Structure and Chemical Reaction Mechanism of LigU, an Enzyme That Catalyzes an Allylic Isomerization in the Bacterial Degradation of Lignin

Article abstract of DOI:10.1021/acs.biochem.9b00549

LigU from Novosphingobium sp. strain KA1 catalyzes the isomerization of (4E)-oxalomesaconate (OMA) to (3Z)-2-keto-4-carboxy-3-hexenedioate (KCH) as part of the protocatechuate (PCA) 4,5-cleavage pathway during the degradation of lignin. The three-dimensional structure of the apo form of the wild-type enzyme was determined by X-ray crystallography, and the structure of the K66M mutant enzyme was determined in the presence of the substrate OMA. LigU is a homodimer requiring no cofactors or metal ions with a diaminopimelate epimerase structural fold, consisting of two domains with similar topologies. Each domain has a central α-helix surrounded by a β-barrel composed of antiparallel β-strands. The active site is at the cleft of the two domains. ¹H nuclear magnetic resonance spectroscopy demonstrated that the enzyme catalyzes the exchange of the pro-S hydrogen at C5 of KCH with D₂O during the isomerization reaction. Solvent-deuterium exchange experiments demonstrated that mutation of Lys-66 eliminated the isotope exchange at C5 and that mutation of C100 abolished exchange at C3. The positioning of these two residues in the active site of LigU is consistent with a reaction mechanism that is initiated by the abstraction of the pro-S hydrogen at C3 of OMA by the thiolate anion of Cys-100 and the donation of a proton at C5 of the proposed enolate anion intermediate by the side chain of Lys-66 to form the product KCH. The 1,3-proton transfer is suprafacial.

Full text of DOI:10.1021/acs.biochem.9b00549

Subscriber access provided by UNIV OF SOUTHERN INDIANA
Article
Structure and Chemical Reaction Mechanism of LigU, an Enzyme that
Catalyzes an Allylic Isomerization in the Bacterial Degradation of Lignin
Tessily N Hogancamp, Seth A Cory, David P Barondeau, and Frank M. Raushel
Biochemistry, Just Accepted Manuscript • DOI: 10.1021/acs.biochem.9b00549 • Publication Date (Web): 24 Jul 2019
Downloaded from pubs.acs.org on July 27, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
Structure and Chemical Reaction Mechanism of LigU, an
Enzyme that Catalyzes an Allylic Isomerization in the
Bacterial Degradation of Lignin
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Tessily N. Hogancamp, Seth A. Cory,
David P. Barondeau*, and Frank M. Raushel*
Department of Chemistry, Texas A&M University,
College Station, Texas, 77843, USA
1
ACS Paragon Plus Environment

Biochemistry
Page 2 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
LigU from Novosphingobium sp. strain KA1 catalyzes the isomerization of (4E)-
oxalomesaconate (OMA) to (3Z)-2-keto-4-carboxy-3-hexenedioate (KCH) as part of the
protocatechuate (PCA) 4,5-cleavage pathway during the degradation of lignin. The three-
dimensional structure of the apo-form of the wild-type enzyme was determined by X-ray
crystallography and the structure of the K66M mutant enzyme was determined in the
presence of the substrate OMA. LigU is a homodimer requiring no cofactors or metal ions
with a diaminopimelate epimerase structural fold, consisting of two domains of similar
topology to one another. Each domain has a central -helix surrounded by a β-barrel
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
composed of antiparallel β-strands. The active site is at the cleft of the two domains. H
NMR spectroscopy demonstrated that the enzyme catalyzes the exchange of the proS
hydrogen at C5 of KCH with D O during the isomerization reaction. Solvent-deuterium
2
exchange experiments demonstrated that mutation of Lys-66 eliminated the isotope
exchange at C5 and that mutation of C100 abolished exchange at C3. The positioning of
these two residues in the active site of LigU is consistent with a reaction mechanism that is
initiated by the abstraction of the proS hydrogen at C3 of OMA by the thiolate anion of Cys-
1
00 and the donation of a proton at C5 of the proposed enolate anion intermediate by the
side chain of Lys-66 to form the product KCH. The 1,3-proton transfer is suprafacial.
2
ACS Paragon Plus Environment

Page 3 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Lignin, the most abundant renewable aromatic compound in the world, is currently
being pursued as a viable starting material for biopolymers. However, it is an underutilized
renewable resource due to its recalcitrant nature as a complex heteropolymer. Lignin is
comprised of phenylpropanoid units (guaiacyl, syringyl, and p-hydroxyphenyl precursors)
that couple to one another through C-C and C-O bonds (1, 2). The natural recycling of lignin
occurs through the combined actions of eukarya, archaea, and bacteria (3). In general,
fungi are responsible for the depolymerization of lignin, and bacteria further metabolize
the monomeric units into viable carbon sources. Many of the known enzymatic pathways
funnel to the intermediate protocatechuate (PCA) where it is further metabolized by three
cleavage pathways (4). In the PCA 4,5-cleavage pathway, the LigAB complex converts PCA
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
and O into 4-carboxy-2-hydroxymuconate-6-semialdehyde (CHMS) (5). LigC then oxidizes
2
+
CHMS with NADP to form 2-pyrone-4,6-dicarboxylate (PDC, 1) (6). LigI hydrolyzes PDC
into (4E)-oxalomesaconate (OMA, 2) (7). LigU isomerizes OMA into (3Z)-2-keto-4-carboxy-
3
-hexendioate (KCH, 3) (8). LigJ hydrates KCH to (S)-4-carboxy-4-hydroxy-2-oxoadipate
(CHA, 4) (9). Finally, LigK cleaves CHA into pyruvate (PYR, 5) and oxaloacetate (OAA, 6),
which are then shuttled into the TCA cycle (Scheme 1) (10, 11).
Previously, we demonstrated that LigU isomerase catalyzes the conversion of OMA
to KCH via a 1,3-allylic isomerization of the double bond from C4/C5 to C3/C4. The
structural characterization of LigJ hydratase complexed with the substrate KCH (3)
provided evidence supporting the migration of the double bond from C4/C5 to C3/C4 prior
to the addition of water to C4 of KCH (8).
3
ACS Paragon Plus Environment

