Stereospecificity of Pseudomonas fluorescens kynureninase for diastereomers of β-methylkynurenine

Article abstract of DOI:10.1016/S0968-0896(99)00088-7

The diastereomers of β-methyl-L-kynurenine were prepared by preparative ozonolysis of the respective diastereomers of β-methyl-L-tryptophan. A practical method for preparative enzymatic resolution of the diastereomers of β-methyltryptophan was developed using carboxypeptidase A digestion of the N-trifluoroacetyl derivatives. The stereochemical assignment was confirmed by X-ray crystal structure determination of (2S,3R)-threo-β-methyl-L-tryptophan. (2S,3S)-erythro-β-Methyl-L-kynurenine is a slow substrate for kynureninase from Pseudomonas fluorescens (k(cat)/K(m)=0.1% that of L-kynurenine), producing anthranilic acid, while (2S,3R)-threo-L-kynurenine is about 390-fold less reactive than erythro. Rapid-scanning stopped-flow measurements show that β-methyl substitution affects the rate of α-deprotonation of the L-kynurenine-pyridoxal-5'-phosphate Schiff's base. This is consistent with the stereoelectronic requirements of the reaction. These results are the first demonstration that β-substituted kynurenines can be substrates for kynureninase, and may be useful in the design of mechanism-based inhibitors. Copyright (C) 1999 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(99)00088-7

Bioorganic & Medicinal Chemistry 7 (1999) 1497±1503
Stereospeci®city of Pseudomonas ¯uorescens Kynureninase for
Diastereomers of ꢀ-methylkynurenine
Lakshmi V. Cyr, M. Gary Newton and Robert S. Phillips*
Department of Chemistry, Department of Biochemistry and Molecular Biology, and Center for Metalloenzyme Studies,
University of Georgia, Athens, GA 30602-2556, USA
Received 25 September 1998; accepted 8 December 1998
AbstractÐThe diastereomers of b-methyl-l-kynurenine were prepared by preparative ozonolysis of the respective diastereomers of
b-methyl-l-tryptophan. A practical method for preparative enzymatic resolution of the diastereomers of b-methyltryptophan was
developed using carboxypeptidase A digestion of the N-tri¯uoroacetyl derivatives. The stereochemical assignment was con®rmed by
X-ray crystal structure determination of (2S,3R)-threo-b-methyl-l-tryptophan. (2S,3S)-erythro-b-Methyl-l-kynurenine is a slow
substrate for kynureninase from Pseudomonas ¯uorescens (k /K =0.1% that of l-kynurenine), producing anthranilic acid, while
cat
m
(
2S,3R)-threo-l-kynurenine is about 390-fold less reactive than erythro. Rapid-scanning stopped-¯ow measurements show that b-
0
methyl substitution a€ects the rate of a-deprotonation of the l-kynurenine-pyridoxal-5 -phosphate Schi€'s base. This is consistent
with the stereoelectronic requirements of the reaction. These results are the ®rst demonstration that b-substituted kynurenines can
be substrates for kynureninase, and may be useful in the design of mechanism-based inhibitors. # 1999 Elsevier Science Ltd. All
rights reserved.
Introduction
Previously, we examined the interaction of P. ¯uor-
escens kynureninase with the diastereomers of dihy-
drokynurenine.11 We found that the (4R)-isomer of
dihydrokynurenine is a substrate for a retro-aldol clea-
vage reaction, while the (4S)-isomer is a potent compe-
l-Kynurenine is an intermediate in the catabolism of l-
tryptophan in plants, animals, and in some species of
bacteria, including Pseudomonas ¯uorescens. In bac-
1
teria, the hydrolytic cleavage of l-kynurenine to give
l-alanine and anthranilic acid is catalysed by kynur-
eninase [EC 3.7.1.3] (eq (1)). In plants and animals, a
similar enzyme preferentially cleaves 3-hydroxy-
titive inhibitor, with a K value of 0.3 mM. This potent
i
inhibition was suggested to be due to the similarity of
these compounds to a postulated gem-diolate inter-
mediate. These results prompted us to prepare and
evaluate a series of S-aryl-l-cysteine S,S-dioxides as
kynureninase inhibitors, due to their similar structures
to the proposed intermediate.12 The most potent of
these latter compounds, S-(2-aminophenyl)-l-cysteine
+
2
kynurenine in the biosynthetic pathway to NAD .
The kynurenine pathway in mammals ultimately results
in the biosynthesis of quinolinic acid, a neurotoxic
compound which has been implicated in the etiology of
a number of neurodegenerative diseases, including
S,S-dioxide, has a K value of 0.070 mM, providing fur-
i
3±10
Huntingdon's chorea and AIDS-related dementia.
ther evidence in support of a gem-diolate intermediate.
