+-HEPES buffer (0.1 M, pH 7.5) containing MgCl2
(3.3 mM) and bovine serum albumin (BSA, 0.005%). The
concentration of (R)-mandelate for assays ranged from 0.25 to
20.0 mM. The concentrations of the inhibitors and MR used
were as follows: 0.81, 1.62, and 3.24 mM TFHEP (150.72 ng/
mL; using a cuvette with a 0.5 cm light path for 3.24 mM
TFHEP), 16.37, 32.74, and 65.49 mM 1-HEP (122.4 ng/mL;
using a cuvette with a 0.5 cm light path), 0.02, 0.05, and 0.10
mM TFHTP (129.8 ng/mL), 0.75, 1.50, and 3.00 mM
tartronate (131.45 ng/mL), and 3.5, 7.0, and 14.0 mM α-
hydroxyisobutyrate (141.55 ng/mL; using a cuvette with a 0.5
cm light path). The apparent kinetic constants Vmax and Km
were determined by fitting the initial velocity data to eq 1 using
nonlinear regression analysis and KaleidaGraph version 4.02
from Synergy Software (Reading, PA)
Site-Directed Mutagenesis. The C92S/C264S/K166C
triple mutant of MR (tmMR) was constructed by performing
three rounds of polymerase chain reaction-based site-directed
mutagenesis using the QuikChange Site-Directed Mutagenesis
Kit (Stratgene, La Jolla, CA) and following the protocols
described by the manufacturer. The pET52b(+)-wtMR
plasmid16 was used as a template to create the C264S MR
mutant plasmid, which was then used as a template in the next
round of site-directed mutagenesis to create a double-mutant
C92S/C264S MR plasmid. Finally, the K166C mutation was
introduced to generate tmMR using the double-mutant plasmid
as the template. Reactions were conducted using PfuTurbo
DNA polymerase (Bio Basic Inc., Markham, ON). The forward
(F) and reverse (R) synthetic deoxyoligonucleotide primers
used to construct the mutants were 5′-CGCAAAACGCTTC-
AGCCTGGCAGGTTATACGGG-3′ (F, C92S), 5′-CCCGT-
ATAACCTGCCAGGCTGAAGCGTTTTGCG-3′ (R, C92S),
5′-GCATTGAGCATCGGTGCAAGCCGGTTGGCTATG-
CC-3′ (F, C264S), 5′-GGCATAGCCAACCGGCTTGCACC-
GATGCTCAATGC-3′ (R, C264S), 5′-GCGGTTAAGACCT-
GCATCGGCTATCCGGC-3′ (F, K166C), and 5′-GCCGG-
ATAGCCGATGCAGGTCTTAACCGC-3′ (R, K166C). The
codons specifying the relevant amino acids are underlined, and
the altered bases are shown in bold. Potential mutant plasmids
were used to transform competent Escherichia coli DH5α cells.
DH5α cells were used for plasmid maintenance and for all
sequencing reactions. After each round of site-directed
mutagenesis, the mutant open reading frame was sequenced
using commercial automated DNA sequencing (Robarts
Research Institute, London, ON) to ensure that no other
alterations of the nucleotide sequence had been introduced. For
X-ray crystallography studies, tmMR was overexpressed in and
purified from E. coli BL21(DE3) cells as described previously
for wild-type MR.12 Replacement of Cys 92 and Cys 264 with
Ser residues did not significantly alter the activity of the enzyme
(M. Nagar and S. L. Bearne, unpublished observations).
Enzyme Purification. Recombinant MR from P. putida was
overproduced in and purified from E. coli BL21(DE3) cells
transformed with the pET-52b(+)-wtMR plasmid, a pET-
52b(+) plasmid (Novagen, Madison, WI) containing the MR
open reading frame.16 This construct encodes the MR gene
product (MASWSHPQFEKGALEVLFQGPGYHM1...MR) as a
fusion protein with an N-terminal StrepII tag (underlined; M1
represents the first amino acid of wild-type MR). The enzyme
was purified by affinity chromatography using Strep-Tactin
V
max[S]
vi =
Km + [S]
(1)
Competitive inhibition constants (Ki) were determined from
plots of the apparent Km/Vmax values versus inhibitor
concentration in accord with eq 2.18
V
max[S]
vi =
[I]
Km 1 +
+ [S]
(
)
Ki
(2)
All kinetic parameters were determined in triplicate, and
average values are reported. The reported errors are standard
deviations. The concentration of wild-type MR was determined
from its absorbance at 280 nm using an extinction coefficient of
53400 M−1 cm−1, which was calculated using the ProtParam
tool available on the ExPASy server (http://web.expasy.org/
protparam).19
Protein Crystallization. Crystals were grown by the
hanging-drop vapor diffusion method against a reservoir
volume of 500 μL. The protein solution and reservoir solution
were mixed in a 1:1 ratio to a final volume of 4 μL. Crystals
grew spontaneously at 21 °C and ∼50% humidity.
