Metallomics

View Article Online

PAPER View Journal

Characterization of a highly ecient antibiotic-

Published on 18 July 2017. Downloaded by ST LAWRENCE UNIVERSITY on 07/08/2017 14:20:42.

degrading metallo-b-lactamase obtained from an

Cite this: DOI: 10.1039/c7mt00195a
uncultured member of a permafrost community
Marcelo Monteiro Pedroso, *a Christopher Selleck,a Charmaine Enculescu,a
Jerey R. Harmer,b Natasa Mitic,c Whitney R. Craig,d Waleed Helweh,e
Philip Hugenholtz,af Gene W. Tyson,af David L. Tierney,d James A. Larrabeee and
Gerhard Schenk *a

Received 4th July 2017,
Accepted 18th July 2017
DOI: 10.1039/c7mt00195a
rsc.li/metallomics
Antibiotic resistance is a major global health problem, one that threatens to derail the benefits garnered from
arguably the greatest success of modern medicine, the discovery of antibiotics. Among the most potent
agents contributing to antibiotic resistance are metallo-b-lactamases (MBLs). The discovery of MBL-like
enzymes in microorganisms that are not in contact with the human population is of particular concern as
these proteins already have the in-built capacity to inactivate antibiotics, even though they may not need MBL
activity for their survival. Here, we demonstrate that a microbiome from a remote and frozen environment in
Alaska harbours at least one highly ecient MBL, LRA-8. LRA-8 is homologous to the B3 subgroup of MBLs
and has a substrate profile and catalytic properties similar to well-known members of this enzyme family,
which are expressed by major human pathogens. LRA-8 is predominantly a penicillinase, but is also active
towards carbapenems, but not cephalosporins. Spectroscopic studies indicate that LRA-8 has an active site
structure similar to that of other MBLs (in particular B3 subgroup representative AIM-1), and a combination of
steady-state and pre-steady-state kinetic data demonstrate that the enzyme is likely to employ a metal ion-
bridging hydroxide to initiate catalysis. The rate-limiting step is the decay of a chromophoric, tetrahedral
intermediate, as is observed in various other MBLs. Thus, studying the properties of such pristine MBL-like
proteins may provide insight into the structural plasticity of this family of enzymes that may facilitate functional
promiscuity, while important insight into the evolution of MBLs may also be gained.

Significance to metallomics
Metal-dependent enzymes (e.g. metallo-b-lactamases (MBLs)) constitute up to 50% of all enzymes and combine the selectivity bestowed upon them by the
complexity of the protein scaold with catalytic eciency that originates from their metallic cofactors. MBLs are major contributors to antibiotic resistance, but
their occurrence is not limited to pathogenic organisms. A metagenomics analysis revealed that microbial communities in frozen environments have several
MBL-like proteins. At least one of these is indeed highly eective in inactivating b-lactam antibiotics, posing an emerging threat to global health, potentially
associated with climate change.

a
School of Chemistry and Molecular Biosciences, The University of Queensland,
St. Lucia, Queensland, 4072, Australia. E-mail: m.pedroso@uq.edu.au,

b
schenk@uq.edu.au
Centre for Advanced Imaging, The University of Queensland, St. Lucia,
Introduction
Queensland, 4072, Australia
c
The successful treatment of common bacterial infections
Department of Chemistry, Maynooth University, Maynooth, Co. Kildare, Ireland
d
Department of Chemistry and Biochemistry, Miami University, Oxford,
has been challenged in an alarming manner. This has been
Ohio 45056, USA achieved by the spread of antibiotic resistance in an increasing
e
Department of Chemistry and Biochemistry, Middlebury College, Middlebury, number of pathogens of clinical relevance, sustained by periods
VT, 05753, USA of haphazard use of antibiotics in health care, combined with
f
Australian Centre for Ecogenomics, The University of Queensland, St. Lucia,
global motility.1 Furthermore, antibiotics are employed in a
Queensland, 4072, Australia
Electronic supplementary information (ESI) available: Fig. S1: Map of recombi-
prophylactic manner in the agricultural industry, often in feed
nant expression system and sequence of LRA-8; Fig. S2: SDS-PAGE gel of purified and water supply.2 Worryingly, a handful of antibiotics used to
LRA-8. See DOI: 10.1039/c7mt00195a treat human pathogens are also used in agriculture, engendering

This journal is The Royal Society of Chemistry 2017 Metallomics


Paper
greater selective pressure in a more widely disseminated
fashion.2,3
Antibiotic resistance determinants are frequently located on
transposable genetic elements. Through horizontal gene transfer
urban and environmental sources of antibiotic resistance are
easily connected.46 In particular, uncultured bacteria in soil
environments have been identified as an important reservoir of
diverse antibiotic resistance determinants,7 including some that
Published on 18 July 2017. Downloaded by ST LAWRENCE UNIVERSITY on 07/08/2017 14:20:42.
have not been subjected to anthropogenic influence.8 The latter
has recently also been demonstrated by an analysis of the gut and
skin microbiome of the Amazonian tribe of the Yanomami, a
people that has not been in previous contact with the outside
world.9 Antibiotic resistance in the absence of human inter-
vention offers invaluable insights regarding the origin and
evolution of antibiotic resistance, which may assist in future
efforts to counteract and prevent this global threat to modern
healthcare.7,9
b-Lactamases are a large family of enzymes, capable of
inactivating most commonly used b-lactam antibiotics such
as penicillins, cephalosporins and carbapenems (Fig. 1). This
family is divided into four classes, A, B, C and D.10,11 Classes A,
C and D are serine-b-lactamases (SBLs) which employ an active
site serine residue as nucleophile to initiate the hydrolysis of
the b-lactam bond.12,13 SBLs have narrow substrate profiles in
addition to a number of clinically available inhibitors.12,14,15
Class B constitutes the group of metallo-b-lactamases (MBLs)
which require one or two zinc(II) ions for the hydrolysis of the
b-lactam bond.1619 The MBLs can be further divided into as
many as four subgroups, i.e. B1B4, depending on sequence
similarity and the number of zinc(II) ions required for
catalysis.2024 In contrast to their SBL counterparts, MBLs have
a high degree of mutational frequency at their active site,
conferring substrate promiscuity.22 Moreover, their rapid
dissemination and a lack of clinical inhibitors represent a
threat to the human population.5,14,25
At an increasing rate antibiotic-degrading activities are
identified in metagenomic analyses of organisms found in
environments that have not been challenged by human activities.
In particular the MBLs play a significant role in this context since
they are catalytically highly ecient while no clinically useful
inhibitor is currently available. Insight into the structure and
function of MBL-like proteins identified in such environments,
including the role(s) metal ions play in promoting the catalytic
proficiency of these enzymes, may thus also inform novel strategies
to design potent inhibitors that may be of clinical relevance to
Fig. 1 Chemical structures of representative antibiotic types, each containing
the defining four-membered b-lactam ring. The R groups are indicative of
modifications for respective antibiotics.
Metallomics
View Article Online
Metallomics
address antibiotic resistance. Examples of such MBL-like proteins
are MIM-1 and MIM-2, produced by the marine microorganisms
Novosphingobium pentaromativorans and Simiduia agarivorans,
respectively.26 Although MIM-1 and MIM-2 are both ecient MBLs
they also have potent lactonase activity, an activity that is associated
with quorum sensing.27 This observed functional promiscuity
provides one explanation why MBL activity may be found in
pristine environments. Similarly, a metagenomics study with
uncultured bacteria collected from soil from a remote region in
Alaska identified a group of 14 candidate genes (LRA; b-Lactam
Resistance in Alaskan soil genes) predicted to encode b-lactamases,
six SBLs and eight MBLs.8 When expressed in Escherichia coli each
of these 14 candidates conferred resistance to the host organism
to a number of b-lactam antibiotics, in some cases to clinically
relevant levels. The most ecient candidate is LRA-8, which
displays minimum inhibitory concentrations (MICs) greater
than 512 mg mL1 for the four antibiotics tested from the
penicillin family. LRA-8, which is homologous to MBLs from
the B3 subgroup (Fig. 2), also confers resistance towards several
cephalosporins (with MICs ranging from 32 to 128 mg mL1; no
carbapenems were tested).8 With the current trend in global
warming, this raises the possibility that genes encoding LRAs
may enter the human population, potentially exacerbating the
spread of antibiotic resistance. Here, we expressed and purified
LRA-8 and demonstrated that it is a potent MBL with catalytic
parameters matching those of its counterparts from pathogenic
organisms.
Results and discussion
LRA-8 is a potent MBL
In order to establish if LRA-8 can act as an MBL, kinetic assays
were carried out with several b-lactam substrates that represent
three major groups of antibiotics that are commonly used
in medical treatments (Fig. 1). The data are summarised
in Table 1. LRA-8 is capable of rapidly hydrolysing the
penam substrates ampicillin (kcat B 500 s1) and penicillin
G (kcat B 150 s1). These substrates do not bind very tightly
to the enzyme with Km values of B870 mM and B320 mM,
respectively, resulting in catalytic eciencies (kcat/Km ratios)
of B0.5 s1 mM1 for both.
The carbapenem representative (biapenem) is turned over
slightly more eciently (kcat/Km B 0.8 s1 mM1), largely due to
its enhanced binding interactions (i.e. low Km value). Among
the cephalosporins tested, only the non-clinical nitrocefin was
hydrolysed; no activity towards cefuroxime or cephalothin was
detected. This may suggest a distinct specificity that may guide
future inhibitor design/drug development studies. However,
with a kcat in the range of that measured for the penam
substrates and a Km similar to that of biapenem the catalytic
eciency towards nitrocefin is the highest among the sub-
strates tested here. It should also be pointed out that while for
some MBLs substrate inhibition was observed at substrate
concentrations above respective Km values no such inhibition
was observed for LRA-8.26,31,33
This journal is The Royal Society of Chemistry 2017
 View Article Online

