The mechanism of a-proton isotope exchange in amino acids catalysed
by tyrosine phenol-lyase
1
What is the role of quinonoid intermediates?
Nicolai G. Faleev
1
, Tatyana V. Demidkina
2
, Marina A. Tsvetikova
1
, Robert S. Phillips
3
and Igor A. Yamskov
1
1
Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences, Moscow, Russia;
2
Engelhardt Institute of
Molecular Biology, Russian Academy of Sciences, Moscow, Russia;
3
Department of Chemistry, Department of Biochemistry and
Molecular Biology, and Center for Metalloenzyme Studies, University of Georgia, Athens, GA, USA
To shed light on the mechanism o f isotopic exchange of
a-protons in amino acids catalyzed by pyridoxal phosphate
(PLP)-dependent enzymes, we studied the kinetics of
quinonoid intermediate formation for the reactions of
tyrosine phenol-lyase with
L
-phenylalanine,
L
-methionine,

L
-methionine-c-lyase
(EC 4.4.1.11). These enzymes are used as very effective
biocatalysts for preparation of enantiomerically pure
a-de uterated (S)-amino acids [5–7].
In the framework of the generally accepted notions of
mechanisms of PLP-dependent enzymes the mechanism of
the i sotopic e xchange traditionally is considered to be
associated with formation of quinonoid intermediates
(Scheme 1). In the holoenzymes (E) the cofactor PLP is
bound in the active site as an Ôinternal aldimineÕ with an
e-amino group of a definite lysine residue. As a result of
interaction with an amino acid substrate, or inhibitor, the
internal aldimine (E) is substituted by an ÔexternalÕ one (ES),
which undergoes the abstraction of the a-proton by a
certain enzyme group, leading to formatio n of a Ôquinonoid
intermediateÕ (EA). The reversibility of the latter transfor-
mation should lead in heavy water to the isotopic exchange
of the a-proton if t he abstracted proton may be easily
exchanged with t he solvent. However, the kinetics o f
quinonoid formation was examined until now only in water
solutions [8–11], while measurements in heavy water, in
conditions identical to those of the isotopic exchange, were
not performed. No attempts to quantitatively estimate the
rates of the exchange of the abstracted proton in the active
site have been reported. We have noted earlier [8] that no
direct correlation was observed between the amount of the
quinonoid intermediate formed under steady-state condi-
tions in reactions of PLP-dependent enzymes with amino
acids and the rates of the enzymatic isotopic exchange for

Abbreviations: PLP, pyridoxal-P-phosphate; TPL, tyrosine phenol-
lyase; SOPC, S-o-nitrophenyl-
L
-cysteine.
Enzymes: tyrosine phenol-lyase (EC 4.1.99.2); tryptophan indole-lyase
(EC 4.1.99.1);
L
-methionine-c-lyase (EC 4.4.1.11); aspartate amino-
transferase (EC 2.6.1.1).
(Received 6 July 2004, revised 7 September 2004,
accepted 8 October 2004)
Eur. J. Biochem. 271, 4565–4571 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04428.x
reaction, while the e xchange of the a-proton is realized by
an alternative, possibly concerted, mechanism.
Materials and methods
Materials
TPL was obtained from Escherichia coli SVS370 cells con-
taining plasmid pTZTPL, which contains the tpl gene from
Citrobacter freundii in pTZ18U (US Biochemical, Cleveland,
OH, USA)
2
, as described [10]. The enzyme obtained was
apparently homogeneous and had specific activity of
4.91 unitsÆmg
)1
. The co ncentration of the active enzyme was
determined by activity measurements, assuming that the pure
enzyme enzyme had a maximum specific activity of 6 unitsÆ
mg
)1

