The importance of the amide bond nearest the thiol group in enzymatic reactions of coenzyme A

Article abstract of DOI:10.1016/j.bioorg.2004.10.002

Analogues of coenzyme A (CoA) and of CoA thioesters have been prepared in which the amide bond nearest the thiol group has been modified. An analogue of acetyl-CoA in which this amide bond is replaced with an ester linkage was a good substrate for the enzymes carnitine acetyltransferase, chloramphenicol acetyltransferase, and citrate synthase, with K_m values 2- to 8-fold higher than those of acetyl-CoA and V_max values from 14 to >80% those of the natural substrate. An analogue in which an extra methylene group was inserted between the amide bond and the thiol group showed less than 4-fold diminished binding to the three enzymes but exhibited less than 1% activity relative to acetyl-CoA with carnitine acetyltransferase and no measurable activity with the other two enzymes. Analogues of several CoA thioesters in which the amide bond was replaced with a hemithioacetal linkage exhibited no measurable activity with the appropriate enzymes. The results indicate that some aspects of the amide bond and proper distance between this amide and the thiol/thioester moiety are critical for activity of CoA ester-utilizing enzymes.

Full text of DOI:10.1016/j.bioorg.2004.10.002

Bioorganic Chemistry 33 (2005) 90–107
www.elsevier.com/locate/bioorg
The importance of the amide bond nearest
the thiol group in enzymatic reactions
of coenzyme A
Jin Xun, Haidong Huang, Kurt W. Vogel1,
Dale G. Drueckhammer*
Department of Chemistry, Stony Brook University, Stony Brook, NY 11794-3400, USA
Received 25 August 2004
Available online 25 November 2004
Abstract
Analogues of coenzyme A (CoA) and of CoA thioesters have been prepared in which the
amide bond nearest the thiol group has been modified. An analogue of acetyl-CoA in which
this amide bond is replaced with an ester linkage was a good substrate for the enzymes carni-
tine acetyltransferase, chloramphenicol acetyltransferase, and citrate synthase, with Km values
2- to 8-fold higher than those of acetyl-CoA and Vmax values from 14 to >80% those of the
natural substrate. An analogue in which an extra methylene group was inserted between the
amide bond and the thiol group showed less than 4-fold diminished binding to the three
enzymes but exhibited less than 1% activity relative to acetyl-CoA with carnitine acetyltrans-
ferase and no measurable activity with the other two enzymes. Analogues of several CoA thi-
oesters in which the amide bond was replaced with a hemithioacetal linkage exhibited no
measurable activity with the appropriate enzymes. The results indicate that some aspects of
the amide bond and proper distance between this amide and the thiol/thioester moiety are crit-
ical for activity of CoA ester-utilizing enzymes.
ꢀ 2004 Elsevier Inc. All rights reserved.
*
Corresponding author. Fax: +1 631 632 7960.
E-mail addresses: ddrueckhamme@notes.cc.sunysb.edu, dale.drueckhammer@sunysb.edu (D.G.
Drueckhammer).
1
Present address: PanVera Inc., Madison, WI 53719, USA.
0045-2068/$ - see front matter ꢀ 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.bioorg.2004.10.002

J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
Keywords: Coenzyme A; Cofactors; Enzyme catalysis
91
1. Introduction
A number of analogues of coenzyme A (CoA) 1 and of acetyl-CoA 2 have been
previously reported from this group (see Fig. 1) [1–8]. For example, the carboxylate
analogue 3, which was designed to mimic the enolate form of acetyl-CoA has been
shown to be a potent inhibitor of citrate synthase, with a Ki value 1000-fold lower
than the Km for acetyl-CoA [1,9]. In more recent work to begin to address the role
of functionality remote from the thiol or thioester moiety, the didemethyl analogue
of CoA 4 was prepared [10]. NMR analysis showed that the methyl groups have
only a moderate effect on the conformation of the part of the molecule in the
vicinity of the geminal methyl groups. The corresponding acetyl thioester 5 was
a fairly good substrate for acetyl-CoA utilizing enzymes with modestly increased
Km values and moderately decreased kcat values relative to the natural substrate
acetyl-CoA.
The work described in this manuscript addresses the importance of the amide
bond nearest the thiol group and the ethylene group linking this amide to the thiol
moiety. In previous work from this group, a modified form 6 of the carboxylate ana-
logue 3 was prepared in which the amide bond was replaced with a pair of methylene
Fig. 1. Structures of coenzyme A and of analogues and derivatives used in this work.

92
J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
groups [7]. This substitution was shown to diminish binding to citrate synthase more
than 10,000-fold. Reported here is an oxoester analogue of CoA 7, in which the
amide bond nearest the thiol group is replaced with an ester linkage, along with stud-
ies of its acetate thioester 8 as a substrate for acetyl-CoA utilizing enzymes. Studies
of the homocysteamine analogue 9 and the corresponding acetyl thioester 10, in
which an extra methylene group is present between the amide bond and thiol group
are also reported. Finally, this manuscript describes studies of CoA ester analogues
11 in which the amide bond is replaced with a reversibly formed hemithioacetal link-
age. The results indicate that the carbonyl group of the amide bond nearest the thiol
or thioester and proper distance between this amide and the thiol or thioester group
are critical for the activity of CoA-utilizing enzymes.
2. Materials and methods
2.1. General methods
Reagents and enzymes were obtained from Sigma–Aldrich and used as supplied.
Wild type [1,11] or recombinant [12,13] phosphopantetheine adenylyltransferase and
dephosphocoenzyme A kinase and recombinant pantothenate kinase [14] were pre-
pared as described previously. Analytical HPLC was performed using a Rainin
Microsorb C-18 column (4.6 · 250 mm, 5 lm) with monitoring at 215 and 260 nm
with a photodiode array detector and a flow rate of 1 mL/min. Solvent conditions
for analytical HPLC were 5% methanol in aq. monopotassium phosphate
(50 mM) for 2 min, followed by a linear gradient to 60% methanol over 12 min
and maintenance at 60% methanol for at least 5 min. Preparative HPLC was per-
formed on a Rainin Microsorb C-18 column (21.4 mm · 25 cm) with monitoring
at 215 and 280 nm and a flow rate of 10 mL/min with solvent gradients as described
for individual compounds. The concentrations of CoA analog solutions were deter-
mined using e260 = 15,400 Mꢀ1 cmꢀ1. Computational studies were conducted using
Gaussian03, with aqueous solvation described by TomasiÕs polarized continuum
model [15].
2.2. Oxoester CoA 7
To a solution of 12 [1] (20 mg, 26 lmol) in water (1 mL) was added b-mercap-
toethanol (140 lL, 2 mmol) and the pH was adjusted to 7.6 with aq. NaOH. After
75 min, analytical HPLC showed predominantly the thioester product at retention
time 12.7 min, with minor oxoester 7 at 14.2 min. The reaction mixture was diluted
to 50 mL with water, the pH was adjusted to 2.5, and the b-mercaptoethanol was
removed via ion exchange chromatography on DEAE–Sepharose (20 · 2.5 cm,
0–0.2 M NaCl, pH 2.8, product eluted at 0.15 M NaCl). The fractions containing
thioester and oxoester 7 were pooled and allowed to stand for 1 month at room
temperature. Analytical HPLC showed near complete conversion of the thioester
to 7. The product was lyophilized and purified by HPLC (5 min at 5% methanol

