Studies on the inactivation of bovine liver enoyl-CoA hydratase by (methylenecyclopropyl)formyl-CoA: Elucidation of the inactivation mechanism and identification of cysteine-114 as the entrapped nucleophile

Article abstract of DOI:10.1021/ja011226k

The inhibitory properties of (methylenecyclopropyl)formyl-CoA (MCPF-CoA), a metabolite derived from a natural amino acid, (methylenecyclopropyl)glycine, against bovine liver enoyl-CoA hydratase (ECH) were characterized. We have previously demonstrated tha

Full text of DOI:10.1021/ja011226k

J. Am. Chem. Soc. 2001, 123, 9749-9759
9749
Studies on the Inactivation of Bovine Liver Enoyl-CoA Hydratase by
(Methylenecyclopropyl)formyl-CoA: Elucidation of the Inactivation
Mechanism and Identification of Cysteine-114 as the Entrapped
Nucleophile
Srikanth Dakoji,† Ding Li,‡ Gautam Agnihotri,§ Hui-qiang Zhou,| and Hung-wen Liu*
Contribution from the DiVision of Medicinal Chemistry, College of Pharmacy, and Department of
Chemistry and Biochemistry, UniVersity of Texas, Austin, Texas 78712, and the Department of Chemistry,
UniVersity of Minnesota, Minneapolis, Minnesota 55455
ReceiVed May 17, 2001
Abstract: The inhibitory properties of (methylenecyclopropyl)formyl-CoA (MCPF-CoA), a metabolite derived
from a natural amino acid, (methylenecyclopropyl)glycine, against bovine liver enoyl-CoA hydratase (ECH)
were characterized. We have previously demonstrated that MCPF-CoA specifically targets ECHs, which catalyze
the reversible hydration of R,â-unsaturated enoyl-CoA substrates to the corresponding â-hydroxyacyl-CoA
products. Here, we synthesized (R)- and (S)-diastereomers of MCPF-CoA to examine the stereoselectivity of
this inactivation. Both compounds were shown to be competent inhibitors for bovine liver ECH with nearly
identical second-order inactivation rate constants (kinact/KI) and partition ratios (kcat/kinact), indicating that the
inactivation is nonstereospecific with respect to ring cleavage. The inhibitor, upon incubation with bovine
liver ECH, labels a tryptic peptide, ALGGGXEL, near the active site of the protein, where X is the amino acid
that is covalently modified. Cloning and sequence analysis of bovine liver ECH gene revealed the identity of
the amino acid residue entrapped by MCPF-CoA as Cys-114 (mature sequence numbering). On the basis of
gHMQC (gradient heteronuclear multiple quantum coherence) analysis with [3-13C]-labeled MCPF-CoA, the
ring cleavage is most likely induced by the nucleophilic attack at the terminal carbon of the exomethylene
group (C2′). We propose a plausible inactivation mechanism that involves relief of ring strain and is consistent
with examples found in the literature. In addition, these studies provide important clues for future design of
more efficient and selective inhibitors to control and/or regulate fatty acid metabolism.
Enoyl-CoA hydratase (ECH) catalyzes the reversible hydra-
tion of R,â-unsaturated enoyl-CoA substrate 1 to the corre-
sponding â-hydroxyacyl-CoA product, 2. This physiologically
important reaction is the second step in the â-oxidation pathway
of fatty acid metabolism and is also required for the catabolism
of branched-chain amino acids.1 ECHs have been isolated from
many sources and have been shown to differ in their preference
toward substrates of varied acyl chain lengths, as well as with
regard to the presence or absence of substituents at R- or â-C.2-4
For example, three types of ECH have been identified in higher
animals. The short chain ECH and a trifunctional protein, which
acts not only as an ECH but also as an (L)-3-hydroxyacyl-CoA
dehydrogenase and a 3-ketoacyl-CoA thiolase, are both found
in mitochondria.5,6 A bifunctional protein which possesses ECH
and 3-hydroxyacyl-CoA dehydrogenase activities is present in
peroxisomes.7 Among them, the mitochondrial short chain ECH,
also known as crotonase (EC 4.2.1.17), is the best studied.
Mitochondrial short chain ECH has been purified from bovine
liver,4 rat liver,8 pig heart,9 pig kidney,10 and a few other sources.
It is an extremely efficient catalyst that processes its optimum
substrate, crotonyl-CoA (C4), at a rate (kcat/Km ) 5 × 107 M-1
s-1)2 approaching the diffusion-controlled limit. The rate of the
bovine liver ECH-catalyzed reaction decreases linearly with the
chain length of the enoyl-CoA substrate up to C16.2 The overall
stereochemistry of ECH-catalyzed bond formation/cleavage at
the R and â positions of the enoyl-CoA substrate has been
established to be syn.11 The reaction occurs at the si face of the
conjugated thioester substrate leading to the formation of an
(S)-3-hydroxy-acyl-CoA as the hydration product in almost all
ECHs examined. However, a (R)-specific ECH involved in the
biosynthesis of polyhydroxy alkanoates (PHA) has recently been
cloned from Aeromonas caViae.12
* To whom correspondence should be addressed. Fax: (512) 471-2746.
E-mail: h.w.liu@mail.utexas.edu.
† Current address: Department of Physiology and Program in Neuro-
science, University of California, San Francisco, CA 94702.
‡ Current address: Department of Biology and Chemistry, City University
of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong.
§ Current address: Anti-Microbials and Host Defense, GlaxoSmithKline
Pharmaceuticals, Collegeville, PA 19426.
(5) Uchida, Y.; Izai, K.; Orii, T.; Hashimoto, T. J. Biol. Chem. 1992,
267, 1034-1041.
(6) Carpenter, K.; Pollitt, R. J.; Middleton, B. Biochem. Biophys. Res.
Commun. 1992, 183, 443-448.
| Current address: Medicinal Chemistry Division, GlaxoSmithKline
Pharmaceuticals, Research Triangle Park, NC 27709.
(7) Ishii, N.; Hijikata, M.; Osumi, T.; Hashimoto, T. J. Biol. Chem. 1987,
262, 8144-8150.
(1) Stern, J. R. In The Enzymes; Boyer, P. D., Lardy, H., Myrback, K.,
Eds., Academic Press: New York, 1961; p 511-529.
(8) Furuta, S.; Miyazawa, S.; Osumi, T.; Hashimoto, T.; Ui, N. J.
Biochem. (Tokyo) 1980, 88, 1059-1070.
(2) Waterson, R. M.; Hass, G. M.; Hill, R. L. J. Biol. Chem. 1972, 247,
5258-5265.
(9) Fong, J. C.; Schulz, H. Methods Enzymol. 1981, 71, 309-318.
(10) Buettner, H. Ph.D. Thesis, University of Konstanz, 1988.
(11) Willadsen, P.; Eggerer, H. Eur. J. Biochem. 1975, 54, 247-252.
(3) Steinman, H. M.; Hill, R. L. J. Biol. Chem. 1973, 248, 892-900.
(4) Steinman, H. M.; Hill, R. L. Methods Enzymol. 1975, 35, 136-151.
10.1021/ja011226k CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/15/2001

9750 J. Am. Chem. Soc., Vol. 123, No. 40, 2001
Dakoji et al.
N-terminal signal sequence, whereas the amino acid residue
numbering for bovine liver ECH refers to the mature sequence
of the protein, minus the signal sequence). The residue Glu144
was inferred by the structural analysis to function as the catalytic
base responsible for the activation of the conserved water
molecule in the active site. The role of Glu144 has since been
confirmed by mutagenesis studies.23 In addition, the crystal
structure also reveals the presence of a strong hydrogen bond
(d ) 2.7 Å) between the CdO of the substrate and the N-H
backbone of Gly141, corroborating the polarization of the
conjugated π-system of substrate in the active site as deduced
by spectroscopic studies.15-18 Because both Glu144 and Glu164
are located on the same side of the trans-enoyl double bond,
such disposition is in agreement with the syn hydration/
dehydration stereochemistry.11,24
While the physiological significance of ECHs is apparent and
its catalytic mechanism is well-documented, very few inhibitors
are known for this class of enzymes.2,25 Examples include
several cinnamoyl-CoA derivatives and acetoacetyl-CoA,2 all
of which are competitive inhibitors against ECHs with Ki values
in the micromolar range. However, these compounds lack
specificity and are toxic against most coenzyme A binding
enzymes involved in fatty-acid metabolism.26-28 Recently, 5,6-
dichloro-7,7,7-trifluoro-4-thia-5-heptenoyl-CoA has been re-
ported to be an active site thiol-directed inactivator for rat liver
ECH.29 Interestingly, (methylenecyclopropyl)formyl-CoA (4,
MCPF-CoA), a metabolite of (methylenecyclopropyl)glycine (3)
isolated from kernels of litchi fruits, has been shown to have
hypoglycemic activity due to its ability to inhibit ECH and thus
interrupt the â-oxidation pathway of fatty acid metabolism.30-33
Our recent study confirmed that MCPF-CoA (4) is an irrevers-
ible inhibitor for bovine liver ECH.34,35 Although the inhibition
was established to be time-dependent, active site-directed, and
irreversible, the details of the inactivation mechanism remained
unclear. To determine whether the two stereoisomers of 4 would
elicit different inhibition behaviors, the (R)- and (S)-epimers of
MCPF-CoA (5 and 6, respectively) were prepared and tested
for their competence as inhibitors against bovine liver ECH.