Biochemistry
Page 4 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 1. Reactions catalyzed by LigI, LigU, LigJ and LigK.
LigU belongs to cog2828, which is also known as the PrpF superfamily. PrpF is
reported to catalyze the isomerization of 4-methyl-cis-aconitate to 2-methyl-cis-aconitate,
an intermediate in the 2-methylcitric acid cycle (12, 13). Structurally, this family shares a
common protein fold, which was first discovered for diaminopimelate epimerase (14),
where there is an apparent domain duplication event and the active site is at the junction of
the two domains (15, 16). The sequence similarity network (SSN) for cog2828 at an E
value cutoff of 10-80 illustrates the sequence relationship among the 158 nonredundant
protein sequences using a 40% identity threshold (Figure S1).
Several aspects of the reaction mechanism for LigU and the PrpF enzyme family
remain unknown. To provide mechanistic details of the LigU catalyzed isomerization
reaction, we have determined the three-dimensional crystal structure of LigU in the
presence and absence of bound ligands. These structures, and the catalytic properties of
mutant enzymes, have led to the identification of active site residues that are required for
the abstraction/donation of protons at C3 and C5 of the substrate and product. The
1
stereochemistry of the proton abstracted from C5 was determined by H NMR
spectroscopy of the reaction products when the reaction was conducted in D O. Based on
2
these findings a detailed chemical reaction mechanism has been proposed for the reaction
4
ACS Paragon Plus Environment

Page 5 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
catalyzed by LigU. In this mechanism, Cys-100 abstracts the proS hydrogen from C3 of
OMA, and Lys-66 donates a proton to the si-face of the double bond at C5 for a net
suprafacial hydrogen transfer.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
MATERIALS and METHODS
Materials. All commercial chemicals used in this investigation were obtained from
Millipore Sigma, unless otherwise stated. 2-Pyrone-4,6-dicarboxylate (PDC, 1) was
synthesized as previously described (17). The LigI hydrolase (UniProt entry: O87170), LigJ
hydratase (UniProt entry: G2IQQ5), and LigK aldolase (UniProt entry: G2IQQ8) were
isolated and purified as previously described (7-9).
Cloning, Expression, and Isolation of LigU. The open reading frame encoding
LigU from Novosphingobium sp. strain KA1 (Sphingomonas sp. strain KA1) (UniProt entry:
Q0KJL4) was chemically synthesized by GenScript and subsequently ligated into a pET-
3
0a(+) vector (EMD Millipore). The protein is 357 residues long plus a C-terminal linker
and His tag added as Leu-Glu-His . The gene was expressed in Escherichia coli BL21(DE3)
6
(
EMD Millipore) and the protein purified as previously described (8). The mutants S21A,
K66Q, K66M, C100S, S254A, K257M, and H281N were constructed using the Agilent
QuikChange protocol and the primers are listed in Table S1. The mutant enzymes were
purified as described previously for the wild-type enzyme (8). To improve the purity of the
enzymes used in the crystallization trials, size-exclusion chromatography was added to the
purification protocol of LigU (8). Protein was injected onto a HiLoad 16/600 Superdex 200
size exclusion chromatography column (GE Healthcare) equilibrated in 20 mM
HEPES/KOH, pH 7.9 and eluted using a flow rate of 0.8 mL/min. The fractions were
5
ACS Paragon Plus Environment

Biochemistry
Page 6 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
checked for purity by Mini-PROTEAN TGX stain-free SDS-PAGE (BIO-RAD). Selected
fractions were combined and concentrated to ~25 mg/mL. The combined fractions were
estimated to be ≥95% pure.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Measurement of Catalytic Activity. The initial velocity for the isomerization of
OMA (2) to KCH (3) by LigU was determined spectrophotometrically by monitoring the
change in absorbance at 265 nm and 30 C in a 96-well UV-STAR (Greiner Bio-One) plate in
a volume of 250 µL. OMA was made in situ by the addition of 1.0 µM LigI hydrolase to
variable concentrations of PDC (0 to 1.2 mM) in 200 mM HEPES/KOH (pH 8.0). Control
reactions were conducted in parallel using the same variable concentrations of PDC to
obtain the nonenzymatic conversion of OMA to KCH. The reaction catalyzed by LigI was
monitored at 312 nm until the reaction was complete. Aliquots of 10 µL of an
appropriately diluted LigU solution were subsequently added and the reaction was
monitored by changes in absorbance at 265 nm. The concentrations of the LigU mutants
used for measurement of the catalytic properties were as follows: K66Q (1.0 µM), K66M
(
2.0 µM), C100S (50 nM), H281N (25 nM), S21A (0.5 nM), K257M (1.0 µM), and S254A (250
pM). The change in the concentration of OMA was determined using a differential
-1
-1
extinction coefficient ( ) of 5900 M cm (8). The kinetic constants were determined
2
65
using an iterative non-linear least square regression fit to eq 1 where  is the initial
velocity, E is the total enzyme concentration, k is the turnover number, A is the substrate
t
cat
concentration, and K is the Michaelis constant. SigmaPlot 11.0 was used to calculate the
m
kinetic constants and standard errors.

