Phenylalanine ammonia-lyase: The use of its broad substrate specificity for mechanistic investigations and biocatalysis - Synthesis of L-arylalanines

Article abstract of DOI:10.1002/1521-3765(20000915)6:18<3386::AID-CHEM3386>3.0.CO;2-5

Several fluoro-and chlorophenylalanines were found to be good substrates of phenylalanine ammonialyase (PAL/EC 4.3.1.5) from parsley. The enantiomerically pure L-amino acids were obtained in good yields by reaction of the corresponding cinnamic acids with 5M ammonia solution (buffered to pH 10) in the presence of PAL. The kinetic constants for nine different fluoro-and chlorophenylalanines do not provide a rigorous proof for but are consistent with the previously proposed mechanism comprising an electrophilic attack of the methylidene-imidazolone cofactor of PAL at the aromatic nucleus as a first chemical step. In the resulting Friedel-Crafts-type σ complex the β-protons are activated for abstraction and consequently the pro-S is abstracted by an enzymic base. Results from semi-empirical calculations combined with a proposed partial active site model showed a correlation between the experimental kinetic constants and the change in polarization of the pro-S C_β-H bond and heat of formation of the σ complexes, thus making the electrophilic attack at the neutral aromatic ring plausible. Furthermore, while 5-pyrimidinylalanine was found to be a moderately good substrate of PAL, 2-pyrimidinylalanine was an inhibitor.

Full text of DOI:10.1002/1521-3765(20000915)6:18<3386::AID-CHEM3386>3.0.CO;2-5

FULL PAPER
Phenylalanine Ammonia-Lyase: The Use of Its Broad Substrate Specificity for
Mechanistic Investigations and BiocatalysisÐSynthesis of l-Arylalanines
[
a]
[b]
Â
[c]
[c, d]
and J a nos R e tey*[a]
Andreas Gloge, Jerzy Zo n , Agnes Köv a ri, L a szl o Poppe,
Abstract: Several fluoro- and chloro-
phenylalanines were found to be good
substrates of phenylalanine ammonia-
lyase (PAL/EC 4.3.1.5) from parsley.
The enantiomerically pure l-amino
acids were obtained in good yields by
reaction of the corresponding cinnamic
acids with 5m ammonia solution (buf-
fered to pH 10) in the presence of PAL.
The kinetic constants for nine different
fluoro- and chlorophenylalanines do not
provide a rigorous proof for but are
consistent with the previously proposed
mechanism comprising an electrophilic
attack of the methylidene-imidazolone
cofactor of PAL at the aromatic nucleus
as a first chemical step.In the resulting
Friedel ± Crafts-type s complex the b-
protons are activated for abstraction and
consequently the pro-S is abstracted by
an enzymic base.Results from semi-
empirical calculations combined with a
proposed partial active site model
showed a correlation between the ex-
perimental kinetic constants and the
change in polarization of the pro-S
C ÀH bond and heat of formation of
b
the s complexes, thus making the elec-
trophilic attack at the neutral aromatic
ring plausible.Furthermore, while 5-pyr-
imidinylalanine was found to be a mod-
erately good substrate of PAL, 2-pyri-
midinylalanine was an inhibitor.
Keywords: chemoenzymatic synthe-
sis ´ enzyme inhibitors ´ enzyme
mechanism ´ halogenated l-phenyl-
alanines ´ lyases
Introduction
the proteins afforded radioactive products which could be
derived from a dehydroalanine residue being the electrophilic
prosthetic group.[
3, 4, 5]
Phenylalanine ammonia-lyase (PAL) catalyzes the reversible
conversion of l-phenylalanine to trans-cinnamic acid.PAL is
the key enzyme in the metabolism of phenylpropanoids in
Its precursors have been found by site-directed mutagenesis
to be Ser202[ and Ser143 in PAL and HAL, respectively.
Since both enzymes have been overexpressed in E. coli cells in
active forms, the post-translational modification takes place
autocatalytically.[
6]
[7]
plants, since cinnamic acid is the precursor of lignins,
flavonoids, and coumarins.[
1, 2]
Early biochemical evidence
8, 9]
indicated a catalytically essential electrophilic group at the
active sites of PAL and the related enzyme histidine
ammonia-lyase (HAL).Inactivation by radiolabelled nucle-
Recently, the 3D structure of HAL has been elucidated by
X-ray analysis which revealed that the electrophilic prosthetic
14
3
[10]
ophiles (K CN and NaB H ) followed by total hydrolyses of
group is methylidene-imidazolone (MIO).
