Nueleophile specificity in anthranilate synthase, aminodeoxychorismate synthase, isochorismate synthase, and salicylate synthase

Article abstract of DOI:10.1021/bi100021x

Anthranilate synthase (AS), aminodeoxychorismate synthase (ADCS), isochorismate synthase (IS), and salicylate synthase (SS) are structurally homologous chorismate-utilizing enzymes that carry out the first committed step in the formation of tryptophan, fo

Full text of DOI:10.1021/bi100021x

Biochemistry 2010, 49, 2851–2859 2851
DOI: 10.1021/bi100021x
Nucleophile Specificity in Anthranilate Synthase, Aminodeoxychorismate Synthase,
Isochorismate Synthase, and Salicylate Synthase
Kristin T. Ziebart and Michael D. Toney*
Department of Chemistry, University of California, Davis, California 95616
Received January 7, 2010; Revised Manuscript Received February 15, 2010
ABSTRACT: Anthranilate synthase (AS), aminodeoxychorismate synthase (ADCS), isochorismate synthase
(IS), and salicylate synthase (SS) are structurally homologous chorismate-utilizing enzymes that carry out the
first committed step in the formation of tryptophan, folate, and the siderophores enterobactin and
mycobactin, respectively. Each enzyme catalyzes a nucleophilic substitution reaction, but IS and SS are
uniquely able to employ water as a nucleophile. Lys147 has been proposed to be the catalytic base that
activates water for nucleophilic attack in IS and SS reactions; in AS and ADCS, glutamine occupies the
analogous position. To probe the role of Lys147 as a catalytic base, the K147Q IS, K147Q SS, Q147K AS, and
Q147K ADCS mutants were prepared and enzyme reactions were analyzed by high-performance liquid
chromatography. Q147K AS employs water as a nucleophile to a small extent, and the cognate activities of
K147Q IS and K147Q SS were reduced ∼25- and ∼50-fold, respectively. Therefore, Lys147 is not solely
responsible for activation of water as a nucleophile. Additional factors that contribute to water activation are
proposed. A change in substrate preference for K147Q SS pyruvate lyase activity indicates Lys147 partially
controls SS reaction specificity. Finally, we demonstrate that AS, ADCS, IS, and SS do not possess choris-
mate mutase promiscuous activity, contrary to several previous reports.
The shikimate pathway occurs in bacteria, plants, fungi, and
apicomplexan parasites (1, 2). Starting with phosphoenol-
pyruvate and D-erythrose 4-phosphate and ending with choris-
In this mechanism, IS and SS employ water as a nucleophile, yet
AS and ADCS do not, despite its abundance (Figure 2). Instead,
AS and ADCS utilize ammonia that is provided from companion
glutamine amidotransferases TrpG and PabA, respectively.
ADCS is unique in its use of an active site lysine (K213) as the
initial nucleophile attacking at C2; in a second SN200 reaction,
nucleophilic attack by ammonia at C4 of the covalent inter-
mediate leads to the final product (14-16). ADCS and IS release
the nucleophile-substituted chorismate as the product, while AS
and SS proceed further by eliminating pyruvate and releasing the
resulting aromatic product.
In AS and ADCS, the companion amidotransferase subunit
delivers ammonia directly to the active site where chorismate
binds and undergoes nucleophilic substitution. An ammonia
tunnel is believed to exist between the amidotransferase and
chorismate-aminating active sites; this is a common mechanism
for avoiding free intracellular ammonia (17, 18). In the absence of
mate, it enables de novo biosynthesis of carbocyclic aromatic
compounds. Chorismate is the common precursor of a variety
of essential metabolites such as phenylalanine, tyrosine, trypto-
phan, folate, siderophores, menaquinones, and others. Produc-
tion of several phenazines, known to impart virulence and
competitive fitness to the opportunistic pathogen Pseudomonas
aeruginosa, also begins with chorismate (3). The central role of
chorismate-utilizing enzymes in the survival of these organisms
and their absence in humans make them attractive targets for
herbicides and antimicrobial drugs.
Figure 1 depicts seven branch points in chorismate metabo-
lism. With the exception of chorismate lyase (CL)1 and choris-
mate mutase (CM), these enzymes are structurally homologous.
X-ray crystal structures of anthranilate synthase (AS), amino-
deoxychorismate synthase (ADCS), isochorismate synthase (IS),
and salicylate synthase (SS) reveal them to be highly similar
proteins with conserved active sites (4-11). These similarities,
along with an abundance of biochemical evidence (12, 13), led He
et al. to propose a common mechanism in which catalysis begins
with nucleophilic attack at C2 of chorismate, concomitant with
displacement of the C4 hydroxyl group in an SN200 reaction (14).
L-glutamine and an amidotransferase subunit, AS and ADCS can
employ (NH4)2SO4 as an ammonia source (14, 19). Although IS
and SS lack a source of ammonia in vivo, each will accept it as a
nucleophile in vitro to form aminodeoxyisochorismate (ADIC)
and anthranilate, respectively (vide infra and ref 16). Since
intracellular ammonia concentrations are generally low, IS and
SS apparently have not needed to develop a means of discrimi-
nating against this nucleophile. However, since water is a poor
nucleophile and is not employed by AS and ADCS, water
activation must exclusively exist in IS and SS to explain the
observed nucleophile specificity among these enzymes.
Sequence alignments indicate a conserved active site lysine
residue exists in all IS and SS sequences (9). A glutamine residue
occupies the analogous position in all ADCS and AS sequences.
Recent crystal structures of MenF, a menaquinone-specific IS,
and Irp9, the SS from Yersinia enterocolitica, reveal this lysine to
*To whom correspondence should be addressed. E-mail: mdtoney@
ucdavis.edu. Telephone: (530) 754-5282. Fax: (530) 752-8995.
1Abbreviations: CL, chorismate lyase; CM, chorismate mutase; AS,
anthranilate synthase; ADCS, 4-amino-4-deoxychorismate synthase;
IS, isochorismate synthase; SS, salicylate synthase; PD, prephenate
dehydratase; ADCL, 4-amino-4-deoxychorismate lyase; MenF, mena-
quinone-specific IS; Irp9, SS from Yersinia enterocolitica; MbtI, SS from
Mycobacterium tuberculosis; ADIC, 2-amino-4-deoxyisochorismate;
ADC, 4-amino-4-deoxychorismate; PHB, p-hydroxybenzoate; HPLC,
high-performance liquid chromatography.
r
2010 American Chemical Society
Published on Web 02/19/2010
pubs.acs.org/Biochemistry

2852 Biochemistry, Vol. 49, No. 13, 2010
Ziebart and Toney
F
IGURE 1: Chorismate is a central branch point compound in plant, bacterial, fungal, and apicomplexan parasite metabolism. With the exception
of chorismate mutase and chorismate lyase, the enzymes shown are structurally similar. Each enzyme catalyzes the first committed step toward
biosynthesis of a carbocyclic aromatic metabolite essential for cell survival and/or virulence.
