An evolutionarily conserved alternate metal ligand is important for activity in α-isopropylmalate synthase from Mycobacterium tuberculosis

Article abstract of DOI:10.1016/j.bbapap.2014.07.013

Members of the DRE-TIM metallolyase superfamily rely on an active-site divalent cation to catalyze various reactions involving the making and breaking of carbon-carbon bonds. While the identity of the metal varies, the binding site is well-conserved at the superfamily level with an aspartic acid and two histidine residues acting as ligands to the metal. Previous structural and bioinformatics results indicate that the metal can adopt an alternate architecture through the addition of an asparagine residue as a fourth ligand. This asparagine residue is strictly conserved in all members of the DRE-TIM metallolyase superfamily except fungal homocitrate synthase (HCS-lys) where it is replaced with isoleucine. The role of this additional metal ligand in α-isopropylmalate synthase from Mycobacterium tuberculosis (MtIPMS) has been investigated using site-directed mutagenesis. Substitution of the asparagine ligand with alanine or isoleucine results in inactive enzymes with respect to α-isopropylmalate formation. Control experiments suggest that the substitutions have not drastically affected the enzyme's structure indicating that the asparagine residue is essential for catalysis. Interestingly, all enzyme variants retained acetyl CoA hydrolysis activity in the absence of α-ketoisovalerate, similar to the wild-type enzyme. In contrast to the requirement of magnesium for α-isopropylmalate formation, hydrolytic activity could be inhibited by the addition of magnesium chloride in wild-type, D81E, and N321A MtIPMS, but not in the other variants studied. Attempts to rescue loss of activity in N321I MtIPMS by mimicking the fungal HCS active site through the D81E/N321I double variant were unsuccessful. This suggests epistatic constraints in evolution of function in IPMS and HCS-lys enzymes.

Full text of DOI:10.1016/j.bbapap.2014.07.013

Biochimica et Biophysica Acta 1844 (2014) 1784–1789
Contents lists available at ScienceDirect
Biochimica et Biophysica Acta
journal homepage: www.elsevier.com/locate/bbapap
An evolutionarily conserved alternate metal ligand is important for
activity in α-isopropylmalate synthase from
☆
Mycobacterium tuberculosis
⁎
Patrick A. Frantom , Yuliya Birman, Brittani N. Hays, Ashley K. Casey
Department of Chemistry, The University of Alabama, 250 Hackberry Lane, Tuscaloosa, AL 35406, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Members of the DRE-TIM metallolyase superfamily rely on an active-site divalent cation to catalyze various
reactions involving the making and breaking of carbon–carbon bonds. While the identity of the metal varies,
the binding site is well-conserved at the superfamily level with an aspartic acid and two histidine residues acting
as ligands to the metal. Previous structural and bioinformatics results indicate that the metal can adopt an alter-
nate architecture through the addition of an asparagine residue as a fourth ligand. This asparagine residue is
strictly conserved in all members of the DRE-TIM metallolyase superfamily except fungal homocitrate synthase
(HCS-lys) where it is replaced with isoleucine. The role of this additional metal ligand in α-isopropylmalate
synthase from Mycobacterium tuberculosis (MtIPMS) has been investigated using site-directed mutagenesis.
Substitution of the asparagine ligand with alanine or isoleucine results in inactive enzymes with respect to α-
isopropylmalate formation. Control experiments suggest that the substitutions have not drastically affected the
enzyme's structure indicating that the asparagine residue is essential for catalysis. Interestingly, all enzyme var-
iants retained acetyl CoA hydrolysis activity in the absence of α-ketoisovalerate, similar to the wild-type enzyme.
In contrast to the requirement of magnesium for α-isopropylmalate formation, hydrolytic activity could be
inhibited by the addition of magnesium chloride in wild-type, D81E, and N321A MtIPMS, but not in the other var-
iants studied. Attempts to rescue loss of activity in N321I MtIPMS by mimicking the fungal HCS active site
through the D81E/N321I double variant were unsuccessful. This suggests epistatic constraints in evolution of
function in IPMS and HCS-lys enzymes.
