Polyene substrates with unusual methylation patterns to probe the active sites of three catalytic antibodies

Article abstract of DOI:10.1016/S0968-0896(01)00402-3

The synthesis of two tetraenes that differ in their methylation pattern from the natural substrate in lanosterol biosynthesis, 2,3-oxidosqualene, and their examination with three catalytic antibodies is described. The design of these novel, linear terpenoid structures was governed by initial results obtained from the characterization of the three catalytic antibodies. These were generated by immunization with a steroidal hapten that mimics multicyclization without the necessity for anti-Markovnikov additions or ring expansions. Such a reaction cascade would represent a more 'primitive' version compared to the oxidosqualene cyclization observed in lanosterol, cycloartenol and β-amyrin biosynthesis and would not require a tail-to-tail connection of the third and fourth isoprene unit as seen in squalene. The first tetraene design (A) only contains trisubstituted double bonds and hence its synthesis starts from farnesol and tris-norgeraniol. The second tetraene design (B) is considered the more precise match to the inducing hapten that generated the antibody collections by exhibiting one disubstituted double bond and its synthesis utilizes a tris-norgeraniol derivative and a symmetrical bis-allylic alcohol as key building blocks. Chromatographic comparison studies lead to the conclusion that the currently studied antibodies also produce monocyclic products from the two substrates as has been formerly observed with a squalene-derived substrate. In contrast, 2,3-oxidosqualene is not accepted by these catalysts supporting the notion that the current substrates are fully bound by recognition of both terminal functional groups.

Full text of DOI:10.1016/S0968-0896(01)00402-3

Bioorganic & Medicinal Chemistry 10 (2002) 1249–1262
Polyene Substrates with Unusual Methylation Patterns to Probe
the Active Sites of Three Catalytic Antibodies
Geun Tae Kim, Marion Wenz, Jong Il Park, Jens Hasserodt* and Kim D. Janda*
Department of Chemistry, The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, CA92037, USA
Received 19 October 2001; accepted 19 October 2001
Abstract—The synthesis of two tetraenes that differ in their methylation pattern from the natural substrate in lanosterol biosyn-
thesis, 2,3-oxidosqualene, and their examination with three catalytic antibodies is described. The design of these novel, linear ter-
penoid structures was governed by initial results obtained from the characterization of the three catalytic antibodies. These were
generated by immunization with a steroidal hapten that mimics multicyclization without the necessity for anti-Markovnikov addi-
tions or ring expansions. Such a reaction cascade would represent a more ‘primitive’ version compared to the oxidosqualene cycli-
zation observed in lanosterol, cycloartenol and b-amyrin biosynthesis and would not require a tail-to-tail connection of the third
and fourth isoprene unit as seen in squalene. The first tetraene design (A) only contains trisubstituted double bonds and hence its
synthesis starts from farnesol and tris-norgeraniol. The second tetraene design (B) is considered the more precise match to the
inducing hapten that generated the antibody collections by exhibiting one disubstituted double bond and its synthesis utilizes a tris-
norgeraniol derivative and a symmetrical bis-allylic alcohol as key building blocks. Chromatographic comparison studies lead to
the conclusion that the currently studied antibodies also produce monocyclic products from the two substrates as has been formerly
observed with a squalene-derived substrate. In contrast, 2,3-oxidosqualene is not accepted by these catalysts supporting the notion
that the current substrates are fully bound by recognition of both terminal functional groups. # 2002 Elsevier Science Ltd. All
rights reserved.
Introduction
The field of catalytic antbodies has already procured
catalysts for a variety of cationic cyclizations.1 How-
ever, only one of them carries out a tandem cyclization.2
It is of particular interest to extend this study to multi-
cyclizations analogous to triterpene cyclizations. Any
obtained catalyst would contribute a plethora of
insights regarding active site requirements to the already
building knowledge on triterpene cyclases3 and may
bear implications for the evolution of cyclase activity.4
Particularly, the hypothesis of ‘minimal assistance’5 by
the enzyme, once initiation of multicyclization has
occurred, can be verified by the catalytic antibody
approach. This is due to the fact that a hapten design
must be realistic and hence monoclonal antibodies gen-
erated will not be able to address every hypothetical
detail of transition states along the reaction coordinate.6
Thus, this ongoing study7 has started with a hapten
design (HA8, Fig. 1) that covers the initiation of the
Figure 1. Different degrees of steric congruency between substrate
redesigns A and B and the inducing transition state analogue/hapten
showing postulated chair–chair–envelope conformations.
*Corresponding authors. Tel.: +1-858-784-2516; fax: +1-858-784-
2595; e-mail: kdjanda@scripps.edu
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(01)00402-3

1250
G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
cascade process by introducing an N-oxide moiety in
ring A of a steroidal framework. This should fulfill two
requirements: mimicking heterolytic epoxide opening of
any suitable substrate (TS analogue approach8) and
inducing acidic amino acid residues in the vicinity that
could act as proton donors for epoxide protonation
(bait-and-switch approach9).
hypothesis that the three catalysts were not able to
enforce a productive fold of the hydrocarbon chain of
49.
Two general polyene substrate designs were targeted in
this study differing only in their methylation pattern.
Design A (Fig. 1) displays four isoprene units with
head-to-tail connections before being linked to the fifth
by a tail-to-tail link thus giving rise to all-Markovnikov
additions in a reaction leading to a 6,6,6,5-steroidal
framework. Substrates adhering to this design may
cyclize more readily under the influence of the current
antibodies, since high-energy anti-Markovnikov addi-
tions or ring expansions are not required.
The main body of the hapten consists of a steroidal
framework with all-trans connections between rings
thus representing a chair–chair–chair–envelope fold of
any suitable polyene substrate. This can be considered
an evolutionary more ‘primitive’ triterpene cyclization
mode compared to eucaryotic oxidosqualene cyclases
that enforce chair–boat–chair folds in addition to facili-
tating methyl and hydrogen migrations.3 Since the
synthesis of hapten HA87b was started from a suitable
commercial steroid—lithocholic acid—its features were
short of one methyl group to make it an ideal starting
point for mimicking a ‘primitive’ substrate design with
head-to-tail connections between the first four isoprene
units thus allowing for all-Markovnikov additions. The
latter contrasts oxidosqualene cyclase (OSC) and bac-
terial cyclases. In order to circumvent the hurdle of an
anti-Markovnikov addition as a direct consequence of
the constitution of squalene, OSC has been shown to
control a five-to-six-membered ring expansion.3
Design B lacks the 10-methyl substitutent that is not
accounted for in the hapten design. This should eradi-
cate any doubts about anti-HA8 antibody combining
sites not being able to accommodate an extra methyl
group. However, such a design implies a somewhat
higher energy barrier for the second electrophilic dou-
ble-bond addition since this double bond is only di-
substituted.
Results and Discussion
Substrate design A was planned to lead to two synthetic
targets differing slightly in the number of methyl sub-
stituents on the epoxide trigger group (1 and 2; Fig. 2).
The retrosynthetic analysis led to two starting materials,
farnesol and geraniol. Both molecules provide all iso-
prenoidal double bonds in configurational purity and
could conceivably be fused using allylic cross-coupling
methodology. In contrast, design B in the form of the
synthetic target 3 carries a disubstituted double bond.
Thus farnesol was ruled out as a suitable starting mat-
From a collection of 25 anti-HA8 antibodies (produced
in sufficient amounts for initial tests via standard hybri-
doma protocol) three catalysts (25A10, 20C7, 15D6)
were selected that performed monocyclization of a
squalene-derived substrate (49, Chart 1).7a These cata-
lysts controlled initiation and formation of ring A of the
steroid nucleus with pronounced preference for one
substrate enantiomer (determined for 25A10), but did
not foster multicyclization. However, this original sub-
strate (49) cannot be considered an ideal match for the
hapten design.7a It was therefore concluded that more
refined substrates had to be created in order to test the
Chart 1. Dashed circles indicate deviations from the original substrate
49.
Figure 2. Fractionation of substrate prototypes A and B into suitable
synthetic starting materials.

G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
1251
erial in favor of a purely synthetical, symmetrical syn-
thon carrying two allylic alcohol moieties (Fig. 2). This
building block was planned to be elongated by a double
bond-bearing four-carbon unit before coupling to the
synthon derived from geraniol. The essential epoxide
moiety was to be installed as late as possible due to its
instability. Based on this consideration the polyenic
nature of the target molecule made sulfur ylid chemistry
the method of choice.
dimethylsilyl ether followed by LAH reduction. The
resulting alcohol 9 was benzylated to provide a protect-
ing group orthogonal to the silyl ether. The latter was
subsequently removed and the resulting allylic alcohol
10 iodinated, followed by an allylic cross-coupling11
with synthon 23 (an allylic sulfone derived from geranyl
actetate; Scheme 2) in excellent yields. It should be
noted that other allylic cross-coupling techniques based
on allylic barium reagents12 and allyl thioethers13 failed
to give any product.14 However, the subsequent removal
of the sulfone substituent turned out to be an unex-
pected challenge. Typical techniques using Li/NH3, Li/
EtNH2, and lithium naphthalide15 failed to give homo-
genous products; 5% Na/Hg/MeOH provided 13 as a
mixture (4:1, 79% yield) with a minor, isomerized pro-
duct exhibiting a disubstituted double bond due to 1,2-
migration. Finally, a palladium-catalyzed reduction
with superhydrideTM (lithium triethyl borohydride)
produced the desired key intermediate 13 in 84%
yield.16 Here, double-bond migration is avoided as the
bulky hydride preferentially attacks on the less hindered
carbon of the intermediate palladium–allyl complex.
Subseqent removal of the benzyl group via the very mild
lithium–naphthalide method followed by PyrSO3 oxi-
dation17 to the aldehyde prepared the molecule for
establishment of the dimethylepoxide moiety in 15 with
the sulfur ylid derived from diphenyl i-propyl sulfonium
tetrafluoroborate.18 Final steps consisted of silyl group
removal and the stepwise oxidation of alcohol 16 to the
corresponding acid 1, the target compound, by use of
PyrSO3 and NaClO2.19 This stepwise oxidation pro-
cedure gave superior yields over direct methods using
stronger oxidants like Ag2O or pyridinium dichromate
in DMF.
The synthesis of substrate 1 adhering to design A
(Scheme 1) started with the regioselective SeO2/
tBuOOH oxidation of farnesol10 acetate (4) to the
mono-acetylated bis-allylic alcohol 5. Iodination of the
unprotected hydroxyl group followed by reaction with
deprotonated t-butyl acetate gives acetate 6. Protecting
groups were next swapped from acetate to t-butyl-
Epoxy acid 2 (the nor derivative of 1) was obtained by
replacing diphenyl iso-propyl sulfonium tetrafluoro-
borate with diphenyl ethyl sulfonium tetrafluoroborate
(Scheme 3) in the reaction with aldehyde 26. Subsequent
steps are identical to those shown in Scheme 1.
The synthesis of substrate 3 (Fig. 2), as based on design
B again consisted of fusion of two fragments via coupl-
ing of an allylic sulfone and an allylic halide. Synthesis
of the sulfone fragment 30 was carried out as follows
Scheme 1. (a) SeO2, t-BuOOH; (b) PPh3, I2, imidazole, CH3CN/Et2O
(2:3); (c) t-butyl acetate, HMPA, lithium cyclohexylisopropylamide,
THF, ꢀ78 ꢁC, 7 (66%) and 6 (11%); (d) MeOH, K2CO3, rt, 81%;
(e) TBSCl, imidazole, DMF, 83%; (f) LiAlH4, THF, 83%; (g)
BnBr, n-Bu4NI, NaH, THF/DMF (4:1); (h) n-Bu4NF, THF (98%); (i)
PPh3, I2, imidazole, CH3CN/Et2O (2:3), 81%; (j) 23, n-BuLi, THF,
ꢀ78 ꢁC, 83%; (k) Pd(dppp)Cl2, LiEt3BH, THF, 0 ꢁC, 84%; (l) lithium
naphthalide, THF, ꢀ25 ꢁC, 100%; (m) SO3Py, DMSO, Et3N, CH2Cl2,
81%; (n) Ph2Si–PrBF4, t-BuLi, THF, ꢀ78 ꢁC; (o) n-Bu4NF, THF,
85% over two steps; (p) NaClO2, NaH2PO4, 2-methyl-2-butene,
t-butanol, H2O, 68%.
Scheme 2. (a) NaBH4; (b) TBDPSCl, imidazole, DMF; (c) MeOH,
K2CO3, 83% over two steps; (d) PPh3, I2, imidazole, CH3CN/Et2O
(2:3), 62%; (e) PhSO2Na, DMF, 92%.