/E = k (A)/(K + A)
(1)
t
cat
m
6
ACS Paragon Plus Environment

Page 7 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
Stereospecific Hydrogen Exchange at C5 of KCH. The stereospecific exchange of
1
protons at C5 of KCH was determined using H NMR spectroscopy with a Bruker Avance III
500 MHz spectrometer equipped with an H-C-N cryoprobe at 305 K. The reaction was
conducted in 200 mM KH PO /KOD (pD 6.9), containing 5.0 mM MgCl , 6.0 mM NADH, 5.0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
4
2
mM PDC, and a lyophilized enzyme cocktail of 1.0 µM each LigI, LigU, LigJ, LigK and malate
dehydrogenase in a volume of 600 µL of D O. Collectively, these enzymes convert the initial
2
1
substrate PDC (1) to malate. The reaction was monitored by H NMR spectroscopy using
the water suppression pulse program Watergate zggpw5 pulse sequence (18).
The deuterated L-malate, made from the combined activities of LigI, LigU, LigJ, LigK,
and malate dehydrogenase was compared to the deuterated L-malate formed from the
2
action of fumarase with fumarate in D O and also with [2,3- H]-fumarate in H O. The
2
2
control reaction was conducted in 200 mM KH PO /KOD (pD 6.6), 5.0 mM fumaric acid,
2
4
and 1.0 µM fumarase from porcine heart (Sigma-Aldrich) in a final volume of 600 µL in D O.
2
The zggpw5 pulse program with water suppression was calibrated at 4.675 ppm. A fully
protonated sample of 5.0 mM malate was prepared in 200 mM KH PO /KOD (pD 7.02) in a
2
4
final volume of 600 µL in ~100% D O.
2
Hydrogen-Deuterium Exchange Experiments. The proton exchange at C5 of KCH
(
3) with solvent D O in the presence of LigU was performed in 200 mM potassium
2
1
phosphate/KOD (pD 7.0) in a volume of 600 µL. The reaction was monitored by H NMR
spectroscopy with a Bruker Avance III 500 MHz spectrometer equipped with an H-C-N
cryoprobe at 305 K. KCH was prepared by mixing 20 mM PDC (1) with 200 mM
NaHCO /KCl, pH 9.6, in the presence of 1.0 µM LigI hydrolase and subsequent titration to
3
pH 9.5 with KOH. The reaction was incubated for 2 h at room temperature, and then LigI
7
ACS Paragon Plus Environment

Biochemistry
Page 8 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
was removed using PALL Nanosep 10 kDa and 3 kDa filters with an Eppendorf centrifuge.
The flow-through was collected and aliquoted (150 μL) into 1.7 mL sterile microcentrifuge
tubes. The samples were lyophilized to dryness and stored at -20C until needed.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The hydrogen-deuterium exchange reactions were initiated by adding 580 µL of D O
2
containing 200 mM KH PO /KOD (pD 6.6 to 7.0) to a lyophilized sample of KCH. A total of
2
4
1
.0 nM LigU and 66 nM bovine serum albumin (BSA) was added to the reaction mixture and
1
the reaction followed by H NMR spectroscopy. The hydrogen-deuterium exchange
reaction was monitored for ~6.4 h using the Watergate zggpw5 pulse program with water
suppression (18). The mutants K66M, C100S, and H281N were also tested in a similar
fashion. A 16 scan acquisition was collected before any enzyme was added to calibrate the
instrument for each of the reactions with the three mutant proteins. The total
concentration of LigU K66M, C100S, and H281N added to each reaction was 111 nM, 11 nM,
and 102 nM, respectively.
Crystallization of LigU Wild-Type and Mutants. Crystals of wild-type LigU were
prepared by sitting drop vapor diffusion (Intelli-Plate 24-4 by Hampton Research) with a
drop and reservoir volume of 4 µL and 500 µL, respectively. The reservoir contained 200
mM CaCl dihydrate, 100 mM HEPES/KOH, pH 6.5, and 35% pentaerythritol ethoxylate
2
(
Molecular Dimensions). The drop was 2.0 µL enzyme (22.3 mg/mL in 20 mM
HEPES/KOH, pH 7.5) and 2.0 µL reservoir solution. Precipitation was observed after ~24
o
h. Crystals were observed after incubation at 17 C for approximately 6 months. These
crystals were supplemented with 5.0 µL of reservoir solution and harvested by a nylon
loop (Hampton Research) prior to flash freezing in liquid nitrogen. Crystals of LigU C100S
were prepared by sitting drop vapor diffusion with a drop and reservoir volume of 4.0 µL
8
ACS Paragon Plus Environment

Page 9 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
and 500 µL, respectively. The reservoir contained 200 mM MgCl hexahydrate, 100 mM
2
HEPES/KOH, pH 7.5, and 25 % (w/v) PEG 3350 (Hampton Research). The drop was 2.0 µL
enzyme (10 mg/mL in 20 mM HEPES/KOH, pH 7.9) and 2.0 µL reservoir solution. Crystals
with the morphology of glass plates were observed after a ~2 week incubation at 17 C.
Flash freezing of crystals was conducted in a similar fashion to wild-type crystals. Sparse
matrix screening of LigU K66M in the presence of OMA was performed at 17 ºC using a
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Mosquito robot (TTP Labtech). Initial crystallization conditions were identified within the
Wizard Classic 1/2 screen (Rigaku) containing 1000 mM sodium citrate tribasic, 100 mM
sodium cacodylate/HCl, pH 6.5 with a total volume of 100 µL. The drop was 200 nL
enzyme (20.4 mg/mL in 20 mM HEPES/KOH, pH 7.9), 100 nL ligand (40 mM OMA, pH 8.2),
and 500 nL reservoir solution. OMA was made by incubating 40 mM PDC with ~4 µM of
LigI hydrolase in 200 mM HEPES/KOH, pH 8.2 using KOH to adjust the pH of the reaction
solution to pH 8.2. LigI was removed by centrifugation using PALL Nanosep 10 KDa and 3
KDa filters and the flow-through was collected. These crystals were incubated at 17C for
approximately 1 week, harvested directly from the drop using a nylon loop, and frozen
directly in liquid nitrogen. Initial LigU C100S (2.5 Å) crystals were prepared by the hanging
drop vapor diffusion method. The reservoir solution contained 0.20 M MgCl , 0.10 M
2
HEPES/KOH, pH 7.5, and 25% (w/v) PEG 3350. LigU C100S was concentrated to 10
mg/mL in 10 mM HEPES/KOH, pH 7.5, and then 2.0 μL of the protein solution was mixed
with 2.0 μL of the well solution and 1.0 μL of a 10 mM solution of OMA. The crystals were
stored at 17 °C and incubated for ~3 months before being submerged in reservoir solution,
harvested, and plunged into liquid nitrogen.
9
ACS Paragon Plus Environment

