Differential effects of bromination on substrates and inhibitors of kynureninase from Pseudomonas fluorescens

Article abstract of DOI:10.1039/b208910f

A series of brominated compounds has been synthesized and evaluated as substrates and inhibitors of kynureninase from Pseudomonas fluorescens. Both 3-bromo-L-kynurenine and 5-bromo-L-kynurenine were found to be substrates with similar k_cat values to L-kynurenine, but the K_m value for 3-bromo-L-kynurenine is very high (ca, 2 mM) compared to that for 5-bromo-L-kynurenine (11 μM) and L-kynurenine (25 μM). Both isomers of bromokynurenine react with kynureninase within the dead time of the stopped-flow instrument (ca, 1 ms) to form quinonoid intermediates with a λ_max of 494 nm that decay with rate constants of 300-600 s^-1, similar to L-kynurenine. The two diastereomers of 5-bromodihydro-L-kynurenine were also prepared, and are more potent inhibitors than dihydro-L-kynurenines. (4R)-5-Bromodihydro-L-kynurenine is one of the most potent inhibitors of P. fluorescens kynureninase found to date (K_i = 55 nM) and also acts as a slow substrate; the (4S)-epimer, on the other hand, shows no measurable substrate activity, but it is a potent competitive inhibitor with a K_i value of 170 nM. In contrast, brominated analogs of (S)-(2-aminophenyl)-L-cysteine S,S-dioxide, (S)-(2-amino-4-bromophenyl)-L-csteine S,S-dioxide and (S)-(2-amino-5-bromophenyl)-L-cysteine S,S-dioxide are competitive inhibitors of kynureninase, with K_i values of about 300 and 400 nm, respectively, about ten-fold higher than the value of 27 nM obtained for the parent compound. These results suggest that the binding modes of substrates and the various classes of inhibitors in the active site of kynureninase are different.

Full text of DOI:10.1039/b208910f

View Article Online / Journal Homepage / Table of Contents for this issue
Differential effects of bromination on substrates and inhibitors of
kynureninase from Pseudomonas fluorescens
Christian Heiss,a Jay Andersona and Robert S. Phillips*a,b,c
a Department of Chemistry, University of Georgia, Athens, Georgia 30602-2556, USA
b Department of Biochemistry and Molecular Biology, Center for Metalloenzyme Studies,
University of Georgia, Athens, Georgia 30602-2556, USA
c Center for Metalloenzyme Studies, University of Georgia, Athens, Georgia 30602-2556, USA
Received 13th September 2002, Accepted 15th October 2002
First published as an Advance Article on the web 9th December 2002
A series of brominated compounds has been synthesized and evaluated as substrates and inhibitors of kynureninase
from Pseudomonas fluorescens. Both 3-bromo--kynurenine and 5-bromo--kynurenine were found to be substrates
with similar kcat values to -kynurenine, but the Km value for 3-bromo--kynurenine is very high (ca. 2 mM) compared
to that for 5-bromo--kynurenine (11 µM) and -kynurenine (25 µM). Both isomers of bromokynurenine react with
kynureninase within the dead time of the stopped-flow instrument (ca. 1 ms) to form quinonoid intermediates with a
λmax of 494 nm that decay with rate constants of 300–600 sϪ1, similar to -kynurenine. The two diastereomers of
5-bromodihydro--kynurenine were also prepared, and are more potent inhibitors than dihydro--kynurenines.
(4R)-5-Bromodihydro--kynurenine is one of the most potent inhibitors of P. fluorescens kynureninase found to date
(Ki = 55 nM) and also acts as a slow substrate; the (4S)-epimer, on the other hand, shows no measurable substrate
activity, but it is a potent competitive inhibitor with a Ki value of 170 nM. In contrast, brominated analogs of (S)-
(2-aminophenyl)--cysteine S,S-dioxide, (S)-(2-amino-4-bromophenyl)--cysteine S,S-dioxide and (S)-(2-amino-5-
bromophenyl)--cysteine S,S-dioxide are competitive inhibitors of kynureninase, with Ki values of about 300 and 400
nm, respectively, about ten-fold higher than the value of 27 nM obtained for the parent compound. These results
suggest that the binding modes of substrates and the various classes of inhibitors in the active site of kynureninase
are different.
Shemyakin.7 In agreement with this hypothesis, we then found
that (S)-(2-aminophenyl)--cysteine S,S-dioxide (SAPCO),
the structure of which resembles that of the postulated gem-
Introduction
Kynureninase [EC 3.7.1.1] is a pyridoxal 5Ј-phosphate (PLP)
dependent enzyme that catalyzes a key step in the catabolism
of tryptophan in Pseudomonas fluorescens, and some other
diolate intermediate, is a very potent inhibitor with a Ki value
of 70 nM.8 In the present study, we have prepared brominated
bacteria,1 the hydrolytic cleavage of -kynurenine to give
analogs of -kynurenine, dihydro--kynurenine, and (S)-(2-
anthranilic acid and -alanine [eqn. (1)]. As part of the bio-
aminophenyl)--cysteine-S,S-dioxide, and we compared the
synthetic pathway from -tryptophan to NADϩ in eucaryotes,
effects of bromination on substrate and inhibitor activity.
kynureninase catalyzes the cleavage of 3-hydroxykynurenine
Unexpectedly, the results of this study suggest that the sub-
to produce 3-hydroxyanthranilate. 3-Hydroxyanthranilate
dioxygenase converts 3-hydroxyanthranilate to quinolinic
acid, which is an N-methyl--aspartate (NMDA) receptor
strates and different classes of inhibitors bind to the active site
of kynureninase in different modes. These results are important
for the design of potent and selective inhibitors of bacterial and
agonist and, hence, a postulated neurotoxin. Elevated levels of
human kynureninase.
quinolinic acid have been implicated in the etiology of a
number of neurological disorders such as Huntington’s disease,
AIDS-related dementia, stroke and other inflammatory brain
Experimental
diseases.2–5 Effective and selective inhibition of human
General
kynureninase thus represents a possible treatment of these
1H- and 13C-NMR spectra were recorded on a Bruker AC 250
diseases.
or AC 300, respectively. Optical rotations were measured with
(1)
an Autopol IV polarimeter from Rudolph Research; [α]2D0 values
are given in units of 10Ϫ1 deg cm2 Ϫ1. Enzyme assays and
g
kinetic experiments were performed with a Varian Cary 1E UV/
visible spectrophotometer. HPLC measurements were carried
out on an instrument with two Rainin Rabbit HP pumps and a
LDC Milton Roy Spectromonitor 3000 variable wavelength
detector using Gilson Unipoint software.