The [3-13C]-labeled analogue of MCPF-CoA (7) was also
synthesized as a probe to help elucidate the chemical nature of
the enzyme-inactivator adduct. The radiolabeled analogue
[3-3H]MCPF-CoA (8) was used to label the protein which, after
tryptic digestion and sequence analysis, allowed the location
of the active site nucleophile being trapped during inactivation
to be determined. Because the protein sequence of bovine liver
ECH is unknown, the corresponding gene was cloned and
Interestingly, studies of the kinetic isotope effects of ECH-
catalyzed reactions have indicated that a single transition state
leads to the formation of both CR-H and Câ-OH bonds,
suggesting that the overall reaction occurs in a concerted
fashion.13,14 Extensive experiments have also established that,
prior to chemical transformation, the substrate and/or its
analogues undergo a reorganization of their conjugated π-elec-
trons in the active site of the enzyme.15-18 This active site
induced polarization of the π-bond effectively enhances the
electrophilicity at Câ of the enoyl-CoA substrate and thus
facilitates the addition of a water molecule across the R,â-
unsaturated thioester. More recently, rat liver ECH has been
shown to add a water molecule across crotonyl-oxyCoA, albeit
with reduced catalytic efficiency.19
The X-ray structures of rat liver ECH complexed with the
tight-binding inhibitors acetoacetyl-CoA and octanoyl-CoA have
been solved to 2.5 and 2.4 Å resolution, respectively.20,21 From
the mutagenesis studies and the crystal structure of this
enzyme,15,22 Glu164 has been firmly assigned as the general
acid that donates a proton to CR of the substrate (the key residues
of rat ECH, such as Gly141, Cys143, Glu144, and Glu164, are
numbered according to the encoded sequence including the
(23) Hofstein, H. A.; Feng, Y.; Anderson, V. E.; Tonge, P. J. Biochem-
istry 1999, 38, 9508-9516.
(12) Fukui, T.; Shiomi, N.; Doi, Y. J. Bacteriol. 1998, 180, 667-673.
(13) Bahnson, B. J.; Anderson, V. E. Biochemistry 1989, 28, 4173-
4181.
(14) Bahnson, B. J.; Anderson, V. E. Biochemistry 1991, 30, 5894-
5906.
(15) D’Ordine, R. L.; Tonge, P. J.; Carey, P. R.; Anderson, V. E.
Biochemistry 1994, 33, 12635-12643.
(16) D’Ordine, R. L.; Bahnson, B. J.; Tonge, P. J.; Anderson, V. E.
Biochemistry 1994, 33, 14733-14742.
(17) Tonge, P. J.; Anderson, V. E.; Fausto, R.; Kim, M.; Pusztai-Carey,
M.; Carey, P. R. Biospectroscopy 1995, 1, 387-394.
(18) Wu, W. J.; Anderson, V. E.; Raleigh, D. P.; Tonge, P. J.
Biochemistry 1997, 36, 2211-2220.
(19) Dai, M.; Feng, Y.; Tonge, P. J. J. Am. Chem. Soc. 2001, 123, 506-
507.
(24) Mohrig, J. R.; Moerke, K. A.; Cloutier, D. L.; Lane, B. D.; Person,
E. C.; Onasch, T. B. Science 1995, 269, 527-529.
(25) He, X. Y.; Yang, S. Y.; Schultz, H. Arch. Biochem. Biophys. 1992,
298, 527-531.
(26) Davidson, B.; Schulz, H. Arch. Biochem. Biophys. 1982, 213, 155-
162.
(27) Powell, P. J.; Lau, S. M.; Killian, D.; Thorpe, C. Biochemistry 1987,
26, 3704-3710.
(28) Schifferdecker, J.; Schulz, H. Life Sci. 1974, 14, 1487-1492.
(29) Baker-Malcolm, J. K.; Lantz, M.; Anderson, V. E.; Thorpe, C.
Biochemistry, 2000, 39, 12007-12018.
(30) Gray, D. O.; Fowden, L. Biochem. J. 1962, 82, 385-389.
(31) Tanaka, K.; Isselbacher, K. J.; Shih, V. Science 1972, 175, 69-71.
(32) Melde, K.; Buettner, H.; Boschert, W.; Wolf, H. P. O.; Ghisla, S.
Biochem. J. 1989, 259, 921-924.
(33) Melde, K.; Jackson, S.; Bartlett, K.; Sherratt, H. S. A.; Ghisla, S.
Biochem. J. 1991, 274, 395-400.
(34) Li, D.; Guo, Z.; Liu, H.-w. J. Am. Chem. Soc. 1996, 118, 275-
276.
(35) Li, D.; Agnihotri, G.; Dakoji, S.; Oh, E.; Lantz, M.; Liu, H.-w. J.
Am. Chem. Soc. 1999, 121, 9034-9042.
(20) Engel, C. K.; Mathieu, M.; Zeelen, J. P.; Hiltunen, J. K.; Wierenga,
R. K. EMBO J. 1996, 15, 5135-5145.
(21) Engel, C. K.; Kiema, T. R.; Hiltunen, J. K.; Wierenga, R. K. J.
Mol. Biol. 1998, 275, 847-859.
(22) Muller-Newton, G.; Janssen, U.; Stoffel, W. Eur. J. Biochem. 1995,
228, 68-73.

InactiVation of BoVine LiVer Enoyl-CoA Hydratase
J. Am. Chem. Soc., Vol. 123, No. 40, 2001 9751
Stratagene (La Jolla, CA). This Lambda cDNA library was screened
with rabbit antibodies generated against bovine liver ECH using the
picoBlue Immunoscreening kit (Stratagene). Two positive clones were
detected following plaque screening (∼ 1 × 106 clones). Subsequent
excision and recircularization of the pBluescript SK(-) phagemid
containing the cloned DNA insert followed protocols recommended
by the manufacturer (Stratagene). The resulting phagemids, designated
as pBluescript SK(()-SDa1 and pBluescript SK(()-SDb1, were char-
acterized by gene walking, and their cDNA was sequenced by an
automated DNA sequencer.
sequenced to facilitate the identification of the modified residue.
Reported in this paper are the results of these experiments and
our comments on the mechanistic insights deduced from these
observations.
Experimental Section
General. The 1H and 13C NMR chemical shifts are reported on the
δ scale relative to an internal standard or appropriate solvent peak with
coupling constants given in hertz. Flash chromatography was performed
in columns of various diameters with J. T. Baker (230-400 mesh) silica
gel by elution with the solvents reported. Analytical thin-layer
chromatography (TLC) was carried out on Merck silica gel 60 G-254
plates (25 mm) and developed with the solvents mentioned. TLC spots
were visualized either with UV light or by heating plates previously
stained with solutions of KMnO4 (1%), vanillin/ethanol/H2SO4 (1:98:
1), or phosphomolybdic acid (7% EtOH solution). Drying agent used
in the routine workup was anhydrous magnesium sulfate. Solvents,
unless otherwise specified, were analytical reagent grade and distilled
once prior to use. For anhydrous reactions, the solvents were pretreated
prior to distillation as follows: tetrahydrofuran (THF) was dried over
sodium and benzophenone; methylene chloride, dimethyl sulfoxide
(DMSO), and dimethyl formamide (DMF) were dried over calcium
hydride; pyridine and triethylamine were dried over KOH.
The concentration of protein was determined by Bradford’s method36
using bovine serum albumin as the standard. However, for routine
purposes, the enzyme concentration was estimated using the extinction
coefficient of 16 000 M-1‚cm-1 at 280 nm.4 The level of expression,
the solubility of the gene product, and the purity of the purified protein
were assessed by sodium dodecyl sulfate-polyacrylamide gel electro-
phoresis (SDS-PAGE) as described by Laemmli.37 The amino acid
sequencing was carried out by the Microchemical Facility at the Institute
of Human Genetics of the University of Minnesota. The DNA
sequencing was performed by the Bioanalytical Center of the University
of Minnesota, St. Paul. DNA sequence analyses were performed with
Wisconsin Sequence Analysis Package of Genetics Computer Group
(Madison, WI) and GeneWorks version 2.5 of IntelliGenetics, Inc.
