Synthetic and mechanistic studies of (2R,3S)-3-vinylmalic acid as a mechanism-based inhibitor of 3-isopropylmalate dehydrogenase

Article abstract of DOI:10.1021/jo982206c

(2R,3S)-3-Vinylmalic acid (VM) was designed as a mechanism-based inhibitor of threo-3-isopropylmalate dehydrogenase (IPMDH), the rate- determining enzyme responsible for the penultimate step in the biosynthetic pathway of the essential amino acid L-leucine. The synthesis of VM was achieved in six steps from diethyl (R)-malate. Besides its weak activity as a substrate, VM was shown to be a mechanism-based inhibitor (K(I) = 1.20 mM) for IPMDH as deduced from the time-dependent and kinetic analyses. 2-Oxo-3- pentenoate was identified as the enzyme reaction product of VM by GCMS analysis and ¹H NMR spectroscopy, but the supplemented 2-oxo-3-pentenoate was not, in contrast, inhibitive to the enzyme reaction. It seems likely from these results that an activated or nucleophilic amino acid of IPMDH by deprotonation during the ordinary enzyme reaction process participates in a strong interaction with the 2-oxo-3-pentenoate product, probably by the formation of a transient covalent bond, which in turn gives rise to inhibition of the enzyme reaction.

Full text of DOI:10.1021/jo982206c

J . Org. Chem. 1999, 64, 6159-6165
6159
Syn th etic a n d Mech a n istic Stu d ies of (2R,3S)-3-Vin ylm a lic Acid a s
a Mech a n ism -Ba sed In h ibitor of 3-Isop r op ylm a la te Deh yd r ogen a se
Akira Chiba,† Tetsuya Aoyama,† Rieko Suzuki,† Tadashi Eguchi,*,†,‡ Tairo Oshima,§ and
Katsumi Kakinuma*,†
Department of Chemistry, Tokyo Institute of Technology, O-okayama, Meguro-ku, Tokyo 152-8551, J apan,
and Department of Molecular Biology, Tokyo University of Pharmacy and Life Science,
Hachioji-shi, Tokyo 192-0392, J apan
Received November 3, 1998
(2R,3S)-3-Vinylmalic acid (VM) was designed as a mechanism-based inhibitor of threo-3-isopropyl-
malate dehydrogenase (IPMDH), the rate-determining enzyme responsible for the penultimate step
in the biosynthetic pathway of the essential amino acid L-leucine. The synthesis of VM was achieved
in six steps from diethyl (R)-malate. Besides its weak activity as a substrate, VM was shown to be
a mechanism-based inhibitor (KI ) 1.20 mM) for IPMDH as deduced from the time-dependent and
kinetic analyses. 2-Oxo-3-pentenoate was identified as the enzyme reaction product of VM by GC-
MS analysis and 1H NMR spectroscopy, but the supplemented 2-oxo-3-pentenoate was not, in
contrast, inhibitive to the enzyme reaction. It seems likely from these results that an activated or
nucleophilic amino acid of IPMDH by deprotonation during the ordinary enzyme reaction process
participates in a strong interaction with the 2-oxo-3-pentenoate product, probably by the formation
of a transient covalent bond, which in turn gives rise to inhibition of the enzyme reaction.
In tr od u ction
Mechanistic enzymology is a field under intense in-
vestigation within the study of chemical biology. We have
been involved in mechanistic and molecular recognition
studies on threo-D-3-isopropylmalate dehydrogenase (IP-
MDH; EC 1.1.1.85), derived from the extremely thermo-
philic bacteria Thermus thermophilus HB8. IPMDH is a
homodimeric enzyme and catalyzes the oxidation and
decarboxylation reaction of (2R,3S)-3-isopropylmalate
(IPM) into 2-oxoisocaproate with the aid of NAD+ in the
penultimate step of the biosynthetic pathway of L-leucine,
as shown in Figure 1.1
The cryptic stereochemistry of the IPMDH reaction has
already been elucidated: the hydride transfer from the
C-2 position of IPM to NAD+ is pro-R (A-site) specific,
and the decarboxylation proceeds with retention of con-
figuration at C-3 of the substrate.2 These stereochemical
features of IPMDH demonstrate significant similarity to
those of threo-D-isocitrate dehydrogenase (ICDH; EC
1.1.1.42) functioning in the tricarboxylic acid cycle.
Genetic comparison of IPMDH and ICDH also suggested
an evolutionary relationship between the two enzymes.3,4
Recently, the separate crystal structures of the enzyme
containing the NAD+ cofactor or the IPM substrate were
analyzed by X-ray crystallography, and the polypeptide
F igu r e 1. IPMDH enzyme reaction.
regions responsible for the binding of NAD+ and IPM
were discussed by Hurley et al. and Tanaka et al.,
respectively.4-6 However, several features of the IP-
MDH-substrate interaction have yet to be clarified, i.e.
(i) the conformation of the nicotinamide nucleotide was
rather mobile in the crystals of IPMDH with NAD+ and
(ii) when the IPMDH crystals were soaked with the IPM
substrate at high concentrations, the enzyme crystals
were destroyed. Tanaka et al. suggested that the enzyme
conformation must be altered rather significantly upon
the substrate binding, and that the crystal structure
deduced with the soaked specimen at low substrate
concentration may not represent the features of the active
IPMDH-substrate complex.6 It is well-established that
most of the enzymes utilizing a nicotinamide coenzyme
bind the coenzyme first, followed by the substrate.
Accordingly, in the complexation of IPMDH with the
cofactor and the substrate, when an enzyme molecule
captures the cofactor, significant conformational changes
are induced to make the resulting enzyme-cofactor
complex prone to accepting the IPM substrate. Further-
more, quite recently, a crystal of Thiobacillus ferrooxi-
dans IPMDH complexed with IPM was analyzed, again
* To whom correspondence should be addressed. Fax: +81-3-5734-
3713. E-mail: kakinuma@chem.titech.ac.jp.
† Tokyo Institute of Technology.
‡ Present address: Department of Chemistry and Materials Science,
Tokyo Institute of Technology.
§ Tokyo University of Pharmacy and Life Science.