In a previous study, we prepared the two diastereomers of di-
hydro--kynurenine, and we found that the (4S)-diastereomer is
a potent competitive inhibitor of bacterial kynureninase with a
Ki value of 300 nM.6 While the (4R)-diastereomer also dis-
played inhibitory activity (Ki = 1.5 µM), it is a slow substrate for
kynureninase-catalyzed retro-aldol reaction to give o-amino-
benzaldehyde and -alanine. These results suggested that the
kynureninase reaction proceeds through a gem-diolate inter-
mediate, as had originally been proposed by Braunstein and
Enzyme
Kynureninase was purified from E. coli cells containing plasmid
pTZ18U with the kynureninase gene of P. fluorescens as
previously described.9 The protein concentration of crude
extracts was measured by the method of Bradford,10 and the
concentration of purified kynureninase was measured by
absorbance at 280 nm (A2801% = 14).11
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 8 8 – 2 9 5
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
288

View Article Online
Kinetic measurements
dried (MgSO4), and evaporated. The remaining acetic acid was
removed by repeated evaporation with toluene. Recrystalliz-
ation from MeOH–H2O gave N α,N-diacetyl-5-bromo--
kynurenine methyl ester (2a) (1.03 g, 80%) as pale yellow
needles, mp 176–178 ЊC (lit.17 mp 178 ЊC); [α]2D0 ϩ85.6 (c = 0.81,
CHCl3); 1H-NMR (CDCl3) δ 11.35 (s, 1H), 8.70 (d, J = 9.0 Hz,
1H), 7.98 (d, J = 2.3 Hz, 1H), 7.66 (dd, J = 9.2, 1.9 Hz, 1H), 6.49
(d, J = 7.6 Hz, 1H), 4.96 (m, 1H), 3.77 (s, 3H), 3.75 (m, 2H),
2.24 (s, 3H), 2.04 (s, 3H).
Kynureninase activity was measured by following the decrease
in absorbance at 360 nm due to conversion of -kynurenine to
anthranilate (∆ε = 4500 MϪ1 cmϪ1). The competitive inhibition
was measured by variation of [-kynurenine] at several fixed
concentrations of inhibitor. Km, Vmax, and Ki values were
determined by fitting of initial rate data to eqns. (2) and (3)
v = Vmax[S]/(Km ϩ [S])
(2)
(3)
(2S )-2-Amino-4-(2Ј-amino-5Ј-bromophenyl)-4-oxobutanoic acid
(5-bromo-L-kynurenine) (3)
v = Vmax[S]/{Km(1 ϩ [I]/Ki) ϩ [S]}
N α,N-Diacetyl-5-bromo--kynurenine methyl ester (0.96 g, 2.5
mmol) was refluxed for 18 h with 6 M HCl (9 mL). The solution
was evaporated and the yellow oily residue was purified by low
pressure reverse phase column chromatography on C18 silica gel
(10% MeOH–H2O) to give 5-bromo--kynurenine (3) (0.57 g,
79%) as a waxy yellow solid, mp 213–217 ЊC (lit.18 mp 233 ЊC
for racemate); [α]2D0 ϩ62.5 (c = 0.28, dioxane–H2O 1 : 1);
1H-NMR (1% DCl–D2O) δ 8.21 (d, J = 2.1 Hz, 1H), 7.77 (dd,
J = 8.6, 2.1 Hz, 1H), 7.22 (d, J = 8.6 Hz, 1H), 4.48 (t, J = 5.0 Hz,
1H), 3.84 (d, J = 5.1 Hz, 2H).
using the compiled FORTRAN programs HYPER and COMP
of Cleland.12 Due to the very low Km values for (4R)-5-bromo-
dihydro--kynurenine, the Km and Vmax values could not be
determined in the conventional manner, and were determined
by running the reaction at a fixed concentration (35 µM) of
substrate and fitting the time course of the increase in
absorbance at 360 nm due to the formation of 2-amino-5-
bromobenzaldehyde (∆ε = 3900 MϪ1 cmϪ1) to the integrated
Michaelis–Menten relationship in eqn. (4).13
Kmln[S]/[S]o ϩ [S] Ϫ [S]o = ϪVmax
t
(4)
(2S,4S )- and (2S,4R)-2-Amino-4-(2Ј-amino-5Ј-bromophenyl)-4-
hydroxybutanoic acid (5-bromodihydro-L-kynurenine) (4a ؉ 4b)
Stopped-flow kinetics
Rapid-scanning stopped flow measurements were performed
on an OLIS RSM-1000 spectrophotometer, as described
previously.14 The reactions contained 1 mg mLϪ1 (22 µM)
kynureninase in 0.05 M potassium phosphate, at pH 8.0 and
25 ЊC. Spectra were collected at a rate of 1000 sϪ1 over the
wavelength range 320–570 nm. The spectra were fitted using
the global analysis program, GlobalWorks, provided by OLIS15
to obtain the singular value decomposition (SVD) spectra of
intermediates and the rate constants for the reaction.