Received 15 March 2014
Received in revised form 14 July 2014
Accepted 16 July 2014
Available online 23 July 2014
Keywords:
Isopropylmalate synthase
DRE-TIM metallolyase superfamily
Metalloenzyme
Enzyme catalysis
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
conserved at the subgroup level are predicted to provide unique func-
tionality to each subgroup such as differences in substrate selectivity
Over the past decade, the identification and characterization of
enzyme superfamilies have provided extraordinary insight into the evo-
lution of enzyme function [1–3]. A superfamily represents a group of
evolutionarily related enzymes that share a common mechanistic
aspect in the stabilization of an intermediate enabled by conserved ac-
tive site architecture [4]. Within a superfamily, subgroups of enzymes
can be identified that catalyze diverse sets of chemical reactions often
complicating functional predictions. Residues conserved throughout
the superfamily can be implicated as critical to a catalytic step in the
mechanism or required for structural stability. In contrast, residues
and regulatory mechanisms.
The DRE-TIM metallolyase superfamily contains enzymes involved
in making and breaking C\C bonds in various metabolic pathways
across all three domains of life [5]. This superfamily includes the en-
zymes isopropylmalate synthase (IPMS1), 3-hydroxy-3-methylgluta-
ryl-CoA lyase (HMG-CoA lyase), 2-hydroxy-4-ketovalerate aldolase,
and pyruvate carboxylase, with each activity represented by a unique
subgroup. Structurally these four enzymes share a TIM barrel catalytic
domain and require a divalent cation for activity. The divalent metal is
liganded by three well-conserved amino acids: an aspartic acid and
two histidine residues in an HXH motif (pink structure, Fig. 1A). In
most members, oxygen atoms from substrate and active site water
molecules contribute the remaining ligands to the metal.
IPMS is a well-characterized member of the DRE-TIM metallolyase
superfamily [6]. The enzyme catalyzes a Claisen-like condensation
between α-ketoisovalerate (KIV) and acetyl-CoA (AcCoA) followed by
hydrolysis of the isopropylmalyl-CoA intermediate resulting in
isopropylmalate (IPM) and CoA (Scheme 1). This is the first step in
the biosynthesis of L-leucine in archaea, bacteria, and some eukaryotes
Abbreviations: AcCoA, acetyl coenzyme A; CMS, citramalate synthase; HCS,
homocitrate synthase; HMG-CoA lyase, 3-Hydroxy-3-methylglutaryl-CoA lyase; IPMS,
α-Isopropylmalate synthase; KIV, α-Ketoisovalerate; MtIPMS, IPMS from Mycobacterium
tuberculosis; ScHCS, homocitrate synthase from Saccharomyces cerevisiae
☆
This work was supported by The University of Alabama and NSF grant MCB-1254077
(P.A.F.)
⁎
Corresponding author at: Department of Chemistry, The University of Alabama, Box
870336, Tuscaloosa, AL 35406, USA. Tel.: +1 205 348 8349; fax: +1 205 348 9104.
E-mail address: pfrantom@ua.edu (P.A. Frantom).
http://dx.doi.org/10.1016/j.bbapap.2014.07.013
1570-9639/© 2014 Elsevier B.V. All rights reserved.

P.A. Frantom et al. / Biochimica et Biophysica Acta 1844 (2014) 1784–1789
1785
Fig. 1. A) Superposition of metal ion binding locations in MtIPMS. PDB files 3u6w (pink) and 3hpx (brown) were superimposed using Matchmaker program in Chimera. The purple sphere
represents a Mn2+ ion from 3u6w, and the green sphere represents a Ni2+ ion from 3hpx. The additional metal ligand is labeled in red. B) Superposition of metal binding residues in the
DRE-TIM metallolyase superfamily. Structures shown include citramalate synthase from L. interrogans (3bli, brown), HMG-CoA lyase from H. sapiens (3mp3, pink), 4-OH-2-oxovalerate
aldolase from Psuedomonas sp. CF600 (1nmv, purple), and pyruvate carboxylase from Rhizobium etli (4jx5, cyan). Superposition was created using Chimera to align atoms contained in
the residues shown. C) Superposition of metal ion ligands in MtIPMS and HCS-lys from Thermus thermophilus (TtHCS). PDB files 3u6w (pink) and 2zyf (green) were used for the figure.
Here, 3u6w is identical to its representation in panel A. The green sphere in 2zyf represents a Mg2+ion. Residue numbers are shown for both MtIPMS and TtHCS. The site of differential
conservation is labeled in red.
[6]. In this enzyme, oxygen atoms from the carbonyl and carboxyl
groups of KIV serve as ligands to the divalent cation. Mechanistically,
the metal is predicted to help polarize the α-keto group to promote nu-
cleophilic attack by the methyl group from AcCoA. A recent structural
report for the catalytic domain of IPMS from Mycobacterium tuberculosis
(MtIPMS) exhibited alternate coordination architecture for the divalent
metal in the absence of substrate [7]. In the absence of KIV, the metal ion
is displaced several angstroms and liganded by an asparagine residue
(N321 in MtIPMS) in addition to the aspartic acid (D81) and histidine
residues (H285/H287) described above (brown structure, Fig. 1A).