(1) (a) Tanaka, T.; Kawano, N.; Oshima, T. J . Biochem. (Tokyo) 1981,
89, 677. (b) Kagawa, Y.; Nojima, H.; Nukiwa, N.; Ishizuka, M.;
Nakajima, T.; Yasuhara, T.; Tanaka, T.; Oshima, T. J . Biol. Chem.
1984, 259, 2956. (c) Yamada, T.; Akutsu, N.; Miyazaki, K.; Kakinuma,
K.; Yoshida, M.; Oshima, T. J . Biochem. (Tokyo) 1990, 108, 449.
(2) (a) Yamada, T.; Kakinuma, K.; Oshima, T. Chem. Lett. 1987,
1745. (b) Yamada, T.; Kakinuma, K.; Endo, T.; Oshima, T. Chem. Lett.
1987, 1749. (c) Kakinuma, K.; Ozawa, K.; Fujimoto, Y.; Akutsu, N.;
Oshima, T. J . Chem. Soc., Chem. Commun. 1989, 1190.
(3) (a) Nakamoto, T.; Vennesland, B. J . Biol. Chem. 1960, 235, 202.
(b) England, S.; Listowsky, I. Biochem. Biophys. Res. Commun. 1963,
12, 356. (c) Hurley, J . H.; Dean, A. M.; Sohl, J . L.; Koshland, D. E.,
J r.; Stroud, R. M. Science 1990, 249, 1012.
(4) Imada, K.; Sato, M.; Tanaka, N.; Katsube, Y.; Matsuura, Y.;
Oshima, T. J . Mol. Biol. 1991, 222, 725.
(5) Hurley, J . H.; Dean, A. M. Structure 1994, 2, 1007.
(6) Kadono, S.; Sakurai, M.; Moriyama, H.; Sato, M.; Hayashi, Y.;
Oshima, T.; Tanaka, N. J . Biochem. (Tokyo) 1995, 118, 745.
10.1021/jo982206c CCC: $18.00 © 1999 American Chemical Society
Published on Web 08/05/1999

6160 J . Org. Chem., Vol. 64, No. 17, 1999
Chiba et al.
by X-ray crystallography.7 The structure was in a fully
closed conformation in contrast to the open conformation
of the Thermus IPMDH-IPM complex. The authors
suggest that the isopropyl group is recognized by a unique
hydrophobic pocket, which includes Glu88, Leu91, and
Leu92 of the subunit 1 and Val193′ of the subunit 2
(corresponding to Glu87, Leu90, Leu91, and Val188′ in
Thermus IPMDH, respectively). In particular, the side
chain, Câ and Cγ moiety, of Glu88 interacts with the
isopropyl group of IPM and plays a central role in the
recognition of the substrate. However, a description of
the structure of the ternary complex of IPM is desirable
since the enzyme-cofactor-substrate interaction has not
been fully rationalized.
F igu r e 2. Enzyme reaction of F-IPM.
One approach to this objective is to prepare an ap-
propriate ternary complex with a highly potent inhibitor,
so that one may obtain appropriate crystals of enzyme-
cofactor-inhibitor complex and thus elucidate the precise
nature of the substrate and cofactor binding in the
enzyme active site. Wittenbach et al.8 and Pirrung et al.9
have reported several inhibitors for IPM. In particular,
O-isobutenyl and O-methyl oxalylhydroxamate were
shown to be potent inhibitors against Salmonella IPMDH
(Ki ) 31 and 15 nM, respectively).8 They suggested that
the hydroxamates could be deprotonated (because of the
acidic amide NH group) and that the deprotonated form,
which mimics the intermediary enolate of the IPMDH
reaction, was bound to the enzyme active site. In other
words, these inhibitors bound to the enzyme in such a
way that the protonation of the enolate intermediate was
inhibited in the later productive step. Therefore, while
these inhibitors are highly potent, it seems rather dif-
ficult to analyze the initial binding of IPMDH with the
substrate and cofactor by use of these hydroxamates.
nism-based inhibitor, anticipating that it can be con-
verted in situ to a Michael acceptor in the enzyme active
site.
VM was expected to be recognized by the enzyme and
be oxidized accordingly as a substrate since 3-ethylmalate
was a good substrate for IPM13 (Figure 3). The second
step of the enzyme reaction may proceed via two possible
pathways. In one pathway (path A), the conjugated
enolate (v), formed after the direct decarboxylation as
in the normal reaction, may abstract a proton of a
neighboring amino acid to give a product (vi). Subse-
quently, the deprotonated residue of the enzyme may
attack the electron-deficient olefin functionality of vi to
give an enzyme-inhibitor complex (ix). In the other
pathway (path B), a double bond migration may prefer-
entially give rise to the ethylidenoxaloacetate (vii). The
electron-deficient conjugated olefin functionality of vii is
susceptible to attack by a nearby nucleophilic group of
the IPMDH active center to give viii, which then
spontaneously decarboxylates to 3-substituted 2-oxopen-
tanoate (ix). In either case, the enzyme reaction can
conceivably result in the formation of a covalent bond
between the reaction product and the enzyme as ix.
We have been involved for quite some time in designing
substrate analogues and inhibitors for IPM.10-13 In this
paper we describe the first successful example of the
design and synthesis of a mechanism-based inhibitor for
IPMDH, (2R,3S)-3-vinylmalate (VM, 1). In our previous
studies (Figure 2), 3-(1-fluoro-1-methylethyl)malic acid
(F-IPM) was found to be converted to iii via oxidation
and decarboxylation by IPMDH, but was not inhibitory
to the enzyme.12 The observation that F-IPM had no
inhibitory activity may be explained in two ways. One
possibility is that the expected Michael addition to an
electron-deficient conjugated olefin functionality cannot
take place due to the steric hindrance of iii. The other
possibility is that HF-elimination of ii to yield iii did not
take place within the enzyme active site. To investigate
the latter possibility, we designed VM as a new mecha-
Resu lts a n d Discu ssion
VM was synthesized from readily available diethyl (R)-
malate 2, as shown in Scheme 1. First, the well-
established diastereoselective R-alkylation reaction of
â-hydroxycarboxylic ester was applied to the starting
material 2.14 Thus, 2 was treated with 2 equiv of LDA at
-78 °C to form an intermediary lithium alkoxide enolate,
which was then alkylated with benzyl 2-iodoethyl ether
at -20 °C in THF-HMPA to give stereoselectively
3-alkylated malate in a 15:1 ratio. The desired major
diastereomer 3, possessing the (2R,3S)-configuration, was
obtained after chromatographic separation in 55% yield.15
After acetylation of the hydroxyl group of 3 (71% yield),
the benzyl group of 4 was removed by catalytic hydro-
genation to afford alcohol 5 in 98% yield. The primary
hydroxyl group of 5 was converted to a phenylselenenyl
group by the method of Grieco et al.,16 in which 5 was
treated with N-phenylselenophthalimide and tributylphos-
phine in THF at 0 °C to give phenylselenenide 6 in 68%
(7) Imada, K.; Inagaki, K.; Matsunami, H.; Kawaguchi, H.; Tanaka,
H.; Tanaka, N.; Namba, K. Structure 1998, 6, 971.