1252
G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
(Scheme 4): Aldehyde 18, derived from geraniol, was
prepared as previousely described20 and protected under
standard conditions to form the acetal 27. Its allylic
acetate moiety was then cleaved under standard condi-
tions and the resulting alcohol 28 was converted into the
allylic bromide 29 via the mesylate by subsequent treat-
ment with LiBr.21 Finally, sulfone 30 was formed under
mild conditions by adding the sodium salt of the corre-
sponding sulfinic acid.22
derivative of 1-TMS-propyne27 yielding the corres-
ponding 1-ene-5-yne system 37.28 The propargylic silane
was deprotected selectively in the presence of the TBS–
ether under mildly basic conditions using K2CO3 in
MeOH.29 Further deprotonation of the resulting
primary alkyne 38 with n-BuLi, to generate the
lithium acetylide in situ, and its subsequent reaction
with p-formaldehyde afforded the propargylic alcohol
39,30 which was selectively transformed into the corre-
sponding (E)-allylic alcohol 40 by treatment with Red-
Al1.31 This alcohol was again transformed into its bro-
mide 41 (vide supra).
For the synthesis of the symmetrical unit of the allylic
halide fragment, the stereochemically pure E,E-bisallylic
alcohol 34 was constructed starting with titanium-
mediated g-dimerization23 of the mixed E/Z-isomers of
silylvinylketene acetal 32, obtained from methyl tiglate
(31) (Scheme 5). This highly stereospecific procedure
proved to be a very efficient process, superior to the
alternative double-Wittig reaction starting from succinic
aldehyde that is known to yield an undesired 9:1mix-
ture of the two E/Z isomers.24 Clean 1,2-reduction of
the crystalline unsaturated diester 33 with DIBAH25
afforded the pure key intermediate 34, bearing the cor-
rect stereochemistry as confirmed by NOE experiments.
Compound 34 was then monosilylated using the sodium
monoalkoxide methodology26 to give 35, and brom-
inated to yield 36.
The allylic sulfone 30 was coupled to the allylic bromide
41 via a deprotonation-addition sequence to give the
secondary allylic sulfone 42 (Scheme 6). Subsequent
reductive desulfonylation was carried out using super-
hydride1 and a catalytic amount of Pd(OAc)2/dppp32
to afford the corresponding disubstituted olefin 43 in a
highly regio- and stereoselective manner.33 The TBS
ether remains uncleaved during this step if the reduction
is stopped as soon as desulfonylation is complete. In
order to extend the polyenic carbon backbone, TBAF
deprotection was followed by dehydrative alkylation
with triethyl methanetricarboxylate (TEMT) under
Mitsunobu conditions34 to form the two carbon-elon-
gated compound 45. Subsequent double decarboxyl-
ation was carried out under Krapcho-like conditions35
Further 3-carbon elongation was achieved by reacting
allylic bromide 36 with the in situ-generated 3-lithio
Scheme 3. (a) Ph2SEtBF4, t-BuLi, THF, ꢀ78 ꢁC; (b) n-Bu4NF,
THF, 21% over two steps; (c) NaClO2, NaH2PO4, 2-methyl-2-
butaene, t-butanol, H2O, 83% over two steps.
Scheme 5. (a) LDA, THF, ꢀ65 ꢁC; TMSCl, THF, rt, 84%; (b) TiCl4,
CH2Cl2, 0 ꢁC, 63%; (c) DIBAL, CH2Cl2, ꢀ78 ꢁC, 94%; (d) NaH,
TBSCl, THF, rt, 55%; (e) (i) MsCl, NEt3, CH2Cl2, ꢀ40 ꢁC; (ii) LiBr,
THF; (f) 1-TMS-propyne, n-BuLi, TMEDA, 0 ꢁC, 67% over two
steps; (g) K2CO3, MeOH, rt, 82%; (h) CH2O, n-BuLi, THF, ꢀ60 ꢁC,
65%; (i) Red-Al, toluene, ꢀ78 ꢁC, 69%.
Scheme 4. (a) Ethylene glycol, p-TosOH, toluene, reflux, 70%; (b)
K2CO3, H2O/MeOH, rt, 96%; (c) MsCl, NEt3, CH2Cl2, ꢀ40 ꢁC; LiBr,
THF, rt; (d) PhSO2Na, DMF, 0 ꢁC, 80% over two steps.

G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
1253
by using excess LiCl as the neutral nucleophile to afford
the monoester 46. The epoxide moiety was introduced
by cleavage of the acetal to give aldehyde 25, and
subsequent treatment with the sulfur ylid generated
from iso-propyl diphenyl sulfonium tetrafluoroborate
and t-BuLi.36 Saponification of the ethylester moiety of
epoxide 48 under standard conditions provided the
desired carboxylic-acid substrate 3.
whose low chromogenicity prohibits precise spectro-
photometric detection. In order to obtain chromato-
grams with the degree of resolution depicted in Figure 3
only selected masses [selected ion monitoring (SIM)]
were monitored. These consisted of the major signals
obtained by initial scans of the respective substrates
between 100 and 800 m and masses obtained by the
addition of 18 m analogous to a H2O addition as part of
one of the two termination pathways of cationic cycli-
zation in aqueous media. Since a rearrangement reac-
tion is studied, it was expected that all products would
be detectable by this method, albeit with varying
response levels.
The thus obtained substrates were tested with catalytic
antibodies HA8-25A10, -20C7, -15D6 and a non-related
control antibody (cab). This assay, that employs the
surfactant TWEEN 80 to solubilize the hydrophobic
substrates in phosphate buffer, was previously used with
substrate 49 (Chart 1) to obtain an initial characteriza-
tion of these catalysts.7a While extraction with for
example methyl t-butyl ether and subsequent TLC ana-
lysis gave a valuable qualitative picture of the reaction,
a more quantitative analysis of the extracts using gas
chromatography was not feasible. It was found that the
substrate carboxylic acids and their corresponding pro-
ducts did not show sufficient volatility, even after
methyl ester formation. However, since it was desirable
to detect products directly in the reaction mixture to
avoid any alteration of the product composition
through extraction and derivatization steps, liquid
chromatography was determined to be the superior
method as it allows for direct analysis of the reaction
mixture. Thus, HPLC coupled with mass spectrometry
detection (LCMS) was used throughout this work
because of the highly sensitive detection of products
According to these LCMS analyses (Table 1and Fig. 3),
antibody catalysts 25A10, 20C7 and 15D6 did not show
marked difference in product formation with the new
substrate structures 1 and 3. The product signals with
retention times of 5.05, 4.75 and 4.95 min for the reac-
tions with 1 and 3 and the original substrate 49, respec-
tively, with all three catalysts assumed virtually identical
Scheme 6. (a) n-BuLi, THF, ꢀ78 ꢁC, 91%; (b) LiBHEt3, Pd(OAc)2,
dppp, THF, 0 ꢁC, 83%; (c) TBAF, SiO2, THF, ꢀ50 ꢁC!rt, 89%; (d)
DEAD, PPh3, TEMT, Et2O, rt, 83%; (e) LiCl, H2O/DMSO, 189 ꢁC,
64%; (f) p-TsOH, H2O/acetone, 50 ꢁC, 77%; (g) [Ph2(i-prop)S]+BF4ꢀ,
t-BuLi, THF, ꢀ78 ꢁC, 62%; (h) NaOH, H2O/MeOH, rt, 99%.
Figure 3. P=monocyclic product, G=glycol (hydrolysis), chromato-
grams a–e were acquired using same mobile phase gradient; for f the
gradient had to be shifted to more apolar mixtures. Conditions: 10 mM
IgG 25A10, 400 mM substrate, 50 mM phosphate buffer, pH 7.0, 22 ꢁC,
135 h.