(Mountain View, CA). Methods and protocols for recombinant DNA
manipulations were according to the manufacturers’ manuals or from
general references.38
Materials. Premade bovine liver cDNA library, Uni-ZAP XR
Library, phagemid vector pBluescript SK((), T3 and T7 sequencing
primers, f1 helper phage (VCSM13 (f1), ExAssist (M13)), picoBlue
Immunoscreening kit, E. coli phage lysate, immunoaffinity-purified goat
anti-rabbit antibodies, nitroblue tertrazolium (NBT), 5-bromo-4-chloro-
3-indoyl phosphate (BCIP), Tween 20, and cloned pfu DNA polymerase
were products of Stratagene (La Jolla, CA). Nitrocellulose membranes,
HindIII digested bacteriophage lambda DNA marker, and agarose were
purchased from GIBCO BRL (Gaithersburg, MD). Geneclean II was
from BIO101, Inc. (La Jolla, CA) and Bradford reagent was from
BioRad (Richmond, CA). Ecoscint A of National Diagnostics (Manville,
NJ) was used as scintillation cocktail. Nucleotide triphosphates (dATP,
dCTP, dGTP, and dTTP (10 mM solution)), Magic Mini Preps, Wizard
Miniprep and Wizard Megaprep DNA purification systems, and buffers
used in PCR (polymerase chain reaction) were obtained from Promega
(Madison, WI). The expression vectors pTrc99A, E. coli JM105, DEAE-
Sepharose (DEAE ) (diethylamino)ethyl), and phenyl-Sepharose CL-6
were purchased from Pharmacia (Uppsala, Sweden). Custom-designed
oligonucleotide primers were made by Integrated DNA Technologies
(Coralville, IA). Rabbit antiserum against the bovine liver ECH was
custom prepared by HTI Biochemicals (Ramona, CA). The molecular
weight standards, biochemicals, and chemicals were purchased from
Sigma or Aldrich (Milwaukee, WI) and were of the highest purity
available.
PCR Amplification and Cloning of Bovine Liver ECH Gene. The
gene was amplified by standard recombinant DNA techniques and was
subcloned into EcoRI and BamHI sites of the expression vector
pTrc99A. The start primer, 5′-GCGGAATTCCATATGAGCGCAG-
CTTTCGAATAC-3′, contained a EcoRI restriction site (in bold), a TA-
rich region, a start codon, and the codons for the first six amino acid
residues of the desired enzyme. The halt primer, 5′-GCGGGATCCT-
CACTGGTCTTTGAAGTT-3′, introduced a HindIII restriction site (in
bold) immediately downstream from the stop codon of the ECH gene.
These primers were used to amplify the desired gene from pBluescript
SK(()-SDa1 by polymerase chain reaction (PCR). The PCR reactions
were run in a total volume of 100 µL and contained 10 µL DMSO, 1
µL glycerol, 10 µL 10× cloned pfu buffer, 4 µL dNTP mix (10 mM
each of dATP, dCTP, dGTP, and dTTP), 0.5 µL each of start and halt
primer (100 pmol/µL), 1 µL pBluescript SK(()-SDa1 DNA (20 ng),
and 1.0 µL cloned pfu DNA polymerase (2.5 units). After mixing, the
reaction mixture was overlaid with a drop of mineral oil and was
subjected to the following thermal cycles: five cycles of one type (94
°C, 5 min; 50 °C, 5 min; 72 °C, 3 min) and thirty cycles of a second
type (94 °C, 1 min; 55 °C, 1 min; 72 °C, 1.5 min), followed by
incubation at 72 °C for an additional 10 min. The PCR product was
purified using GeneClean II, digested with the appropriate restriction
enzymes, and ligated into the EcoRI/HindIII sites of the transcription
vector pTrc99A. This recombinant plasmid was isolated and was used
to transform E. coli JM101. The plasmid DNA from positive clones
was used to transform E. coli JM105.
Purification of Recombinant Bovine Liver ECH. An overnight
culture of E. coli JM105/pSD6 (30 mL), grown in Luria-Bertani (LB)
medium supplemented with ampicillin (100 µg/mL) at 37 °C, was
evenly divided and added to three 1 L portions of the same LB/
ampicillin medium in Pyrex Erlenmeyer flasks. The flasks were shaken
at 250 rpm at 25 °C until the OD600 reached 0.66. Thereafter, isopropyl
â-D-thiogalactoside (IPTG) was added to the culture to a final
concentration of 1.0 mM, and the incubation was continued for an
additional 6 h at 25 °C. The cells were harvested by centrifugation at
4000g for 15 min at 4 °C. Subsequent operation to purify the enzyme
was carried out at 4 °C and is described below.
Step 1: Crude Extract. The cell pellet was resuspended in 100
mL of 50 mM potassium phosphate buffer containing 3 mM EDTA
and 1 mM DTT (dithiothreitol) (pH 7.4). This suspension was sonicated
with five 1 min bursts with a 1 min cooling period between each burst
to disrupt the cells. Cell debris was removed by centrifugation (17 000g,
20 min). The supernatant was collected and designated the crude extract.
Step 2: Ammonium Sulfate Fractionation. The crude extract (75
mL) from step 1 was treated with solid ammonium sulfate (18.22 g),
added portionwise over 1 h up to 40% saturation. This suspension was
centrifuged at 17 000g for 25 min to remove the precipitates, and to
the supernatant was added more solid ammonium sulfate (24.5 g) to a
final concentration of 85% saturation. This cloudy solution was stirred
for an additional hour, and the precipitated proteins were collected by
centrifugation (17 000g, 25 min). The protein pellet was resuspended
in a minimal amount of 50 mM potassium phosphate buffer containing
3 mM EDTA, pH 7.4, and dialyzed against the same buffer overnight
with three changes of buffer.
Isolation of Bovine Liver ECH cDNA Clones. A bovine liver
Lambda cDNA library in a Uni-ZAP XR vector was purchased from
Step 3: DEAE-Sepharose Chromatography. The dialysate from
step 2 was clarified by centrifugation (17 000g, 20 min) and applied
to a DEAE-Sepharose column (1 cm × 17 cm) preequilibrated with
50 mM potassium phosphate buffer containing 3 mM EDTA, pH 7.4.
The column was eluted with the same buffer (500 mL total volume)
with a flow rate of 1 mL/min. The active fractions that were not retained
were pooled and concentrated by ultrafiltration on an Amicon concen-
(36) Bradford, M. M. Anal. Biochem. 1976, 72, 248-254.
(37) Laemmli, U. K. Nature 1970, 227, 680-685.
(38) Sambrook, J.; Fritsch, E. F.; Maniatis, T. Molecular Cloning: A
Laboratory Manual; Cold Spring Harbor Laboratory Press: Cold Spring
Harbor, NY, 1989.

9752 J. Am. Chem. Soc., Vol. 123, No. 40, 2001
Dakoji et al.
trator using a YM-10 membrane. The concentrated solution was then
dialyzed against 2 L of 50 mM potassium phosphate buffer containing
0.5 M NaCl and 3 mM EDTA, pH 7.4, overnight, with four changes
of buffer.
13. The crude products were purified repeatedly by flash chromatog-
raphy (30% ethyl acetate in hexanes) to separate the diastereomers.
The combined yield of purified compounds 12 (1.2 g) and 13 (1.1 g)
was 48%. The enantiomeric purity of each preparation was determined
by HPLC using a Spheris ORB silica gel (1 cm × 25 cm) that was
eluted with 10% 2-propanol in methylene chloride (flow rate 4 mL/
min, monitoring wavelength 254 nm). Under these conditions, the
retention times of isomers 12 and 13 were found to be 5.9 and 6.8
min, respectively. The enantiomeric purities of both isomers were found
to be greater than 99%. 12: 1H NMR (CDCl3) δ 7.37-7.28 (5H, m,
Ar H), 6.43 (1H, br s, NH), 5.54 (2H, br s, dCH2), 5.01 (1H, dt, J )
12.1 and 5.0, H-1′), 3.82 (2H, t, J ) 5.0, H-2′), 2.92 (1H, br s, OH),
2.16 (1H, m, H-1), 1.73 and 1.58 (1H each, m, H-3); 13C NMR (CDCl3)
δ 171.7 (CdO), 139.0 (Ar C), 130.6 (C-2), 128.9, 127.9, 126.7 (Ar),
105.3 (dCH2), 66.5 (C-2′), 56.1 (C-1′), 20.0 (C-1), 11.2 (C-3). 13: 1H
NMR (CDCl3) δ 7.38-7.27 (5H, m, Ar H), 6.33 (1H, br s, NH), 5.55
(2H, br s, dCH2), 5.04 (1H, dd, J ) 11.5 and 5.0, H-1′), 3.85 (2H, m,
H-2′), 2.77 (1H, br s, OH), 2.17 (1H, m, H-1), 1.72 and 1.62 (1H each,
m, H-3). 13C NMR (CDCl3) δ 171.7 (CdO), 139.1 (Ar C), 130.7 (C-
2), 128.9, 127.9, 126.7 (Ar C), 105.2 (dCH2), 66.4 (C-2′), 56.0 (C-1′),
20.0 (C-1), 11.2 (C-3). High resolution FAB-MS: calcd for C13H16-
NO2 (M + 1)+, 218.1181; found, 218.1165.
