Benzimidazole analogs of l-tryptophan are substrates and inhibitors of tryptophan indole lyase from Escherichia coli

Article abstract of DOI:10.1111/febs.12205

Tryptophan indole lyase (TIL), an enzyme found in Escherichia coli and related enterobacteria, produces indole from l-tryptophan (l-Trp). Indole is a signaling molecule in bacteria, affecting biofilm formation, pathogenicity and antibiotic resistance. β-(Benzimidazol-1-yl)-l-alanine (BZI-Ala), 2-amino-4-(benzimidazol-1-yl)butyric acid (homo-BZI-Ala) and 2-amino-5-(benzimidazol-1-yl)pentanoic acid (bishomo-BZI-Ala) were synthesized and tested as substrates and inhibitors of TIL. BZI-Ala is a good substrate of TIL, with K_m = 300 μm, k_cat = 5.6 s^-1 and k_cat/K_m = 1.9 × 10⁴, similar to l-Trp. BZI-Ala is also a good substrate for H463F mutant TIL, which has very low activity with l-Trp. In contrast, homo-BZI-Ala was found to be a potent competitive inhibitor of TIL, with a K_i of 13.4 μm. However, the higher homolog, bishomo-BZI-Ala, was inactive as an inhibitor of TIL at a concentration of 600 μm, and is thus a much weaker inhibitor. The reaction of TIL with BZI-Ala was too fast to be observed in the stopped-flow spectrophotometer, and shows an aldimine intermediate in the steady state. However, H463F TIL shows equilibrating mixtures of aldimine and quinonoid complexes in the steady state. The spectra of the reaction of TIL with homo-BZI-Ala show a rapidly formed intermediate absorbing at 340 nm, probably a gem-diamine, that decays slowly to form a quinonoid complex absorbing at 494 nm. The potent binding of homo-BZI-Ala may be due to it being a 'bi-product' analog of the indole-α-aminoacrylate complex. These results demonstrate that an amino acid substrate may be converted to a potent inhibitor of TIL simply by homologation, which may be useful in the design of other potent TIL inhibitors. β-(Benzimidazol-1-yl)-l-alanine (BZI-Ala), 2-amino-4-(benzimidazol-1-yl) butyric acid (homo-BZI-Ala), and 2-amino-5-(benzimidazol-1-yl)pentanoic acid (bishomo-BZI-Ala) were synthesized and tested as substrates and inhibitors of tryptophan indole-lyase (TIL), an enzyme found in Escherichia coli and related enterobacteria. BZI-Ala is a good substrate of TIL, homo-BZI-Ala is a potent competitive inhibitor of TIL, with K_i of 13.4 μM, but bishomo-BZI-Ala, was inactive as an inhibitor of TIL. 2013 The Authors Journal compilation

Full text of DOI:10.1111/febs.12205

Benzimidazole analogs of L-tryptophan are substrates and
inhibitors of tryptophan indole lyase from Escherichia coli
Austin P. Harris1 and Robert S. Phillips1,2
1 Department of Chemistry, University of Georgia, Athens, GA, USA
2 Department of Biochemistry and Molecular Biology, University of Georgia, Athens, GA, USA
Keywords
Tryptophan indole lyase (TIL), an enzyme found in Escherichia coli and
related enterobacteria, produces indole from L-tryptophan (L-Trp). Indole
is a signaling molecule in bacteria, affecting biofilm formation, pathogenic-
enzyme mechanism; pyridoxal-5′-phosphate;
steady-state kinetics; stopped-flow kinetics;
b-elimination
ity and antibiotic resistance. b-(Benzimidazol-1-yl)-L-alanine (BZI-Ala),
2-amino-4-(benzimidazol-1-yl)butyric acid (homo-BZI-Ala) and 2-amino-5-
Correspondence
(benzimidazol-1-yl)pentanoic acid (bishomo-BZI-Ala) were synthesized and
tested as substrates and inhibitors of TIL. BZI-Ala is a good substrate of
TIL, with Km = 300 lM, kcat = 5.6 sꢀ1 and kcat/Km = 1.9 9 104, similar to
L-Trp. BZI-Ala is also a good substrate for H463F mutant TIL, which has
very low activity with L-Trp. In contrast, homo-BZI-Ala was found to be a
potent competitive inhibitor of TIL, with a Ki of 13.4 lM. However, the
higher homolog, bishomo-BZI-Ala, was inactive as an inhibitor of TIL at
a concentration of 600 lM, and is thus a much weaker inhibitor. The reac-
tion of TIL with BZI-Ala was too fast to be observed in the stopped-flow
spectrophotometer, and shows an aldimine intermediate in the steady state.
However, H463F TIL shows equilibrating mixtures of aldimine and quino-
noid complexes in the steady state. The spectra of the reaction of TIL with
homo-BZI-Ala show a rapidly formed intermediate absorbing at 340 nm,
probably a gem-diamine, that decays slowly to form a quinonoid complex
absorbing at 494 nm. The potent binding of homo-BZI-Ala may be due to
it being a ‘bi-product’ analog of the indole-a-aminoacrylate complex. These
results demonstrate that an amino acid substrate may be converted to a
potent inhibitor of TIL simply by homologation, which may be useful in
the design of other potent TIL inhibitors.
R. S. Phillips, Department of Chemistry,
Athens, GA 30602-2556, USA
Fax: +1 706 542 9454
Tel: +1 706 5421996
E–mail: plp@uga.edu
(Received 31 December 2012, revised 14
February 2013, accepted 18 February 2013)
doi:10.1111/febs.12205
Introduction
Indole-positive Gram-negative enterobacteria such as
Escherichia coli, Klebsiella oxytoca, Providencia stuart-
ii, Citrobacter koseri, Morganella morganii and Hae-
mophilus influenza have long been recognized as a
threat to public health. These indole-positive microbes
are characterized by their ability to degrade L-Trp to
indole and ammonium pyruvate, using the enzyme
tryptophan indole lyase (tryptophanase, EC 4.1.99.1,
TIL)
(1)
The ability of these microbes to produce indole is
correlated to their pathogenicity [1,2]. Until recently,
the role that TIL plays in the metabolism of indole-
positive microbes was considered to be solely cata-
bolic, to provide energy from excess tryptophan [3].
Abbreviations
bishomo-BZI-Ala, (2S)-2-amino-5-(benzimidazol-1-yl)pentanoic acid; BZI-Ala, 3-b-(benzimidazol-1-yl)-L-alanine; homo-BZI-Ala,
(2S)-4-(benzimidazol-1-yl)-2-aminobutanoic acid; PLP, pyridoxal-5′-phosphate; TIL, tryptophan indole lyase (tryptophanase) [EC 4.1.99.1].