Cocrystallization of MR with TFHTP. For the MR crystals
grown in the presence of TFHTP, the reservoir solution
consisted of PEG 1500 [20% (w/v)], glycine (120 mM), KNO3
(70 mM), and Tris-HCl buffer (100 mM, pH 8.0). The protein
solution consisted of 5.9 mg/mL MR purified as described
above, MgCl2 (3.3 mM), TFHTP (1 mM), and HEPES buffer
(50 mM, pH 7.5). The resulting cubelike crystals (∼100 μm ×
60 μm × 50 μm) grew to full size within 10−20 days. Crystals
were harvested and transferred to a synthetic stabilizing
solution consisting of PEG 1500 [24% (w/v)], glycine (132
mM), KNO3 (70 mM), TFHTP (1 mM), and Tris-HCl buffer
(80 mM, pH 8.0). These stabilized crystals were equilibrated in
the synthetic stabilizing solution for 5−10 min and then were
transferred directly to a cryoprotectant solution consisting of
PEG 1500 [40% (w/v)], glycine (138 mM), KNO3 (70 mM),
TFHTP (1 mM), and Tris-HCl buffer (100 mM, pH 8.0). The
Superflow resin (IBA GmbH, Gottingen, Germany) as
̈
described previously.16 Upon elution from the column, the
enzyme was dialyzed against storage buffer {HEPES buffer (100
mM, pH 7.5) containing MgCl2 (3.3 mM), NaCl (200 mM),
and glycerol [10% (v/v)]} and stored at −20 °C. The purity of
1171
dx.doi.org/10.1021/bi401703h | Biochemistry 2014, 53, 1169−1178

Biochemistry
Article
a
Table 1. Data Collection and Refinement Statistics
MR and tartronate
4M6U
MR and TFHTP
4FP1
tmMR and benzilate
PDB entry
4HNC
space group
I422
I422
I422
cell dimensions
a, b, c (Å)
149, 149, 178
90, 90, 90
149, 149, 170
90, 90, 90
149, 149, 170
90, 90, 90
α, β, γ (deg)
resolution range (Å)
redundancy
50.0−1.80 (1.83−1.80)
10.1 (8.7)
50.0−1.68 (1.71−1.68)
7.5 (5.6)
50.0−1.89 (1.92−1.89)
7.6 (5.5)
completeness (%)
no. of unique reflections
Rmerge (%)
98.8 (99.8)
90212
99.9 (99.3)
107280
99.5 (98.9)
75403
7.3 (37.6)
7.7 (39.8)
7.6 (40.0)
average I/σ
29.5 (5.0)
24.2 (4.5)
21.5 (3.4)
Refinement
32.38−1.80 (1.85−1.80)
0.147 (0.200)
0.171 (0.264)
5522
resolution range (Å)
Rcryst
50.0−1.68 (1.72−1.68)
50.0−1.89 (1.94−1.89)
0.169 (0.209)
0.193 (0.233)
5974
0.170 (0.207)
0.198 (0.228)
5912
Rfree
no. of protein atoms
no. of water molecules
Wilson B value (Å2)
average B factor (Å2)
protein
637
445
504
13.6
16.5
16.8
19.7
23.2
10.4
28.8
16.0
31.6
15.3
23.8
17.2
18.9
12.8
25.6
ligands
Mg2+
solvent
Ramachandran plot (%)
most favored
90.8
8.4
90.7
8.7
90.9
8.4
additionally allowed
generously allowed
disallowed
0.3
0.3
0.3
0.5
0.3
0.3
root-mean-square deviation
bond lengths (Å)
bond angles (deg)
0.017
1.754
0.014
1.553
0.013
1.460
a
Values in parentheses are for the highest-resolution bin.
cryoprotected crystals were flash-cooled in a nitrogen gas
stream at 100 K.
consisted of 6.6 mg/mL tmMR purified as described above,
MgCl2 (3.3 mM), benzilate (20 mM), and HEPES buffer (50
mM, pH 7.5). The resulting cubelike crystals (∼160 μm × 90
μm × 120 μm) grew to full size within 10−20 days. Crystals
were harvested and transferred directly to a synthetic stabilizing
solution consisting of PEG 1500 [24% (w/v)], glycine (165
mM), sodium tartronate (20 mM), NaCl (50 mM), and
HEPES buffer (80 mM, pH 7.5). These stabilized crystals were
subsequently transferred to a solution containing PEG 1500
[42% (w/v)], glycine (175 mM), benzilate (20 mM), NaCl (60
mM), and HEPES buffer (100 mM, pH 7.5), incubated for 5
min, and then flash-cooled in a nitrogen gas stream at 100 K.
Data Collection, Structure Determination, and Refine-
ment. X-ray diffraction data for MR cocrystallized with
tartronate were collected at Advanced Photon Source (APS)
beamline LS-CAT-21-ID-G on a Rayonix MarMosaic 300 CCD
detector, with an X-ray wavelength of 0.979 Å. X-ray diffraction
data for MR cocrystallized with TFHTP were collected at APS
beamline LS-CAT-21-ID-D on a Rayonix MarMosaic 300 CCD
detector, with an X-ray wavelength of 1.13 Å. X-ray diffraction
data for tmMR cocrystallized with benzilate were collected at
APS beamline LS-CAT-21-ID-F on a Rayonix MarMosaic 225
CCD detector, with an X-ray wavelength of 0.979 Å. Diffraction
images were processed using HKL2000.20 The structures were
determined by molecular replacement using the wild-type MR
enzyme [Protein Data Bank (PDB) entry 2MNR] as the search
model with Phaser.21 The molecular replacement models were
Cocrystallization of MR with Tartronate. For the MR
crystals grown in the presence of tartronate, the reservoir
solution consisted of PEG 1500 [14% (w/v)], glycine (100
mM), and triethanolamine buffer (100 mM, pH 8.5). The
protein solution consisted of 6.0 mg/mL MR purified as
described above, MgCl2 (3.3 mM), sodium tartronate (20
mM), and HEPES buffer (50 mM, pH 7.5). The resulting
cubelike crystals (∼70 μm × 70 μm × 50 μm) grew to full size
within 20−30 days. Approximately 50 days after crystallization
had been initiated, the crystals were harvested and transferred
to a synthetic stabilizing solution consisting of PEG 1500 [19%
(w/v)], glycine (110 mM), sodium tartronate (20 mM, pH
7.5), and triethanolamine buffer (100 mM, pH 8.5). These
stabilized crystals were equilibrated in the synthetic stabilizing
solution for 5−10 min and then were transferred directly to a
cryoprotectant solution consisting of PEG 1500 [39.5% (w/v)],
glycine (110 mM), sodium tartronate (20 mM), and triethanol-
amine buffer (100 mM, pH 8.5). After being equilibrated in the
cryoprotectant solution for 7 min, the crystals were flash-cooled
in a nitrogen gas stream at 100 K.