MIO can be
4
regarded as a modified dehydroalanine.The ring structure
and the geometry dictated by the protein conformation
prevents the lone pairs of the imidazole nitrogens to deloc-
alize into the a/b-unsaturated carbonyl system, thus enhanc-
ing the electrophilicity of the latter.The driving force for the
formation of MIO from the tripeptide Ala142Ser143Gly144 is
[
a] Dr.A.Gloge, Prof.Dr.J.Re Âtey
Institut für Organische Chemie der Universität Karlsruhe
Richard Willstätter Allee 2, 76128 Karlsruhe (Germany)
Fax : (‡49)721-608-4823
E-mail: biochem@ochhades.chemie.uni-karlsruhe.de
unknown but not without precedence.The chromophoric
[
b] Dr.J.Zon Â
imidazolone of the green fluorescent protein[
formed by a similar autocatalytic process.
11, 12, 13]
must be
Institute of Organic Chemistry, Biochemistry and Biotechnology
Wroclaw University of Technology Wybzeze Wyspianskiego 27
5
Â
0-370 Wroclaw (Poland)
The discovery of the electrophilic prosthetic group was
followed by the proposal that it reacts with the amino group of
the substrate in a type of Michael addition. It was suggested
that this reaction would enhance the leaving ability of the
amino group, that is facilitate the elimination of ammonia, but
it leaves the question open how the ªnon-acidic b-protonº
could be abstracted by an enzymic base.
[
c] A.KövaÂri, Dr.L.Poppe
Institute for Organic Chemistry
Budapest University of Technology and Economics
[14]
1
111 Budapest, Gell e rt t e r 4 (Hungary)
[
d] Dr.L.Poppe
Chemical Research Center of the Hungarian Academy of Sciences
H-1025 Budapest, Pusztaszeri u t 59-67 (Hungary)
3386
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3386 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 18

3
386±3390
Recently, experimental evidence accumulated in favor of
trations of ammonia.[18] For preparative conversions the
halogenated cinnamic acids were solved in half concentrated
ammonia solution whose pH was brought to 10 by bubbling
an electrophilic attack at the aromatic nucleus as shown in
Scheme 1 for the PAL reaction.[
15, 16]
In the Friedel ± Crafts-
type s complex the b-protons of the side chain are activated
for abstraction by an enzymic base.The proton transfer is then
followed by ammonia elimination concomitant with restora-
tion of the prosthetic group and the aromaticity of the six-
membered ring.
CO into the solution.The reaction was started by addition of
2
recombinant PAL (1 ± 2 iU).After incubation overnight at
378C the enzyme was removed by boiling and acidification
(pH 1.5) followed by filtration. Chromatography on an acidic
cation exchange column afforded the enantiomerically pure
substituted phenylalanines in moderate to excellent yields.
Their ee was determined by chromatography on a chiral
column to be more than 99%.The separability of the
enantiomeric phenylalanine derivatives was checked also on
chiral thin layer plates.The halogenated cinnamic acids and
phenylalanines were characterized by their NMR-, and UV
spectra, as well as their R values.Their structures together
f
with the yields of the isolated phenylalanines are shown in
Scheme 2.
Scheme 1.Postulated mechanism, partial putative active site/substrate,
and active site/s-complex models for the reaction catalyzed by phenyl-
alanine ammonia-lyase.
The question is legitimate, why the b-proton is abstracted
leading to an exocyclic double bond while abstraction of the
ring proton would straightforwardly lead to re-aromatization.
Although the latter would preferentially take place in
solution, PAL prevents this reaction by excluding any basic
group in the binding site for the phenyl group.Simultaneously,
a basic group is correctly positioned to abstract the pro-S b-
proton of the substrate.
In the present communication we describe the enantiospe-
cific synthesis of various fluoro- and chlorophenylalanines by
reversal of the PAL reaction.Kinetic measurements with
these new substrates and the results from theoretical calcu-
lations and a proposed partial active site model are discussed
in terms of the Friedel ± Crafts-type mechanism of the PAL
reaction.
