Inhibition of Escherichia coli tryptophan indole-lyase by tryptophan homologues

Article abstract of DOI:10.1016/j.abb.2014.07.027

We have designed, synthesized and evaluated homotryptophan analogues as possible mechanism-based inhibitors for Escherichia coli tryptophan indole-lyase (tryptophanase, TIL, E.C. 4.1.99.1). As a quinonoid structure is an intermediate in the reaction mechanism of TIL, we anticipated that homologation of the physiological substrate, l-Trp would provide analogues resembling the transition state for β-elimination, and potentially inhibit TIL. Our results demonstrate that l-homotryptophan (1a) is a moderate competitive inhibitor of TIL, with K_i = 67 μM, whereas l-bishomotryptophan (1b) displays more potent inhibition, with K_i = 4.7 μM. Pre-steady-state kinetics indicated the formation of an external aldimine and quinonoid with 1a, but only the formation of an external aldimine for 1b, suggesting differences in the inhibition mechanism. These results demonstrate that formation of a quinonoid complex is not required for strong inhibition. In addition, the Trp analogues were evaluated as inhibitors of Salmonella typhimurium Trp synthase. Our results indicate that compound 1b is at least 25-fold more selective toward TIL than Trp synthase. We report that compound 1b is comparable to the most potent inhibitor previously reported, while displaying high selectivity for TIL. Thus, 1b is a potential lead for the development of novel antibacterials.

Full text of DOI:10.1016/j.abb.2014.07.027

Archives of Biochemistry and Biophysics 560 (2014) 20–26
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics
journal homepage: www.elsevier.com/locate/yabbi
Inhibition of Escherichia coli tryptophan indole-lyase by tryptophan
homologues
a
a
a
a,b,
⇑
Quang T. Do , Giang T. Nguyen , Victor Celis , Robert S. Phillips
a
Department of Chemistry, University of Georgia, Athens, GA 30602, USA
Department of Biochemistry and Molecular Biology, University of Georgia, Athens, GA 30602, USA
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 23 May 2014
and in revised form 14 July 2014
Available online 1 August 2014
We have designed, synthesized and evaluated homotryptophan analogues as possible mechanism-based
inhibitors for Escherichia coli tryptophan indole-lyase (tryptophanase, TIL, E.C. 4.1.99.1). As a quinonoid
structure is an intermediate in the reaction mechanism of TIL, we anticipated that homologation of the
physiological substrate,
and potentially inhibit TIL. Our results demonstrate that
itive inhibitor of TIL, with K = 67 M, whereas -bishomotryptophan (1b) displays more potent inhibi-
tion, with K = 4.7 M. Pre-steady-state kinetics indicated the formation of an external aldimine and
L-Trp would provide analogues resembling the transition state for b-elimination,
L
-homotryptophan (1a) is a moderate compet-
Keywords:
i
l
L
L-Homotryptophan
i
l
L
-Bishomotryptophan
0
quinonoid with 1a, but only the formation of an external aldimine for 1b, suggesting differences in the
inhibition mechanism. These results demonstrate that formation of a quinonoid complex is not required
for strong inhibition. In addition, the Trp analogues were evaluated as inhibitors of Salmonella typhimu-
rium Trp synthase. Our results indicate that compound 1b is at least 25-fold more selective toward TIL
than Trp synthase. We report that compound 1b is comparable to the most potent inhibitor previously
reported, while displaying high selectivity for TIL. Thus, 1b is a potential lead for the development of
novel antibacterials.
Pyridoxal-5 -phosphate
Steady-state kinetics
Pre-steady-state kinetics
Ó 2014 Elsevier Inc. All rights reserved.
1
Tryptophan indole-lyase (tryptophanase, TIL, E.C. 4.1.99.1) is a
plasmid stability [6], pathogenicity [7], and antibiotic resistance
[8,9]. Since TIL is a bacterial enzyme not found in eukaryotes, inhi-
bition of its activity presents a selective and attractive approach for
an antibiofilm and antibacterial treatment. Thus, potent inhibitors
of TIL are of interest as possible antibiotics.
0
pyridoxal-5 -phosphate (PLP)-dependent bacterial enzyme that cata-
lyzes the reversible hydrolytic cleavage of the C AC bond of L-Trp to
b c
yield indole and ammonium pyruvate (Eq. (1)) [1]. For many years,
this source of indole was considered to be a bacterial metabolic
waste product of L-Trp. However, recent reports have suggested that
metabolic indole has a critical role in formation of biofilm in
In addition to
tion of a series of substrates, including
S-alkyl- -serines, b-chloroalanine, 2,3-diaminopropionate, O-acyl-
-serines, and S-(o-nitrophenyl)- -cysteine (SOPC), since all of these
L-Trp, TIL can also catalyze the in vitro b-elimina-
L
-serine, O-alkyl- -serines,
L
L
L
L
substrates have a reasonable leaving group suitable for b-elimina-
tion [10–12]. However, the cleavage mechanism for its physiolog-
ical substrate,
L-Trp, is intriguing as indole is formally an
ð1Þ
unactivated aromatic carbon leaving group in this reaction. Never-
theless, previous work from our laboratory [13–18] and that of
others [19,20] has provided valuable insights into the catalytic
mechanism of TIL (Scheme 1).