Metallomics Paper
Published on 18 July 2017. Downloaded by ST LAWRENCE UNIVERSITY on 07/08/2017 14:20:42.

Fig. 2 Top: Partial sequence alignment including LRA-8 and the well-characterised B3 MBLs L1, AIM-1 and FEZ-1 using ENDscript software; the
secondary structure elements were obtained from the FEZ-1 structure.28 Highly conserved residues are highlighted in red and homologous residues are
in yellow. Metal binding residues are indicated with the appropriate label. Bottom: Active site of AIM-1 (PDB 4AWY).29,30 The grey spheres represent zinc
ions with the numbers indicating metal ion binding sites 1 and 2, respectively.

Table 1 Kinetic parameters of LRA-8. Data were collected in 50 mM are summarised. The table also contains information for two
TrisHCl, pH 8.5, containing 150 mM ZnCl2, using substrates representative MBL-like proteins identified in non-pathogenic marine organisms,
for major b-lactam groups
i.e. MIM-1 and MIM-2 from N. pentaromativorans and S. agarivorans,
kcat (s1) Km (mM) kcat/Km (s1 mM1) respectively. LRA-8 is approximately one order of magnitude
Ampicillin 490  40 870  140 0.55  0.15 less ecient than these other MBLs, largely due to weaker
Penicillin G 150  15 320  75 0.45  0.10 binding (i.e. larger Km values) of the substrates. Nonetheless,
Biapenem 35 5 40 5 0.8  0.2 LRA-8 displays a similar preference for penam-type substrates
Cefuroximea
Cephalothina (i.e. ampicillin, penicillin G) as these MBLs from the B3 subgroup.
Nitrocefin 90 5 50 5 1.8  0.5 Currently there is no clinically useful inhibitor for MBLs.14
a Captopril is, however, an in vitro inhibitor that has been useful
No hydrolysis was observed using substrate concentrations ranging
from 10 mM to 1280 mM. in functional studies of MBLs; while L-captopril is a drug used
to treat hypertension and in some cases of heart failure, in
particular the stereoisomer D-captopril is a potent inhibitor of
In Table 2 corresponding catalytic parameters for two several MBLs.34 Reported inhibition constants (Ki values) range
well-characterised MBLs from the B3 subfamily, AIM-1 and L1, from B6 mM for the B3 subgroup MIM-1 to B70 mM for the

This journal is The Royal Society of Chemistry 2017 Metallomics

 View Article Online

Paper Metallomics

Table 2 Kinetic parameters for the reactions of two known MBLs (AIM-1 and L1) and two MBL-like proteins (MIM-1 and MIM-2) with dierent b-lactam
substrates and standard deviations are shown, where available. Each of these proteins belongs to the B3 subgroup of MBLs

AIM-1 L1 MIM-1 MIM-2

kcat kcat/Km kcat kcat/Km kcat kcat/Km kcat kcat/Km
Ampicillin 150  5 6.2 580  20 1.9 665  40 2.9 230  5 1.6
Penicillin G 590  30 5.3 410  20 5.5 125  5 2.4 120  2 3.3
Biapenem 235  60 8.1 65  5 0.9 55  5 1.0 5.5  0.3 0.06
Cefuroxime 170  5 4.8 55  10 0.4 30  1 1.8 40  2 0.7
Published on 18 July 2017. Downloaded by ST LAWRENCE UNIVERSITY on 07/08/2017 14:20:42.

Data were reproduced from Leiros et al.,29 Selleck et al.,31 Crowder et al.32 and Miraula et al.26 for AIM-1, L1, MIM-1 and MIM-2, respectively.
The kcat values are in (s1) and the values of kcat/Km are in (s1 mM1).

MichaelisMenten-type behaviour was observed for each

Fig. 3 Inhibition of LRA-8 activity by D-captopril using penicillin G as
substrate. The inhibitor concentrations are zero (circle), 100 mM (squares),
150 mM (triangles) and 300 mM (diamonds). Individual data points were
recorded in triplicates and the averaged values and standard deviations are
indicated. The inhibition constants were obtained by a fit of the experi-
mental data to eqn (2) (see Materials and methods section).
B2-subgroup CphA.26 D-Captopril also inhibits LRA-8 activity,
albeit somewhat weaker than other MBLs (Fig. 3); nonetheless,
similar to other MBLs studied to date the mode of inhibition is
competitive (Kic B 110 mM, Kiuc 4 7 mM). Thus, in summary
the above substrate preference and D-captopril inhibition data
confirm that LRA-8 is an ecient MBL with catalytic para-
meters resembling those of well-characterised B3-type MBLs
such as AIM-1 or L1.
substrate at each pH in the range between 6.0 and 11.0
(data not shown). For each substrate kcat increases as the pH
increases until a maximal value is reached; a further increase in
pH does not lead to further changes (Fig. 4; top panel).
This behaviour is characteristic of a monoprotic system with
only one relevant protonation equilibrium. The data were thus
fit to eqn (3) (see Materials and methods section). It is noted
that biapenem is most rapidly hydrolysed only at a pH larger
than 9.0 (kcat,max B 125 s1 versus 35 s1 at pH 8.5; see Table 1).
In contrast to kcat, the pH dependence of the kcat/Km ratio varies
between dierent substrates (Fig. 4; bottom panel); while no
dependence is apparent for penicillin G, a single protonation
equilibrium is detected for the reaction with biapenem. These
data were thus also fit to eqn (3). For the reaction with
nitrocefin very high pH values lead to a decrease in the catalytic
eciency. The resulting bell-shaped profile is characteristic
of a diprotic system and was consequently fitted to eqn (4)
(see Materials and methods section).

