Identification of active site cysteine residues that function as general bases: Diaminopimelate epimerase [9]

Full text of DOI:10.1021/ja001193t

6122
J. Am. Chem. Soc. 2000, 122, 6122-6123
Identification of Active Site Cysteine Residues that
Function as General Bases: Diaminopimelate
Epimerase
†
‡
‡
Carolyn W. Koo, Andrew Sutherland, John C. Vederas, and
John S. Blanchard*,†
Department of Biochemistry
Albert Einstein College of Medicine
Bronx, New York 10461
Figure 1. Reaction catalyzed by diaminopimelate epimerase.
Department of Chemistry, UniVersity of Alberta
Edmonton, Alberta, Canada T6G 2G2
ReceiVed April 5, 2000
D,L-meso-Diaminopimelate (DAP), a precursor to L-lysine in
bacterial lysine biosynthesis, is a diamino acid which is incor-
porated into the pentapeptide of the Gram-negative peptidoglycan
1
moiety. Diaminopimelate epimerase (EC 5.1.1.7), a member of
the pyridoxal phosphate-independent amino acid racemases,
catalyzes the interconversion of L,L and D,L-meso-diaminopimelate
2
(
Figure 1). Mechanistically similar to the well-studied PLP-
3
independent amino acid racemases, especially proline racemase
and glutamate racemase,4,5 DAP epimerase uses two cysteine
residues. The thiolate form of one of the cysteines functions as
the general base, and the other cysteine thiol functions as the
general acid in one direction, while these ionization states, and
functions, must be necessarily reversed for the other direction.
An early mechanistic study of the Escherichia coli diaminopime-
Figure 2. Proposed mechanism of hydrogen fluoride elimination by DAP
epimerase. The product after HF elimination of L,L- and D,L-3-fluoro-
DAP (1 and 2) forms the enamine (3), which spontaneously hydrolyzes
and cyclizes to generate tetrahydrodipicolinic acid (4).
2
late epimerase supported a two-base mechanism. The irreversible
inhibition of the E. coli enzyme due to the alkylation of Cys73
by an active site-directed inhibitor, 2-(4-amino-4-carboxybutyl)-
aziridine-2-carboxylate (aziDAP), suggested that this cysteine
values of 4 µM and 25 µM, respectively), the enzyme was also
found to rapidly eliminate hydrogen fluoride (HF) from 2,
epimerize 1, and slowly eliminate HF from 1. The product formed
from this elimination is the enamine (3), which spontaneously
cyclizes (possibly via hydrolysis to the intermediate 2-keto-6-
aminopimelate) to tetrahydrodipicolinic acid (4, Figure 2).
Monitoring hydrogen fluoride release by 19F NMR, both elimina-
tion and epimerization products were identified with wild-type
H. influenzae DAP epimerase (data not shown). Single mutants,
C73A, C73S, C217A, and C217S, and the double mutant, C73S/
C217S, were prepared and purified from E. coli.10 Each mutant
was incubated with 1 and 2 and monitored continuously for HF
elimination (Table 1).
6
residue is present at or near the active site. The three-dimensional
structure of the Haemophilus influenzae diaminopimelate epime-
rase has been determined, and the 274-amino acid monomeric
enzyme has the two conserved cysteine residues, Cys73 and
Cys217, in disulfide linkage at the interface of two structurally
7
superimposable domains. Kinetic and isotopic studies of the
reduced, active enzyme suggest these two cysteine residues are
8
the catalytic acid and base. The assignment of the two active-
site cysteines responsible for proton abstraction in the L,L f D,L
and D,L f L,L directions has been investigated using diasteriomeric
3
-fluoro-DAP substrates and cysteine site-directed mutants of H.
Although the single mutants of diaminopimelate epimerase,
C73A and C217A, were found to be inactive as epimerases, these
enzymes were able to catalyze the elimination of hydrogen
fluoride via abstraction of the C-2 hydrogen. The C73A mutant
was able to rapidly catalyze elimination of the D,L-3-fluoro-DAP
analogue (2) and was essentially unable to catalyze elimination
with the L,L-3-fluoro-DAP analogue (1, Figure 3a and 3b). This
trend was reversed with the C217A mutant, which catalyzed HF
elimination from L,L-3-fluoro-DAP but was incapable of cata-
lyzing HF elimination from D,L-3-fluoro-DAP (Figure 3c and 3d).