Escherichia coli [2–4]. In addition, it has been shown that indole is a
bacterial signaling molecule that regulates gene expression [5],
The active form of E. coli TIL is a tetramer with PLP bound to Lys-
270 in the active site through a Schiff’s base linkage [21]. Nucleo-
philic attack by the amino acid ligand generates a gem-diamine,
which releases the Lys-270 and forms a substrate-bound external
aldimine, observed at kmax = 422 nm (Scheme 1) [14]. Previously,
⇑
Corresponding author at: Department of Chemistry, University of Georgia,
Athens, GA 30602, USA. Fax: +1 706 542 9454.
E-mail address: plp@uga.edu (R.S. Phillips).
1
0
Abbreviations used: PLP, pyridoxal-5 -phosphate; SOPC, S-(o-nitrophenyl)-L-cys-
0
teine; DEAM, diethyl acetamidomalonate; L-Homotryptophan, (2S)-4-(3 -indolyl)-2-
aminobutanoic acid; L-Bishomotryptophan, (2S)-5-(3 -indolyl)-2-aminopentanoic
acid; TIL, tryptophan indole-lyase [tryptophanase, EC 4.1.99.1].
it was suggested that an indolenine tautomer of
L-Trp was an inter-
0
mediate in this reaction [22,23]. Further studies in our laboratory
http://dx.doi.org/10.1016/j.abb.2014.07.027
0
003-9861/Ó 2014 Elsevier Inc. All rights reserved.

Q.T. Do et al. / Archives of Biochemistry and Biophysics 560 (2014) 20–26
21
Scheme 1. Proposed mechanism of tryptophan indole-lyase.
bition and pre-steady-state kinetics evaluation of 1a and 1b as
suggested that following deprotonation of C
indole was accomplished by C protonation to form a quinonoid
intermediate, absorbing at 505 nm followed by C AC bond cleav-
a
, the departure of
inhibitors of TIL and Trp synthase.
c
b
c
age (Scheme 1) [15]. Decay of the quinonoid intermediate in the
presence of benzimidazole, an uncompetitive inhibitor, demon-
strated the presence of an aminoacrylate intermediate, with
Results and discussion
Synthesis of L-homotryptophan and L-bishomotryptophan
k
max = 345 nm (Scheme 1) [15]. The aminoacrylate intermediate
then releases the ammonium pyruvate product. The transition
state for indole elimination is expected to have an extended C AC
bond length. Hence, we proposed that -homotryptophan (1a) and
-bishomotryptophan (1b) (Scheme 2) might resemble the transi-
tion state for indole elimination, and thus could potentially act as
potent competitive inhibitors for TIL [24]. In addition, -Trp ana-
Syntheses of racemates of both 1a [26] and 1b [27] were previ-
ously reported, but to our knowledge, no attempts at optical reso-
lution of these compounds were reported in the literature. Elegant
asymmetric syntheses of 1a were also reported, using the Schol-
lkopf chiral auxiliary [28] and Larock’s heteroannulation [29].
However, both methods for asymmetric synthesis suffered from
lengthy procedures to introduce the chiral center. Following a
modified procedure, we were able to obtain, in three steps, the
racemic N-acetyl derivatives, 5a and 5b, which can be enzymati-
cally resolved conveniently to yield enantiomeric 1a or 1b
(Scheme 3). Tryptophol (2a) and homotryptophol (2b) were
selected as starting materials since they can be obtained from com-
b
c
L
L
L
logues that are potent inhibitors of TIL are also potent inhibitors
of Trp synthase, an enzyme widely distributed in plants, fungi
and bacteria that catalyzes the last two steps of
23,25]. Therefore, we also evaluated the inhibitory activity of 1a
and 1b toward Trp synthase. We present herein the synthesis, inhi-
L-Trp biosynthesis
[

2
2
Q.T. Do et al. / Archives of Biochemistry and Biophysics 560 (2014) 20–26
Scheme 2. Structures of Trp and AzaTrp homologues.