(8) (a) Wittenbach, V. A.; Rayner, D. R.; Schloss, J . V. Curr. Top.
Plant Physiol. 1992, 7, 69. (b) Wittenbach, V. A.; Teaney, P. W.; Hanna,
W. S.; Rayner, D. R.; Schloss, J . V. Plant Physiol. 1994, 106, 321.
(9) (a) Pirrung, M. C.; Han, H.; Nunn, D. S. J . Org. Chem. 1994,
59, 2423. (b) Pirrung, M. C.; Han, H.; Ludwig, R. T. J . Org. Chem.
1994, 59, 2430. (c) Pirrung, M. C.; Han, H.; Chen, J . J . Org. Chem.
1996, 61, 4527.
(10) Terasawa, H.; Miyazaki, K.; Oshima, T.; Eguchi, T.; Kakinuma,
K. Biosci. Biotechnol. Biochem. 1994, 58, 870.
(11) Chiba, A.; Eguchi, T.; Oshima, T.; Kakinuma, K. Tetrahedron
1997, 53, 3537.
(12) Aoyama, T.; Eguchi, T.; Oshima, T.; Kakinuma, K. J . Chem.
Soc., Perkin Trans 1 1995, 1905.
(13) (a) Kakinuma, K.; Terasawa, H.; Li, H.-Y.; Miyazaki, K.;
Oshima, T. Biosci. Biotechnol. Biochem., 1993, 57, 1916. (b) Miyazaki,
K.; Kakinuma, K.; Terasawa, H.; Oshima, T. FEBS Lett, 1993, 332,
35.
(14) Seebach, D.; Aebi, J .; Wasmuth, D. Organic Syntheses; Wiley:
New York, 1990; Collect. Vol. VII, p 153 and references therein.
(15) The diastereoselectivity was estimated from the 1H NMR
spectra of a crude reaction mixture, and the stereochemistry of 3 was
deduced from the 1H NMR coupling constant (3J H2,H3 ) 3 Hz); see refs
10 and 13.
(16) Grieco, P. A.; Gilman, S.; Nishizawa, M. J . Org. Chem. 1976,
41, 1485.

(2R,3S)-3-Vinylmalic Acid as an Inhibitor
J . Org. Chem., Vol. 64, No. 17, 1999 6161
F igu r e 3. Possible pathways of the enzyme reaction of VM.
Sch em e 1a
a
Reagents and conditions: (a) 2 equiv of LDA, BnOCH2CH2I, THF-HMPA, -20 °C; (b) Ac2O-DMAP/Py; (c) H2/Pd-C, EtOH; (d)
PhSeNPht, nBu3P, THF; (e) 30% H2O2-THF; (f) 6 M HCl-THF.
yield.17 Oxidative elimination of the phenylselenenyl
group of 6 was accomplished by treatment with hydrogen
peroxide in THF to give a 3-vinylmalate derivative 7 in
82% yield. Finally, acid hydrolysis of 7 with 6 M HCl in
THF afforded, after chromatographic purification and
recrystallization, the desired VM (1) in 32% yield.
The synthesized VM was incubated with thermophilic
IPMDH derived from T. thermophilus HB8 as described
previously.10-13 VM showed a very weak substrate activ-
ity (Km ) 1.85 mM, kcat ) 5.10 × 10-2 s-1) and a potent
inhibition activity against IPMDH. The inhibition ap-
peared to be mechanism-based since incubation of IP-
MDH with VM resulted in a time-dependent inactivation
as shown in Figure 4.19 To obtain KI and kinact values for
VM, the half-time (t1/2) for inactivation at each inhibitor
concentration was plotted against 1/[VM], referred to as
a Kitz and Wilson plot20 (Figure 5) and VM was found to
be remarkably inhibitory (KI ) 1.20 mM, kinact ) 8.16 ×
10-2 s-1). The partition ratio for VM (kcat/kinact ) 0.625)
suggested an extremely strong interaction between the
inactivator and the enzyme, probably by the formation
of a covalent bond. Interestingly, however, we found that
the enzyme activity recovered after dialysis of a VM
incubation mixture. These results may be rationalized
in two possible ways. Covalent bond formation may take
place as anticipated, but the bond formation between the
F igu r e 4. Progress curves for the inactivation of IPMDH by
VM. The reaction was initiated by adding enzyme to an assay
mixture containing IPM (0.1 mM), NAD+ (5 mM), MgCl2 (5
mM), KCl (100 mM), and VM (0-2 mM) in 50 mM HEPES-
NaOH (pH 7.8) at 60 °C. VM concentrations were 0 (open
square), 0.5 (closed square), 1.0 (open circle), and 2.0 mM
(closed circle).
enzyme and the active species is reversible. Alternatively,
the chemically feasible double bond migration from the
isolated to the conjugated double bond after the initial
oxidation might be faster than the decarboxylation. This
would lead to formation of vii, which afterward resides
in the active site of IPMDH for inhibition.
We first examined the latter possibility by comparing
the reaction kinetics between VM and (2R)-3-ethyl-
idenemalic acid 8. Compound 8 was anticipated to be
(17) Nicolaou, K. C.; Claremon, D. A.; Barnette, W. E.; Seitz, S. P.
J . Am. Chem. Soc. 1979, 101, 3704.
(18) Rambaud, M.; Bakasse, M.; Duguay, G.; Villieras, J . Synthesis
1988, 564.
(19) Morrison, J . F.; Walsh, C. T. Adv. Enzymol. 1988, 61, 201.