1254
G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
positions with regard to the substrate peak (6.5, 6.05
and 6.45 min, respectively) making monocyclization for
these altered substrate structures the most likely reac-
tion outcome in congruency with the established pro-
duct spectrum from 49.7a A multicyclized structure like
the tetracylic carbocylic acid derived from lanosterol,
which is a possible cyclization product from 49 by an
OSC-like cyclase, elutes 1.8 min earlier than the mono-
cyclic antibody products (chromatogram e; Fig. 3). It
even elutes 0.6 min earlier than the glycol (peak G at
3.25 min in chromatograms a, b, and d) formed by
background hydrolysis from the original substrate 49
despite the fact that it carries only two polar functional
groups (one hydroxyl and one carboxyl) instead of three
(two hydroxyls and one carboxyl). This illustrates the
considerable difference in interaction with the immobile
phase resulting from the largely differing rigidity of
monocyclic versus tetracylic systems. However, chro-
matograms a, b and d (Fig. 3) show distinctive minor
product components with retention times in the area
where dihydroxylated species may be expected resulting
from a termination pathway that consists of water
addition to the terminal carbocation. The individual
mass spectra support this observation. Due to the low
abundance of these minor products, actual isolation and
NMR characterization was not attempted.
recognition of the transition state of substrate 3 by vir-
tue of its largest rate. This may serve as an indication
that indeed substrate design B carrying no methyl sub-
stituent at C10 has superior congruency with the hapten
structure.
Substrate 2 (a mixture of two diastereomeric sets of
enantiomers) exhibits the same methylation pattern as
1. It was hoped that by exclusion of one of the two
methyl substituents on the epoxide moiety one may
prevent any repulsive interactions with the antibody
combining site that may be prevalent in the interaction
with substrates 49, 1, and 3 due to a less space-consum-
ing hapten (Fig. 1). However, any positive effects
resulting from this design alteration were over-
compensated by the inherently lower basicity of a
disubstituted epoxide moiety which results in a marked
drop in reactivity as is evident from Figure 3, chroma-
togram c. This phenomenon has been observed during
the probing of yeast oxidosqualene cyclase with similar
trigger group variations.38 That a comparable mode of
reaction is still promoted by the current antibodies can
be seen from the low-intensity peak at 4.59 min (visible
only after zooming in on the baseline of chromatogram
c) which corresponds to the peak at 4.95 min (mono-
cyclic product7a) in chromatogram d (Fig. 3).
Table 1provides a crude determination of rates of pro-
duct (monocyclic) formation from substrates 1, 2, 3, 49
with all three catalytic antibodies and a non-related
control antibody. The applied substrate concentration
(400 mM) can be regarded as comparatively low for
these catalysts since attempts to reach their saturation
under the current assay conditions has been impossible.
At least part of the reason may be that [S]0 may only be
an apparent substrate concentration due to the fact that
the hydrophobic substrates are mostly contained in the
micelle interior (TWEEN 80!). It can be assumed that
their respective partition constant for diffusion into the
aqueous exterior where the antibody catalyst resides is
rather low. Consequently, this is a situation of low sub-
strate concentration where the Michaelis–Menten equa-
tion assumes the following simpler form:37 v=kcat/KM
[E][S], where [E] is the concentration of free or unbound
enzyme. Since [E] and [S] where held constant in this
study (Table 1), the deviations of the shown rates have
to be attributed to differences in the specificity constant
kcat/KM. Thus, the antibodies show improved molecular
The substrate 3, deprived of one of the four methyl
groups present in 49 and 1, shows a more pronounced
formation of dihydroxylated species. It may be specu-
lated that this has resulted from the substrate leaving
more room for water molecules in the active site. Table
1reports relative velocities for monocyclic product for-
mation by all three catalytic antibodies. While IgG
25A10 turns out to be the most proficient catalyst, all
three antibodies exhibit almost identical product spec-
tra/peak distributions. This may indicate a close genetic
relation as has been observed with other catalytic anti-
bodies obtained by standard hybridoma protocol.39
Importantly, parallel experiments carried out with
2,3-oxidosqualene (50) (Chart 1, Fig. 3), did not show
any activity with the current antibody catalysts. This
illustrates how successfully the hapten design (Fig. 1)
instructed the antibodies to recognize and bind the
entirety of these linear and conformationally quite
mobile molecules including the carboxylic-acid tail.
To obtain some initial insight into the catalytic
mechanism of antibody 25A10, a pH rate profile with
substrate 49 (obtainable in largest amounts in this
study) was constructed. The bell-shaped dependency
indicates the presence of two ionizable groups being
involved in the rate-determining step of this catalytic
antibody.40 The optimum performance was observed at
pH 6.3, rather close to the value seen with natural tri-
terpene cyclases.41 The crystal structure of squalene-
hopene cyclase has revealed the presence of an aspartic
acid residue that—for various structural reasons and
because it was in close contact with the aminoxide moi-
ety of a bound inhibitor42—was made resonsible for
initiation by epoxide or double bond protonation.
Whether such an aspartate or glutamate is a likely
Table 1. Crude rates of monocyclic product formation (relative to
total amount of initial substrate and products expressed in peak area
percent after 135 h reaction time)a
1
3
2
(399, 417, (385, 403, (385, 403, (399, 417,
49
433, 439,
455)b
421, 425,
443)b
425)b
433, 439,
455)b
25A10
20C7
15D6
Control antibody
24.8
3.5
0.5
0
35.3
18.2
11.9
0
0.7
1.5
0.4
0
30.4
2.4
21.2
0
aConditions: 10 mM antibody (IgG), 400 mM racemic substrate, 50 mM
phosphate buffer, pH 7.0, 22 ꢁC, 0.1% TWEEN 80, 135 h.
bIons monitored.

G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
1255
candidate for one of these residues has to await a
detailed kcat–pH dependency study for further support
even though such interpretations generally bear various
pitfalls.43 Substrate 49 carries an ionizable carboxylic
acid group but the impact of its pKa of ca. 4–5 on the
present pH study is believed to be small since at most
studied pH values the substrate exists in its anionic form.
Plates (60F-254), fractions being visualized by UV light,
phosphomolybdic acid solution (or other staining solu-
tions as indicated) with subsequent heat application.
Column chromatography was carried out with Mal-
linckrodt SilicAR 60 silicagel (40–63 mM). Reagent
grade solvents for chromatography were obtained from
Fisher Scientific. Reagent and anhydrous solvents were
obtained from Aldrich Chemical Co. and used without
further purification. All reactions were carried out
under anhydrous conditions and an atmosphere of
argon, unless otherwise noted. Reported yields were
determined after purification to homogenous material.
Conclusion
The substrate structures presented in this contribution
have been chosen to further probe the active site of
three catalytic antibodies raised against a steroidal hap-
ten. The hypothesis that a higher congruency of sub-
strate and hapten structure would enable the catalysts
to enforce a more productive conformation of the poly-
ene chain for multicyclization could not be verified. In
light of the current findings, the three antibodies appear
to lack the capability to properly align the second dou-
ble bond to the first to accomplish bicyclization. None-
theless, these antibodies have been taught to stabilize
the initial TS and to bind the entire length of the sub-
strates most probably in a tightly folded fashion in
congruency with the compactness of the hapten design.
The steroidal framework of the hapten makes a pseudo-
chair–chair–chair fold likely while precision of this fold
has apparently not been accomplished. The pH rate
dependency and optimum of antibody 25A10 also sug-
gests the presence of an acidic residue characteristic for
triterpene cyclases.
Ester (7). To a solution of 5 (2.69 g, 9.60 mmol), imida-
zole (981mg, 14.4 mmol) and triphenylphosphine
(3.78 g, 14.4 mmol) in a 2:3 mixture of acetonitrile and
Et2O (50 mL) was added iodine (3.65 g, 14.4 mmol)
portionwise over 30 min with 10 min intervals at 0 ꢁC.
After 1h, excess iodine was quenched with 15% sodium
thiosulfate and the mixture was extracted with Et2O.
The combined organic extracts were dried over MgSO4
and concentrated under reduced pressure. The residue
was purified by column chromatography (EtOAc/hex-
ane, 1:20) to afford the corresponding iodide (2.98 g,
80%) as a colorles oil.
n-BuLi (10.15 mL, 2.5 M in hexane, 25.37 mmol) was
added slowly to
a solution of N-isopropylcyclo-
hexylamine (4.17 mL, 25.38 mmol) in THF (21 mL) at
0 ꢁC and stirring was continued at this temperature for
20 min before cooling to ꢀ78 ꢁC. The mixture was
slowly treated with freshly distilled t-butyl acetate over
15 min. Stirring was continued for 30 min before the
above iodide (3.30 g, 8.46 mmol) in THF (10 mL) was
added through a double-tipped needle. After 10 min, the
reaction mixture was treated with hexamethylpho-
sphoramide (HMPA, 4.41mL). After a further 30 min,
the reaction was quenched with satd NH4Cl. The
resulting mixture was warmed to room temperature and
extracted with Et2O. The combined extracts were dried
over MgSO4 and evaporated. The residue was purified
Considering the complexity of the task of multi-
cyclization, the required fine tunings to achieve it were
left to a considerable extent to the diversity of the
immune response rather than to hapten design. Thus, it
appears paramount to significantly increase the pool of
possible candidates from the currently tested set of 25
monoclonal antibodies that were obtained by tradi-
tional hybridoma protocols. This may either be done by
variegating the already exisiting catalytic machinery of
the current catalytic antibodies by cloning and various
directed evolution techniques or by deriving an anti-
body library from the immunization with hapten HA8
(Fig. 1).
using
column
chromatography
(EtOAc/hexane,
1:20!1:5) to give 1.883 g (66%) of 7 and 377 mg (11%)
1
of 6 as colorless oils. H NMR (600 MHz, CDCl3) d
1.39 (9H, s), 1.55 (3H, s), 1.56 (3H, s), 1.64 (3H, s),
1.92–1.94 (2H, m), 1.99–2.09 (6H, m), 2.19–2.28 (4H,
m), 4.11 (2H, d, J=6.9 Hz), 5.05–5.36 (2H, m), 5.37–
5.38 (1H, m); 13C NMR (100 MHz, CDCl3) d 15.88,
15.94, 16.23, 26.19, 26.46, 28.06, 28.09, 34.35, 34.77,
39.48, 59.30, 80.01, 123.39, 123.86, 124.74, 133.43,
135.11, 139.49, 172.90; HRMS (MALDI-FTMS) calcd
for C21H36O3 (M+Na+) 359.2562; found 359.2554.
Experimental
General procedures
If not stated otherwise, 1H NMR and 13C NMR spectra
were recorded on Bruker DRX-600, 500 instrument at
600, 500 and 150, 125 MHz, repectively. Chemical shifts
(d) are given in ppm relative to CHCl3 in CDCl3
Conversion of acetate 6 to 7. Acetate 6 (418 mg,
1.10 mmol) was stirred with K2CO3 (15 mg, 0.11 mmol)
in MeOH (5 mL) at room temperature After 15 h, H2O
(20 mL) was added and the mixture was extracted with
Et2O. Solvent evaporation followed by column chro-
matography (EtOAc/hexane, 1:4) gave 303 mg (81%) of
7 as a colorless oil.
1
(7.27 ppm, H: 77.00 ppm, 13C). Signals are quoted as s
(singlet), d (doublet), t (triplet), q (quartet), and m
(multipet). High-resolution mass spetra (HRMS) were
recorded at The Scripps Research Institute on VG ZAB-
ZSE mass spectrometer.
All reactions were monitored by thin-layer chromato-
graphy (TLC), using 0.25 mM Merck Silicagel Glass
Alcohol (9). To a solution of 7 (2.18 g, 6.49 mmol) and
imidazole (1.06 g, 15.5 mmol) in DMF (20 mL) was