FEBS Journal (2013) ª 2013 The Authors Journal compilation ª 2013 FEBS
1

Benzimidazoles with tryptophan indole lyase
A. P. Harris and R. S. Phillips
However, a number of recent studies have provided
support for the hypothesis that indole, produced
through the degradation of L-Trp in Eqn (1) by TIL,
is probably involved in extracellular regulatory
signaling in bacteria [4]. Indole formation affects the
production of biofilm in indole-positive Gram-nega-
tive bacteria [5,6]. When inhibitors of TIL are intro-
duced into an indole-positive Gram-negative bacteria
growth assay, the formation of bacterial biofilm is
reduced [6]. Because formation of bacterial biofilm is
an essential step in the bacterial infection process, the
preparation of TIL inhibitors is of interest as an
alternative to existing antibiotics [7]. Indole has also
been shown to play an essential role in acid resistance
[8] and in the stability of high-copy-number plasmids
in E. coli [9].
an indolenine intermediate, and cleavage of the Cb–Cc
bond follows. Pre-steady-state kinetic isotope effect
experiments showed that the isotope effects from deu-
terium substitution at the a- and b-carbons are addi-
tive, consistent with a concerted mechanism suggesting
it is a transition state rather than an intermediate [16].
The formation and breakdown of this indolenine inter-
mediate/transition state is facilitated by hydrogen
bonding of the indole NH, possibly with His463 or
Asp137, and by proton transfer to C3 from Tyr74
[19,20]. Indolenine is a tautomer of indole that is
structurally similar to benzimidazole, with an imine
C=N bond rather than an enamine at N1–C2, and,
due to this structural similarity, we hypothesized
that analogs of L-Trp with the indole ring replaced
by benzimidazole may be substrates and/or potent
inhibitors of TIL. In this paper, we synthesized
three novel benzimidazole analogs of L-Trp and
L-homotryptophan: 3-b-(benzimidazol-1-yl)-L-alanine
(BZI-Ala, 4a), 4-(benzimidazol-1-yl)-2-aminobutanoic
acid, (homo-BZI-Ala, 4b) and 5-(benzimidazol-1-yl)-2-
amino-pentanoic acid (bishomo-BZI-Ala, 4c), and per-
formed steady-state and pre-steady-state kinetic studies
on these analogs.
The mechanism of the reaction of L-Trp with TIL
has been extensively studied in our laboratory [10–16].
A key species proposed to be formed during the degra-
dation of L-Trp by TIL is the indolenine intermediate
(Scheme 1) [16–18]. After reaction of L-Trp with
enzyme-bound pyridoxal-5′-phosphate (PLP), the
external aldimine is deprotonated, forming a quino-
noid complex. Proton transfer to the indole ring forms
Scheme 1. Proposed mechanism for TIL.
2
FEBS Journal (2013) ª 2013 The Authors Journal compilation ª 2013 FEBS

A. P. Harris and R. S. Phillips
Benzimidazoles with tryptophan indole lyase
Scheme 2. Synthesis of benzimidazole analogs of L-Trp.
of the next higher homolog, 2-amino-5-(benzimidazol-
1-yl)pentanoic acid (bishomo-BZI-Ala) (4c), was simi-
lar, but started with readily available L-ornithine,
which was converted to Na-Cbz-Nd-Boc-L-ornithine, as
described by Wiejak et al. [22]. Removal of the t-butyl-
oxycarbonyl group gave Na-Cbz-L-ornithine (1c),
which was then converted to the benzimidazole under
the conditions shown in Scheme 2. However, the yield
for this transformation was much less than for the
shorter-chain homologs.
Results
Synthesis
BZI-Ala (4a) and homo-BZI-Ala (4b), which may be
considered as 3-aza analogs of L-Trp and HomoTrp,
respectively, were synthesized according to Scheme 2.
Na-Cbz-L-Asn and Na-Cbz-L-Asn were subjected to
Hoffmann rearrangement with diacetoxyiodobenzene
to produce the corresponding b- and c-amines [21].
N-arylation was then achieved using 2-fluoronitroben-
zene in dimethylformamide, followed by reduction
with Zn, formylation and cyclization in refluxing for-
mic acid, in one pot, and subsequent acid hydrolysis
of the carbobenzyloxy group. Catalytic reduction of
the o-amino nitro group and formylation in formic
acid under standard conditions with 10% palladium
on carbon was found to be very slow under 0.34 MPa
H2 at ambient temperature, and required heating to
100 °C. The reduction with Zn in formic acid was
more convenient and gave similar results. Preparation
Steady-state kinetics
The results from the steady-state kinetics studies of
the benzimidazole analogs of L-Trp shown in
Scheme 3 are given in Table 1. BZI-Ala (4a) was
found to be a good substrate for E. coli TIL, with
Km = 315 lM, slightly higher than the Km for L-Trp
of 150 lM. The kcat value is comparable to that of
L-Trp, so the kcat/Km for L-Trp is approximately 1.5
times that of BZI-Ala. In contrast, the kcat/Km of
Scheme 3. Comparison of structures of L-Trp and benzimidazole analogs.
FEBS Journal (2013) ª 2013 The Authors Journal compilation ª 2013 FEBS
3

Benzimidazoles with tryptophan indole lyase
A. P. Harris and R. S. Phillips
Table 1. Steady-state kinetic parameters for reaction of analogs with TIL.
ꢀ1
Substrate/enzyme
Ki (lM)
Km (lM)
kcat (sꢀ1
)
kcat/Km (M ꢂsꢀ1
)
L-Trpa/wild-type
–
150
4.0
0.5
2.7 9 104
2.5 9 103
b-indazolyl-L-Alaa/wild-type
ΒZI-Ala (4a)/wild-type
BZI-Ala (4a)/H463F
ND
200
301 ꢁ 1
ND
315 ꢁ 64
5.6 ꢁ 0.6
(1.8 ꢁ 0.2) 9 104
23.6 ꢁ 2.9
0.22 ꢁ 0.01
(9.3 ꢁ 1.0) 9 103
Homo-BZI-Ala (4b)/wild-type
Bishomo-BZI-Ala (4c)/wild-type
13.4 ꢁ 1.5
> 600
–
–
–
–
–
–
aData from [15].
BZI-Ala is approximately six times that of b-indazol-
yl-L-alanine, another aza analog containing an addi-
tional nitrogen in the pyrrole ring of the indole
(Scheme 3). The next higher homolog, homo-BZI-Ala
(4b), was found not to be a substrate of E. coli TIL,
but instead to be a potent inhibitor of E. coli TIL,
with Ki = 13.4 lM. In contrast, the higher homolog,
bishomo-BZI-Ala (4c), was found to be inactive as
an inhibitor of E. coli TIL at a concentration of
600 lM, and thus the Ki must be > 600 lM. H463F
TIL is a mutant enzyme with much lower elimination
activity for L-Trp, < 0.1% of wild-type TIL [19].