(20) (a) Kitz, R.; Wilson, I. B. J . Biol. Chem. 1962, 237, 3245. (b)
Silverman, R. B. Mechanism Based Enzyme Inactivation: Chemistry
and Enzymology; CRC Press Inc.: Boca Raton, FL, 1988.

6162 J . Org. Chem., Vol. 64, No. 17, 1999
Chiba et al.
F igu r e 5. Kitz and Wilson plot for the inactivation of IPMDH
by VM. A preincubated solution of IPMDH containing an
inhibitor (0-4 mM) was diluted 100-fold into the enzyme assay
mixture containing IPM (0.1 mM), NAD+ (5 mM), MgCl2 (5
mM), and KCl (100 mM) in 50 mM HEPES buffer (pH 7.8).
The half-time (t1/2) for inactivation at each inhibitor concentra-
tion was plotted against 1/[VM].
F igu r e 6. Progress curves for the inactivation of IPMDH by
8. The reaction was initiated by adding enzyme to an assay
mixture containing IPM (0.1 mM), NAD+ (5 mM), MgCl2 (5
mM), KCl (100 mM), and 8 (0-150 mM) in 50 mM HEPES-
NaOH (pH 7.8) at 60 °C. The concentrations of 8 were 0 (open
square), 50 (closed square), 100 (open circle), and 150 mM
(closed circle).
Sch em e 2a
a
Reagents and conditions: (a) 2 equiv of LDA, Etl, THF-
HMPA, -20 °C; (b) equiv of LDA, PhSeBr; (c) 30% H2O2-THF;
(d) LiOH, THF-H2O.
recognized and oxidized by IPMDH to give vii. The
synthesis of 8 also started from (R)-malate 2 (Scheme
2). First, (R)-malate 2 was treated with 2 equiv of LDA
and then alkylated with ethyl iodide to give (2R,3S)-3-
ethylmalate 9 in 9:1 ratio.14 3-Ethylmalate 9 was in turn
treated with 2 equiv of LDA, and a phenylselenenyl group
was introduced into the 3-position of 9. Subsequent
treatment with H2O2 and hydrolysis gave (2R)-3-ethyl-
idenemalic acid 8 (E:Z ) 25:1). Regarding the geometrical
configuration of 8, the major compound was found to have
(E)-configuration since an NOE was observed between
the CH3 of the 3-ethylidene group and the proton at C-2.
The synthesized 8 was incubated with IPMDH as in
the case of VM as described above. A very weak substrate
activity (Km ) 234 mM, kcat ) 8.54 × 10-3 s-1) and a
modest inhibitory activity (KI ) 153 mM, kinact ) 8.13 ×
10-2 s-1, kcat/kinact ) 0.105) were observed. Since the time-
dependent inactivation of IPMDH was observed (Figure
6),19 the inhibition appeared to be mechanism-based and
3-ethylidenemalate was found to be converted to vii by
IPMDH. Further, it seems less likely that the inhibition
of VM took place through the formation of 8, because the
kcat and kcat/kinact values of 8 were distinctly different and
far lower from those of VM. Thus, the enzyme reaction
of the powerful inhibitor VM proceeds via path A.
We next analyzed the enzyme reaction product of VM
by incubating on a large scale. The resulting reaction
mixture was divided into two parts, and each was
F igu r e 7. GC-MS analysis of the enzyme reaction product
of VM. (A) Mass spectrum of the nondeuterated product at the
retention time of 7.0 min. (B) Mass spectrum of the deuterated
product at the same retention time.
separately reduced, with either NaBH4 or NaBD4, and
dried by lyophilization. Both residues were treated with
diazomethane, and the ethereal extracts thereof were
subsequently analyzed by GC-MS. The rationale of this
protocol was that, once VM was oxidized by IPMDH, the
product might well contain a carbonyl group. The ex-
pected carbonyl group must then be reduced with NaBH4
or NaBD4 to yield a nondeuterated or deuterated alcohol
derivative. The difference in the products should be only
one mass unit in MS, while the GC retention time would
be the same. The GC-MS ion trace showed one peak at
7 min that exhibited the desired 1 amu difference in m/z
(Figure 7). Since the retention time of this peak in the
GC column was shorter than that of VM dimethyl ester
(tR ) 12 min), the enzyme reaction product appeared to
have a lower molecular weight than VM. A possible
candidate was methyl 2-hydroxy-3-pentenoate 11. Au-
thentic 11 was synthesized as described in Scheme 3 and
compared with the derivative from the enzyme reaction
by GC-MS analysis. The retention time (tR ) 7 min) and
fragment ions of the authentic sample were identical to

(2R,3S)-3-Vinylmalic Acid as an Inhibitor
J . Org. Chem., Vol. 64, No. 17, 1999 6163
role in the recognition of the isopropyl group of IPM.7
Hence, the carboxylate of Glu87 may be a candidate to
form the transient bond as ix, although there are some
other possible Glu and Asp residues in the active site.
In conclusion, VM is the first successful example of a
mechanism-based inhibitor that shows a very potent
inhibition toward IPMDH. The involvement of ethylide-
noxaloacetate seemed less likely as the results of incuba-
tion of the synthesized (2R)-3-ethylidenemalic acid. The
enzyme reaction with VM was followed by GC-MS and
NMR spectroscopic studies. These studies showed that
VM was recognized by the enzyme and oxidized normally
as a substrate to yield 2-oxo-3-pentenoate (vi). Though
the enzyme-inhibitor complex ix was not directly de-
tected by MALDI-TOF mass spectrometry of the enzyme
reaction, transient bond formation between IPMDH and
a product of VM was strongly suggested by the time-
dependent and kinetic analyses. The development of a
highly potent inhibitor as VM may allow us to obtain
crystals of the IPMDH-inhibitor-NAD+ complex ap-
propriate for crystallographic analysis, and efforts are
being directed toward this goal.
Sch em e 3a
a
Reagent and conditions: (a) DIBAH, Et2O, -78 °C.