Scheme 3. Synthesis of
L-Trp homologues.
mercial sources. Bromination of compounds 2 using phosphorus
tribromide, in our hands, only gave modest yields, consistent with
previous reports for this reaction [27]. Alternatively, the Appel’s
salt is a milder brominating agent for primary alcohols [30]. When
bromination of 2 was carried out using the Appel’s salt, a clean
conversion with high yield was achieved. Alkylation of 3 using
knowledge, this is the first report of the resolution of 1a and 1b
using Aspergillus acylase. Our specific rotation for
phan (1a), +35.7°, is very similar to that reported previously for
-homotryptophan obtained by asymmetric synthesis, +37.5°
[28]. The resolution and specific rotation of -bishomotryptophan
L-homotrypto-
L
L
1
eq. of diethyl acetamidomalonate afforded 4 in modest yield
(1b) has not been reported previously, although the synthesis of
[
27]. Hydrolysis using 5 eq. of NaOH, as reported by Snyder [27],
DL-bishomotryptophan has been reported [26].
resulted in an excess amount of salt recovered during the workup
and did not provide the decarboxylated product 5. However, when
using 1.2 eq. of NaOH, ester hydrolysis and decarboxylation was
achieved in one step to yield the N-acetyl derivatives, 5. Enantiose-
lective hydrolysis by Aspergillus acylase under the conditions previ-
Enzyme inhibition
Our results indicate that 1a exhibits moderate competitive inhi-
bition against TIL with K = 67 ± 12 lM (Table 1). However, by add-
ing an additional methylene between the amino acid and the
indole ring, compound 1b displayed an increase in potency of more
than an order of magnitude as an inhibitor of TIL, with
i
ously reported for N-acetyl-a-amino acids [31] was difficult due to
low solubility of 5 in water. Alternatively, when this step was car-
ried out in phosphate buffer at pH 8, we found that compound 5
solubility was greatly enhanced. Hydrolysis of 5a and 5b under this
Ki = 4.7 ± 0.5
yl- -alanine and 2,3-dihydrotryptophan are potent competitive
inhibitors of both TIL (K = 2.5 and 4.5 M, respectively) and trypto-
phan synthase [24,25]. These compounds remain the most potent
lM (Table 1). Previously, we reported that 2-oxindol-
condition yielded the
L-enantiomers, 1a and 1b, which were conve-
L
niently recovered through simple filtration, as both product com-
pounds have very low solubility in phosphate buffer. To our
l
i

Q.T. Do et al. / Archives of Biochemistry and Biophysics 560 (2014) 20–26
23
Table 1
bance peak at kmax = 500 nm, was observed (Fig. 1A). This result is
consistent with the accepted mechanism for TIL reaction with -Trp
Competitive inhibitors of tryptophan indole-lyase.
L
Compound
K
i
(l
M)
shown in Scheme 1, in which the initial formation of an external
aldimine with kmax = 422 nm and subsequent formation of quino-
noid intermediate at kmax = 505 nm are observed. The reaction
200a
L
-Tryptophan
3S)-2,3-Dihydro-
Oxindolyl- -alanine
b
(
L-Trp
2
c
L
5
requires 3 exponentials to obtain an adequate fit. In the fast phase,
13.6d
ꢀ1
(
(
2S)-2-Amino-4-(benzimidazol-1-yl)butanoic acid (HomoBZI-Ala)
2S)-2-Amino-5-(benzimidazol-1-yl)pentanoic acid (BishomoBZI-
Ala)
with 1/s
1
= 1.29 ± 0.02 s , there is a decrease in the 420 nm peak
>600d
and an increase at 500 nm, with an isosbestic point at 438 nm
ꢀ1
(
Fig. 1B). In the second phase, with 1/
s
2
= 0.67 ± 0.12 s , and the
L
L
-Homotryptophan (1a)
67 ± 12
ꢀ
1
third phase, with 1/
3
s = 0.06 ± 0.01 s , there is an increase at
-Bishomotryptophan (1b)
4.7 ± 0.5
5
00 nm, but little change in the absorbance at 420 nm, and no isos-
a
b
c
K
m
, from Ref. [12].
bestic point. Therefore, we conclude that compound 1a initially
complexes with PLP at the active site to form an external aldimine,
which occurs within the dead time of the stopped-flow device, fol-
From Ref. [36].
From Ref. [35].
From Ref. [33].
d
a
lowed by deprotonation of C to yield a quinonoid intermediate,
which then undergoes two subsequent slow conformational
changes (Eq. (2)) [14,15,18]. The rate constant for quinonoid inter-
inhibitors reported to date for TIL. Parola et al. also reported the
evaluation of several -Trp and anthraquinone derivatives as inhib-
itors for TIL, with moderate inhibition (K = 48–174 M) [32]. More
recently, our group reported the evaluation of several benzimid-
azole homologues of -Trp as inhibitors, with the most efficient
compound being (S)-4-(benzimidazol-1-yl)-2-aminobutanoic acid
Scheme 2), a benzimidazole derivative of homotryptophan, with
= 13 M [34]. To our knowledge, 1b is the only compound thus
far to achieve a K for TIL on the same order as 2-oxindolyl- -ala-
mediate formation from 1a is much slower than
which forms a 505 nm intermediate at about 500 s [14,15].