Cocrystallization of tmMR with Benzilate. Crystals of the
MR triple mutant (C92S/C264S/K166C, tmMR) were grown
in the presence of benzilate. The reservoir solution consisted of
PEG 1500 [15% (w/v)], glycine (150 mM), NaCl (50 mM),
and HEPES buffer (100 mM, pH 7.5), and the protein solution
1172
dx.doi.org/10.1021/bi401703h | Biochemistry 2014, 53, 1169−1178

Biochemistry
Article
Table 2. Binding Constants for Substrates, Ground State Analogues, and Transition State Analogues
R1R2R3COH
a
R1
R2
-H
R3
ligand
Ki (mM)
Substrates and Substrate Analogues
(R)-mandelate
b
-C6H5
-CF3
-COO−
-COO−
1.2 0.2
1.0 0.1
1.2 0.2
b
(S)-mandelate
b
-H
(R)-trifluorolactate [(R)-TFL]
(S)-trifluorolactate [(S)-TFL]
tartronate
b
1.74 0.08
1.8 0.1
-COO−
-CH3
-H
-H
-COO−
-COO−
c
(R)-lactate
32.0 1.4
c
(S)-lactate
26.1 3.1
Substrate-Product Analogues
benzilate
d
-C6H5
-CF3
-C6H5
-CF3
-COO−
-COO−
-COO−
0.67 0.12
3,3,3-trifluoro-2-hydroxy-2-(trifluoromethyl)propanoate (TFHTP)
α-hydroxyisobutyrate
0.027 0.004
5.5 0.6
-CH3
-CH3
Transition State Analogues
(R,S)-α-hydroxybenzylphosphonate
(R,S)-2,2,2-trifluoro-1-hydroxyethylphosphonate (TFHEP)
(R,S)-1-hydroxyethylphosphonate (1-HEP)
e
2−
-C6H5
-CF3
-H
-H
-H
-PO3
0.0047 0.0007
0.67 0.13
2−
-PO3
2−
-CH3
-PO3
40.2 4.8
a
Unless indicated otherwise, values are competitive inhibition constants. Values are means of triplicate trials, and reported errors are the standard
b
c
d
e
deviations. Km values from ref 4. From ref 5. From ref 6. From ref 25.
extended by several rounds of manual model building with
COOT22 and refinement with REFMAC23 using a geometric/
X-ray weighting term of 0.2. Noncrystallographic restraints
between each monomer were applied for the first round of
refinement but were relieved for subsequent rounds. Water
molecules were added to the model in COOT with subsequent
manual verification. Ligand coordinates for TFHTP were
generated and optimized for structure refinement using
eLBOW (electronic ligand building and optimization work-
bench).24 Data collection and processing statistics are listed in
Table 1.
benzohydroxamate (BzH; Ki = 9.3 μM),25 N-hydroxyformani-
lide (Ki = 2.8 μM),26 and Cupferron (Ki = 2.7 μM).26
Inhibition by Transition State/Intermediate Ana-
logues. Because trifluoromethyl groups imbued ground state
analogues with higher binding affinity, we explored the
possibility of incorporating a trifluoromethyl group into the
structure of a transition state/intermediate analogue inhibitor.
Previously, we demonstrated that (R,S)-α-hydroxybenzyl-
phosphonate was a potent inhibitor of MR (Table 2).25
Therefore, we prepared (R,S)-TFHEP as an analogue of the
putative aci-carboxylate intermediate formed during the
racemization of trifluorolactate.4 Interestingly, MR bound
(R,S)-TFHEP with an affinity that was only ∼2-fold greater
than the affinity exhibited for TFL (Table 2), in contrast to the
∼230-fold higher affinity exhibited for (R,S)-α-hydroxybenzyl-
phosphonate relative to mandelate. Although it appears that, in
this case, the trifluoromethyl group does not make a significant
contribution to the enhanced binding affinity, such is not the
case. First, the overall binding affinity of TFHEP is expected to
be reduced relative to that of α-hydroxybenzylphosphonate
because of the lower transition state stabilization energy
afforded by MR during the racemization of the corresponding
fluorinated substrate. The transition state stabilization afforded
by MR when mandelate is the substrate is −26 kcal/mol,2 while
the transition state stabilization is reduced to −20 kcal/mol
when TFL is the substrate.4 Consequently, one would expect
MR to bind a transition state/intermediate analogue based on
the structure of the aci-carboxylate of TFL with approximately
2.5 × 104-fold less affinity than it would bind the corresponding
transition state/intermediate analogue based on the structure of
the aci-carboxylate of mandelate. The observed decrease in
affinity for (R,S)-TFHEP (Ki = 0.67 mM), relative to that for
(R,S)-α-hydroxybenzylphosphonate (Ki = 4.7 μM), is in accord
with this expectation, although the magnitude of the decrease in
binding affinity (∼143-fold) is certainly much smaller than
expected. Second, the nonfluorinated analogue of (R,S)-
TFHEP, i.e., (R,S)-1-HEP, is bound with an affinity that is
60-fold lower than that exhibited for the corresponding
fluorinated intermediate/transition state analogue. Thus, the
presence of the trifluoromethyl group in the phosphonate-
RESULTS AND DISCUSSION
■
Inhibition by Ground State Analogues. MR binds the
enantiomers of mandelate and TFL with similar affinities
(Table 2). However, the enzyme binds the enantiomers of
lactate with approximately 23-fold less affinity than it exhibits
for mandelate and TFL. This observation indicates that the
trifluoromethyl group on the glycolate moiety contributes an
additional ∼1.8 kcal/mol to the free energy of binding relative
to the methyl group. On the basis of the similar binding
affinities that MR exhibits for mandelate and TFL, it appears
that the binding determinants within the hydrophobic pocket of
MR’s active site bind the phenyl ring of mandelate and the
trifluoromethyl group of TFL with approximately equal affinity.