5-Bromo--kynurenine (143 mg, 0.5 mmol) was suspended in
H2O (50 mL) and treated with NaBH4 (50 mg, 1.3 mmol). The
mixture was stirred overnight, brought to pH 3 by addition of
1 M HCl, and loaded onto a Dowex-50 ion exchange column.
Elution with 1 M NH3 yielded a 2 : 1 mixture of (2S,4S) and
(2S,4R)-diastereomers (100 mg, 69%) (Anal. Calcd. for
C10H13BrN2O3ؒ0.5H2O: C, 40.29; H, 4.73; N, 9.40. Found: C,
40.04; H, 4.50; N, 9.17%). The mixture was separated by HPLC
(C18, 2.5 × 25 cm, isocratic at 5% MeOH–0.1%TFA) in 4 mg
portions. (2S,4S)-Isomer (4b): mp > 360 ЊC; [α]2D0 Ϫ53.5 (c =
Synthesis of brominated substrates and inhibitors
1
0.40, 1% NaOH); H-NMR (1.0% NaOD–D2O) δ 7.36 (d, J =
2.1 Hz, 1H), 7.21 (dd, J = 8.6, 2.4 Hz, 1H), 6.69 (d, J = 8.6 Hz,
1H), 3.31 (dd, J = 8.0, 5.0 Hz, 1H), 2.07 (m, 1H), 1.75 (m, 1H);
13C-NMR (1.0% D2O) δ 185.3, 162.6, 145.0, 133.3, 131.6, 121.7,
113.2, 69.9, 55.9, 43.4. (2S,4R)-Isomer (4a): mp > 360 ЊC; [α]2D0
ϩ8.2 (c = 0.44, 1% NaOH); 1H-NMR (1.0% NaOD–D2O)
δ 7.39 (d, J = 2.1 Hz, 1H), 7.25 (dd, J = 8.6, 2.3 Hz, 1H), 6.74
(d, J = 8.6 Hz, 1H), 3.27 (t, J = 6.9 Hz, 1H), 2.00 (m, 1H), 1.89
(m, 1H); 13C-NMR (1.0% D2O) δ 185.2, 162.5, 145.1, 133.4,
131.6, 121.6, 113.1, 71.0, 56.4, 43.4.
(2S )-Methyl 2-acetamido-4-(2Ј-acetamidophenyl)-4-
oxobutanoate (Nꢀ,N-diacetyl-L-kynurenine methyl ester) (1a)
A stream of ozone was passed through a solution of N α-
acetyl--tryptophan methyl ester (2.28 g, 8.76 mmol) in MeOH
(100 mL) at Ϫ78 ЊC. After the reaction was completed (10% KI-
trap), excess oxidizing species were destroyed by treatment
with sat. NaHSO3 solution (100 mL). The mixture was
extracted with CH2Cl2, dried (MgSO4), and evaporated.
The residue was taken up in MeOH (80 ml) and treated with
trifluoroacetic acid (5 mL). After stirring at room temperature
overnight, the mixture was evaporated and the resulting oil
taken up in CHCl3 (100 mL). Acetic anhydride (10 mL) was
added, the solution was stirred for 1 h, washed with saturated
NaHCO3 (10 × 30 mL), dried (MgSO4), and evaporated.
Recrystallization from ethyl acetate–hexanes gave N a,N-
diacetyl--kynurenine methyl ester (1a) (1.36 g, 53%) as white
needles, mp 163–164 ЊC (lit. 134–136 ЊC);16 [α]2D0 ϩ141 (c = 0.5,
CHCl3) (Anal. Calcd for C15H18N2O5: C, 58.82; H, 5.92; N,
(2S )-Methyl 2-acetamido-4-(2Ј-amino-5Ј-iodophenyl)-4-
oxobutanoate (Nꢀ-acetyl-5-iodo-L-kynurenine methyl ester) (2b)
N α-Acetyl--kynurenine methyl ester (1b) (1.5 g, 5.7 mmol)19
was dissolved in acetic acid (50 mL) containing fused sodium
acetate (3.75 g) and treated with iodine (3.6 g, 14.2 mmol).
After 1 h at room temperature, the mixture was poured into
ice–water, treated with 10% NaHSO3 (20 mL), extracted with
CHCl3, dried (MgSO4), and evaporated. Residual acetic acid
was removed by repeated evaporation with toluene. Passage of
the crude product through a short column of silica gel and
elution with ethyl acetate gave N α-acetyl-5-iodo--kynurenine
methyl ester (2b) (1.15 g, 52%) as pale yellow needles, mp
1
9.15. Found: C, 58.86; H, 5.94; N, 8.97%); H-NMR (CDCl3)
δ 11.45 (s, 1H), 8.75 (d, J = 8.7 Hz, 1H), 7.88 (d, J = 7.8 Hz, 1H),
7.57 (t, J = 7.8 Hz, 1H), 7.12 (t, J = 7.3 Hz, 1H), 6.54 (d, J = 7.7
Hz, 1H), 4.97 (m, 1H), 3.76 (s, 3H), 3.75 (m, 2H), 2.23 (s, 3H),
2.04 (s, 3H).
156–157 ЊC; [α]2D0 ϩ131 (c = 0.56, CHCl3); H-NMR (CDCl3)
1
δ 7.93 (d, J = 1.9 Hz, 1H), 7.49 (dd, J = 8.8, 1.9 Hz, 1H), 6.56
(d, J = 7.9 Hz, 1H), 6.46 (d, J = 8.8 Hz, 1H), 4.95 (m, 1H), 3.74
(s, 3H), 3.71 (dd, J = 3.9, 18.2 Hz, 1H), 3.52 (dd, J = 18.2, 3.9
Hz, 1H), 2.02 (s, 3H).