Mechanism of action of LRA-8

Studying the pH dependence of the kinetic parameters of an
enzyme-catalysed reaction is a powerful approach to gain insight
into the details of the mechanism of action. In particular,
mechanistically relevant protonation equilibria may be uncovered
that allow the identification of residues that are important for the
chemical step (i.e. the hydrolytic step in the case of MBLs),
substrate binding, or both. While relatively few pH dependence
studies have been reported for MBLs, it appears that these
enzymes operate most eciently at pH values close to neutral.17
Furthermore, dierent substrates may lead to dierences in
observed pKa values, possibly a reflection of dierent binding
interactions and subtle mechanistic variations. Here, penicillin G,
biapenem and nitrocefin, substrates that represent three major Fig. 4 pH dependence of kcat (top panel) and kcat/Km (bottom panel) for
classes of b-lactams substrates (Fig. 1), were used to investigate the LRA-8-catalysed hydrolysis of penicillin G (blue), biapenem (black) and
the eect of pH on both kcat and kcat/Km (Fig. 4). nitrocefin (red).

Metallomics This journal is The Royal Society of Chemistry 2017


Metallomics
Table 3 Relevant pKa values for the LRA-8-catalysed hydrolysis of penicillin G,
biapenem and nitrocefin
Penicillin G Biapenem Nitrocefin
pKes 6.5  0.3 8.4  0.2 6.7  0.2
pKe1 6.6  0.4 6.9  0.1
pKe2 10.1  0.2
Published on 18 July 2017. Downloaded by ST LAWRENCE UNIVERSITY on 07/08/2017 14:20:42.
According to the model used to derive eqn (3) and (4) the
pH dependence of kcat provides an estimate for catalytically
relevant pKa values for the enzymesubstrate (ES) complex
(i.e. pKes), while the pH dependence of kcat/Km provides an estimate
for pKa values associated with the free enzyme (i.e. pKe).35,36
Relevant pKa values are summarised in Table 3. The residue likely
to be associated with pKes (and pKe1 in the reaction with biapenem
and nitrocefin) is a metal ion-bridging water molecule (m-H2O).
Deprotonation of this water molecule would strongly enhance its
nucleophilicity. Indeed, on related studies with other MBLs the
m-OH moiety was proposed to act as the hydrolysis-initiating
nucleophile.16,17,19,31 It thus appears that LRA-8 is likely to employ
a catalytic mechanism similar to that of well-characterised MBLs.
However, the data also indicate that different substrates may
interact differently with the active site of LRA-8. For the reaction
with nitrocefin, substrate binding has little effect on the pKa of the
bridging water molecule (pKe1 B pKes), while for the reaction with
biapenem the acidity of the bridging water is reduced by nearly
two orders of magnitude (pKe1 = 6.6 vs. pKes = 8.4).
A similar observation has been reported for other dinuclear
metallohydrolases such as purple acid phosphatases or some
organophosphate-degrading triesterases.3739 In those enzymes
the alkaline shift upon substrate binding was interpreted in
terms of a move of the bridging water ligand into a pseudo-
monodentate position, likely accompanied by an elongation
of the metalmetal distance in the active site. It is thus
hypothesised that in LRA-8 substrates such as biapenem and
nitrocefin bind in different modes to the metal ions in the
active site, bridging in the case of biapenem and monodentate
in the case of nitrocefin. No information about the binding of
penicillin G can be gained from the pH dependence of the catalytic
parameters, except that the affinity for this substrate decreases as
pH is increased, thus offsetting the beneficial effect of the depro-
tonation of the bridging water on the catalytic reactivity.
pKe2, only observed for the reaction with nitrocefin (Table 3
and Fig. 4), is tentatively assigned to the deprotonation of a
water molecule that is bound to only one divalent metal ion in
the active site, an assignment that is consistent with expected pKa
values for aquo complexes of transition metal ion complexes.40
In its deprotonated state this water ligand is slowly exchanged,
thus interfering with ecient substrate binding and/or product
release. In agreement with this interpretation this protonation
equilibrium is absent in the enzymesubstrate complex, likely
to be due to the replacement of the water ligand by the bound
substrate.
To probe the individual steps of the catalytic mechanism
further stopped-flow measurements were carried out. Nitrocefin
was used as substrate due to its beneficial spectroscopic
This journal is The Royal Society of Chemistry 2017
View Article Online
Paper
Fig. 5 Rapid-scan UV-Vis spectra from 300800 nm for the reaction of
50 mM of nitrocefin with 70 mM of LRA-8 in 50 mM TrisHCl, pH 8.5, at
25 1C. Top: The arrows indicate the loss of substrate (at 390 nm) and the
accumulation of product (at 485 nm). Bottom: The emergence and
disappearance of a spectral feature at B665 nm is shown. The time course
for the reactants as shown in Fig. 6 was determined from the baseline-
corrected absorbances at 390 nm (substrate), 485 nm (product) and
665 nm (intermediate).
properties; the intact molecule has a characteristic absorbance
maximum at 390 nm (e = 11 500 M1 cm1), while hydrolyzed
nitrocefin (i.e. after opening of the b-lactam ring) has a max-
imum at 485 nm (e = 17 420 M1 cm1).41 The reaction of 70 mM
of LRA-8 with 50 mM nitrocefin was monitored with a photo
diode array detector under single turnover conditions between
300800 nm (Fig. 5). As expected the substrate is consumed
rapidly as product is formed (indicated by arrows at 390 nm
and 485 nm, respectively, in Fig. 5). An additional peak at
665 nm emerged in the first 25 ms of the reaction and decayed
over the subsequent 150 ms (Fig. 5 and 6). The combined
spectroscopic features in Fig. 5, characterized by isosbestic
points around B430 and B590 nm, are indicative of a reaction
that involves the transient formation of a reaction intermediate,
as was previously reported for some MBLs.31,4245 Plots of the
absorption intensities of the substrate, intermediate and product
versus time are shown in Fig. 6.
The simplest model for the reaction mechanism employed
by LRA-8 is shown in Scheme 1 and includes the formation of a
short-lived intermediate (EI complex). A full numeric evaluation
of the model would require eight rate constants, four for the
forward and four for the reverse reactions (i.e. ki and ki, with
i = 14; Scheme 2). Several simplifications may be applied.
Firstly, the rapid association between the enzyme (E) and the
substrate (S) or product (P) were assumed to be limited by
diffusion (i.e. k1 and k4 B 108 M1 s1, fixed). Secondly, the
Metallomics
 View Article Online

Paper Metallomics
Published on 18 July 2017. Downloaded by ST LAWRENCE UNIVERSITY on 07/08/2017 14:20:42.