Because MR binds the substrate-product analogue benzilate
with an affinity that is similar to that exhibited for mandelate,
we hypothesized that MR should bind TFHTP with similar
affinity. To our surprise, MR bound TFHTP with an affinity (Ki
= 27 μM) that exceeded that exhibited for mandelate and TFL
by ∼54-fold (Table 2). The corresponding nonfluorinated
analogue, α-hydroxyisobutyrate, however, was bound only
weakly by MR (Ki = 5.5 mM), indicating that the presence
of the six fluorine atoms in TFHTP increased the binding
affinity by 204-fold and contributed an additional 3.1 kcal/mol
to the free energy of binding. TFHTP is the most potent
ground state analogue inhibitor of MR reported to date,
binding with an affinity that is only slightly lower than those
reported for the binding of the transition state/intermediate
analogues (R,S)-α-hydroxybenzylphosphonate (Ki = 4.7 μM),25
1173
dx.doi.org/10.1021/bi401703h | Biochemistry 2014, 53, 1169−1178

Biochemistry
Article
based transition state/intermediate analogue does enhance
binding compared to that observed for the corresponding
nonfluorinated analogue. Such enhanced binding arising from
substitution of a trifluoromethyl group for a methyl group is
not unexpected. For example, α-methylacyl-CoA racemase
binds the α-methyl group of ligands at a hydrophobic pocket
within the enzyme’s active site,27,28 and replacement of the α-
methyl group in α-methylmyristoyl-CoA with an α-trifluor-
omethyl group leads to a 152-fold greater binding affinity.29
Sructure of MR Complexed with TFHTP. The X-ray
crystal structure of wild-type MR complexed with TFHTP was
determined to 1.68 Å resolution. As with previous crystal
structures of MR, the enzyme crystallized in space group I422
and the asymmetric unit consists of two tightly associating
monomers.7−11 Each monomer of the stable dimer contributes
residues to the neighboring active site, and the dimer forms the
biologically active octamer through crystallographic symmetry.
There are no significant structural changes observed for the
ligand-bound structure8 relative to the ligand-free structure
(Figure 1A).
The electron density for the TFHTP ligand varies between
active sites, and while both active sites contain acceptable B
factors, the electron density for the ligand is most complete for
monomer B (Figure 2A). This discrepancy in ligand binding
appears to stem from the inherent disorder associated with a
segment of the N-terminal capping domain (Asn 15−Val 35),
i.e., the 20s loop, that closes over the active site, constricting the
active site cavity and enhancing coordination of the ligand with
the metal center.12 The electron density for the 20s loop of
monomer A is weak relative to that of monomer B, and
therefore, the observed electron density for TFHTP is more
poorly defined in the active site of monomer A. The
descriptions of ligand binding are, consequently, limited to
the active site of monomer B.
TFHTP does not bind in the expected “catalytic” orientation,
with the glycolate moiety chelating the Mg2+ ion as observed
previously in the complexes of MR with (S)-mandelate and (S)-
atrolactate.7,9−11 Instead, TFHTP interacts indirectly with the
active site Mg2+ through a bridging water molecule (Figure 2A).
The surprisingly high binding affinity of TFHTP must,
therefore, arise for reasons other than its ability to chelate the
active site metal ion. First, the two trifluoromethyl groups are
intimately packed within the hydrophobic pocket, packing
against residues of the closed active site flap (Leu 18, Val 22,
Thr 24, and Val 29), residues within the α/β-barrel (Phe 52
and Tyr 54), Leu 298 (on a loop between the strand β10 and
α-helix8), residues at the end of the β11 strand (Leu 319 and
Leu 321), and Leu 93 from the adjacent monomer within the
dimer (Figure 2D). Second, the carboxylate group is locked in
place by several H-bonds (Figure 2E). Asn 197 and Lys 166
form H-bonds ∼3.2 Å in length to one oxygen of the
carboxylate. Nε2 of His 297 interacts with the other oxygen of
the carboxylate group via an H-bond of 2.6 Å, and Glu 317
interacts with the same carboxylate oxygen via a bridging water
molecule (not shown). Thus, rather than chelating the Mg2+
ion, the carboxylate of TFHTP forms a bridge between the two
active site Brønsted acid−base catalysts His 297 and Lys 166.
As a consequence of its binding orientation, the carboxylate
group assumes a binding orientation such that the plane of the
carboxylate group is roughly perpendicular to the orientation of
the hydroxamate moiety of BzH (Figure 2E). In addition, Nε2
of His 297 interacts with the α-OH group of TFHTP via an H-
bond of 3.3 Å. Considering that the pKa of the hydroxyl group
Figure 1. Alignment of the structures of the MR−TFHTP, MR−
tartronate, and tmMR−benzilate complexes with ligand-free MR (PDB
entry 2MNR8). The structures of the MR−TFHTP (A), MR−
tartronate (B), and tmMR−benzilate (C) complexes exhibit no
appreciable structural changes with respect to the structure of ligand-
free MR. The root-mean-square deviation is 0.3 Å upon alignment of
the Cα atoms of the biological dimer from MR−TFHTP (monomer
A, green; monomer B, cyan) and MR−tartronate (monomer A,
yellow; monomer B, purple) complexes with the equivalent dimer
from ligand-free MR (maroon). Alignment of the dimer from the
tmMR−benzilate complex (monomer A, light blue; monomer B, blue)
with the equivalent dimer from ligand-free MR (maroon) reveals that
the 20s loop of the N-terminal capping domain (green) is not closed
over the active site. In all structures, the divalent metal ion is shown in
space-filling representation (olive) and the ligands are shown in stick
representation.