J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
93
in 10 mM potassium phosphate, pH 4.5, followed by a linear gradient to 35%
methanol over 45 min). The fractions eluted at 30–35 min were combined and lyoph-
ilized to give 7 (3.1 mg, 4.0 lmol, and 15% yield). kmax = 260 nm. Four hundred
1
megahertz NMR (D2O) d 0.68 (s, 3H), 0.80 (s, 3H), 2.55 (t, 2H, J = 6.4 Hz), 2.67
(t, 2H, J = 6.4 Hz), 3.04 (t, 2H, J = 5.4 Hz), 3.48 (dd, 1H, J = 4.8, 10.0 Hz), 3.75
(dd, 1H, J = 4.8, 10.0 Hz), 3.93 (s, 1H), 4.10 (t, 2H, J = 6.4 Hz), 4.16 (br s, 2H),
4.52 (s, 1H), 6.12 (d, 1H, J = 6.0 Hz), 8.27 (s, 1H), 8.52 (s, 1H). HRMS (FAB):
[M ꢀ H] calcd for C21H34N6O17P3S m/z 767.091. Found 767.095.
2.3. S-Acetyl-oxoester-CoA 8
To a solution of 7 (2.7 mg, 3.5 lmol) in Hepes buffer (1 mL, 25 mM, pH 8.2) was
added S-thiophenyl acetate (0.1 mL, 0.75 mmol) in acetonitrile (0.9 mL). The turbid
solution was capped and stirred vigorously at room temperature. The reaction was
monitored by HPLC (product tr = 14.8 min) and was complete in 1 h. The aqueous
layer was extracted with ether (5 · 4 mL), the pH was adjusted to 4.5 with aq. HCl,
and the product was purified by HPLC (10 min 5% methanol in 0.01 M sodium
phosphate, pH 4.5, followed by a linear gradient to 40% methanol over 40 min).
The product eluted at 39–46 min, the fractions containing product were combined
and lyophilized to give 8 (1.2 mg, 1.5 lmol, and 42% yield). kmax = 260 nm. Four
1
hundred megahertz NMR (D2O) d 0.70 (s, 3H), 0.83 (s, 3H), 2.29 (s, 3H), 2.53 (t,
2H, J = 6.4 Hz), 3.06 (t, 2H, J = 6.2 Hz), 3.35–3.45 (m, 2H), 3.46–3.52 (m, 1H),
3.75–3.82 (m, 1H), 3.96 (s, 1H), 4.13 (t, 2H, J = 6.2 Hz), 4.19 (br s, 2H), 4.55 (br
s, 1H), 6.14 (d, 1H, J = 6.4 Hz), 8.28 (s, 1H), 8.54 (s, 1H).
2.4. Homocysteamine CoA 9
A suspension of homocysteamine hydrochloride (4.4 g, 10 mmol) in water (5 mL)
was added to a solution of 12 (100 mg, 130 lmol) in water (5 mL) and the pH was
adjusted to 10 with 6 M aq. NaOH. The reaction was stirred at room temperature
under nitrogen and monitored by HPLC. Complete disappearance starting material
(tr = 17.3 min) and appearance of product (tr = 13.7 min) was observed after 4 h.
The reaction mixture was filtered, the pH was adjusted to 4.5 with aq. HCl, and
the product was partially purified by HPLC (5 min 5% methanol in 0.01 M potas-
sium phosphate, pH 4.5, followed by a linear gradient to 45% methanol over
40 min). The product, eluted at 25–32 min, was further purified by ion exchange
chromatography on DEAE–Sepharose (20 · 2.5 cm, 0–0.2 M NaCl, pH 2.8). The
product eluted at 0.13 M NaCl. Fractions containing product were combined and
lyophilized to give 9 (45 mg, 58 lmol, and 45% yield). kmax = 260 nm. Four hundred
megahertz 1NMR (D2O) d 0.72 (s, 3H), 0.85 (s, 3H), 1.65–1.72 (m, 2H), 2.39 (t, 2H,
J = 6.4 Hz), 2.45 (t, 2H, J = 7.0 Hz), 3.17 (t, 2H, J = 6.8 Hz), 3.41 (t, 2H,
J = 6.2 Hz), 3.52 (dd, 1H, J = 4.8, 10.0 Hz), 3.79 (dd, 1H, J = 4.8, 10.0 Hz), 3.96
(s, 1H), 4.20 (br s, 2H), 4.55 (br s, 1H), 6.16 (d, 1H, J = 6.0 Hz), 8.31 (s, 1H),
8.56 (s, 1H). HRMS (FAB): [M ꢀ H] calcd for C22H38N7O16P3S m/z 780.191. Found
780.126.