1256
G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
added t-butyldimethylsilyl chloride (1.17 g, 7.79 mmol)
at 0 ꢁC. After 2 h, H2O (50 mL) was added and the
mixture was extracted with Et2O (3ꢂ50 mL). The com-
bined extracts were dried over MgSO4 and evaporated
under reduced pressure. The residue was purified by
column chromatography (EtOAc/hexane, 1:40) to give
2.43 g (83%) of 8 as a colorless oil.
(1.419 g, 5.41 mmol) in CH3CN/Et2O 2:3 (25 mL) was
added iodine (1.373 g, 5.41 mmol) portionwise over
30 min with 10 min intervals at 0 ꢁC. After 30 min, the
excess iodine was quenched with 15% sodium thio-
sulfate and the mixture was extracted with Et2O. The
combined extracts were dried over MgSO4 and con-
centrated under reduced pressure. The residue was
purified by column chromatography (EtOAc/hexane,
1:20) to give 1.36 g (81%) of 11 as a colorless oil.
8 (2.74 g, 6.09 mmol) in Et2O (20 mL) was added drop-
wise to a solution of lithium aluminum hydride (346 mg,
9.11 mmol) in Et2O (20 mL) over 15 min at 0 ꢁC. The
reaction mixture was warmed to room temperature,
stirred for 2 h and cooled in an ice bath. H2O (0.35 mL)
was added carefully over 10 min before the addition of
15% NaOH (0.35 mL) and H2O (1.05 mL). This resulted
in a grain-like texture of the solid that was filtered off
through Celite. The filtrate was evaporated to afford an
oil which was purified by column chromatogaphy
(EtOAc/hexane, 1:4) to give 2.21 g (83%) of 9 as a col-
n-BuLi (1.74 mL, 4.35 mmol) was added dropwise to 23
(2.15 g, 4.37 mmol) in THF (15 mL) at ꢀ78 ꢁC. After
1h, 11 (1.36 g, 12.91 mmol) in THF (10 mL) was added
dropwise to the reaction mixture. After the mixture was
stirred for 7 h at ꢀ78 ꢁC, satd NH4Cl was added and the
mixture was extracted with Et2O. The combined
extracts were dried over MgSO4 and concentrated under
reduced pressure. The residue was purified by column
chromatography (EtOAc/hexane, 1:8) to give 2.02 g
(83%) of 12 as a colorless oil. 1H NMR (400 MHz,
CDCl3) d 1.04 (9H, s), 1.13 (3H, s), 1.47–1.50 (2H, m),
1.54 (3H, m), 1.56 (3H, s), 1.58 (3H, s), 1.65–1.70 (2H,
m), 1.91–2.05 (12H, m), 2.28–2.36 (1H, m), 2.84–2.88
(1H, m), 3.42 (2H, t, J=6.4 Hz), 3.57 (2H, t, J=6.0 Hz),
3.70 (1H, dt, J=2.8, 10.4 Hz), 4.47 (2H, s), 4.92–4.97
(2H, m), 5.02–5.09 (2H, m), 7.24–7.45 (13H, m), 7.53–
7.57 (1H, m), 7.63–7.64 (4H, m), 7.79–7.81 (2H, m); 13C
NMR (100 MHz, CDCl3) d 15.86, 15.92, 16.33, 16.47,
19.15, 26.28, 26.55, 26.60, 26.80, 27.96, 30.77, 35.91,
35.99, 39.64, 39.67, 63.35, 64.77, 69.97, 72.82, 116.94,
118.46, 123.83, 124.39, 127.42, 127.56, 127.59, 128.29,
128.62, 129.04, 129.56, 133.24, 133.85, 134.37, 135.09,
135.47, 138.01, 138.60, 138.67, 145.03; HRMS
(MALDI-FTMS) calcd for C53H70O4SSi (M+Na+)
853.4662; found 853.4673.
1
orless oil. H NMR (600 MHz, CDCl3) d 0.04 (6H, s),
0.88 (9H, s), 1.40 (1H, brs), 1.57 (3H, s), 1.58 (3H, s),
1.60 (3H, s), 1.61–1.66 (2H, m), 1.94–2.00 (4H, m),
2.01–2.09 (6H, m), 3.60 (2H, t, J=6.5 Hz), 4.17 (2H, d,
J=5.7 Hz), 5.07–5.13 (2H, m), 5.27–5.29 (1H, m), 13C
NMR (150 MHz, CDCl3) d ꢀ5.05, 15.84, 15.94, 16.34,
18.42, 26.00, 26.25, 26.48, 30.70, 35.97, 39.51, 39.59,
60.34, 62.79, 124.09, 124.35, 124.73, 134.58, 134.99,
136.86; HRMS (MALDI-FTMS) calcd for C23H44SiO2
(M+Na+) 403.3008; found 403.3000.
Benzyl ether (10). Benzyl bromide (0.86 mL, 7.27 mmol)
was added to a solution of 9 (1.393 g, 3.65 mmol),
sodium hydride (219 mg, 60% in a mineral oil,
5.47 mmol) and n-tetrabutylammonium iodide (67 mg,
0.18 mmol) in a 4:1 mixture of THF and DMF at 0 ꢁC.
After 20 min the mixture was warmed to room tem-
perature and stirred for 10 h. Satd NH4Cl was added
and the mixture was extracted with Et2O. The combined
extracts were dried over MgSO4 and evaporated. The
residue was purified by column chromatography
(EtOAc/hexane, 1:20) to afford the corresponding ben-
zyl ether that was contaminated with benzyl bromide.
Benzyl ether (13). Lithium triethylborohydride (super
hydrideTM) (5.34 mL, 5.34 mmol) was added portion-
wise over 2 h to a solution of 12 (2.02 g, 2.43 mmol) and
bis(diphenylphosphino)propanepalladium(II)dichloride
[(dppp)PdCl2)]44 (72 mg, 0.122 mmol) in THF (26 mL)
at 0 ꢁC. After the mixture was stirred for 12 h, satd
NH4Cl was added and the mixture was extracted with
Et2O. The combined extracts were dried over MgSO4
and concentrated under reduced pressure. The residue
was purified by column chromatography (EtOAc/hex-
ane, 1:40) to afford 1.413 g (84%) of 13 as a colorless
The crude benzyl ether was stirred with n-tetrabutyl-
ammonium fluoride (5.5 mL, 1.0 M in THF, 5.5 mmol)
in THF (8 mL) at room temperature After 12 h, the
mixture was diluted with H2O and extracted with Et2O.
The combined extracts were dried over MgSO4 and
evaporated. The residue was purified by column chro-
matography (EtOAc/hexane, 1:8!1:5) to give 1.284 g
(98%) of 10 over two steps. 1H NMR (400 MHz,
CDCl3) d 1.58 (6H, s), 1.65 (3H, s), 1.68–1.72 (2H, m),
1.94–2.11 (10H, m), 3.43 (2H, t, J=6.4 Hz), 4.11 (2H, d,
J=6.1Hz), 4.48 (2H, s), 5.07–5.11 (2H, m), 5.36–5.42
(1H, m), 7.24–7.28 (5H, m); 13C NMR (100 MHz,
CDCl3) d 15.85, 15.92, 16.20, 26.19, 26.45, 27.92, 35.94,
39.46, 39.55, 59.22, 69.94, 72.79, 123.36, 123.77, 124.39,
127.41, 127.56, 128.26, 134.31, 135.17, 138.52, 139.40;
HRMS (MALDI-FTMS) calcd for C24H36O2
(M+Na+) 379.2613; found 379.2610.
1
oil. H NMR (600 MHz, CDCl3) d 1.05 (9H, s), 1.54
(3H, s), 1.56 (9H, s), 1.59–1.73 (4H, m), 1.96–2.08 (16H,
m), 3.45 (2H, t, J=6.6 Hz), 3.65 (2H, t, J=6.6 Hz), 4.49
(2H, m), 5.12–5.13 (4H, m), 7.24–7.41 (11H, m), 7.66–
7.68 (4H, m); 13C NMR (150 MHz, CDCl3) d 15.89,
15.96, 16.00, 16.04, 19.21, 26.65, 26.69, 26.86, 28.02,
28.25, 28.28, 30.97, 35.81, 36.04, 39.68, 39.75, 63.60,
70.04, 72.86, 124.25, 124.29, 124.38, 124.58, 127.44,
127.55, 127.59, 128.31, 129.45, 134.15, 134.32, 134.76,
134.84, 135.10, 135.56, 138.70; HRMS (MALDI-
FTMS) calcd for C47H66O2Si(M+Na+) 713.4730;
found 713.4721.
Alcohol (14). Lithium naphthalide solution (0.3 M) was
prepared as follows. Granulated lithium (51mg,
7.34 mmol) was immersed in MeOH to clean the surface,
Sulfone (12). To a solution of 10 (1.284 g, 3.06 mmol),
imidazole (368 mg, 5.41mmol) and triphenylphosphine