(R)-(Methylenecyclopropyl)formic Acid (14) and (S)-(Methyl-
enecyclopropyl)formic Acid (15). An equal volume of 4 N sulfuric
acid (25 mL) was added to a solution of 12 (1.2 g, 5.3 mmol) in THF
(25 mL). After refluxing for 4 h, the solution was cooled to room
temperature and extracted with ether. The combined organic extracts
were dried over anhydrous magnesium sulfate and filtered. The filtrate
was concentrated in vacuo, the residue was purified by flash chroma-
tography (10% ethyl acetate in hexanes), and 14 was obtained in a
final yield of 89%. 1H NMR (CDCl3) δ 5.52 (2Η, d, J ) 1.8, dCH2),
2.22 (1H, m, H-1), 1.85 and 1.68 (1H each, m, H-3); 13C NMR (CDCl3)
δ 179.0 (CdO), 129.9 (C-2), 105.0 (dCH2), 17.9 (C-1), 12.2 (C-3).
High resolution CI-MS (CH4): calcd for C5H7O2 (M + 1)+, 99.0446;
found, 99.0445. Compound 15 was prepared by a similar procedure in
87% yield. The NMR spectra of 15 are identical to those obtained for
14.
(R)-(Methylenecyclopropyl)formyl-CoA (5) and (S)-(Methyl-
enecyclopropyl)formyl-CoA (6). To a solution of 14 in methylene
chloride was added an equimolar amount of triethylamine under an
argon atmosphere. After the mixture was stirred for 10 min, an
equimolar amount of isobutyl chloroformate was added dropwise at 0
°C. The reaction was agitated vigorously, and fuming was noted during
the process. The reaction was stirred at room temperature for 1 h.
Thereafter, the solvent was removed under reduced pressure, and the
residue containing the mixed anhydride was redissolved in THF (5 mL)
to give a cloudy solution. Meanwhile, a solution of coenzyme A was
prepared by dissolving its sodium salt (50 mg, 50 µmol) in distilled
water (5 mL) that had been deoxygenated by repeated freeze-thaw
cycles under vacuum. The pH of this solution was brought up to 8.0
by the addition of 1 N NaOH. To this solution was slowly added the
mixed anhydride solution via a cannula using positive argon pressure.
The pH of the resulting mixture was adjusted to 8.0, and the reaction
mixture was stirred for 10 min. The pH of the reaction was then changed
to 5.0-5.5 by the addition of dilute perchloric acid. The organic solvent
(THF) was removed under reduced pressure, and the remaining aqueous
solution was extracted twice with ether to remove any residual organic
soluble materials. The aqueous solution was then lyophilized. The crude
product was purified by HPLC using a C18 column (10 mm × 250
mm, 5 µm) and eluted with 30% methanol in 50 mM potassium
phosphate buffer (pH 5.3, flow rate 3 mL/min, monitoring wavelength
260 nm). The fractions containing the product were pooled, concentrated
under reduced pressure to remove methanol, and lyophilized. The
resulting residue was desalted on the same column, which was washed
with water (3 mL/min) for 20 min followed by methanol. The eluant
containing the product was evaporated under reduced pressure and then
lyophilized, providing (R)-MCPF-CoA (5) as a white powder. The
Step 4: Phenyl-Sepharose Chromatography. The dialyzed solution
from step 3 was applied to a phenyl-Sepharose column (1 cm × 17
cm) that had been equilibrated with 50 mM potassium phosphate buffer
containing 0.5 M NaCl and 3 mM EDTA, pH 7.4. After loading, the
column was washed with 20 mL of the same buffer, followed by elution
with a linear gradient of NaCl from 0.5 to 0 M in 50 mM potassium
phosphate buffer containing 3 mM EDTA, pH 7.4 (1 L total volume).
The flow rate was 1 mL/min, and 10 mL fractions were collected. The
active fractions were combined, concentrated by ultrafiltration (YM-
10), and stored at -80 °C for later analysis. The level of protein
expression and the purity was assessed by SDS-PAGE analysis.
Enzyme Assays. The activity of ECH was assayed as described by
Steinman and Hill,4 except for the omission of ovalbumin from the
reaction mixture and the use of 10 mM potassium phosphate buffer
(pH 7.0). Crotonyl-CoA (80 µM) was used as the substrate, and changes
in absorbance at 263 nm (ꢀ 6700 M-1‚cm-1), which corresponds to
the R,â-unsaturated moiety of the substrate, were monitored. A unit of
activity is defined as the amount of ECH required to catalyze the
hydration of 1 µmol of substrate per minute.
Kinetic Analysis of Inactivation. In a typical inactivation experi-
ment, an appropriate amount of MCPF-CoA was added to the bovine
liver ECH (9 µM) in 250 µL of 50 mM potassium phosphate buffer,
pH 7.6, at 25 °C. At various time intervals, aliquots (10 µL) were
removed and diluted into 1 mL of the same buffer containing crotonyl-
CoA (10 µM), and the residual enzyme activity was determined as
described above.
Determination of Partition Ratio. A series of sample solutions each
containing 9 µM bovine liver ECH and an appropriate amount of (R)-
or (S)-MCPF-CoA were prepared in 400 µL of 50 mM potassium
phosphate buffer, pH 7.6, to give [I]o/[E]o ratios ranging from 1.0 to
50. These samples were incubated at 25 °C for 2 h. Subsequently, an
aliquot of 10 µL was measured from each sample, and the residual
enzyme activity was determined as described above. The control
experiments were carried out in the same manner without inactivator.
The partition ratio was deduced by plotting the residual enzyme activity
versus [I]o/[E]o.
Isolation and Sequence Analysis of Labeled Peptides. Bovine liver
ECH (1.5 mg, 56 nmol) was incubated with 20 equiv of [3H]MCPF-
CoA (8; 0.56 mCi/mmol) in 50 mM potassium phosphate buffer (pH
7.6) at room temperature overnight. The inactivated enzyme was
concentrated by Microcon (Amicon) and repeatedly washed with the
same potassium phosphate buffer to remove the excess MCPF-CoA
until the radioactivity of the filtrates matched the background readings.
The radioactivity of the labeled protein was measured and its concentra-
tion determined. The labeled protein was then digested using trypsin
(60 µg, porcine pancreas type IX) under nitrogen overnight in the dark.
The tryptic fragments, after treatment with 1 mM DTT and filtered
through Microcon, were separated by HPLC on a Vydac C18 (4.6 mm
× 25 cm) column using a 0-60% (B/A) linear gradient formed from
0.1% (v/v) trifluoroacetic acid (TFA) in water (solvent A) and 0.1%
TFA in acetonitrile (solvent B) over a period of 2 h. The elution was
monitored at 220 nm, and the flow rate was 1 mL/min. The two
radioactive peaks (peaks A and B) were separately pooled and further
purified under the same HPLC conditions. The two radiolabeled tryptic
fragments were sequenced by automated Edman degradation. The
effluents obtained after each cycle were also counted for radioactivity.
N-[(R)-1′-Phenyl-2′-hydroxyethyl]-(1R)-methylenecyclopropylcar-
boxamide (12) and N-[(R)-1′-Phenyl-2′-hydroxyethyl]-(1S)-methyl-
enecyclopropylcarboxamide (13). The synthesis of (methylenecyclo-
propyl)formic acid (11) has been described previously.35 To a solution
of acid 11 (1.84 g, 18.7 mmol) and triethylamine (2.6 mL, 18.7 mmol)
in THF (70 mL) was added isobutyl chloroformate (2.43 mL, 18.7
mmol) at 0 °C. The solution was stirred at room temperature for 1 h,
which was followed by the addition of (R)-phenylglycinol (2.6 g, 18.7
mmol) in 10 mL THF.39,40 The resulting mixture was stirred for an
additional 3 h at room temperature and then filtered. The filtrate was
concentrated under reduced pressure to obtain the crude amides 12 and
(39) Helmchen, G.; Nill, G.; Flockerzi, D.; Schuhle, W.; Youssef, M. S.
K. Angew. Chem., Int. Ed. Engl. 1979, 18, 63-64.
(40) Baek, D.-J.; Daniels, S. B.; Reed, P. E.; Katzenellengbogen, J. A.