Scheme 2.Preparative conversion of various halogenated phenylalanines
by reversal of the phenylalanine ammonia-lyase reaction.
Determination of the kinetic constants for the halogenated
phenylalanines: The prerequisits for the quantitative meas-
urement of the concentrations of the cinnamic acid products
was the determination of their extinction coefficients (e) at
the wavelengths at which a maximal difference existed
between those of the arylalanines and the corresponding
cinnamic acids (see Experimental Section).The values were
Results and Discussion
Enantioselective synthesis of l-phenylalanines halogenated in
the ring: Contrary to early reports that the PAL reaction is
irreversible[ it can be reversed by applying high concen-
17]
Chem. Eur. J. 2000, 6, No. 18
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3387 $ 17.50+.50/0
3387

J.ReÂtey et al.
FULL PAPER
measured in 0.1m Tris-buffer pH 8.8. These were the con-
ditions also for the kinetic measurements.In columns 4 and 5
of Table 1 the Vmax values relative to that of phenylalanine and
the activating effect on the b-protons by the positive charge in
the ring were carried out.To facilitate the problem, semi-
empirical calculations were applied to simplified models of
the putative s complexes containing the substrates in their
zwitterionic form.In these studies two types of models were
investigated.Simple models contained a methyl group which
corresponds to the carbon of the methylidene moiety of MIO
(Scheme 3), while the more sophisticated models were
derived from a putative partial active site model (Scheme 1).
Conformational analysis of the simple model from l-Phe
indicated that the most decisive conformation factor for the
molecular properties is the dihedral angle (F) between the
C À C and the pro-S C ÀH bonds (Scheme 3).Results from
the K values, respectively, are listed.
m
Inspection of the relative Vmax values reveals that phenyl-
alanines halogenated in the 3'-position react on PAL signifi-
cantly faster than the parent compound.This is consistent
with an electrophilic attack of MIO at the aromatic nucleus.A
halogen substituent in 3'-position facilitates such an attack in
the ortho- or para position by stabilisation of the cationoid
transition state.Both positions are available for the attack by
the sterically fixed MIO due to free rotation of the phenyl ring
before substrate binding.This effect is similar to that
previously found with a 3'-hydroxyphenylalanine (m-tyrosine)
as substrate.[ Halogen substituents in the phenyl ring may
also influence the acidity of the b-protons of the side chain.
While this effect should activate the b-protons in the neutral
ground state, the opposite effect is expected for the cationoid
intermediate.In the latter, delocalization of the positive
charge to the halogen atoms will decrease electron deficiency
in the ring and hence diminish the acidity of the side chain b
protons.The acidifying effect of the halogen substituents is,
however, much less than that of a nitro group.Apart from
that, the b-proton abstraction does not seem to be the rate
1
'
2'
b
semiempirical calculations are compiled in columns 2 and 3 of
Table 1.
15]
With respect to the s complex formation and scissible bond
polarization the 2'F, 2'Cl and 4'Cl substituted compounds,
particularly the B-type sigma complexes are similar to the
unsubstituted l-Phe.Accordingly, their kinetic behavior is
similar to that of the natural substrate.
The putative active site model suggests that the aromatic
ring of the substrate is sandwiched between MIO and a phenyl
ring of Phe399.The arrangement as shown in Scheme 1 allows
maximal overlap between the p system of MIO and the
determining step for good substrates as shown by deuterium
aromatic C position.The 2 ' H of the aromatic moiety might
2
labelling.[
19]
While the attack by MIO is partially rate
be stabilized in a charge transfer or p complex type manner
without the possibility of rearomatization by its removal.The
conserved nature of the analogous phenylalanine in all PAL
and HAL sequences (e.g., the GGNFH segment is present in
the HAL enzymes of microorganisms such as P. putida or B.
subtilis as well as in mammalian HAL enzymes like mice, rat
or human) supports the importance of this amino acid.
Inspection and comparison of data with the simple s-
complex models (Scheme 3, Table 1) and partial active site/s-
complex models (Scheme 1, Table 1) reveals some important
determining, another rate-limiting step seems to be product
release.The kinetic significance of these two steps is
substrate-dependent.The Km values vary only moderately,
but roughly increase with increasing numbers of the halogenic
substituents.