1174
dx.doi.org/10.1021/bi401703h | Biochemistry 2014, 53, 1169−1178

Biochemistry
Article
Figure 2. Active site architecture of the MR−TFHTP, MR−benzilate, and tmMR−benzilate complexes. Representative simulated annealing omit
maps for the Mg2+−TFHTP (A) and Mg2+−tartronate (B) complexes in the active site of MR and the Mg2+−benzilate complex in tmMR (C) are
shown. The electron density maps are contoured at 3.0σ and extend to a distance of 5 Å from the ligand. (D) A stereoview shows the close packing
of the trifluoromethyl groups of TFHTP within the hydrophobic cavity of the active site of MR. The two active site Brønsted acid−base catalysts Lys
166 and His 297, Mg2+ (labeled M), water (labeled W), and TFHTP are shown in space-filling representation (with carbon, oxygen, nitrogen,
fluorine, and Mg2+ atoms colored cyan, red, blue, “limon”, and gray-green, respectively). Hydrophobic residues are shown in stick representation,
including Leu 93 from the adjacent subunit within the dimer (green). The purple contour shows the van der Waals surface of the hydrophobic
residues. (E) A stereoview shows the superposition of the structures of the MR complexes with bound benzohydroxamate (BzH, purple; PDB entry
3UXK12), TFHTP (green), and tartronate (orange). The Mg2+ ion is shown in space-filling representation, while the ligands and Asn 197, His 297,
and Lys 166 are shown in stick representation. Putative hydrogen bonds are shown as dashes, and the measurements are in angstroms.
of TFHTP should be markedly reduced relative to that of
aliphatic alcohols (pKa values are 17.1 for 2-propanol30 and 9.3
for hexafluoro-2-propanol,31 bearing in mind that a hydro-
phobic environment may increase the pKa from that expected
in aqueous solution), it is plausible that the enhanced binding
of TFHTP may also arise, in part, from the transfer of the
proton from the hydroxyl group to His 297. Such a scenario is
reminiscent of the time-of-flight neutron structure of the D-
xylose isomerase−product complex reported by Langan and co-
workers32 in which O5 of D-xylulose was not protonated but
formed an H-bond (3.07 Å) with a doubly protonated His
residue. Hence, it is the effect of the optimal hydrophobic
packing interactions “pulling” the trifluoromethyl groups of
TFHTP toward the active site flap, the steric bulk of the
trifluoromethyl groups that resides outside the plane of the
phenyl ring of BzH, and the bridging interactions between the
active site bases that produce the higher binding affinity of
TFHTP even though the glycolate moiety failed to chelate the
Mg2+ ion. Consequently, the Mg2+ ion is coordinated by a water
molecule (Figure 2A) rather than the α-OH group and
carboxylate of TFHTP.
Inhibition by Tartronate. Recognizing that binding of a
ligand could be enhanced when a ligand positions a carboxylate
between the active site Brønsted acid−base catalysts, we
anticipated that tartronate might be an inhibitor of MR.
Tartronate is a structural mimic of mandelate in which the
glycolate moiety is retained but the phenyl ring is replaced with
a carboxylate group (Scheme 1). Interestingly, tartronate
inhibited MR with a binding affinity that was similar to that
observed for mandelate (Table 2), despite having the polar
carboxylate group in place of the nonpolar phenyl ring. In fact,
the 15−18-fold difference in binding affinities between lactate
and tartronate reveals that the carboxylate group of tartronate
contributes an additional 1.7 kcal/mol in binding energy
relative to the methyl group of lactate. This observation
suggests that inhibition by tartronate does indeed arise, in part,
through interaction of the carboxylate with the active site
Brønsted acid−base catalysts.
Structure of MR Complexed with Tartronate. The X-
ray crystal structure of wild-type MR complexed with tartronate
was determined to 1.80 Å resolution. Unlike TFHTP,
tartronate is bound in a substratelike orientation, with the
1175
dx.doi.org/10.1021/bi401703h | Biochemistry 2014, 53, 1169−1178

Biochemistry
Article
Figure 3. Tartronate and benzilate bound in the active site. Bound tartronate in monomer A (A) and monomer B (B) are shown with electron
density maps contoured at 3.0σ, and extending to a distance of 5 Å from each ligand. The orientation of tartronate in monomer A may mimic that of
(R)-mandelate, while the orientation of tartronate in monomer B mimics that of (S)-mandelate. (C) An overlay of bound tartronate in monomers A
(green) and B (yellow), centered on the active site, shows the distances between the α-carbon and His 297 and Lys 166, and (D) an expanded view
shows the two conformations adopted by the 20s loop upon tartronate binding. In monomer A, the 20s loop adopts both an open (light green) and a
closed (green) conformation, while in monomer B, the 20s loop has a closed conformation (yellow). (E) A stereoview shows the superposition of
the structures of the MR−tartronate (monomer A) and tmMR−benzilate complexes. The 20s loop of the tmMR−benzilate complex (cyan) adopts
an open conformation, similar to that of the 20s loop in monomer A of the MR−tartronate complex [light green; cf. the closed conformation
(green)]. The Brønsted acid−base catalysts Lys 166 and His 297 are shown in stick representation, and Mg2+ (green) and a water molecule (red) are
shown in space-filling representation.
glycolate moiety chelating the Mg2+ ion in the active site
(Figure 2B). As predicted, the second carboxylate group
bridges the two active site Brønsted acid−base catalysts. The
carboxylate oxygens of tartronate are positioned 2.7 Å from His
297 and 4.0 Å from Lys 166 (Figure 2E). Interestingly, in all of
the ligand-bound structures of MR, Lys 166 assumes an
identical rotamer, irrespective of the identity of the bound
ligand or the H-bonding potential with different functional
groups on these ligands (Figure 2E). Nevertheless, the
structure supports our hypothesis that the greater than
expected binding affinity of tartronic acid results from the
interaction of the second carboxylate group with the two
Brønsted acid−base catalysts.