However, there have been no reports of the substrate or
inhibitory activity of kynurenines with b-substituents
with kynureninase. We have now prepared both threo
and erythro diastereomers of b-methyl-l-kynurenine,
and we have examined them as substrates and inhibitors
of kynureninase. We ®nd that the erythro isomer of b-
methyl-l-kynurenine exhibits signi®cantly greater sub-
strate activity with kynureninase than does the threo
isomer. This result is consistent with and provides
ꢀ
1†
additional support for our previously postulated
In the course of these
Key words: Kynureninase; b-methyl-l-kynurenine; b-methyl-l-trypto-
phan; pyridoxal-5 -phosphate; carboxypeptidase.
kynureninase mechanism.1
1±14
studies, we developed a procedure for the enzymatic
0
*
Corresponding author.
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(99)00088-7

1
498
L. V. Cyr et al. / Bioorg. Med. Chem. 7 (1999) 1497±1503
resolution of the threo and erythro diastereomers of
b-methyl-l-tryptophan.
5 mM. The reaction of kynureninase is conveniently
monitored by following the absorbance decrease at
3
60 nm due to loss of the ortho-aminophenylketone
chromophore. When we performed these measure-
ments with the b-methylkynurenines as substrates, we
found that (2S,3S)-erythro-b-methyl-l-kynurenine readily
showed evidence for reaction (Fig. 2). There is an iso-
sbestic point between the 364 nm peak of the b-
methylkynurenine and the 310 nm peak of the anthra-
nilic acid product, indicating that these are the only
species in the reaction mixture (Fig. 2). However, the
Results
Synthesis of ꢀ-methylkynurenines. The diastereomers of
b-methyl-dl-tryptophan were synthesized by a mod-
i®cation of the procedure of Snyder and Matteson.15
However, to our knowledge, an ecient procedure for
the enzymatic resolution of both b-methyltryptophan
diastereomers on a preparative scale has not been pre-
viously reported, although (2S,3R)-threo-b-methyl-
tryptophan is a natural product. The racemic threo and
erythro diastereomers of N-acetyl-b-methyltryptophan
were separated by crystallization, then subjected to acid
hydrolysis to obtain the respective racemic amino
K value of kynureninase for (2S,3S)-erythro-b-methyl-
m
l-kynurenine is about 4.8 mM, about 200-fold higher
1
1
than that for l-kynurenine, which is 25 mM. The k
cat
for (2S,3S)-erythro-b-methyl-l-kynurenine, 0.9 s 1, is
�
about 15% that of l-kynurenine; thus, kcat/Km is
�
1
� 1
5
� 1 � 1
s
195 M
s
, compared to 2Â10 M
for l-kynur-
1
5
13
acids. These were then converted to the N-tri¯uoro-
acetyl derivatives and resolved by carboxypeptidase A-
catalysed hydrolysis, as we have utilized before for a
enine. Hence, the catalytic eciency of kynureninase
has been reduced about 1000-fold by the presence of the
erythro b-methyl group. In contrast, (2S,3R)-threo-b-
methyl-l-kynurenine showed no change in UV spec-
trum under the reaction conditions of Fig. 2, even
with larger amounts of enzyme and with longer periods
of incubation. However, when reaction mixtures of
(2S,3R)-threo-b-methyl-l-kynurenine were analysed by
HPLC with ¯uorescence detection, time-dependent for-
mation of anthranilic acid was observed in the presence
of kynureninase. At a concentration of 2 mM (2S,3R)-
threo-b-methyl-l-kynurenine, the apparent rate con-
6
±8
number of other tryptophan analogues.
The asym-
metric synthesis of the four isomers of b-methyl-
tryptophan as the N -mesitylenesulfonyl protected
In
1
6
derivatives was reported by Boteju et al., but the pre-
paration of the free amino acids was not described. The
preparation of (2R,3S)-threo-b-methyltryptophan by an
asymmetric hydrogenation reaction has been recently
17
reported by Hoerner et al. Based on the rotation of
+
ꢀ
methyltryptophan of 91% ee, our observed rotation for
this product of +29.8 indicates that our material has
27.1 reported by Hoerner for (2R,3S)-threo-b-
�
1
stant of formation of anthranilic acid is 0.001 s
Assuming a K value of 5 mM, based on the K value of
.
ꢀ
m
i
greater than 99% ee. The stereochemical assignment of
the b-methyl-l-tryptophans was con®rmed by X-ray
crystal structure determination of (2S,3R)-threo-b-
methyl-l-tryptophan (Fig. 1). Both diastereomers of b-
methyl-l-kynurenine were then prepared by ozonolysis
of the respective b-methyl-l-tryptophans in 1 M HCl.