-Phe. The
procedure for preparation of a-deuterated
L
-Met was the
same, except initially 0.7 g
L
-Met was dissolved in 15 mL
D
2
O, and the time of incubation was 80 h.
Stopped-flow measurements
Prior to performing rapid kinetic experiments, the stock
enzyme was incubated with 1 m
M
pyridoxal-P for 1 h at
30 °C at pH 7.0 and then separated from excess pyridoxal-P
using a short desalting column (PD-10, Pharmacia) equili-
brated with 0.1
M
potassium phosphate pH 8.7. For
experiments in D
2
O the enzyme solution was concentrated
to a minimal volume by ultrafiltration and diluted with
0.1
M
potassium phosphate in D
2
O pD 8.7. To determine
pD values an allowance was made for the isotope effect of

the respective kinetic parameters were calcu-
lated using Eqns (3–5), and are presented in Table 2.
Scheme 1.
Fig. 1. The concentration dependence for quinonoid formation rates for
the reaction of TPL with a-deuterated
L
-phenylalanine in D
2
O. d,
Experimental data; solid line, calculated fit to Eqn (1) w ith K
S
, k
f
and
k
r
given in Table 1.
4566 N. G. Faleev et al. (Eur. J. Biochem. 271) Ó FEBS 2004
Isotope exchange experiments
The reaction with
L
-phenylalanine was run in 0.1
M
potassium phosphate solution in D
2
O pD 8.7, containing
33.94 m
ML
-Phe, 0.1 m
M

bial sources have been established by X-ray studies [14–16].
It was shown that the cofactor, PLP, occupies a strictly
determined position in the active site. According t o Pletnev
et al. and Sundararaju et al. [15,16], for TPL from Citro-
bacter freundii, Arg404 is the best candidate for t he binding
of the a-carboxylate group of the s ubstrate, when the
external aldimine is formed. The anchoring of a-carboxylate
and a-amino group in the external aldimine defines
automatically the positions of the a-proton and the side
chain of any bound amino ac id. The lability of the a-proton
observed for a large number of amino acids [5] under the
action of TPL gives evidence for the orthogonal orientation
of the a-proton with respect to the cofa ctor plane [17], and
shows that t he pattern of binding is the same for a variety of
amino acids. It has been established [5] that for a number of
amino acid i nhibitors bearing nonbranched substituents
without functional groups, the hydrophobicity of the side
chain is the main factor controlling K
i
. Amino acids that
contain nucleophilic side chains (
L
-aspartic acid,
L
-homo-
serine,
L
-methionine,
L
-glutamic acid) exhibit enhanced

L
-phenylalanine,
L
-methionine, and their deuterated
analogs.
Substrate Solvent K
S
(m
M
) k
f
(s
)1
) k
r
(s
)1
)
L
-Phenylalanine
2
H
2
O 1.14 ± 0.11 14.9 ± 0.35 3.1 ± 0.24
[a-
2
H]-
L
-Phenylalanine
2

substrate.
Substrate a K
m
(m
M
) k
cat
(s
)1
) K
p
(m
M
)
L
-Phenylalanine 0.294 0.196 0.748 0.370
L
-Methionine 0 0.0915 0.015
a
0.340
L
-Tyrosine – 0.2 [10] 3.5 [10] –
a
Maximum possible value.
Scheme 3.
Ó FEBS 2004 Enzymatic a-proton exchange in amino acids (Eur. J. Biochem. 271) 4567
k
obs
¼
1

rðDÞ
ð2Þ
The relative contributions of these processes are described
by a partition coefficient a, which is determined by: (a) the
rates of the isotopic exchange between the enzyme func-
tional group having abstracted the a-proton, and existing as
a conjugate acid, and surrounding groups, capable of
isotopic exchange, and s olvent molecules p resent in the
active site; (b) the statistical factor taking account of the
ratio of protons and deuterons on the considered group
when the latter is polyprotic; (c) the degree of restriction of
the free rotation of the considered group in the active site.
For t he reaction with
L
-phenylalanine the value of k
r
for
nondeuterated substrate is more than for the a-deuterated
one by a factor of 2.4. This indicates the presence of a
considerable internal return. The value of k
r(D)
, character-
izing t he deuteration process, corresponds to the k
r
value f or
the a-deuterated
L
-phenylalanine. We a ssumed that t he
value of k
r(H)