(2S )-Methyl 2-acetamido-4-(2Ј-acetamido-5Ј-bromophenyl)-4-
oxobutanoate (Nꢀ,N-diacetyl-5-bromo-L-kynurenine methyl
ester) (2a)
N α,N-Diacetyl--kynurenine methyl ester (1.02 g, 3.32 mmol)
was dissolved in acetic acid (20 mL) containing fused sodium
acetate (1.5 g) and treated with bromine (4 mL). After 1 h at
room temperature, the mixture was poured into ice–water,
treated with 10% NaHSO3 (20 mL), extracted with CHCl3,
(2S )-2-Amino-4-(2Ј-amino-3Ј-bromophenyl)-4-oxobutanoic acid
(3-bromo-L-kynurenine) (6a)
N α-Acetyl-5-iodo--kynurenine methyl ester (400 mg, 1.03
mmol) was dissolved in acetic acid (8 mL) containing fused
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 8 8 – 2 9 5
289

View Article Online
Scheme 1
(S)-(2-Amino-5-bromophenyl)-L-cysteine S,S-dioxide (12)
sodium acetate (600 mg) and treated with bromine (0.051 mL,
160 mg, 1.0 mmol). After 30 min at room temperature, the
mixture was poured into ice–water and the precipitate (380 mg)
was filtered and dried. The product was a mixture of N α-acetyl-
3-bromo-5-iodo--kynurenine methyl ester (5a) (242 mg, 50%)
and N α-acetyl-3,5-dibromo--kynurenine methyl ester (5b) (135
mg, 31%). The mixture (309 mg) was refluxed in 6 M HCl
overnight. Low pressure reverse phase chromatography on
C18-silica gel (10–20% MeOH–H2O) gave (2S)-2-amino-4-
(2Ј-amino-3Ј-bromophenyl)-4-oxobutanoic acid (3-bromo--
kynurenine) (6a) (52 mg, 43%), mp 210–212 ЊC (lit.20 >175 ЊC);
Triethylamine (3.5 mL) was added dropwise to a solution of
2.24 g -cysteine hydrochloride hydrate, 2.00 g 1-bromo-
3,4-dinitrobenzene,22 and 9 mL DMF. After 40 minutes, a red–
orange solid formed, which was broken up in 10 mL of water
and filtered to give the crude product. Recrystallization of the
crude product from 400 mL hot water per g gave 1.3 g (28.5%)
bright yellow crystals. MS (ESI): m/z 321, 323 (M ϩ H)ϩ. This
product was found to be a mixture of (S)-(5-bromo-2-nitro-
phenyl)--cysteine (10) and (S)-(4-bromo-2-nitrophenyl)--
1
cysteine (7) by NMR. H-NMR of 10: (1% DCl–D2O) δ 8.04
[α]2D0 ϩ63.2 (c = 0.5, dioxane–H2O, 1 : 1); H-NMR (1% DCl–
1
(d, J = 8.4 Hz, 1H), 7.81 (d, J = 2.0 Hz, 1H), 7.56 (dd, J = 8.8,
2.0 Hz, 1H), 4.37 (dd, J = 3.6, 6.8 Hz, 1H), 3.75 (dd, J = 4.8,
16 Hz, 1H), 3.60 (dd, J = 8.0, 16 Hz, 1H).
D2O) δ 7.85 (d, J = 8.1 Hz, 1H), 7.75 (d, J = 7.7 Hz, 1H), 6.71
(t, J = 7.9 Hz, 1H), 4.34 (t, J = 4.9 Hz, 1H), 3.81 (d, J = 5.0 Hz,
1H).
The above mixture (0.200 g) was added to a solution of 5 mL
of trifluoroacetic acid and 1 mL of 30% H2O2. After 48 hours,
the pale yellow solution was evaporated in vacuo at 30 ЊC.21
Trituration with water resulted in pale yellow-to-white crystals,
MS (ESI): m/z 353, 355 (M ϩ H)ϩ. The product was dissolved
in 25 mL of glacial acetic acid, 1.00 g Zn powder was added,
and the mixture was stirred for 24 hours. The reaction mixture
was then filtered through Celite to remove remaining Zn and
precipitated Zn(OAc)2, and the filtrate was evaporated in vacuo.
Trituration of the residue with water and methanol gave 0.080 g
(43.7%) of a tan powder. HPLC analysis (C18, 10% MeOH)
showed that this product was a 2 : 1 mixture of (S)-(2-amino-5-
bromophenyl)--cysteine S,S-dioxide (9) and (S)-(2-amino-5-
bromophenyl)--cysteine S,S-dioxide (12). MS (ESI): m/z 323,
325 (M ϩ H)ϩ; 1H-NMR of 12: (1% DCl–D2O) δ 7.77 (d,
J = 2.2 Hz, 1H), 7.55 (dd, J = 8.8, 2.4 Hz, 1H), 6.86 (d, J = 8.4
Hz, 1H), 4.54 (dd, J = 6.4, 4.4 Hz, 1H), 3.99 (m, 2H).
(S)-(4-Bromo-2-nitrophenyl)-L-cysteine (7)
Triethylamine (1.4 mL) was added dropwise to a solution of
1.0 g -cysteine hydrochloride hydrate, 0.894 g 4-bromo-2-
nitrofluorobenzene, and 4 mL of DMF. After 40 minutes, the
resultant yellow solid was broken up in 20 mL of water and
filtered. The crude product was recrystallized from 300 mL of
hot water per g of product to yield 1.18 g (58%) bright yellow
crystals of (S)-(4-bromo-2-nitrophenyl)--cysteine (7). Anal.
Calcd. for C9H9BrN2O4S: C, 33.66; H, 2.82; N, 8.72. Found:
1
C, 33.83; H, 2.85; N, 8.68%. H-NMR (1% DCl–D2O) δ 8.26
(d, J = 2.2 Hz, 1H), 7.72 (dd, J = 8.7, 2.2 Hz, 1H), 7.45 (d, J = 8.7
Hz, 1H), 4.24 (t, J = 4.7 Hz, 1H), 3.66 (dd, J = 4.0, 15 Hz, 1H),
3.47 (dd, J = 7.7, 15 Hz, 1H).