Under these conditions, the initial ozonolysis product,
N-formyl-b-methyl-l-kynurenine, hydrolyzes rapidly,
and the hydrochloride salt of b-methyl-l-kynurenine is
obtained by low pressure reverse phase column chro-
matography and evaporation.
the erythro isomer as a competitive inhibitor, kcat/Km
for threo-b-methyl-l-kynurenine can be estimated as
0.5 M
b-methyl-l-kynurenines is highly diastereospeci®c, 390-
fold greater for the erythro than the threo isomer.
�
1
� 1
s . Thus, the reaction of kynureninase with the
Rapid scanning stopped-¯ow kinetic studies of kynurenine
and ꢀ-methylkynurenines. When kynureninase is mixed
with 1 mM l-kynurenine, a quinonoid intermediate with
lmax at 494 nm is completely formed within the dead
Reaction of ꢀ-methylkynurenines with kynureninase.
Both isomers of b-methyl-l-kynurenine were found to
be weak competitive inhibitors, with K values of about
i
Figure 2. Reaction of (2S,3S)-erythro-b-methyl-l-kynurenine with
kynureninase. The reaction mixture contained 0.15 mM (2S,3S)-ery-
thro-b-methyl-l-kynurenine in 50 mM potassium phosphate, pH 7.8,
ꢀ
Figure 1. Structure of (2S,3R)-threo-b-methyl-l-kynurenine hydrate.
and 2.8 mM kynureninase at 25 C. Scans were taken every 15 min.

L. V. Cyr et al. / Bioorg. Med. Chem. 7 (1999) 1497±1503
1499
time of the stopped-¯ow instrument (ca. 2 ms) (Scheme
). Thus, the rate constant for deprotonation of the a-
no detectable formation of quinonoid intermediates
absorbing at 500 nm for the threo isomer, it is possible
that either a-deprotonation, external aldimine forma-
tion, or both, is rate-determining for the reaction of
(2S,3R)-threo-b-methyl-l-kynurenine.
1
carbon of the l-kynurenine external aldimine must be
�
1
greater than 1300 s . This quinonoid intermediate then
decays rapidly in a biphasic manner, with apparent rate
�
1 14
constants of 750 and 20 s
.
The rate determining step
in the reaction of l-kynurenine appears to be deproto-
nation of the pyruvate ketimine intermediate at C4 ,
0
13,14
Discussion
so there is very little of the quinonoid intermediate
observeable in the steady-state of the reaction. In con-
trast, when 2 mm (2S,3S)-erythro-b-methyl-l-kynur-
enine is mixed with kynureninase, a weakly absorbing
peak at 500 nm is formed at an apparent rate constant
Implications of diastereospeci®city for the reaction
mechanism
In the present study, we have found that kynureninase is
highly stereospeci®c for reaction with diastereomers of
b-methyl-l-kynurenine. Recent pH dependence results
suggest that only a single base is involved in the
�
1
of 0.05 s and persists in the steady-state (data not
shown). Thus, the apparent a-deprotonation rate
constant has been reduced more than 20,000-fold by
substitution of a methyl on the b-carbon of kynurenine.
Since the rate constant of quinonoid intermediate
formation is comparable in magnitude to the value
measured for kcat, a-deprotonation is likely to be at least
partly rate determining in the reaction of (2S,3S)-ery-
thro-b-methyl-l-kynurenine. When the rapid-scanning
stopped-¯ow experiment was performed in the presence
of benzaldehyde, an intermediate with an absorbance
peak forms at 496 nm, with a rate constant of
13
mechanism of kynureninase, in agreement with the
results of Palcic et al.18 Thus, all proton transfers must
occur from the same face of the PLP complex. A
mechanism for the reaction of b-methyl-l-kynurenines
with kynureninase is shown in Scheme 1. After removal
of the a-proton from the external aldimine to form the
0
quinonoid intermediate, C-4 is protonated by the con-
jugate acid to give a ketime intermediate (Scheme 1).
The hydrolytic cleavage of the b±g C� C bond then
takes place. b-Protonation of the enamine intermediate
subsequently forms a pyruvate ketimine. In the presence
of benzaldehyde, the enamine intermediate can be trap-
ped to form a stable quinonoid intermediate with
lmax=496 nm.14 The deprotonation of the pyruvate
�
3
� 1
8
.5Â10
s
intermediate forms in the presence of benzaldehyde,
(Fig. 3). With l-kynurenine, a similar
�
1 14
with a rate constant of 67.4 s
of this intermediate is about 8000-fold slower with
2S,3S)-erythro-b-methyl-l-kynurenine. In contrast, the
.