F
IGURE 2: Chorismate undergoes nucleophilic attack at C2 by ammonia (AS), Lys213 (ADCS), or water (IS and SS). SS and IS employ NH3 as a
nucleophile if supplied exogenously, but AS and ADCS are unable to activate water as a nucleophile. K213A ADCS activity is rescued by
(NH4)2SO4.
be well-positioned between an ordered water molecule and C2
of chorismate (4, 11). On the basis of the available structural
evidence and the fact that a K147A IS mutant was inactive,
Kolappan et al. proposed Lys147 to be the general base catalyst
that activates water for nucleophilic attack in IS and SS reac-
tions (9).

Article
Biochemistry, Vol. 49, No. 13, 2010 2853
FIGURE 3: Abell and co-workers’ proposed mechanism for promiscuous formation of phenylpyruvate by AS and SS (19). The diaxial
conformation of chorismate undergoes pericyclic rearrangement to form prephenate, followed by decarboxylation and general acid-assisted
loss of water. Glu197 is conserved in all AS and SS sequences. Data presented here demonstrate that these reactions are not catalyzed by AS
and SS.
Seeking to provide additional biochemical insight into the
catalytic role of Lys147, we mutated it here to a glutamine in MbtI
(the SS from Mycobacterium tuberculosis) and EntC (the enter-
obactin-specific IS); the corresponding glutamine to lysine mu-
tants of the ammonia-utilizing AS and ADCS were also prepared.
Kerbarh et al. performed similar experiments with AS and Irp9
SS (20). In that study, enzyme-catalyzed reactions were conducted
K274A/Q211K mutagenic primers were 50-CGGTGATTGC-
TATAAGGTGAATCTCGCC (forward) and 50-GGCGAGA-
TTCACCTTATAGCAATCACCG (reverse). K274A/Q211K
ADCS was expressed and purified in a manner similar to that
of wild-type (WT) ADCS. The protein yield was 190 mg/23 g of
cell paste.
For K147Q IS, the mutation was created in the pET28a-entC
construct using the following mutagenic primers: 50-GCCG-
CAGGTCGACCAAGTGGTGTTG (forward) and 50-CAA-
CACCACTTGGTCGACCTGCGGC (reverse). K147Q IS was
overexpressed and purified in a manner similar to that of WT IS.
The protein yield was 182 mg/26 g of cell paste.
1
at pH 7 and product mixtures were analyzed by H NMR; a
reciprocal switch in nucleophile specificity was not observed. The
major result of Kerbarh et al. was that AS and Irp9 mutants
apparently catalyze formation of phenylpyruvate. As illustrated
in Figure 3, Kerbarh et al. proposed that SS, K147Q SS, and
Q147K AS catalyze two successive promiscuous reactions: peri-
cyclic rearrangement of chorismate to form prephenate, followed
by decarboxylation and loss of water to form phenylpyruvate.
This work substantially builds upon previous results and add-
resses more thoroughly the issue of nucleophile specificity in the
following ways. (1) Four enzymes, AS, ADCS, IS, and SS, are
under study instead of just two. (2) Reaction mixtures are analyzed
by HPLC, which affords a detection limit of reaction products lower
Open reading frame Rv2386c of M. tuberculosis genomic DNA
encodes MbtI, a protein identified as a salicylate synthase (10).
The E. coli codon-optimized synthetic gene (procured from
Mr. Gene) was subcloned into a pET28a expression vector
(Novagen) using NdeI and BamHI restriction sites. After E. coli
BL21(DE3) cells harboring a pET28a-mbtI plasmid had been
grown in LB medium at 37 °C to OD600 of 2, SS overexpression
proceeded for 7 h at 30 °C after induction with 0.5 mM IPTG.
Cells were harvested by centrifugation and resuspended in lysis
buffer [20 mM Na2HPO4, 500 mM NaCl, 10 mM imidazole
(pH 7.4), 0.5 mg/mL lysozyme, and 0.2 unit/mL DNase I] prior to
disruption by sonication. Cell debris was pelleted by centrifuga-
tion at 14000 rpm for 45 min. The supernatant was incubated at
4 °C for 45 min with Chelating Sepharose Fast Flow resin
(Pharmacia) that had been charged with Ni2þ. The resin was
loaded into a column and washed with 10 column volumes of
starting buffer [20 mM Na2HPO4, 500 mM NaCl, and 10 mM
imidazole (pH 7.4)]. Protein was eluted with a linear gradient of
500 mL from 10 to 300 mM imidazole in starting buffer. The
purest fractions, as judged by SDS-PAGE analysis, were con-
centrated by ultrafiltration and dialyzed against 20 mM KPi
(pH 7.5), 50 mM KCl, and 1 mM DTT. Purified enzyme was
flash-frozen and stored at -80 °C. Protein concentrations were
measured with the Lowry protein assay kit (Bio-Rad) using IgG
as a standard. The protein yield was 25 mg/33 g of cell paste.
For K205Q SS, the mutation was created in the pET28a-mbtI
construct using the following mutagenic primers: 50-CTGGT-
CGTTACCACCAAGTTATTCTGTCCC (forward) and 50-GG-
GACAGAATAACTTGGTGGTAACGACCAG (reverse).
K205Q SS was overexpressed and purified in a manner similar
to that of WT SS (note that K205 is analogous to K147 in the
EntC IS sequence). The protein yield was 40 mg/31 g of cell paste.
A small portion of each His-tagged protein that had been
purified by nickel affinity chromatography was subjected to a
second purification by anion exchange. Protein, which had been
stored in 20 mM KPi (pH 7.5), 50 mM KCl, and 1 mM DTT at
1
than that for H NMR. (3) (NH4)2SO4-dependent reactions are
conducted at pH 8 instead of pH 7, thus increasing the concentra-
tion of NH3 10-fold. (4) The source of promiscuous chorismate
mutase and prephenate dehydratase activities is explained.
EXPERIMENTAL PROCEDURES
Materials. All chemical reagents were purchased from Sigma-
Aldrich and used without further purification. Lysozyme was
purchased from Sigma. Glutamate dehydrogenase (GDH) was
purchased from Calzyme. Lactate dehydrogenase (LDH) and
DNaseI were purchased from Roche. Chorismate was prepared
according to a literature procedure (21).