J. Org. Chem. 1989, 54, 3963-3972.

InactiVation of BoVine LiVer Enoyl-CoA Hydratase
J. Am. Chem. Soc., Vol. 123, No. 40, 2001 9753
1
overall yield was 87%. Similarly, compound 6 was prepared in a yield
of 84%. 1H NMR (2H2O) of 5 (the chemical shifts of the MCPF signals
are shown italicized): δ 8.58, 8.32 (1H each, s, adenine H), 6.20 (1H,
d, J ) 6.0, ribose anomeric H), 5.62 and 5.56 (1H each, d, J ) 2.2,
dCH2), 4.93-4.78 (1H, buried under 2HOH peak), 4.71 and 4.62 (1H
each, s, ribose H), 4.27 (2H, s, ribose CH2O), 4.04 (2H, s), 3.83 (1H,
m), 3.60 (1H, m), 3.47 (2H, t, J ) 10.8), 3.38 (2H, t, J ) 10.8), 3.02
(2H, m), 2.79 (1H, t, J ) 6.5, H-1), 2.44 (2H, t, J ) 10.0), 1.89 (2H,
m, H-3), 0.91, 0.78 (3H each, s, Me); the measured sample was
repeatedly dissolved in 2H2O and lyophilized prior to 1H NMR analysis.
High resolution FAB-MS calcd for C26H40N7O17P3S (M + 1)+,
848.1492; found, 848.1431. The spectra of 5 and 6 are superimposable.
desired acid 20 in 67% yield (44 mg). H NMR (CDCl3) δ 5.56 (2H,
m, CH2dC), 2.26 (1H, m, 1-H), 1.88 (1H, ddddd, J ) 166.2, 9, 4.5,
2.4, and 2.4, H-3), 1.71 (1H, dtt, J ) 166.2, 9, and 2.4, H-3); 13C NMR
(CDCl3) δ 12.27 (C-3).
[3-13C1](Methylenecyclopropyl)formyl-CoA (7). Compound 7 was
prepared from 20 (6.0 mg, 0.06 mmol) in 62% yield (32.0 mg) by
coupling with coenzyme A (50 mg, 65.1 mmol) according to the same
1
procedure used to synthesize 5 as described above. H NMR (D2O) δ
8.38, 8.09 (1H each, s, adenine H), 6.00 (1H, d, J ) 6.5, ribose anomeric
H), 5.37 (2H, m, CH2dC), 4.44 (1H, s, HOCHCMe2), 4.09 (2H, br s,
C(Me)2CH2O), 3.87 (1H, s, HOCHCMe2), 3.67 (1H, dd, J ) 10 and 5,
ribose), 3.40 (1H, dd, J ) 10 and 5, ribose), 3.28 (2H, t, J ) 6.5), 3.16
(2H, m), 2.85 (3H, m, overlap CHCO2H, CoA), 2.44 (1H, dm, JH-C
)
(Z)-3-Iodo-2-butenol (17). To a solution of 2-butyn-1-ol (16, 1.0
g, 14.3 mmol) in 15 mL of anhydrous ether was added dropwise a
solution of Red-Al (7.1 mL, 22.8 mmol) at -78 °C under argon. The
reaction was allowed to warm to room temperature, and stirring was
continued overnight. Subsequently, the mixture was cooled to -78 °C
and quenched with I2 (11.2 g, 44.3 mmol). The resulting solution was
gradually warmed to room temperature with continuous stirring within
30 min. It was concentrated under reduced pressure and chromato-
graphed directly on a silica gel column (1:1 ether/pentane) to afford
167.5, H-3), 2.25 (2H, m), 1.71 (1H, dm, JH-C ) 167.5, H-3), 0.73,
0.60 (3H each, s, Me); 13C NMR (D2O) δ 14.14 (13CH2). FAB-MS
calcd for 13C1C25H40N7O16P3S (M + H)+, 849.2; found, 849.2.
Two-Dimensional H{13C} gHMQC Analysis. Bovine liver ECH
1
(11.2 mg, 0.4 µmol) was incubated with 4 molar equiv of 13C-labeled
MCPF-CoA (7) in 50 mM potassium phosphate buffer (pH 7.6) at room
temperature. The reaction was incubated for 6 h to ensure complete
inactivation and then ultrafiltered through an Amicon YM-10 filter to
remove unreacted MCPF-CoA and any possible turnover products. The
resulting mixture was concentrated to 500 µL, to which 50 µL of 2H2O
was added, transferred to an NMR tube, and subjected to gHMQC
(gradient heteronuclear multiple quantum coherence) analysis on a
Varian U-500 spectrometer (number of transients ) 256; temperature
) 23 °C; time ∼ 12 h).
1
17, as a colorless oil, in 73% yield (3.3 g). H NMR (CDCl3) δ 5.77
(1H, tq, J ) 6.0 and 1.2, H-2), 4.15 (2H, dq, J ) 6.0 and 1.2, H-1),
2.54 (3H, q, J ) 1.2, Me).
[3-13C1](2-Iodo-2-methylcyclopropyl)methanol (18). To a solution
of 17 (0.25 g, 1.27 mmol) and [13C]CH2I2 (0.34 g, 1.27 mmol) in 10
mL of methylene chloride was added dropwise a 2 M solution of
trimethylaluminum in hexanes (1.34 mL, 2.68 mmol) over 15 min at
0 °C under nitrogen. The resultant mixture was stirred at room
temperature for 24 h. The reaction was then chilled to 0 °C and
quenched with 5 mL of 1N NaOH. After being stirred for an additional
hour, the organic layer was separated, and the aqueous layer was washed
with 10 mL of methylene chloride twice. The organic extracts were
combined, washed with water, dried, filtered, and concentrated under
reduced pressure. The crude product was purified by flash column
chromatography (1:4 ether/pentane) to give 18 in 74% yield (0.2 g).
Because this product is a mixture of a few diastereomers, multiple sets
of signals made the spectrum too complicated to be analyzed. 1H NMR
(CDCl3) δ 3.97 (dd, J ) 12.2, 4.7), 3.73 (t, J ) 4.7), 3.48 (m, CH2OH,
diastereomeric mixture), 1.96 (s), 1.94 (s, Me, diastereomeric mixture),
1.67 (1H, br s, CH2OH), 1.02 (1H, dt, JHa-C ) 59.3, JHa-C-H ) JHa-CC-H
) 6.9, CH2), 0.95 (1H, dm, JHb-C ) 161.5, CH2), 0.55 (1H, m, H-1);
13C NMR (CDCl3) δ 22.4 (C-3).
Results
Preparation of (R)- and (S)- MCPF-CoA and Inactivation
of Bovine Liver ECH. As mentioned above, MCPF-CoA (4)
is a specific inactivator against mitochondrial short chain ECHs,
such as those from bovine liver and pig kidney.34,35 Recent
experiments have shown that the toxicity of MCPF-CoA toward
these enzymes results from covalent trapping of an enzyme
active site nucleophile by this methylenecyclopropyl derivative.
Furthermore, the inhibitory effect of MCPF-CoA is more
pronounced against the bovine liver ECH as compared to the
pig kidney enzyme.35 Surprisingly, studies with rat liver ECH
revealed that MCPF-CoA is a competitive inhibitor of this
enzyme with a Ki of 30 µM and that the loss of activity is
reversible in nature.35 Evidently, despite the fact that a high
degree of sequence homology (g80%)41,42 has been noted for
all ECHs regardless of their origins,8 subtle structural differences
must exist between bovine liver and rat liver ECH as reflected
by the disparity of the modes of inhibition caused by MCPF-
CoA. This result not only highlighted the importance of effective
binding for the action of inhibition but also prompted us to
consider that the two stereoisomers of MCPF-CoA (5 and 6)
may experience deviate chiral discrimination imposed by the
bovine liver enzyme and may thus exhibit distinct inhibitory
activities. Considering the significant role of ECHs in fatty acid
metabolism and amino acid catabolism, selective inhibitors for
different ECHs may emerge as a useful means to control and/
or regulate these physiological processes for therapeutic inter-
vention. With this hope in mind, we decided to examine the
inhibitory effects of (R)- and (S)-MCPF-CoA on the activity of
bovine liver ECH.
[3-13C1](Methylenecyclopropyl)methanol (19). To a solution of
potassium t-butoxide (0.2 g, 1.69 mmol) in 5 mL of anhydrous dimethyl
sulfoxide (5 mL) was added neat 18 over a period of 30 min at 70 °C.
The resultant mixture was vigorously stirred for 7 h, and subsequently,
it was cooled to room temperature and poured over ice. The crude
product was extracted in ethyl ether (3 × 25 mL), and the combined
organic layers were washed with water, dried, filtered, and carefully
concentrated in vacuo. The crude product was purified by flash column
chromatography (3:1 pentane/ethyl ether) to give the desired product,
19, in 72% yield (56 mg). 1H NMR (CDCl3) δ 5.48 (1H, ddd, J ) 9.6,
2.2, and 0.9, CH2dC), 5.44 (1H, q, J ) 2.1, CH2dC), 3.65, 3.60 (1H
each, t, J ) 5.4, 5.7, CH2OH), 3.51 (2H, m, CH2OH, diastereomeric
mixture), 1.81 (1H, m, CHCH2OH), 1.41, 1.21 (1H each, dm, JH-C
)
102, CH2); 13C NMR (CDCl3) δ 7.2 (C-3). High resolution FAB-MS
calcd for 13C1C4H9O (M + 1)+, 86.0653; found, 86.0668.