Hence, the formation
(
0
reaction of (2S,3R)-threo-b-methyl-l-kynurenine did not
show any measureable absorbance increase at 500 nm
over a period up to 2 min, consistent with the extremely
slow kcat value for this isomer. Additionally, no absor-
bance increase at 500 nm was seen for up to 4 h when
benzaldehyde was included in incubations. Since there is
ketimine intermediate at C-4 appears to be rate-deter-
14,15
mining in the reaction of l-kynurenine.
The major
kinetic e€ect of the addition of the b-methyl group
appears to be on the rate of formation of the quinonoid
intermediate. The b-g carbon±carbon bond must be
oriented perpendicular to the plane of the PLP-p-system,
Scheme 1.

1
500
L. V. Cyr et al. / Bioorg. Med. Chem. 7 (1999) 1497±1503
according to Dunathan's postulate,19 for the cleavage
reaction to proceed. Thus, there must be a rotation
around the a±b bond concomitant with a-deprotona-
tion to orient the aryl group for the subsequent hydra-
tion and cleavage. In the case of (2S,3R)-threo-b-
methyl-l-kynurenine, this rotation will eclipse the
methyl and the carboxylate of the substrate in the tran-
sition state, resulting in a large increase in the activation
energy for quinonoid intermediate formation (Scheme
Implications for the design of kynureninase inhibitors
Kynureninase is a potential target for the design of
drugs for the treatment of neurological disorders, due to
its role in the biosynthesis of quinolinic acid, a known
neurotoxin.3
±10
Our results demonstrate for the ®rst
time that b-substituted kynurenines, especially those
with the erythro con®guration, can be substrates for
kynureninase. Replacement of the b-methyl group with
other groups that can be activated during catalysis
may thus result in speci®c `suicide substrates' of
kynureninase.
2
). Based on the di€erences in kcat/Km for l-kynurenine
and (2S,3R)-threo-b-methyl-l-kynurenine, there is a
5
di€erence in catalytic eciency of about 2Â10 . This
corresponds to an activation energy di€erence of
7.2 kcal/mol at 25 C. In a simple model, 2-ethylacrylic
ꢀ
acid, the conformation in which the methyl group is
eclipsed with the carboxylic acid is 5.3 kcal/mol higher
in energy than the gauche conformation, based on
molecular mechanics calculations. For (2S,3S)-erythro-
b-methyl-l-kynurenine, there are no strongly unfavor-
able eclipsing interactions in the transition state for a-
deprotonation. However, the transition state strain due
to the presence of the gauche methyl group still drama-
tically reduces the rate of a-deprotonation. This could
be due to steric restriction of the enzyme active site.
Experimental
General
Melting points are uncorrected. Column chromato-
graphy was performed on silica gel (70±230 mesh). All
reactions with air and moisture sensitive compounds
were conducted in oven dried glassware under an
atmosphere of dry nitrogen.
Instruments
1
H NMR spectra were recorded at 300 MHz with TMS
as an internal standard, and 13C NMR spectra were
recorded at 75 MHz with CDCl as an internal stan-
3
dard, using a Bruker AC-300 spectrometer. Optical
rotations were measured with an Autopol IV from
Rudolph Research. UV/visible spectra and enzyme
assays were performed with a Cary 1 spectro-
photometer. Rapid-scanning stopped-¯ow studies were
performed on an RSM instrument from OLIS, Inc.
equipped with a stopped-¯ow mixer, as described else-
where.14 High pressure liquid chromatography (HPLC)
was performed on an instrument with two Gilson 502
pumps and a Gilson ®lter ¯uorometer detector, using
Figure 3. (A) Rapid-scanning stopped-¯ow spectra of the reaction of
1
mM (2S,3S)-erythro-b-methyl-l-kynurenine and 10 mM benzalde-
hyde with 7 mM kynureninase in 50 mM potassium phosphate, pH
.8. Scans were collected at 1 s intervals. The scans shown were col-
lected at (1) 3 s, (2) 120 s, (3) 240 s, (4) 480 s, (5) 960 s, and (6) 1800 s.
B) Time course of the reaction at 496 nm.
7
(
Scheme 2.

L. V. Cyr et al. / Bioorg. Med. Chem. 7 (1999) 1497±1503
1501
Gilson Unipoint software. The excitation ®lter had an
upper cut-o€ of 350 nm and the emission ®lter had a
bandpass of 390±600 nm. Molecular mechanics calcula-
tions were performed using SPARTAN (Wavefunction,
Inc.) running on the Origin2000 in the Research Com-
puting Resource of the University of Georgia. X-ray
di€raction data were collected on an Enraf±Nonius
CAD4 di€ractometer with graphite monochromated
Cu-Ka radiation.