L
-tryptophan,
ꢀ
1
L
i
l
fast
E þ L-HTrp () E ꢀ L-HTrpð420nmÞ
L
ꢀ1
1:29 s
(
) E ꢀ L-HTrpð500nmÞ
ꢀ
1
(
K
0:67 s
(
) E ꢀ L-HTrpð500nmÞ
i
l
ꢀ1
0
:06 s
i
L
() E ꢀ L-HTrpð500nmÞ
ð2Þ
nine and 2,3-dihydrotryptophan. It is interesting to note that from
our previous work with benzimidazole derivatives [33], by replac-
With compound 1b, the absorbance peak at kmax = 422 nm is
observed, corresponding to the formation of an external aldimine
ing the C-3 carbon of
a substrate for TIL. By extending the C
lene, the aza-analogue of homotryptophan showed high efficiency,
= 13 M, as a competitive inhibitor (Table 1). However, further
L-Trp with a nitrogen, the compound was also
(Fig. 2A), but the 500 nm absorbance peak of the quinonoid inter-
a
position with one methy-
mediate is not seen. The course of the reaction can be fit with 2
ꢀ1
ꢀ1
1 2
exponentials, with 1/s = 16.1 ± 2.4 s and 1/s = 0.52 ± 0.09 s .
K
i
l
In the first phase, there is a red-shift to 422 nm and small increase
in intensity of the aldimine peak. In the second phase, there is a
further increase in the 422 nm aldimine peak and a very small
shoulder at about 500 nm, concomitant with a decrease in absor-
bance at 338 nm (Fig. 2B). Thus, it appears that the initial rapid
reaction of TIL with 1b forms an equilibrating mixture of gem-dia-
mine and external aldimine complexes, and a slow conformational
change results in conversion of the gem-diamine to the external
aldimine. In our previous study of benzimidazole aza-analogues
extension with a second methylene to the aza analogue of bishom-
otryptophan resulted in a great decrease in inhibitory potency.
Based on these results, our initial expectation was that 1a would
show greater inhibition than 1b. Unexpectedly, 1b was discovered
i
to have a K of more than 10-fold lower than 1a. As indole and
benzimidazole are isosteric and isoelectronic, this difference is
likely due to the differences in the structure of the N-3 nitrogen
in the two rings. In the tryptophan homologue series, the indole
nitrogen has a hydrogen on it, while in the azatryptophan homo-
logue series, the benzimidazole nitrogen is an unprotonated imine.
Thus, in the tryptophan homologue series, N-3 is an H-bond donor,
while in the azatryptophan homologues, N-3 is an H-bond acceptor
of L-Trp (Scheme 2), we found that the homologues also showed
evidence for gem-diamine intermediates, which are not seen in
the reaction of natural amino acids [33]. Furthermore, homophe-
nylalanine had also been found to exhibit a transient gem-diamine
intermediate in binding to TIL [34]. Thus, having 2 or more meth-
ylenes between the amino acid moiety and the aromatic ring must
introduce steric restrictions which slow down the formation of
external aldimines.
(Scheme 2).
Tryptophan synthase is also a PLP-dependent enzyme that
catalyzes the last two steps of
-Trp analogues that are potent inhibitors of TIL also show potent
inhibition of Trp synthase [24,25]. However, the diastereomers of
,3-dihydro- -tryptophan are stereoselective for inhibition of TIL
L-Trp biosynthesis. In our experience,
L
2
L
and Trp synthase. With regard to antibacterial development, an
ideal inhibitor is one which efficiently and effectively inhibits TIL
but not Trp synthase, as Trp synthase is widely distributed in
plants, fungi and bacteria. Assays for compounds 1a and 1b with
Trp synthase displayed no measurable inhibition at concentrations
Mechanistic implications
Interestingly, by extending
L-Trp by one additional methylene
at the C position (1a), the reaction is mechanistically similar to
a
that of the physiological substrate, but carbon–carbon cleavage
up to 100
lM, suggesting that both compounds are more selective
cannot occur. In compound 1a, the indole ring is positioned on
for TIL than Trp synthase, and compound 1b is at least 25-fold
more selective for TIL.