G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
1257
rinsed with anhydrous THF, and then added to a solu-
tion of naphthalene (926 mg, 7.23 mmol) in THF
(24 mL) under Ar at room temperature. The mixture
was stirred vigorously for 3 h resulting in a dark green
solution that was immediately used for the following.
for the formation of 17 (81%) from 16. A solution of
sodium chlorite (NaClO2, 64 mg, 80%, 0.566 mmol) and
soduim dihydrogenphosphate (68 mg, 0.566 mmol) in
H2O (1.0 mL) was added to a solution of 17 (38 mg,
0.095 mmol) in t-BuOH (1mL) and 2-methyl-2-butene
(0.5 mL) at room temperature After 30 min, H2O was
added and the mixture was extracted with Et2O. The
combined extracts were dried over MgSO4 and con-
centrated under reduced pressure. The residue was
purified by column chromatography (EtOAc/hexane,
Lithium naphthalide (5 mL, 0.3 M in THF, 1.5 mmol)
was added to a solution of 13 (500 mg, 0.723 mmol) in
THF (5 mL) at ꢀ25 ꢁC. After 30 min, satd NH4Cl was
added and the mixture was extracted with Et2O. The
combined extracts were dried over MgSO4 and con-
centrated under reduced pressure. The residue was
purified by column chromatography (EtOAc/hexane,
1:20!1:6) to afford 449 mg (100%) of 14 as a colorless
1
1:3!1:1) to give 1 (27 mg, 68%) as a colorless oil. H
NMR (600 MHz, CDCl3) d 1.22 (2H, s), 1.25 (3H, s),
1.56 (6H, s), 1.59 (6H, s), 1.60–1.66 (2H, m), 1.94–2.07
(13H, m), 2.10–2.14 (1H, m), 2.28 (2H, t, J=8.4 Hz),
2.42 (2H, t, J=8.0 Hz), 2.70 (1H, dd, J=6.1, 6.1 Hz),
5.07–5.16 (4H, m); 13C NMR (150 MHz, CDCl3) d
15.94, 15.96, 15.97, 16.01, 18.70, 24.84, 26.58, 26.61,
27.37, 28.02, 28.18, 32.89, 34.30, 36.26, 39.61, 39.68,
58.59, 64.30, 124.02, 124.29, 124.91, 125.27, 133.08,
133.91, 134.77, 135.25, 179.00; HRMS (MALDI-
FTMS) calcd for C27H44O3 (M+Na+) 439.3188; found
439.3177.
1
oil. H NMR (500 MHz, CDCl3) d 1.03 (9H, s), 1.55
(3H, s), 1.60 (3H, s), 1.64–1.65 (4H, m), 1.97–2.05 (16H,
m), 3.59–3.64 (4H, m), 5.10–5.14 (4H, m), 7.34–7.42
(6H, m), 7.75–7.77 (4H, m); 13C NMR (150 MHz,
CDCl3) d 15.83, 15.93, 15.95, 16.03, 19.19, 26.54, 26.64,
26.83, 28.22, 28.25, 30.66, 30.92, 35.79, 35.98, 39.62,
39.72, 62.80, 63.58, 124.24, 124.35, 124.37, 124.80,
127.54, 129.45, 134.09, 134.56, 134.74, 134.77, 135.09,
135.54; HRMS (MALDI-FTMS) calcd for C40H60O2Si
(M+Na+) 623.4260; found 623.4258.
Acetate (20). Alcohol 19 was prepared using a literature
procedure.20 The same procedure of conversion of 7 to 8
was used for the formation of 20 (100%) from 19 except
using t-butyldiphenylsilyl chloride instead of t-butyldi-
methylsilyl chloride. 1H NMR (400 MHz, CDCl3) d 1.05
(9H, s), 1.67 (3H, s), 1.68–1.70 (2H, m), 2.04 (3H, s),
2.13 (2H, t, J=7.6 Hz), 3.63–3.67 (2H, m), 4.57 (2H, d,
J=7.3 Hz), 5.32–5.35 (1H, m), 7.35–7.42 (6H, m), 7.65–
7.68 (4H, m); 13C NMR (100 MHz, CDCl3) d 16.35,
19.15, 21.00, 26.80, 30.44, 35.68, 61.30, 63.29, 118.25,
127.56, 129.50, 133.91, 135.50, 142.03, 171.04; HRMS
(MALDI-FTMS) calcd for C25H34O3Si (M+Na+)
433.2175; found 433.2167.
Alcohol (16). Pyridine–sulfur trioxide complex (180 mg,
1.13 mmol) in DMSO (1.5 mL) was added to a solution
of 14 (100 mg, 0.166 mmol) and triethylamine (0.24 mL,
1.72 mmol) in CH2Cl2 at 0 ꢁC. After 2 h, H2O was added
and the mixture was extracted with Et2O. The combined
extracts were dried over MgSO4 and concentrated under
reduced pressure. The residue was purified by column
chromatography (EtOAc/hexane, 1:20) to give 90 mg
(90%) of the aldehyde 14a as a colorless oil.
t-BuLi (1.7 M in pentane) was added dropwise to a solu-
tion of isopropyldiphenyl-sulfonium tetrafluoroborate in
THF at ꢀ78 ꢁC. The above aldehyde in THF was added
and the mixture was stirred for 30 min, before being
quenched with sat. NH4Cl. The combined ethereal
extracts of this mixture were dried over MgSO4 and
concentrated under reduced pressure. The remaining resi-
due was purified by column chromatography to give 15.
Alcohol (21). Acetate 20 (19.0 g, 46.3 mmol) was stirred
with potassium carbonate (640 mg, 4.63 mmol) in
MeOH at room temperature After 12 h, 1 N HCl was
added to neutralized the mixture which was evaporated
under reduced pressure. The residue was diluted with
H2O and then extracted with Et2O. The combined
extracts were dried over MgSO4 and concentrated under
reduced pressure. The residue was purified by column
chromatography (EtOAc/hexane, 1:4) to afford 14.1 g
(83%) of 21 as a colorless oil. 1H NMR (400 MHz,
CDCl3) d 1.04 (9H, s), 1.16 (1H, brs), 1.63 (3H, s), 1.65–
1.70 (2H, m), 2.09 (2H, t, J=8.2 Hz), 3.64 (2H, t,
J=6.4 Hz), 4.09–4.11 (2H, m), 5.35–5.39 (1H, m), 7.34–
7.43 (6H, m), 7.64–7.67 (4H, m); 13C NMR (100 MHz,
CDCl3) d 16.20, 19.18, 26.83, 26.86, 30.56, 35.65, 59.31,
63.35, 123.36, 127.53, 127.56, 129.51, 133.96, 135.53,
139.49; HRMS (MALDI-FTMS) calcd for C23H32O2Si
(M+Na+) 391.2069; found 391.2069.
The silyl ether 15 was stirred with n-tetrabutylammonium
fluoride in THF at room temperature After 12 h, water
was added and the mixture was extracted with Et2O. The
combined extracts were dried over MgSO4 and con-
centrated under reduced pressure. The remaining oil
was purified by column chromatography to afford 16
1
(85% over two steps). H NMR (600 MHz, CDCl3) d
1.22 (3H, s), 1.27 (3H, s), 1.56 (6H, s), 1.58 (6H, s),
1.60–1.66 (4H, m), 1.94–2.06 (15H, m), 2.09–2.14 (1H,
s), 2.67 (1H, t, J=6.6 Hz), 3.59 (2H, t, J=6.2 Hz), 5.07–
5.15 (4H, m); 13C NMR (150 MHz, CDCl3) d 15.86,
15.95, 15.96, 16.01, 18.69, 24.86, 26.58, 26.59, 27.39,
28,12, 28.18, 30.70, 35.97, 36.25, 39.59, 39.69, 58.37,
62.74, 64.20, 124.10, 124.29, 124.75, 124.87, 133.92,
134.72, 134.74, 135.18; HRMS (MALDI-FTMS) calcd
for C27H46O2 (M+Na+) 425.3390; found 425.3392.
Sulfone (23). The same procedure of conversion of 5 to
the corresponding iodide was used for the preparation
of 22 (62%). Compound 22 (5.67 g, 11.86 mmol) was
stirred with benzenesulfonic acid, sodium salt (4.5 g,
ꢁ
27.41mmol) in DMF (50 mL) at 0 C. After 30 min, the
Carboxylic acid (1). The same procedure of conversion
of 14 to the corresponding aldehyde 14a was employed
mixture was warmed to room temperature and H2O
(100 mL) was added. The mixture was extracted with

1258
G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
Et2O and the combined extracts were dried over
MgSO4. Evaporation of solvent followed by column
chromatography gave 5.36 g (92%) of 23 as a colorless
oil. 1H NMR (400 MHz, CDCl3) d 1.03(9H, s), 1.25
(3H, s), 1.51–1.58 (2H, m), 2.04 (2H, t, J=7.6 Hz), 3.58
(2H, t, J=6.1Hz), 3.76 (2H, d, J=7.9 Hz), 5.13–5.17
(1H, m), 7.34–7.41 (6H, m), 7.44–7.48 (2H, m), 7.56–
7.60 (1H, m), 7.62–7.64 (4H, m), 7.81–7.83 (2H, m), 13C
NMR (100 MHz, CDCl3) d 16.04, 19.14, 26.78, 30.51,
35.85, 55.99, 63.19, 110.27, 127.57, 128.45, 128.88,
129.54, 133.46, 133.81, 135.47, 138.49, 146.23; HRMS
(MALDI-FTMS) calcd for C29H36O3SSi (M+Na+)
515.2052; found 515.2067.
22.914 mmol) and p-toluene sulfonic acid (0.073 g,
0.382 mmol). The resulting solution was heated under
reflux until water evolution ceased (3 h). The mixture
was washed with satd Na2CO3 (50 mL), the aqueous
layer was separated and extracted with ethyl acetate
(3ꢂ100 mL). The combined organic layers were dried
(MgSO4), concentrated and purified by column chro-
matography (EtOAc/hexane, 1:2) to give 27 (1.234 g,
70%) as a colorless oil.
Alcohol (28). Dioxolane 27 (0.756 g, 3.284 mmol) in
MeOH (10 mL) was treated with satd K2CO3 (10 mL) at
room temperature under vigorous stirring (1h). The
reaction mixture was diluted with ethyl acetate
(100 mL), the aqueous layer separated and extracted
with ethyl acetate (3ꢂ50 mL). The combined organic
layers were dried over MgSO4, concentrated and puri-
fied by column chromatography (EtOAc/hexane, 2:1) to
give 28 (0.545 g, 96%) as a colorless oil. 1H NMR
(400 MHz, CDCl3) d 1.68 (3H, s), 1.78 (2H, m), 2.14
(2H, m), 3.85 (2H, m), 3.97 (2H, m), 4.15 (2H, d,
J=6.8 Hz), 4.86 (1H, t, J=4.7 Hz), 5.44 (1H, dt, J=1.2/
6.8 Hz); 13C NMR (125 MHz, CDCl3) d 15.94, 31.69,
33.34, 58.64, 64.52, 103.84, 123.59, 137.68; HRMS
(MALDI-FTMS) calcd for C9H16O3 (M+Na+)
195.0992; found 195.0994.
Alcohol (25). t-BuLi (1.7 M in pentane) was added
dropwise to a solution of ethyldiphenylsulfonium tetra-
fluoroborate in THF at ꢀ78 ꢁC. The aldehyde 14a (see
above) in THF was added and the mixture was stirred
for 30 min, before being quenched with satd NH4Cl.
The combined ethereal extracts of this mixture were
dried over MgSO4 and concentrated under reduced
pressure. The remaining residue was purified by column
chromatography to give 24.
The silyl ether 24 was stirred with n-tetrabutyl-
ammonium fluoride in THF at room temperature After
1 2 h, H O was added and the mixture was extracted
2
with Et2O. The combined extracts were dried over
MgSO4 and evaporated. The obtained residue was
purified by column chromatography to afford 25 (1:1
mixture of diastereomers, 21% over two steps). 1H
NMR (600 MHz, CDCl3) d 1.22 (3H, s), 1.24 (3H, s),
1.57 (9H, s), 1.58 (3H, s), 1.63–1.65 (4H, m), 1.94–2.06
(16H, m), 2.58–2.60 (0.5H, m), 2.71–2.72 (0.5H, m),
2.85–2.88 (0.5H, m), 2.99–3.02 (0.5H, m), 3.58–3.60
(2H, m), 5.08–5.15 (4H, m), 13C NMR (150 MHz,
CDCl3) d 13.18, 15.86, 15.90, 15.97, 16.02, 17.67, 26.06,
26.57, 26.58, 28.12, 28.15, 28.16, 28.18, 29.66, 30.51,
30.70, 35.79, 35.98, 36.19, 39.57, 39.59, 39.69, 52.73,
54.73, 56.80, 59.54, 62.76, 124.10, 124.29, 124.33,
124.78, 124.84, 124.94, 133.85, 133.87, 134.73, 134.75,
135.20; HRMS (MALDI-FTMS) calcd for C26H44O2
(M+Na+) 411.3233; found 411.3240.
Bromide (29). Compound 28 (1.249 g, 7.252 mmol) in
CH2Cl2 (60 mL) under argon at ꢀ40 ꢁC was consecutively
treated with triethylamine (2.022 mL, 14.504 mmol) and
mesyl chloride (10.878 mmol, 0.858 mL) and the result-
ing mixture was stirred at ꢀ40 ꢁC for 1h. LiBr
(72.520 mmol, 6.298 g) in THF (120 mL) was added and
the resulting mixture was stirred at room temperature
for 1h generating a white precipitate. Et 2O (100 mL)
and satd NH4Cl (5 mL) were added. The aqueous layer
was separated and extracted with Et2O (3ꢂ100 mL).
The combined organic layers were dried over MgSO4
and concentrated to give 29 (1.703 g) as a yellowish oil,
that was introduced into the next step without further
purification. 1H NMR (400 MHz, CDCl3) d 1.71 (3H, s),
1.75 (2H, m), 2.16 (2H, m), 3.83 (2H, m), 3.94 (2H, m),
3.99 (2H, d, J=8.2 Hz), 4.83 (1H, t, J=4.7 Hz), 5.54
(1H, dt, J=1.2/8.2 Hz); 13C NMR (100 MHz, CDCl3) d
15.89, 29.38, 31.75, 33.52, 64.84, 103.80, 120.57, 142.70.
Carboxylic acid (2). The same procedure of conversion
of 16 to 1 was used for the formation of 2 (1:1 mixture
1
of diastereomers, 83%) from 25. H NMR (600 MHz,
Sulfone (30). Compound 29 (1.703 g, 7.243 mmol) in
DMF (36 mL) under argon at 0 ꢁC was treated with
PhSO2Na (1.784 mL, 10.865 mmol) in one portion and
stirring was continued at 0 ꢁC for 20 min. The reaction
mixture was washed with water (15 mL) and the aqu-
eous layer was extracted with Et2O/hexane (1:1)
(3ꢂ100 mL). followed by column chromatography
(EtOAc/hexane, 1:3) gave 30 (1.719 g, 80%) as a color-
less oil. 1H NMR (400 MHz, CDCl3) d 1.31 (3H, s), 1.67
(2H, m), 2.10 (2H, m), 3.80 (2H, d, J=7.9 Hz), 3.85
(2H, m), 3.96 (2H, m), 4.81(1H, t, J=4.7 Hz), 5.22 (1H,
dt, J=1.2/7.9 Hz), 7.53 (2H, t, J=7.4 Hz), 7.64 (1H, t,
J=7.4 Hz), 7.85 (2H, d, J=7.4 Hz); 13C NMR
(125 MHz, CDCl3) d 16.09, 31.82, 33.71, 55.98, 64.87,
103.75, 110.63, 128.48, 128.95, 133.54, 138.49, 145.61;
HRMS (MALDI-FTMS) calcd for C15H20O4S
(M+Na+) 319.0974; found 319.0973.
CDCl3) d 1.24 (3H, s), 1.25 (3H, s), 1.57 (9H, s), 1.59
(3H, s), 1.60–1.66 (2H, m), 1.94–2.10 (14H, m), 2.26–
2.29 (2H, m), 2.40–2.43 (2H, m), 2.60–2.61(0.5H, m),
2.73–2.74 (0.5H, m), 2.87–2.90 (0.5H, m), 3.03–3.04
(0.5H, m), 5.07–5.16 (4H, m); 13C NMR (150 MHz,
CDCl3) d 13.16, 15.90, 15.94, 15.97, 16.01, 17.65, 26.02,
26.57, 26.60, 28.01, 28.18, 30.49, 32.85, 34.30, 35.81,
36.20, 39.59, 39.62, 39.68, 52.87, 54.86, 56.92, 59.66,
124.02, 124.28, 124.32, 124.87, 124.98, 125.28, 133.07,
133.82, 133.86, 134.76, 134.78, 135.25; HRMS
(MALDI-FTMS) calcd for C26H42O3 (M+Na+)
425.3026; found 425.3033.
Dioxolane (27). Compound 18 (prepared using a litera-
ture procedure20) (1.300 g, 7.638 mmol) in toluene
(125 mL) was treated with ethylene glycol (1.278 mL,