Enzyme Preparation. The preparation of plasmid constructs
bearing the genes pabA, pabB, and entC was described previously
(14, 16). Overexpression and purification of the proteins encoded
by these three genes, PabA, ADCS, and IS, respectively, were
conducted according to literature procedures (14, 16). Protein
yields were as follows: 83 mg of PabA/15 g of cell paste, 330 mg of
ADCS/20 g of cell paste, and 213 mg of IS/42 g of cell paste.
Mutations were introduced using Stratagene’s QuikChange
site-directed mutagenesis kit. Construction of the ADCS double
mutant (K274A/Q211K) used pET28a-pabB-K274A as the tem-
plate strand (note that Q211 is the residue analogous to K147,
and K274 is analogous to A213 in the EntC IS sequence).2 The
preparation of K274A ADCS was described previously (14). The
2The numbering used throughout is based on the EntC protein from
Escherichia coli.

2854 Biochemistry, Vol. 49, No. 13, 2010
Ziebart and Toney
-80 °C, was thawed and loaded onto a 50 mL Q-Sepharose Fast
Flow (Pharmacia) column that had been equilibrated with start
buffer [10 mM triethanolamine hydrochloride (TEA-HCl)
(pH 7.8), 10 mM mercaptoethanol, 5 mM MgCl2, and 1 mM
EDTA]. Protein was eluted with a linear gradient of 500 mL from
0 to 300 mM KCl in start buffer. The purest fractions, as judged
by SDS-PAGE, were concentrated by ultrafiltration and dia-
lyzed against 20 mM KPi (pH 7.5), 50 mM KCl, and 1 mM DTT.
The purified enzyme was flash-frozen and stored at -80 °C. The
activity of each doubly purified protein was the same as that
measured for the singly purified protein.
The plasmid construct bearing the genes for the partial com-
plex of anthranilate synthase (TrpE2-TrpG2) was a gift from
R. Bauerle (University of Virginia, Charlottesville, VA). Partial
complex anthranilate synthase (AS) from Salmonella typhimur-
ium consists of a TrpE2-TrpG2 heterotetramer. Plasmid pSTC25
contains two tandem stop codons engineered into the trpD gene,
such that only the region encoding amidotransferase activity is
expressed; the polypeptide translated from this truncated gene is
named TrpG (19).
AS was prepared in the following manner. E. coli cells (strain
CB694) harboring the pSTC25 construct were grown in LB
medium at 37 °C to an OD600 of 2.5-3. Cells were harvested by
centrifugation, resuspended in 10 mM TEA (pH 7.8), 10 mM
mercaptoethanol, 5 mM MgCl2, 1 mM EDTA, 0.5 mg/mL
lysozyme, and 0.2 unit/mL DNaseI, and then disrupted by soni-
cation. Cell debris was pelleted by centrifugation at 14000 rpm.
Ammonium sulfate was added to 23% saturation to the soluble
extract; the resulting precipitate was resuspended in a minimum
volume of start buffer [10 mM TEA (pH 7.8), 10 mM mercap-
toethanol, 5 mM MgCl2, and 1 mM EDTA] and loaded onto a
50 mL Q-Sepharose Fast Flow column. Protein was eluted with a
linear gradient of 500 mL from 0 to 300 mM KCl in start buffer.
The purest fractions, as judged by SDS-PAGE, were con-
centrated by ultrafiltration and dialyzed against 20 mM KPi
(pH 7.5), 50 mM KCl, and 1 mM DTT. The purified enzyme was
flash-frozen and stored at -80 °C. The protein concentration was
measured with the Lowry protein assay kit (Bio-Rad), using IgG
as a standard. The yield was 142 mg of purified protein/10 g of
cell paste.
Introduction of the Q263K mutation into the trpE gene within
pSTC25 used the following mutagenic primers: 50-GGCGAGA-
TATTTAAGGTGGTGCCGTC (forward) and 50-GACGG-
CACCACCTTAAATATCTCGCC (reverse) (note that Q263
is the residue analogous to K147 in EntC IS). Q263K AS was
overexpressed and purified in a manner similar to that of WT AS.
The protein yield was 173 mg/22 g of cell paste.
Lacking a His tag, AS and Q263K AS could not be purified by
nickel affinity chromatography. Therefore, these were repurified
via hydroxyapatite chromatography. Protein, which had been
stored in 20 mM KPi (pH 7.5), 50 mM KCl, and 1 mM DTT at
-80 °C, was thawed and loaded onto a column containing CHT
Ceramic Hydroxyapatite resin (Bio-Rad) that had been equili-
brated with start buffer [5 mM sodium phosphate and 150 mM
NaCl (pH 6.8)]. Protein was eluted with a linear gradient of
500 mL from 5 to 500 mM sodium phosphate in start buffer. The
purest fractions, as judged by SDS-PAGE, were combined and
concentrated by ultrafiltration and dialyzed against 20 mM KPi
(pH 7.5), 50 mM KCl, and 1 mM DTT. Purified enzyme was
flash-frozen and stored at -80 °C. The activity of the doubly
purified protein was the same as that measured for the singly
purified protein.
Spectrofluorometric AS Activity Assays. These were
performed on a PerkinElmer LS 50B luminescence spectro-
photometer at 25 °C. The enzyme activity assay for AS was based
on a previously described method (19). The assay directly
monitors anthranilate emission at 390 nm using excitation at
313 nm. Reactions were initiated by addition of anthranilate syn-
thase. The standard glutamine-dependent assay contains 100 mM
phosphate buffer (pH 7.0), 20 mM L-glutamine, 5 mM MgCl2,
and 2.2 nM WT AS. The standard NH4þ-dependent assay
contains 100 mM TEA-HCl (pH 8.0), 100 mM (NH4)2SO4,
5 mM MgCl2, and 2.2 nM WT AS. The same assays were used to
monitor Q147K AS-catalyzed reactions. The protein concentra-
tion of Q147K AS was increased 500 nM to compensate for the
decreased activity of the mutant.
Spectrophotometric Activity Assays. Kinetic assays were
performed on a Kontron Uvikon 930 UV-vis spectrophoto-
meter at 25 °C. Amidotransferase activity was measured in a
coupled assay with excess glutamate dehydrogenase (GDH) that
was based on a previously described method (22). PabA gluta-
minase activity was measured while it was in complex with either
WT ADCS or K213A/Q147K ADCS. Each 500 μL reaction
mixture contained 100 mM TEA-HCl (pH 9.0), 5 mM MgCl2,
4 mM 3-acetylpyridine adenine dinucleotide (APAD), 20 units of
GDH, 0.5 μM PabA, and 0.75 μM ADCS (same concentration
used for both WT and K213A/Q147K reactions). The
L-gluta-
mine concentration was varied from 0.1 to 4.0 mM.