Theoretical calculations: To demonstrate the feasibility of the
electrophilic attack by MIO at the halogenated aromatic rings
semiempirical calculations for the heat of formation and for
Table 1.Kinetic constants for the various phenylalanine analogues and calculated properties of their active complex models.
[d]
max/Vmax-Phe[e]
Substrate
s complex model)
DDHs-g[a]
DDC
[M.c.u]
b
ÀHs-g
V
K
m
[mm]
(
kcalmol ][b]
À1
[kcalmol ][c]
À1
[
l-Phe
0.0
15.7
7.8
3.1
1.5
7.6
21.9
4.8
29.9
7.4
5.2
0.0
12.4
3.6
3.9
1.8
3.1
14.6
4.1
15.0
23.0
11.0
15.6
À0.1
1.9
0.184
0.135
0.129
0.109
0.116
0.128
0.160
0.037
0.141
0.117
0.115
0.100
0.102
0.126
1.00
1.14
0.033
0.065
2
'F-l-Phe
'F-l-Phe
'F-l-Phe
(A)
B)
(A)
B)
(
3
2.09
0.079
(
4
2
3
2
2
0.56
0.85
2.72
0.16
1.03
0.010
0.085
0.159
0.076
0.050
',6'F
',5'F
2
-l-Phe
-l-Phe
2
',3',4',5',6'F
'Cl-l-Phe
5
-l-Phe
(A)
B)
(A)
B)
(
3
4
'Cl-l-Phe
'Cl-l-Phe
À 1.7
À 4.2
3.2
2.01
0.82
0.094
0.045
4.2
(
b-(5-pyrimidinyl)-d,l-Ala
b-(2-pyrimidinyl)-d,l-Ala
25.8
À 24.0
10.0
18.8
0.229
À 0.314
0.80
0.00
(K
i
ˆ 7mm)
[
a] The difference for heat of formation calculated for the active complex model for l-Phe and the ground state l-Phe substrate was substracted from the
corresponding difference for the substituted models and substrates.[b] Values calculated for simple models (fixed F ˆ À908 torsion angle, compare
Scheme 3) using PM3 method[ in vacuum.[c] Values for the partial active site models (c .f .Scheme 1) were obtained using MM ‡ optimization (with fixed
25]
[
26]
position of the methyl groups of the MIO and Phe399 models and carboxylate of the substrate) followed by single point PM3 calculation of the optimized
[
25]
structure. [d] The difference in Mulliken charges on the C
(
b
ÀH atoms in the substrate substracted from the corresponding difference in the simple models
fixed F ˆ À908 torsion angles, compare Scheme 3).[e] Relative Vmax values of the analogues compared with Vmax with l-Phe.
3388
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3388 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 18

Phenylalanine Ammonia-Lyase
3386±3390
Scheme 4.The postulated unreactive pyrimidinium complex with 2-pyr-
imidinylalanine 1 and the reaction with 5-pyrimidinylalanine 2.(Although
for the experiments the racemic compounds were used, only the l-
enantiomers are shown here which are presumed to be the enzymatically
active species).
Synthesis of and kinetic measurements with pyrimidinylala-
nines: The synthesis of b-(2-pyrimidinyl)-d,l-alanine (1) was
carried out in six steps following the procedure of Haggerty
et al.[
22]
A chemoenzymatic strategy was applied to the
preparation of b-(5-pyrimidinyl)-l-alanine (2).First, b-(5-
pyrimidinyl)-acrylic acid (3) was synthesised starting from
Scheme 3.Conformational properties of the simplified model and the
postulated s complex as intermediate of the phenylalanine ammonia-lyase
reaction.
5
-bromo-pyrimidine and tert-butylacrylate under Heck con-
ditions.Treatment of the product with trifluoro acetic acid
afforded the free acid 3, which was converted in 57% yield
into the corresponding enantiomerically pure l-alanine 2
using PAL as a catalyst.