Upon the addition of KIV to the crystal the metal transitions back to
the canonical coordination.
Recent bioinformatics analysis of the DRE-TIM metallolyase super-
family indicates that this alternate metal binding architecture is con-
served at the superfamily level, suggesting that it may play a role in
catalysis or structural stability [8]. With the exception of homocitrate
synthase sequences predicted to be involved in L-lysine biosynthesis
(HCS-lys), the asparagine (or glutamine in the case of pyruvate carbox-
ylase) residue acting as an additional ligand to the metal is found in a
structurally similar position (Fig. 1B) and is one of a small number of
residues strongly conserved throughout the DRE-TIM metallolyase su-
perfamily (Figure S1). An alternate metal coordination architecture sim-
ilar to that identified in MtIPMS has also been structurally documented
in DRE-TIM metallolyase superfamily member HMG-CoA lyase [9]. Con-
servation of this asparagine at the superfamily level suggests that it may
play an important role.
alternate metal binding site is present [8]. In contrast, HCS-lys enzymes
contain a differentially conserved isoleucine in place of the conserved as-
paragine residue. Isoleucine is not capable of coordinating the metal ion
suggesting that the alternate metal coordination site is not accessible to
these enzymes. The N → I substitution is accompanied by a glutamic
acid in place of the conserved aspartic acid, with the glutamic acid acting
as a ligand to the divalent cation (green structure, Fig. 1C). The lack of this
site in fungal HCS enzymes suggests that the ligand is not essential to the
catalytic function of the scaffold or may play a role in additional function-
ality not required by HCS-lys enzymes.
In order to investigate the catalytic necessity of the alternate metal
ion site, site-directed mutagenesis has been employed to create enzyme
variants at the conserved aspartic acid and asparagine positions. The
resulting enzyme variants have been characterized with respect to
KIV-dependent isopropylmalate formation, AcCoA hydrolysis activity,
and the ability of Mg2+ to inhibit hydrolysis. Overall, the results are con-
sistent with a requirement for the conserved asparagine residue in the
condensation activity of MtIPMS. Additionally, variants of MtIPMS that
mimic the active site of HCS-lys do not catalyze the condensation
reaction.
2. Materials and methods
2.1. Materials
Interestingly, fungal and euglena HCS enzymes in the α-aminoadipate
pathway for the biosynthesis of L-lysine are the exception within the
superfamily. HCS catalyzes an identical reaction to IPMS using α-
ketoglutarate instead of KIV as the α-keto acid substrate. This reaction is
the first step in the biosynthesis of L-lysine (HCS-lys) in fungal and
euglena organisms and in the production of homocitrate for the nitroge-
nase cofactor in nitrogen-fixing organisms (HCS-NifV). As identified by
the protein similarity network, sequences predicted to be HCS-NifV en-
zymes contain the conserved asparagine residue, suggesting that the
For the site-directed mutagenesis of the MtIPMS gene, oligonucleo-
tides from Eurofins MWG Operon were acquired (Huntsville, AL and
Louisville, KY). 4,4′-Dithiopyridine (DTP) was purchased from Acros Or-
ganics. Acetyl CoA (AcCoA) and α-ketoisolvalerate (α-KIV) were pur-
chased from Sigma-Aldrich. All other buffers and reagents were of the
highest purity grade available and were acquired from VWR Interna-
tional, LLC. The QuikChange Lightning Site-Directed Mutagenesis Kit
was obtained from Stratagene. XL Gold cells were purchased from
Stratagene, and competent Escherichia coli BL21(DE3)pLysS cells were
Scheme 1. Reaction catalyzed by IPMS..

1786
P.A. Frantom et al. / Biochimica et Biophysica Acta 1844 (2014) 1784–1789
purchased from Novagen. The Plasmid DNA Mini Kit I was acquired
from Omega. The HisTrap HP column was obtained from GE Healthcare.
2.6. Data analysis
Steady-state kinetic parameters for AcCoA hydrolysis reactions were
determined by a fit of the initial velocities at varying substrate concen-
trations to the Michaelis–Menten equation (Eq. 1). Here, v is the initial
velocity, Et is the total concentration of enzyme in the reaction, kcat is
the maximal velocity of the reaction, KM is the Michaelis constant for
the varied substrate, and [S] is the concentration of the varied substrate.