Scheme 2 Hydrolysis of nitrocefin (top) and kinetic model used for

modelling and fitting of the experimental data.

Fig. 6 Stopped-flow data analysis. The experimental data were analysed

Table 4 Rate constants for the reaction of LRA-8 with nitrocefin. The rate
using the mechanistic model presented in Scheme 2 and the corres-
constants were obtained using the mechanistic scheme illustrated in
ponding fits to the time course for the substrate, product and intermediate
Scheme 2. kcat and Km represent the theoretical values calculated using
are shown in red.
the KingAltman approach to derive rate equations for the reactions
illustrated in Schemes 1 and 2

rate constants k2 and k3 were set to zero because the hydrolysis k1 (M1 s1) 1.0  108 (fixed)
of nitrocefin is essentially irreversible. Furthermore, estimates k1 (s1) 1.0  104
k2 (s1) 485  11
for k2 and k3 were obtained from the rates of substrate depletion k3 (s1) 90  5
and product formation. These approximations were used to k4 (s1) 1.1  0.1  105
simulate the reaction using the program KINSIM.46 These k4 (M1 s1) 1.0  108 (fixed)
kcat (s1) 76  3
simulations indicate that the shape of calculated concentration KM (mM) 18  5
versus time plots (see Fig. 6) is not sensitive towards the
magnitude of k1, k4 and k4, but depends strongly on the values
of k1, k2 and k3. Table 4 summarises rate constants that were calculated from the constants in Table 4 and using the KingAltman
obtained from a reasonable simulation of the experimental approach35 are 76 s1 and 18 mM, also in good agreement with the
data. The parameters in Table 4 indicate that the slowest step steady-state parameters (i.e. B90 s1 and 50 mM).
in the reaction is the conversion of the intermediate to the Thus, LRA-8 is likely to employ a mechanism as depicted in
product, characterized by the rate constant k3 (B90 s1). Scheme 1, similar to that of a number of other MBLs such as
Based on studies with biomimetics and other MBLs it was NDM-1, BcII and L1, with the decay of a reaction intermediate
previously proposed that the structure of this intermediate as the rate-limiting step.43,45
is an anionic adduct where the nucleophilic oxygen forms
a bond to the electrophilic carbon atom of the b-lactam bond Insight into the active site structure of LRA-8
(Scheme 1).16,40,47 The discussion above demonstrates that LRA-8 is an enzyme
It thus follows that breakage of the b-lactam bond is the with catalytic properties characteristic of MBLs. In the absence
rate-limiting step in the reaction. It should be pointed out that of crystallographic information about the structure of this
k3 is of similar magnitude as kcat, the first order rate constant enzyme we employed spectroscopic techniques to compare its
measured under steady-state conditions (B90 s1 at pH 8.5; active site structure with other MBLs for which corresponding
Table 1). Furthermore, the theoretical kcat and Km values data are available. For this purpose the native Zn(II) cofactor

Scheme 1 Proposed hydrolysis mechanism of nitrocefin by LRA-8, modified from a model previously reported for an MBL from Bacteroides fragilis;47
the rate limiting step is the decay of the anionic reaction intermediate. The hydrolysis-initiating nucleophile is a hydroxide activated by the two metal ions
in the active site.

Metallomics This journal is The Royal Society of Chemistry 2017

 View Article Online

Metallomics Paper

was replaced with the paramagnetic Co(II) ion, providing the Table 5 Fitted electronic parameters for MCD spectra and VTVH MCD
basis for both Magnetic Circular Dichroism (MCD) and Electron data
Paramagnetic Resonance (EPR) measurements. Similar to the Parameter 464 nm 491 nm 509 nm
B3-type MBL AIM-1 the replacement of Zn(II) by Co(II) leads to 1
J (cm ) 0.29 0.31 0.27
a B50% decrease in catalytic activity and a moderate eect on Da (6-coordinate) (cm1) Z50 Z50 Z50
substrate (penicillin) binding (i.e. kcat B150 s1 and B60 s1 Db (6-coordinate) (cm1)
and Km B300 mM and B350 mM, respectively, for the Zn(II) and E (a,b metal) (cm1) 0.0 0.0 0.0
Co(II) derivatives of LRA-8 at pH 8.5). Similarly, the competitive
Published on 18 July 2017. Downloaded by ST LAWRENCE UNIVERSITY on 07/08/2017 14:20:42.

inhibitory eect of D-captopril is also reduced from B100 mM for

Table 6 Summary of ligand field calculations
the Zn(II) derivative to B250 mM for the Co(II) derivative. Despite
these catalytic changes the Co(II) derivative represents an active dd transition (nm)
form of LRA-8, suitable for spectroscopic investigations of the Origin in Oh Observed Calculated
structure and mechanism of this enzyme. In Fig. 7 the MCD
6-Coordinate species 4
T1g - 2
T1g(G) 464 460
spectrum of LRA-8 is shown, together with VTVH MCD 4
T1g - 4
T1g(P) 491 495
data recorded at three of the major transitions. The spectrum
4
T1g - 4
T1g(P) 502 503
4
T1g - 4
T1g(P) 509 517
could be fitted to four transitions between 450 to 550 nm. The
VTVH MCD data at 464 nm, 491 nm and 509 nm were
extracted and analysed as described in detail elsewhere.48 This interpretation was further probed by comparing the
In brief, the data were fit to the dimer model, resulting
EPR spectra of these two enzymes (Fig. 8).31 Their respective
in relevant fitting parameters summarised in Table 5. continuous wave (CW) EPR spectra are similar and characteristic
The bands at 491 nm and 509 nm are the main spin-allowed of a bimetallic Co(II) system with five or six ligands (the data are
4
T1g(P) transitions from the two six-coordinate Co(II) ions not sensitive enough to distinguish between five or six ligands in
in the active site. Both metal ions can also have a spin-doublet the coordination environment of the metal centres).31,44,45 Fig. 8
around 464 nm (Table 6). Consistently, the VTVH MCD analysis also shows that there is little dierence in the spectral features
of the 464 nm band indicates that the Co(II) ion(s) associated upon the addition of the competitive inhibitor D-captopril; the
with this transition is weakly and ferromagnetically exchange- low field shoulder for both spectra exhibit splittings from a
coupled ( J = 0.29 cm1) to another six-coordinate Co(II) ion
cobalt hyperfine interaction, which are essentially identical.
that is associated with either the 491 nm or 509 nm transition While there is a reduction in the intensity of the sharp feature
(which have virtually identical electronic structural parameters; around 340 mT upon the addition of D-captopril, overall the data
Table 5). suggest that the inhibitor interacts only weakly with the active
The only MBL that has previously been analysed by MCD in site metal ions. This interpretation is again in agreement with a
a similar manner is AIM-1,31 where both Co(II) species are also similar observation reported for AIM-1.31
likely to be six-coordinate with a similar weak ferromagnetic The spectra were simulated using XSophe (Bruker Biospin),
coupling. Thus, it is evident that the active site structures of assuming H0 = bB0gS/h  + SDS, where S = 3/2, |D| c |bgBS/h  |.
both LRA-8 and AIM-1 are very similar.