F igu r e 8. 1H NMR spectrum of the enzyme reaction mixture
of VM (0.80 mmol, final 3.98 mM) for 3 h at 25 °C. The asterisk
peaks were derived from 2-oxo-3-pentenoate.
those of the enzyme reaction product. Similarly, the
enzyme reaction product was identified by 1H NMR
spectroscopy. VM (0.80 mmol, 3.98 mM) was incubated
with a large quantity of IPMDH (0.08 mmol, 0.39 mM)
at 25 °C in an NMR tube. The signals of 2-oxo-3-
pentenoate (δH 2.0, 6.2, and 7.0 ppm) clearly appeared
in the H NMR spectrum (Figure 8). It appeared, there-
fore, that 2-oxo-3-pentenoate was an end product from
VM.
Exp er im en ta l Section
1
Melting points are uncorrected. H and 13C NMR chemical
shifts are reported in δ value based on internal TMS (δH ) 0)
or dioxane (δC ) 66.5), or solvent signal (CDCl3, δC ) 77.0;
HOD, δH ) 4.8) as reference. Silica gel column chromatography
was carried out with Kieselgel 60 (70-230 mesh, Merck). All
reactions, except for catalytic hydrogenation reaction, were
carried out in an inert (Ar or N2) atmosphere. Tetrahydrofuran
was distilled immediately prior to use from sodium/benzophe-
none ketyl. Pyridine was distilled from potassium hydroxide.
Toluene was distilled from calcium hydride. HMPA was
distilled from calcium hydride under reduced pressure and
stored over molecular sieves 4A under Ar.
Dieth yl (2R,3S)-3-(2-Ben zyloxyeth yl)m a la te (3). A solu-
tion of BuLi (2.0 mL, 3.2 mmol, 1.60 M) was added dropwise
to a stirred solution of diisopropylamine (0.54 mL, 3.8 mmol)
and 2,2′-dipyridyl (5 mg) in THF (6.3 mL) at 0 °C. After 30
min, the mixture was cooled to -78 °C. A solution of diethyl
(R)-malate 2 (301 mg, 1.58 mmol) in THF (0.2 mL) was added
dropwise at -78 °C. After completion of addition, the mixture
was warmed to room temperature and stirred for 1 h. HMPA
(1.3 mL) was added to the solution. Then the solution was
recooled to -20 °C, and a solution of benzyl 2-iodoethyl ether
(500 mg, 1.90 mmol) in THF (0.2 mL) was added. After 15 min,
the reaction was quenched by addition of water, and the
mixture was extracted with ethyl acetate (50 mL). The organic
phase was subsequently washed with saturated aqueous NH4-
Cl solution and brine, dried over Na2SO4, and evaporated. The
resulting residue was chromatographed over silica gel with
hexanes-ether (5:1-1:1) to give 3 (280 mg, 55%) as an oil:
IR (film) 1740, 1640 cm-1;1H NMR (500 MHz, CDCl3) δ 1.25
(t, J ) 7 Hz, 3H), 1.34 (t, J ) 7 Hz, 3H), 2.00 (m, 1H), 2.23 (m,
1H), 3.19 (dt, J ) 3 and 8 Hz, 1H), 3.33 (d, J ) 8 Hz, 1H), 3.63
(m, 2H), 4.15 (m, 2H), 4.28 (m, 2H), 4.33 (dd, J ) 3 and 8 Hz,
1H), 4.51 (s, 2H), 7.35 (m, 5H); 13C NMR (125.8 MHz, CDCl3)
δ 14.0, 14.1, 28.1, 45.4, 60.8, 61.8, 67.5, 70.9, 72.9, 127.6, 128.3,
138.2, 172.3, 173.3. Anal. Calcd for C17H24O6: C, 62.95; H, 7.46.
Found: C, 62.86; H, 7.65.
1
Path A is the most likely for the formation of 2-oxo-3-
pentenoate from VM. This pathway involves the normal
oxidation and decarboxylation processes of the IPMDH
reaction. It was already reported that (2R,3S)-ethyl-
malate was a very good substrate for IPM.13b The only
structural difference between the products of VM and
ethylmalate is the conjugated enone moiety, which must
therefore be responsible for the inhibitory reaction of VM.
The most realistic scenario is as follows. In the final step
of enolate quenching, the conjugated dienolate v may
abstract a proton, from the enzyme or from the environ-
ment, and be protonated at the terminal methylene
position. The amino acid residue, probably activated by
the proton abstraction, may subsequently attack vi to
form a transient covalent bond. Attempted MALDI-TOF
analysis to detect an expected enzyme-inhibitor complex
has so far been unsuccessful.
With regard to the other possible inhibition mechanism
of VM, either the enolate v or the enzyme reaction
product vi might inhibit the normal enzyme reaction
without covalent bond formation. If the inhibition was
due to the enolate v residing in the active site, it would
imply that v is as chemically stable as the deprotonated
forms of the oxalylhydroxamates, which is unlikely. The
intact or free vi is similarly unlikely for the inhibition,
since it would be expelled out to the outside of the enzyme
as a regular reaction product. To figure out this possibil-
ity, the supplemented 2-oxo-3-pentenoate was incubated
with IPMDH, but no inhibition was observed. Accord-
ingly, the transient bond formation is the most likely
mechanism for the inhibition by VM, although the
anticipated complex ix has not yet been directly detected.
The most recent X-ray analysis of the Thiobacillus
IPMDH-IPM complex suggested by analogy that Câ and
Cγ of Glu87 in Thermus IPMDH may play an important
Dieth yl (2R,3S)-2-O-Acetyl-3-(2-ben zyloxyeth yl)m a la te
(4). A mixture of 3 (2.69 g, 8.29 mmol), DMAP (5 mg), and
acetic anhydride (5.4 mL) in pyridine (27 mL) was stirred for
30 min. The reaction was quenched by addition of water at 0
°C. The reaction mixture was extracted with ethyl acetate (150
mL × 2). The combined organic phase was subsequently
washed with 6 M HCl, saturated aqueous NaHCO3 solution,
and brine, dried with Na2SO4, and evaporated. The resulting
residue was chromatographed over silica gel with hexanes-
ether (5:1-1:1) to give 4 (2.16 g, 71%) as an oil: IR (film) 1740

6164 J . Org. Chem., Vol. 64, No. 17, 1999
cm-1 1H NMR (270 MHz, CDCl3) δ 1.24 (t, J ) 7 Hz, 3H),
1.29 (t, J ) 7 Hz, 3H), 1.82 (dq, J ) 14 and 6 Hz, 1H), 2.03 (s,
3H), 2.05 (m, 1H), 3.27 (ddd, J ) 5, 6, and 8 Hz, 1H), 3.55 (t,
J ) 6 Hz, 2H), 4.15 (m, 2H), 4.23 (m, 2H), 4.49 (s, 2H), 5.29
(d, J ) 5 Hz, 1H), 7.32 (m, 5H);13C NMR (50.3 MHz, CDCl3) δ
14.1, 20.6, 27.6, 43.4, 61.0, 61.4, 67.3, 72.2, 72.9, 127.5, 128.3,
168.2, 170.0, 170.3. Anal. Calcd for C19H26O7: C, 62.28; H, 7.15.