2.2. Construction of enzyme variant-containing plasmids
Point mutations were created in the pET28b::LeuA plasmid using
the QuikChange Lightning Site-Directed Mutagenesis Kit. The primers
used are listed in Table S1. The DNA products were sequenced by Oper-
on to confirm the existence of the desired mutation.
v
kcat ꢁ ½Sꢀ
:
¼
ð1Þ
½Etꢀ KM þ ½Sꢀ
2.3. Expression and purification of enzyme variants
IC50 values for metal inhibition of AcCoA hydrolysis were deter-
mined by fitting the initial velocities obtained at varying concentrations
of magnesium chloride to Eq. 2. Here, v is the initial velocity, v0 is the ini-
tial velocity in the absence of inhibitor, I is the concentration of metal
ion, IC50 is the concentration of inhibitor that gives 50% inhibition, and
nH is the Hill coefficient.
Enzymes were recombinantly expressed in autoinduction media as
previously described [10]. Briefly, the plasmids were transformed into E.
coli BL21(DE3)pLysS cells. The E. coli BL21(DE3)pLysS cells were inoculat-
ed into 25 mL of LB (30 μg/mL kanamycin, 35 μg/mL chloramphenicol,
0.5% glucose). The cells were then grown at 37 °C for 12–16 h. Approxi-
mately 12.5 mL of the overnight culture was added to 400 mL of
autoinduction media and grown at 25 °C for 24–36 h. Once the overnight
culture reached an OD600 of 10 to 15, the cells were isolated by centrifu-
gation (6340 g, 4 °C, 10 min). The cell pellet was stored at −80 °C. The
cell pellet was resuspended in lysis buffer (20 mM TEA (pH 7.8), 300
mM KCl, 50 mM imidazole). 1 mM of phenylmethanesulfonylfluoride
(PMSF) protease inhibitor, DNaseI, 0.25 mg/mL of lysozyme, and MgCl2
were added to the lysis buffer for a total volume of 50 mL. As previously
described, the cells were lysed by sonication for 10 min (2.5 min on, 2.5
min off, repeat four times) (Branson Sonifier). The supernatant was re-
trieved after centrifugation (34,540 g, 4 °C, 30 min) and loaded onto a
5 mL HisTrap HP column. The protein was eluted with a linear gradient
of 0–250 mM imidazole over 20 to 30 column volumes. An SDS-PAGE
gel was run to check the protein purity, and the fractions containing puri-
fied enzyme were pooled. The protein was then concentrated to a final
volume of approximately 10 mL using an Amicon ultrafiltration apparatus
(MWCO of 30 kDa). After treatment with 2 mM EDTA and 1 mM DTT
(4 °C, 30 min), the protein was dialyzed three times against 1 L of 20
mM TEA (pH 7.8) for four hours. The purified protein was stored in 10%
glycerol at −20 °C. Results from atomic absorption spectroscopy in
Chelex-treated 20 mM TEA (pH 7.8), 50 mM KCl buffer indicate that the
final enzyme preparation contains ~0.1 equivalents of magnesium.
Â
Ã
nH
v ¼ v0= 1 þ ð½Iꢀ=IC50
Þ
:
ð2Þ
Approximate molecular weights for the IPMS variants were deter-
mined from size exclusion chromatography results. The data was ana-
lyzed by plotting the log MW of each standard protein versus the
calculated Kav partition coefficient value for protein MW standards to
create a standard curve, which was then used to calculate molecular
weight values for the IPMS variants.
2.7. Multiple sequence alignment for representative DRE-TIM metallolyase
superfamily members
A structure-based sequence alignment was created using the
Needleman–Wunsch algorithm [11] as implemented with the Match-
maker [12] program in Chimera [13] using the HMGL-like catalytic do-
mains for four representative members of the DRE-TIM metallolyase
superfamily: MtIPMS (PDB ID: 3u6w [7]), pyruvate carboxylase from
Rhizobium etli (PDB ID: 4jx5 [14]), 4-hydroxy-2-ketovalerate aldolase
from Pseudomonas sp. CF600 (PDB ID: 1nvm [15]), and 3-hydroxy-3-
methylglutaryl CoA lyase from Homo sapiens (PDB ID: 3mp3 [16]).
This alignment was used as a seed to generate the full alignment of
representative sequences using MAFFT version 7 [17].
3. Results
2.4. Enzymatic assays
3.1. Kinetic parameters of AcCoA hydrolysis
Initial velocities for the kinetic assays were determined by fol-
lowing the formation of CoA using 4,4′-dithiopyridine (DTP) at
324 nm (ε = 19.8 mM−1 cm−1). A standard reaction mixture
contained 50 mM potassium phosphate buffer (pH 7.5), 12 mM
MgCl2, 10 μM DTP, and varying concentrations of AcCoA or KIV.