(S)-(2-Amino-4-bromophenyl)-L-cysteine S,S-dioxide (9)
Results
(S)-(4-Bromo-2-nitrophenyl)--cysteine (0.300 g) was added to
a mixture of 10 mL 88% formic acid and 2 mL 30% H2O2. The
solution was stirred for one week, then the solvent was removed
in vacuo at 30 ЊC.21 The syrupy product was triturated with
water to give 0.331 g (82%) pale yellow-to-white crystals of
(S)-(4-bromo-2-nitrophenyl)--cysteine S, S-dioxide (8). (S)-
(4-Bromo-2-nitrophenyl)--cysteine S,S-dioxide (0.301 g) was
dissolved in 25 mL of glacial acetic acid, 1.00 g Zn powder was
added, and the mixture was stirred for 24 hours. The reaction
mixture was filtered through Celite to remove remaining Zn and
precipitated Zn(OAc)2, and the filtrate was evaporated in vacuo.
Trituration with water and methanol gave 0.16 g (67%) of a
cream-colored powder of (S)-(2-amino-4-bromophenyl)--
cysteine S,S-dioxide (9). MS (ESI): m/z 323, 325 (M ϩ H)ϩ.
1H-NMR (1% DCl–D2O) δ 7.47 (d, J = 8.7 Hz, 1H), 7.13 (d,
J = 1.7 Hz, 1H), 6.98 (dd, J = 8.7, 1.7 Hz, 1H), 4.53 (t, J = 4.7
Hz, 1H), 3.96 (m, 2H).
Synthesis of brominated substrates and inhibitors
5-Bromo--kynurenine (3) was prepared as summarized in
Scheme 1, by acid hydrolysis of N α,N-diacetyl-5-bromo--
kynurenine methyl ester (2a), obtained by a procedure similar
to that described by Casnati et al.17 5-Bromo--kynurenine was
then reduced to 5-bromodihydro--kynurenine with NaBH4
following a method we previously described for preparation of
dihydro--kynurenine.5 A 2 : 1 mixture of (4R)-5-bromodi-
hydro--kynurenine (4a) and (4S)-5-bromodihydro--kynure-
nine (4b) was obtained, which was separated using preparative
HPLC. N α-Acetyl-5-iodo--kynurenine methyl ester (2b) was
prepared from N α-acetyl--kynurenine methyl ester (1b)
according to Scheme 1. Attempts to obtain 5-iodo--kynure-
nine by hydrolysis of N α-acetyl-5-iodo--kynurenine methyl
ester were unsuccessful, because the iodine was substituted by a
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 8 8 – 2 9 5
290

View Article Online
Scheme 2
proton under the strongly acidic conditions needed for
tained 67% of the (S)-(2-amino-4-bromophenyl)--cysteine
hydrolysis. This observation, however, indicated to us that
iodine could be used as an easily removable protecting group in
the convenient synthesis of 3-bromo--kynurenine (Scheme 1).
Thus, N α-acetyl-5-iodo--kynurenine methyl ester (2b) was sub-
jected to bromination conditions, and the product was found to
be a 2 : 1.1 mixture of N α-acetyl-3-bromo-5-iodo--kynurenine
methyl ester (5a) and N α-acetyl-3,5-dibromo--kynurenine
methyl ester (5b). The mixture was subjected to hydrolysis, and
the resulting mixture of 3-bromo--kynurenine and 3,5-
dibromo--kynurenine was readily separated by atmospheric
pressure reverse phase liquid chromatography on C18 silica gel.
This synthesis of 3-bromo--kynurenine is shorter and more
efficient than the only other one reported in the literature.20
A mixture of (S)-(2-amino-4-bromophenyl)--cysteine S,S-
dioxide (4-Br-SAPCO) (9) and (S)-(2-amino-5-bromophenyl)-
-cysteine S,S-dioxide (5-Br-SAPCO) (12) was prepared by
the reaction of -cysteine with 1-bromo-3,4-dinitrobenzene
(obtained by nitration of 3-nitrobromobenzene by the method
of Mangini22), followed by oxidation with H2O2 in tri-
fluoroacetic acid (Scheme 2). The final step, reduction of the
nitro group, was performed with Zn in glacial acetic acid,
instead of Pd and H2 in formic acid, as we had used previously,8
to avoid hydrogenolysis of the aryl bromide. This reduction
method was also found to produce a cleaner product than
catalytic hydrogenation in the preparation of SAPCO. We have
found that the catalytic reduction performed in formic acid, as
described in our original procedure,8 gives variable amounts
of N 2-formyl SAPCO. We had hoped that the nucleophilic sub-
stitution reaction of cysteine with 1-bromo-3,4-dinitrobenzene
would show regioselectivity in favor of the desired 5-isomer.
However, NMR analysis of the brominated o-nitrophenyl--
cysteine product clearly showed two isomeric products. HPLC
analysis of the final product mixture demonstrated that it con-
S,S-dioxide (9) and only 33% of the desired (S)-(2-amino-5-
bromophenyl)--cysteine S,S-dioxide (12), since we had the
authentic 4-bromo isomer as a standard. (S)-(2-Amino-4-
bromophenyl)--cysteine S,S-dioxide (9) was prepared as shown
in Scheme 2 from commercially available 4-bromo-2-nitro-
fluorobenzene.