However, BZI-Ala is an excellent substrate for
H463F TIL, with a kcat/Km that is approximately
half that of wild-type TIL (Table 1).
intermediates was seen in the spectra in the steady state,
as only an aldimine-type spectrum was observed, with
kmax at approximately 420 nm (data not shown). In
contrast to wild-type TIL, the reaction of H463F TIL
with BZI-Ala in the stopped-flow spectrophotometer
shows formation of a prominent quinonoid band at
494 nm, and decay of the external aldimine at 423 nm,
with a clear isosbestic point at 433 nm between the
external aldimine and quinonoid intermediates
(Fig. 1A), over a period of approximately 2 s (Fig. 1B).
The global fit requires two exponentials to obtain an
adequate fit, with 1/τ1 = 3.1 ꢁ 0.2 sꢀ1 and 1/τ2 =
1.6 ꢁ 0.2 sꢀ1. This spectrum is quasi-stable during the
steady state of the reaction, decaying after the substrate
is depleted in 1–2 min.
The rapid-scanning stopped-flow spectra resulting
from mixing 2 mM homo-BZI-Ala with 2 mgꢂmLꢀ1
wild-type TIL are shown in Fig. 2A. Homo-BZI-Ala
initially forms a peak at 340 nm, within the instrument
dead time, that decays slowly to give peaks at 388 and
494 nm. The isosbestic point between the peaks at 340
and 494 nm suggests that this peak reacts to form the
494 nm complex, without accumulation of an interme-
Pre-steady-state kinetics
Rapid-scanning stopped-flow kinetic measurements
taken for the reaction of BZI-Ala with TIL revealed
that the reaction reached steady-state within the mixing
time of the instrument (approximately 2 ms). No evi-
dence for accumulation of quinonoid or aminoacrylate
8
A
B
7
494 nm
0.25
6
0.25
0.2
0.2
5
4
0.15
0.1
0.15
0.1
3
2
Fig. 1. Reaction of E. coli H463F TIL
0
0.05
0
0.05
0
(0.5 mgꢂmLꢀ1, 9.8 lM) with 2 mM BZI-Ala.
(A) Selected spectra are shown: 0, enzyme
spectrum before mixing; 1, 0.010 s; 2,
0.060 s; 3, 0.120 s; 4, 0.240 s; 5, 0.480 s;
6, 0.960 s; 7, 1.50 s; 8, 1.92 s. (B) Time
courses of the reaction at 423 and 494 nm.
423 nm
1.5
1
350
400
450
500
5500
0.5
1
Wavelength (nm)
Time (s)
4
FEBS Journal (2013) ª 2013 The Authors Journal compilation ª 2013 FEBS

A. P. Harris and R. S. Phillips
Benzimidazoles with tryptophan indole lyase
0.5
494 nm
A
C
B
7
0.4
0.3
0.2
6
0.4
0.2
5
4
3
2
335 nm
0.1
Fig. 2. Reaction of E. coli TIL (2 mgꢂmLꢀ1
38.5 lM) with 2 mM homo-BZI-Ala. (A)
;
388 nm
1
0
0
0.002
D
Selected spectra are shown: 1, 0.008 s; 2,
0.120 s; 3, 0.500 s; 4, 1.00 s; 5, 4.00 s; 6,
8.00 s; 7, 12.00 s. (B) Time courses for the
reaction at 335, 388 and 494 nm. (C) SVD
spectra from the global fit of the stopped-
flow data. Solid line, spectrum 1; dashed
line, spectrum 2; dashed/dotted line,
0.6
0
–0.002
–0.004
–0.006
0.4
0.2
0
spectrum 3; dotted line, spectrum 4. (D)
Residuals from the global fit of the data to
two exponential processes (dotted line) or
three exponential processes (solid line).
300 350 400 450 500 550 600
Wavelength (nm)
0.1
1
Time (s)
diate. In addition, a weak shoulder at approximately
388 nm forms simultaneously with the 494 nm peak.
A global fit of the kinetic data at 1.6 mM homo-BZI-
Ala revealed the reaction to be triphasic, comprising
four species (Fig. 2B), whose corresponding rate
constants are 1/τ1 = 3 ꢁ 0.1 sꢀ1, 1/τ2 = 4.5 ꢁ 0.2 9
reduction/cyclization gave high yields for the products
from Asn and Gln, but unexpectedly gave very low
yields for the reaction sequence starting with Orn. In
the latter case, a black solid was obtained, suggesting
that oxidation of the o-diaminobenzene intermediate
occurs faster than formylation and cyclization to the
benzimidazole. A very small amount of the desired
product was obtained by reverse-phase chromatogra-
phy. As we observed no inhibition of TIL by this com-
pound at a concentration of 600 lM, we did not
attempt to optimize the synthesis further. These
tryptophan analogs may be interesting as mechanistic
probes or inhibitors of other tryptophan-metabolizing
enzymes.
10ꢀ1 sꢀ1 and 1/τ3 = 8.0 ꢁ 0.4 9 10ꢀ2
s
ꢀ1. A biphasic
fit with only three species gave much worse residuals
(Fig. 2D). These three phases correspond to decay of
the 340 nm intermediate and formation of the 494 nm
peak, as may be seen from the time courses (Fig. 2C).
The rate constants for all three phases of the reaction
were found not to be significantly concentration-
dependent over the concentration range from 0.1 to
1.6 mM.
Steady-state kinetics
Discussion
Of the compounds tested, only BZI-Ala was expected
and found to be a substrate. The reaction rate con-
stants calculated for BZI-Ala are shown in Table 1.
The Ki of BZI-Ala was found to be 301 lM, while
the Km was found to be 315 lM, an excellent agree-
ment. The Km of L-Trp and BZI-Ala (Table 1) differ
by less than a factor of two. The enzyme catalytic
efficiency for the two substrates also differs only
slightly, as the kcat/Km for L-Trp was determined to
Synthesis
We prepared the benzimidazole analogs of L-Trp (BZI-
Ala), homoTrp (homo-BZI-Ala) and bishomoTrp (bis-
homo-BZI-Ala) by a novel stereospecific route starting
from L-Asn, L-Gln and L-Orn, respectively, as shown
in Scheme 2. BZI-Ala has been made previously only
as a racemate, by malonic ester synthesis [23], while
homo-BZI-Ala and bishomo-BZI-Ala have not been
described. The aromatic ring was introduced on the
primary amine by reaction with 2-fluoronitrobenzene,
followed by reduction with Zn, formylation of the
resulting o-diaminobenzene in formic acid, and cycliza-
tion, and finally hydrolysis in refluxing 6 M HCl to
give the substituted benzimidazole amino acid. The
ꢀ1
be 27 mM ꢂsꢀ1 whilꢀe1 that for BZI-Ala was deter-
mined to be 18 mM ꢂsꢀ1 (Table 1). Similarly, the
values of kcat for L-Trp and BZI-Ala are comparable.