Each reaction was initiated by the addition of 10 to 20 nM of the en-
zyme. Hydrolysis activity was determined using similar conditions
in the absence of KIV. Metal-inhibition assays were performed
under similar conditions with a range of magnesium chloride con-
centrations (0–120 mM). The effect of EDTA on AcCoA hydrolysis
was determined by the addition of 1 mM EDTA to the reaction con-
ditions described above. All reactions were performed at 25 °C
maintained by a circulating water bath.
Overall, five enzyme variants were constructed: D81 was substituted
with alanine [8] and glutamic acid, N321 was substituted with alanine
and isoleucine, and the double substitution D81E/N321I was created
to mimic the active site of HCS-lys enzymes. All five enzymes were
successfully expressed and purified as described above. None of the
enzyme variants were capable of catalyzing KIV-dependent
isopropylmalate formation. Size-exclusion chromatography results
indicate native molecular weights of 152–178 kDa for the variant
enzymes in comparison to the native molecular weight of 171 kDa for
wild type MtIPMS (Table S2). The calculated molecular weight for
monomeric MtIPMS is approximately 72 kDa. These results suggest
that amino acid substitution has not disrupted the dimeric nature of
the enzyme. Circular dichroism spectroscopy was also used to compare
secondary structure content for the variants with the wild-type enzyme
(Figure S2). Spectra from the variants overlaid well with the spectrum
for wild-type enzyme suggesting that amino acid substitutions have
not significantly disrupted secondary structure elements. Initial
attempts to determine kinetic parameters for the variants using the
thiol-capture assay (which detects the production of CoA using
dithiodipyridine) were complicated when increases in KIV
2.5. Size exclusion chromatography
A GL Superdex 200 10/300 column was equilibrated with 20 mM
TEA (pH 7.8) and 100 mM KCl at a flow rate of 0.25 mL/min. The column
was calibrated using a molecular weight standard kit (BioRad). After
calibration, 100 μL of each enzyme (1 mg/mL concentration) was
injected onto the column.

P.A. Frantom et al. / Biochimica et Biophysica Acta 1844 (2014) 1784–1789
1787
concentration slowed the rate of reaction. Maximal rates of CoA forma-
tion were seen in the absence of KIV, suggesting KIV was not involved in
the formation of CoA and instead the variants catalyze the hydrolysis of
AcCoA. This activity has been previously reported to occur in wild-type
MtIPMS in the absence of KIV [18]. For wild-type MtIPMS, when KIV is
present no uncoupled hydrolysis of AcCoA is detected. The enzyme var-
iants display Michaelis–Menten kinetics (Figure S3) with respect to the
AcCoA-dependent hydrolysis reaction and the kinetic parameters deter-
mined for each enzyme are shown in Table 1. The KAcCoA values are ap-
proximately 5–10-fold larger compared for the KAcCoA value determined
in the KIV-dependent condensation reaction.
shown to be inactive, most likely due to the disruption of the metal
binding site [8]. Here, the D81E substitution retains a side chain capable
of acting as a ligand (as shown by HCS sequences), but IPMS activity is
not rescued. Substitutions at N321 also abolish IPMS activity. The
N321 side chain is ~3 Å from the side chain of D81 in both metal archi-
tectures of MtIPMS (and in other superfamily members). Thus, one ex-
planation for the results is that N321 may be responsible for proper
orientation of D81 in catalysis, however further structural studies are re-
quired to test this hypothesis.
4.2. AcCoA hydrolysis is not affected by the substitution of D81 or N321
All of the enzyme variants were capable of catalyzing the hydrolysis
of AcCoA in the absence of KIV. The catalytic machinery necessary for
hydrolysis in the isopropylmalate formation reaction or the AcCoA hy-
drolysis reaction has yet to be identified. Normal solvent kinetic isotope
effects are measured when pyruvate is used as an alternate substrate for
MtIPMS ruling out a metal-promoted water molecule in the IPMS reac-
tion, suggesting the involvement of side chains in the active site [18].
Consistent with this hypothesis, the AcCoA hydrolysis reaction cata-
lyzed by the MtIPMS variants was determined to be independent of ex-
ogenous magnesium and insensitive to the addition of EDTA.