Found: C, 62.46; H, 6.85.
Chiba et al.
;
dropwise at -78 °C. After completion of addition, the mixture
was warmed to -20 °C, and HMPA (80 mL) was added. After
1 h, a solution of ethyl iodide (3.4 mL, 42.5 mmol) in THF (3.4
mL) was added dropwise to the mixture at -20 °C. After 10
min, the reaction was quenched by addition of saturated
aqueous NH4Cl solution. The solution was diluted with ethyl
acetate and 2 M HCl. The organic phase was separated. The
aqueous layer was extracted twice with ethyl acetate. The
combined organic phase was subsequently washed with satu-
rated aqueous NaHCO3 solution and brine, dried over Na2-
SO4, and evaporated. The resulting residue was chromato-
graphed over silica gel with hexanes-ethyl acetate (5:1) to give
Dieth yl (2R,3S)-2-O-Acetyl-3-(2-h yd r oxyeth yl)m a la te
(5). A mixture of 4 (811 mg, 2.21 mmol) and 10% Pd/C (80
mg) in ethanol (10 mL) was vigorously stirred for 6 h under
hydrogen atmosphere. The catalyst was filtered off through a
Celite pad and washed with ethanol. The filtrate and washings
were combined and evaporated to give 5 (631 mg, 98%) as an
9 (2.90 g, 63%) in a 9:1 ratio: IR (neat) 3500, 2980, 1740 cm-1
;
1H NMR of the major product (300 MHz, CDCl3) δ 1.02 (t, J )
7 Hz, 3H), 1.25 (t, J ) 7 Hz, 3H), 1.31 (t, J ) 7 Hz, 3H), 1.70
(m, 1H), 1.89 (m, 1H), 2.78 (dt, J ) 3 and 7 Hz, 1H), 3.22 (d,
J ) 7 Hz, 1H), 4.15 (q, J ) 7 Hz, 2H), 4.27 (m, 3H); 13C NMR
of the major product (75.4 MHz, CDCl3) δ 11.9, 14.1, 21.4, 50.2,
60.7, 61.8, 70.8, 172.7, 173.5. Anal. Calcd for C10H18O5: C,
55.03; H, 8.31. Found: C, 55.29; H, 8.49.
1
oil: IR (film) 3550, 1740 cm-1; H NMR (200 MHz, CDCl3) δ
1.27 (t, J ) 7 Hz, 3H), 1.29 (t, J ) 7 Hz, 3H), 1.81 (m, 1H),
2.03 (m, 1H), 2.15 (s, 3H), 3.12 (dt, J ) 9 and 5 Hz, 1H), 3.74
(t, J ) 6 Hz, 2H), 4.20 (m, 2H), 4.22 (m, 2H), 5.33 (d, J ) 5
Hz, 1H); 13C NMR (50.3 MHz, CDCl3) δ 14.1, 20.6, 30.3, 43.6,
60.3, 61.2, 61.7, 72.4, 168.2, 170.0, 171.2. Anal. Calcd for
C
12H20O7: C, 52.17; H, 7.30. Found: C, 52.21; H, 7.25.
Dieth yl (2R,3S)-2-O-Acetyl-3-(2-p h en ylselen en yleth yl)-
Dieth yl (2R)-3-Eth ylid en em a la te (10). A solution of BuLi
(50.5 mL, 80.8 mmol, 1.60 M) was added dropwise to a stirred
solution of diisopropylamine (11.9 mL, 84.9 mmol) and 2,2′-
dipyridyl (5 mg) in THF (180 mL) at 0 °C. After 30 min, the
mixture was cooled to -78 °C, and a solution of 9 (8.39 g, 38.4
mmol) in THF (20 mL) was added dropwise at -78 °C. After
completion of addition, the mixture was warmed to 0 °C and
stirred for 2 h. The mixture was recooled to -78 °C, and a
solution of phenylselenenyl bromide (13.6 g, 57.6 mmol) in
THF (25 mL) was added dropwise at -78 °C. After 20 min,
the reaction was quenched by addition of saturated aqueous
NH4Cl solution. The solution was diluted with ethyl acetate
and 2 M HCl. The organic phase was separated. The aqueous
layer was extracted twice with ethyl acetate. The combined
organic phase was subsequently washed with saturated aque-
ous NaHCO3 solution and brine, dried over Na2SO4, and
evaporated. The resulting residue was chromatographed over
silica gel with hexanes-ethyl acetate (10:1-3:1) to give the
phenylselenenyl compound (2.81 g).
m a la te (6). N-(Phenylseleno)phthalimide (1.0 g, 3.35 mmol)
was added to a stirred solution of 5 (615 mg, 2.23 mmol) and
Bu3P (0.98 mL, 4.8 mmol) in THF (12 mL), and the mixture
was stirred for 1.5 h. The reaction was quenched by addition
of saturated aqueous NH4Cl solution. The mixture was ex-
tracted with ethyl acetate (100 mL). The organic phase was
subsequently washed with brine, dried with Na2SO4, and
evaporated. The resulting residue was chromatographed over
silica gel with hexanes-ethyl acetate (5:1) to give 6 (628 mg,
68%) as an oil: IR (film) 1740 cm-1; 1H NMR (270 MHz, CDCl3)
δ 1.23 (t, J ) 7 Hz, 3H), 1.24 (t, J ) 7 Hz, 3H), 1.84 (m, 1H),
2.09 (s, 3H), 2.19 (m, 1H), 2.88 (dt, J ) 13 and 8 Hz, 1H), 3.01
(ddd, J ) 6, 9, and 13 Hz, 1H), 3.22 (dt, J ) 9 and 4 Hz, 1H),
4.15 (m, 4H), 5.25 (d, J ) 5 Hz, 1H), 7.25 (m, 3H), 7.49 (m,
2H); 13C NMR (67.9 MHz, CDCl3) δ 14.00,14.05, 20.5, 25.0,
27.9, 46.1, 61.1, 61.6, 72.0, 127.1, 129.1, 132.9, 168.1, 170.0,
170.5. Anal. Calcd for C18H24O6Se: C, 52.05; H, 5.82. Found:
C, 52.14; H, 5.77.