94
J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
2.5. 3-Acetyl-homocysteamine-CoA 10
A solution of 9 (40 mg, 51 lmol) in water (1.5 mL) adjusted to pH 8.2 with aq.
NaOH was added to a solution of S-phenyl thioacetate (0.2 mL, 1.5 mmol) in aceto-
nitrile (2 mL). The resulting biphasic mixture was stirred vigorously at room temper-
ature and the reaction was monitored by HPLC. The reaction was complete in 1 h, at
which time stirring was stopped and the acetonitrile layer was discarded. The aque-
ous layer was extracted with ether (5 · 4 mL), the pH was adjusted to 4.5 with aq.
HCl, and the product was purified by HPLC (5 min 5% methanol in 0.02 M sodium
phosphate, pH 4.5, followed by a linear gradient to 45% methanol over 45 min). The
product eluted at 30–38 min. Fractions containing product were combined and
lyophilized to give 10 (23 mg, 28 lmol, and 55% yield). kmax = 260 nm. Four hun-
1
dred megahertz NMR (D2O) d 0.72 (s, 3H), 0.85 (s, 3H), 1.63–1.70 (m, 2H), 2.29
(s, 3H), 2.41 (t, 2H, J = 6.4 Hz), 2.78 (t, 2H, J = 7.2 Hz), 3.13 (t, 2H, J = 6.8 Hz),
3.43 (t, 2H, J = 6.0 Hz), 3.53 (dd, 1H, J = 4.8, 10.0 Hz), 3.80 (dd, 1H, J = 4.8,
10.0 Hz), 3.98 (s, 1H), 4.22 (br s, 2H), 4.56 (br s, 1H), 6.15 (d, 1H, J = 6.0 Hz),
8.26 (s, 1H), 8.54 (s, 1H). HRMS (FAB): [M ꢀ H] calcd for C24H39N7O17P3S m/z
822.134. Found 822.138.
2.6. 3-Aminopropionaldehyde diethyl acetal 16
A solution of 15 (9.96 g, 60 mmol) and sodium azide (4.9 g, 75 mmol) in dimethyl
sulfoxide (60 mL) was stirred for 16 h at 50 ꢁC. The mixture was poured into ice
water (50 mL) and extracted with ethyl acetate (3 · 30 mL). The combined organic
layers were washed with sat. aq. sodium chloride (30 mL), dried over sodium sulfate,
and concentrated under reduced pressure to give 3-azidopropionaldehyde diethyl
acetal as a pale yellow oil which was used in the next step without further purifica-
1
tion (9.9 g, 57 mmol, and 95% yield). Two hundred megahertz H NMR (CDCl3) d
1.19 (t, 6H, J = 5.5 Hz), 1.86 (m, 2H), 3.35 (m, 2H), 3.47 (m, 2H), 3.65 (m, 2H), 4.59
(m, 1H). To a solution of 3-azidopropionaldehyde diethyl acetal (9.9 g, 57 mmol) in
methanol (100 mL) was added 5% Pd/C (0.5 g). The apparatus was charged with
hydrogen (50 psi) and shaken for 2 h at room temperature. The solution was filtered
through celite and the solvent was removed under reduced pressure to give 16 as a
1
yellow oil (8.2 g, 56 mmol, and 98% yield). Three hundred megahertz H NMR
(CDCl3) d 1.19 (t, 6H, J = 5.3 Hz), 1.79 (q, 2H, J = 5.0 Hz), 2.75 (t, 2H,
J = 5.3 Hz), 3.49 (m, 2H), 3.64 (m, 2H), 4.56 (t, 1H, J = 5.2 Hz).
2.7. (R)-N-(3,3-Diethoxypropyl)-3,3-dimethyl-2,4-dihydroxybutanamide
To a solution of 16 (2 g, 13.6 mmol) in toluene (10 mL) was added (R)-pantolac-
tone 17 (1.3 g, 10 mmol) and the mixture was heated at reflux for 16 h. The solvent
was removed under reduced pressure and the product was purified by chromatogra-
phy on silica gel (9:1 ethyl acetate/hexane) to give the title compound as a viscous
1
yellow oil (2.38 g, 8.6 mmol, and 86% yield). Three hundred megahertz H NMR
(D2O) d 0.79 (s, 3H), 0.82 (s, 3H), 1.09 (t, 6H, J = 4.7 Hz), 1.72 (m, 2H), 3.19 (t,

J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
95
2H, J = 3.8 Hz), 3.31 (m, 2H), 3.51 (m, 2H), 3.62 (m, 2H), 3.86 (s, 1H), 4.56 (t, 1H,
J = 3.9 Hz). 13C NMR (D2O) d 9.85, 14.71, 16.01, 28.22, 30.19, 34.15, 58.36, 63.86,
71.25, 97.10.
2.8. (R)-N-(3-Oxopropyl)-3,3-dimethyl-2,4-dihydroxybutanamide 18
A solution of (R)-N-(3,3-diethoxypropyl)-3,3-dimethyl-2,4-dihydroxybutanamide
(0.41 g, 1.5 mmol) in aq. HCl (10 mL, 0.1 M) was stirred at room temperature for
16 h. The solvent was removed under reduced pressure to give 18 as a pale yellow
oil which was used in the next step without further purification (0.26 g, 1.3 mmol,
1
and 87%). Three hundred megahertz H NMR (D2O) d 1.01 (s, 6H), 1.89 (m, 2H),
3.39–3.60 (m, 4H), 3.92 (s, 1H).
2.9. P1-Adenosine-50-diphosphate-P2-[3(R)-hydroxy-4-(2- formylethylamino)-2,
2-dimethyl-4-oxobutyl] ester 13
To a solution of 18 (0.20 g, 1.0 mmol) in water (10 mL) was added phosphoenol-
pyruvate (1.04 mmol), ATP (0.20 mmol), MgCl2 (0.80 mmol), KCl (0.40 mmol), and
triethanolamine hydrochloride (18 mg, 0.10 mmol). The pH was adjusted to 7.8 and
the volume was adjusted to 20 mL by addition of water. Pantothenate kinase (20 U),
pyruvate kinase (20 U), and sodium azide (5 mg) were added and the mixture was
allowed to stand at room temperature. HPLC analysis and enzymatic determination
of phosphoenolpyruvate showed the reaction to be complete after 16 h. ATP
(1.0 mmol) was added and the pH was adjusted to 7.0. Phosphopantetheine adenyl-
yltransferase (20 U) and inorganic pyrophosphatase (20 U) were added and the mix-
ture was left at room temperature. HPLC analysis and enzymatic determination of
ATP showed the reaction to be complete after 96 h. The solution was centrifuged
to remove insoluble matter and the pH was adjusted to 4.0 with 2 N aq. HCl. The
solution was loaded onto a DEAE–Sepharose column (2.5 · 18 cm), which had been
equilibrated with 3 mM aq. HCl. The column was washed with 100 mL of 3 mM
HCl followed by a linear gradient of LiCl (0–0.2 M) in 3 mM HCl (total volume
450 mL). Compound 13 eluted at 0.11–0.13 M LiCl, as determined by HPLC anal-
ysis. Fractions containing product were combined, adjusted to pH 4.0, and lyophi-
1
lized to give 13 (465 mg, 0.67 mmol, and 67% yield). Three hundred megahertz H
NMR (D2O, approximately 1:1 aldehyde and hydrate) d 0.81 (s, 3H), 0.92 (s, 3H),
1.87 (m, 1H), 2.80 (t, 1H, J = 11.9 Hz), 3.29 (m, 1H), 3.61 (m, 2H), 3.85 (m, 1H),
4.03 (s, 1H), 4.29 (m, 3H), 4.42 (s, 1H), 4.58 (s, 1H), 5.1 (t, 0.5 H, J = 11.0 Hz),
6.12 (d, 1H, J = 10.2 Hz), 8.16 (s, 1H), 8.45 (s, 1H), 9.71 (s, 0.5H). MS: [M ꢀ H]
calcd for C23H35N7O14P2 m/z 696.1798. Found 696.1.
2.10. S-(2-Mercaptoethyl)thioacetate 14a
To a solution of 1,2-ethanedithiol (40 mmol) in pyridine (10 mL) and dichlorom-
ethane (10 mL) was added acetic anhydride (40 mmol). The reaction was stirred at
room temperature for 16 h. The solvent was removed under reduced pressure and