opposing orientations (Figure 3). The bound tartronate in
monomer B is best modeled with the α-carbon in the position
that would be occupied by the α-carbon of (S)-mandelate,
while in monomer A, the bound tartronate is best modeled with
the α-carbon in the position that we anticipate would be
occupied by the α-carbon of (R)-mandelate (Figure 3A,B). The
orientation of tartronate in monomer A, therefore, may offer
insight into how (R)-mandelate is oriented in the active site,
because all attempts to cocrystallize MR with (R)-mandelate or
(R)-mandelate analogues have, thus far, been unsuccessful or
have caused crystals to crack in soaking-in experiments
(unpublished observations). An overlay of chains A and B,
centered on the active site, reveals the difference in the α-
carbon position adopted by tartronate in the two active sites
(Figure 3C). The difference in the α-carbon position is
consistent with our previous conclusions that the positions of
the functional groups chelating Mg2+ remain relatively fixed
While the α-carbon of tartronic acid is not a stereogenic
center, the electron densities observed for the ligands in
monomers A and B of the structure of MR cocrystallized with
tartronate suggest that the ligand binds in the active site in two
1176
dx.doi.org/10.1021/bi401703h | Biochemistry 2014, 53, 1169−1178

Biochemistry
Article
during catalysis and that the α-carbon and phenyl group move
during racemization.6,12 The two Brønsted acid−base catalysts
His 297 and Lys 166 occupy identical positions in these two
active sites (Figure 3C). However, exclusively in monomer A
where the α-carbon of tartronate mimics the (R)-mandelate
orientation, there are two conformations for the 20s loop that
forms part of the hydrophobic cavity within the active site. The
20s loop of monomer A adopts both the open and closed
conformations, each at 50% occupancy, while the 20s loop in
monomer B fully occupies the closed conformation (Figure
3D). The difficulties associated with cocrystallizing and/or
soaking the enzyme with (R)-mandelate or corresponding
analogues suggest that there may be considerable conforma-
tional changes associated with the binding of the substrate in
the R stereochemistry.
Structure of the tmMR−Benzilate Complex. As
mentioned above, structures of MR with (R)-mandelate or
(R)-mandelate analogues are not yet available; hence, “snap-
shots” of the entire reaction catalyzed by MR are also not
available. We reasoned that benzilate, a substrate-product
analogue,6 might be more amenable to crystallization and
would provide insights into the orientation and position of the
aromatic rings corresponding to both (R)- and (S)-mandelate,
recognizing, of course, that the 7.128 Å separation between the
para carbon atoms of the phenyl rings observed in the crystal
structure of potassium benzilate33 probably exceeds the
distance that the para carbon atom of mandelate moves during
MR-catalyzed racemization (∼1.6−2.4 Å) and that the ipso
carbon remains relatively fixed during MR-catalyzed racemiza-
tion.6,12 Unfortunately, despite several cocrystallization and
soaking experiments, we were unable to obtain the structure of
benzilate within the active site of wild-type MR, possibly
because the hydrophobic pocket cannot accommodate two
aromatic rings in a conformation that remains amenable to
crystallization. However, as part of our ongoing studies of MR,
we had prepared the C92S/C264S/K166C triple mutant of MR
(tmMR). We anticipated that truncation of the side chain of
Lys 166 might afford additional room in the active site,
permitting benzilate to bind in a catalytically relevant
orientation. The X-ray crystal structure of tmMR complexed
with benzilate was determined to 1.89 Å resolution. As
observed for the previously described wild-type structures,
tmMR crystallized in space group I422 with the asymmetric
unit consisting of two monomers associating as a biological
dimer. The overall structure of tmMR remains largely
unperturbed when it is aligned with the ligand-free structure
(Figure 1C).
aromatic ring. Specifically, when a ligand positions its α-carbon
and α-proton like those of (R)-mandelate (i.e., tartonate in
monomer A of wild-type MR) or its aromatic ring in a position
expected to be similar to that of (R)-mandelate (i.e., one of the
two aromatic rings of benzilate in tmMR), the 20s loop adopts
a more open conformation (Figure 3E). Thus, the 20s loop of
MR likely undergoes a significant conformational change upon
binding (R)-mandelate.
CONCLUSIONS
■
At first blush, it is unlikely that one would conceive of either
TFHTP or tartronate as an inhibitor of MR. However, our
observations that MR can bind TFL as a substrate with an
affinity similar to that exhibited for mandelate4 and that MR is
capable of also binding the substrate-product analogue benzilate
with a similar affinity6 led us to anticipate that MR might be
inhibited by TFHTP. The fact that TFHTP was bound only 3-
fold less tightly than the most potent transition state analogue
inhibitors reported to date led us to examine the X-ray crystal
structure of the MR−TFHTP complex, which revealed an
unexpected binding orientation. The high binding affinity of
TFHTP arises from efficient packing of the two trifluoromethyl
groups within the hydrophobic pocket of the enzyme’s active
site and through the participation of the carboxylate group in a
bridge between the two active site Brønsted acid−base catalysts
His 297 and Lys 166. This observation suggests that inhibitors
may be designed to capitalize on the recognition of the active
site Brønsted acid−base catalysts as binding determinants, as
we showed for tartronate. Consequently, this inhibitor design
strategy should allow the development of inhibitors of other
enzymes of the enolase superfamily, such as those in the MR,
MLE, and MAL subgroups.34−36
ASSOCIATED CONTENT
■
Accession Codes
The atomic coordinates of the MR−3,3,3-trifluoro-2-hydroxy-2-
(trifluoromethyl)propanoic acid, MR−tartronate, and tmMR−
benzilate complexes have been deposited as Protein Data Bank
entries 4FP1, 4M6U, and 4HNC, respectively.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: sbearne@dal.ca. Phone: (902) 494-1974. Fax: (902)
494-1355.