Steady-state kinetic studies
5-Bromo--kynurenine is an excellent substrate for kynureni-
nase, with a kcat/Km value about half that of -kynurenine,
whereas 3-bromo--kynurenine also has a high turnover, but
kcat/Km is reduced about 100-fold due to a very high Km (Table
1). Thus, the bromo substitutent at the 5-position has much
less effect on activity than the 3-bromo substitutent. Of the
diastereomers of 5-bromodihydro--kynurenine, only (4R)-5-
bromodihydro--kynurenine acts as a slow substrate, with a low
kcat value of 4.6 × 10Ϫ3 Ϫ1. Similarly, we found previously that
s
(4R)-dihydro--kynurenine is a substrate, while the (4S)-dia-
stereomer is an inhibitor.5 However, both diastereoisomers of
5-bromodihydro--kynurenine are potent competitive inhibitors
of kynureninase, with Ki values of 170 and 55 nM, respectively,
for the (4S) and (4R)-epimers (Table 1). In contrast, (4S)-di-
hydro--kynurenine is a more potent inhibitor (Ki = 300 nM)
than the (4R)-epimer (Ki = 1.5 mM).5 The Ki value of 4-Br-
SAPCO is 300 nM (Table 1), while that of the 2 : 1 mixture of
4- and 5-bromo isomers is 327 nM. Assuming that the observed
Ki is a weighted average for the 2 : 1 mixure of the 4-bromo and
5-bromo isomers, the Ki for 5-Br-SAPCO was estimated to be
400 nM (Table 1). In this study, we also repeated the measure-
ment of the Ki for the parent compound, SAPCO, and we found
it to be 27 nM, somewhat lower than we reported previously (70
nM).8 This may be due to the change in the reduction procedure
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 8 8 – 2 9 5
291

View Article Online
Table 1 Kinetic parameters for kynureninase inhibitors and substrates
Compound
Ki/nM
kcat/sϪ1
Km/µM
(kcat/Km)/MϪ1 sϪ1
-Kynurenine
5-Br--Kyn
3-Br--Kyn
—
—
—
300
1500
170
55
27
300
400
6.9
2.1
11
—
25
12
2.8 × 105
1.8 × 105
5.5 × 103
—
2.0 × 103
(4S)-H2--Kyna
(4R)-H2--Kyna
(4S)-5-Br-H2--Kyn
(4R)-5-Br-H2--Kyn
SAPCO
—
1.5
0.6
4.0 × 105
—
—
—
4.6 × 10Ϫ3
5.2 × 10Ϫ2
8.8 × 104
—
—
—
—
—
—
—
—
4-Br-SAPCO
5-Br-SAPCO
—
a From ref. 6.
noted above, which avoids formation of the N 2-formyl
derivative.
these reactions results in the rapid formation (kobs = 78 10 sϪ1
of a stable absorbance peak at 496 nm, similar to that we
)
14
observed previously with -kynurenine (kobs = 65 sϪ1
)
(Fig.
1B). This new intermediate is formed by rapid trapping of the
enamine intermediate, after elimination of 5-bromoanthra-
nilate, and subsequent deprotonation at C4Ј to form a
quasi-stable quinonoid complex. Rapid-scanning stopped-flow
kinetic experiments with 1 mM 3-bromo--kynurenine also
showed the rapid formation, within the dead time, and sub-
sequent decay, of an intermediate at 494 nm, with two phases,
kobs = 300 sϪ1 and 21 sϪ1 (Fig. 2A). However, the amplitude of
Stopped-flow kinetic studies
The reactions of 3-bromo- and 5-bromo--kynurenine with
kynureninase were examined by rapid scanning stopped-flow
spectrophotometry. In the reaction of 5-bromo--kynurenine,
there is an absorption peak at 494 nm (Fig. 1A) due to a quino-
Fig. 1 A. Rapid-scanning stopped flow fitted SVD spectra of the
reaction of kynureninase with 0.64 mM 5-bromo--kynurenine. Solid
line, initial spectrum; dashed line, intermediate spectrum; dashed and
dotted line, final spectrum. B. Rapid-scanning stopped flow fitted SVD
spectra of the reaction of kynureninase with 0.64 mM 5-bromo--
kynurenine in the presence of 10 mM benzaldehyde. Solid line, initial
spectrum; Dashed line, first intermediate spectrum; dashed and dotted
line, second intermediate spectrum; dashed and double dotted line, final
spectrum.
Fig. 2 A. Rapid-scanning stopped flow fitted SVD spectra of the
reaction of kynureninase with 1 mM 3-bromo--kynurenine. Solid line,
initial spectrum; dashed line, intermediate spectrum; dashed and dotted
line, final spectrum. B. Rapid-scanning stopped flow SVD spectra of
the reaction of kynureninase with 1 mM 3-bromo--kynurenine in the
presence of 10 mM benzaldehyde. Solid line, initial spectrum; Dashed
line, intermediate spectrum; dashed and dotted line, second
intermediate spectrum; dashed and double dotted line, final spectrum.
noid intermediate, formed within the dead time (ca. 1 ms) of the
stopped-flow instrument, which undergoes rapid decay in two
phases, at about 600 sϪ1, and 22 sϪ1, very similar to the behavior
of -kynurenine that we observed previously.14 These rate con-
stants do not change significantly over the concentration range
of 5-bromo--kynurenine from 0.08 to 0.64 mM. Addition of
10 mM benzaldehyde and 5-bromo--kynurenine together in
the absorbance changes with 3-bromo--kynurenine is smaller
than with 5-bromo--kynurenine. Addition of 10 mM benz-
aldehyde to the reactions of 3-bromo--kynurenine also results
in the formation of a new quinonoid intermediate with kobs = 56
5 sϪ1 (Fig. 2B).
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 8 8 – 2 9 5
292

View Article Online
SAPCO also reacts with kynureninase to form a prominent
absorption band at about 500 nm, but at a slower rate than -
kynurenine (Fig. 3A).8 This intermediate is stable in the steady-
rapidly to form a PLP ketimine (EK1). Upon hydration of
the carbonyl, cleavage of the Cβ–Cγ bond can occur to give
anthranilate and a PLP enamine (EE). It is this enamine species
which can react in an aldol-like manner with benzaldehyde.6,25
All of the steps up to the formation of anthranilate, the first
irreversible step, can contribute to kcat/Km. Thus, for the
simplified kinetic mechanism shown in eqn. (5), kcat/Km is given
by eqn. (6).