The activity of TrpG, measured while in a tetrameric complex
with either WT TrpE or Q147K TrpE, was measured by the same
assay. Each 500 μL reaction mixture contained 100 mM TEA-
HCl (pH 9.0), 5 mM MgCl2, 4 mM 3-acetylpyridine adenine
dinucleotide (APAD), 25 units of GDH, and 1.0 nM AS (same
concentration used for both WT and Q147K reactions). The
L-glutamine concentration was varied from 0.5 to 8.0 mM.
HPLC Analysis. The standard reaction mixture for HPLC
analysis contained 100 mM bicine (pH 8.0) [AS-Gln reactions
used 100 mM phosphate buffer (pH 7.0)], 5 mM chorismate,
either 20 mM L-glutamine or 100 mM (NH4)2SO4, 5 mM MgCl2
(IS and SS reaction mixtures contained 1 mM MgCl2), and
20 μM enzyme(s).
After 3 h at room temperature, the samples were ultrafiltered
to remove protein(s) and injected onto an Agilent 1100 HPLC
system. Conditions for control reactions (no enzyme) were
identical. A 5 μm Supelco Supelcosil (4.6 mm ꢀ 250 mm) C18
column was used to separate reaction mixtures. Mobile phase A
consisted of a 5% acetic acid/water mixture; mobile phase B
consisted of acetonitrile. Reaction mixtures were eluted isocrati-
cally at 95% A and 5% B for 15 min, followed by gradient elution
in the following manner: 5 to 50% B from 15 to 18 min, 50% B
from 18 to 25 min, and 50 to 5% B from 25 to 30 min. Two
detectors were used. The first monitored absorbance at 280 nm.
The second monitored fluorescence emission at 400 nm from
excitation at 300 nm. The flow rate was 1.25 mL/min.
RESULTS
HPLC Analysis of Wild-Type and Mutant Enzyme-Cata-
lyzed Reactions. HPLC data are summarized in Tables 1-4.
For each reaction condition, a nonenzymatic control reaction
was performed. Peak area integration of each signal was correc-
ted for the control sample. Adjusted peak areas were then conver-
ted to a concentration value by correlation with measured res-
ponse factors. HPLC peak assignments were based on a comparison
of retention times with those of authentic standards. Additional

Article
Biochemistry, Vol. 49, No. 13, 2010 2855
Table 1: IS Product Distributionsa
Table 3: AS Product Distributionsa
-Gln
þ
þ
NH4
H2O
WT enzyme
L
NH4
H2O
WT enzyme
K147Q
enzyme
K147Q
enzyme
WT
Q147K
WT
Q147K
WT
Q147K
enzyme enzyme enzyme enzyme enzyme enzyme
chorismate
conversion (%)
isochorismate
ADIC
59
48
38
3
chorismate
conversion (%)
isochorismate
ADIC
100
0.3
8
100
30
0.6
0
0.4
21
79
41
59
4
100
100
96
0.4
2
salicylate
16
76
aReaction mixtures were incubated for 3 h at 25 °C and pH 8.0 prior to
separation by RP-HPLC. Each product is reported as a fraction of all pro-
ducts formed. Chorismate conversion is the percentage of 5 mM chorismate
enzymatically converted to products. All values were corrected for nonen-
zymatic controls run under identical conditions.
anthranilate
100
100
97
aReaction mixtures were incubated for 3 h at 25 °C and pH 8.0 prior to
separation by RP-HPLC; -Gln reactions were conducted at pH 7.0. Each
product is reported as a fraction of all products formed. Chorismate
conversion is the percentage of 5 mM chorismate enzymatically converted
to products. All values were corrected for nonenzymatic controls run under
identical conditions.
L
Table 2: SS Product Distributionsa
þ
NH4
WT enzyme
H2O
WT enzyme
indicate that formation of these two products by K147Q SS
occurs ∼50-fold slower relative to WT SS.
K147Q
enzyme
K147Q
enzyme
The presence of peaks due to ADIC and anthranilate in
Figure 5 indicates that both WT SS and K147Q SS employ
ammonia as a nucleophile. In previous work with SS, Kerbarh
et al. did not observe anthranilate in SS reaction mixtures that
included NH4Cl (20). However, the inclusion of (NH4)2SO4 in
WT SS reaction mixtures inhibits its isochorismate pyruvate lyase
activity. This is illustrated by the fact that the salicylate:isochoris-
mate product ratio formed by WT SS when NH3 is added is
∼6-fold smaller than that observed in WT SS-H2O reactions
(Table 2).
chorismate
conversion (%)
isochorismate
ADIC
62
54
84
7
1.9
44
17
42
40
1
0.8
77
salicylate
1.2
21
93
56
anthranilate
aReaction mixtures were incubated for 3 h at 25 °C and pH 8.0 prior to
separation by RP-HPLC. Each product is reported as a fraction of all pro-
ducts formed. Chorismate conversion is the percentage of 5 mM chorismate
enzymatically converted to products. All values were corrected for none-
nzymatic controls run under identical conditions.
The data in Table 2 show interesting differences in the dis-
tribution of elimination products, i.e., anthranilate and salicylate,
between WT SS and K147Q SS. An altered substrate preference
for the pyruvate lyase activity is observed in reactions catalyzed
by K147Q SS, relative to those catalyzed by WT SS. WT SS
prefers isochorismate over ADIC as a substrate for its lyase acti-
vity. Despite a high concentration of ADIC (1.4 mM) available to
WT SS, only 2% is converted to anthranilate. On the other hand,
WT SS forms 1.2 mM salicylate, which represents a 20%
conversion of 5 mM chorismate. The opposite trend is observed
in K147Q SS reactions. Relative to WT SS, K147Q SS forms
18-fold more anthranilate. This equates to a 21% conversion of
2.1 mM ADIC. Conversely, if supplied with 2 mM isochorismate
as the substrate in the absence of (NH4)2SO4, K147Q SS converts
only 7% to salicylate.
The Gln f Lys Mutation Does Not Engender High IS
Activity in AS and ADCS. In the simplest scenario, the Q f K
mutation was expected to confer AS and K213A ADCS with
salicylate synthase and isochorismate synthase activity, respec-
tively. For ADCS, the K213A mutation was introduced to render
it incapable of forming a Lys213-chorismate covalent inter-
mediate (16). HPLC chromatograms of WT and mutant-cata-
lyzed reactions, conducted in the absence of ammonia, are shown
in Figure 6. The 100-fold expansion inset in the Q147K AS trace
reveals the presence of a small peak due to isochorismate at
5.8 min and salicylate at 20.7 min (fluorescence signal displayed).