simultaneously. The reaction mixture was stirred over-
night at ambient temperature. The solution was then
poured into water and acidi®ed to pH 1 with con-
ꢀ
centrated HCl at 0 C. The product was extracted with
ethyl acetate and dried over anhydrous MgSO . The
4
solution was ®ltered and concentrated to obtain 4.6 g
ꢀ
(91%) of crude product (mp 166±168 C). IR (KBr pel-
let) n 3415, 3399, 3306, 2974, 1700, 1552, 1459, 1433,
1
1198, 1168, 741, 676. H NMR (CDCl -DMSO) d 1.15±
3
1
.18 (d, J=7.23 Hz), 3.55±3.65 (m, 1H), 4.49±4.54 (dd,
J=8.66, 4.59, 1H), 6.71±6.85 (m, 3H), 7.06±7.09 (d,
J=8.04 Hz, 2H), 7.36±7.39 (d, J=7.73 Hz), 10.2 (s, 1H);
Enzymes
13
Kynureninase was isolated from cells of Pseudomonas
uorescens grown on 0.2% l-tryptophan as described
C NMR (CDCl -DMSO) d 170.5, q (155.9, 155.4),
3
¯
135.3, 125.3, 121.2, 120.1, 117.7, 117.6, 113.2, 110.5,
56.2, 31.9, 17.2.
1
1
previously, or from E. coli DH5a containing the plas-
mid pTZKYN. Carboxypeptidase A was obtained
2
0
from Sigma as a crystalline slurry in H O.
2
(2S,3R)-2-Amino-3-(3-indolyl)-butanoic acid (ꢀ-methyl-
L-tryptophan, isomer A). N-Tri¯uoroacetyl-b-methyl-
dl-tryptophan (4.9 g, 15.6 mmol) was suspended in
water (25 mL) and the pH of the solution was adjusted
to 7.5 with triethylamine. Carboxypeptidase A (1.2 mL)
was then added and the reaction was stirred for 16 h.
The pH of the reaction mixture was then adjusted to
1 with concentrated H SO . The unreacted N-tri-
Kinetic measurements
The kynureninase activity was determined by follow-
ing the decrease in absorbance at 360 nm (Áe=
�
1
� 1
11,13
�
4500 M cm ) as previously described.
thro-b-methylkynurenine, activity was determined at
00 nm due to the high Km value which led to absor-
For ery-
2
4
4
¯uoroacetyl-b-methyl-d-tryptophan was extracted with
three 20 mL portions of ethyl acetate. The organic por-
tion was dried, ®ltered and concentrated to provide 2.2 g
of crude material. The water layer was also con-
centrated to dryness and redissolved in a minimum
volume of water. The pH was then adjusted to 4.8±5.
The b-methyl-l-tryptophan was obtained as white crys-
bances greater than 3 at 360 nm. For the stopped-¯ow
measurements, kynureninase was preincubated for
3
0 min with excess PLP, then passed through a small
desalting column (Pharmacia, PD-10) equilbrated with
.05 M potassium phosphate, pH 7.8 to remove excess
0
PLP. The stopped-¯ow reactions were then performed
in 0.05 M potassium phosphate, pH 7.8 with the syr-
ꢀ
tals, 1.37 g (81%), mp 266±268 C. IR (KBr pellet) n
ꢀ
inges in a water bath held at 25 C.
3436, 3334, 3096, 2975, 2025, 1633, 1585, 1485, 1450,
1
1
388, 1349, 1220, 1159, 1085, 968, 906, 749. H NMR
2
¯
-Tri¯uoroacetamido-3-(3-indolyl)-butanoic acid (N-tri-
uoroacetyl-ꢀ-methyltryptophan, isomer A). 3.5 g
(D O+DCl) d 0.47±0.5 (d, 3H, J=7.44 Hz), 2.86±2.94
2
(m, 1H), 3.45±3.47 (d, 1H, J=4.31 Hz), 6.21±6.34 (m,
2H), 6.36 (s, 1H), 6.56±6.59 (d, 1H, J=8.14 Hz), 6.69±
15
(
16.1 mmol) b-methyl-dl-tryptophan (A-isomer), was
dissolved in dry methanol (15 mL) in a 50 mL round
bottomed ¯ask. Then, triethylamine (2.3 g, 22.6 mmol)
and ethyl tri¯uoroacetate (2.5 g, 18.1 mmol) were simul-
taneously added. The reaction mixture was stirred
overnight at ambient temperature. The solution was
poured into water and acidi®ed to pH 1 with con-
6.72 (d, 1H, J=7.73 Hz); 13C NMR (D O+DCl) d
170.6, 136.1, 124.9, 123.8, 121.9, 119.1, 117.9, 111.9,
2
2
4
ꢀ
Anal. calcd for C H N O .H O: C, 61.0, H, 6.83, N,
111.8, 56.8, 31.0, 13.4. [a] � 30.6 (c 1.69, 0.1 M HCl).
d
12
14
2
2
2
11.86. Found: C, 60.89, H, 6.86, N, 11.78.
ꢀ
centrated HCl at 0 C. The product was extracted with
ethyl acetate and dried over anhydrous MgSO . The
(2R,3S)-2-Amino-3-(3-indolyl)-butanoic acid (ꢀ-methyl-
D-tryptophan, isomer A). The unreacted N-tri¯uoro-
acetyl derivative (2.2 g, 7 mmol) obtained from the
above reaction was suspended in 10 mL of water.