conjugate acid, bearing a positive charge and containing
two deuteriums and one hydrogen at the nitrogen atom. If
rotation around the C–Nbond is not restricted, the statis-
tical factor for the internal return of the proton is equal to
0.33. For the reaction of
L
-phenylalanine the observed value
of the internal return coefficient (a ¼ 0294) is only slightly
less. Consequently, it is reasonable to conclude that the
transfer of the proton (or deuteron) from the amino group
to the a-carbon ato m of the quinonoid intermediate should
go faster than the isotopic exchange of the proton in the
active site. For the reaction of
L
-methionine, w here no
internal return is observed, on the contrary, the isotopic
exchange goes faster, which seems natural because the
deuteration of the quinonoid intermediate proceeds much
slower than in the
L
-phenylalanine reaction. Thus, we may
estimate the va lue of the isotope exchange rate from the
protonated amino group as being considerably more than
the k
r(D)
value for the reaction with
L
-methionine
(% 0.01 s
)1

are
formally nonidentical because for the former the protona-
tion (internal return), leading to regeneration of the initial
nondeuterated substrate is still possible, while the latter can
be only deuterated. Thus, quinonoid intermediate EA
D
is
off the reaction pathway responsible for the principal
transformation.
Values of K
SH
and k
f(H)
correspond to K
D
and k
f
for the
reaction of nondeuterated substrate, and K
SD
, k
f(D)
and
k
r(D)
are equal, respectively, to K
D
, k
f
and k

k
fðHÞ
k
fðHÞ
þ ak
rðHÞ
þð1 À aÞk
rðDÞ
ð4Þ
The suggested mechanism implies also t hat the isotopic
exchange reaction should be inhibited by the deuterated
product. The respective i nhibition constant (K
p
) is described
by Eqn (5).
K
p
¼
KS
D
1 þ
k
fðDÞ
k
rðDÞ
ð5Þ
The theoretical kinetic parameters calculated in this way are
presented in Table 2.
For enzymatic reactions where inhibition by product is
observed the dependence of product concentration on time

L
-methio-
nine, calculated with t he use o f t he kinetic p arameters
presented in Table 2 are compared with the experimental
data. For the reaction of
L
-phenylalanine, the experimental
4568 N. G. Faleev et al. (Eur. J. Biochem. 271) Ó FEBS 2004
points at longer times lie somewhat below the theoretical
curve, which may be due to some inac tivation o f t he enzyme
during the reaction. In general, however, the deviations of
the experimental values from the calculated ones are not
significant. We believe therefore that for this reaction the
traditional mechanism of isotopic exchange, involving the
formation of a quinonoid species as a principal intermediate
structure, agrees satisfactorily with the experimental results.
The r ate o f isotopic e xchange i s mainly determined b y
deuteration of the quinonoid intermediate.
On the other hand, it is obvious from Fig. 3 that for
reaction of
L
-methionine the experimental data can in no
way be reconciled with the theoretically expected results.
The experimental values are much higher than the calcula-
ted ones, and the initial rate of exchange (k
ex
¼ 0.37 s
)1
)is
by a f actor of 2 2.5 faster than the highest possible k

quinonoid intermediate. For a simple kinetic scheme
(Scheme 4) the observed exchange rate may be described
by Eqn (7):
k
ex
¼
k
Ã
1þ
k
f
k
r
ð7Þ
and the k* value estimated in this way should be equal to
230–240 s
)1
.
Considering alternative mechanisms of the isotopic
exchange we should note that although numerous exam-
ples of apparent stepwise mechanisms in reactions of PLP-
dependent enzymes are known, in some cases an interesting
tendency to utilize concerted mechanisms was observed.
Julin and Kirsch [21] h ave shown for the reaction o f
cytosolic aspartate aminotransferase that the proton
transfer from the C
a
to the C
4
, position of the cofactor