96
J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
the product was purified by chromatography on silica gel using dichloromethane/
hexanes 1:1 to give 14a as a light yellow oil (4.13 g, 30.4 mmol, and 76% yield). Three
hundred megahertz 1H NMR (CDCl3) d 1.59 (t, 1H, J = 10.2 Hz), 2.30 (s, 3H), 2.63
(q, 2H, J = 7.9 Hz), 3.03 (m, 2H).
2.11. S-(2-Mercaptoethyl)-3-oxothiobutanoate 14b
To a solution of ethanedithiol (1.9 mL, 2.1 g, and 22 mmol) in tetrahydrofuran
(10 mL) was added diketene (1.6 mL, 1.7 g, and 20 mmol) and dimethylaminopyri-
dine (2 mmol). The solution was stirred for 18 h at room temperature. Diethyl ether
(70 mL) was added and the resulting solution was washed with aq. sodium bisulfate
(5%, 2 · 10 mL), sat. aq. sodium bicarbonate (2 · 10 mL), and sat. aq. sodium chlo-
ride (2 · 10 mL). The organic layer was dried over magnesium sulfate, concentrated
under reduced pressure, and purified by chromatography on silica gel (ethyl acetate/
hexanes/acetic acid 2:8:0.5) to give 14b as a yellow oil (2.88 g, 16.2 mmol, and 81%
yield). Three hundred megahertz 1H NMR (CDCl3) d 1.69 (m, 1H), 2.32 (s, 3H), 2.79
(m, 2H), 3.20 (m, 2H), 3.76 (s, 2H).
2.12. (2-Mercaptoethyl)-3-oxobutanoate 14c
To a solution of 2-mercaptoethanol (1.5 mL, 1.7 g, and 22 mmol) in tetrahydro-
furan (10 mL) was added diketene (1.6 mL, 1.7 g, and 20 mmol) and dimethylamino-
pyridine (0.24 g, 2.0 mmol). The solution was stirred for 18 h at room temperature.
Diethyl ether (70 mL) was added and the resulting solution was washed with aq.
sodium bisulfate (5%, 2 · 10 mL), sat. aq. sodium bicarbonate (2 · 10 mL), and
sat. aq. sodium chloride (2 · 10 mL). The organic layer was dried over magnesium
sulfate, concentrated under reduced pressure, and purified by chromatography on
silica gel (ethyl acetate/hexanes 2:3) to give 14c as a yellow oil (2.01 g, 12.4 mmol,
and 62% yield). Three hundred megahertz 1H NMR (CDCl3) d 1.59 (t, 1H,
J = 7.9 Hz), 2.37 (s, 3H), 2.80 (m, 2H), 3.53 (s, 2H), 4.29 (t, 2H, J = 6.2 Hz).
2.13. N-(S-2-Mercaptoethyl thio)methyl phenylacetamide 26
A solution of 25 [16] (0.87 g, 5.3 mmol) in trifluoroacetic acid (10 mL) was added
dropwise to a solution of ethanedithiol (1 g, 10.2 mmol) in trifluoroacetic acid
(15 mL) over 10 min. The reaction was stirred at room temperature for 1.5 h. Triflu-
oroacetic acid was removed under reduced pressure. The residue was dissolved in
50 mL methylene chloride and washed with 0.2 M sodium bicarbonate solution
(3 · 20 mL) and brine 920 mL). The organic layer was dried over magnesium sulfate.
The solvent was removed by rotary evaporation. The crude product was dissolved in
2 mL methylene chloride, and purified by chromatography using 1:1 hexane/ethyl
acetate as the eluent. The solvent was removed to give 26 as a colorless liquid
1
(0.56 g, 2.3 mmol, and 43%). The product was stored under N2. H NMR (CDCl3,
300 MHz) 7.2–7.4 (m, 5H) 5.7–5.8 (s, broad, 1H), 4.36 (d, 2H), 3.61 (s, 2H), 2.73
(m, 4H), 1.61 (t, 1H).

J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
97
2.14. Compound 28
To a solution of 26 (0.56 g, 2.3 mmol) and triethylamine (0.25 g, 0.35 mL, and
2.3 mmol) in anhydrous methylene chloride (20 mL) at 0 ꢁC, a solution of c-meth-
oxycrotonyl chloride [17] 27 (0.31 g, 2.3 mmol) in methylene chloride (10 mL) was
added dropwise. The ice bath was then removed, the brown reaction mixture was
stirred at room temperature for 25 min and washed with 0.1 M potassium acetate/
acetic acid buffer (pH 4, 2 · 20 mL) and brine (20 mL). The organic layer was dried
over magnesium sulfate and the solvent was evaporated. The product was purified by
chromatography on silica gel with 2:1 hexanes/ethyl acetate followed by 1:2 hexanes/
ethyl acetate. The solvent was removed to give 28 as an orange liquid (0.40 g,
1
1.2 mmol, and 51% yield). H NMR (CDCl3, 250 MHz) 7.2–7.4 (m, 5H), 6.85 (dt,
1H), 6.35 (dt, 1H), 6.3–6.5 (br s, 1H, triplet at high concentration), 4.41 (d, 2H),
4.08 (q, 2H), 3.60 (s, 2H), 3.41 (s, 3H), 3.03 (t, 2H), 2.67 (t, 2H). 13C NMR (CDCl3)
189.35, 170.73, 140.75, 134.52, 128.99, 128.51, 127.22, 126.87, 60.02, 43.20, 40.82,
30.56, 28.60.
2.15. Compound 14d
To a solution of 28 (0.10 g, 0.29 mmol) in methanol (10 mL) was added aq. Tris
buffer (0.1 M, pH 7.5) until a precipitate began to form. Penicillin amidohydrolase
(50 U) was added, the pH was readjusted to 7.5 by addition of 0.1 M aq. HCl,
and the solution was stirred at room temperature. The reaction was monitored for
thiol formation by addition of a 20 lL sample to Tris buffer (1 mL, pH 8) containing
DTNB (0.01 mM) and measuring the absorbance at 412 nM. Thiol concentration
was calculated using e420 = 13.6 · 103 Mꢀ1 cmꢀ1. A maximum thiol concentration
of about 0.22 mmol was observed after 3 h. The resulting solution was frozen and
aliquots were used directly in assays for crotonase activity.
2.16. Enzyme assays
Chloramphenicol acetyltransferase activity was determined as previously de-
scribed [1] in Tris buffer (1 mL, 0.1 M, pH 7.0) containing chloramphenicol
(0.1 mM), DTNB (0.2 mM), acetyl-CoA (0.01–0.08 mM) or analogue 8 (0.02–
0.15 mM), and chloramphenicol acetyltransferase from Escherichia coli (0.01 U).
The reactions were monitored at 412 nm using e412 = 13.6 · 103 Mꢀ1 cmꢀ1. The Ki
value for 10 was determined from assays at 0, 50, 100, and 200 lM concentrations
of 10. Activity with 11a was determined using 30% ethanol with the Tris buffer, with
0.2 U enzyme and 14a (50 mM), and was compared to activity with natural substrate
under the same solvent conditions. The assay was initiated by addition of 13
(0.1 mM) and monitored for disappearance of the thioester absorbance at 233 nm.
Assays with natural substrate in the presence of 13 (0.1 mM) were performed to test
for inhibition.
Carnitine acetyltransferase activity was determined as previously described [18] in
potassium phosphate buffer (1.0 mL, 25 mM, pH 7.5) containing carnitine acetyl-