Sodium hydroxide was then added (1.2 equiv, 8.4 mmol,
0.336 g). The reaction was stirred for 18 h. The reaction
was stopped by adjusting the pH 4.6±5.0 with 4 N sul-
furic acid. The product which crystallized from the
solution was ®ltered and dried to provide 1.3 g (85%),
4
solution was ®ltered and concentrated to obtain 4.2 g
ꢀ
(
85%) of crude product (mp 160±163 C). IR (KBr pel-
let) n 3395, 31114, 3070, 2973, 1745, 1719, 1689, 1543,
458, 1414, 1281, 1222, 1168, 1105, 1008, 839, 748, 674.
1
1
H NMR (CDCl -DMSO) d 0.81±0.83 (d, 3H, J=
3
6
6
.86 Hz), 2.47±3.19 (m, 1H), 4.13±4.19 (m, 1H), 6.38±
.51 (m, 3H), 6.74±6.77 (d, 1H, J=7.92 Hz), 6.97±6.99
ꢀ
(
(
d, 1H, J=7.58 Hz), 8.37±8.40 (d, 1H, J=8.51 Hz), 9.96
1
mp 246±248 C. IR (KBr pellet) n 3618, 3417, 3073,
2970, 2612, 2438, 1602, 1534, 1399, 1172, 1093, 921,
3
s, 1H); C NMR (CDCl -DMSO) d 170.6, q (156.0,
3
1
1
55.5, 154.8), 152.2, 125.4, 121.2, 120.2, 117.6, 117.3,
799, 751; H NMR (D O+DCl) d 0.94±0.96 (d, 3H,
2
1
14.1, 110.6, 56.7, 31.5, 15.9.
J=6.92 Hz), 3.38±3.48 (m, 1H), 3.90±3.92 (d, 1H,
J=4.37 Hz), 6.61±6.76 (m, 2H), 6.81 (s, 1H), 6.98±
7.01(d, 1H, J=8.32 Hz), 7.16±7.19 (d, 1H, J=7.67 Hz);
2
-Tri¯uoroacetamido-3-(3-indolyl)-butanoic acid (N-
tri¯uoroacetyl-ꢀ-methyltryptophan, isomer B). 3.5 g
13
C NMR (D O) d 170.9, 136.4, 125.2, 124.0, 122.2,
2
1
5
24
ꢀ
(
16.1 mmol) b-methyl-dl-tryptophan (B-isomer) was
119.4, 118.2, 122.2, 112.0, 56.9, 31.3, 13.6. [a]d +29.9
(c 1.66, 0.1 M HCl). Anal. calcd for C H N O .H O:
dissolved in dry methanol (15 mL) in a 50 mL round
bottomed ¯ask. Then, triethylamine (2.3 g, 22.6 mmol)
and ethyl tri¯uoroacetate (2.5 g, 18.1 mmol) were added
12
14
2
2
2
C, 61.0, H, 6.83, N, 11.86. Found: C, 60.91, H, 6.83, N,
11.86.

1
502
L. V. Cyr et al. / Bioorg. Med. Chem. 7 (1999) 1497±1503
(
2S,3S)-2-Amino-3-(3-indolyl)-butanoic acid (ꢀ-methyl-
(m, 1H), 7.64±7.67 (d, J=8.27 Hz, 1H); 13C NMR
(DMSO) d 205.9, 173.0, 153.9, 136.2, 132.2, 118.7,
116.3, 115.9, 57.4, 35.9, 12.9. MS (ESI) m/z 223
L-tryptophan isomer B). N-Tri¯uoroacetyl-b-methyl-dl-
tryptophan, isomer B, (6.8 g, 21.7 mmol) was suspended
in water (30 mL) and the pH of the solution was adjus-
ted to 7.5 with triethylamine. Carboxypeptidase A
+
(M+H ). UV (H O), l
227 nm (log e=4.29),
2
max
290 nm (log e=3.82), 364 nm (log e=3.61).
(
1
1.5 mL) was then added and the reaction was stirred for
6 h. The pH of the reaction solution was then adjusted
to 1 with concentrated H SO . The unreacted N-tri-
(3S,2S)-erythro-ꢀ-Methyl-L-kynurenine hydrochloride
(B isomer). The b-methyl-l-tryptophan B-isomer (0.3 g,
1.38 mmol) was dissolved in 10 mL of 1 M HCl, and
ozone was passed through the solution in a slow stream.