Isochorismate was not detected in K213A/Q147K ADCS reac-
tion mixtures.
chemical identification of ADC, ADIC, and isochorismate peaks
has been reported previously (14, 16). Product distributions listed
in Tables 1-4 reflect the fraction of total products formed by the
specified enzyme.
Removal of Lys147 Does Not Eliminate the Formation of
Isochorismate by IS and SS. If K147 is the base responsible for
activating water as a nucleophile in the IS- and SS-catalyzed
reactions, then the K f Q mutation is expected to render these
enzymes unable to catalyze formation of isochorismate. Further,
the ability to catalyze formation of ADIC and anthranilate by
K f Q mutants of IS and SS, respectively, was expected to be
unchanged relative to that of WT IS and SS if exogenous ammo-
nia was supplied (Figure 2). HPLC chromatograms of reaction
mixtures formed by WT IS and K147Q IS in the presence and
absence of (NH4)2SO4 are shown in Figure 4. Surprisingly, the
K147Q mutant forms isochorismate despite losing its presumed
catalytic base, at a rate reduced by ∼25-fold relative to the WT
rate. The HPLC data for WT IS reflect an equilibrium mixture;
the isochorismate to chorismate concentration ratio is 0.56, which
is identical to the Keq reported by Liu et al. (23). As expected,
K147Q IS retains the ability to catalyze formation of ADIC, and
the HPLC data in Table 1 indicate that rates of formation of
ADIC by WT and K147Q IS are similar.
HPLC chromatograms of product mixtures formed by WT SS
and K147Q SS in the presence and absence of (NH4)2SO4 are
shown in Figure 5. Loss of the presumed catalytic base, Lys147,
did not completely remove SS’s ability to use water as a nucleo-
phile, as evidenced by the presence of peaks due to isochorismate
and salicylate in K147Q SS reaction mixtures. The data in Table 2
The Q f K mutation was not anticipated to affect AS’s ability
to use ammonia as a nucleophile, since WT IS and WT SS, which
possess an analogous lysine residue, display high ADIC synthase
activity. L-Glutamine and (NH4)2SO4 were both used as ammonia

2856 Biochemistry, Vol. 49, No. 13, 2010
Ziebart and Toney
Table 4: ADCS Product Distributionsa
þ
þ
PabA þ
L-Gln
PabA þ NH4
NH4
WT enzyme
K213A
enzyme
K213A/Q147K
WT
K213A
enzyme
K213A/Q147K
enzyme
WT
K213A
K213A/Q147K
enzyme
enzyme
1.9
enzyme
enzyme
enzyme
chorismate
conversion (%)
ADC
PABAb
56
23
55
49
8.4
45
98
2
42
43
57
1.3
33
69
31
12
45
43
75
25
24
50
26
85
15
ADIC
100
67
isochorismate
aReaction mixtures were incubated for 3 h at 25 °C and pH 8.0 prior to separation by RP-HPLC. Each product is reported as a fraction of all products
formed. Chorismate conversion is the percentage of 5 mM chorismate enzymatically converted to products. All values were corrected for nonenzymatic
controls run under identical conditions. bPABA is formed by ADCL, which is present as a contaminant in PabA. Reactions were also conducted in the absence
of an ammonia source; those results are not shown because no products were formed.
F
IGURE 6: HPLC chromatograms of reactions catalyzed in the ab-
sence of an ammonia source by AS and its Q147K mutant and ADCS
and its two mutants, K213A and K213A/Q147K. Chorismate (5 mM)
was the substrate in each reaction. Insets (placed at the correct
position on the time axis) in the Q147K AS trace emphasize the
peaks due to isochorismate at 5.8 min and salicylate at 20.7 min.
Small differences in chorismate retention times are due to slight
variations in chromatographic conditions over the period of time
during which the data were collected.
F
IGURE 4: HPLC chromatograms of reactions catalyzed by IS or its
mutant K147Q, in the presence and absence of 100 mM (NH4)2SO4
(pH 8), with 5 mM chorismate as the substrate. PHB is p-hydro-
xybenzoate, a nonenzymatic decomposition product of chorismate.
sources. Surprisingly, with L-glutamine, Q147K AS did not exhi-
bit anthranilate synthase activity. However, using (NH4)2SO4,
29% of the WT activity was restored to Q147K AS, with
anthranilate representing 97% of all products (see Table 3). This
result indicates that exogenous ammonia has greater access to
C2 of chorismate than the ammonia that is released from TrpG-
mediated L-glutamine hydrolysis.
ADCS-catalyzed reactions were conducted under four differ-
ent conditions: (1) PabA-Gln, (2) PabA-NH4þ, (3) NH4þ, and (4)
H2O (Table 4). The results of the reactions conducted in the
absence of an ammonia source are not listed in Table 4 because
no products were formed. A small amount of activity is observed
in ammonia-dependent reactions, regardless of the source [i.e.,
PabA with L-Gln or (NH4)2SO4]. Previous work has demon-
strated that exogenous ammonia rescues the activity of K213A
ADCS; in fact, when (NH4)2SO4 is the ammonia source, the kcat
values of WT and K213A ADCS are nearly identical (14).
Therefore, the low activity of the double mutant is due to the
Q147K mutation. With exogenous ammonia, ADC and ADIC
were formed by K213A/Q147K ADCS at a rate 32-fold slower
compared to that of K213A ADCS and 35-fold slower compared
F
IGURE 5: HPLC chromatograms of reactions catalyzed by SS and
its Q147K mutant in the presence and absence of 100 mM (NH4)2SO4
(pH 8). Chorismate (5 mM) was the substrate in each reaction. In
cases of weak UV absorbance due to a low analyte concentration,
fluorescence signals are included as insets, which are placed at the
correct position on the time axis.

Article
Biochemistry, Vol. 49, No. 13, 2010 2857
Table 5: Amidotransferase Steady State Kinetic Data
KL-Gln (mM)
kcat (s-1
)
WT ADCS
0.52 ( 0.11
0.89 ( 0.07
2.5 ( 0.4
0.67 ( 0.05
0.60 ( 0.02
138 ( 9
K213A/Q147K ADCS
WT AS
Q147K AS
3.6 ( 0.9
152 ( 18
to that of WT ADCS. The product distribution (ADC vs ADIC)
in the double mutant is similar to that of K213A ADCS, which
indicates that other catalytic steps are responsible for the slow
turnover by K213A/Q147K ADCS. In other words, the nucleo-
philic addition steps may proceed at similar rates, but other
processes, such as product release, may be slower for the double
mutant, relative to K213A ADCS.