[3-13C1](Methylenecyclopropyl)formic Acid (20). Compound 19
(56 mg, 0.66 mmol) was dissolved in acetone (10 mL) and treated
with Jones reagent that was prepared by mixing chromium oxide (26.72
g) with concentrated sulfuric acid (23 mL), followed by water dilution
to a final volume of 100 mL. Addition of Jones reagent was continued
until the red color persisted for at least 1 min. The resulting mixture
was stirred at room temperature for 30 min to ensure the completion
of oxidation. The excess oxidizing reagent was quenched with 2-pro-
panol. The reaction solution was then diluted with water, followed by
repeated extraction with ether. The combined organic extracts were
dried, filtered, and concentrated in vacuo. The crude product was
purified by flash chromatography (10% ethyl acetate/hexanes) to give
Preparation of pure diastereoisomers of MCPF-CoA (5 and
6) was accomplished according to a procedure developed earlier
to synthesize (R)- and (S)-(methylenecyclopropyl)acetyl-CoA.43
As shown in Scheme 1, the synthesis is initiated by a rhodium
(41) Minami-Ishii, N.; Taketani, S.; Osumi, T.; Hashimoto, T. Eur. J.
Biochem. 1989, 185, 73-78.
(42) Kanazawa, M.; Ohtake, A.; Abe, H.; Yamamoto, S.; Satoh, Y.;
Takayanagi, M.; Niimi, H.; Mori, M.; Hashimoto, T. Enzyme Protein 1993,
47, 9-13.

9754 J. Am. Chem. Soc., Vol. 123, No. 40, 2001
Dakoji et al.
Scheme 1
acetate catalyzed cyclopropanation of 2-bromopropene (9) with
ethyl diazoacetate. Derivatization of 11 with (R)-2-phenylgly-
cinol led to a diastereomeric mixture (12 and 13), which could
be readily separated by flash chromatography.39,40 The resolved
amide was converted to (methylenecyclopropyl)formic acid (14
and 15) by acid hydrolysis. Subsequent coupling to coenzyme
A43 afforded the desired (1R)- and (1S)-MCPF-CoA (5 and 6,
respectively).
The effect of the MCPF-CoA isomers on the catalytic activity
of ECH was analyzed by successive titration of bovine liver
ECH with aliquots of each isomer. A plot of the observed
residual activity versus the total equivalents of inhibitor added
gave the same partition ratio of 10 for both (1R)- and (1S)-
MCPF-CoA (Figure 1). The time-dependent loss of enzyme
activity follows Michaelis-Menten saturation kinetics. As
illustrated in Figure 2, a double reciprocal plot of the observed
first-order rate (kobs) versus inhibitor concentration led to the
inactivation rate constant kinact of 3.36 × 10-3 min-1 and the
apparent dissociation constant KI of 49.2 µM for the (1R)-isomer
(5). The values of kinact and KI for the (1S)-isomer (6) were
similarly determined to be 2.65 × 10-3 min-1 and 57.1 µM,
respectively. The nearly identical kinetic parameters and parti-
tion ratios strongly suggest that (1R)- and (1S)-MCPF-CoA are
equally competent inactivators against bovine liver ECH.
Therefore, the inactivation of bovine liver ECH by MCPF-CoA
is nonstereospecific.
Identification of Modified Residue in Bovine Liver ECH.
Previous study using 3H-labeled MCPF-CoA had fully substan-
tiated the covalent nature of the inactivation of bovine liver
ECH.34 To further characterize the mechanism of this process,
we proceeded to determine the identity of the amino acid residue
entrapped by the methylenecyclopropyl moiety. As described
in the Experimental Section, incubation of bovine liver ECH
with [3H]MCPF-CoA (8) yielded an inactive protein containing
nearly 1 equiv of [3H]MCPF-CoA per monomer, which is
consistent with the labeling of one amino acid residue in the
active site of the enzyme. Tryptic digestion of the labeled protein
gave a mixture of peptides that was fractionated by reversed
phase HPLC (Figure 3). In principle, only one tryptic fragment
should be labeled; however, radioactivity was found in two
separate fractions. The two radioactive peptides were isolated
Figure 1. Effect of (R)- and (S)-MCPF-CoA on the catalytic activity
of bovine liver ECH. These figures show the percentage of residual
activity versus the ratio of MCPF-CoA to ECH.
and sequenced by standard Edman degradation. Interestingly,
the first eight amino acid residues were identical for both
peptides and had the sequence ALGGGXEL, indicating that they
were formed from incomplete proteolysis catalyzed by a
contaminating protease, possibly chymotrypsin. It should be
noted that a majority of the radioactivity was released during
one cycle of the Edman degradation of each fragment and was
found to be associated with the sixth amino acid (Figure 4),
which was designated X. However, the lack of nucleotide and/
or peptide sequence information on bovine liver ECH precluded
the determination of the identity of X. Nevertheless, the deduced
peptide sequence aligned perfectly with a well conserved region
(43) Lai, M.-t.; Liu, L.-d.; Liu, H.-w. J. Am. Chem. Soc. 1991, 113,
7388-7397.

InactiVation of BoVine LiVer Enoyl-CoA Hydratase
J. Am. Chem. Soc., Vol. 123, No. 40, 2001 9755
Because X should be nucleophilic, cysteine is a prime candidate
for the entrapped active site residue.
Isolation and Nucleotide Sequence Analysis of cDNA
Clones of Bovine Liver ECH. To gain more information about
the identity of X, we proceeded to isolate the cDNA of bovine
liver ECH. A bovine liver Lambda cDNA library (Stratagene)
was screened by polyclonal antibodies raised against this enzyme
using the dot blot test as described in the Experimental Section.
Two positive clones were identified after screening ∼1 × 106
clones from the bovine liver library. These positive clones were
subjected to in vivo excision in the host E. coli SOLR strain,
and the resulting constructs were labeled as pBluescript SK-
(()-SDa1 and pBluescript SK(()-SDb1. Both clones carried a
cDNA fragment of the same size (∼1.6 kb), and sequencing
both ends of the cloned DNA revealed that these two clones
are identical. Both strands of the coding region were sequenced
by the gene walking technique with custom designed primers.
The resulting DNA sequence along with the translated amino
acid sequence for bovine liver ECH is shown in Figure 6
(GenBank accession number AY049961).
The translated sequence of this enzyme is highly homologous
to the mammalian mitochondrial ECHs. The overall identity of
bovine liver ECH to rat liver and human liver ECHs is 87 and
85%, respectively, at the amino acid sequence level. As shown
in Figure 6, it is clear that the isolated bovine liver ECH gene
is truncated at the 5′-end, missing the signal sequence and the
start codon.44 Fortunately, the entire coding sequence for the
mature protein remains intact in the cloned constructs (pBlue-
script SK(()-SDa1 and SK(()-SDb1), as indicated by the
presence of the N-terminal amino acid sequence which was
deduced by N-terminal sequencing a commercial sample of
bovine liver ECH. Most importantly, sequence alignment of the
tryptic digested labeled peptide fragment with the wild-type
bovine liver ECH has allowed the unambiguous identification
of residue X as Cys114 (mature sequence numbering).
Figure 2. Effect of (R)- and (S)-MCPF-CoA on the catalytic activity
of bovine liver ECH. Panels A and B are plots of kobs as a function of
(R)- and (S)-MCPF-CoA concentration, respectively. The inset shows
the double reciprocal plot of kobs versus MCPF-CoA concentration.
Purification of Bovine Liver ECH from Esherichia coli
JM105/pSD6. To ensure that the cloned sequence mentioned
above indeed corresponds to bovine liver ECH, the cDNA
encoding the mature protein (261 amino acids) in pBluescript
SK(()-SDa1 was amplified by PCR and subcloned into a
pTrc99A expression vector. A start codon was engineered
flanking the 5′-end of the mature sequence of the protein.
Expression in Escherichia coli JM105 of the resulting construct,
pSD6, was induced by IPTG, and a significant portion of the
in the catalytic core of the mitochondrial short-chain ECHs from
rat liver,41 human liver,42 and many others from different species
(Figure 5). For example, this fragment corresponds to residues
138-145 in the rat liver enzyme, which includes Gly141 as
well as Glu144 (rat liver ECH numbering), both of which have
been shown to be important for catalysis. Interestingly, the
alignment reveals that the counterpart of X in these sequences
could be either cysteine, leucine, phenylalanine, or methionine.
Figure 3. Fractionation of the tryptic digest of [3H]MCPF-CoA (8) labeled bovine liver ECH by HPLC. See Experimental Section for details.