the
in the b-elimination of indole. In the crystal structure of the highly
homologous (ca. 40% identity) tyrosine phenol-lyase with -Ala and
b c
c-carbon prohibiting the participation of C AC p-electrons
L
Pre-steady-state kinetics
pyridine-N-oxide bound, which mimics the complex of aminoacry-
late and the phenol leaving group, there is a distance of 3.0 Å
Rapid-scanning stopped-flow experiments were conducted to
determine the inhibitor–enzyme binding interactions. Pre-steady-
state kinetics evaluation demonstrated significant differences in
the inhibition mechanism of both compounds. For 1a, an initial
absorbance peak at kmax = 420 nm, followed by a second absor-
between the C
matic ring [35]. Furthermore, the aromatic ligand is found to be
positioned directly over the C AC bond of the Ala. It is likely that
TIL would have a similar geometry for the elimination of indole
from -tryptophan. In the case of 1b, with the extension of -tryp-
b 4
of the Ala quinonoid complex and the C of the aro-
a
b
L
L

2
4
Q.T. Do et al. / Archives of Biochemistry and Biophysics 560 (2014) 20–26
Fig. 1. Rapid-scanning stopped-flow spectra for TIL reaction with
L
-homoTrp (1a). (A) The scans are shown at the following times after mixing: 1, 0.002 s; 2, 0.0345 s; 3,
0
.0875 s; 4, 2.81 s; 5, 7.60 s; 6, 15.9 s. (B) Time courses at 420 and 500 nm.
Fig. 2. Rapid-scanning stopped-flow spectra for TIL reaction with L-bishomoTrp (1b). (A) The scans are shown at the following times after mixing: 1, 0.002 s; 2, 0.046 s; 3,
0
.120 s; 4, 0.261 s; 5, 0.955 s; 6, 1.914 s. (B) Time courses at 338, 394 and 458 nm.
tophan by two additional methylenes, the compound apparently
must adopt a different conformation for the external aldimine, pos-
sibly with the indole ring of the inhibitor occupying a remote site
which the nascent indole product would occupy transiently on its
way out. This conformation must not allow the quinonoid interme-
so we were interested in the design of mechanism-based inhibitors
of TIL. As an alternative to asymmetric syntheses, we developed a
modified method of Snyder et al. to conveniently obtain the opti-
cally active isomer of compounds 1a and 1b. Evaluation of 1a in
inhibition assays indicated that it is a moderate inhibitor for TIL
a a b
diate to adopt the planar geometry of N AC AC necessary for
deprotonation to occur, which explains why the absorbance peak
at 500 nm for the quinonoid intermediate is negligible. Thus, it is
likely that 1b is a byproduct analogue of TIL. All of the other known
potent inhibitors of TIL form quinonoid complexes [24,25], but the
present results demonstrate that formation of a quinonoid com-
plex is not necessary for potent inhibition.
with K
was observed in pre-steady-state evaluation. In contrast, 1b dis-
played potent inhibition for TIL with K = 4.7 M and formation of
a quinonoid intermediate was not observed. No inhibition was
observed with Trp synthase for both compounds up to 100 M,
indicating higher selectivity for TIL in comparison with Trp syn-
thase. To our knowledge, -bishomotryptophan (1b) is the first
TIL inhibitor with efficiency on the same order as 2-oxindolyl-
alanine and 2,3-dihydro- -tryptophan while displaying high selec-
tivity for TIL in preference to Trp synthase.
i
= 67 lM and the presence of a quinonoid intermediate
i
l
l
L
Conclusion
L-
L
Recent reports suggest bacterial TIL is a novel attractive target
for development of novel antibiotics targeting biofilm formation,

Q.T. Do et al. / Archives of Biochemistry and Biophysics 560 (2014) 20–26
25
Materials and methods
7.30–7.32 (d, 1H), 7.48–7.49 (d, 1H), 8.22–8.24 (1H), 12.51 (broad,
H).
1
Materials
L
-Homotryptophan (1a)
Methylene chloride and ethanol were dried over CaH
2
and Mg,
In a 50 ml culture tube, compound 5a (150 mg) was dissolved in
respectively, prior to use. Tryptophol (Acros), homotryptophol
Acros), triphenylphosphine (Aldrich), carbon tetrabromide (Janssen
Chimica), diethyl acetamidomalonate ester (DEAM) (Aldrich),
potassium phosphate buffer (9 ml, 100 mM) and the final pH was
adjusted to 8 using 6 M NaOH. Then Aspergillus acylase I (45 mg)
was added to the mixture and it was allowed to incubate overnight
at 37 °C, with shaking speed of 250 rpm (C25, New Brunswick
Scientific). The resulting mixture was cooled, filtered and washed
with cold water to give pure 1a as a crystalline solid. No additional
product was observed after readjusting the pH of the filtrate to 6
using 6 M HCl. Attempts to readjust pH of the filtrate to 8, intro-
ducing additional enzyme and longer incubation time also did
not result in additional product. Yield: 49.2 mg (78%). Unreacted
5a can be recovered by acidifying the filtrate to pH 2 using 6 M
(
Aspergillus acylase
I
(0.43 unit/mg, Aldrich), indole (Aldrich),
0
L-serine (USB Corp.), pyridoxal-5 -phosphate (PLP) (USBiochemical
Corp.) and all other reagents (Fisher) were used without further
purification. S-(o-Nitrophenyl)- -cysteine (SOPC) used in enzyme
L
assays was prepared as previously described [36]. Enzyme assays
were performed using distilled deionized water. All NMR data were
collected on a 400 MHz Varian Mercury Plus NMR instrument and
data were processed by MNova NMR processing software. ESI-MS
experiments were performed on a Perkin Elmer Sciex API I Plus.