Fig. 8 X-band (9.36 GHz) CW EPR data recorded in perpendicular (top)

Fig. 7 MCD spectrum (at 1.3 K and 7 T) of LRA-8 at pH 7.0 (top) and fitted
VTVH MCD data (bottom) for the 464, 491 and 509 nm transitions. Tem-
peratures of red, green, blue, yellow, grey, black represent data observed at
1.4, 3.0, 6.0, 12.0, 24.0 and 48.0 K, respectively.
and parallel (bottom) mode. Shown are spectra for LRA-8 (blue) and LRA-8
in the presence of the inhibitor D-captopril (red), and for comparison the
corresponding spectra from the enzyme AIM-1 (cyan and magenta). The
simulation for a Co(II) high-spin species is shown (black). The measure-
ments were carried out at 20 K under non-saturating conditions using a
modulation amplitude of 1 mT and modulation frequency of 100 kHz.

This journal is The Royal Society of Chemistry 2017 Metallomics


Paper
Note that when D 4 0 the MS = 1/2 Kramers doublet lies
lowest and all observed EPR transitions are from this doublet;
if D o 0 the MS = 3/2 Kramers doublet lie lowest. The
simulation parameters were D = 14 GHz, E = 0.95 GHz, with
gx, gy, gz = 2.10, 2.29, 2.18, respectively, with the corresponding
linewidths of 1.2, 0.61 and 0.35 (modelled by g-strain). The
model assumes that the magnetic interaction between the two
cobalt ions is small (consistent with the MCD data) and not
Published on 18 July 2017. Downloaded by ST LAWRENCE UNIVERSITY on 07/08/2017 14:20:42.
resolved by the EPR spectrum. Not surprisingly, these para-
meters are similar to those obtained from the simulation of the
corresponding EPR data for the enzyme AIM-1.
Thus, in summary the comparison of the spectral para-
meters of LRA-8 and AIM-1 indicate that both enzymes have a
conserved active site geometry, as would be expected based on
the similarity of their kinetic parameters.
Conclusions
Antibiotic resistance is a major global health concern, one that
threatens to derail the benefits garnered from arguably the
greatest success of modern medicine, the discovery of antibiotics.
MBLs are a major contributor to the emergence and spread of
resistance, partly because there are currently no clinically
useful inhibitors available, and partly because these enzymes
are frequently located on mobile genetic elements that may be
easily transferred from one microorganism to another.
The discovery of MBL-like enzymes in microorganisms that
are not in contact with the human population is thus of
particular concern as these proteins already have the in-built
capacity to inactivate antibiotics, even though they may not
need MBL activity for their survival. These enzymes are likely to
have dierent primary functions, relevant to the organisms
metabolism. Thus, studying the properties of such MBL-like
proteins may provide insight into the structural plasticity of
this family of enzymes that may facilitate functional promiscuity,
while important insight into the evolution of MBLs may also be
gained. The structural plasticity may also be a major factor that
has contributed to the current lack of clinically useful inhibitors
for MBLs. In this respect, the demonstration that LRA-8 is indeed
a potent MBL that is likely to employ a reaction mechanism
similar to that of well-known virulent MBLs is of great relevance
for inhibitor design, but also for functional investigations into
mechanistic factors that may contribute to promiscuity. Eorts
towards elucidating the crystal structure of the enzyme are
currently in progress.
Materials and methods
The recombinant expression system containing the full-length
LRA-8-encoding sequence, modified by the addition of an
N-terminal hexa-histidine tag, in the vector pJ411 was purchased
from DNA 2.0 (see Fig. S1 (ESI) for a map of the construct). E. coli
BL21 (DE3) cells used for protein expression were purchased from
New England BioLabs. All chemicals used were purchased from
Sigma-Aldrich unless stated otherwise.
Metallomics
View Article Online
Metallomics
Protein expression and purifications
All media used for the expression of LRA-8 were supplemented
with 50 mg ml1 of kanamycin. The LRA-8-encoding construct
LRA8::pJ411 was used to transform competent E. coli BL21
(DE3) cells by heat shock. The culture was grown at 37 1C and
then transferred into TB medium (1 L) and grown at 37 1C until
an optical density (OD600) of 0.40.6 was reached. The tempera-
ture was then reduced to 18 1C and the culture were incubated
for an additional 48 hours. Subsequently, the cells were
harvested by centrifugation at 4 1C. The cells were resus-
pended in lysis buffer (50 mM TrisHCl, pH 8.5, 5 mM
b-mercaptoethanol, 150 mM NaCl, a protease inhibitor cock-
tail and 1 mg mL1 lysozyme), and then disrupted by sonica-
tion. The supernatant was loaded onto a 25 mL HiTrapQ FF
Ni(II) affinity column (Ni(II)-IMAC resin; GE Healthcare), equi-
librated with 50 mM TrisHCl buffer, pH 8.5, containing
150 mM NaCl, 5 mM b-mercaptoethanol and 20 mM imida-
zole. The protein was eluted isocratically using 500 mM
imidazole. The enzyme was further purified using a S-300
gel filtration column equilibrated with 50 mM TrisHCl,
pH 8.5, containing 0.2 M NaCl and 0.15 mM of ZnCl2.
An average expression resulted in a yield of 12 mg of purified
LRA-8 per litre of culture medium. The purity of the enzyme
was estimated by SDS-PAGE analysis to be at least 95%;
the molecular weight of the isolated enzyme is estimated to
B34 kDa, in good agreement with both the calculated mass and
the weight recorded by mass spectrometry (Fig. S2, ESI). The
metal ion content was determined by atomic absorption spectro-
scopy and confirmed the presence of two (2.1  0.2) Zn ions per
active site. It is noted that no activity was detected in the flow-
through of the Ni(II)-IMAC column after loading the enzyme
sample, indicating that the full-length recombinant protein
remained intact (i.e. no posttranslational modification did
occur), in agreement with the molecular weight estimated from
the SDS-PAGE gel (Fig. S2, ESI).
Steady state enzyme kinetics
All kinetic measurements were carried out with a Cary 60 Bio
Varian UV-Vis spectrophotometer. The hydrolysis of ampicillin
(e235nm = 890 M1 cm1), penicillin G (e243nm = 936 M1 cm1),
cefuroxime (e235nm = 9320 M1 cm1), cephalothin (e293nm =
8790 M1 cm1), biapenem (e293nm = 7600 M1 cm1) and
nitrocefin (e485nm = 17 400 M1 cm1) was monitored in 50 mM
TrisHCl (pH 8.5), containing 150 mM ZnCl2 (or 100 mM CoCl2 in
case of the Co(II) derivative of LRA-8), at 25 1C. The buffers were
treated prior to use with Chelex (Bio-Rad). The MichaelisMenten
parameters (kcat, kcat/Km) were determined by non-linear regression
Recombinant LRA-8 contains a non-cleavable hexa-histidine tag. However, its
presence is unlikely to aect the metal ion content and catalytic properties of the
enzyme significantly. This interpretation is supported by the observation that in
the Co(II) derivative of LRA-8 the metal ions exclusively bind in the close proximity
provided by the catalytic active site (Fig. 7 and 8). Furthermore, the closely related
enzyme AIM-1 was purified in the absence of a hexa-histidine tag and the
spectroscopic properties are indeed similar to those of LRA-8.31
This journal is The Royal Society of Chemistry 2017