9756 J. Am. Chem. Soc., Vol. 123, No. 40, 2001
Dakoji et al.
Scheme 2
by pBluescript SK(()-SDa1 is indeed the desired bovine liver
ECH gene.
Characterization of the Covalent Adduct Between Bovine
Liver ECH and MCPF-CoA. To perform structural charac-
terization of the inhibitor-enzyme adduct to glean more insights
into the inactivation mechanism, we decided to chemically
synthesize [3-13C1]-labeled MCPF-CoA (7). As delineated in
Scheme 2, reduction of 2-butynol (16) with Red-Al followed
by quenching with I2 gave 3-iodo-2-butenol (17). Cyclopropa-
nation of the resulting crotyl alcohol (Al(Me)2/13CH2I2) intro-
duced the stable isotope at C-3 of the ring carbon to give a
racemic alcohol 18 in 54% overall yield.45 Alcohol 18 was
subjected to dehydroiodination, followed by Jones oxidation of
the resulting (methylenecyclopropyl)methyl alcohol (19), to give
(methylenecyclopropyl)formic acid (20). The corresponding
CoA thioester was prepared by the mixed anhydride method
and purified by HPLC to afford final product 7 in 62% yield.
The labeled carbon (C-3) in 7 has a chemical shift of 14.4 ppm.
Four equivalents of 7 was incubated with bovine liver ECH
(0.4 µmol) in 50 mM potassium phosphate buffer (pH 7.6) at
25 °C, and the progress of the inactivation was monitored using
the standard assay protocol. After 6 h, at which time further
inactivation was insignificant, the incubation mixture was
concentrated to 500 µL by ultrafiltration and then subjected to
the gHMQC (gradient heteronuclear multiple quantum coher-
ence) experiment.46 As illustrated in Figure 7, the gHMQC
spectrum showed one pertinent signal at 105 ppm in the 13C
dimension and the corresponding cross-peaks at δ 4.8 and 5.0
Figure 4. Determination of amino acid sequence for peptides A and
B by automatated Edman degradation. After each cycle, the effluents
were collected, and an aliquot of each was counted for radioactivity.
Profiles A and B show the data obtained for peptides A and B,
respectively. X represents the modified amino acid residue.
1
in the H dimension. The carbon and proton chemical shifts
Figure 5. Amino acid sequence alignment of representative ECHs from
different sources with the labeled peptide derived from bovine liver
ECH.
are characteristic for a terminal olefinic methylene group. Thus,
the gHMQC results are most consistent with a nucleophilic
attack on C-2′, followed by ring fragmentation to yield a
thioether adduct (21) (Scheme 4, route A).
protein was produced in soluble form. The expressed protein
was purified to near homogeneity after DEAE-Sepharose and
Phenyl-Sepharaose column chromatography. The recombinant
protein has a molecular mass of 28 kDa per monomer and
resembles that of the wild-type protein isolated from bovine
liver.4 Its catalytic activity is also similar to the wild-type
enzyme, with a kcat of 2.06 × 103 s-1 and a Km of 15.6 ( 1.3
µM. In addition, the N-terminal sequence of the recombinant
protein was determined to be identical to that of the wild-type
enzyme. Thus, the above results confirmed that the cDNA borne
Discussion
Cyclopropyl compounds have proven to be valuable mecha-
nistic probes in the study of enzyme-catalyzed reactions,
especially where free radical chemistry is involved. Our interest
in this class of compounds as inhibitors for fatty-acid degrada-
tion enzymes evolved from a study of the toxicity of hypoglycin
A (22), which is the causative agent of Jamaican Vomitting
Sickness.47-49 This compound is metabolized in vivo to (me-
thylenecyclopropyl)acetyl-CoA (MCPA-CoA, 23), which in-
(44) Ozasa, H.; Furuta, S.; Miyazawa, S.; Osumi, T.; Hashimoto, T.;
Mori, M.; Miura, S.; Tatibana, M. Eur. J. Biochem. 1984, 144, 453-458.
(45) Russo, J. M.; Price, W. A. J. Org. Chem. 1993, 58, 3589-3590.
(46) Martin, G. E.; Zektzer, A. S. In Two-Dimensional NMR Methods
for Establishing Molecular ConnectiVity; VCH Publishers: New York, 1988.
(47) Tanaka, K. In Handbook of Clinical Neurology; Vinken, P. J., Bruyn,
G. W., Eds.; Elsevier/North Holland: Amsterdam, 1979; Vol. 37, Chapter
17, 511-539.
(48) Sherratt, H. S. A. Trends Pharmacol. Sci. 1986, 7, 186-191.

InactiVation of BoVine LiVer Enoyl-CoA Hydratase
J. Am. Chem. Soc., Vol. 123, No. 40, 2001 9757
Figure 6. Nucleotide and the translated amino acid sequence of bovine liver and rat liver ECHs (GenBank accession number AY049961).
Scheme 3
taneous.53-57 Thus, it is not surprising that both C-1 epimers of
MCPA-CoA are effective inhibitors for acyl-CoA dehydroge-
nases, because the bond cleavage at â-C of MCPA-CoA is not
enzyme-catalyzed. Interestingly, MCPF-CoA (4), a lower ho-
mologue of MCPA-CoA, does not exhibit inhibitory activity
against acyl-CoA dehydrogenases, but specifically inhibits
bovine liver ECH in a time-dependent, irreversible fashion.34,35
The inactivation of bovine liver ECH occurs at the active site
of the enzyme and involves covalent modification by 1 equiv
of MCPF-CoA.34
Figure 7. 13C-1H-detected gHMQC spectrum (500 MHz, 2H2O)
showing the pertinent resonances of the covalent adduct between bovine
liver ECH and [3-13C1](methylenecyclo-propyl)formyl-CoA (7).
The crystal structures of rat liver ECH and several other
proteins that bind coenzyme A thioesters have been solved.58
The predominant theme in these enzymes is that the coenzyme
A moiety serves as the primary binding determinant. Therefore,
with the coenzyme A scaffold locked into the active site of
bovine liver ECH, it is reasonable to assume that the cyclopropyl
ring, where the inactivation occurs, would acquire a unique
spatial orientation within the active site dictated by the stereo-
chemistry at C-1. Therefore, one isomer of MCPF-CoA is
hibits both short-chain and medium-chain acyl-CoA dehydro-
genases.50-52 The inactivation is initiated by the cleavage of
the cyclopropyl ring followed by covalent modification of the
active site bound FAD coenzyme (Scheme 3). The ring-opening
step is likely induced by a cyclopropylcarbinyl radical (24), and
the rapid ring opening renders this process practically spon-
(49) Tanaka, K.; Ikeda, Y. In Fatty Acid Oxidation: Clinical, Biochemical
and Molecular Aspects; Tanaka, K., Coates, P. M., Eds.; Alan R. Liss,
Inc.: New York, 1990, 167-172.
(50) Ghisla, S.; Wenz, A.; Thorpe, C. In Enzyme Inhibitors; Brodbeck,
U., Ed.; Verlag Chim: Weinheim, Germany, 1980; p 43-60.
(51) Wenz, A.; Thorpe, C.; Ghisla, S. J. Biol. Chem. 1981, 256, 9809-
9812.
(52) Ghisla; S., Melde, K.; Zeller, H. D.; Boschert, W. In Fatty Acid
Oxidation; Clinical, Biochemical, and Molecular Aspects; Tanaka, K.,
Coates, P. M., Eds.; Alan R. Liss, Inc.: New York, 1990; p 185-192.
(53) Lai, M.-t.; Liu, H.-w. J. Am. Chem. Soc. 1990, 112, 4034-4035.
(54) Lai, M.-t.; Liu, L.-d.; Liu, H.-w. J. Am. Chem. Soc. 1991, 113,
7388-7397.
(55) Lai, M.-t.; Liu, H.-w. J. Am. Chem. Soc. 1992, 114, 3160-3162.
(56) Lai, M.-t.; Li, D.; Oh, E.; Liu, H.-w. J. Am. Chem. Soc. 1993, 115,
1619-1628.
(57) Li, D.; Zhou, H.-q.; Dakoji, S.; Shin, I.; Oh, E.; Liu, H.-w. J. Am.
Chem. Soc. 1998, 120, 2008-2017.
(58) Engel, C.; Wierenga, R. Curr. Opin. Struct. Biol. 1996, 6, 790-
797.

9758 J. Am. Chem. Soc., Vol. 123, No. 40, 2001
Dakoji et al.
Scheme 4
expected to exert a greater inhibitory potency toward the enzyme
than the other and may in fact be the sole contributor to the
inactivation of the enzyme. To investigate the stereochemical
preference of this inactivation, diastereomerically pure (R)- and
(S)-MCPF-CoA were prepared. To our surprise, both (R)- and
(S)-MCPF-CoA display similar kinetic behavior and therefore
are equally competent in causing the inactivation of bovine liver
ECH. As mentioned earlier, the lack of stereospecificity
observed in the MCPA-CoA-mediated inactivation of acyl-CoA
dehydrogenases could be readily explained by invoking the
presence of a radical intermediate in the reaction mechanism.