HCl and extraction using ethyl acetate. ESI-MS M+1 (m/z): 219.
1
H NMR (D
2
O/NaOD) d (ppm): 1.72–1.88 (m, 2H), 2.62–2.66 (t,
H), 3.12–3.16 (t, 1H), 6.97–7.01 (t, 1H), 7.06–7.09 (t, 2H), 7.32–
.34 (d, 1H), 7.53–7.55 (d, 1H). [ = +35.7° (Lit., +37.5° [38]) (CH3-
H, c = 0.28, 25.9 °C).
2
7
CO
Synthesis of L-homotryptophan and L-bishomotryptophan
a]
D
2
3
-(2-Bromoethyl)-1H-indole (3a)
In a three-neck flask, tryptophol (2.00 g, 1 eq.) and triphenyl-
phosphine (4.23 g, 1.3 eq.) was dissolved in dry CH Cl (15 ml).
In an addition funnel, carbon tetrabromide (5.35 g, 1.3 eq.) dis-
solved in dry CH Cl (5 ml) was added drop-wise, under inert
L
-Bishomotryptophan (1b)
Compound 1b was obtained following the procedure described
above for compound 1a. Starting with homotryptophol 2b, bromin-
ation using the Appel’s salt yielded 3b as a yellow oil (yield: 98%).
Alkylation of 3b with diethyl acetamidomalonate gave 4b as an oil
(yield: 58%). Ester hydrolysis and decarboxylation of 4b in aqueous
NaOH gave crude 5b as a yellow solid which upon purification by
recrystallization in water gave a white solid (yield: 55%). Enantio-
selective hydrolysis of 5b using Aspergillus acylase I, as described
above, gave 1b as a white crystalline solid (yield: 71%). 1H NMR
2
2
2
2
atmosphere at 0 °C, until the addition was complete. The reaction
was allowed to stir at room temperature for an additional 3 h, or
until complete disappearance of starting materials by TLC. Solvent
was removed under reduced pressure and the residue was purified
by column chromatography using hexane and ethyl acetate to yield
3
1
a as an off-white solid. Yield: 2.64 g (95%). (Appendix) H NMR
(D
3.10 (t, 1H), 6.96–7.00 (t, 1H), 7.04–7.09 (t, 2H), 7.32–7.34 (d,
1H), 7.53–7.55 (d, 1H). [ = +43.0° (CH CO H, c = 0.232, 26.3 °C).
2
O/NaOD) d (ppm): 1.45–1.57 (m, 4H), 2.61–2.64 (t, 2H), 3.07–
(
CDCl
3
) d (ppm): 3.32–3.36 (t, 2H), 3.62–3.66 (t, 2H), 7.10 (s, 1H),
7
.12–7.16 (t, 1H), 7.19–7.23 (t, 1H), 7.37–7.39 (d, 1H), 7.59–7.61
a
]
D
3
2
(
d, 1H), 8.04 (s, 1H).
Enzymes and enzyme assays
TIL was purified from E. coli JM101 containing plasmid pMD6, as
previously described [37]. Enzyme assays were performed on a
Cary 1 UV–visible spectrophotometer equipped with a Peltier tem-
perature-controlled six-cell changer. Enzyme activity was rou-
tinely determined as previously described [12], by following the
Diethyl 2-(2-(1H-indol-3-yl)ethyl)-2-acetamidomalonate (4a)
In a three-neck flask, diethyl acetamidomalonate (2.56 g, 1 eq.)