The single mutants, C73S and C217S were able to catalyze both
epimerization of DAP and HF elimination of the fluoro-DAP
analogues (Table 1). Qualitatively, the C73S mutant was able to
influenzae diaminopimelate epimerase.
The 3-fluoro analogue of L,L-DAP, (2S,3R,6S)-3-fluoro-2,6-
diaminopimelic acid (1) and the 3-fluoro analogue of D,L-DAP,
the 2R,3S,6S isomer (2), were synthesized and tested with the E.
coli DAP epimerase.9 Although 1 and 2 were competitive
inhibitors of wild-type E. coli DAP epimerase, (exhibiting K
i
*
To whom correspondence may be addressed. Telephone: (718) 430-
3
096. Fax: (718) 430-8565. E-mail: blanchar@aecom.yu.edu.
†
‡
Albert Einstein College of Medicine.
University of Alberta.
(
1) (a) Scapin, G.; Blanchard, J. S. AdV. Enzymol. 1998, 72, 279-324. (b)
Cox, R. J. Nat. Prod. Rep. 1996, 13, 29-43. (c) Cox, R. J.; Sutherland, A.;
Vederas, J. C. Bioorg. Med. Chem. 2000, in press.
(
(
2) Wiseman, J. S.; Nichols, J. S. J. Biol. Chem. 1984, 259, 8907-8914.
3) (a) Rudnick, G.; Abeles, R. H. Biochemistry 1975, 14, 4515-4522.
(
b) Belasco, J. G.; Albery, W. J.; Knowles, J. R. Biochemistry 1986, 25, 2552-
(9) (a) Gelb, M. H.; Lin, Y.; Pickard, M. A.; Song, Y.; Vederas, J. C. J.
Am. Chem. Soc. 1990, 112, 4932-4942. (b) Sutherland, A.; Vederas, J. C. J.
Chem. Soc., Chem. Commun. 1999, 1739-1740.
2
3
3
558.
(
4) Gallo, K. A.; Tanner, M. E.; Knowles, J. R. Biochemistry 1993, 32,
991-3997.
(10) Mutations at the Cys73 and Cys217 positions of the H. influenzae
diaminopimelate epimerase (dapF gene) were introduced by PCR (Perkin-
Elmer). Sequencing of each of the plasmid DNA confirmed successful
mutation(s) of the enzyme. Expression and purification of the mutant enzymes
(
5) Tanner, M. E.; Gallo, K. A.; Knowles, J. R. Biochemistry 1993, 32,
998-4006.
(
6) (a) Higgins, W.; Tardif, C.; Richaud, C.; Krivanek, M. A.; Cardin, A.
8
Eur. J. Biochem. 1989, 186, 137-143. (b) Gerhart, F.; Higgins, W.; Tardif,
were accomplished following the protocol for the wild-type enzyme.
C.; Ducep, J.-B. J. Med. Chem. 1990, 33, 2157-2162.
Electrospray ionization mass spectrometry confirmed the mass of the mutant
epimerases as well as the purity of the samples. Further details of the cloning,
expression, and purification of the mutant enzymes will be published
separately.
(
7) Cirilli, M.; Zheng, R.; Scapin, G.; Blanchard, J. S. Biochemistry 1998,
3
7, 16452-16458.
(
8) Koo, C. W.; Blanchard, J. S. Biochemistry 1999, 38, 4416-4422.
1
0.1021/ja001193t CCC: $19.00 © 2000 American Chemical Society
Published on Web 06/08/2000

Communications to the Editor
J. Am. Chem. Soc., Vol. 122, No. 25, 2000 6123
Table 1. Relative Activity of H. Influenzae DAP Epimerase
Enzymes
epimerization (%)a
L,L-DAP
HF eliminationb
L,L-3-F (1)
D,L-3-F (2)
wild-type
C73A
C217A
C73S
C217S
100
+++
-
++
++++
-
<0.05
<0.05
3
+
++++
++
++
++++
+
2
C73S/C217S
<0.05
++
a
Activity was measured with the coupled enzyme activity assay in
the L,L f D,L direction.8 Normalized to the same protein concentra-
tion. ++++ complete HF elimination in <1 h; +++ <2 h; ++ <6
h; + <24 h.
b
interferes with epimerization of the L,L-3-fluoro-DAP analogue.