The reaction was followed using UV-spectroscopy
monitoring the decrease in indole absorbance at
280 nM. When the reaction was complete, the solution
was evaporated to dryness. The crude product was
chromatographed on a low pressure C18 reverse-phase
silica gel column in aqueous methanol and evaporated
to give a light-yellow solid (0.19 g, 53%). IR (KBr pel-
let) n 3486, 3464, 3366, 3340, 3162, 2853, 2572, 2019,
2
4
¯
uoroacetyl-b-methyl-d-tryptophan was extracted with
three 20 mL portions of ethyl acetate. The organic por-
tion was dried, ®ltered and concentrated to provide 3.2 g
of crude material. The water layer was then con-
centrated down to dryness and redissolved in a mini-
mum volume of water. The pH was then adjusted to
4
.8±5.0. The b-methyl-l-tryptophan was obtained as
ꢀ
white crystals, 1.27 g (66%), mp 244±246 C. IR (KBr
pellet) n 3400, 3053, 1617, 1458, 1408, 1321, 1246, 1227,
1
1186, 1098, 1014, 818, 764, 741. H NMR (D O+DCl)
2
1
d 0.64±0.61 (d, 3H, J=7.34 Hz), 3.011±2.90 (m, 1H),
.42±3.39 (d, 1H, J=4.95 Hz), 6.43±6.25 (m, 2H), 6.46
s, 1H), 6.66±6.63 (d, 1H, J=8.60 Hz), 6.78±6.75 (d, 1H,
1652, 1576, 1484, 1388, 1553, 1216, 1159, 979, 744. H
3
(
NMR (D O) d 7.56±7.59 (d, 1H, J=7.78 Hz), 7.12±7.17
2
(m, 1H), 7.01±7.06 (m, 1H), 6.89±6.92 (d, 1H, J=8 Hz),
3.84±3.91 (m, 1H), 3.77±3.79 (d, 1H, J=4.6 Hz), 0.75±
1
3
J=7.96 Hz); C NMR (D O+DCl) d 170.7, 135.8,
2
1
25.1, 123.8, 121.7, 119.0, 118.1, 111.6, 111.2, 57.1, 31.5,
0.77 (d, 3H, J=7.5 Hz); 13C NMR (D O) d 203.8, 170.1,
135.6, 131.7, 130.0, 129.8, 126.3, 125.5, 54.0, 41.5, 13.4.
2
2
4
ꢀ
C H N O 2/5H O:C, 63.93, H, 6.62, N, 12.42 Found:
16.8. [a] +47.6 (c 1.70, 0.1 M HCl). Anal. calcd for
d
+
MS (ESI) m/z 223 (M+H ). UV (H O), l
227 nm
1
2
14
2
2
2
2
max
C, 63.95, H, 6.59, N, 12.47.
(log e=4.19), 290 nm (log e=3.72), 364 nm (log
e=3.58).
(
2R,3R)-2-Amino-3-(3-indolyl)-butanoic acid (ꢀ-methyl-
D-tryptophan, isomer B). The unreacted N-tri¯uroacetyl
derivative (3.2 g, 10.2 mmol) obtained from the above
reaction was suspended in 20 mL of water followed by
the addition of 1.2 equiv (12.2 mmol, 0.489 g) of sodium
hydroxide. The reaction was stirred for 18 h. The reac-
tion was stopped by adjusting the pH 4.6±5 with 4 N
H SO . The product crystallised out upon cooling to
Acknowledgements
This work was partially supported by a grant from the
National Institutes of Health (GM42588) and by a
contract with Pharmacia.
2
4
ꢀ
give 1.9 g (85%), mp 241±243 C. IR (KBr pellet) n 3396,
3
052, 2936, 2495, 2032, 1615, 1458, 1409, 1320, 1227,
1
References
1186, 1098, 1021, 922, 818, 742; H NMR (D O+DCl)
2
d 0.89±0.92 (d, 3H, J=7.33 Hz), 3.20±3.31 (m, 1H),
.67±3.69 (d, 1H, J=5.14 Hz), 6.49±6.65 (m, 2H), 6.72
s, 1H), 6.87±6.91 (d, 1H, J=8.29 Hz), 7.02±7.05 (d, 1H,
1
. Hayaishi, O.; Stanier, R. Y. J. Bacteriol. 1951, 62, 691.
. Soda, K.; Tanizawa, K. Advances Enzymol. Rel. Areas Mol.
3
(
2
Biol. 1979, 49, 1.