was added to a solution of dry ethanol (20 ml) containing dissolved
sodium metal (0.271 g, 1 eq.) at 0 °C. The mixture was stirred for an
additional 30 min at 0 °C. Compound 3a (2.64 g, 1 eq.) was then
added, and the solution was allowed to reflux under inert atmo-
sphere for an additional 15 h. Solvent was removed under reduced
pressure and the residue was purified by column chromatography
ꢀ1
ꢀ1
decrease in absorbance of SOPC at 370 nm (
in 50 mM phosphate buffer, pH 8.0, at 25 °C. Enzyme concentration
D
e
= ꢀ1860 M cm ),
in solution was estimated spectrophotometrically from the absor-
1%
bance of the holoenzyme at 278 nm (A = 9.19) [37]. Trp synthase
using hexane and ethyl acetate to yield 4a as a white solid. Yield:
from Salmonella typhimurium was purified as previously described
1
1
3
6
7
.69 g (40%). H NMR (CDCl
3
) d (ppm): 1.18–1.22 (t, 3H), 2.00 (s,
[
38]. The concentration of Trp synthase was determined from the
H), 2.63–2.67 (m, 2H), 2.79–2.83 (m, 2H), 4.07–4.21 (m, 4H),
.85 (s, 1H), 6.97 (s, 1H), 7.09–7.12 (t, 1H), 7.16–7.20 (t, 1H),
.33–7.35 (d, 1H), 7.54–7.56 (d, 1H), 8.00 (s, 1H).
1%
absorbance at 278 nm (A = 6.0), and enzyme activity were deter-
mined by following the increase in absorbance of -Trp at
k = 290 nm (De = 1800 M cm ) [38].
L
ꢀ1
ꢀ1
2
-Acetamido-4-(1H-indol-3-yl)butanoic acid (5a)
In a round bottom flask, NaOH (0.225 g, 1.2 eq.) was dissolved
Enzyme inhibition assays
A typical enzyme inhibition assay contained potassium phos-
phate (50 mM), PLP (40 M), with varying concentrations of SOPC,
1a or 1b in a total volume of 600 l. Similarly, Trp synthase assays
were performed in phosphate buffer (50 mM) containing PLP
(40 M), indole (0.1 mM), with varying concentration of serine,
1a or 1b in a total volume of 600 l. Experimental velocities from
inhibition assays were fitted to Eq. (3) and the inhibition constant,
, was calculated using the FORTRAN program, COMPO, of Cleland
[39].
in aqueous tetrahydrofuran (2:1, THF:H
2
O, 45 ml) solution. Then
l
compound 4a (1.69 g, 1 eq.) was added in and the solution was
allowed to reflux for 15 h. The solvent was then removed under
reduced pressure and the residue was taken up in water and ethyl
acetate. The aqueous layer was acidified to pH 2 using 6 M HCl and
extracted with ethyl acetate. The organic layer was washed with
l
l
l
water and dried over MgSO
4
. Removal of solvent under reduced
K
i
pressure give crude 5a as a yellow solid which can be further puri-
fied by recrystallization in water to give pure 5a as an off-white
needle-shaped solid. Yield: 0.684 g (56%). 1H NMR (DMSO) d
ꢁ
m
¼ Vmax ꢁ ½Sꢂ=ðK ð1 þ ½Iꢂ=K
i
Þ þ ½SꢂÞ
ð3Þ
m
(
4
ppm): 1.87 (s, 3H), 1.98–2.01 (m, 2H), 2.63–2.74 (m, 2H), 4.15–
.20 (m, 1H), 6.93–6.97 (t, 1H), 7.02–7.06 (t, 1H), 7.07 (s, 1H),

2
6
Q.T. Do et al. / Archives of Biochemistry and Biophysics 560 (2014) 20–26
[11] R.S. Phillips, Arch. Biochem. Biophys. 256 (1987) 302–310.
Rapid-scanning stopped-flow experiments
[
[
12] C.H. Suelter, J. Wang, E.E. Snell, FEBS Lett. 66 (1976) 230–232.
13] R.S. Phillips, T.V. Demidkina, N.G. Faleev, Biochim. Biophys. Acta 1647 (2003)
Prior to use, the enzyme was incubated with 0.5 mM PLP for
167–172.
3
0 min and subsequently eluted with phosphate buffer (20 mM,
[14] R.S. Phillips, J. Am. Chem. Soc. 122 (2000) 1008–1014.
[
[
[
15] R.S. Phillips, Biochemistry 30 (1991) 5927–5934.
pH 8, 0.16 M KCl) through a gel filtration column (PD-10, Pharma-
cia) to remove excess PLP. Rapid-scanning stopped-flow experi-
ments were performed in the same buffer on an RSM-1000
spectrophotometer (OLIS, Inc.), equipped with a stopped-flow cell
mixer compartment of 1 cm path length capable of collecting up
to 1000 scans per second with a dead time of 2 ms. The spectra
were fitted with Global Works software provided by OLIS.
16] R.S. Phillips, P. Gollnick, FEBS Lett. 268 (1990) 213–216.
17] R.S. Phillips, I. Richter, P. Gollnick, P. Brzovic, M.F. Dunn, J. Biol. Chem. 266
(1991) 18642–18648.
[
[
18] M. Lee, R.S. Phillips, Bioorg. Med. Chem. 3 (1995) 195–205.