Thus L-Trp and BZI-Ala are both excellent substrates
for TIL, which may be attributed to their structural
similarities. For H463F TIL, BZI-Ala is also an
excellent substrate, with a kcat/Km approximately half
FEBS Journal (2013) ª 2013 The Authors Journal compilation ª 2013 FEBS
5

Benzimidazoles with tryptophan indole lyase
A. P. Harris and R. S. Phillips
that of wild-type TIL with this substrate (Table 1).
As H463F TIL was found in a previous study to
have very low activity with L-Trp [19], < 0.1%, it
may be concluded that His463 does not play as criti-
cal a role in the reaction of BZI-Ala with wild-type
TIL as it does for L-Trp.
visible absorption peak. gem-Diamines are essential
intermediates in the transaldimination reaction of sub-
strates, but usually do not accumulate, and are rarely
observed in PLP-dependent enzyme reactions. How-
ever, with 1-aminocyclopropane-1-carboxylate deami-
nase, a relatively stable gem-diamine intermediate
forms [24]. It is also interesting that we previously
observed a transient gem-diamine intermediate absorb-
ing at 340 nm in the reaction of homophenylalanine
with TIL [12]. Thus, it appears that gem-diamines are
stabilized in the reaction of TIL with substrate analogs
with extended chains. The gem-diamine then decays
slowly to form a spectrum dominated by the 494 nm
absorption assigned to a quinonoid species (Fig. 2A,
B), without accumulation of an external aldimine
intermediate. A global fit of the time course of decay
of the gem-diamine at 340 nm and formation of the
quinonoid species at 494 nm revealed that the reaction
occurs by three observable steps, with the singular
value decomposition (SVD) spectra shown in Fig. 2C
and the residuals shown in Fig. 2D. Based on the L-
Trp/TIL mechanism (Scheme 1), and the pre-steady
state data, we propose that the reaction of homo-BZI-
Ala consists of seven steps, as shown in Scheme 4. The
first step in the reaction involves the fast unobserved
process of homo-BZI-Ala reacting with TIL (E + I) to
form the gem-diamine (G). The rate constant for this
step, k1, must be > 2000 sꢀ1 because formation of the
gem-diamine intermediate was complete within the
instrumental dead time (approximately 2 ms). Next, G
is expected to inter-convert by elimination of the
Lys270 amino group to give an external aldimine (A)
with an absorption maximum at approximately
420 nm. However, we observed conversion of G to a
spectrum that is a mixture of an intense band at
It is interesting to compare b-indazolylalanine to
BZI-Ala, because both substrates possess an amino
acid side chain that is attached to a nitrogen atom in
their respective isomeric heterocyclic rings (Scheme 3).
ꢀ1
The kcat/Km for BZI-Ala (18 mM ꢂsꢀ1) is approximately
six times greater than that of b-indazolylalanine
ꢀ1
(2.7 mM ꢂsꢀ1). This may be due to BZI-Ala having a
nitrogen atom capable of hydrogen bonding at
position 1 of the heterocyclic ring, similar to L-Trp.
Since these aromatic rings are poor leaving groups as
the strongly basic N-anions (pKa values of approxi-
mately 13), protonation of the second nitrogen must
occur prior to or concomitant with elimination. Hence,
we attribute the 10-fold difference between the kcat val-
ues of TIL with b-indazolylalanine and BZI-Ala to the
fact that the benzimidazole ring (pKa of 5.53) in BZI-
Ala is a better leaving group than the indazole ring
(pKa = 1.25) present in b-indazolylalanine. This
conclusion is also consistent with accumulation of a
quinonoid intermediate in the reaction of b-indazolyl-
alanine with wild-type TIL [15], but not for BZI-Ala,
showing that elimination is at least partially rate-deter-
mining for b-indazolylalanine but not for BZI-Ala.
Pre-steady-state kinetics
The reaction of BZI-Ala with wild-type TIL did not
show accumulation of any reaction intermediates
except the external aldimine, which was formed within
the dead time of the stopped-flow instrument. This
implies that quinonoid intermediate formation must be
concerted with or slower than elimination. In contrast,
the mutant H463F TIL forms a prominent band at
494 nm with BZI-Ala, indicating that the quinonoid
intermediate accumulates during turnover, and hence
elimination is slower than deprotonation. Thus,
although the kcat/Km parameters for the H463T
mutant and wild-type enzymes with this substrate are
similar, the rate-determining step is clearly different,
with deprotonation probably being rate-limiting for
wild-type TIL and elimination probably being rate-lim-
iting for H463F TIL.
494 nm and
a weak shoulder at approximately
388 nm. The position of the shoulder is blue-shifted
from the normal position of the external aldimine, at
approximately 420–425 nm [10–16]. However, if the
external aldimine is strained due to twisting and can-
not be protonated on the imine nitrogen, the absorp-
tion maximum of the external aldimine may be blue-
shifted to 360–390 nm. In aspartate aminotransferase
and aspartate b-decarboxylase, the internal aldimines
are bent and the absorption maxima are at approxi-
mately 360 nm [25–27]. Hence, extension of the sub-
The reaction of wild-type TIL with homo-BZI-Ala
shows formation of an intermediate absorbing at
340 nm within the dead time of the stopped-flow
instrument. This is likely to be a gem-diamine, based
on its rapid rate of formation and the position of the
Scheme 4. Kinetic mechanism for reaction of TIL with analogs.
FEBS Journal (2013) ª 2013 The Authors Journal compilation ª 2013 FEBS
6

A. P. Harris and R. S. Phillips
Benzimidazoles with tryptophan indole lyase
strate structure in homoBZI-Ala by an additional car-
bon causes strain in the external aldimine C=N bond.
Because neither formation nor decay of A is observa-
ble as a separate phase in the reaction, we conclude
that the rate constant, k2, for this step is slow. As fast
as it forms, A equilibrates with a quinonoid species
(Q) at a rate, k3, that is faster than k2. These rate con-
stants suggest that k2 is the rate-limiting step between
conversion of the gem-diamine to the quinonoid, and
thus is the only rate observed in the inter-conversion
process. Hence, we propose that the steps G?A?Q
are observed as one step by the global fit, and corre-
spond to the observed rate constant 3.0 9 10 sꢀ1. The
amplitude of this phase is relatively low, corresponding
to > 10% of the absorbance increase at 494 nm
(Fig. 2C, dashed line).
a-aminoacrylate [13,16]. As further homologation gave
a compound, bishomo-BZI-Ala, with no observable
inhibition at 600 lM, its binding must be much
weaker, possibly due to the steric restrictions of the
active site.