While the 5-pyrimidinyl analogue 2 turned out to be a
moderately good substrate of PAL, its 2-pyrimidinyl counter-
part 1 was a competitive inhibitor.The corresponding kinetic
constants are shown in Table 1.The Km value of 2 is
differences.The phenyl ring of Phe399 in these s complexes
adopts nearly perpendicular arrangement to the Csp
3
ÀH bond
axis and the hydrogen points toward the central point of the
ring indicating strong interaction with the aromatic p-
electrons.For the 2 '-halogen substituted analogues the B-
type intermediates are substantially favored, especially in the
comparable to the K value of 1.
i
These results support the postulated ortho-attack of MIO at
the aromatic nucleus.When both ortho-positions are occupied
by a nitrogen atom such an attack leads to a pyrimidinium ion
which is so stable that no further reaction occurs (see
Scheme 4).The competitive nature of the inhibition requires,
however, that the binding of 2 is reversible.
2
'-chloro case.The presence of the Phe399 phenyl ring
substantially lowers the relative energy of the intermediate
cations for the polyfluorinated analogues, especially for 2',6'-
F -l-Phe and 2',3',4',5',6'-F -l-Phe.Thus, the maximal energy
requirement (2',3',4',5',6'-F -l-Phe: 29.9 kcalmol ) calculated
in vacuum in the absence of Phe399 for formation of the s
2
5
À1
5
complex intermediates substantially dropped when the aro-
matic ªprotective groupº was present (2',3',4',5',6'-F -l-Phe:
5
À1
1
5.0 kcalmol ).In addition, the pro-S C ÀH bond in the
b
Experimental Section
ground state of the 3',5'F substituted compound is signifi-
2
cantly more polarized (ca.0 0. 40 Mulliken charge units) then
in all the other models.
For the 5-pyrimidinyl compound relatively high energy is
required for formation of the s complex but the polarization
Recombinant PAL: The lyase was overexpressed in E. coli and purified as
[
6]
described, first according to Schuster and R e tey and later using the
improved method of Baedeker and Schulz.[
23]
Chloro- and fluoro-l-phenylalanines: Chlorocinnamic acids and fluorocin-
namic acids were purchased from Fluka and from Lancaster.Reaction of
the chloro- and fluorocinnamic acids with ammonia catalyzed by PAL
afforded the corresponding chloro- and fluoro-l-phenylalanines.A con-
centrated aqueous ammonia solution (10 mL) was diluted with water
of the scissible pro-S C ÀH is relatively strong.Accordingly, a
b
high Km value but a normal V is observed.The inhibition
max
observed with the 2-pyrimidinyl compound is a consequence
of the stability of the cationic pyrimidinium complex and the
unfavorable proton abstraction from the negatively polarized
(
10 mL).The pH of the solution was adjusted to pH 10 by bubbling CO
2
through the solution.Cinnamic acid (100 mg, 0 6. 75 mmol) and recombi-
nant wtPAL (1 U) (Petroselinum crispum) were added to this solution.The
reaction mixture was agitated overnight at 378C.The solution was acidified
with 5% HCl to pH 1.5, heated to boiling, filtered, and applied to a
Dowex 50 cation exchange resin column.The elution occurred with diluted
pro-S C ÀH bond (Scheme 4).
b
All these correlations are consistent with the intermediacy
of the Friedel ± Crafts-type s complexes and the kinetic
significance of their formation as partially rate-determining
step.Further support for our mechanism was provided by
recent publications.[
ammonia solution.The crude product was purified with HPLC [Nucleosil
À1
1
00 C18, 7 mm, 250 Â 20 mm (Macherey ± Nagel); flow rate: 5 mLmin
load: 20 mg; mobile phase: 0 ± 15 min: 99.9% H
linear increasing gradient to 99.9% CH CN/0.1% TFA, retention times:
;
2
O/0.1% TFA, 15 ± 90 min:
20, 21]
3
Chem. Eur. J. 2000, 6, No. 18
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3389 $ 17.50+.50/0
3389

J.ReÂtey et al.
FULL PAPER
l-phenylalanine: 41.7 min, 2'-fluoro-l-phenylalanine: 42.4 min, 3'-fluoro-l-
phenylalanine: 44.1 min, 4'-fluoro-l-phenylalanine: 43.9 min, 2',6'-di-
fluoro-l-phenylalanine: 42.7 min, 3',5'-difluoro-l-phenylalanine: 46.5 min,
arise by rotation of the substrate prior to the active complex formation,
model A represents the structure where the halogen substituent is closer to
the sp center in the s complex.