Glutamine Amidotransferase Activity of TrpG and PabA.
AS is a heterotetramer, consisting of a TrpE-TrpG heterodimer.
ADCS is a heterodimer of PabA and PabB subunits. To verify
that the Q f K mutation in PabB or TrpE does not affect the
glutamine amidotransferase activity of PabA or TrpG, respec-
tively, production of free ammonia was monitored using gluta-
mate dehydrogenase. PabA is a conditional amidotransferase;
only when in complex with ADCS does it possess activity (22).
Amidotransferase kinetic data, obtained with WT and mutant
TrpE and PabB, confirm that neither TrpG nor PabA activity
was affected by the mutation(s) in TrpE or PabB, respectively.
These data are summarized in Table 5.
AS, ADCS, IS, and SS Do Not Possess Chorismate
Mutase and Prephenate Dehydratase Promiscuous Activi-
ties. Before they were subjected to a second purification step,
every enzyme under study here, including mutants, exhibited
chorismate mutase (CM) promiscuous activity. Additionally,
WT AS, Q147K AS, K213A ADCS, and K213A/Q147K ADCS
exhibited prephenate dehydratase (PDT) activity as well. That is,
they formed phenylpyruvate when supplied with prephenate as
the substrate. However, these promiscuous activities were either
completely removed or significantly reduced after the proteins
were subjected to an additional orthogonal purification step (e.g.,
anion-exchange purification after nickel affinity purification).
Representative HPLC data are shown for singly and doubly
purified Q263K AS and WT SS in Figure 7. The peak at 21.8 min
is due to phenylpyruvate. Phenylpyruvate is the acid-catalyzed
decomposition product of prephenate. HPLC separations were
performed at pH 4. Therefore, phenylpyruvate appears in the
HPLC chromatogram regardless of whether the enzymes possess
PDT promiscuous activity in addition to CM promiscuous
activity.
F
IGURE 7: HPLC traces of singly (---) and doubly (;) purified SS
(A) and Q147K AS (B). Chorismate mutase activity, indicated by the
presence of phenylpyruvate, is significantly reduced after a second
purification.
DISCUSSION
singly purified proteins, but it was abolished or significantly
reduced in doubly purified proteins. Therefore, the observed CM
activity, originally thought to be a promiscuous activity of
ADCS, AS, IS, and SS, is due to authentic chorismate mutase
present in singly purified proteins that were expressed by an
E. coli lab strain [BL21(DE3)]. The E. coli chorismate mutase is a
bifunctional enzyme, possessing mutase and prephenate dehy-
dratase activities, with a kcat of 50 s-1 (26). Such an enzyme could
generate detectable product after 3 h even if present at nanomolar
to picomolar concentrations. Indeed, on the basis of this kcat
value and the concentration of phenylpyruvate measured by HPLC,
the CM concentration in the AS preparation was calculated to be
9 nM, which translates to a 0.05% contamination. The smallest
The survival of bacteria, plants, fungi, and apicomplexan para-
sites depends on the activity of several chorismate-utilizing enzymes,
which catalyze the first committed step in the biosynthesis of
aromatic amino acids, folate, siderophores, quinones, and others
(1, 2). These enzymes are attractive targets for antimicrobial
drugs, but they have not yet been exploited for this purpose.
Understanding more fully their catalytic mechanisms may assist
drug discovery efforts. AS, ADCS, IS, and SS each catalyze a
nucleophilic substitution reaction, and new information for charac-
terizing this step in catalysis is presented.
An important finding of this work is that AS, ADCS, IS, and
SS do not possess CM promiscuous activity. This contradicts
several previous studies (20, 24, 25). CM activity was observed in

2858 Biochemistry, Vol. 49, No. 13, 2010
Ziebart and Toney
concentration of CM present in any protein sample was 10 pM.
The four singly purified proteins originally thought to possess
PDT promiscuous activity (WT AS, Q263K AS, K213A ADCS,
and K213A/Q147K ADCS) in addition to CM promiscuous
activity all contained nanomolar levels of CM. Removal of the
contaminating E. coli CM protein by secondary purification
schemes facilitated efforts to investigate nucleophile specificity
in AS, ADCS, IS, and SS. In some mutant enzymes, nearly
100% of chorismate was converted to phenylpyruvate. Thus, the
removal of the competing CM activity allowed detection of
mutant enzyme-catalyzed reaction products by HPLC.
The retention of cognate activity in K147Q IS and K147Q SS
activity was a surprising result that contrasts with earlier reports
(see Figures 4 and 5) (9, 20, 24). It is unlikely that this activity is
due to a contamination of WT IS and SS in the mutant pre-
parations. After each mutant had been subjected to an additional
purification step, the isochorismate synthase activity increased
slightly. Furthermore, the HPLC data in Figure 6 indicate that
the Q f K mutation enables AS to activate water as a nucleo-
phile, albeit to a small extent. With a calculated HPLC detection
limit of ∼0.3 μM for isochorismate, introduction of Lys147
into AS enhanced its ability to activate water at pH 8 by at least
100-fold compared to that of WT. The WT reactions showed no
detectable amounts of isochorismate or salicylate at pH 7, 8, or 9,
while the Q f K mutant gave 10- and 2-fold increases on going
from pH 7 to 8 and from pH 8 to 9, respectively (data not shown).
These changes as a function of pH are best interpreted as depro-
tonation of a general base catalyst in the active site (assumed
here to be the introduced Lys); they are inconsistent with
hydroxide acting as a nucleophile since only a 2-fold increase is
observed on going from pH 8 to 9, while a 10-fold increase is
observed on going from pH 7 to 8.
While the results described above qualitatively agree with
Kolappan’s assertion that Lys147 is the catalytic base, they
strongly suggest that Lys147 is not the sole determinant of water
activation as a nucleophile given the high activity remaining in
the K f Q mutants. The results also suggest that in IS and SS a
different residue may act as a general base catalyst in the absence
of Lys147; however, its identification is not possible on the basis
of the available data. An alternative explanation is that Lys147
increases the reactivity of water 25-, 50-, and 100-fold in IS, SS,
and Q147K AS, respectively.
The failure to introduce IS activity into ADCS with a Q f K
mutation and the smaller than expected amount of SS activity
imparted to AS by the same mutation prompted a close inspec-
tion of multiple-sequence (54 total) and structural (6 total)
alignments of AS, ADCS, IS, and SS. This analysis revealed
two additional structural features that correlate with the ability to
activate water as a nucleophile: (1) the presence of two hydrogen
bond acceptors, one on either side of Lys147, to position
appropriately the ε-amino group toward C2 of chorismate and
(2) the presence of either leucine or isoleucine at position 244
instead of methionine (Figure 8), which occupies the analogous
position in all AS and ADCS sequences.