98
J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
transferase from pigeon breast muscle (0.026 U with acetyl-CoA and 8, 3.4 U with
10), -carnitine (2.0 mM), DTNB (0.2 mM), and acetyl-CoA (0.01–0.1 mM) or ana-
logue 8 (0.007–0.050 mM) or 10 (0.014–0.125 mM). The reactions were monitored at
412 nm using e412 = 13.6 · 103 Mꢀ1 cmꢀ1
L
.
Citrate synthase activity was determined as previously described [1] in Tris buffer
(1 mL, 0.1 M, pH 8.0) containing citrate synthase from Porcine heart (0.01 U), oxa-
loacetate (0.5 mM), DTNB (0.1 mM), acetyl-CoA (3–30 lM) or analogue 8 (0.02–
0.15 mM). The reactions were monitored at 412 nm using
e412 = 13.6 ·
103 Mꢀ1 cmꢀ1. The Ki value for 10 was determined from assays at 0, 20, and
40 lM concentrations of 10. Activity with 11a was determined using 30% ethanol
in the Tris buffer, with
L
-malate (3 mM), NAD (0.22 mM), citrate synthase
(0.2 U), malate dehydrogenase (12 U), and 14a (50 mM). The assay was initiated
by addition of 13 (0.1 mM) and monitored for appearance of NADH at 340 nm.
For comparison, activity with natural substrate under the same solvent conditions
was determined as was activity with natural substrate in the presence of the same
concentration of 13.
Phosphotransacetylase activity was determined in Tris buffer (1.0 mL, 0.1 M, pH
7.4) containing acetyl phosphate (7.4 mM), ammonium sulfate (19 mM), coenzyme
A (28–86 lM) or homocysteamine-CoA 9 (28–160 lM), and phosphotransacetylase
from Bacillus stearothermophilus (0.03 U). Initial rates were measured from the in-
crease in absorbance due to the thioester bond at 233 nm (e233 = 4400 Mꢀ1 cmꢀ1).
Assays of hemithioacetal substrate analogue 11a with citrate synthase and chlor-
amphenicol acetyltransferase were performed as described above. Assays of 3-
hydroxyacyl-CoA dehydrogenase were done in Tris buffer (1.0 mL, 50 mM, pH
7.3) containing 30% ethanol and NADH (0.3 mM), 3-hydroxyacyl-CoA dehydroge-
nase from Porcine heart (1.0 U) and 14b (0.17–5 mM) or 14c (50 mM). The baseline
was recorded and the assay was initiated by addition of 13 (0.1 or 0.5 mM). The reac-
tion was monitored at 340 nm using e340 = 6.22 · 103 Mꢀ1 cmꢀ1. Comparative as-
says with natural substrate contained acetoacetyl-CoA (0.1 mM) in place of 13
and 14b or 14c under the same solvent conditions and used only 0.02 U enzyme.
Assays of crotonase were done in Tris buffer (1.0 mL, 50 mM, pH 7.5) containing
crotonase from Bovine liver (1.0 U) and 14d (0.5 mM). The baseline was recorded
and the assay was initiated by addition of 13 (0.2 mM). The reaction was monitored
at 280 nm using e280 = 3.6 · 103 Mꢀ1 cmꢀ1 [19]. Comparative assays with natural
substrate contained crotonyl-CoA (0.1 mM) in place of 13 and 14 d and used only
0.02 U enzyme.
3. Results
The ester analogue of CoA 7 and the corresponding acetate thioester 8 were pre-
pared as shown in Scheme 1. Reaction of the CoA analogue synthon 12 with mercap-
toethanol formed 7 by initial formation of the thioester followed by rearrangement
to the oxoester. 12 was prepared using the enzymes that catalyze the final steps in
CoA biosynthesis as previously reported [1]. 7 was converted to the acetyl thioester

J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
99
Scheme 1. Synthesis of ester and homocysteine analogues of acetyl-CoA.
8 by reaction with S-phenyl thioacetate. The previously reported homocysteamine
analogue 9 [5] was prepared by reaction of 3-aminopropanethiol with 12. 9 was also
converted to the acetyl thioester 10 by reaction with S-phenyl thioacetate.
Hemithioacetal analogues 11 were generated in situ from the truncated aldehyde
analogue of CoA 13 and the appropriate thiol 14 (Scheme 2). 13 was prepared as
shown in Scheme 3. Chloropropionaldehyde diethyl acetal 15 was converted to the
azide, which was reduced to the amine 16. Reaction of 16 with pantolactone 17 fol-
lowed by acid-catalyzed hydrolysis of the acetal formed the truncated aldehyde ana-
logue of pantetheine 18. 18 was phosphorylated enzymatically to form the
phosphopantetheine analogue 19 catalyzed by pantothenate kinase, using ATP along
with the ATP regeneration system of phosphoenolpyruvate and pyruvate kinase [20].
Scheme 2. In situ generation of hemithioacetal analogues of CoA esters.

100
J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
Scheme 3. Synthesis of a truncated aldehyde analogue of CoA.
19 was coupled with ATP catalyzed by phosphopantetheine adenylyltransferase to
form the dephospho-CoA analogue 13. Inorganic pyrophosphatase was included
in this final step to hydrolyze the pyrophosphate produced, thus making the reaction
irreversible. Attempts to phosphorylate 13 using dephospho-CoA kinase were
unsuccessful.
Thioesters were prepared as shown in Scheme 4 for formation of hemithioacetal
analogues of CoA esters upon reaction with 13. The acetyl thioester 14a was pre-
pared by reaction of excess ethanedithiol 21 with acetic anhydride 20. The acetoac-
etyl thioester 14b was prepared by reaction of ethanedithiol 21 with diketene 22 [21].
The corresponding oxoester 14c was prepared by reaction of mercaptoethanol 23
with diketene. The c-methoxycrotonyl thioester 14d was prepared as shown in
Scheme 5. Reaction of phenylacetamide 24 with formaldehyde and base formed
the N-(hydoxymethyl) amide 25 [16]. Reaction of 25 with ethanedithiol and trifluo-
Scheme 4. Synthesis of acyl derivatives of ethanedithiol and mercaptoethanol.