3
2
',3',4',5',6'-pentafluoro-l-phenylalanine: 48.4 min, 2'-chloro-l-phenylala-
The putative active site model was built from a sequence segment of PAL
nine: 45.3 min, 3'-chloro-l-phenylalanine: 48.0 min, 4'-chloro-l-phenylala-
nine: 48.7 min]. For the determination of the enantiomeric excess we used a
chiral column from astec/ict (Handels-GmbH, 61352 Bad Homburg,
(
SP: P24481) containing the ASGDL active site motif and the F399 by
[27]
[10]
homology-modelling using HAL structure (PDB: 1B8F ) as folding
template, followed by Gromos and Amber energy minimizations.The
ASG structure of the model was finally replaced by MIO taken from the
[27]
[26]
Germany) (Chirobiotic T, 5 mm, 250 Â 4.6 mm; mobile phase: 70% H
2
O/
À1
3
0% EtOH; flow rate: 0.8 mLmin ; load: 5 mg; retention times: l-
1
B8F HAL structure (less than 0.4 Š deviations in the corresponding
phenylalanine: 7.22 min, 2'-fluoro-l-phenylalanine: 6.87 min, 3'-fluoro-l-
phenylalanine: 6.98 min, 4'-fluoro-l-phenylalanine: 7.20 min, 2',6'-difluoro-
l-phenylalanine: 6.42 min, 3',5'-difluoro-l-phenylalanine: 6.75 min,
atomic positions).
Partial active site/s complex models (c.f. Scheme 1) were obtained using
MM ‡ optimization (with fixed position of the methyl groups of the MIO
2
',3',4',5',6'-pentafluoro-l-phenylalanine: 5.89 min, 2'-chloro-l-phenylala-
[
26]
and Phe399 models and carboxylate of the substrate) followed by single
point PM3 calculation of the optimized structure.[
nine: 9.06 min, 3'-chloro-l-phenylalanine: 8.22 min, 4'-chloro-l-phenylala-
nine: 9.25 min). In all cases enantiomeric excess over 99% was determined.
The d-enantiomers were eluted between 1 and 2 min later, as described in
the astec/ict prospect.
25]
Acknowledgements
b-(2-Pyrimidinyl)-d,l-alanine (1): Preparation as reported by W.J.Hagg-
[
22]
erty, R.H.Springer, and C.C.Cheng.
We thank M.Baedeker and G.Schulz (Universität Freiburg) for the
improved expression system for PAL .Financial support by the Deutsche
Forschungsgemeinschaft, the Fonds der Chemischen Industrie, the Hun-
garian OMFB and the Polish Science Foundation Council through the
Wroclaw University of Technology Fund is gratefully acknowledged.
b-(5-Pyrimidinyl)-l-alanine (2): b-(5-Pyrimidinyl)-acrylic acid (3) obtained
by Heck coupling between 5-bromopyrimidine and tert-butyl acrylate,
followed by ester-cleavage in TFA.[
[23]
24]
A concentrated, aqueous ammonia-solution (10 mL) was diluted with
distilled water (10 mL).The pH was adjusted to 10 0. by bubbling CO
2
into
the solution.Acrylic acid 3 (64.3 mg, 0.428 mmol) and PAL (1 U) were
added to this solution.After 24 h agitation at 37 8C the enzyme was
denatured by heat and removed by filtration.The clear solution was
applied to a Dowex 50 cation exchange resin column.The elution occurred
with diluted ammonia solution.The crude product was purified with HPLC
[
[
1] K.R.Hanson, E.A H. avir, Recent Adv. Phytochem. 1978, 12, 91 ± 137.
2] K.Hahlbrock, D.Scheel, Annu. Rev. Plant Physiol. Plant Mol. Biol.
1
989, 40, 347 ± 369.
[
[
[
[
[
3] K.R.Hanson, E.A.Havir, Arch. Biochem. Biophys. 1970, 141, 1 ± 77.
4] K.R.Hanson, E.A.Havir, Biochemistry 1973, 12, 1583 ± 1591.