Biochemistry
Page 10 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Data Collection and Processing. Diffraction data was collected at Stanford
Synchrotron Radiation Lightsource (SSRL) BL 14-1 on an Eiger 16M detector at 100 K.
Data was integrated in either HKL2000 or iMosflm (19, 20). Scaling and merging was
conducted in either HKL2000 or Aimless (19, 21). Merged observed intensities were then
imported into Phenix reflection file editor where an R-free set was assigned using the
default settings (2000 reflections) (22).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Structure Determination. Using a lower resolution dataset, initial phases for LigU
(
(
C100S) were determined by molecular replacement using chain A of PDB entry: 3G7K
methylitaconate--isomerase, Eubacterium barkeri with 41% sequence identity) as a
search model in the Phaser-MR module of Phenix (23, 24). Several rounds of model
building and refinement were conducted using AutoBuild. Additional rounds of manual
rebuilding were then conducted in Coot and refinement of the coordinates performed in
phenix.refine (25, 26). A monomer from this model was used for molecular replacement
against the wild type and LigU(K66M) + OMA datasets. For the high resolution LigU
(
C100S) dataset, the original R-free flags were transferred and extended to the higher
resolution for further refinement. In the LigU (K66M) + OMA dataset, translational NCS
was identified by Phenix Xtriage (22). However, the translational NCS did not appear to
have an impact on molecular replacement or refinement. OMA was built in JLigand and
optimized in REEL (27, 28). Non-crystallographic symmetry restraints were utilized in the
early stages of refinement but were released toward the ending stages. Final refinement of
coordinates was conducted in phenix.refine with the following settings enabled: xyz
(
reciprocal and real), individual b-factors, TLS, occupancies, X-ray/stereochemistry
weights, and X-ray/ADP weights. Real space xyz refinement was disabled during the last
10
ACS Paragon Plus Environment

Page 11 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
cycle of refinement for LigU due to significant disorder in chain B of the structure. Model
validation was conducted in Phenix and Molprobity prior to deposition (29). Data
collection and refinement statistics can be found in Table 1. The RMSD calculations were
determined using Chimera to align the residue pairs with refitting iterations until full
population convergence with a 5.0 Å cutoff for the maximum C-C pairwise distance (30).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
11
ACS Paragon Plus Environment

Biochemistry
Page 12 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1. Data Collection and Refinement Statistics for Wild-Type LigU and Mutants.
Structure
PDB entry
LigU
6P3J
LigU (C100S)
LigU (K66M) + OMA
6P3K
6P3H
Data Collection Parameters
Beamline
Wavelength (Å)
Resolution (Å)
Space Group
SSRL BL14-1
1.1271
50.00 - 2.02
SSRL BL14-1
1.1271
50.00 - 1.88
C2
SSRL BL14-1
1.1271
39.53 - 1.62
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
P2 2 2
1 1 1
1 1
P2 2 2
a, b, c (Å)
α, β, γ ( )
Completeness (%)
62.78, 67.36, 161.07
90.00, 90.00, 90.00
97.8 (86.5)
45,097
97.42, 142.33, 60.35
90.00, 111.73, 90.00
87.3 (47.1)
53,939
84.88, 134.00, 168.37
90.00, 90.00, 90.00
97.5 (94.2)
236,378
o
a
Unique reflections
a
I/σ
30.2 (1.1)
59.3
5.2 (2.8)
5.7 (59.7)
2.6 (38.8)
22.1 (2.1)
91.5
2.9 (1.8)
4.6 (24.2)
3.0 (20.0)
8.8 (2.6)
44.6
5.1 (5.2)
9.1 (59.7)
CC1/2 last shell (%)
a
Redundancy
a
R
R
merge
a
pim
4.4 (27.6)
Refinement
resolution (Å)
a
42.00 - 2.02 (2.07 - 2.02) 42.03 - 1.88 (1.92 - 1.88) 39.47 - 1.62 (1.66 - 1.62)
a
no. of reflections (free)
44,713 (2,003)
18.6 (30.1)
53,866 (2,597)
17.3 (23.2)
236,265 (2,000)
15.7 (28.7)
a
R
R
work
a
free
21.2 (31.1)
21.5 (29.8)
18.7 (28.3)
Molecules in asymmetric unit
TLS groups
2
5
2
4
4
8
No. of atoms
protein
Ions
5,144
12
5,140
6
10,505
6
Ligand
Water
NA
224
NA
388
56
1,516
B-factors
All
Protein
Ions
Ligand
61.3
61.8
59.9
NA
43.0
43.1
37.7
NA
29.4
28.2
35.8
23.8
38.4
Water
50.0
41.9
Validation
Ramachandran preferred (%)
Ramachandran allowed (%)
Ramachandran outliers (%)
RMSD for bond lengths (Å)
RMSD for bond angles ( )
Rotamer outliers (%)
Clash score
96.8
3.2
0.0
0.003
0.596
0.9
96.7
3.3
0.0
0.014
1.249
0.6
98.0
1.9
0.1
0.017
1.394
0.8
o
2.9
5.6
4.0
a
Number in parentheses indicate values for the highest resolution shell.
12
ACS Paragon Plus Environment