However, ECHs catalyze a stereospecific hydration reaction
without the assistance of any cofactors. Therefore, the basis for
the lack of stereospecificity in the inactivation of bovine liver
ECH by MCPF-CoA is not immediately obvious, although it is
possible that a rate-determining step independent of the bound
inactivator, such as the ionization of an active site residue, may
be involved in the inactivation process.
Intrigued by the apparent lack of discrimination displayed
by bovine liver ECH toward the stereoisomers of MCPF-CoA,
experiments were carried out to determine which residue(s) at
the enzyme active site is (are) modified during the inactivation.
Accordingly, bovine liver ECH was incubated with excess
tritiated racemic MCPF-CoA (8), and the resulting inactivated
enzyme was subjected to proteolytic digestion. Upon separation
of the peptide fragments by reverse-phase HPLC, only two
radiolabeled peptides were detected. Although the two peptides
were not classical tryptic peptides, both had an identical
N-terminal sequence, ALGGGXEL, and a significant amount
of radiolabel was associated with single residue “X”. These
results indicate that covalent modification of bovine liver ECH
by both diastereomers of MCPF-CoA occurs at a single site
and are consistent with the observation that this inactivation is
nonstereospecific. Although the cDNA sequence of bovine liver
ECH was not available at the time, sequence alignment of
crotonases isolated from different sources showed a high degree
of homology, and the deduced short peptide sequence aligned
perfectly with a well conserved segment which corresponds to
residues 138-145 of the rat liver ECH (Figure 5). Therefore,
our initial speculation was that a cysteine residue may be
involved in the capture of MCPF-CoA by bovine liver ECH.
Interestingly, the X-ray structure of the rat liver ECH indicates
the importance of this cysteine residue in the formation of a
putative H-bond with Gln161 (rat liver ECH numbering), an
essential element for the proper positioning of the catalytic
residues.20,21
It should be noted that bovine liver ECH contains five
cysteines per subunit and none are implicated in disulfide bond
formation.4,59 Earlier reports had shown that thiol-directing
reagents such as p-chloromercuribenzoate and iodoacetate
inactivate the bovine liver ECH, and it was suggested on this
basis that a thiol group may be catalytically essential for this
enzyme. However, later studies indicated that none of the five
cysteinyl residues are crucial for catalytic activity.3 Interestingly,
an affinity reagent, p-bromoacetamido-trans-cinnamoyl-CoA
(BCC), inactivates the bovine liver ECH by steric blockage of
the substrate-binding site and by alkylation of a proximal
cysteine residue.3 It is conceivable that MCPF-CoA, like BCC,
exhibits an analogous mode of binding and that the same
cysteinyl residue is the target of covalent modification. Although
these results are intriguing, further characterization of the
identity of the putative cysteine residue toward the inactivation
was prudent in order to reach definitive conclusions. Thus,
(59) Hass, G. M.; Hill, R. L. J. Biol. Chem. 1969, 244, 6080-6086.

InactiVation of BoVine LiVer Enoyl-CoA Hydratase
J. Am. Chem. Soc., Vol. 123, No. 40, 2001 9759
cloning and sequencing of the gene encoding bovine liver ECH
were performed.
that in case of C-3 attack, the existing olefinic bond is retained
between C-2 and C-2′, while C-3 and C-1 are transformed into
allylic carbons. On the contrary, for the mechanism involving
a C-2′ attack, an olefinic bond is formed between C-2 and C-3,
while C-2′ and C-1 are transformed into allylic carbons. To
distinguish between these two pathways, [3-13C]MCPF-CoA (7)
was synthesized as a probe to determine the fate of C-3 of
MCPA-CoA after the inactivation had occurred. As described
previously, the NMR data of the inactivated sample are most
consistent with the assignment of the 13C label as an alkene
carbon in the adduct. Therefore, the molecular basis of the
inactivation likely involves the nucleophilic attack of Cys-114
at the C-2′ olefinic carbon of MCPF-CoA, leading to ring
scission and the permanent modification of the enzyme.
Several cyclopropane-containing mechanism-based inactiva-
tors in which the target enzyme activates the cyclopropane for
nucleophilic addition by oxidation or protonation of the ap-
pended groups rendering them more electron withdrawing have
been reported.60,61 However, no such catalytic activation of
MCPF-CoA appears to be necessary as the initial step for
inactivation. Although the regioselectivity and mechanism of
the methylenecyclopropyl ring fragmentation is poorly defined,
the main driving force must arise from the relief of the great
ring strain (40.8 kcal/mol).62 Substrate activation via electron
polarization of the π-bond to enhance the electrophilicity of
â-C is a prerequisite of catalysis for ECH.15-18 Given the
preponderance of the ECH active site toward π-bond electron
reorganization, it is conceivable that MCPF-CoA is activated
by a similar push-pull mechanism, thus facilitating the entrap-
ment of the active site nucleophile. Hence, the inactivation
chemistry appears to be dictated by a normal turnover mech-
anism with the help of proper contacts and binding modes within
the enzyme active sites. Considering the role of ECH in the
metabolism of fatty acids and the catabolism of branched-chain
amino acids, its inhibitors may emerge as a useful means to
control and/or regulate these physiological processes for thera-
peutic intervention. Thus, the results described herein may
provide important clues for future design of more efficient and
selective inhibitors to control and/or regulate fatty acid me-
tabolism.
On the basis of the sequence of the cDNA clone for bovine
liver ECH, we were able to establish that the amino acid
corresponding to “X” labeled with MCPF-CoA is indeed a
cysteine residue that is located at position 114 in the mature
protein. It is clear that bovine liver ECH and its rat liver
counterpart share a very high degree of sequence identity (Figure
6). In fact, the major differences in the amino acid sequence
are primarily limited to the N- and C-terminal domains. The
core catalytic domains of the two enzymes are virtually identical,
with only a few amino acid substitutions, all of which are
conservative in nature. It is reasonable to assume that bovine
liver ECH and rat liver ECH have similar structures, catalytic
properties, and kinetic parameters. However, MCPF-CoA acts
as a time-dependent irreversible inhibitor for the bovine liver
enzyme and as a competitive, reversible inhibitor for the rat
liver enzyme. The high level of sequence identity of the two
enzymes makes these differences in inhibition patterns all the
more intriguing. However, in the absence of structural informa-
tion about the bovine liver enzyme, we can only speculate that
there may be subtle differences in the active site architecture
of the two enzymes that govern the interaction of each with
MCPF-CoA.
As far as the chemical basis of inactivation is concerned, three
possible mechanisms of inactivation of bovine liver ECH by
MCPF-CoA had been proposed (Scheme 4). However, two of
them were considered unlikely on the basis of our early
experiments.34 For instance, a transesterification mechanism
involving the formation of an acyl-enzyme adduct and the
liberation of coenzyme A (Scheme 4, route B) was ruled out
on the basis of the resistance of the enzyme-inhibitor adduct
toward alkaline hydrolysis and on the association of the adenine
chromophore with the inactivated enzyme despite prolonged
dialysis. Similarly, a mechanism involving C-3 deprotonation
followed by anion-induced ring opening (Scheme 4, route C)
was rejected on the basis of the lack of tritium washout from
[3-3H]MCPF-CoA (8) when it was incubated with the enzyme.
Therefore, a nucleophile-induced ring-opening mechanism ap-
pears to be the only viable alternative to explain the irreversible
inactivation of the enzyme by MCPF-CoA.
Interestingly, such a mechanism could conceivably operate
via two different pathways. As shown in Scheme 4, the
nucleophilic attack could occur at C-3, leading to the cleavage
of the C1-C3 bond to form the enzyme-inhibitor covalent
adduct. Alternatively, the nucleophilic attack could occur at the
olefinic C-2′ carbon, resulting in the formation of a C2-C3
double bond and the cleavage of the C1-C3 bond. Both
inactivation routes result in an identical structure of the ultimate
covalent adduct, irrespective of whether the initial nucleophilic
attack occurred at C-3 or C-2′. However, it should be noted
Acknowledgment. This work was supported in part by
National Institutes of Health Grant GM40541.
JA011226K
(60) Suckling, C. J. Angew. Chem., Int. Ed. Engl. 1988, 27, 537-552.
(61) Liu, H.-w.; Walsh, C. T. In The Chemistry of the Cyclopropyl Group;
Pappoport, Z., Ed.; Wiley-Interscience: New York, 1987; part 2, 969-
1025.
(62) Benson, S. W.; Cruickshank, F. R.; Golden, D. M.; Haugen, G. R.;
O’Neal, H. E.; Rodgers, A. S.; Shaw, R.; Walsh, R. Chem. ReV. 1969, 69,
279-324.