As predicted, the removal of the side chain of Lys 166 in
tmMR permitted benzilate to bind in a substratelike orientation
similar to that of tartronate and the substrate analogue (S)-
atrolactate (Scheme 1),9 chelating the catalytic Mg2+ ion with
the α-hydroxyl oxygen and one of the carboxylate oxygens. The
20s loop of the N-terminal capping domain could be modeled
for only monomer A because the observed electron density for
this region in monomer B was irregular and discontinuous,
suggesting multiple conformations. Interestingly, the 20s loop
of the N-terminal capping domain of monomer A occupies the
same open conformation that was observed in monomer A of
the MR−tartronate structure described above (Figures 1C and
3E). Considered together, the structures of MR described in
this study suggest that the conformation of the 20s loop
depends on two features that are expected to be unique to (R)-
mandelate: the position of the α-carbon and the position of the
Funding
This work was supported by a Discovery Grant from the
Natural Sciences and Engineering Research Council (NSERC)
of Canada (S.L.B.) and National Science Foundation MRI-R2
Grant DBI-0959442 (M.St.M.). A.D.L. is supported by a
GAANN award (Graduate Assistance in Areas of National
Need) from the U.S. Department of Education. Use of the
Advanced Photon Source was supported by U.S. Department of
Energy, Office of Science, Office of Basic Energy Sciences,
under Contract DE-AC02-06CH11357. Use of LS-CAT Sector
21 was supported by the Michigan Economic Development
Corp. and the Michigan Technology Tri-Corridor for the
support of this research program (Grant 085P1000817).
Notes
The authors declare no competing financial interest.
1177
dx.doi.org/10.1021/bi401703h | Biochemistry 2014, 53, 1169−1178

Biochemistry
Article
(17) Sharp, T. R., Hegeman, G. D., and Kenyon, G. L. (1979) A
direct kinetic assay for mandelate racemase using circular dichroic
measurements. Anal. Biochem. 94, 329−334.
ABBREVIATIONS
■
BzH, benzohydroxamate; CD, circular dichroism; 1-HEP,
(R,S)-1-hydroxyethylphosphonate; HEPES, 4-(2-
hydroxyethyl)piperazine-1-ethanesulfonic acid; MR, mandelate
racemase; TFL, trifluorolactate; TFHEP, (R,S)-2,2,2-trifluoro-
1-hydroxyethylphosphonate; TFHTP, 3,3,3-trifluoro-2-hy-
droxy-2-(trifluoromethyl)propanoate.
(18) Segel, I. H. (1975) Enzyme Kinetics, John Wiley and Sons, Inc.,
New York, pp 100−111.
(19) Gasteiger, E., Gattiker, A., Hoogland, C., Ivanyi, I., Appel, R. D.,
and Bairoch, A. (2003) ExPASy: The proteomics server for in-depth
protein knowledge and analysis. Nucleic Acids Res. 31, 3784−3788.
(20) Otwinowski, Z., and Minor, W. (1997) Processing of X-ray
diffraction data collected in oscillation mode. Methods Enzymol. 276,
307−326.
(21) McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M.
D., Storoni, L. C., and Read, R. J. (2007) Phaser crystallographic
software. J. Appl. Crystallogr. 40, 658−674.
(22) Emsley, P., Lohkamp, B., Scott, W. G., and Cowtan, K. (2010)
Features and development of Coot. Acta Crystallogr. D66, 486−501.
(23) Vagin, A. A., Steiner, R. S., Lebedev, A. A., Potterton, L.,
McNicholas, S., Long, F., and Murshudov, G. N. (2004) REFMAC5
dictionary: Organisation of prior chemical knowledge and guidelines
for its use. Acta Crystallogr. D60, 2284−2295.
(24) Moriarty, N. W., Grosse-Kunstleve, R. W., and Adams, P. D.
(2009) Electronic Ligand Builder and Optimization Workbench
(eLBOW): A tool for ligand coordinate and restraint generation.
Acta Crystallogr. D65, 1074−1080.
(25) St. Maurice, M., and Bearne, S. L. (2000) Reaction intermediate
analogues for mandelate racemase: Interaction between Asn 197 and
the α-hydroxyl of the substrate promotes catalysis. Biochemistry 39,
13324−13335.
(26) Bourque, J. R., Burley, R. K., and Bearne, S. L. (2007)
Intermediate analogue inhibitors of mandelate racemase: N-Hydrox-
yformanilide and Cupferron. Bioorg. Med. Chem. Lett. 17, 105−108.
(27) Bhaumik, P., Schmitz, W., Hassinen, A., Hiltunen, J. K.,
Conzelmann, E., and Wierenga, R. K. (2007) The catalysis of the 1,1-
proton transfer by α-methylacyl-CoA racemase is coupled to a
movement of the fatty acyl moiety over a hydrophobic, methionine-
rich surface. J. Mol. Biol. 367, 1145−1161.
(28) Sharma, S., Bhaumik, P., Schmitz, W., Venkatesan, R., Hiltunen,
J. K., Conzelmann, E., Juffer, A. H., and Wierenga, R. K. (2012) The
enolization chemistry of a thioester-dependent racemase: The 1.4 Å
crystal structure of a reaction intermediate complex characterized by
detailed QM/MM calculations. J. Phys. Chem. B 116, 3619−3629.
(29) Carnell, A. J., Hale, I., Denis, S., Wanders, R. J., Isaacs, W. B.,
Wilson, B. A., and Ferdinandusse, S. (2007) Design, synthesis, and in
vitro testing of α-methylacyl-CoA racemase inhibitors. J. Med. Chem.
50, 2700−2707.
(30) Murto, J. (1964) Nucleophilic reactivity of alkoxide ions toward
2,4-dinitrofluorobenzene and the acidity of alcohols. Acta Chem. Scand.
18, 1043−1053.
(31) Dyatkin, B. L., Mochalina, E. P., and Knunyants, I. L. (1965)
The acidic properties of fluorine-containing alcohols, hydroxylamines
and oximes. Tetrahedron 21, 2991−2995.