(5)
kcat/Km = kϩ1kϩ2/(kϪ1 ϩ kϩ2
)
(6)
In this mechanism the formation of the external aldimine and
quinonoid intermediates are combined in kϩ1, since both take
place within the deadtime of the stopped-flow instrument. It is
interesting that there is very little effect of bromine substitution
at either the 3- or 5-position on the rate of formation and decay
of the 494 nm quinonoid intermediate in stopped-flow kinetic
experiments (Figs. 1 and 2). The initial formation of the
external aldimine is very fast for the C-3 brominated com-
pound, since the quinonoid intermediate is still completely
formed in the dead time of the stopped-flow mixer, so it is
unlikely that the kϩ1 for external aldimine formation has been
decreased significantly. Furthermore, kϩ2, the formation of
anthranilate, is not significantly affected by the presence of
bromine, as determined by the stopped-flow experiments in the
presence of benzaldehyde, which directly measure the rate of
carbon–carbon bond cleavage by trapping the enamine.14
Thus, the off rate, kϪ1, for release of substrate from the external
aldimine must be increased at least 50-fold by the presence of
the 3-bromo substitutent in order to cause the observed kcat/Km
Fig. 3 A. Spectra of kynureninase in the presence of SAPCO and 4-
Br-SAPCO. The spectra are the fitted SVD spectra from the rapid-
scanning stopped-flow experiment. B. Time courses at 500 nm for the
reaction of SAPCO and 4-Br-SAPCO with kynureninase. Circles:
SAPCO; squares, 4-BrSAPCO. The lines are the fitted curves from the
exponential fits with k = 260 sϪ1
.
to decrease from 2.8 × 105 MϪ1 sϪ1 to 5.5 × 103 MϪ1 sϪ1
.
We demonstrated previously that steady-state turnover for
-kynurenine is limited by the slow release of -alanine from
the pyruvate ketimine (EK2 in Scheme 3);24 hence, it is not
surprising that all of the kynurenines examined exhibit similar
values of kcat (Table 1).
state, but decays after consumption of SAPCO, which is a very
slow substrate for β-elimination.8 4-Br-SAPCO forms a similar
absorption band, with λmax at 505 nm (Fig. 3B). The rate con-
stants of formation of the quinonoid intermediates are essen-
tially identical for both SAPCO and 4-Br-SAPCO, kobs = 260
We next reduced 5-bromo--kynurenine in order to make a
more active inhibitor for potential use in our X-ray structural
analysis. In previous work, we showed that the diastereoisomers
of dihydro--kynurenine differ in their properties, in that the
(4R)-epimer acts as a substrate (Km = 1.5 mM) and the (4S)-
epimer is a potent competitive inhibitor (Ki = 300 nM).6 The
behavior of the two diastereomers of 5-bromodihydro--
kynurenine parallels that of the unsubstituted dihydro--
kynurenines in that the (4R)-epimer acts as a substrate and the
(4S)-epimer does not. Surprisingly, in contrast to the minimal
effect on kcat/Km observed for substrate bromination at C-5,
both brominated dihydrokynurenines are significantly more
potent inhibitors than the parent compounds. Apparently, the
presence of bromine results in favorable dipolar or van der
Waals forces that enhance binding affinity of the dihydro--
kynurenines. Furthermore, the brominated (4R)-epimer is a
more potent inhibitor (Ki = 55 nM) than the (4S)-epimer (Ki =
170 nM), as shown in Table 1. Thus, there is a 27-fold increase
in the binding of the (4R)-epimer with 5-bromination, but only
a 1.8-fold increase in binding for the (4S)-epimer. This suggests
that the aromatic rings of the diastereomeric 5-bromodi-
hydrokynurenines have different modes of binding in the active
site.
s
Ϫ1 (compare curves in Fig. 3B).
Discussion
We initially prepared 3-bromo- and 5-bromo--kynurenine as
potential heavy atom derivatives to use in the determination of
the three-dimensional structure of P. fluorescens kynureninase
by X-ray crystallography.23 Our method of preparation of
3-bromokynurenine is more direct and efficient than the
previously reported method.20 We determined substrate activity
in solution to see if the binding of the brominated compounds
to kynureninase crystals was possible. We found that 5-bromo-
-kynurenine has kcat and kcat/Km values comparable to those of
-kynurenine (Table 1), while the kcat/Km value for 3-bromo--
kynurenine is 127-fold lower, primarily due to the very high
value of Km for the latter compound (2 mM). This result is
consistent with previous studies, where we demonstrated that
3-hydroxy--kynurenine has a kcat/Km value about 80-fold less
than that of -kynurenine.24 Thus, a bulky substitutent such as
bromo or hydroxy at the C-3 position is not well accommodated
by bacterial kynureninase, but a corresponding bromo substi-
tutent at C-5 has very little effect on substrate activity (Table 1).
This implies that there is adequate space in the active site of
bacterial kynureninase near the 5-position of bound -kynure-
nine to accommodate a bulky substituent, but not the 3-pos-
ition. The mechanism of kynureninase, based on previous
steady-state and stopped-flow experiments,14,24 is shown in
Scheme 3. The kynurenine quinonoid intermediate (EQ1) is
formed within the dead time of the mixer (ca. 1 ms), and decays
(S)-(2-Aminophenyl)--cysteine S,S-dioxide (SAPCO) was
previously reported to be a potent inhibitor of bacterial
kynureninase.8 We were interested to see if bromination would
also enhance the activity of this compound in the same way as it
does for dihydro--kynurenine. If so, we expected to observe a
Ki value in the low nM range for 5-Br-SAPCO. We prepared
pure 4-Br-SAPCO and a 2 : 1 mixture of the 4-Br and 5-Br
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 8 8 – 2 9 5
293

View Article Online
Scheme 3
isomers of SAPCO. From this, we were able to estimate that the
Ki values for these compounds are 300 and 400 nM, respectively
(Table 1), compared to 27 nM for the unsubstituted compound.