The correct orientation of Lys147 with respect to C2 of choris-
mate is critical since it is expected to act as a general base (i.e.,
removing a proton from water simultaneous with the O attacking
C2). This is achieved via two hydrogen bonds, one on each side of
the ε-amino group. In the Irp9 SS structure, these hydrogen bond
acceptors are Ser245 and the backbone carbonyl of Glu241
(Figure 8A). Ser245 is conserved in all SS sequences, while
leucine, isoleucine, and valine exist at the homologous position
F
IGURE 8: SS, IS, AS, and ADCS active sites. EntC IS residue num-
bering is used throughout. (A) Overlay of Irp9 SS residues (gray
carbons) and MenF IS (peach carbons). IS residues differing from SS
are indicated in parentheses. Irp9 SS crystallized in the presence of its
reaction products salicylate and pyruvate. Lys147 is the presumed
catalytic base that activates water as a nucleophile. (B) AS residues
(gray carbons) and the analogous ADCS residues (peach carbons).
Residues unique to ADCS are indicated in parentheses. Chorismate
was superimposed onto the benzoate molecule present in the AS crys-
tal structure.
in AS, ADCS, and IS sequences, respectively. In the MenF IS
structure, the backbone carbonyl of Ala303, which is ∼3 A from
˚
Lys147, is positioned to replace Ser245 as the hydrogen bond
acceptor. Analogous hydrogen bond interactions are not avail-
able to position correctly the introduced lysine residue in Q f K
AS and ADCS mutants. This may partly explain their low IS acti-
vity. Each IS and SS sequence contains either leucine or isoleu-
cine at position 244. Leu244 exists in the two SS structures and
one IS structure available. Leu244 may also assist the proper
positioning of Lys147 in IS and SS; from the structures, it appears
to make van der Waals interactions with the methylene carbons
of the Lys147 side chain.
The conserved methionine residue at position 244 in AS and
ADCS sequences is related to their specificity for ammonia as a
nucleophile. As shown in Figure 8B, a hydrogen bond between
the methionine sulfur atom and the Gln147 amide side chain is
possible. This interaction keeps the side chain of Gln147 pointed
toward Met244. When Gln147 is held in such a position, it creates
an appropriately sized cavity through which ammonia, arriving via
the tunnel from the TrpG active site, may access C2 of chorismate.
A pairwise positional correlation analysis of 54 AS, ADCS, IS,
and SS sequences, performed using the online CRASP program
(27), corroborated the supposition that the identity of residue
244 is a critical component of the nucleophile specificity displayed
by these enzymes. In the CRASP analysis, the “energy of transfer
from water to ethanol” parameter was chosen with a variability
threshold of 2 and a significance output value of 99.99%. Under
these parameters, Lys147 was found to be correlated to Leu244
with a linear correlation value of 0.98. Leu244 corresponds to
methionine in all AS and ADCS sequences.
Kerbarh et al. suggested that Lys147 plays a dual role in SS
and IS nucleophile specificity, both as an activator of water and,
via steric and/or electronic forces, as a repressor of ammonia as a
nucleophile (20). However, the formation of ADIC by WT IS,
WT SS, and K213A/Q147K ADCS and the formation of anthra-
nilate by Q147K AS argue against this assertion. Since intracel-
lular concentrations of ammonia are generally low, IS and SS
apparently have not developed a means of discriminating against
this nucleophile.
The presence of an amidotransferase subunit is the most
important factor that determines which enzymes will employ

Article
Biochemistry, Vol. 49, No. 13, 2010 2859
ammonia as a nucleophile. The ∼450-fold reduction of cognate
Serratia marcescens crystallized in the presence of (i) its substrates,
chorismate and glutamine, and a product, glutamate, and (ii) its end-
activity in Q147K AS when L-glutamine is the source of ammonia
product inhibitor,
6021–6026.
L-tryptophan. Proc. Natl. Acad. Sci. U.S.A. 98,
indicates that the identity of residue 147 is important insofar as it
must not block the access of ammonia to the TrpE active site. An
∼30-fold reduction in K213A/Q211K ADCS cognate activity
9. Kolappan, S., Zwahlen, J., Zhou, R., Truglio, J. J., Tonge, P. J., and
Kisker, C. (2007) Lysine 190 is the catalytic base in MenF, the
menaquinone-specific isochorismate synthase from Escherichia coli:
Implications for an enzyme family. Biochemistry 46, 946–953.
10. Harrison, A. J., Yu, M., Gardenborg, T., Middleditch, M., Ramsay,
R. J., Baker, E. N., and Lott, J. S. (2006) The structure of MbtI from
Mycobacterium tuberculosis, the first enzyme in the biosynthesis
of the siderophore mycobactin, reveals it to be a salicylate synthase.
J. Bacteriol. 188, 6081–6091.
was observed under L-Gln conditions. This suggests that the
PabA-PabB interface, and, hence, the ammonia tunnel that is
believed to connect the two active sites, is oriented slightly diffe-
rently than the TrpG-TrpE interface observed in the AS crystal
structure (5). The ADCS crystal structure was not determined in
the presence of PabA (7); therefore, the precise arrangement of
PabA and PabB subunits is unknown.
11. Parsons, J. F., Shi, K. M., and Ladner, J. E. (2008) Structure of
isochorismate synthase in complex with magnesium. Acta Crystallogr.
D64, 607–610.
The data in Tables 2 and 3 suggest that SS pyruvate lyase
activity is partially controlled by Lys147. In WT SS-NH3 and
K147Q SS-NH3 reactions, significant build-up of ADIC was
observed (Figure 5). Notably, ADIC does not accumulate in WT
AS reactions, and 96% of ADIC is converted to anthranilate by
Q147K AS when (NH4)2SO4 is the ammonia source. Morollo
and Bauerle demonstrated that ADIC is an intermediate in the
AS reaction (28). If Lys147 were the sole determinant of reaction
specificity (i.e., formation of anthranilate vs salicylate), then a
K f Q mutation would render SS equivalent to AS, since each
possesses pyruvate lyase activity. However, WT SS forms little
anthranilate, despite high concentrations of ADIC in the reaction
mixture. On the other hand, anthranilate represents 97% of all
products formed in Q263K AS-NH3 reactions. These contrast-
ing results suggest that SS reaction specificity is controlled
by factors beyond the selection of water as a nucleophile. The
improved ability of K147Q SS to aromatize ADIC and a
weakened ability to aromatize isochorismate (relative to WT
SS) additionally support this notion. Future efforts will be aimed
at uncovering the forces that dictate elimination specificity
among AS, SS, ADCS, and IS.