G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
1259
TMS-Enol ether (32). Diisopropylamine (3.222 mL,
22.998 mmol) in THF (22 mL) at ꢀ78 ꢁC was treated
with n-BuLi (9.637 mL, 2.5 M in hexane, 24.093 mmol)
for 15 min, and stirring was continued for another
30 min at 0 ꢁC. The mixture was cooled to ꢀ65 ꢁC, 31
(commercially available) (3.000 g, 21.903 mmol) in THF
(5 mL) was added dropwise and the mixture was stirred
at this temperature for further 2 h. TMSCl (3.336 mL,
26.284 mmol) in THF (5 mL) was added dropwise and
the resulting mixture was allowed to warm to room
temperature over a 90-min period. After addition of
pentane (200 mL) and cold satd Na2CO3 (50 mL), the
aqueous layer was separated and extracted with pentane
(3ꢂ100 mL). The combined organic layers were dried
over NaSO4, concentrated and purified by vacuum dis-
tillation (full oil pump vacuum, bp 42–52 ꢁC) to give 32
(2.795 g, 84%) as a colorless oil.
10:1) to give monosilyl ether 35 (0.516 g, 55%) as a col-
1
orless oil. H NMR (500 MHz, CDCl3) d 0.04 (6H, s),
0.89 (9H, s), 1.58 (3H, s), 1.64 (3H, s), 2.08 (4H, m), 3.96
(2H, s), 3.99 (2H, s), 5.38 (1H, sbr), 5.40 (1H, sbr), 13C
NMR (125 MHz, CDCl3) d ꢀ5.34, 13.61, 15.16, 18.41,
25.89, 68.54, 68.80, 123.98, 125.78, 134.71, 135.00.
Bromide (36). Compound 35 (1.000 g, 3.515 mmol) was
transformed to 36 with the same method used for the
formation of bromide 29. Bromide 36 (1.175 g) was
obtained as a yellow oil that was introduced into the
next step without further purification. 1H NMR
(600 MHz, CDCl3) d 0.05 (6H, s), 0.90 (9H, s), 1.58 (3H,
s), 1.74 (3H, s), 2.08 (4H, m), 3.95 (2H, s), 3.99 (2H, s),
5.36 (1H, sbr), 5.59 (1H, sbr); 13C NMR (100 MHz,
CDCl3): d ꢀ5.32, 13.47, 14.67, 18.41, 25.94, 26.95,
28.27, 41.83, 68.42, 123.38, 131.10, 132.20, 135.08; GC–
MS: [MꢀBr]+=267 (2%).
Analytical data were in accord with the literature
values.23
Compound 37. A solution of trimethysilylpropyne
(1.496 mL, 10.103 mmol) in Et2O (7 mL) under argon
was cooled to ꢀ5 ꢁC and TMEDA (1.677 mL,
11.114 mmol) was added in one portion. To this mixture
a solution of n-buthyl lithium (6.947 mL, 1.6 M in hex-
ane, 11.114 mmol) was added dropwise. and stirring was
continued for 15 min. Now, 36 (1.170 g, 3.368 mmol) in
Et2O (3 mL) was added dropwise(!) and the resulting
mixture was stirred at 0 ꢁC overnight. Addition of satd
NH4Cl (5 mL) followed by extraction with Et2O
(3ꢂ10 mL) gave a combined organic phase that was
dried over MgSO4, concentrated and purified by column
chromatography (EtOAc/hexanes, 1:50) to give 37
Dimethyl ester (33). TMS–enol ether 32 (2.781g,
18.504 mmol) in CH2Cl2 (18.5 mL) under argon at 0 ꢁC
was treated dropwise with TiCl4 (18.504 mL, 1 M in
CH2Cl2, 18.504 mmol) under vigorous stirring. The
resulting gray-green suspension was stirred for further
2 h. Quenching with H2O (18.5 mL) was followed by
extraction with CH2Cl2 (3ꢂ100 mL). The combined
organic layers were dried over MgSO4, concentrated
and purified by column chromatography (EtOAc/hex-
ane, 1:5) to give 33 (1.320 g, 63%) as white crystals.
1
Analytical data were in accord with the literature
values.23
(0.856 g, 67%) as a colorless oil. H NMR (500 MHz,
CDCl3) d 0.06 (6H, s), 0.14 (9H, s), 0.91 (9H, s), 1.54
(3H, s), 1.60 (3H, s), 2.05 (4H, m), 2.19 (2H, t,
J=7.5 Hz), 2.30 (2H, t, J=7.5 Hz), 4.00 (2H, s), 5.19
(1H, sbr), 5.39 (1H, sbr); 13C NMR (125 MHz, CDCl3)
d ꢀ5.27, 0.15, 13.43, 15.82, 18.41, 19.19, 25.96, 27.74,
27.93, 38.68, 68.65, 84.50, 107.35, 124.30, 125.42,
133.65, 134.53; GC–MS: [M]+=378 (4%).
Diol (34). Compound 33 (27.693 mmol, 6.266 g) in
CH2Cl2 (90 mL) under argon at ꢀ78 ꢁC was treated
dropwise with DIBAL (49.231mmol, 73.846 mL, 1.5 M
in toluene). The temperature was slowly raised to
ꢀ10 ꢁC over 2 h when the reaction was quenched with
satd sodium potassium tartrate solution (18.5 mL).
Stirring was continued for 2 h before the aqueous layer
was separated and extracted with CH2Cl2 (3ꢂ300 mL).
The combined organic layers were dried over MgSO4
and concentrated to give 34 (4.593 g, 94%) as a colorless
oil that was introduced into the next step without fur-
Monosubstituted alkine (38). Compound 37 (0.791g,
2.077 mmol) was dissolved in MeOH (10 mL), potas-
sium carbonate (0.432 g, 3.128 mmol) was added and the
resulting mixture was stirred at room temperature
overnight. The mixture was diluted with water (6 mL)
and extracted with hexane (3ꢂ25 mL). The combined
organic layers were dried over MgSO4, concentrated
and purified by column chromatography (EtOAc/hex-
anes 1:50) to give 38 (0.525 g, 82%) as a colorless oil. 1H
NMR (600 MHz, CDCl3) d 0.06 (6H, s), 0.91(9H, s),
1.59 (3H, s), 1.60 (3H, s), 1.94 (1H, sbr), 2.06 (4H, sbr),
2.20 (2H, t, J=7.5 Hz), 2.27 (2H, t, J=7.5 Hz), 4.00
(2H, s), 5.20 (1H, sbr), 5.39 (1H, sbr); 13C NMR
(150 MHz, CDCl3) d ꢀ5.28, 13.45, 15.80, 17.54, 18.41,
25.92, 27.65, 27.87, 38.42, 68.30, 68.61, 84.38, 124.23,
125.45, 133.43, 134.51; GC–MS: [M]+=306 (1%).
1
ther purification. H NMR (250 MHz, CDCl3) d 1.66
(6H, s), 2.11(4H, m), 4.00 (4H, s), 5.41(2H, m), NOE
(CDCl3) d 4.00: 1.66, 5.41; HRMS (MALDI-FTMS)
calcd for C10H18O2 (M+Na+) 193. 1199; found
193.1206.
Mono silyl ether (35). To a suspension of sodium
hydride (0.013 g, 3.270 mmol) in THF (5 mL) under
argon was added dropwise 34 (0.559 g, 3.270 mmol) in
THF (2 mL). After stirring for 45 min, TBS chloride
(0.493 g, 3.270 mmol) was added in one portion and
stirring was continued for 45 min The reaction mixture
was poured into Et2O (50 mL) and washed with 10%
K2CO2 (20 mL). The aqueous layer was separated and
extracted with Et2O (3ꢂ50 mL). The combined organic
layers were dried over Na2SO4, concentrated and puri-
fied by column chromatography (Hex.!EA/MeOH,
Alkinol (39). Compound 38 (0.520 g, 1.696 mmol) in
THF (2 mL) under argon at ꢀ60 ꢁC was treated drop-
wise with n-butyllithium (0.780 mL, 2.5 M in hexane,
1.951 mmol), and the resulting mixture was stirred at
ꢀ60 ꢁC for 15 min before being allowed to warm to