3. Sei, S.; Saito, K.; Stewart, S. K.; Crowley, J. S.; Brouwers,
P.; Kleiner, D. E.; Katz, D. A.; Pizzo, P. A.; Heyes, M. P. J.
Infect. Dis. 1995, 172, 638.
1
3
J=7.95 Hz); C NMR (D O-DCL) d 171.1, 136.1,
2
1
25.4, 124.1, 122.1, 119.3, 118.4, 111.9, 111.6, 57.4, 31.9,
2
4
ꢀ
C H N O .1/2.H O:C, 63.42, H, 6.65, N, 12.33.
17.1. [a] � 49.0 (c 1.65, 0.1 M HCl). Anal. calcd for
d
4. Sei, Y.; Paul, I. A.; Saito, K.; Layar, R.; Hartley, J. W.;
Morse, H. C. R.; Skolnick, P.; Heyes, M. P. J. Neurochem.
1
2
14
2
2
2
Found: C, 63.46, H, 6.37, N, 12.36.
1
5
996, 66, 296.
. Saito, K.; Crowley, J. S.; Markey, S. P.; Heyes, M. P. J.
(
3R,2S)-threo-ꢀ-Methyl-L-kynurenine hydrochloride (A
isomer). The b-methyl-l-tryptophan A-isomer (0.3 g,
.38 mmol) was dissolved in 10 mL of 1 M HCl, and
Biol. Chem. 1993, 268, 15496.
. Saito, K.; Nowak, T. S., Jr.; Markey, S. P.; Heyes, M. P. J
Neurochem 1993, 60, 180.
. Saito, K.; Nowak, T. S., Jr.; Suyama, K.; Quearry, B. J.;
6
1
ozone was passed through the solution in a slow stream.
The reaction was followed using UV-spectroscopy,
monitoring the indole absorbance at 280 nm. When the
reaction was complete, the reaction mixture was evapo-
rated to dryness. The crude product was chromato-
graphed on a low pressure C18 reverse-phase silica gel
column in aqueous methanol to provide 0.19 g (53%) of
light-yellow solid after evaporation. IR (KBr pellet) n
7
Saito, M.; Crowley, J. S.; Markey, S. P.; Heyes, M. P. J.
Neurochem 1993, 61, 2061.
8. Heyes, M. P.; Brew, B.; Martin, A.; Markey, S. P.; Price, R.
W.; Bhalla, R. B.; Salazar, A. Adv. Exp. Med. Biol. 1991, 294,
6
9
87.
. Heyes, M. P.; Papagapiou, M.; Leonard, C.; Markey, S. P.;
Auer, R. N. Adv. Exp. Med. Biol. 1991, 294, 679.
1
1
1
0. Achim, C. L.; Heyes, M. P.; Wiley, C. A. J. Clin. Invest.
993, 91, 2769.
1. Phillips, R. S.; Dua, R. K. J. Am. Chem. Soc. 1991, 113,
3
436, 3334, 3096, 2974, 2025, 1633, 1585, 1485, 1450,
1
1388, 1349, 1220, 1159, 1085, 968.5, 906.6, 749.8. H
NMR (D O+DCl) d 1.02±1.05 (d, J=7.59 Hz, 3H),
2
7385.
12. Dua, R. K.; Taylor, E. W.; Phillips, R. S. J. Am. Chem.
Soc. 1993, 115, 1264.
3.89±3.91 (d, J=3.43 Hz, 1H), 4.02±4.11 (m, 1H), 6.57±
6.64 (m, 1H), 6.66±6.69 (d, J=8.3 Hz, 1H), 7.17±7.24

L. V. Cyr et al. / Bioorg. Med. Chem. 7 (1999) 1497±1503
1503
1
3. Koushik, S. V.; Moore, J. A., III; Sundararaju, B.; Phillips,
R. S. Biochemistry 1998, 37, 1376.
4. Phillips, R. S.; Sundararaju, B.; Koushik, S. V. Biochem-
istry 1998, 37, 8783.
17. Hoerner, R. S.; Askin, D.; Volante, R. P.; Reider, P. J.
Tetrahedron Lett. 1998, 39, 3455.
18. Palcic, M. M.; Antoun, M.; Tanizawa, K.; Soda, K.; Floss,
H. J. Biol. Chem. 1985, 260, 5248.
1
1
7
1
5. Snyder, H. R.; Matteson, D. S. J. Am. Chem. Soc. 1957,
9, 2217.
6. Boteju, L. W.; Wegner, K.; Qiuan, X.; Hruby, V. J. Tet-
19. Dunathan, H. C. Adv. Enzymol. Rel. Areas Mol. Biol.
1971, 35, 79.
20. Koushik, S. V.; Sundararaju, B.; Phillips, R. S. Arch. Bio-
chem. Biophys. 1997, 344, 301.
rahedron 1994, 50, 2391.