Metallomics
analysis (eqn (1)35) of the experimental data using GraphPad
Prism 6 software.
Vmax  S
v (1)
Km S
Here, v, Vmax and Km represent the measured catalytic rate, the
maximal rate and the Michaelis constant, respectively, while S
indicates the substrate concentration.
Published on 18 July 2017. Downloaded by ST LAWRENCE UNIVERSITY on 07/08/2017 14:20:42.
Inhibition assays were conducted in the same manner as
described above using the substrate penicillin G and various
concentrations of the inhibitor D-captopril. The data were fitted
to a general inhibition equation that allows for both a compe-
titive and uncompetitive mode of inhibitor binding (eqn (2)).35
Kic and Kiuc represent the inhibitor dissociation constants for
competitive and uncompetitive binding, respectively, and I is
the inhibitor concentration.
Vmax  S
v     (2)
I I
Km 1 1 S
Kic Kiuc
The pH profiles of these parameters were determined using
nitrocefin as the substrate. The pH for these assays ranged from
7.0 to 11 using a 20 mM acetate, 20 mM MES, 20 mM HEPES,
20 mM CHES and 20 mM CAPS multi-component buer system.
The experimental data were fitted using equations ((3) and (4),
respectively) for mono- or diprotic systems as described
elsewhere.35,36
0 1
B c C
Log y LogB
@
C (3)
HA
1
K1
0 1
B c C detail elsewhere.31,48,5357
Log y LogB
@
C (4)
H K2 A
1
K1 H
Here, H is the proton concentration, Ki represents the acid
dissociation constant for either the enzymesubstrate complex
(ES) or the free enzyme (E), c is the pH independent values of
y, and y is kcat or kcat/Km.
Pre-steady state kinetics measurements
All pre-steady state kinetics measurements were carried out
in 50 mM TrisHCl, pH 8.5, at 25 1C using an Applied
Photophysics SX-18 spectrometer, coupled with a photodiode
array detector. Data were collected under single turnover
conditions using a solution of 70 mM of LRA-8 in one syringe
and a solution of nitrocefin (50 mM) in the other. Absorbance
changes were monitored with the photodiode array detector
(for 5 seconds over the wavelength range of 300800 nm) as
described elsewhere.31,4245,49
Data from at least five reproducible experiments were
collected, averaged and corrected for the instrument dead time
(1.5 ms). The kinetic mechanism for the reaction was then
analysed using the program KINSIM and the model shown
This journal is The Royal Society of Chemistry 2017
View Article Online
Paper
in Scheme 2.47,50,51 The data were then fitted using Reactlab
software.
Spectroscopic characterization of LRA-8
The active site structure of LRA-8 was probed using MCD and
CW EPR. In order to use these methods Zn(II) ions were
replaced by paramagnetic Co(II). The preparation of the apo-
enzyme was performed following a previously published
procedure31,52,53 by adding 10 mM EDTA in 20 mM Hepes
buer, pH 7.0, to the purified enzyme. After 24 h incubation
at 4 1C the chelating agent was removed using a desalting
column (Econo-Pac 10DG from Bio-Rad) that was equilibrated
with 20 mM Hepes buer, pH 7.0. Subsequently, three equi-
valents of Co(II) were added to the metal-free protein solution
and the mixture was incubated over night at 4 1C. The excess of
Co(II) was removed using again a desalting column, pre-
equilibrated with 50 mM TrisHCl buffer, pH 8.0. The concen-
tration of the resulting sample was 1.75 mM with nearly two
equivalents of bound Co(II) (1.8  0.1, determined by atomic
absorption spectroscopy). For spectroscopic measurements this
solution was diluted with glycerol to a final concentration
of 0.7 mM (i.e. a 3 : 2 glycerol : buffer mixture).
For MCD measurements an aliquot was transferred to a
0.62 cm path length nickel-plated copper sample cell with quartz
windows. The MCD system used has a JASCO J815 spectropolari-
meter and an Oxford Instruments SM4000 cryostat/magnet. Data
were collected at 7.0 T and 1.4 K. Variable-temperature, variable-
field (VTVH) data were collected at increments of 0.5 T from 0
to 7.0 T and at temperatures of 1.4, 3, 6, and 12 K. The
experimental spectra were plotted as a function of wavenumbers
and fitted to a minimum number of Gaussian peaks to achieve
the final composite spectrum using the GRAMS AI software
package. The data were subsequently analysed as described in
EPR measurements were carried out with the same sample
used for MCD experiments. Low temperature X-Band CW EPR
spectra were recorded on a Bruker EMX EPR spectrometer
equipped with an Oxford Instruments ESR900 helium flow cryostat.
Spectra were recorded at 9.64 GHz (B0>B1) and 9.38 GHz (B08B1)
using an ER4116DM dual-mode cavity, with 10 G (1 mT) magnetic
field modulation at 100 kHz. Spin Hamiltonian parameters were
estimated from computer simulations carried out using XSophe
(Bruker Biospin), assuming H0 = bB0gS/h  + SDS, where S = 3/2,
|D| c |bgBS/h  |, and where D 4 0 implies the MS = 1/2
Kramers doublet lies lowest and all observed EPR transitions
are from this doublet, and D o 0 implies the MS = 3/2
Kramers doublet lie lowest and all observed EPR transitions
are from this doublet.
Statement of contributions
MP, CS and CE were responsible for most of the data collection
and the initial draft of the paper; JH, WC and DT contributed to
EPR data collection and analysis; NM, WH and JL contributed
to MCD data collection and analysis; PH and GT contributed to
Metallomics