5] E.A.Havir, P.D.Reid, H.V.March, Plant Physiol. 1971, 48, 130 ± 136.
6] B.Schuster, J.Re Âtey, FEBS Lett. 1994, 349, 252 ± 254.
(
5
1
Nucleosil 100 C18, 7 mm, 250 Â 20 mm; Macherey ± Nagel, flow rate:
À1
mLmin ; load: 20 mg; mobile phase: 0 ± 15 min: 99.9% H
2
O/0.1% TFA,
3
5 ± 90 min: linear increasing gradient to 99.9% CH CN/0.1% TFA,
7] M.Langer, G.Reck, J.Reed, J.Re Âtey, Biochemistry 1994, 33, 6462 ±
retention times: 23.5 min). The solvent was carefully removed under
reduced pressure to give a white solid (50 mg, 0.246 mmol, 57%).
6
467.
[
[
8] K.R.Hanson, E.A.Havir, Biochem. Plants 1981, 7, 577 ± 625.
9] W.Schulz, H.G.Eiben, K.Hahlbrock, FEBS Lett. 1989, 258, 335 ± 338.
Determination of Vmax- and K
m
values: The kinetic constants were
determined by measuring the UV absorption of the produced
[
10] T .F .Schwede, J.Re Âtey, G.E. Schulz, Biochemistry 1999, 38, 5355 ±
361.
11] M.Ormö, A B. .Cubitt, K.Kallio, L A. .Gross, R Y. .Tsien, S J. .
Remington, Science 1996, 273, 1392 ± 1395.
12] K.J.Luebke, Chem. Biol. 1998, 5, 317 ± 322.
13] G.J.Palm, A.Zdanov, G.A.Gaitanaris, R.Stauber, G.N.Pavlakis, A.
Wlodawer, Nat. Struct. Biol. 1997, 4, 361 ± 365.
À1
À1
acrylates (trans-cinnamic acid: e290 ˆ 10000 Lcm mol ; 2'-fluorocinnamic
5
À1
À1
acid:
e
280 ˆ 12550 Lcm mol
;
3'-fluorocinnamic
acid:
e
280
À1
ˆ
[
À1
À1
À1
1
3850 Lcm mol
;
4'-fluorocinnamic acid:
e
À1
280 ˆ 15320 Lcm mol
;
À1
2
',6'-difluorocinnamic acid: e290 ˆ 3960 Lcm mol ; 3',5'-difluorocinnamic
[
[
À1
À1
acid: e290 ˆ 5060 Lcm mol ; 2',3',4',5',6'-pentafluorocinnamic acid: e280
ˆ
À1
À1
À1
À1
7
910 Lcm mol
;
2'-chlorocinnamic acid:
e
285 ˆ 10770 Lcm mol
;
À1
À1
3
'-chlorocinnamic acid:
e
285 ˆ 10680 Lcm mol
290 ˆ 15790 Lcm mol
1110 Lcm mol ) using 0.05 ± 10mm of the corresponding amino acid
;
4'-chlorocinnamic
[
14] T .A .Smith, F .H .Cordelle, R .H .Abeles,
Arch. Biochem. Biophys.
À1
À1
acid:
1
e
;
b-(5-pyrimidinyl)-acrylic acid:
e
270
ˆ
1
967, 120, 724 ± 725.
À1
À1
[
[
[
[
15] B.Schuster, J.ReÂtey, Proc. Natl. Acad. Sci. USA 1995, 92, 8433 ± 8437.
16] J.ReÂtey, Naturwissenschaften 1996, 83, 439 ± 447.
17] A.Peterkofsky, J. Biol. Chem. 1962, 237, 787 ± 795.
18] M.Yanaka, U.Ura, A.Takahashi, N.Fukuhara, Mitsui Toatsu
Chemicals, Jpn.Kokai Tokyo Koho JP06, 113870 [94113870]
Cl.C12P13/06), 1994 [Chem. Abstr. 1994, 121, 155941].
19] J.D. Hermes, P.M. Weiss, W.W. Cleland, Biochemistry 1985, 24,
959 ± 2967.
20] A.Gloge, B.Langer, L.Poppe, J.Re Âtey, Arch. Biochem. Biophys.