1260
G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
ꢀ25 ꢁC for 1.5 h. Upon renewed cooling to ꢀ60 ꢁC
p-formaldehyde (0.268 g, 8.481mmol) was added in one
portion. The reaction mixture was allowed to warm to
room temperature and was stirred overnight. Addition
of water (2 mL), followed by extration with Et2O
(3ꢂ10 mL) gave a combined organic phase that was
dried over MgSO4, concentrated and purified by column
chromatography (EtOAc/hexanes, 1:4) to give 39
(EtOAc/hexanes, 1:3) to give 42 (0.361g, 91%) as a
1
colorless oil. H NMR (500 MHz, CDCl3) d 0.06 (6H,
s), 0.91 (9H, s), 1.21 (3H, s), 1.26 (2H, m), 1.56 (6H, s),
1.64 (2H, m), 1.95 (2H, m), 2.00–2.10 (8H, m), 2.32 (1H,
m), 2.84 (1H, m), 3.74 (1H, dt, J=3.1/10.4 Hz), 3.85 (2H,
m), 3.95 (2H, m), 4.00 (2H, s), 4.79 (1H, t, J=4.5 Hz),
5.00 (1H, d, J=10.4Hz), 5.11 (1H, t, J=7.6 Hz), 5.23
(1H, dt, J=6.8/7.5 Hz), 5.38 (1H, t, J=6.6 Hz), 5.49 (1H,
dt, J=6.8/7.5 Hz), 7.51(2H, t, J=7.6 Hz), 7.61(1H, t,
J=7.6 Hz), 7.82 (2H, d, J=7.6 Hz); HRMS (MALDI-
FTMS) calcd for C35H56O5SSi (M+Na+) 639.3510;
found 639.3511.
1
(0.372 g, 65%) as a colorless oil. H NMR (600 MHz,
CDCl3) d 0.04 (6H, s), 0.88 (9H, s), 1.57 (3H, s), 1.58
(3H, s), 2.04 (4H, sbr), 2.15 (2H, t, J=7.5 Hz), 2.27 (2H,
t, J=7.5 Hz), 3.98 (2H, s), 4.19 (2H, s), 5.17 (1H, sbr),
5.38 (1H, sbr); 13C NMR (150 MHz, CDCl3) d ꢀ5.37,
13.36, 15.74, 17.81, 18.30, 25.84, 27.57, 27.76, 38.48,
51.02, 68.51, 78.68, 85.81, 124.13, 125.30, 133.52,
134.37; HRMS (MALDI-FTMS) calcd for C20H34O2Si
(M+Na+) 359.2382; found 359.2380.
Key intermediate (43). A solution of Pd(OAc)2 (0.022 g,
0.098 mmol) and dppp (0.040 g, 0.098 mmol) in degassed
THF (2 mL) under argon was strirred at room tem-
perature for 30 min during which a fine solid pre-
cipitates. This mixture was cooled to 0 ꢁC and 42
(0.605 g, 0.981mmol) in THF (5 mL) was added drop-
wise. After stirring for 5 min, a solution of super hy-
drideTM (1.962 mL, 1 M in THF, 1.962 mmol) was
added very slowly (ca. 2 h). The reaction progress was
followed by TLC and stopped immediately upon com-
pletion to prevent significant cleavage of the allylic
TBS–ether. Best results were obtained when the whole
amount of super hydrideTM was kept under argon in a
dry scintillation vial, sealed by a septum. Little portions
were taken out, using a fresh plastic syringe with a new
disposable needle (since the super hydrideTM solution
seemed to be corrosive). It was made sure to empty the
volume of the needle after each injection. The resulting
mixture was stirred at 0 ꢁC for 2 h before being warmed
to room temperature (1h). Quenching was carried out
with sat. NH4Cl (10 mL). The mixture was extracted
with Et2O (3ꢂ30 mL). The combined organic layers
were dried over MgSO4, concentrated and purified by
column chromatography (EtOAc/hexanes, 1:5) to give
43 (0.389 g, 83%) as a colorless oil. 1H NMR (600 MHz,
CDCl3) d 0.06 (6H, s), 0.91 (9H, s), 1.58 (3H, s), 1.59
(3H, s), 1.60 (3H, s), 1.75 (2H, m), 1.97–2.11 (14H, m),
3.84 (2H, m), 3.97 (2H, m), 4.00 (2H, s), 4.85 (1H, t,
J=4.8 Hz), 5.13 (1H, sbr), 5.16 (1H, sbr), 5.39 (2H, sbr);
13C NMR (150 MHz, CDCl3) d ꢀ5.28, 13.43, 16.04,
18.41, 25.95, 27.82, 28.08, 30.28, 31.24, 32.43, 32.72,
33.87, 39.74, 64.82, 68.67, 104.34, 124.17, 124.28,
124.45, 129.86, 130.21, 134.27, 134.39, 135.01; HRMS
(MALDI-FTMS) calcd for C29H52O3Si (M+Na+)
499.3578; found 499.3580.
Allylic alcohol (40). A solution of Red-Al1 (0.808 mL,
65 wt-% in toluene, 2.690 mmol) in toluene (8 mL)
under argon at ꢀ78 ꢁC was treated dropwise with a
solution of 39 (0.283 g, 0.841mmol) in toluene (4 mL).
The resulting mixture was stirred at ꢀ78 ꢁC for 1h and
was allowed to warm to room temperature overnight.
Quenching was carried out with ethyl acetate (1mL)
and MeOH (1mL). A satd sodium–potassium tartrate
solution (5 mL) was added carefully and the resulting
mixture stirred vigorously for 1h. The aqueous layer
was separated and extracted with diethyl ether
(3ꢂ10 mL). The combined organic layers were dried
over magnesium sulfate, concentrated and purified by
column chromatography (EtOAc/hexanes, 1:5) to give
40 (0.198 g, 69%) as a colorless oil. 1H NMR (600 MHz,
CDCl3) d 0.06 (6H, s), 0.91 (9H, s), 1.58 (3H, s), 1.59
(3H, s), 2.04 (6H, m), 2.14 (2H, m), 4.00 (2H, sbr), 4.08
(2H, sbr), 5.15 (1H, sbr), 5.39 (1H, sbr), 5.66 (2H, m);
13C NMR (150 MHz, CDCl3) d ꢀ5.31, 13.41, 15.97,
18.37, 25.88, 27.71, 27.83, 30.70, 39.13, 63.63, 68.58,
124.28, 124.50, 128.94, 132.86, 134.38, 134.56.
Bromide (41). Compound 40 (0.550 g, 1.624 mmol) was
transformed to 41 with the same method used for the
formation of bromide 29. Bromide 41 (0.616 g, 94%)
was obtained as a yellow oil that was introduced into
the next step without further purification. 1H NMR
(500 MHz, CDCl3) d 0.06 (6H, s), 0.91(9H, s), 1.59 (6H,
s), 2.05 (6H, sbr), 2.16 (2H, m), 3.94 (2H, d, J=7.3 Hz),
4.01 (2H, s), 5.14 (1H, sbr), 5.39 (1H, sbr), 5.72 (2H, m);
13C NMR (150 MHz, CDCl3) d ꢀ5.31, 13.42, 15.96,
18.37, 25.91, 27.70, 27.84, 30.53, 33.45, 38.73, 68.57,
124.25, 124.78, 126.30, 134.14, 134.43, 136.15.
Alcohol (44). Silicagel (1.632 g, 2 g/mmol) was added to
a solution of compound 43 (0.389 g, 0.816 mmol) in
THF (16 mL) under argon and the mixture was cooled
to ꢀ50 ꢁC. TBAF (1.036 mL, 1 M in THF, 1.036 mmol)
was added dropwise and the resulting mixture was stir-
red for 1h, before being warmed to room temperature
over 30 min and then heated to 50 ꢁC for 2 h. Upon
addition of satd NaCl (20 mL) at room temperature, the
aqueous layer was separated and extracted with Et2O
(3ꢂ50 mL). The combined organic layers were dried
over MgSO4, concentrated and purified by column
chromatography (EtOAc/hexanes, 1:5) to give 44
Allylic sulfone (42). A solution of sulfone 30 (0.646 g,
2.179 mmol) in THF (8 mL) under argon at ꢀ78 ꢁC was
treated dropwise with n-BuLi (1.318 mL, 1.6 M in hex-
ane, 2.109 mmol) and stirring was continued for 1 h.
Now, bromide 41 (0.244 g, 0.703 mmol) in THF (2 mL)
was added dropwise and the mixture was stirred at
ꢀ78 ꢁC for 10 h before being slowly warmed to room
temperature. Addition of sat. NH4Cl (10 mL) was fol-
lowed by extraction with Et2O (3ꢂ30 mL). The com-
bined organic layers were dried over MgSO4,
concentrated and purified by column chromatography
1
(0.264 g, 89%) as a colorless oil. H NMR (500 MHz,
CDCl3) d 1.52 (3H, s), 1.59 (3H, s), 1.67 (3H, s), 1.75