Surprisingly, in wild-type, D81E, and N321A MtIPMS, AcCoA hydrolysis
activity is inhibited by the addition of magnesium chloride, with appar-
ent IC50 values of 5, 9, and 3 mM, respectively (Fig. 2). This is in stark
contrast to the absolute requirement for a divalent ion in the
isopropylmalate formation reaction [19]. The addition of magnesium
did not affect AcCoA hydrolysis in the remaining enzyme variants
(D81A, N231I, and D81E/N321I), suggesting that these substitutions
have altered the ability of the metal to interact with the enzyme.
One possible explanation for these results is that metal at the canon-
ical (Asp/Glu)/HXH site is responsible for the inhibition of AcCoA hydro-
lysis. While not a direct measurement of metal affinity, IC50 values
determined for metal ions in the AcCoA hydrolysis reaction with wild-
type and variant MtIPMS enzymes are identical to the Kact value (con-
centration of magnesium resulting in 50% activation) determined for
magnesium in the condensation reaction with wild-type MtIPMS. [19]
Thus, the susceptibility of AcCoA hydrolysis to metal inhibition in
D81E and N321A MtIPMS suggests that the site of metal-inhibition, pre-
sumably the primary (Asp/Glu)/HXH metal binding site, is unperturbed
in these enzyme variants.
3.2. Metal inhibition of enzyme catalyzed AcCoA hydrolysis
While attempting to identify conditions for maximal rates of AcCoA
hydrolysis, an observation was made that decreasing concentrations of
the divalent cation (essential for KIV-dependent condensation activity)
lead to higher rates of hydrolysis. A systematic approach was taken to
evaluate the effect of magnesium chloride concentration on the hydro-
lysis activity of the enzyme variants. As can be seen in Fig. 2, hydrolysis
activity of wild-type and two enzyme variants is inhibited by increasing
concentrations of magnesium chloride with IC50 values in the low milli-
molar range (Table 2). The lack of requirement for the metal ion in the
AcCoA hydrolysis reaction was confirmed by the addition of 1 mM
EDTA to an assay for wild-type MtIPMS; this had no effect on the rate
of the reaction. Three of the enzyme variants (D81A, N321I, and D81E/
N321I) were insensitive to increases in magnesium chloride concentra-
tion up to 50 mM. Proteolytic removal of the N-terminal (His)6-tag did
not affect the metal inhibition parameters for wild-type MtIPMS, ruling
out a contribution from this additional divalent metal binding site
(Figure S4).
4. Discussion
Analysis of enzyme superfamilies containing diverse functionalities
can allow for the identification of common catalytic strategies. In the
case of the DRE-TIM metallolyase superfamily, members are proposed
to stabilize a common enolate intermediate using a conserved active
site arginine residue. All members of the superfamily also require a di-
valent cation for full activity; however, the exact role of the metal ion
in catalysis is not clear for this superfamily. The recent structural and
bioinformatics results indicate the presence of an alternate metal archi-
tecture using a conserved asparagine residue to act as an additional li-
gand to the metal ion.
4.3. MtIPMS cannot use HCS-like active site architecture
If the evolutionarily conserved asparagine is critical for catalysis, it is
of interest to understand why HCS-lys enzymes can catalyze a similar
reaction while lacking the alternate metal binding site. As described
above, fungal HCS enzymes have a pair of conserved substitutions
with a D → E substitution for the metal binding carboxylate. In order
to accommodate the extra methylene group from glutamic acid, HCS-
lys enzymes have a conserved compensatory substitution of the smaller
isoleucine at the N321 position. The doubly substituted D81E/N321I
MtIPMS is unable to support isopropylmalate formation activity, and
MgCl2 does not inhibit the enzyme's hydrolysis activity. One explana-
tion for this result is that additional changes to the active site are re-
quired for MtIPMS to be active in the absence of the alternate metal
binding site. However, closer inspection of the active site architecture
for the two enzymes shows no readily apparent changes that could be
proposed.
4.1. D81 and N321 are critical to MtIPMS function
MtIPMS variants with substitutions at the asparagine and aspartic
acid positions are inactive with respect to the KIV-dependent
isopropylmalate formation reaction suggesting that these residues are
important in catalysis or for structural integrity. Combined results
from size-exclusion chromatography and circular dichroism spectrosco-
py suggest that the substitutions have not grossly affected the structure
of the enzyme, ruling out a large-scale structural perturbation in the in-
activity of the variants. The D81A substitution has previously been
Table 1
Kinetic parameters for AcCoA hydrolysis for wild-type MtIPMS and variants.