998, 359, 1 ± 7.
as substrate.Conditions: 0 0. 5m m ± 10mm substrate in 0.1m Tris-HCl buffer
pH 8.8 at 308C.The reaction was initiated by addition of 20 mg (0.02 U)
1
PAL. H-NMR (250 MHz; D
5
2
O, 25 8C): l-phenylalanine: d ˆ 7.28 ± 7.43 (m,
H), 3.96 (t, 1H), 3.27 (dd, 1H), 3.08 (dd, 1H), 2'-fluoro-l-phenylalanine:
d ˆ 7.34 (m, 2H), 7.17 (m, 2H), 4.26 (t, 1H), 3.40 (dd, 1H), 3.20 (dd, 1H), 3'-
(
fluoro-l-phenylalanine: d ˆ 7.38 (m, 1H), 7.08 (m, 3H), 4.19 (t, 1H), 3.32
[
[
[
[
(
(
dd, 1H), 3.16 (dd, 1H), 4'-fluoro-l-phenylalanine: d ˆ 7.26 (m, 2H), 7.10
m, 2H), 4.28 (t, 1H), 3.29 (dd, 1H), 3.15 (dd, 1H), 2',6'-difluoro-l-
2
phenylalanine: d ˆ 7.36 (m, 1H), 6.99 (m, 2H), 4.28 (t, 1H), 3.37 (dd, 1H),
1
3
1
.26 (dd, 1H), 3',5'-difluoro-l-phenylalanine: d ˆ 6.90 (m, 3H), 4.32 (t,
21] A.Lewandowicz, J.Jemielity, M.Kan Âska, J.ZonÂ, P.Paneth, Arch.
H), 3.35 (dd, 1H), 3.17 (dd, 1H), 2',3',4',5',6'-pentafluoro-l-phenylala-
Biochem. Biophys. 1999, 370, 216 ± 221.
nine: d ˆ 4.01 (t, 1H), 3.26 (dd, 1H), 3.19 (dd, 1H), 2'-chloro-l-phenyl-
22] W.J. Haggerty, R.H. Springer, C.C. Cheng,
964, 2, 1 ± 6.
23] M.Baedeker, G.E.Schulz, FEBS Lett. 1999, 457, 57 ± 60.
24] J.E. Plevyak, J.E. Dickerson, R.F. Heck, J. Org. Chem. 1979, 44,
078 ± 4080.
[25] PC Spartan Pro, Wavefunction, Inc., 18401 Von Karman, Suite 370,
Irvine, CA 92612, USA.
[26] HyperChem 5.1, Hypercube, Inc., 1115 NW 4th Street, Gainesville,
32601, FL, USA.
J. Heterocycl. Chem.
alanine: d ˆ 7.48 (m, 1H), 7.34 (m, 3H), 4.05 (t, 1H), 3.45 (dd, 1H), 3.16 (dd,
1
1
1
2
H), 3'-chloro-l-phenylalanine: d ˆ 7.36 (m, 3H), 7.21 (m, 1H), 4.20 (t,
H), 3.31 (dd, 1H), 3.16 (dd, 1H), 4'-chloro-l-phenylalanine: d ˆ 7.40 (d,
H), 7.26 (d, 2H), 3.97 (t, 1H), 3.23 (dd, 1H), 3.10 (dd, 1H).
[
[
4
Theoretical calculations
Simple s-complex models for l-phenylalanine, the halogenated l-phenyl-
alanine, and pyrimidinyl-l-Ala derivatives were calculated in vacuo using
[
25, 26]
AM1 and PM3 methods.
Conformational analysis showed two favored
‡
states for zwitterionic structures with antiperiplanar pro-S H
b
ÀNH
3
and
[27] Swiss-PdbViewer, 3 5. 1, N.Guex, M.C.Peitsch,
Electrophoresis 1997,
zig ± zag arrangement (Scheme 3).All further calculations refer to the
arrangement with a fixed F ˆ À908 torsion angle.Since the parallel results
from the two methods were similar, values derived from PM3 calculations
are indicated only.In the cases where two distinct s complex models may
18, 2714 ± 2723.
Received: October 18, 1999
Revised version: March 3, 2000 [F2095]
3390
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3390 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 18