Mechanistic implications
Based on the similarities between BZI-Ala and L-Trp,
two mechanisms may be hypothesized for the BZI-Ala
reaction with TIL. One mechanism is similar to that of
L-Trp shown in Scheme 1. BZI-Ala first reacts with
TIL to form an external aldimine. The aldimine is then
deprotonated, leading to a quinonoid intermediate.
The quinonoid intermediate is then protonated by
Tyr74 at the proximal nitrogen of the benzimidazole,
forming an aza-indolenine, which then collapses to
form benzimidazole and the aminoacrylate. However,
this mechanism is not consistent with the rapid-scan-
ning stopped-flow data, which do not show accumula-
tion of a quinonoid intermediate during catalysis.
In Scheme 5, an alternative mechanism is proposed
for the reaction of BZI-Ala with TIL. As in Scheme 1,
the first step in the mechanism is formation of an
external aldimine of the Trp analog. The external aldi-
mine is then deprotonated to form a quinonoid inter-
mediate concomitant with protonation of the distal
nitrogen of the benzimidazole ring by an acid and
b-elimination. Thus, for this substrate, the mechanism
is apparently changed from an E1Cb-type, for L-Trp,
to an E2-type. The accumulation of the quinonoid
intermediate in the H463F TIL reaction with BZI-Ala
(Fig. 1) is consistent with proton transfer to the distal
nitrogen of the benzimidazole ring of this substrate by
His463 in wild-type TIL, thus changing the mechanism
back to an E1Cb-type, due to slow protonation of the
distal nitrogen of the benzimidazole, either by solvent
or Tyr74.
There is a second slower phase that also corresponds
to a decrease at 340 nm and an increase at 494 nm,
but with a significantly larger amplitude (Fig. 2C,
dashed/dotted line). The observed rate constant for
this step is 4.5 9 10ꢀ1 ꢀ1. This may be the result of
s
either two forms of the gem-diamine, G and G*, which
react at different rates to give Q, or, alternatively, the
two steps, G?A and A?Q are kinetically coupled,
and the two relaxations result from this coupling. If
the former is the case, the predominant form of the
gem-diamine is G*. Finally, after Q forms, there is an
additional very slow process of increasing absorbance
at 494 nm, with a relatively large amplitude (Fig. 2C,
dotted line). It is likely that this results from Q under-
going a conformational change to a structurally differ-
ent quinonoid species (Q*) at a very slow rate
(k4 = 8.0 9 10ꢀ2 ꢀ1).
s
Homo-BZI-Ala is a relatively potent competitive
inhibitor of TIL, with a Ki of 13.4 lM. Only oxindolyl-
L-alanine (Ki = 5 lM [17,28]) and 2,3-dihydro-L-Trp
(Ki = 2 lM [17,18]) are better inhibitors. The strong
binding of this homotryptophan analog, together with
the slow kinetics, suggest that it may force the inhibi-
tor to adopt a folded conformation resembling the
product complex after the Cb–Cc bond cleavage. The
additional carbon atom in the side chain may allow
the benzimidazole ring to bind at the site where the
nascent indole binds upon C–C bond cleavage. Stereo-
electronic considerations suggest that the product
indole ring is positioned over the a-aminoacrylate,
similar to the Ala complex with pyridine-N-oxide in
the homologous enzyme tyrosine phenol lyase [29].
Thus, it is possible that this inhibitor binds as a
‘bi-product’ analog that mimics the indole-a-aminoac-
rylate product complex (Scheme 1). In this regard, it
should be noted that benzimidazole is an uncompeti-
tive inhibitor of TIL, binding to and stabilizing the
Conclusions
This study was performed in order to determine
whether benzimidazole analogs of L-Trp may serve as
substrates or inhibitors of TIL. Steady-state kinetic
testing revealed BZI-Ala to be a substrate of TIL with
a catalytic efficiency similar to that of wild-type TIL
for L-Trp and much better than that of H463F TIL for
L-Trp. We next hypothesized that, by increasing the
amino acid side chain of BZI-Ala by one methylene
group to homo-BZI-Ala, we would be able to produce
a benzimidazole analog of L-Trp, which functioned as a
more potent inhibitor of Trp than BZI-Ala. The
steady-state kinetic data revealed that homo-BZI-Ala is
FEBS Journal (2013) ª 2013 The Authors Journal compilation ª 2013 FEBS
7

Benzimidazoles with tryptophan indole lyase
A. P. Harris and R. S. Phillips
Scheme 5. Proposed mechanism for
reaction of BZI-Ala with TIL.
a competitive inhibitor of TIL, with a Ki of 13.4 lM,
and is one of the most potent competitive inhibitors
known for TIL. Pre-steady-state kinetics studies of
homo-BZI-Ala with wild-type TIL revealed that the
reaction rapidly forms a gem-diamine, which slowly
inter-converts to a quinonoid intermediate. Finally, we
increased the amino acid side chain by an additional
methylene group, generating bishomo-BZI-Ala. Steady-
state kinetics studies revealed that this compound was
not an inhibitor of TIL at 600 lM.
Synthesis
3-(benzimidazol-1-yl)-L-alanine (BZI-Ala)
Na-benzyloxycarbonyl-L-asparagine (1a) (US Biochemical
Corp.) (1 g, 3.75 mmol) and 50 mL tetrahydrofuran were
combined in a 100 mL round-bottomed flask. While stir-
ring, the solution was cooled to 10 °C using an ice bath.
Iodobenzene diacetate (Acros) (1.6 g, 4.88 mmol) was
added to the tetrahydrofuran solution, and the reaction
was stirred at room temperature for 8 h, yielding a white
precipitate. The precipitate was filtered, and the precipitate
was washed twice with tetrahydrofuran, yielding 0.625 g
(70%) of Na-benzyloxycarbonyl-b-amino-L-alanine (2a)
[21].