The alternative explanation would be that HF elimination is faster
than epimerization for these mutants (in contrast to the wild-type
enzyme), but this appears improbable since the conformational
orientation of the fluorine in the active site (unfavorable for
9a
elimination) is likely to remain the same. Epimerization activity
of the double mutant, C73S/C217S was not detectable by the
8
coupled enzyme activity assay nor the circular dichroism assay.
However, the double mutant was able to slowly eliminate HF
from both 1 and 2.
For glutamate racemase, substrate analogues, threo-3-chloro-
glutamate and N-hydroxyglutamate, were used to assign Cys73
as the general base for proton abstraction in the D f L direction
5
and Cys184 as the general base for the L f D directions.
Glutamate racemase mutants, C73A and C184A, which were
inactive as racemases, catalyzed hydrogen chloride elimination
of the L- and D-glutamate analogues, 2S,3S- and 2R,3R-3-
chloroglutamate, respectively. The cysteine to serine mutants,
C73S and C184S, were able to eliminate water more readily from
11
L-N-hydroxyglutamate and D-N-hydroxyglutamate, respectively.
For H. influenzae diaminopimelate epimerase, on the basis of the
observations of the ability or inability of the mutants to catalyze
elimination of HF from either 1 or 2, we assign Cys73 as the
general base responsible for the abstraction of the C-2 hydrogen
of the L,L-3-fluoro-DAP, and C217 is responsible for proton
abstraction of the D,L-3-fluoro-DAP (Figure 2). Assuming that
the 3-fluoro-DAP analogues bind in a fashion similar to that of
the corresponding diaminopimelate substrates, the results of these
experiments support the assignments of Cys73 and Cys217 as
the general base for proton abstraction in the L,L f D,L and D,L
f L,L directions, respectively. Currently under investigation are
the steady-state kinetic and isotope effects studies of the C73S
and C217S diaminopimelate epimerase mutants.
Figure 3. 19F NMR time course of the reaction of the C73A mutant
diaminopimelate epimerase with (a) 1 and (b) 2, and of the C217A mutant
with (c) 1 and (d) 2. Method: The reaction was initiated by the addition
of C73A or C217A to 1 (1.8 mM solution, final enzyme concentrations
were 10 and 1 µM, respectively) or to 2 (1.4 mM and 2.0 mM solution,
respectively; final enzyme concentration was 3 µM), 50 mM potassium
phosphate, and D
2
O (10%). The final volume was 330 µL. The experiment
Acknowledgment. We thank Dr. Sean Cahill for expert technical
assistance with the NMR spectroscopy and Edward Nieves of the
Laboratory for Macromolecular Analysis and Proteomics for the elec-
trospray ionization mass spectrometry data. This work was supported by
the National Institute of Health (Grant AI33696 to J.S.B. and Grant
GM20332 to C.W.K.), Alberta Heritage Foundation for Medical Research
was carried out at pH 7.8 and 25 °C in a 5 mm symmetrical microtube
matched with D O (Shigemi). F NMR spectra (proton decoupled) were
2
recorded on a Bruker DRX 300 spectrometer. Acquisition time ) 1.05
s; number of scans ) 128 (4.4 m); spectral width ) 62.5 kHz. No window
function was applied prior to Fourier transformation and phasing.
1
9
(Fellowship to A.S.), and the Natural Sciences and Engineering Research
catalyze elimination from 2 more readily than from 1, and the
reverse trend was seen with the C217S mutant. Interestingly, the
fluoro-epimerization product of 1 was not detected for either of
these mutants, suggesting that the cysteine to serine mutation
Council of Canada.
JA001193T
(11) Glavas, S.; Tanner, M. E. Biochemistry 1993, 32, 4106-4113.