An alternate explanation rooted in the superfamily-based evolution
is that HCS activity in the absence of the alternate metal binding site is
due to different functional properties of the two enzymes. Indeed bio-
chemical characterization of MtIPMS and HCS from Saccharomyces
cerevisiae (ScHCS) indicates several functional differences. Most striking
is the difference in regulatory properties between the two enzymes.
MtIPMS, which is feedback inhibited by L-leucine, contains a C-
MtIPMS variant
kcat (min−1
)
KAcCoA (μM)
WT
D81A
D81E
N321A
N321I
D81E/N321I
19
3
2
1
1
1
1
1
253
97
233
123
262
163
48
13
40
4
7
9
9
55
28
8

1788
P.A. Frantom et al. / Biochimica et Biophysica Acta 1844 (2014) 1784–1789
Fig. 2. Inhibition of AcCoA hydrolysis by MgCl2 for wildtype and variant MtIPMS. Experimental conditions are as described in Materials and methods. The solid line is from a fit of the data to
Eq. 2.
terminal regulatory domain and is inhibited allosterically. In contrast
ScHCS, which is feedback inhibited by L-lysine, is subject to competitive
inhibition [20]. Mutational studies also suggest differences in the identi-
ty of conserved residues responsible for acid–base chemistry. Alanine
substitutions for active site residues H379 and E218 (MtIPMS number-
ing) cause either loss of or significant decreases in HCS activity [21]
while identical substitutions in MtIPMS result in only 4-to-30-fold de-
creases in activity [22].
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.bbapap.2014.07.013.
References
[1] J.A. Gerlt, P.C. Babbitt, M.P. Jacobson, S.C. Almo, Divergent evolution in enolase su-
perfamily: strategies for assigning functions, J. Biol. Chem. 287 (2012) 29–34.
[2] K.N. Allen, D. Dunaway-Mariano, Markers of fitness in a successful enzyme super-
family, Curr. Opin. Struct. Biol. 19 (2009) 658–665.
[3] M.A. Hicks, A.E. Barber 2nd, L.A. Giddings, J. Caldwell, S.E. O'Connor, P.C. Babbitt, The
evolution of function in strictosidine synthase-like proteins, Proteins 79 (2011)
3082–3098.
[4] J.A. Gerlt, P.C. Babbitt, Divergent evolution of enzymatic function: mechanistically
diverse superfamilies and functionally distinct suprafamilies, Annu. Rev. Biochem.
70 (2001) 209–246.
[5] F. Forouhar, M. Hussain, R. Farid, J. Benach, M. Abashidze, W.C. Edstrom, S.M. Vorobiev,
R. Xiao, T.B. Acton, Z. Fu, J.J. Kim, H.M. Miziorko, G.T. Montelione, J.F. Hunt, Crystal struc-
tures of two bacterial 3-hydroxy-3-methylglutaryl-CoA lyases suggest a common cata-
lytic mechanism among a family of TIM barrel metalloenzymes cleaving carbon–carbon
bonds, J. Biol. Chem. 281 (2006) 7533–7545.
[6] P.A. Frantom, Structural and functional characterization of alpha-isopropylmalate
synthase and citramalate synthase, members of the LeuA dimer superfamily, Arch.
Biochem. Biophys. 519 (2012) 202–209.
[7] F.H. Huisman, N. Koon, E.M. Bulloch, H.M. Baker, E.N. Baker, C.J. Squire, E.J. Parker,
Removal of the C-terminal regulatory domain of alpha-isopropylmalate
synthase disrupts functional substrate binding, Biochemistry 51 (2012)
2289–2297.
5. Conclusions
Overall, the results are consistent with the asparagine residue respon-
sible for the alternate metal architecture in MtIPMS being required for full
catalytic activity. One possible explanation for this result is that N321 is
responsible for properly orienting D81 in the active site, though future
structural studies are required to test this hypothesis. The ability of
N321A MtIPMS to catalyze AcCoA hydrolysis with kinetic parameters
similar to the wild-type enzyme suggests that the adjacent site is not in-
volved in the hydrolytic step following the condensation reaction. Finally,
due to the superfamily level of conservation, it will be of interest to deter-
mine if this site is critical for activity in other superfamily members.
Acknowledgements
[8] A.K. Casey, H., M. A., J.L. Johnson, P.C. Babbitt, P.A. Frantom, Mechanistic and
bioinformatic investigation of
a conserved active site helix in alpha-
isopropylmalate synthase form Mycobacterium tuberculosis, a member of the
DRE-TIM metallolyase superfamily, Biochemistry 53 (2014) 2915–2925.