Experimental procedures
Materials
The product 2a (0.625 g, 2.63 mmol) and 25 mL DMF
were combined in a 50 mL round-bottomed flask together
with 0.797 g (7.89 mmol) triethylamine. While stirring,
0.371 g (2.63 mmol) 2-fluoronitrobenzene (Acros) was
added dropwise to the reaction mixture. The reaction
mixture was then stirred at 100 °C for 48 h, cooled to
room temperature, and 100 mL water was added to the
solution. The pH of the solution was adjusted to 2.0 with
Lactate dehydrogenase (from rabbit muscle lyophilized) and
NADH disodium salt were obtained from US Biochemical
Corp. (Cleveland, OH, USA). S-(o-Nitrophenyl)-L-cysteine
for enzyme assays was prepared from L-cysteine and 2-flu-
oronitrobenzene as previously described [30–32]. Buffers
and inorganic reagents were products of Fisher Scientific
Co. (Norcross, GA, USA)
8
FEBS Journal (2013) ª 2013 The Authors Journal compilation ª 2013 FEBS

A. P. Harris and R. S. Phillips
Benzimidazoles with tryptophan indole lyase
1 N HCl, and the reaction mixture was extracted with three
15 mL portions of ethyl acetate. The ethyl acetate extrac-
tions were then combined and dried over sodium sulfate.
The ethyl acetate was next removed, under reduced pres-
sure, at 60 °C, yielding an orange oil: 2-amino-3-(2′-nitro-
phenylamino)propanoic acid (3a) (0.360 g, 60.8%).
Following reflux, the reaction mixture was cooled and fil-
tered. The filtrate was then collected, and the formic acid
was removed in vacuo, yielding a clear oil that was then
dissolved in 6 N HCl and refluxed at 100 °C for 12 h. Fol-
lowing reflux, the solvent was evaporated in vacuo, yielding
homo-BZI-Ala dihydrochloride (4b), (0.510 g, 2.32 mmol,
91%) as a white crystalline solid. The MS (ESI) data are
m/z 220 (100, M + 1); 1H-NMR: d (D2O) 9.00 (Ar-H, 1H,
s), 7.60 (Ar-H, 1H, dd), 7.54 (Ar-H, 1H, dd), 7.36 (Ar-H,
2H, m), 4.60 (c-CH2, 2H, t), 3.88 (a-CH, 1H, dd), 2.35
(b-CH2, 2H, m).
Without further purification, 3a (0.360 g, 1.60 mmol)
was dissolved in 30 mL formic acid. Zinc dust (1.0 g)
was added to the solution, and the solution was refluxed
for 24 h. The reaction mixture was then cooled and
filtered. The filtrate containing the formic acid was then
collected, and the formic acid was removed in vacuo,
yielding a clear oil. The oil was then dissolved in 6 N
HCl and refluxed at 100 °C for 12 h. Following reflux,
the acid was removed in vacuo, producing 0.308 g (94%)
of BZI-Ala dihydrochloride (4a) as a white crystalline
solid. The MS (ESI) data are m/z 206 (100, M + 1);
1H-NMR: d (D2O) 9.19 (Ar-H, 1H, s), 7.79 (Ar-H, 1H,
dd), 7.74 (Ar-H, 1H, dd), 7.56 (Ar-H, 2H, m), 4.94
(b-CH2, 2H, t), 4.34 (a-CH, 1H, t).
2-amino-5-(benzimidazol-1-yl)pentanoic acid
(bishomo-BZI-Ala)
Na-Benzyloxycarbonyl-Nd-tert-butoxycarbonyl-L-ornithine
[22] (3.0 g, 8.57 mmol) was dissolved in a mixture of
50 mL methylene chloride and 10 mL trifluoroacetic acid.
The solution was stirred for 30 min, and the solvent was
removed in vacuo. The resulting clear oil was then dis-
solved in 50 mL H2O and purified using a DOWEX-50
(The Dow Chemical Co., Midland, MI, USA) cation
exchange column, yielding 2b (1.98 g, 7.92 mmol, 92%).
The product 2b (1 g, 4.00 mmol) and 25 mL DMF
2-amino-4-(benzimidazol-1-yl)-propionic acid
(homo-BZI-Ala)
Na-Benzyloxycarbonyl-L-glutamine (1b) (Sigma-Aldrich
Chemical Co., St Louis, MO, USA) (2.0 g, 7.14 mmol) and
50 mL THF were combined in a 100 mL round-bottomed
flask. While stirring, the solution was cooled to 10 °C using
an ice bath. Iodobenzene diacetate (3.05 g, 9.29 mmol) was
then added to the THF solution, and the reaction was stir-
red at room temperature for 8 h. The THF was removed
under reduced pressure. The resulting clear oil was then
dissolved in water and extracted three times with chloro-
form. The chloroform extract was discarded, and the water
from the aqueous layer was then removed under reduced
pressure, yielding 1.5 g (57.6%) of a light brown solid: 2-
(benzyloxycarbonylamino)-4-amino
were combined in
a
50 mL round-bottomed flask
together with 1.21 g (12.00 mmol) triethylamine. While
stirring, 2-fluoronitrobenzene (0.564 g, 4.00 mmol) was
added dropwise to the reaction mixture. The reaction
mixture was stirred at 100 °C for 48 h. The reaction mix-
ture was then allowed to cool to room temperature, and
100 mL water was added to the solution. The pH of the
solution was adjusted to 2.0 with 1 N HCl, and the reac-
tion mixture was extracted three times with 15 mL por-
tions of ethyl acetate. The ethyl acetate extractions were
combined and dried over sodium sulfate. The ethyl ace-
tate was next removed in vacuo, yielding Na-Cbz-5-(2-ni-
trophenylamino)-L-ornithine (3c) (0.804 g, 52%) as an
orange oil.
butanoic acid (2b).
The product 2b (1.5 g, 5.95 mmol) and 25 mL DMF
were combined in a 50 mL round-bottomed flask together
with triethylamine (1.8 g, 17.85 mmol). While stirring,
2-fluoronitrobenzene, (0.839 g, 5.95 mmol) was added drop-
wise to the reaction mixture. The reaction mixture was stirred
at 100 °C for 48 h. The reaction mixture was then allowed
to cool to room temperature, and 100 mL water was
added. The pH of the solution was adjusted to 2.0 with
1 N HCl, and the reaction mixture was extracted three
times with 15 mL portions of ethyl acetate. The combined
ethyl acetate layers were dried over sodium sulfate, and the
solvent was then removed in vacuo, yielding 2-(benzyloxy-
carbonylamino)-4-(2′-nitrophenylamino)butanoic acid (3b)
(0.954 g, 43%) as an orange oil.
Without further purification, 3c (0.804 g, 2.08 mmol) was
dissolved in 30 mL formic acid and combined with 1.0 g
Zn dust. The reaction mixture was refluxed by heating at
101ꢀC for 24 h. Following reflux, the reaction mixture was
cooled and filtered. The filtrate was then collected, and the
formic acid was removed in vacuo, yielding a brown oil. The
crude oil was then dissolved in 6 N HCl and refluxed at
100 °C for 12 h. Following reflux, the solvent was removed
in vacuo, producing a black solid, which was purified on a
C18 reversed-phase column (Sorbent Technologies, Inc.,
Norcross, GA) using a 20% aqueous methanol mobile
phase. Upon removal of the mobile phase, 2-amino-5-(ben-
zimidazol-1-yl)pentanoic acid (bishomo-BZI-Ala, 4c) was
obtained (0.001 g, 0.004 mmol, 0.2%). The 1H-NMR data
are d, D2O, 8.00 (Ar-H, 1H, s), 7.55 (Ar-H, 1H, d), 7.44
(Ar-H, 1H, d), 7.19 (Ar-H, 2H, m), 4.10 (C-H, 2H, t), 3.02
(C-H, 1H, t), 1.73 (C-H, 2H, m), 1.35 (C-H, 2H, m).