Page 13 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
RESULTS
Stereochemistry of Proton Transfer at C5 by LigU. During the allylic
isomerization catalyzed by LigU, a proton is abstracted from C3 of OMA (2) and another
proton (or perhaps the same proton) is added to C5 to form KCH (3) (Scheme 1). To
determine whether the proton transfer at C5 of KCH occurs with either the proR or proS
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
hydrogen, the isomerization reaction was conducted in D O in the presence of LigU, LigJ,
2
and LigK to drive the reaction forward toward the formation of oxaloacetate (6).
Simultaneously, the oxaloacetate was reduced to L-malate (7) using NADH with malate
dehydrogenase (MDH) (Scheme 2). C3 of the enzymatically formed L-malate corresponds
with C5 of KCH, and it will be bonded to one hydrogen and one deuterium (from solvent).
Scheme 2. Stereochemistry of the LigU catalyzed isomerization reaction.
Proton NMR spectroscopy was used to determine the specific stereochemical
arrangement of the hydrogen and deuterium at C3 of L-malate (7) formed from PDC in the
presence of LigI, LigU, LigJ, LigK, and MDH when the reactions were conducted in D O. In
2
1
the H NMR spectrum of unlabeled L-malate, the proR and proS hydrogens at C3 are
separated by 0.31 ppm (Figure 1a). The chemical shift assignments for these two
2
hydrogens were determined by catalyzing the formation of L-[3- H]-malate by the addition
of D O to fumarate with fumarase (Figure 1b). It is known that fumarase catalyzes the
2
13
ACS Paragon Plus Environment

Biochemistry
Page 14 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
abstraction of the proR hydrogen of L-malate at C3 and thus the pair of partially resolved
triplets at 2.309 ppm in Figure 1b is due to the proS hydrogen (31, 32). The L-malate
formed from the LigU catalyzed reaction in D O exhibits an unresolved triplet at 2.605 ppm
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
and thus LigU catalyzes the exchange of the proS hydrogen from C5 of KCH (Figure 1c).
The stereochemical transformations are summarized in Scheme 2.
1
Figure 1: H NMR spectra of L-malate for the hydrogen(s) attached to C3. (a) L-malate
purchased commercially with H from C3 a doublet of doublets centered at 2.629 ppm with
R
2
3
J = 15.3 Hz and J = 3.1 Hz, and H from C3 a doublet of doublets centered at 2.323 ppm
S
2
3
with J = 15.3 Hz and J = 10.1 Hz; (b) L-malate formed from the enzymatic hydration of
fumarate with fumarase in D O with H from C3 a doublet of triplets centered at 2.309 ppm
with J = 10.2 Hz; (c) L-malate formed enzymatically from PDC (1) in the presence of LigI,
2
S
3
LigU, LigJ, LigK and L-malate dehydrogenase in D O with H from C3 an unresolved triplet
2
R
centered at 2.605 ppm.
Mutation of Active Site Residues. The kinetic constants for the isomerization of
OMA to KCH were determined for wild-type LigU and seven single-site mutant enzymes
(
Table 2). The residues Lys-66 and Cys-100 were mutated based on previous studies
suggesting they were catalytic residues in homologous enzymes. The other four residues
14
ACS Paragon Plus Environment

Page 15 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
(
Ser-21, Ser-254, Lys-257, and His-281) were chosen based on homology models to be in
the active site pocket. The wild-type enzyme is quite active with values of k and k /K
cat
cat
m
-1
6
-1 -1
-1 -1
of 1300 s and 7.7 x 10 M s , respectively. No activity (k /K < 14 M s ) was
cat
m
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
observed with the K66Q, K66M, or K257M mutants and significant losses in activity were
observed for the C100S and H281N mutants.
Table 2. Kinetic Constants for Isomerization of OMA to KCH by LigUa
k
(s )
K
(M)
m
k /K
cat m
(M s )
cat
-
1
-1 -1
enzyme
wild-type 1300 ± 120 170 ± 30
(8 ± 2) x 106
(5 ± 1) x 105
< 4
S21A
K66Q
60 ± 5
--
120 ± 30
--
--
K66M
C100S
S254A
K257M
H281N
--
< 4
1.6 ± 0.2
150 ± 10
--
140 ± 50 (12 ± 3) x 103
100 ± 20 (14 ± 2) x 105
--
< (14 ± 3)
380 ± 110 (11 ± 2) x 103
4.3 ± 0.6
aAt 30 C, pH 8.0, and OMA concentrations varied from 0 to 0.8 mM. The kinetic
parameters were calculated using an iterative non-linear least square regression fitting
analysis and standard errors were obtained by the matrix inversion method using
Sigmaplot 11.0.
Hydrogen-Deuterium Exchange at C5 of KCH. In the LigU-catalyzed
isomerization of OMA, a proton from C3 must be abstracted and a proton must
1
subsequently be added to C5 to form KCH. H NMR spectroscopy was used to monitor the
exchange of hydrogens at C5 of KCH in D O during the LigU catalyzed reaction in an
2
15
ACS Paragon Plus Environment