J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
101
Scheme 5. Synthesis of a methoxycrotonyl thioester of ethanedithiol.
roacetic acid formed 26 [16,22]. 26 was reacted with c-methoxycrotonyl chloride 27,
which was prepared from ethyl crotonate as previously described [17], to form 28.
The thiol was deprotected enzymatically using penicillin acylase to initially form
the intermediate 29, which spontaneously hydrolyzed to the thiol 14d [16]. The thiol
was not characterized due to its instability, but was used directly in an enzymatic
assay.
The CoA ester analogues were tested as substrates for some commercially avail-
able CoA ester utilizing enzymes. Table 1 shows results for the acetyl-CoA oxoester
analogue 8 and the homocysteine analog 10 along with activities for natural acetyl-
Table 1
Kinetic data for acetyl-CoA analogues as substrates and inhibitors of acetyl-CoA utilizing enzymes.
Substrate
Carnitine acetyltransferase
Chloramphenicol
acetyltransferase
Citrate synthase
Km (lM)
Relative Vmax
Km (lM)
Relative Vmax
Km (lM)
Relative Vmax
2a
8
60
35
116
1
32
67
116 (Ki)
1
7
54
20 (Ki)
1
0.83
.006
0.14
<10ꢀ5
0.46
<10ꢀ5
10
a
Relative Vmax values of 1 correspond to 38 lmol/min/mg enzyme for carnitine acetyltransferase,
45 lmol/min/mg enzyme for chloramphenicol acetyltransferase, and 32 lmol/min/mg enzyme for citrate
synthase.

102
J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
CoA 2 for three enzymes. The oxoester 8 was a fairly good substrate for chloram-
phenicol acetyltransferase, carnitine acetyltransferase, and citrate synthase with
Km values slightly lower to 8-fold higher than those of acetyl-CoA and Vmax values
from 14 to >80% those of the natural substrate. The homocysteamine analogue 10
exhibited less than 1% the activity of the natural substrate with carnitine acetyltrans-
ferase and gave no measurable activity with chloramphenicol acetyltransferase or cit-
rate synthase. Km or Ki values were about 2- to 3.5-fold higher than the Km for
acetyl-CoA. The homocysteine CoA analog 9 was also tested as a substrate for phos-
photransacetylase. The Vmax for 9 was just less that 0.1% the activity with CoA 1,
while the Km for 9 was 53 lM, compared to 83 lM for CoA.
Hemithioacetal formation with the truncated aldehyde analog was studied by
NMR, using the phosphopantetheine analog 19 and mercaptoethanol 23 as a model.
At 0.1 M mercaptoethanol and less than 10 mM aldehyde, 90% of the aldehyde was
converted to the hemithioacetal, with the free aldehyde and hydrate accounting for
10%. Efforts to use hemithioacetals of ethanedithiol 21 as substrates for CoA-utiliz-
ing enzymes were not successful due to the insolubility of ethanedithiol, even in the
presence of organic cosolvents. The acetyl thioester 11a was generated from 14a and
13 in 30% aq. ethanol to achieve solubility of 14a, and using at least 50 mM 14a to
insure conversion of most of 13 to the hemithioacetal 11a. No measurable activity
was observed with the acetyl-CoA utilizing enzymes chloramphenicol acetyltransfer-
ase and citrate synthase under conditions in which less than 0.1% of the activity of
the natural substrate under the same solvent conditions would have been detectable.
No inhibition of either enzyme by 13 was observed at the concentration of 13 used in
the assays. The acetoacetyl thioester 11b is the hemithioacetal analogue of the
natural substrate for b-hydroxybutyryl-CoA dehydrogenase. Again, 30% aqueous
ethanol was necessary to achieve sufficient solubility of the thiol 14b. The more
water-soluble oxoester 14c was also prepared and tested as a substrate for
b-hydroxybutyryl-CoA dehydrogenase at a concentration of 50 mM in the presence
of 0.1 mM 13. However, no measurable activity was observed with 14b or 14c, indi-
cating less than 0.1% of the activity of the natural substrate. 14d was designed as a
water-soluble compound for which the hemithioacetal 11d was viewed as a potential
substrate for crotonase. However, no crotonase activity was observed with 14d in the
presence of 13, again indicating less than 0.1% of the activity of the natural substrate.
Computational studies were conducted on the simple hemithioacetal 30 to model
the structure and conformation of the hemithioacetal linkage (Fig. 2). Energy min-
ima were found for the three staggered conformations, which were optimized at the
B3LYP/6-311+Gd,p level with an aqueous solvation model. Conformer 30a, in
which the two methyl groups are anti was determined to be about 0.4 kcal/mol lower
Fig. 2. Three staggered conformations of a simple model hemithioacetal.