8.7, containing 95.23 m
ML
-Met, 0.1 m
M
PLP, and 2.64 unitsÆmL
)1
TPL. j, Experimental data; solid line, t he experimental curve calcu-
latedusingEqn(6)andkineticparametersfromTable2.
Fig. 2. Isotopic exchange of
L
-phenylalanine under the action of TPL.
Thereactionwasrunin0.1
M
potassium phosphate buffer in D
2
O
pD ¼ 8.7, containing 33.94 m
ML
-Phe, 0.1 m
M
PLP, and 1.27 unitsÆ
mL
)1
TPL. d, Experimental data; solid line, the experimental curve
calculated using Eqn (6) and kinetic parameters from Table 2.
Ó FEBS 2004 Enzymatic a-proton exchange in amino acids (Eur. J. Biochem. 271) 4569
by O-acetylserine sulfhydrylase. By analogy w ith these
findings, a concerted mechanism of isotopic exchange may
be considered as a possible alternative. The concerted
mechanism, involving t he Lys257 amino group and t he

k
f
¼ 5.85 s
)1
(Table 1) and k* ¼ 230–240 s
)1
it follows
that the putative concerted isotopic exchange should go
faster by a factor of 4 0 t han the ÔnormalÕ a-proton
abstractioninthecomplexofTPLwith
L
-methionine.
Acknowledgments
This research was supported by grants from the Russian Foundation
for Basic Researches (0 4-04-49370 ) to N.G.F. and Fogarty Interna-
tional Center (TW00106) to R.S.P. and T.V.D.
References
1. Konnikova, A.S., Dobbert, N.N. & Braunstein, A.E. (1947)
Labilization of a-hydrogen of amino acids under the action of
aminoferase. Biokhimia 12, 556–568 (in Russian).
2. Esaki, N., Nakayuma, T., Sawada, S., Tanaka, H. & Soda, K.
(1985) Proton NMR studies of substrate hydrogen exchange
reactions catalyzed by 1-m ethionine-c-lyase. Biochenistry 24,
3857–3862.
3. Kainosho, M., Ajisaka, K., Kamisaku, M. & Murai, A. (1975)
Conformational analysis of amino acids and peptides using spe-
cific isotope substitution. 1. Conformation of 1-phenylalanylgly-
cine. Biochem. Biophys. Res. Commun. 69,429.
4. Morino, Y. & Snell, E. (1967) The relation of spectral changes and
tritium exchange reactions to the mechanism of tryptophanase-

13. Phillips, R.S. & Gollnick, P.D. (1989) Evidence that cysteine 298 i s
in the active site of tryptophan indole-lyase. J. Biol. Chem. 264,
10627–10632.
14. Antson, A.A., Demidkina, T.V., Gollnick, P., Dauter, Z., Von
Tersch, R., Long, J., Berezhnoy, S.N., Phillips, R.S., W ilson, K.S.
& Harutyunyan, E. (1993) Three-dimentional structure of tyrosine
phenol-lyase. Biochemistry 32, 4195.
15. Pletnev, S.V., Antson, A.A., Sinitsina, N.I., Dauter, Z., Khurs,
E.N., Faleev, N.G., W ilson, K.S., Dodson, G.G., Demidkina,
T.V. & Harutyunyan, E.H. (1997) Crystallographic study of
tyrosine phen ol-lyase from Erwinia herbicola. Krist allografia
(in Russian) 42, 877–888.
16. Sundararaju, B., Antson, A.A., Phillips, R.S., Demidkina, T.V.,
Barbolina, M.V., Gollnick, P., Dodson, G.G. & Wilson, K.S.
(1997) The crystal structure of Citrobacter freundii tyrosine phe-
nol-lyase complexed with 3-(4-hydroxyphenyl) propionic acid,
together with site-directed mutagenesisis andx kinetic analysis,
demonstrates that arginine 381 i s required for substrate specificity.
Biochemistry 36, 6502–6510.
17. Dunathan, H.C. (1971) Stereochemical asp ects of pyrido xal
phosphate catalysis. Adv. Enzymol. Relat. Areas Mol. Biol. 85,79–
134.
18. Mouratou, B., Kasper, P., Gehring, H. & Christen, P. (1999)
Conversion of tyrosine phenol-lyase to dicarboxylic amino acid
b-lyase, an enzyme not found in nature. J. Biol. Chem. 247, 1320–
1325.
Fig. 4. The putative concerted mechanism of isotopic exchange involving
the formation of a six-membered transition state.
4570 N. G. Faleev et al. (Eur. J. Biochem. 271) Ó FEBS 2004
19. Strickland, S., Palmer, G. & Massey, V. (1975) Determination of