Biochemistry
Page 16 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
attempt to determine which residues are responsible for these protons transfers (Scheme
1
3
). In the H NMR spectrum of KCH (3a), the two hydrogens at C5 of KCH resonate as a
doublet at 3.209 ppm. These two hydrogens are spin coupled to the hydrogen at C3 (6.862
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
ppm) with a J = 1.4 Hz (Figure 2a). After the addition of wild-type LigU, the intensity of
the doublet at 3.209 ppm decreases and is gradually replaced by an unresolved triplet at
3.191 ppm with an apparent coupling constant of ~1.9 Hz. The new triplet originates from
the proR hydrogen at C5 that is spin coupled to the geminal proS deuterium of KCH (3b and
3d). Also observed is the formation of a new singlet at 3.207 ppm that partially overlaps
with the original doublet centered at 3.209 ppm. This new resonance must be due to the
exchange of the hydrogen at C3 (3c) of KCH with solvent deuterium and the accompanying
1
1
absence of the H- H spin coupling. The loss of the original hydrogen at C3 is only possible
via the transient enzymatic formation of OMA and the subsequent nonenzymatic exchange
of the labile hydrogens adjacent to the carbonyl group at C2 .
The hydrogen-deuterium exchange reactions were similarly conducted with the
K66M, C100S, and H281N mutants of LigU. When the K66M mutant of LigU is added to
1
KCH there was no observable change in the intensity of the H NMR spectrum for the C5
protons of KCH (Figure 2b). However, significant exchange of a proton attached to C5 of
KCH was observed when the C100S mutant was utilized (Figure 2c). The doublet for the
protons attached to C5 at 3.211 ppm diminishes in intensity while the unresolved triplet at
3.194 ppm increases in intensity until the doublet completely vanishes. There was no
indication for the appearance of a new singlet slightly upfield from the doublet at 3.207
ppm as was observed with wild-type LigU. The solvent exchange reaction for the H281N
mutant of LigU was similar to that of the wild type LigU except that higher enzyme
16
ACS Paragon Plus Environment

Page 17 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
concentrations were required to observe the changes in the NMR spectrum of KCH (Figure
2d).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 3. Hydrogen-deuterium exchange at C5 of KCH catalyzed by LigU isomerase.
1
Figure 2: H NMR spectra of 1.0 mM KCH showing the exchange of hydrogen at C5 with
solvent deuterium after the addition of wild-type LigU or one of the mutant enzymes. (a)
wild-type LigU (1.0 nM) before the addition of enzyme (black), after 10 min (red), and after
160 min (blue); (b) K66M (111 nM) before the addition of enzyme (black) and after 60 min
(
red); (c) C100S (11 nM) before the addition of enzyme (black), after 25 min (red), and
after 350 min (blue); (d) H281N (113 nM) before the addition of enzyme (black), after 20
min (red), and after 320 min (blue). Additional details are provided in the text.
17
ACS Paragon Plus Environment

Biochemistry
Page 18 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Three-Dimensional Structure of LigU. The structure of wild-type LigU was
determined to a resolution of 2.0 Å. LigU crystallized in the space group P2 2 2 and
1
1 1
contains two monomers per asymmetric unit with a RMS difference of 0.7 Å over the fully
populated 344 residue pairs (Figure 3). Each of the two subunits is composed of an N-
terminal (residues 1-171) and a C-terminal (172-357) domain. The two monomers
interact with one another near the N-terminal domain via residues that originate from the
last -strand (Ala-347 through Phe-356) from the C-terminal domain. The protein fold of
each monomer consists of 16 β-strands, and 10 -helices (Figure 4). Each of the two
domains has a central α-helix that is wrapped by antiparallel β-strands that forms a β-
barrel.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. (a) Ribbon model of the homodimer of wild-type LigU (PDB entry: 6P3J) at 2.0 Å
resolution. One of the two monomers is colored silver, and the second is colored purple for
the N-terminal -helices, sea green for the N-terminal -strands, orange for the C-terminal

-helices, and pink for the C-terminal -strands. β-Strand 16 is colored black from chain A
and β-strand 16 from chain B is colored yellow. (b) Ribbon model rotated 90 to show the
interactions from the last β-strands of each monomer. The figures were made using
Chimera (33).
18
ACS Paragon Plus Environment

Page 19 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Topology model for the two domains of LigU chain A where the -helices and -
strands of the N-domain are depicted as purple cylinders and sea green arrows,
respectively. The -helices and -strands of the C-domain are depicted as orange cylinders
and pink arrows, respectively. -Strand 16 is shown as a black arrow. The figure was
made using Affinity Designer 1.6.1.
The N- and C-terminal domains that compose each monomeric unit are structurally
similar to one another with an average RMS difference of 2.1 Å over the fully populated 118
residue pairs, but the amino acid sequence identity between the two domains is only 10%
(Figures 5 and S2). This value was calculated using chain A where the N-terminal domain
(
residues 8-171) was compared with residues from the C-terminal domain (172-345). In
addition to the central -helices, 2 and 4, there are also external -helices in each
domain (Figure 4). These external -helices and nearby loops create a solvent channel
that leads to the active site. The active site is located at the junction of these two domains
where the internal -helices contact one another at their N-terminal ends. The putative
catalytic residues, Lys-66 and Cys-100, are located on loops within the N-terminal domain,
whereas other highly conserved residues (Lys-257 and His-281) that interact with the
substrate (see below) are from the C-terminal domain.
19
ACS Paragon Plus Environment

Biochemistry
Page 20 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. (a) Structural alignment of the N- and C-terminal domains of LigU. The N-
terminal domain (residues 8-171) is shown in pink and the C-terminal domain (residues
1
72-345) is presented in light blue from chain A. (b) Ribbon model rotated 90. The
figures were made using Chimera (33) and Affinity Designer 1.6.1.
Structure of C100S Mutant of LigU. The structure of the C100S mutant of LigU
was determined to a resolution of 1.9 Å by molecular replacement (PDB entry: 6P3K). This
mutant crystallized into the space group C2 with unit cell dimensions a = 97.42 Å, b =
142.33 Å, c = 60.35 Å, and contained two monomers per asymmetric unit (Table 1). The
RMS differences between wild-type LigU and the C100S is 0.5 Å for chain A over the fully
populated 351 residue pairs. There are no significant changes between the wild-type and
20
ACS Paragon Plus Environment