J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
103
in energy than 30b in which the S-methyl group is anti to the hydroxyl group but
about 0.5 kcal/mol higher in energy than 30c in which the S-methyl group is anti
to hydrogen. The same order but somewhat larger energy differences were observed
in MP2 calculations using the same basis set and in calculations without the solva-
tion model. The dihedral angle between methyl groups in conformer 30a is 171ꢁ, the
methyl being skewed 9ꢁ away from the OH and toward H. The C–S bond lengths are
˚
about 1.8 A and the C–S–C angle is about 100ꢁ in all three structures.
4. Discussion
Syntheses of the oxoester 7 and homocysteamine 9 analogues of CoA were based
on the previously reported general CoA analogue synthon 12 [1]. In this synthon, the
amide bond nearest the thiol group is replaced with a thioester and for synthetic con-
venience the thiol group is replaced with a methyl group, or in more recent work by a
hydrogen atom. Reaction of 12 with mercaptoethanol resulted initially in formation
of the thioester. The characteristic thioester absorbance at 230–235 nm was observed
as a shoulder on the adenine absorbance peak centered at 260 nm. Rearrangement to
the oxoester was observed by monitoring the disappearance of this shoulder in the
UV spectrum. The equilibrium between S-acetyl thiopropanol and O-acetyl thiopro-
panol is reported to favor the oxoester by a factor of 56 [23], and the equilibrium
constant between 7 and the thioester isomer is expected to be similar. The yield of
7 was only about 15%, apparently due to substantial hydrolysis of the thioester
and/or oxoester over the course of the rearrangement reaction. This rearrangement
was conducted under acidic conditions to avoid base-catalyzed hydrolysis and thiol
oxidation. The homocysteamine analogue 9 was prepared by the more standard ami-
nolysis reaction employed previously in the synthesis of several other CoA ana-
logues. The yield in the conversion of 7 to the S-acetyl derivative of only 42% was
also apparently due largely to partial hydrolysis of the ester under the basic condi-
tions of the acetylation reaction. The synthesis of the truncated aldehyde analogue
of dephospho CoA 13 followed standard reactions for the synthesis of the aldehyde
analogue of pantothenic acid 18. 18 was efficiently phosphorylated by pantothenate
kinase and the resulting phosphopantothenic acid analogue 19 was efficiently con-
verted to the dephospho-CoA analogue 13. However, attempts to phosphorylate
13 catalyzed by dephospho-CoA kinase were not successful. Addition of mercap-
toethanol or ethanolamine to convert the aldehyde to a hemithioacetal or imine,
respectively, bearing the hydroxyethyl group as a mercaptoethyl mimic also failed
to facilitate phosphorylation of 13. The synthesis of acetyl and acetoacetyl deriva-
tives of ethanedithiol and mercaptoethanol were straightforward as shown in
Scheme 2. Synthesis of the methoxycrotonyl thioester 14 was more complicated.
Reaction of the acid chloride 27 with one equivalent or greater of ethanedithiol re-
sulted in thiol addition to the double bond in addition to reaction with the acid chlo-
ride. Reaction with less than one equivalent of ethanedithiol resulted in formation of
a mixture of products including unacylated, monoacylated, and diacylated dithiol,
and attempts to isolate the desired monoacylated product were unsuccessful.

104
J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
Mono-t-BOC protected ethanedithiol was prepared and reacted with the acid chlo-
ride 27 to form the thioester. However, removal of the t-BOC protecting group un-
der acidic conditions was accompanied by conjugate addition of the thiol group and
other unidentified side reactions. Finally, an enzymatically cleaved thiol protecting
group was employed as shown in Scheme 5. The thioester 28 was purified and char-
acterized. Penicillin acylase was then used to generate the thiol 14d, which was not
isolated or characterized [22].
Previous observation of a more than 10,000-fold decrease in affinity of a carbox-
ylate analogue of acetyl-CoA to citrate synthase upon substitution of the outer-
most amide bond with a pair of methylene groups indicated the potential
importance of this amide bond in the function of CoA [7]. The oxoester is perhaps
the most conservative substitution for this amide bond. Relative to an amide, the
oxoester is less conformationally rigid, lacks the hydrogen bond donating amide
hydrogen, and the carbonyl oxygen should be less basic and thereby a poorer
hydrogen bond acceptor [24]. The results of Table 1 indicate that at least with
the three enzymes studied, the effect of the amide to ester substitution is quite
modest. Only with citrate synthase is the effect on substrate binding more than
about 2-fold and only with chloramphenicol acetyltransferase is there more than
about a 2-fold decrease in Vmax
.
The homocysteamine analogue 9 was reported previously and was shown to serve
as a substrate for the holo-acyl carrier protein synthase from E. coli and as an inhib-
itor of the phosphotransacetylases from E. coli and from C. kluyveri [5,25]. The
S-acetyl derivative 10 has apparently not been previously reported. In studies of
acetoacetyl-CoA thiolase, an acetoacetyl derivative of a homopantetheine analogue
was previously shown to exhibit a less than 2-fold decreased Vmax and an approxi-
mately 3-fold increased Km relative to the normal acetoacetyl pantetheine derivative
[26]. The activity of 9 with phosphotransacetylase observed in this work is in contrast
to the absence of activity observed in previous work [25]. However, this may simply
reflect a greater sensitivity of our assays by use of larger quantities of enzyme. The
similar Km value for 9 relative to CoA is in agreement with the similar Ki value pre-
viously reported for the E. coli enzyme, though a much higher Ki value was reported
for the phosphotransacetylase from C. kluyveri [25]. The kinetic results with 9 and 10
indicate that the extra methylene group has little effect on enzyme binding. However,
the observation of little or no enzyme activity indicates that precise positioning of the
thiol or acetylthio group in an enzyme active site is essential for activity, and this
positioning is perturbed by the extra methylene group. These observations contrast
the previous observations with acetoacetyl-CoA thiolase, where the more complex
acetoacetyl group perhaps facilitates its own proper positioning in the active site
despite the extra methylene group in the cysteamine moiety.
It was envisioned that if the hemithioacetal analogues 11 were accepted as sub-
strates for CoA thioester utilizing enzymes, the truncated CoA aldehyde 13 could
be used to activate simple thioesters as substrates for CoA thioester utilizing en-
zymes. As hemithioacetal formation is reversible, only a catalytic amount of 13 could
be sufficient for quantitative enzymatic processing of the simple thioester. However,
this potential strategy failed as none of the substrate analogues exhibited activity in

J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
105
assays with their target enzymes. The generally limited aqueous solubility of the sim-
ple thioesters also proved problematic.
Calculations of the simple model hemithioacetal 30 indicate that the most stable
conformation is represented by 30c rather than 30a, the conformation that best mim-
ics the trans geometry of the amide bond. The higher energy of 30b may be attributed
to the loss of the stereoelectronic effect between the sulfur lone pair and the
anti-bonding orbital of the C–O bond. The initially surprising greater stability of
30c relative to 30a is supported by the greater stability of the gauche relative to
the anti-conformation of 2-thiobutane (CH3–S–CH2–CH3) in the gas phase and
similar observations in the solid state [27]. This may be due to a favorable van der
Waal distance between methyl groups in the gauche conformation, imparted by
the greater length of the C–S bonds relative to a C–C bond.
Hydrogen bonding between the N–H of the amide bond nearest the thiol group
and the protein is consistently observed in enzyme–CoA complexes, including those
of enzymes studied in this work. This N–H forms a hydrogen bond to the backbone
carbonyl oxygens of Leu 273 in citrate synthase [28], Ala149 in chloramphenicol ace-
tyltransferase [29], Gln555 in carnitine acetyltransferase [30], and Ala96 in crotonase
[31]. Hydrogen bonding of the amide carbonyl oxygen of CoA is less consistently
apparent but is observed with the side chain of Asn373 and the backbone amide
N–H of Gly317 in citrate synthase [28] and a bound water molecule in chloramphen-
icol acetyltransferase [29]. Though hydrogen bonding to the amide N–H is always
observed, a striking result of the studies with the oxoester analogue is that this
hydrogen bond is not critical for substrate binding or catalysis. Replacement of
the amide bond with the hemithioacetal linkage introduces much greater changes
in structure. The bond angles at both the sulfur atom and the hemithioacetal carbon
are substantially smaller than the angles expected for the amide, and the carbon-sul-
˚
fur bonds are about 0.3 A longer than the corresponding bonds of the amide. The
anti-conformer that best mimics the amide bond is not the most stable conformation
of the hemithioacetal and thus the concentration of the potentially active conformer
is only a fraction of the total concentration. Hemithioacetal formation also intro-
duces a chiral center, which further diminishes the concentration of any single struc-
ture, though either stereoisomer could potentially be accepted as a substrate. The
hydroxyl oxygen of the hemithioacetal could potentially hydrogen bond to groups
that form hydrogen bonds to the carbonyl oxygen of the natural amide functionality
at this position. However, the hemithioacetal oxygen is certainly positioned differ-
ently due to the change in carbon hybridization and thus it is uncertain how effective
such hydrogen bonding may be. Some combination of these factors result in a com-
plete absence of observable activity with the hemithioacetal-linked substrate
analogues.
5. Conclusion
A previously reported CoA ester analogue having the amide bond nearest the
thiol/thioester group replaced with a pair of methylene groups and newly reported