Paper
target selection and, together with NM, DT and JL, they also
contributed to the refinement of the draft manuscript; GS was
responsible for the overall project, the outline and refinement
of the manuscript and the allocation of individual tasks of the
project.
Acknowledgements
Published on 18 July 2017. Downloaded by ST LAWRENCE UNIVERSITY on 07/08/2017 14:20:42.
This research was supported by Project Grants from the NH&MRC
(APP1084778) and Australian Research Council (DP150104358)
and a Future Fellowships (FT120100694) awarded to GS. CS
acknowledges financial support in form of an APA Scholarship
from the University of Queensland and NM is grateful to the
Science Foundation Ireland for funding in form of a President of
Ireland young Researcher Award (SFI-PIYRA). JAL and WH thank
the National Science Foundation (USA) for financial support from
grants CHE1303852 and CHE0820965 (MCD instrument). The
authors are grateful to the reviewers for their constructive support
in the preparation of this manuscript.
References
1 P. M. Hawkey, Multidrug-resistant Gram-negative bacteria:
a product of globalization, J. Hosp. Infect., 2015, 89, 241247.
2 H. K. Allen, J. Donato, H. H. Wang, K. A. Cloud-Hansen,
J. Davies and J. Handelsman, Call of the wild: antibiotic
resistance genes in natural environments, Nat. Rev. Microbiol.,
2010, 8, 251259.
3 S. Thiele-Bruhn, Pharmaceutical antibiotic compounds in
soils a review, J. Plant Nutr. Soil Sci., 2003, 166, 145167.
4 G. Dantas, M. O. Sommer, R. D. Oluwasegun and G. M.
Church, Bacteria subsisting on antibiotics, Science, 2008,
320, 100103.
5 S. Gottig, A. G. Hamprecht, S. Christ, V. A. Kempf and
T. A. Wichelhaus, J. Antimicrob. Chemother., 2013, 68,
17371740.
6 K. J. Forsberg, A. Reyes, B. Wang, E. M. Selleck, M. O.
Sommer and G. Dantas, The shared antibiotic resistome
of soil bacteria and human pathogens, Science, 2012, 337,
11071111.
7 C. S. Riesenfeld, R. M. Goodman and J. Handelsman,
Uncultured soil bacteria are a reservoir of new antibiotic
resistance genes, Environ. Microbiol., 2004, 6, 981989.
8 H. K. Allen, L. A. Moe, J. Rodbumrer, A. Gaarder and
J. Handelsman, Functional metagenomics reveals diverse
b-lactamases in a remote Alaskan soil, ISME J., 2009, 3,
243251.
9 J. C. Clemente, E. C. Pehrsson, M. J. Blaser, K. Sandhu,
Z. Gao, B. Wang, M. Magris, G. Hidalgo, M. Contreras,
O. Noya-Alarcon, O. Lander, J. McDonald, M. Cox, J. Walter,
P. L. Oh, J. F. Ruiz, S. Rodriguez, N. Shen, S. J. Song, J. Metcalf,
R. Knight, G. Dantas and M. G. Dominguez-Bello, The
microbiome of uncontacted Amerindians, Sci. Adv., 2015,
1, e1500183.
Metallomics
View Article Online
Metallomics
10 L. P. Kotra and S. Mobashery, b-Lactam antibiotics,
b-lactamases and bacterial resistance, Bull. Inst. Pasteur,
1998, 96, 139150.
11 S. Ghafourian and N. Sadeghifard, Extended spectrum beta-
lactamases: definition, classification and epidemiology,
Curr. Issues Mol. Biol., 2015, 17, 1121.
12 B. G. Hall and M. Barlow, Evolution of the serine
b-lactamases: past, present and future, Drug Resist. Updates,
2004, 7, 111123.
13 I. E. Crompton, B. K. Cuthbert, G. Lowe and S. G. Waley,
b-Lactamase inhibitors. The inhibition of serine b-lactamases
by specific boronic acids, Biochem. J., 1988, 251, 453459.
14 R. P. McGeary, G. Schenk and L. W. Guddat, The applica-
tions of binuclear metallohydrolases in medicine: recent
advances in the design and development of novel drug leads
for purple acid phosphatases, metallo-b-lactamases and
arginases, Eur. J. Med. Chem., 2014, 76, 132144.
15 M. Shahid, F. Sobia, A. Singh, A. Malik, H. M. Khan, D. Jonas
and P. M. Hawkey, b-Lactams and b-lactamase-inhibitors
in current- or potential-clinical practice: a comprehensive
update, Crit. Rev. Microbiol., 2009, 35, 81108.
16 M. W. Crowder, J. Spencer and A. J. Vila, Metallo-b-lactamases:
novel weaponry for antibiotic resistance in bacteria, Acc. Chem.
Res., 2006, 39, 721728.
17 E. K. Phelan, M. Miraula, C. Selleck, D. L. Ollis, G. Schenk
and N. Mitic, Metallo-b-Lactamases: A Major Threat to
Human Health, Am. J. Mol. Biol., 2014, 4, 89104.
18 A. I. Karsisiotis, C. F. Damblon and G. C. Roberts, A variety
of roles for versatile zinc in metallo-b-lactamases, Metallomics,
2014, 6, 11811197.
19 N. Mitic, M. Miraula, C. Selleck, K. S. Hadler, E. Uribe,
M. M. Pedroso and G. Schenk, Catalytic mechanisms of
metallohydrolases containing two metal ions, Adv. Protein
Chem. Struct. Biol., 2014, 97, 4981.
20 C. Bebrone, Metallo-b-lactamases (classification, activity,
genetic organization, structure, zinc coordination) and their
superfamily, Biochem. Pharmacol., 2007, 74, 16861701.
21 G. Cornaglia, H. Giamarellou and G. M. Rossolini, Metallo-
b-lactamases: a last frontier for b-lactams, Lancet, 2011, 11,
381393.
22 F. von Nussbaum and G. Schier, Aspergillomarasmine A,
an inhibitor of bacterial metallo-beta-lactamases conferring
blaNDM and blaVIM resistance, Angew. Chem., Int. Ed.,
2014, 53, 1169611698.
23 C. F. D. Hou, E. K. Phelan, M. Miraula, D. L. Ollis, G. Schenk
and N. Mitic, Unusual metallo-b-lactamases may constitute
a new subgroup in this family of enzymes, Am. J. Mol. Biol.,
2014, 4, 1115.
24 P. Vella, M. Miraula, E. Phelan, E. W. Leung, F. Ely,
D. L. Ollis, R. P. McGeary, G. Schenk and N. Mitic, Identifi-
cation and characterization of an unusual metallo-b-
lactamase from Serratia proteamaculans, J. Biol. Inorg. Chem.,
2013, 18, 855863.
25 A. Makena, J. Brem, I. Pfeer, R. E. Geen, S. E. Wilkins,
H. Tarhonskaya, E. Flashman, L. M. Phee, D. W. Wareham
and C. J. Schofield, Biochemical characterization of New
This journal is The Royal Society of Chemistry 2017