Without further purification, 3b (0.954 g, 2.56 mmol)
was dissolved in 30 mL formic acid. Zinc dust (1.0 g) was
added to the flask, and the solution was refluxed for 24 h.
FEBS Journal (2013) ª 2013 The Authors Journal compilation ª 2013 FEBS
9

Benzimidazoles with tryptophan indole lyase
A. P. Harris and R. S. Phillips
Enzyme
References
Tryptophan indole lyase was purified from cells of E. coli
JM101 containing plasmid pMD6, with the E. coli tnaA
gene under natural regulation, as described previously [33],
except that the Sepharose CL-4B (GE Healthcare
Biosciences, Pittsburgh, PA, USA) treatment was per-
formed in a column rather than batchwise. H463F TIL was
prepared as described previously [19]. Enzyme concentra-
tions were estimated from the absorbance of the holoen-
zyme at 278 nm (A1% = 9.19) [31] using a subunit molar
mass of 52 kDa [34].
1 Martin K, Morlin G, Smith A, Nordyke A,
Eisenstark A & Golomb M (1998) The tryptophanase
gene cluster of Haemophilus influenzae type b:
evidence for horizontal gene transfer. J Bacteriol 180,
107–118.
2 Anyanful A, Dolan-Livengood JM, Lewis T, Sheth S,
Dezalia MN, Sherman MA, Kalman LV, Benian GM &
Kalman D (2005) Paralysis and killing of
Caenorhabditis elegans by enteropathogenic Escherichia
coli requires the bacterial tryptophanase gene. Mol
Microbiol 57, 988–1007.
3 Snell EE (1975) Tryptophanase: structure, catalytic
activities, and mechanism of action. Adv Enzymol Relat
Areas Mol Biol 42, 287–333.
Steady-state kinetic measurements
Steady-state kinetic measurements were performed at 25 °C
as described previously [28] using a Cary 1 UV/Vis spectro-
photometer (Agilent Technologies, Santa Clara, CA, USA)
equipped with a Peltier thermoelectric cell block for temper-
ature control. The activities of substrates were measured
using the lactate dehydrogenase-coupled assay [35], and
the decrease in absorbance was followed at 340 nm
4 Wang D, Ding X & Rather PN (2001) Indole can act as
an extracellular signal in Escherichia coli. J Bacteriol
183, 4210–4216.
5 Di Martino P, Merieau A, Phillips R, Orange N &
Hulen C (2002) Isolation of an Escherichia coli strain
mutant unable to form biofilm on polystyrene and to
adhere to human pneumocyte cells: involvement of
tryptophanase. Can J Microbiol 48, 132–137.
6 Di Martino P, Fursy R, Bret L, Sundararaju B &
Phillips RS (2003) Indole can act as an extracellular
signal to regulate biofilm formation in Escherichia coli
and in other indole-producing bacteria. Can J Microbiol
49, 443–449.
ꢀ1
(De = ꢀ6220 M ꢂcmꢀ1) in 0.1 M potassium phosphate, pH
8.0, at 25 °C. Inhibition activity assays were performed with
S-(o-Nitrophenyl)-L-cysteine as the substrate in 0.1 M potas-
sium phosphate, pH 8.0, at 25 °C, and the decrꢀea1se in absor-
bance was followed at 370 nm (De = ꢀ1860 M ꢂcmꢀ1) [36].
Steady-state kinetic data were analyzed using the FOR-
TRAN programs of Cleland [37], HYPERO and COMPO,
to fit the data to Eqns (2) and (3), respectively:
7 Lee HH, Molla MN, Cantor CR & Collins JJ (2010)
Bacterial charity work leads to population-wide
resistance. Nature 467, 82–85.
V ¼ Vmax ꢃ ½Sꢄ=Km þ ½Sꢄ
ð2Þ
ð3Þ
8 Hirakawa H, Hayashi-Nishino M, Yamaguchi A &
Nishino K (2010) Indole enhances acid resistance in
Escherichia coli. Microb Pathog 49, 90–94.
V ¼ Vmax ꢃ ½Sꢄ=Kmð1 þ ½Iꢄ=KiÞ þ ½Sꢄ
9 Field CM & Summers DK (2012) Indole inhibition of
ColE1 replication contributes to stable plasmid
maintenance. Plasmid 67, 88–94.
Pre-steady-state kinetic measurements
Rapid scanning stopped-flow kinetic experiments were per-
formed using an RSM-1000 spectrophotometer and
stopped-flow compartment with 10 mm path-length
10 Roy M, Miles EW, Phillips RS & Dunn MF (1988)
Detection and identification of intermediates in the
reactions of tryptophan synthase with oxindolyl-L-
alanine and 2,3-dihydro-L-tryptophan via rapid-
scanning and single wavelength stopped-flow studies.
Biochemistry 27, 8661–8669.
a
a
observation cell from OLIS Inc. (Bogart, GA, USA). Rapid
scanning measurements were performed at 1 kHz. Prior to
the rapid kinetics experiments, the stock enzyme was sepa-
rated from excess PLP on a short desalting column (PD-10;
GE Healthcare Biosciences, Pittsburg, PA, USA) equili-
brated with 0.020 M potassium phosphate, pH 8.0, and
0.16 M KCl, and reactions were performed in the same buf-
fer. The stopped-flow kinetic measurements were performed
at room temperature. Generally, the enzyme solutions were
mixed with the solutions of the corresponding tryptophan
analog in the same buffer. In control experiments, the
enzyme was mixed with buffer without the tryptophan ana-
log. The data obtained from the scans were fitted by SVD
analysis using OLIS GLOBALWORKS software [38].
11 Phillips RS (1989) The mechanism of tryptophan
indole-lyase: insights from pre-steady-state kinetics and
substrate and solvent isotope effects. J Am Chem Soc
111, 727–730.
12 Phillips RS, Bender SL, Brzovic P & Dunn MF (1990)
The mechanism of binding of substrate analogues to
tryptophan indole-lyase: studies using rapid-scanning
and single wavelength stopped-flow spectrophotometry.
Biochemistry 29, 8608–8614.