To a solution of the phenylselenenyl compound (2.81 g) in
CH2Cl2 (30 mL) was added 30% aqueous H2O2 solution (1.50
mL, 15.2 mmol) at 0 °C. After 5 h, the mixture was diluted
with ethyl acetate and water. The organic phase was sepa-
rated. The aqueous layer was extracted twice with ethyl
acetate. The combined organic phase was subsequently washed
with saturated aqueous Na2SO3 solution and brine, dried over
Na2SO4, and evaporated. The resulting residue was chromato-
graphed over silica gel with hexanes-ethyl acetate (10:1-3:
1) to give 10 (1.06 g, 13%) as the E-Z mixture (E:Z ) 10:1):
Dieth yl (2R,3S)-2-O-Acetyl-3-vin ylm a la te (7). Aqueous
30% H2O2 (1.62 mL) was added to a stirred solution of 6 (599
mg, 1.44 mmol) in THF (6 mL) at 0 °C. The mixture was stirred
for 1.5 h at 0 °C, and then at room temperature for 5 h. The
mixture was extracted with ethyl acetate (50 mL × 2). The
combined organic phase was subsequently washed with satu-
rated aqueous Na2SO3 solution and brine, dried with Na2SO4,
and evaporated. The resulting residue was chromatographed
over silica gel with benzene-ethyl acetate (100:1) to give 7
1
1
(492 mg, 82%) as an oil: IR (film) 1740, 1640 cm-1; H NMR
IR (neat) 1740, 1650 cm-1; H NMR of (E)-isomer (300 MHz,
(270 MHz, CDCl3) δ 1.26 (t, J ) 7 Hz, 3H), 1.27 (t, J ) 7 Hz,
3H), 2.12 (s, 3H), 3.60 (dd, J ) 7 and 9 Hz, 1H), 4.20 (m, 2H),
4.21 (m, 2H), 5.28 (dt, J ) 17 and 1 Hz, 1H), 5.28 (d, J ) 10
Hz, 1H), 5.28 (d, J ) 7 Hz, 1H), 5.85 (ddd, J ) 9, 10, and 17
Hz, 1H); 13C NMR (67.9 MHz, CDCl3) δ 14.0, 20.4, 51.5, 61.3,
61.5, 72.7, 120.6, 130.2, 168.1, 169.7, 169.9. Anal. Calcd for
CDCl3) δ 1.26 (t, J ) 7 Hz, 3H), 1.28 (t, J ) 7 Hz, 3H), 1.98 (d,
J ) 7 Hz, 3H), 3.62 (d, J ) 7 Hz, 1H), 4.17 (qd, J ) 11 and 7
Hz, 1H), 4.18 (qd, J ) 11 and 7 Hz, 1H), 4.26 (qd, J ) 11 and
7 Hz, 1H), 4.27 (qd, J ) 11 and 7 Hz, 1H), 5.03 (d, J ) 7 Hz,
1H), 7.12 (q, J ) 7 Hz, 1H);13C NMR of (E)-isomer (75.4 MHz,
CDCl3) δ 14.06, 14.07, 14.3, 60.9, 61.9, 66.2, 130.8, 142.5, 165.8,
172.9. Anal. Calcd for C10H16O5: C, 55.55; H, 7.46. Found: C,
55.74; H, 7.76.
C
12H18O6: C, 55.80; H, 7.03. Found: C, 55.66; H, 6.80.
(2R,3S)-3-Vin ylm a lic Acid (1). To a stirred solution of 7
(2.50 g, 9.67 mmol) in THF (75 mL) was added 6 M HCl (35
mL). The mixture was stirred for 26 h at 60 °C and concen-
trated to dryness. The residue was chromatographed over
LiChroprep RP-18 with water. Crystallization from CH3CN-
CHCl3 gave 1 (496 mg, 32%) as a colorless powder: mp 115-
(2R)-3-Eth ylid en em a lic Acid (8). To a stirred solution of
10 (991 mg, 4.58 mmol) in THF (25 mL) and water (25 mL)
was added LiOH‚H2O (968 mg, 23.1 mmol). After 3 h, the
mixture was acidified by addition of 6 M HCl at 0 °C. The
solution was evaporated. The residue was chromatographed
over LiChroprep RP-18 with water. Crystallization from CH3-
CN-CHCl3 gave 8 (442 mg, 60%) as the E-Z mixture (E:Z )
1
118 °C; IR (KBr) 3000, 1710, 1650 cm-1; H NMR (300 MHz,
D2O) δ 3.65 (dd, J ) 5 and 9 Hz, 1H), 4.56 (d, J ) 5 Hz, 1H),
5.35 (d, J ) 16 Hz, 1H), 5.36 (d, J ) 11 Hz, 1H), 5.93 (dt, J )
16 and 9 Hz, 1H); 13C NMR (67.9 MHz, D2O) δ 53.4, 71.3, 120.7,
1
25:1): mp 131-134 °C; IR (KBr) 1720, 1680, 1640 cm-1; H
NMR of the major product (300 MHz, D2O) δ 1.94 (d, J ) 7
Hz, 3H), 5.22 (s, 1H), 7.22 (q, J ) 7 Hz, 1H); 13C NMR of the
major product (75.4 MHz, D2O) δ 13.9, 65.0, 130.0, 146.7,
168.9, 176.3. Anal. Calcd for C6H8O5: C, 45.01; H, 5.04.
Found: C, 45.11; H, 5.05.
Meth yl (E)-2-Hyd r oxy-3-p en ten oa te (11). A solution of
diisobutylaluminum hydride in toluene (4.0 mL, 1 M) was
added dropwise to a solution of methyl (E)-2-oxo-3-pentenoate
(12; 337 mg, 2.63 mmol), prepared from 1-propenylmagnesium
130.8, 174.8, 175.4; [a]27 -43.7° (c ) 1.04, H2O). Anal. Calcd
D
for C6H8O5: C, 45.01; H, 5.04. Found: C, 44.91; H, 5.06.