[9] Z. Fu, J.A. Runquist, F. Forouhar, M. Hussain, J.F. Hunt, H.M. Miziorko, J.J. Kim, Crystal
structure of human 3-hydroxy-3-methylglutaryl-CoA Lyase: insights into catalysis
and the molecular basis for hydroxymethylglutaric aciduria, J. Biol. Chem. 281
(2006) 7526–7532.
[10] A.K. Casey, J. Baugh, P.A. Frantom, The slow-onset nature of allosteric inhibition in
alpha-isopropylmalate synthase from Mycobacterium tuberculosis is mediated by
a flexible loop, Biochemistry 51 (2012) 4773–4775.
This work was supported by NSF grant MCB-1254077 (P.A.F.).
Molecular graphics and analyses were performed with the UCSF Chimera
package. Chimera is developed by the Resource for Biocomputing,
Visualization, and Informatics at the University of California, San Francisco
(supported by NIGMS P41-GM103311). We thank Dr. Holly Ellis (Auburn
University) for her assistance in circular dichroism spectroscopy.
[11] S.B. Needleman, C.D. Wunsch, A general method applicable to the search for
similarities in the amino acid sequence of two proteins, J. Mol. Biol. 48
(1970) 443–453.
[12] E.C. Meng, E.F. Pettersen, G.S. Couch, C.C. Huang, T.E. Ferrin, Tools for integrat-
ed sequence-structure analysis with UCSF Chimera, BMC Bioinforma. 7 (2006)
339.
Table 2
Apparent inhibition constants determined for the inhibition
of AcCoA hydrolysis by divalent metal for wild-type MtIPMS
and variantsa.
[13] Z. Yang, K. Lasker, D. Schneidman-Duhovny, B. Webb, C.C. Huang, E.F. Pettersen, T.D.
Goddard, E.C. Meng, A. Sali, T.E. Ferrin, UCSF Chimera, MODELLER, and IMP: an
integrated modeling system, J. Struct. Biol. 179 (2012) 269–278.
[14] A.D. Lietzan, M. St Maurice, Insights into the carboxyltransferase reaction of pyru-
vate carboxylase from the structures of bound product and intermediate analogs,
Biochem. Biophys. Res. Commun. 441 (2013) 377–382.
[15] B.A. Manjasetty, J. Powlowski, A. Vrielink, Crystal structure of a bifunctional aldol-
ase–dehydrogenase: sequestering a reactive and volatile intermediate, Proc. Natl.
Acad. Sci. U. S. A. 100 (2003) 6992–6997.
[16] Z. Fu, J.A. Runquist, C. Montgomery, H.M. Miziorko, J.J. Kim, Functional insights into
human HMG-CoA lyase from structures of Acyl-CoA-containing ternary complexes,
J. Biol. Chem. 285 (2010) 26341–26349.
MtIPMS variant
IC50 (mM)
WT
D81A
D81E
N321A
N321I
D81E/N321I
5
–
9
3
–
–
1
b
1
1
a
Experimental conditions described in Materials and
methods.
b
No inhibition observed up to 50 mM.

P.A. Frantom et al. / Biochimica et Biophysica Acta 1844 (2014) 1784–1789
1789
[17] K. Katoh, D.M. Standley, MAFFT multiple sequence alignment software ver-
sion 7: improvements in performance and usability, Mol. Biol. Evol. 30
(2013) 772–780.
[20] B. Andi, A.H. West, P.F. Cook, Regulatory mechanism of histidine-tagged homocitrate
synthase from Saccharomyces cerevisiae. I. Kinetic studies, J. Biol. Chem. 280 (2005)
31624–31632.
[18] L.P. de Carvalho, J.S. Blanchard, Kinetic and chemical mechanism of alpha-
isopropylmalate synthase from Mycobacterium tuberculosis, Biochemistry 45 (2006)
8988–8999.
[21] J. Qian, J. Khandogin, A.H. West, P.F. Cook, Evidence for a catalytic dyad in the active site
of homocitrate synthase from Saccharomyces cerevisiae, Biochemistry 47 (2008)
6851–6858.
[19] L.P. de Carvalho, J.S. Blanchard, Kinetic analysis of the effects of monovalent cations and
divalent metals on the activity of Mycobacterium tuberculosis alpha-isopropylmalate
synthase, Arch. Biochem. Biophys. 451 (2006) 141–148.
[22] L.P. de Carvalho, P.A. Frantom, A. Argyrou, J.S. Blanchard, Kinetic evidence for
interdomain communication in the allosteric regulation of alpha-isopropylmalate syn-
thase from Mycobacterium tuberculosis, Biochemistry 48 (2009) 1996–2004.