13 Phillips RS (1991) The reaction of indole and
benzimidazole with amino acid complexes of E. coli
10
FEBS Journal (2013) ª 2013 The Authors Journal compilation ª 2013 FEBS

A. P. Harris and R. S. Phillips
Benzimidazoles with tryptophan indole lyase
tryptophan indole-lyase: detection of a new reaction
intermediate. Biochemistry 30, 5927–5934.
important than electrostatic interaction in controlling
the pKa of the catalytic group in aspartate
14 Lee M & Phillips RS (1995) The mechanism of
Escherichia coli tryptophan indole-lyase: substituent
effects on steady-state and pre-steady-state kinetic
parameters for aryl-substituted tryptophan derivatives.
Bioorg Med Chem 3, 195–205.
aminotransferase. Biochemistry 40, 353–360.
26 Lima S, Sundararaju B, Huang C, Khristoforov R,
Momany C & Phillips RS (2009) The crystal structure
of the Pseudomonas dacunhae aspartate-b–decarboxylase
dodecamer reveals an unknown oligomeric assembly
for a pyridoxal-5′–phosphate-dependent enzyme. J Mol
Biol 388, 98–108.
15 Sloan MS & Phillips RS (1996) Effects of a-deuteration
and of aza and thia analogs of L-tryptophan on
formation of intermediates in the reaction of
Escherichia coli tryptophan indole-lyase. Biochemistry
35, 16165–16173.
27 Phillips RS, Lima S, Khristoforov R & Sudararaju B
(2010) Insights into the mechanism of Pseudomonas
dacunhae aspartate b-decarboxylase from rapid-scanning
stopped-flow kinetics. Biochemistry 49, 5066–5073.
28 Kiick DM & Phillips RS (1988) Mechanistic
deductions from multiple kinetic and solvent
deuterium isotope effects and pH studies of pyridoxal
phosphate dependent carbon–carbon lyases:
Escherichia coli tryptophan indole-lyase. Biochemistry
27, 7339–7344.
16 Phillips RS, Sundararju B & Faleev NG (2000) Proton
transfer and carbon–carbon bond cleavage in the
elimination of indole catalyzed by Escherichia coli
tryptophan indole-lyase. J Am Chem Soc 122,
1108–1114.
17 Phillips RS, Miles EW & Cohen LA (1984)
Interactions of tryptophan synthase, tryptophanase
and pyridoxal phosphate with oxindolyl-L-alanine and
2,3-dihydro-L-tryptophan: support for an indolenine
intermediate in tryptophan metabolism. Biochemistry
23, 6228–6234.
ꢀ
29 Milic D, Demidkina TV, Faleev NG, Phillips RS,
ꢁ
ꢀ
ꢀ
Matkovic-Calogovic D & Antson AA (2011)
Crystallographic snapshots of tyrosine phenol-lyase
show that substrate strain plays a role in C-C bond
cleavage. J Am Chem Soc 133, 16468–16476.
30 Phillips RS, Ravichandran K & Von Tersch RL (1989)
Synthesis of L-tyrosine from phenol and
18 Phillips RS, Miles EW & Cohen LA (1985) Differential
inhibition of tryptophan synthase and of tryptophanase
by the two diastereoisomers of 2,3-dihydro-L-
tryptophan: implications for the stereochemistry of the
reaction intermediates. J Biol Chem 260, 14665–14670.
19 Phillips RS, Johnson N & Kamath AV (2002)
Formation in vitro of hybrid dimers of H463F and
Y74F mutant Escherichia coli tryptophan indole-lyase
rescues activity with L-tryptophan. Biochemistry 41,
4012–4019.
S-(o-nitrophenyl)-L-cysteine catalyzed by tyrosine
phenol-lyase. Enzyme Microb Technol 11, 80–83.
31 Dua RK, Taylor EW & Phillips RS (1993) S–aryl-l–
cysteine S, S–dioxides: design and evaluation of a new
class of mechanism based inhibitors of kynureninase.
J Am Chem Soc 115, 1264–1270.
32 Heiss C, Anderson J & Phillips RS (2003) Differential
effects of bromination on substrates and inhibitors of
kynureninase from Pseudomonas fluorescens. Org
Biomol Chem 1, 288–295.
20 Demidkina TV, Zakomirdina LN, Kulikova VV,
Dementieva IS, Faleev NG, Ronda L, Mozzarelli A,
Gollnick PD & Phillips RS (2003) Role of aspartate-
133 and histidine-458 in the mechanism of tryptophan
indole-lyase from Proteus vulgaris. Biochemistry 42,
11161–11169.
33 Phillips RS & Gollnick PD (1989) Evidence that
cysteine-298 is in the active site of tryptophan indole-
lyase. J Biol Chem 254, 10627–10632.
21 Zhang L-H, Kauffman GS, Pesti JA & Yin J (1997)
Rearrangement of Na-protected L-asparagines with
iodosobenzene diacetate. A practical route to b-amino-
L-alanine derivatives. J Org Chem 62, 6918–6920.
22 Wiejak S, Masiukiewicz E & Rzeszotarska B (2001)
Improved scalable syntheses of mono- and bis-urethane
derivatives of ornithine. Chem Pharm Bull (Tokyo) 49,
1189–1191.
34 Kagamiyama H, Wada H, Matsubara H & Snell EE
(1972) The chemical structure of tryptophanase from
Escherichia coli. II. The structure of tryptophanase
monomer. J Biol Chem 247, 1571–1575.
35 Morino Y & Snell EE (1970) Tryptophanase
(Escherichia coli B). Methods Enzymol 17A, 439–446.
36 Suelter CH, Wang J & Snell EE (1976) Direct
spectrophotometric assay of tryptophanase.
23 Mamalis P, Petrow V & Sturgeon B (1950) Some
benziminazolylalanines. J Chem Soc 1950, 1600–1603.
24 Thibodeaux CJ & Liu HW (2011) Mechanistic studies
of 1-aminocyclopropane-1-carboxylate deaminase:
characterization of an unusual pyridoxal 5′-phosphate-
dependent reaction. Biochemistry 50, 1950–1962.
25 Mizuguchi H, Hayashi H, Okada K, Miyahara I,
Hirotsu K & Kagamiyama H (2001) Strain is more
FEBS Lett 66, 230–232.
37 Cleland WW (1979) Statistical analysis of enzyme
kinetic data. Methods Enzymol 63, 103–138.
38 Matheson IBC (1987) The method of successive
integration: a general technique for recasting kinetic
equations in a readily soluble form which is linear in
the coefficients and sufficiently rapid for real time
instrumental use. Anal Instrum 16, 345–373.
FEBS Journal (2013) ª 2013 The Authors Journal compilation ª 2013 FEBS
11