Dieth yl (2R)-3-Eth ylm a la te (9). A solution of BuLi (27.6
mL, 44.1 mmol, 1.60 M) was added dropwise to a stirred
solution of diisopropylamine (6.5 mL, 46.4 mmol) and 2,2′-
dipyridyl (5 mg) in THF (80 mL) at 0 °C. After 30 min, the
mixture was cooled to -78 °C, and a solution of diethyl (R)-
malate 2 (4.01 g, 21.1 mmol) in THF (4 mL) was added

(2R,3S)-3-Vinylmalic Acid as an Inhibitor
J . Org. Chem., Vol. 64, No. 17, 1999 6165
bromide and dimethyl oxalate according to the method of
Rambaud et al,18 in diethyl ether (20 mL) at -78 °C, and the
mixture was stirred for 1.5 h at the same temperature. The
reaction was quenched by addition of water. The mixture was
diluted with diethyl ether and acidified with 2 M HCl. The
organic phase was separated, and the aqueous layer was
extracted with diethyl ether. The combined organic phase was
washed successively with 2 M HCl and brine, dried over Na2-
SO4, and evaporated. The residue was chromatographed over
silica gel with hexane-diethyl ether (2:1) to give 11 (34 mg,
concentration was plotted against 1/[VM], referred to as a Kitz
and Wilson plot, and KI and kinact values were obtained.20b All
standard errors are less than 20% of the estimates.
GC-MS An a lysis of th e En zym e Rea ction P r od u ct of
VM. 3-Vinylmalic acid (128 µL, 1.25 mmol, 10 mM) was added
to an assay mixture (total volume 800 µL) in 50 mM HEPES
buffer (pH 7.8) containing 5 mM NAD+, 5 mM MgCl2, and 100
mM KCl. The enzyme reaction was started by addition of the
enzyme (30 µL, 0.31 mmol, 10.3 mM). At 1 h after the
incubation started at 25 °C, NaBH4 (6.7 mg, 177 mmol) or
NaBD4 (7.6 mg, 182 mmol) was added, and the mixture was
stirred for 4 h at room temperature. The mixture was lyoph-
ilized. After acetic acid (0.2 mL) and methanol (1 mL) were
added to the residue, the mixture was treated with an excess
of ethereal diazomethane and evaporated. The oily residue was
dissolved in diethyl ether and filtered. The filtrate was
evaporated to give a sample for GC-MS analysis. Gas chro-
matography-mass spectra were taken on a J EOL J MS-AX
505HA mass spectrometer using a column of HP-5. The column
temperature was increased at the rate of 10 °C/min from 70
°C, and the flow rate of He gas was 50 mL/min.
1
10%): IR (CHCl3) 3500, 3000, 1720 cm-1; H NMR (300 MHz,
CDCl3) δ 1.73 (ddd, J ) 7, 2 and 2 Hz, 3H), 2.96 (d, J ) 6 Hz,
1H), 3.81 (s, 3H), 4.61 (dddd, J ) 6, 6, 2 and 2 Hz, 1H), 5.54
(ddq, J ) 15, 6 and 2 Hz, 1H), 5.90 (dqd, J ) 15, 7 and 2 Hz,
1H); 13C NMR (75.4 MHz, CDCl3) δ 17.6, 52.6, 71.3, 127.1,
129.7, 174.1. Anal. Calcd for C6H10O3: C, 55.37; H, 7.74.
Found: C, 55.29; H, 7.51.
En zym e a n d Su bstr a te Assa y. The thermophilic IPMDH
derived from T. thermophiles HB8 was prepared and purified
as described previously.1 IPMDH reaction was monitored by
measuring the NADH absorption at 340 nm on a Shimadzu
UV-160A UV-vis recording spectrophotometer or a Gilford
Responce spectrometer. Kinetic measurements were performed
at 60 °C in an assay mixture (total volume 700 µL) containing
50 mM HEPES buffer (pH 7.8), 5 mM NAD+, 5 mM MgCl2,
and 100 mM KCl. In inhibition assay, the reaction was started
by addition of the enzyme (0.68 µg) to the reaction mixture
with all required components including IPM (5-100 mM) and
the inhibitor (0.01-500 µM). The formation of NADH was
measured for 1 min as described above. Data were graphically
analyzed by Lineweaver-Burk double reciprocal plots. In
substrate assay, the reaction was started by addition of the
enzyme (6.8-97 µg) to the reaction mixture with all required
components including VM (1-10 µM) and 8 (100-250 µM),
and the enzyme reaction was monitored by the above method.
For the inactivation assay, a preincubated solution (7 µL)
of IPMDH (0.7 µg) containing an inhibitor (0-20 mM) was
diluted 100-fold into an assay mixture (total volume 700 µL)
containing 50 mM HEPES buffer (pH 7.8), 5 mM NAD+, 5 mM
MgCl2, 100 mM KCl, and 100 mM IPM at 60 °C, and the
formation of NADH was measured for 1 min as described
above. The half-time (t1/2) for inactivation at each inhibitor
A P u r su it of th e En zym e Rea ction of VM by 1H NMR
Sp ectr oscop y. VM 1 (79.6 µL, 0.796 mmol, 10.0 mM) was
added to an assay mixture (total volume 200 µL) in 50 mM
HEPES buffer (pH 7.8) containing 5 mM NAD+, 5 mM MgCl2,
and 100 mM KCl. The assay solution was lyophilized. The
residue was dissolved in D2O (20 µL)-H2O (150 µL). The
enzyme reaction was started by addition of the enzyme (30
µL, 0.08 mmol, 2.7 mM) to the solution at 25 °C in an NMR
tube. The enzyme reaction was monitored by accumulation of
1H NMR spectral data for 3 h.
Ack n ow led gm en t. This work was supported in part
by a Grant-in-Aid for Scientific Research in Priority
Area from the Ministry of Education, Science, Sports
and Culture, J apan.
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
spectra of the synthesized inhibitors 1 and 8. This material is
available free of charge via the Internet at http://pubs.acs.org.
J O982206C