Thus, bromination at either C-4 or C-5 results in a 10-fold
reduction in binding affinity for SAPCO, rather than the
increase in binding seen with dihydro--kynurenines. This
suggests that the binding mode of the aromatic ring for these
compounds is different again from those for both the dihydro--
kynurenines and -kynurenine. This is surprising, since we
initially designed SAPCO as a transition state analog inhibitor
based on the properties of the dihydrokynurenines.6 However,
SAPCO accumulates a quinonoid intermediate (Fig. 3), while
the kynurenines only exhibit a transient quinonoid species
(Figs. 1 and 2). In our previous work, we applied COMFA to
develop a model of inhibitor and substrate binding to the active
site of kynureninase.8 This model required that the aromatic
rings of the substrate and different classes of inhibitors are
bound in the same way in the active site. However, the present
data clearly cannot be accommodated by a single mode of
binding of the aromatic rings of these compounds.
Unfortunately, attempts to observe the bound brominated
compounds in kynureninase crystals by soaking the crystals
with the inhibitor have been unsuccessful to date. The three
dimensional structure of bacterial kynureninase has been
et al. reported that 2-amino-4-(3Ј-hydroxyphenyl)-4-hydroxy-
butanoic acid, which is similar in structure to the dihydro-
kynurenines examined in this work, is a potent inhibitor of
human and rat kynureninase, with Ki values around 100 nM.27
It will be of interest to see if halogenation also results in an
increase in the inhibitory activity of dihydrokynurenines for
human kynureninase.
Acknowledgements
This work was partially supported by grant GM-42588 to R.S.P
from the National Institutes of Health.
References and notes
1 O. Hayashi and R. Y. Stanier, J. Bacteriol., 1951, 62, 691.
2 M. F. Beal, N. W. Kowall, D. W. Ellison, M. F. Mazurek,
K. J. Swartz and J. B. Martin, Nature, 1986, 321, 168.
3 M. P. Heyes, B. J. Brew, A. Martin, R. W. Price, A. M. Salazar,
J. J. Siditis, J. A. Yergey, M. M. Mouradian, A. E. Sadler and
J. Keilp, Ann. Neurol., 1991, 29, 202.
4 K. Saito, J. S. Crowley, S. P. Markey and M. P. Heyes, J. Biol. Chem.,
1993, 268, 15496.
5 R. Carpenedo, A. Chiarugi, P. Russi, G. Lombardi, V. Carlá,
R. Pellicciari, L. Mattoli and F. Moroni, Neuroscience, 1994, 61,
237.
6 R. S. Phillips and R. K. Dua, J. Am. Chem. Soc., 1991, 113, 7385.
7 A. E. Braunstein and M. M. Shemyakin, Biokhimiya, 1953, 18,
393.
8 R. K. Dua, E. W. Taylor and R. S. Phillips, J. Am. Chem. Soc., 1993,
115, 1264.
9 S. V. Koushik, B. Sundararaju, R. A. McGraw and R. S. Phillips,
Arch. Biochem. Biophys., 1997, 344, 301.
10 M. M. Bradford, Anal. Biochem., 1976, 72, 248.
11 M. Moriguchi and K. Soda, Biochemistry, 1973, 12, 2974.
12 W. W. Cleland, Methods Enzymol., 1979, 63, 103.
13 K. M. Plowman, Enzyme Kinetics, McGraw-Hill, New York, 1972,
p 18.
ultimately solved using
a combination of heavy atom
derivatives and molecular replacement.23
The results of this work show that bromine, and likely other
halogen, substitution on dihydrokynurenines can enhance the
kynureninase inhibitory activity significantly. The effects of the
bromo substituent may be due to dipolar effects, van der Waals
effects, or a combination of both, with active site residues. A
series of substitutents will need to be examined in order to
determine the nature of these effects. The inhibition of mam-
malian kynureninase by SAPCO and some substituted analogs
was reported by Drysdale and Reinhard.26 In that study, fluor-
ine or methoxy substitution in the 3-position resulted in slightly
improved inhibition, but the best activity (IC50 = 11 µM) was
observed with the 5-methyl derivative. More recently, Walsh
14 R. S. Phillips, B. Sundararaju and S. V. Koushik, Biochemistry, 1998,
255, 508.
15 I. B. C. Matheson, Comput. Chem., 1990, 14, 49.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 8 8 – 2 9 5
294

View Article Online
16 R. Ritchie and J. E. Saxton, Tetrahedron, 1981, 37, 4295.
17 G. Casnati, G. P. Gardini and G. Palla, J. Chem. Soc., Pekin Trans. 1,
1974, 2397.
18 M. Varasi, A. DellaTorre, F. Heidempergher, P. Pevarello and
P. Speciale, Eur. J. Med. Chem., 1996, 31, 11.
22 A. Mangini, Gazz. Chim. Ital., 1936, 66, 343.
23 V. Levdikov, L. Blagova, C. Momany, and R. S. Phillips,
unpublished results.
24 S. V. Koushik, J. A. Moore, B. Sundararaju and R. S. Phillips,
Biochemistry, 1998, 37, 1376.
19 G. Gellerman, A. Rudi and Y. Kashman, Tetrahedron, 1994, 50,
12959.
20 S. Röver, A. M. Cesura, P. Huguenin, R. Kettler and A. Szente,
J. Med. Chem., 1997, 40, 4378.
25 G. S. Bild and J. C. Morris, Arch. Biochem. Biophys., 1984, 235,
41.
26 M. J. Drysdale and J. F. Reinhard, Bioorg. Med. Chem. Lett., 1998,
8, 133.
21 The solution may contain residual peracid, which is an explosion
hazard and should not be heated above 30 ЊC during evaporation.
27 H. A. Walsh, P. L. Leslie, K. C. O’Shea and N. P. Botting, Bioorg.
Med. Chem. Lett., 2002, 12, 361.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 8 8 – 2 9 5
295