106
J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
analogues having a hemithioacetal linkage in place of this amide bond or an extra
methylene group between the amide and thioester all exhibited greatly diminished
binding and/or activity with the appropriate CoA ester utilizing enzymes. In con-
trast, an acetyl-CoA analogue having an ester linkage in place of the amide nearest
the thioester was a very good substrate for the three acetyl-CoA utilizing enzymes
tested. These results suggest that the carbonyl group and/or the double bond char-
acter and planarity of the amide or ester bond and proper distance between this bond
and the thioester are necessary and sufficient for efficient substrate recognition and
catalysis by CoA ester utilizing enzymes.
Acknowledgments
This work was supported by NIH Grant GM45831. NMR facilities were made
possible by NSF CHE-0131146. We thank William Shaw and Suzanne Jackowski
for providing expression plasmids.
References
[1] D.P. Martin, R.T. Bibart, D.G. Drueckhammer, J. Am. Chem. Soc. 116 (1994) 4660–4668.
[2] B. Schwartz, D.G. Drueckhammer, K.C. Usher, S.J. Remington, Biochemistry 34 (1995) 15459–
15466.
[3] B. Schwartz, D.G. Drueckhammer, J. Am. Chem. Soc. 118 (1996) 9826–9830.
[4] B. Schwartz, K.W. Vogel, D.G. Drueckhammer, J. Org. Chem. 61 (1996) 9356–9361.
[5] A.M. Gehring, R.H. Lambalot, K.W. Vogel, D.G. Drueckhammer, C.T. Walsh, Chem. Biol. 4
(1997) 17–24.
[6] K.W. Vogel, D.G. Drueckhammer, J. Am. Chem. Soc. 120 (1998) 3275–3283.
[7] R.T. Bibart, K.W. Vogel, D.G. Drueckhammer, J. Org. Chem. 64 (1999) 2903–2909.
[8] P.K. Mishra, D.G. Drueckhammer, Chem. Rev. 100 (2000) 3283–3309.
[9] K.C. Usher, S.J. Remington, D.P. Martin, D.G. Drueckhammer, Biochemistry 33 (1994) 7753–7759.
[10] K.W. Vogel, L.M. Stark, P.K. Mishra, W. Yang, D.G. Drueckhammer, Bioorg. Med. Chem. 8
(2000) 2451–2460.
[11] D.P. Martin, D.G. Drueckhammer, Biochem. Biophys. Res. Commun. 192 (1993) 1155–1161.
[12] A. Geerlof, A. Lewendon, Biochemistry 274 (1999) 27105–27111.
[13] P.K. Mishra, P. Park, D.G. Drueckhammer, J. Bacteriol. 183 (2001) 2774–2778.
[14] R.B. Calder, R.S.B. Williams, G. Ramaswamy, C.O. Rock, E. Campbell, S.E. Unkles, J.R.
Kinghorn, S. Jackowski, J. Biol. Chem. 274 (1999) 2014–2020.
[15] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, J.A.
Montgomery, T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V.
Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H. Nakatsuji, M. Hada, M.
Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H.
Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian, J.B. Cross, C. Adamo, J. Jaramillo, R.
Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y.
Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich,
A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman,
J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A.
Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez,
J.A. Pople, Gaussian 03 Revision A.1, Gaussian, Pittsburgh, PA, 2003.

J. Xun et al. / Bioorganic Chemistry 33 (2005) 90–107
107
[16] M. Royo, J. Alsina, E. Griralt, U. Slomcyznska, F. Alberico, J. Chem. Soc., Perkin Trans. I. (1995)
1095–1102.
[17] K.E. Harding, B.A. Clement, L. Moreno, J. Peter-Katalinic, J. Org. Chem. 46 (1981) 940–948.
[18] R.R. Ramsay, P.K. Tubbs, FEBS Lett. 45 (1975) 21–25.
[19] H.M. Steinman, R.L. Hill, Methods Enzymol. 35 (1975) 136–151.
[20] B.L. Hirschbein, G.M. Mazenod, G.M. Whitesides, J. Org. Chem. 47 (1982) 3765–3766.
[21] R.J. Clemens, Chem. Rev. 86 (1986) 241–318.
[22] Y. Kiso, M. Yoshida, T. Kimura, Y. Fujiwara, M. Shimokuru, Tetrahedron Lett. 30 (1989) 1979–
1982.
[23] W.P. Jencks, M. Gilchrist, J. Am. Chem. Soc. 86 (1964) 4651–4654.
[24] M.H. Abraham, J.A. Platts, J. Org. Chem. 66 (2001) 3484–3491.
[25] M. Shimizu, Methods Enzymol. 18 (1971) 322–338.
[26] J.T. Davis, R.N. Moore, B. Imperiali, A.J. Pratt, K. Kobayashi, S. Masamune, A.J. Sinskey, C.T.
Walsh, T. Fukui, K. Tomita, J. Biol. Chem. 262 (1987) 82–89.
[27] S.H. Gellman, Biochemistry 30 (1991) 6633–6636.
[28] S. Remington, G. Wiegand, R. Huber, J. Mol. Biol. 158 (1982) 111–152.
[29] A.G.W. Leslie, P.C.E. Moody, W.V. Shaw, Proc. Natl. Acad. Sci. USA 85 (1988) 4133–4137.
[30] Y.-S. Hsiao, G. Jogl, L. Tong, J. Biol. Chem. 279 (2004) 31584–31589.
[31] C.K. Engel, T.R. Kiema, J.K. Hiltunen, R.K. Wierenga, J. Mol. Biol. 275 (1998) 847–859.