Page 21 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
mutant enzymes except for the orientation of the hydroxyl group in C100S, relative to the
thiol group in the wild-type enzyme. The thiol of Cys-100 is hydrogen bonded to Lys-66 at
a distance of 3.2 Å in the wild-type enzyme, whereas the distance between Ser-100 and
Lys-66 is 5.0 Å in the mutant enzyme (Figure S3).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Structure of the Ligand-Bound Complex with K66M Mutant of LigU. The K66M
mutant of LigU was crystallized in the presence of 10 mM OMA (2). The protein
crystallized in the space group P2 2 2 with unit cell dimensions a = 84.88 Å, b = 134.00 Å, c
1
1
=
168.37 Å with four monomers per asymmetric unit (Table 1). The simulated annealing
omit map contoured to 3  has defined electron density for the orientation of all functional
groups of the substrate in each monomer (Figure 6a). Even though the electron density
shows high occupancy of the ligand, a resolution of 1.6 Å does not provide discrete peaks
for individual atoms. Thus, we were unable to distinguish the double bond location and
have modelled the substrate, OMA, in the active site. The modelling is supported by both
functional data showing K66M has no observable activity, and the equilibrium constant of
0
.3 favoring OMA over the product KCH in solution (8). In the ligand-bound complex there
is a large network of hydrogen bond interactions with the backbone and side chains of
nearby residues to the carboxylate and carbonyl groups of the substrate (Figure 6b). The
C1 carboxylate group is within hydrogen bonding distance of a water molecule and the
amide backbone of Gly-288, Leu-287, Val-286, and Gly-101. Additional hydrogen bonds to
the C2 carbonyl oxygen are formed via interactions with the backbone nitrogens of Asn-
1
02, Val-286, and Gly-101. The C6 carboxylate forms hydrogen bonds with two water
molecules, Gln-98, and Lys-257. Lastly, the C7 carboxylate forms hydrogen bonds with a
21
ACS Paragon Plus Environment

Biochemistry
Page 22 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
water molecule, His-281, Asn-102, and Ser-21. Overall, the substrate forms a substantial
number of hydrogen bonds within the active site.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Model representations of LigU K66M co-crystallized with substrate (PDB entry:
6
P3H, chain A) where the ligand is depicted in yellow. (a) Simulated annealing-omit mFo-
DF map (black mesh) contoured at 3.0  shows unbiased electron density for substrate
c
OMA (2) in the active site. (b) Model showing hydrogen bonding and ionic interactions
between the substrate and protein.
Conformational Changes in LigU Complexed with Substrate. A conformational
change between apo- and ligand-bound complexes is observed. Chains A and B from wild-
type LigU were aligned with chains A and C of the K66M mutant. The RMS difference
across 695  carbon atoms of the dimer was 3.8 Å. The majority of this difference comes
22
ACS Paragon Plus Environment

Page 23 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
from the C-terminal domain, specifically -helices 5’ (Arg-221 to Asp-225) and 3 (Ala-
228 to Val-241). Arg-221 from 5’ changes orientation to be more fully extended in the
active site channel to occlude the opening with movements of 6.3-7.0 Å in each subunit
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
Figure 7a). Arg-279 on the loop after -strand-12 also moves inward helping to block the
channel by moving distances of 5.2 - 7.9 Å in each subunit. The apo-form of LigU has a large
solvent channel for the binding of substrate. In the bound structure the solvent channel
disappears as the -helices and loops from the flexible C-terminus move inward to give a
closed conformation (Figure 7b and 7c). An overlay of the monomeric units is shown in
Figure S4. The conformational differences between wild-type LigU and the K66M mutant
are most likely due to the presence of the substrate bound in the active site of the K66M
mutant rather than any differences caused by the mutation of Lys66 to methionine.
23
ACS Paragon Plus Environment

Biochemistry
Page 24 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. (a) Superposition of wild-type LigU and the K66M mutant dimers. The wild-type
enzyme is colored tan, the K66M mutant is colored sky blue and the ligand colored medium
blue. Arg-221 from 5’ is colored pink in wild-type and yellow in the K66M mutant with
the distance moved labeled for each monomer. (b) Surface rendition of wild-type enzyme
shown in the superimposed structures of panel a. (c) Surface rendition of K66M shown in
the superimposed image of panel a.
24
ACS Paragon Plus Environment

Page 25 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
DISCUSSION
Chemical Reaction Mechanism. The mechanism of action for the 1,3-allylic
isomerization of OMA (2) to KCH (3) catalyzed by LigU was addressed using NMR
spectroscopy, site-directed mutagenesis, and X-ray crystallography. Six residues contained
within the active site of LigU were mutated to residues that cannot function as general
acid/base catalysts for the transfer of protons between C3 and C5 of OMA and KCH. Of the
mutants made for this investigation, the most significant reductions in catalytic activity
were obtained for Lys-66, Cys-100, Lys-257, and His-281 (Table 2). In the substrate-
bound complex of the K66M mutant, the ε-amino group of Lys-257 makes a hydrogen bond
to the C6-carboxylate and is positioned too far from either C3 or C5 to be involved in a
proton transfer (Figure 6b). Mutants of the remaining three residues (K66, C100, and
H281) were subsequently interrogated for their ability to catalyze hydrogen/deuterium
exchange reactions between KCH and solvent.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The wild-type enzyme catalyzes the exchange of one of the two prochiral hydrogens at
C5 of KCH with solvent and also facilitates the exchange of the proton at C3 via transient
formation of OMA (Figure 2a). The exchange pattern for the H281N mutant was identical
to that observed for the wild-type enzyme. The K66M mutant, however, was unable to
catalyze the exchange of the hydrogen at C5 with deuterium from solvent (Figure 2b).
The C100S mutant was functionally able to catalyze hydrogen/deuterium exchange at C5
but was unable to facilitate the exchange at C3 (Figure 2c). These results support the
conclusion that Cys-100 functions to catalyze the removal of a proton from C3 and that Lys-
66 functions to donate a proton to C5 during the conversion of OMA to KCH.
25
ACS Paragon Plus Environment