(32) Kovalevsky, A. Y., Katz, A. K., Carrell, H. L., Hanson, L.,
Mustyakimov, M., Fisher, S. Z., Coates, L., Schoenborn, B. P., Bunick,
G., Glusker, J. P., and Langan, P. (2008) Hydrogen location in stages
of an enzyme-catalyzed reaction: Time-of-flight neutron structure of D-
xylose isomerase with bound D-xylulose. Biochemistry 47, 7595−7597.
(33) Rojas, L. S., Ramírez, B. M., Mora, A. J., Delgado, G. E., and de
Delgado, G. D. (2003) Redetermination of potassium benzilate. Acta
Crystallogr. E59, m647−m651.
(34) Gerlt, J. A., and Babbitt, P. C. (2001) Divergent evolution of
enzymatic function: Mechanistically diverse superfamilies and func-
tionally distinct suprafamilies. Annu. Rev. Biochem. 70, 209−246.
(35) Gerlt, J. A., Babbitt, P. C., and Rayment, I. (2005) Divergent
evolution in the enolase superfamily: The interplay of mechanism and
specificity. Arch. Biochem. Biophys. 433, 59−70.
REFERENCES
■
(1) Gerlt, J. A. (1998) Enzyme-catalyzed proton transfer reactions to
and from carbon. In Bioorganic Chemistry: Peptides and Proteins (Hecht,
S. M., Ed.) pp 279−311, Oxford University Press, New York.
(2) Bearne, S. L., and Wolfenden, R. (1997) Mandelate racemase in
pieces: Effective concentrations of enzyme functional groups in the
transition state. Biochemistry 36, 1646−1656.
(3) St. Maurice, M., and Bearne, S. L. (2002) Kinetics and
thermodynamics of mandelate racemase catalysis. Biochemistry 41,
4048−4058.
(4) Nagar, M., Narmandakh, A., Khalak, Y., and Bearne, S. L. (2011)
Redefining the minimal substrate tolerance of mandelate racemase.
Racemization of trifluorolactate. Biochemistry 50, 8846−8852.
(5) St. Maurice, M., and Bearne, S. L. (2004) Hydrophobic nature of
the active site of mandelate racemase. Biochemistry 43, 2524−2532.
(6) Siddiqi, F., Bourque, J. R., Jiang, H., Gardner, M., St. Maurice, M.,
Blouin, C., and Bearne, S. L. (2005) Perturbing the hydrophobic
pocket of mandelate racemase to probe phenyl motion during
catalysis. Biochemistry 44, 9013−9021.
(7) Kallarakal, A. T., Mitra, B., Kozarich, J. W., Gerlt, J. A., Clifton, J.
G., Petsko, G. A., and Kenyon, G. L. (1995) Mechanism of the
reaction catalyzed by mandelate racemase: Structure and mechanistic
properties of the K166R mutant. Biochemistry 34, 2788−2797.
(8) Neidhart, D. J., Howell, P. L., Petsko, G. A., Powers, V. M., Li, R.
S., Kenyon, G. L., and Gerlt, J. A. (1991) Mechanism of the reaction
catalyzed by mandelate racemase. 2. Crystal structure of mandelate
racemase at 2.5-Å resolution: Identification of the active site and
possible catalytic residues. Biochemistry 30, 9264−9273.
(9) Landro, J. A., Gerlt, J. A., Kozarich, J. W., Koo, C. W., Shah, V. J.,
Kenyon, G. L., Neidhart, D. J., Fujita, S., and Petsko, G. A. (1994) The
role of lysine 166 in the mechanism of mandelate racemase from
Pseudomonas putida: Mechanistic and crystallographic evidence for
stereospecific alkylation by (R)-α-phenylglycidate. Biochemistry 33,
635−643.
(10) Mitra, B., Kallarakal, A. T., Kozarich, J. W., Gerlt, J. A., Clifton, J.
G., Petsko, G. A., and Kenyon, G. L. (1995) Mechanism of the
reaction catalyzed by mandelate racemase: Importance of electrophilic
catalysis by glutamic acid 317. Biochemistry 34, 2777−2787.
(11) Schafer, S. L., Barrett, W. C., Kallarakal, A. T., Mitra, B.,
Kozarich, J. W., Gerlt, J. A., Clifton, J. G., Petsko, G. A., and Kenyon,
G. L. (1996) Mechanism of the reaction catalyzed by mandelate
racemase: Structure and mechanistic properties of the D270N mutant.
Biochemistry 35, 5662−5669.
(12) Lietzan, A. D., Nagar, M., Pellmann, E. A., Bourque, J. R.,
Bearne, S. L., and St. Maurice, M. (2012) Structure of mandelate
racemase with bound intermediate analogues benzohydroxamate and
Cupferron. Biochemistry 51, 1160−1170.
(13) Freeman, S., Irwin, W. J., and Schwalbe, C. H. (1991) Synthesis
and hydrolysis studies of phosphonopyruvate. J. Chem. Soc., Perkin
Trans. 2, 263−267.
(14) Plaga, W., Frank, R., and Knappe, J. (1988) Catalytic-site
mapping of pyruvate formate lyase. Eur. J. Biochem. 178, 445−450.
(15) Chen, T., Shen, P., Li, Y., and He, H. (2006) The synthesis and
herbicidal evaluation of fluorine-containing phenoxyacetoxyalkyl-
phosphonate derivatives. Phosphorus, Sulfur Silicon Relat. Elem. 181,
2135−2145.
(16) Narmandakh, A., and Bearne, S. L. (2010) Purification of
recombinant mandelate racemase: Improved catalytic activity. Protein
Expression Purif. 69, 39−46.
(36) Gerlt, J. A., Babbitt, P. C., Jacobson, M. P., and Almo, S. C.
(2012) Divergent evolution in the enolase superfamily: Strategies for
assigning functions. J. Biol. Chem. 287, 29−34.
1178
dx.doi.org/10.1021/bi401703h | Biochemistry 2014, 53, 1169−1178