12. Walsh, C. T., Liu, J., Rusnak, F., and Sakaitani, M. (1990) Molecular
Studies on Enzymes in Chorismate Metabolism and the Enterobactin
Biosynthetic-Pathway. Chem. Rev. 90, 1105–1129.
13. Kozlowski, M. C., Tom, N. J., Seto, C. T., Sefler, A. M., and Bartlett,
P. A. (1995) Chorismate-utilizing enzymes isochorismate synthase,
anthranilate synthase, and p-aminobenzoate synthase: Mecha-
nistic insight through inhibitor design. J. Am. Chem. Soc. 117,
2128–2140.
14. He, Z., Stigers Lavoie, K. D., Bartlett, P. A., and Toney, M. D. (2004)
Conservation of mechanism in three chorismate-utilizing enzymes.
J. Am. Chem. Soc. 126, 2378–2385.
15. Bulloch, E. M. M., and Abell, C. (2005) Detection of covalent
intermediates formed in the reaction of 4-amino-4-deoxychorismate
synthase. ChemBioChem 6, 832.
16. He, Z., and Toney, M. D. (2006) Direct detection and kinetic analysis
of covalent intermediate formation in the 4-amino-4-deoxychoris-
mate synthase catalyzed reaction. Biochemistry 45, 5019–5028.
17. Raushel, F. M., Thoden, J. B., and Holden, H. M. (2003) Enzymes
with molecular tunnels. Acc. Chem. Res. 36, 539–548.
18. Huang, X. Y., Holden, H. M., and Raushel, F. M. (2001) Channeling
of substrates and intermediates in enzyme-catalyzed reactions. Annu.
Rev. Biochem. 70, 149–180.
19. Bauerle, R., Hess, J., and French, S. (1987) Anthranilate synthase-
anthranilate phosphoribosyltransferase complex and subunits of
Salmonella typhimurium. Methods Enzymol. 142, 366–386.
20. Kerbarh, O., Ciulli, A., Chirgadze, D. Y., Blundell, T. L., and Abell,
C. (2007) Nucleophile selectivity of chorismate-utilizing enzymes.
ChemBioChem 8, 622–624.
21. Grisostomi, C., Kast, P., Pulido, R., Huynh, J., and Hilvert, D. (1997)
Efficient in Vivo Synthesis and Rapid Purification of Chorismic Acid
Using an Engineered Escherichia coli Strain. Bioorg. Chem. 25, 297–
305.
REFERENCES
1. Roberts, F., Roberts, C. W., Johnson, J. J., Kyle, D. E., Krell, T.,
Coggins, J. R., Coombs, G. H., Milhous, W. K., Tzipori, S., Ferguson,
D. J., Chakrabarti, D., and McLeod, R. (1998) Evidence for the
shikimate pathway in apicomplexan parasites. Nature 393, 801–805.
2. Dosselaere, F., and Vanderleyden, J. (2001) A metabolic node in
action: Chorismate-utilizing enzymes in microorganisms. Crit. Rev.
Microbiol. 27, 75–131.
3. Mavrodi, D. V., Ksenzenko, V. N., Bonsall, R. F., Cook, R. J.,
Boronin, A. M., and Thomashow, L. S. (1998) A seven-gene locus for
synthesis of phenazine-1-carboxylic acid by Pseudomonas fluorescens
2-79. J. Bacteriol. 180, 2541–2548.
4. Kerbarh, O., Chirgadze, D. Y., Blundell, T. L., and Abell, C. (2006)
Crystal structures of Yersinia enterocolitica salicylate synthase and its
complex with the reaction products salicylate and pyruvate. J. Mol.
Biol. 357, 524–534.
5. Morollo, A. A., and Eck, M. J. (2001) Structure of the cooperative
allosteric anthranilate synthase from Salmonella typhimurium. Nat.
Struct. Biol. 8, 243–247.
6. Knochel, T., Ivens, A., Hester, G., Gonzalez, A., Bauerle, R.,
Wilmanns, M., Kirschner, K., and Jansonius, J. N. (1999) The crystal
structure of anthranilate synthase from Sulfolobus solfataricus: Func-
tional implications. Proc. Natl. Acad. Sci. U.S.A. 96, 9479–9484.
7. Parsons, J. F., Jensen, P. Y., Pachikara, A. S., Howard, A. J., Eisenstein,
E., and Ladner, J. E. (2002) Structure of Escherichia coli aminodeoxy-
chorismate synthase: Architectural conservation and diversity in choris-
mate-utilizing enzymes. Biochemistry 41, 2198–2208.
22. Roux, B., and Walsh, C. T. (1992) p-Aminobenzoate synthesis in
Escherichia coli: Kinetic and mechanistic characterization of the
amidotransferase PabA. Biochemistry 31, 6904–6910.
23. Liu, J., Quinn, N., Berchtold, G. A., and Walsh, C. T. (1990) Over-
expression, purification, and characterization of isochorismate
synthase (EntC), the first enzyme involved in the biosynthesis of
enterobactin from chorismate. Biochemistry 29, 1417–1425.
24. Zwahlen, J., Kolappan, S., Zhou, R., Kisker, C., and Tonge, P. J.
(2007) Structure and mechanism of MbtI, the salicylate synthase from
Mycobacterium tuberculosis. Biochemistry 46, 954–964.
25. Zhou, R., Zwahlen, J., Subramaniapillar, K., Kisker, C., and Tonge,
P. (2006) Comparative structural and biochemical studies of choris-
mate binding enzymes, MenF, EntC and MbtI. FASEB J. 20,
A463.
26. Lee, A., Stewart, J. D., Clardy, J., and Ganem, B. (1995) New insight
into the catalytic mechanism of chorismate mutases from structural
studies. Chem. Biol. 2, 195–203.
27. Afonnikov, D. A., Oshchepkov, D. Yu., and Kolchanov, N. A. (2001)
Detection of conserved physico-chemical characteristics of proteins
by analyzing clusters of positions with co-ordinated substitutions.
Bioinformatics 17 (11), 1035–1046.
28. Morollo, A. A., and Bauerle, R. (1993) Characterization of composite
aminodeoxyisochorismate synthase and aminodeoxyisochorismate
lyase activities of anthranilate synthase. Proc. Natl. Acad. Sci. U.S.A.
90, 9983–9987.
8. Spraggon, G., Kim, C., Nguyen-Huu, X., Yee, M. C., Yanofsky, C.,
and Mills, S. E. (2001) The structures of anthranilate synthase of