Biochemistry
Page 26 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The stereochemical outcomes for the proton transfers at C3 and C5 of OMA (2) were
determined by the positioning of the active site residues in the three-dimensional structure
of LigU complexed with substrate, and isotopic labeling of the product KCH (3). When the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
LigU-catalyzed reaction is conducted in D O, deuterium is found exclusively in the proS
2
position at C5 (Figure 1). This result is fully consistent with the presumptive positioning
of Lys-66 in the active site of the K66M mutant of LigU where the terminal methyl group of
the methionine is 3.8 Å from the si-face of the double bond at C5 (Figure 8). The
stereochemical outcome for the proton transfer at C3 could not be determined directly, but
the only residue in the active site that is positioned to abstract a proton from C3 of OMA is
Cys-100. The thiol group of Cys-100 is 3.4 Å from the proS hydrogen at C3. Since Lys-66
and Cys-100 are on the same face of the bound substrate, the 1,3-allylic isomerization
catalyzed by LigU is suprafacial (Scheme 4). In the proposed reaction mechanism, the
thiolate of Cys-100 abstracts the proS hydrogen from C3 of OMA in a reaction that is
stabilized by charge delocalization to the C2 carbonyl. KCH is formed by the protonation of
C5 via proton transfer from the side chain of Lys-66.
Figure 8. Model representation of the K66M mutant of LigU co-crystallized with OMA (PDB
entry: 6P3H, chain A) where OMA is depicted in yellow.
26
ACS Paragon Plus Environment

Page 27 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 4. Proposed reaction mechanism for LigU.
Comparisons between LigU and PrpF. PrpF and aconitate isomerase catalyze the
allylic rearrangement of cis-aconitate to trans-aconitate (12, 34, 35). PrpF from Shewanella
oneidensis MR-1 has also been reported to catalyze the isomerization of (S)-4-methyl-cis-
aconitate to 2-methyl-cis-aconitate (13). Klinman and Rose demonstrated that aconitate
isomerase from Pseudomonas putida catalyzes the 1,3-allylic isomerization of cis-aconitate
via the abstraction of the proS hydrogen from C4 and suprafacial donation of a proton to C2
(35). These results are fully in accord with the results obtained with LigU (Scheme 4). The
three-dimensional structure of PrpF has been determined with trans-aconitate bound in
the active site (PDB entry: 2PW0) (12). A structural comparison of the bound ligands in the
27
ACS Paragon Plus Environment

Biochemistry
Page 28 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
active sites of these two enzymes is illustrated in Figures 9 and S5. The C1 carboxylate
group of trans-aconitate is positioned similarly as the C7 carboxylate group of OMA and
interacts with the side chain of His-317. The C5 carboxylate of trans-aconitate overlays
quite closely with the C6 carboxylate of OMA. The side chains of the proposed catalytic
residues (Lys-73 and Cys-107) are aligned quite closely with those for LigU and are
positioned on the same face of the ligand. The C1 and C2 functional groups of OMA are
situated deeper within the active site pocket where backbone interactions occur.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 9. Superimposed models of PrpF (tan) complexed with trans-aconitate (blue) (PDB
entry: 2PW0, chain a) and LigU (green) complexed with ligand (OMA or KCH) (pink) (PDB
entry: 6P3H, chain a).
Conclusions. We have determined the three-dimensional structure of LigU in the
presence and absence of bound substrate. Solvent deuterium exchange experiments and
kinetic analysis of site-directed mutants have demonstrated that LigU catalyzes the
isomerization of OMA (2) to KCH (3) via the abstraction of the proS hydrogen from C3 by
Cys-100 and donation of a proton to C5 to the same face of the enolate intermediate by Lys-
66 (Scheme 4).
28
ACS Paragon Plus Environment

Page 29 of 34
Biochemistry
1
2
3
4
5
6
7
8
9
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the ACS Publications website at
DOI: XXXX, This information includes (i) DNA primers used for the construction of LigU
mutants; (ii) sequence similarity network for cog2828; (iii) sequence alignment of the N-
and C- terminal domains of LigU; (iv) ribbon model of the monomer unit of wild-type and
mutant C100S superimposed with one another; (v) surface models of the monomer unit of
wild-type LigU and the K66M mutant enzyme superimposed with one another; and (vi)
ribbon model of the monomers of the K66M LigU mutant and PrpF with ligand bound
superimposed.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Accession Codes
The X-ray coordinates and structure factors of LigU have been deposited in the Protein
Data Bank as entries 6P3J (wild-type), 6P3K (apo C100S mutant), and 6P3H (K66M mutant
with ligand bound).
AUTHOR INFORMATION
Corresponding Authors
Frank M. Raushel
e-mail: raushel@tamu.edu
David P. Barondeau e-mail: barondeau@tamu.edu
Accession Codes
LigU UniProt entry: Q0KJL4
ORCID
Tessily N. Hogancamp: 0000-0002-7610-1217
Frank M. Raushel: 0000-0002-5918-3089
Seth A. Cory: 0000-0003-0886-1502
David P. Barondeau: 0000-0002-6422-9053
29
ACS Paragon Plus Environment