Metallomics
Delhi metallo-beta-lactamase variants reveals dierences
in protein stability, J. Antimicrob. Chemother., 2015, 70,
463469.
26 M. Miraula, J. J. Whitaker, G. Schenk and N. Mitic, b-Lactam
antibiotic-degrading enzymes from non-pathogenic marine
organisms: a potential threat to human health, J. Biol. Inorg.
Chem., 2015, 20, 639651.
27 M. Miraula, G. Schenk and N. Mitic, Promiscuous metallo-b-
Published on 18 July 2017. Downloaded by ST LAWRENCE UNIVERSITY on 07/08/2017 14:20:42.
lactamases: MIM-1 and MIM-2 may play an essential role
in quorum sensing networks, J. Inorg. Biochem., 2016, 162,
366375.
28 X. Robert and P. Gouet, Deciphering key features in protein
structures with the new ENDscript server, Nucleic Acids Res.,
2014, 42, 4.
29 H. K. Leiros, P. S. Borra, B. O. Brandsdal, K. S. Edvardsen,
J. Spencer, T. R. Walsh and O. Samuelsen, Antimicrob. Agents
Chemother., 2012, 56, 43414353.
30 C.-F. D. Hou, J.-W. Liu, C. Collyer, N. Mitic, M. Pedroso,
G. Schenk and D. L. Ollis, Insights into an evolutionary
strategy leading to antibiotic resistance, Sci. Rep., 2017,
7, 40357.
31 C. Selleck, J. A. Larrabee, J. Harmer, L. W. Guddat, N. Mitic,
W. Helweh, D. L. Ollis, W. A. Craig, D. L. Tierney, M. M.
Pedroso and G. Schenk, AIM-1: an antibiotic-degrading
metallohydrolase that displays mechanistic flexibility,
Chem. Eur. J., 2016, 22(49), 1770417714.
32 M. W. Crowder, T. R. Walsh, L. Banovic, M. Pettit and
J. Spencer, Overexpression, purification, and characteriza-
tion of the cloned metallo-b-lactamase L1 from Stenotropho-
monas maltophilia, Antimicrob. Agents Chemother., 1998, 42,
921926.
33 M. P. Yanchak, R. A. Taylor and M. W. Crowder, Mutational
analysis of metallo-b-lactamase CcrA from Bacteroides
fragilis, Biochemistry, 2000, 39, 1133011339.
34 J. Brem, S. S. Berkel, D. Zollman, S. Y. Lee, O. Gileadi, P. J.
McHugh, T. R. Walsh, M. A. McDonough and C. J. Schofield,
Structural basis of metallo-b-lactamase inhibition by captopril
stereoisomers, Antimicrob. Agents Chemother., 2016, 60,
142150.
35 I. H. Segel, Enzyme kinetics: Behavior and analysis of rapid
equilibrium and steady-state enzyme systems, John Wiley &
Sons, USA, 1993.
36 A. Kantacha, R. Buchholz, S. J. Smith, G. Schenk and
L. R. Gahan, Phosphate ester cleavage promoted by a
tetrameric iron(III) complex, J. Biol. Inorg. Chem., 2011, 16,
2532.
37 G. Schenk, N. Mitic, L. R. Gahan, D. L. Ollis, R. P. McGeary
and L. W. Guddat, Binuclear metallohydrolases: complex
mechanistic strategies for a simple chemical reaction, Acc.
Chem. Res., 2012, 45, 15931603.
38 G. Schenk, N. Mitic, G. R. Hanson and P. Comba, Purple
acid phosphatase: a journey into the function and mecha-
nism of a colorful enzyme, Coord. Chem. Rev., 2013, 257,
473482.
39 F. Ely, K. S. Hadler, L. R. Gahan, L. W. Guddat, D. L. Ollis
and G. Schenk, The organophosphate-degrading enzyme
This journal is The Royal Society of Chemistry 2017
View Article Online
Paper
from Agrobacterium radiobacter displays mechanistic flex-
ibility for catalysis, Biochem. J., 2010, 432, 565573.
40 L. J. Daumann, G. Schenk, D. L. Ollis and L. R. Gahan,
Spectroscopic and mechanistic studies of dinuclear metal-
lohydrolases and their biomimetic complexes, Dalton
Trans., 2014, 43, 910928.
41 C. H. OCallaghan, A. Morris, S. M. Kirby and A. H. Shingler,
Novel method for detection of b-lactamases by using a
chromogenic cephalosporin substrate, Antimicrob. Agents
Chemother., 1972, 1, 283288.
42 H. Yang, M. Aitha, A. M. Hetrick, T. K. Richmond, D. L.
Tierney and M. W. Crowder, Mechanistic and spectroscopic
studies of metallo-b-lactamase NDM-1, Biochemistry, 2012,
51, 38393847.
43 H. Yang, M. Aitha, A. R. Marts, A. Hetrick, B. Bennett, M. W.
Crowder and D. L. Tierney, Spectroscopic and mechanistic
studies of heterodimetallic forms of metallo-b-lactamase
NDM-1, J. Am. Chem. Soc., 2014, 136, 72737285.
44 M. J. Hawk, R. M. Breece, C. E. Hajdin, K. M. Bender, Z. Hu,
A. L. Costello, B. Bennett, D. L. Tierney and M. W. Crowder,
Dierential binding of Co(II) and Zn(II) to metallo-b-lactamase
Bla2 from Bacillus anthracis, J. Am. Chem. Soc., 2009, 131,
1075310762.
45 J. D. Garrity, B. Bennett and M. W. Crowder, Direct evidence
that the reaction intermediate of metallo-b-lactamase L1 is
metal bound, Biochemistry, 2005, 44, 10781087.
46 Q. Dang and C. Frieden, New PC versions of the kinetic-
simulation and fitting programs, KINSIM and FITSIM,
Trends Biochem. Sci., 1997, 22, 317.
47 Z. Wang, W. Fast, A. M. Valentine and S. J. Benkovic,
Metallo-b-lactamase: structure and mechanism, Curr. Opin.
Chem. Biol., 1999, 3, 614622.
48 J. A. Larrabee, G. Schenk, N. Mitic and M. J. Riley, Use of
magnetic circular dichroism to study dinuclear metallo-
hydrolases and the corresponding biomimetics, Eur. Biophys.
J., 2015, 44, 393415.
49 S. McManus-Munoz and M. W. Crowder, Kinetic mecha-
nism of metallo-b-lactamase L1 from Stenotrophomonas
maltophilia, Biochemistry, 1999, 38, 15471553.
50 C. Frieden, Numerical integration of rate equations by
computer, Trends Biochem. Sci., 1993, 18, 5860.
51 C. Frieden, Numerical integration of rate equations by
computer: an update, Trends Biochem. Sci., 1994, 19,
181182.
52 M. M. Pedroso, F. Ely, T. Lonhienne, L. R. Gahan, D. L. Ollis,
L. W. Guddat and G. Schenk, Determination of the catalytic
activity of binuclear metallohydrolases using isothermal
titration calorimetry, J. Biol. Inorg. Chem., 2014, 19, 389398.
53 M. M. Pedroso, F. Ely, N. Mitic, M. C. Carpenter, L. R.
Gahan, D. E. Wilcox, J. L. Larrabee, D. L. Ollis and G. Schenk,
Comparative investigation of the reaction mechanisms
of the organophosphate-degrading phosphotriesterases from
Agrobacterium radiobacter (OpdA) and Pseudomonas diminuta
(OPH), J. Biol. Inorg. Chem., 2014, 19, 12631275.
54 F. Ely, K. S. Hadler, M. Mitic, L. Gahan, D. L. Ollis,
J. A. Larrabee and G. Schenk, Electronic and geometric
Metallomics

Paper
structure of the organophosphate-degrading enzyme from
Agrobacterium radiobacter, J. Biol. Inorg. Chem., 2011, 16,
777787.
55 K. S. Hadler, N. Mitic, F. Ely, G. R. Hanson, L. R. Gahan,
J. A. Larrabee, D. L. Ollis and G. Schenk, Structural
Flexibility Enhances the Reactivity of the Bioremediator
Glycerophosphodiesterase by Fine-Tuning Its Mechanism of
Hydrolysis, J. Am. Chem. Soc., 2009, 131, 1190011908.
Published on 18 July 2017. Downloaded by ST LAWRENCE UNIVERSITY on 07/08/2017 14:20:42.
56 K. S. Hadler, E. A. Tanifum, S. H.-C. Yip, N. Mitic, L. W.
Guddat, C. J. Jackson, L. R. Gahan, K. Nguyen, P. D. Carr,
Metallomics
View Article Online
Metallomics
D. L. Ollis, A. C. Hengge, J. A. Larrabee and G. Schenk,
Substrate-Promoted Formation of a Catalytically Compe-
tent Binuclear Center and Regulation of Reactivity in a
Glycerophosphodiesterase from Enterobacter aerogenes,
J. Am. Chem. Soc., 2008, 130, 1412914138.
57 K. S. Hadler, N. Mitic, S. H.-C. Yip, L. R. Gahan, D. L. Ollis,
G. Schenk and J. A. Larrabee, Electronic Structure Analysis
of the Dinuclear Metal Center in the Bioremediator Glycero-
phosphodiesterase (GpdQ) from Enterobacter aerogenes,
Inorg. Chem., 2010, 49, 27272734.
This journal is The Royal Society of Chemistry 2017