19] T.V. Demidkina, L.N. Zakomirdina, V.V. Kulikova, I.S. Dementieva, N.G. Faleev,
L. Ronda, A. Mozzarelli, P.D. Gollnick, R.S. Phillips, Biochemistry 42 (2003)
11161–11169.
[
[
20] V.V. Kulikova, L.N. Zakomirdina, N.P. Bazhulina, I.S. Dementieva, N.G. Faleev,
P.D. Gollnick, T.V. Demidkina, Biochemistry (Moscow) 68 (2003) 1181–1188.
21] S. Ku, P. Yip, P.L. Howell, Acta Crystallogr. D62 (2006) 814–823.
Appendix A. Supplementary data
[22] L. Davis, D.E. Metzler, The Enzymes, third ed., Academic Press, New York, 1972.
pp 33–74.
[
[
23] R.S. Phillips, E.W. Miles, L.A. Cohen, Biochemistry 23 (1984) 6228–6234.
24] V.L. Schramm, Ann. Rev. Biochem. 67 (1998) 693–720.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.abb.2014.07.027.
[25] R.S. Phillips, E.W. Miles, L.A. Cohen, J. Biol. Chem. 260 (1985) 14665–14670.
[
26] F. Lingens, K.H. Weiler, Justus Liebigs Annalen der Chemie 662 (1963) 139–
46.
1
References
[
[
27] H.R. Snyder, F.J. Pilgrim, J. Am. Chem. Soc. 70 (1948) 1962–1963.
28] C. Ma, S. Yu, X. He, X. Liu, J.M. Cook, Tetrahedron Lett. 41 (2000) 2781–2785.
[
[
1] E.E. Snell, Adv. Enzymol. Relat. Areas Mol. Biol. 42 (1975) 287–333.
2] P. Di Martino, A. Merieau, R. Phillips, N. Orange, C. Hulen, Can. J. Microbiol. 48
[29] K. Goswami, S. Paul, S.T. Bugde, S. Sinha, Tetrahedron 68 (2012) 280–286.
[30] R. Appel, Angew. Chem. Int. Ed. Engl. 14 (1975) 801–811.
[31] H.K. Chenault, J. Dahmer, G.M. Whitesides, J. Am. Chem. Soc. 111 (1989) 6354–
6364.
[32] R. Scherzer, G.Y. Gdalevsky, Y. Goldgur, R. Cohen-Luria, S. Bittner, A.H. Parola, J.
Enzyme Inhib. Med. Chem. 24 (2009) 250–355.
(
2002) 132–137.
3] P. Di Martino, R. Fursy, L. Bret, B. Sundararaju, R. Phillips, Can. J. Microbiol. 49
2003) 443–449.
4] J. Lee, T. Maeda, S.H. Hong, T.K. Wood, Appl. Environ. Microbiol. 75 (2009)
703–1716.
[
[
[
(
1
[33] A.P. Harris, R.S. Phillips, FEBS J. 280 (2013) 1807–1817.
[34] R.S. Phillips, S.L. Bender, P. Brzovic, M.F. Dunn, Biochemistry 29 (1990) 8608–
5] A. Collet, S. Vilain, P. Cosette, G.A. Junter, T. Jouenne, R.S. Phillips, P. Di Martino,
Antonie van Leeuwenhoek 91 (2007) 71–85.
6] E.L. Chant, D.K. Summers, Mol. Microbiol. 63 (2007) 35–43.
7] A. Anyanful, J.M. Dolan-Livengood, T. Lewis, S. Sheth, M.N. Dezalia, M.A.
Sherman, L.V. Kalman, G.M. Benian, D. Kalman, Mol. Microbiol. 57 (2005) 988–
8614.
[
[
[35] D. Mili c´ , T.V. Demidkina, N.G. Faleev, R.S. Phillips, D. Matkovi c´ - Cˇ alogovi c´ , A.A.
Antson, J. Am. Chem. Soc. 133 (2011) 16468–16476.
[36] R.S. Phillips, K. Ravichandran, R.L.V. Tersch, Enzyme Microb. Technol. 11
(1989) 80–83.
1
007.
[
[
8] H.H. Lee, M.N. Molla, C.R. Cantor, J.J. Collins, Nature 467 (2010) 82–85.
9] N.M. Vega, K.R. Allison, A.S. Khalil, J.J. Collins, Nat. Chem. Biol. 8 (2012) 431–
[37] R.S. Phillips, P.D. Gollnick, J. Biol. Chem. 264 (1989) 10627–10632.
[38] L.H. Yang, S.A. Ahmed, E.W. Miles, Expr. Purif. 8 (1996) 126–136.
[39] W.W. Cleland, Methods Enzymol. 63 (1979) 103–138.
4
33.
[
10] T. Watanabe, E.E. Snell, J. Biochem. 82 (1997) 733–745.