G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
1261
(2H, m), 1.97–2.13 (14H, m), 3.85 (2H, m), 3.97 (2H,
m), 4.00 (2H, s), 4.85 (1H, t, J=4.8 Hz), 5.13 (1H, sbr),
5.16 (1H, sbr), 5.40 (2H, sbr) 5.42 (1H, sbr); 13C NMR
(125 MHz, CDCl3) d 13.69, 16.07, 27.83, 27.92, 28.09,
31.23, 32.47, 32.72, 33.88, 39.73, 64.85, 69.03, 104.38,
123.99, 124.28, 126.13, 129.94, 130.16, 134.32, 134.86,
135.27; HRMS (MALDI-FTMS) calcd for C23H38O3
(M+Na+) 385.2713; found 385.2701.
(EtOAc/hexanes 1:5) to give 47 (0.006 g, 77%) as a col-
1
orless oil. H NMR (600 MHz, CDCl3) d 1.24 (3H, t,
J=7.1), 1.58 (3H, s), 1.60 (6H, s), 1.95–2.07 (12H, m),
2.27–2.35 (4H, m), 2.36–2.41(2H, m), 2.48–2.52 (2H, m),
4.11 (2H, q, J=7.1 Hz), 5.11 (1H, sbr), 5.16 (1H, sbr), 5.17
(1H, sbr), 5.39 (2H, sbr), 9.15 (1H, s); 13C NMR
(125 MHz, CDCl3) d 14.23, 15.91, 16.01, 16.10, 28.02,
28.19, 28.20, 31.23, 31.84, 32.59, 33.26, 34.69, 39.71, 42.14,
60.18, 124.16, 125.08, 125.20, 129.65, 130.37, 133.07,
133.42, 135.00, 173.47, 202.61; HRMS (MALDI-FTMS)
calcd for C25H40O3 (M+Na+) 411.2869; found 411.2868.
Tricarboxylic ester (45). Alcohol 44 (0.025 g, 0.069 mmol)
in Et2O (1mL) under argon was treated with triphenyl
phosphine (0.054 g, 0.206 mmol) and triethyl methane-
tricarboxylate (TEMT) (0.034 g, 0.166 mmol) and
cooled to 0 ꢁC. A solution of DEAD (0.032 mL, 0.6 M
in Et2O, 0.206 mmol) was added dropwise and the
resulting mixture was stirred at room temperature for
3 h. After addition of 5% NaHCO3 (3 mL) and extrac-
tion with Et2O (3ꢂ10 mL), the combined organic layers
were dried over MgSO4, concentrated and purified by
column chromatography (EtOAc/hexanes 1:5) to give
45 (0.033 g, 83%) as a colorless oil. 1H NMR (600 MHz,
CDCl3) d 1.26 (9H, t, J=7.1), 1.56 (3H, s), 1.57 (3H, s),
1.59 (3H, s), 1.75 (2H, m), 1.95–2.07 (14H, m), 2.96 (2H,
s), 3.84 (2H, m), 3.96 (2H, m), 4.21(6H, q, J=7.1Hz),
4.84 (1H, t, J=4.8 Hz), 5.10 (1H, sbr), 5.16 (1H, sbr),
5.23 (1H, sbr), 5.38 (2H, sbr); 13C NMR (150 MHz,
CDCl3) d 13.86, 16.01, 16.45, 27.81, 28.08, 28.33, 31.24,
32.42, 32.72, 33.87, 39.71, 43.07, 61.87, 62.41, 64.84,
104.33, 123.99, 124.27, 129.89, 129.92, 130.16, 130.19,
134.27, 135.15, 166.50; HRMS (MALDI-FTMS) calcd
for C33H52O8 (M+Na+) 599.3554; found 599.3558.
Epoxide (48). [Ph2(i-prop)S]+BF4ꢀ (0.115 g, 0.386 mmol)
in THF (1mL) at ꢀ78 ꢁC was treated dropwise with
t-BuLi (0.227 mL, 1.7 M in pentane, 0.386 mmol) and
stirring was continued for 1h. Aldehyde 47 (0.030 g,
0.077 mmol) in THF (0.5 mL) was added and the
resulting solution stirred at ꢀ78 ꢁC for further 30 min
The reaction was quenched with brine (3 mL), the aqu-
eous layer was separated and extracted with Et2O
(3ꢂ10 mL). The combined organic layers were dried
over MgSO4, concentrated and purified by column
chromatography (EtOAc/hexanes 1:5) to give 26
1
(0.020 g, 62%) as a colorless oil. H NMR (500 MHz,
CDCl3) d 1.25 (3H, t, J=7.1), 1.26 (3H, s), 1.30 (3H, s),
1.60 (3H, s), 1.66 (6H, s), 1.95–2.21 (16H, m), 2.29 (2H,
m), 2.39 (2H, m), 2.70 (1H, t, J=6.2), 4.12 (2H, q,
J=7.1 Hz), 5.12 (1H, sbr), 5.16 (1H, sbr), 5.17 (1H, sbr),
5.40 (2H, sbr); 13C NMR (100 MHz, CDCl3) d 14.24,
15.91, 15.96, 16.02, 18.73, 24.89, 26.39, 27.42, 28.05,
28.11, 28.20, 29.68, 30.29, 31.25, 32.73, 33.08, 34.69,
39.74, 43.44, 58.33, 60.20, 64.19, 124.14, 124.68, 125.08,
129.84, 130.25, 133.42, 134.21, 135.01, 173.53; HRMS
(MALDI-FTMS) calcd for C28H46O3 (M+Na+)
453.3339; found 453.3344.
Monocarboxylic ester (46). Lithium chloride (0.054 g,
1.284 mmol) in water (0.007 g, 0.428 mmol) was added
to a solution of compound 23 (0.247 g, 0.428 mmol) in
DMSO (0.5 mL) in a pressure bottle. The sealed bottle
was heated to 189 ꢁC for 9 h. After cooling, hexane
(1mL) was added and the aqueous layer extracted with
hexane (3ꢂ50 mL). The combined organic layers were
dried over MgSO4, concentrated and purified by column
chromatography (EtOAc/hexanes, 1:5) to give 46
Carboxylic acid (3). Ethyl ester 48 (0.016 g, 0.037 mmol)
was dissolved in MeOH (1mL) and five droplets of a
1M NaOH were added. The mixture was stirred at
room temperature overnight, diluted with Et2O (10 mL)
and washed with brine (3 mL). The aqueous layer was
separated, neutralized with three droplets of citric acid
and extracted with Et2O (10 mL each). The combined
organic layers were dried over MgSO4, concentrated
and purified by column chromatography (EtOAc/hex-
1
(0.118 g, 64%) as a colorless oil. H NMR (500 MHz,
CDCl3) d 1.25 (3H, t, J=7.1), 1.56 (3H, s), 1.58 (3H, s),
1.59 (3H, s), 1.75 (2H, m), 1.95–2.07 (14H, m), 2.29 (2H,
t, J=7.2), 2.39 (2H, t, J=7.2), 3.85 (2H, m), 3.96 (2H,
m), 4.11 (2H, q, J=7.1Hz), 4.85 (1H, t, J=4.6 Hz), 5.11
(1H, sbr), 5.16 (2H, sbr), 5.40 (2H, sbr); 13C NMR
(100MHz, CDCl3) d 14.24, 15.93, 16.03, 16.06, 28.04,
28.08, 28.20, 31.26, 32.43, 32.73, 33.25, 33.87, 34.69, 39.74,
60.22, 64.84, 104.34, 124.12, 124.27, 125.09, 129.89, 130.19,
133.42, 134.28, 135.04, 173.55; HRMS (MALDI-FTMS)
calcd for C27H44O4 (M+Na+) 455.3132; found 455.3132.
1
anes, 1:5) to give 3 (0.015 g, 99%) as a colorless oil. H
NMR (600 MHz, CDCl3) d 1.27 (3H, s), 1.31 (3H, s),
1.58 (3H, s), 1.56–1.72 (2H, m), 1.61 (3H, s), 1.62 (3H,
s), 1.95–2.12 (13H, m), 2.15 (1H, m), 2.31 (2H, t,
J=7.6 Hz), 2.45 (2H, t, J=7.6 Hz), 2.73 (1H, t, J=6.2),
5.11 (1H, sbr), 5.17 (1H, sbr), 5.18 (1H, sbr), 5.40 (2H,
sbr); 13C NMR (125 MHz, CDCl3) d 15.96, 16.03, 18.73,
24.85, 27.37, 27.98, 28.12, 28.16, 31.19, 32.60, 32.73,
34.37, 36.30, 39.71, 58.65, 64.32, 124.18, 124.78, 125.30,
129.89, 130.26, 133.13, 134.16, 135.03, 177.98; HRMS
(MALDI-FTMS) calcd for C26H42O3 (M+Na+)
425.3026; found 425.3008.
Aldehyde (47). Monocarboxylic ester 46 (0.008 g,
0.002 mmol) was dissolved in acetone (1mL) and one
droplet of water was added. To this solution p-toluene-
sulfonic acid (0.001g) was added and the mixture was
heated to 50 ꢁC for 5 h. After cooling to room tempera-
ture the mixture was diluted with Et2O (10 mL) and
washed with satd NaHCO3 (3 mL). The aqueous layer
was separated and extracted with Et2O (3ꢂ10 mL). The
combined organic layers were dried over MgSO4, con-
centrated and purified by column chromatography
Acknowledgements
Financial support from the National Institutes of
Health (GM-43858, K.D.J.), The Skaggs Institute for

1262
G. T. Kim et al. / Bioorg. Med. Chem. 10 (2002) 1249–1262
Chemical Biology (K.D.J.) and The Scripps Research
Institute (J.H.) is gratefully acknowledged.
16. Mohri, M.; Kinoshita, H.; Inomata, K.; Kotake, H.
Chem. Lett. 1985, 451.
17. Hamada, Y.; Shibata, M.; Sugiura, T.; Kato, S.; Shioiri,
T. J. Org. Chem. 1987, 52, 1252.
18. Corey, E. J.; Lin, K.; Jautelat, M. J. Am. Chem. Soc. 1968,
90, 2724.
References and Notes
19. Bal, B. S.; Childers, W. E., Jr.; Pinnick, H. W. Tetra-
hedron 1981, 37, 2091.
20. Vidari, G.; Dapiaggi, A.; Zanoni, G.; Garlaschelli, L.
Tetrahedron Lett. 1993, 34, 6485.
1. (a) Li, T.; Janda, K. D.; Ashley, J. A.; Lerner, R. A.
Science 1994, 264, 1289. (b) Hasserodt, J.; Janda, K. D.; Ler-
ner, R. A. J. Am. Chem. Soc. 1996, 118, 11654. (c) Li, T.;
Janda, K. D.; Lerner, R. A. Nature 1996, 379, 326. (d) Li, T.;
Lerner, R. A.; Janda, K. D. Acc. Chem. Res. 1997, 30, 115.
2. (a) Hasserodt, J.; Janda, K. D.; Lerner, R. A. J. Am. Chem.
Soc. 1997, 119, 5993. (b) Hasserodt, J.; Janda, K. D. Tetra-
hedron 1997, 53, 11237. (c) Paschall, C.; Hasserodt, J.; Jones,
T.; Lerner, R. A.; Janda, K. D.; Christianson, D. W. Angew.
Chem., Int. Ed. Engl. 1999, 38, 1743.
3. (a) Wendt, K. U.; Schulz, G. E.; Corey, E. J.; Liu, D. R.
Angew. Chem., Int. Ed. 2000, 39, 2812. (b) Abe, I.; Rohmer,
M.; Prestwich, G. D. Chem. Rev. 1993, 93, 2189. (c) Full, C.;
Poralla, K. FEMS Microbiol. Lett. 2000, 183, 221. (d) Herrera,
J. B. R.; Wilson, W. K.; Matsuda, S. P. T. J. Am. Chem. Soc.
2000, 122, 6765. (e) Kushiro, T.; Shibuya, M.; Masuda, K.;
Ebizuka, Y. J. Am. Chem. Soc. 2000, 122, 6816.
4. (a) Ourisson, G.; Nakatani, Y. Tetrahedron 1999, 55, 3183.
(b) Ourisson, G.; Nakatani, Y. Chem. Biol. 1994, 1, 1 1 .
5. (a) Caspi, E. Acc. Chem. Res. 1980, 13, 97. (b) Rohmer, M.;
Bouvier, P.; Ourrisson, G. Proc. Natl. Acad. Sci. U.S.A. 1979,
76, 847.
6. (a) DeSilva, B. S.; Wilson, G. S.; Schowen, R. L. In Che-
mical Immunology; Paul, S., Ed.; Karger: Basel, 2000; Vol. 77,
p 33. (b) Jenson, C.; Jorgensen, W. L. J. Am. Chem. Soc.
1997, 119, 10846.
21. Corey, E. J.; Cheng, H.; Baker, C. H.; Matsuda, S. P. T.;
Li, D.; Song, X. J. Am. Chem. Soc. 1997, 119, 1277.
22. Horita, K.; Tanaka, K.; Yonemitsu, O. Chem. Pharm.
Bull. 1993, 41, 2044.
23. Hirai, K.; Ojima, I. Tetrahedron Lett. 1983, 24, 785.
24. Lindel, T.; Franck, B. Tetrahedron Lett. 1995, 36, 9465.
25. Narasaka, K.; Sakakura, T.; Uchimaru, T.; Guedin-
Vuong, D. J. Am. Chem. Soc. 1984, 106, 2954.
26. McDougal, P. G.; Rico, J. G.; Oh, Y.-I.; Condon, B. D. J.
Org. Chem. 1986, 51, 3388.
27. Corey, E. J.; Kirst, H. A. Tetrahedron Lett. 1968, 48, 5041.
28. Corey, E. J.; Katzenellenbogen, J. A.; Gilman, N. W.;
Roman, S. A.; Erickson, B. W. J. Am. Chem. Soc. 1968, 90, 561 8.
29. (a) Ziessel, R.; Suffert, J.; Youinou, M.-T. J. Org. Chem.
1996, 61, 6535. (b) Gevorgyan, V.; Quan, L. G.; Yamamoto,
Y. J. Org. Chem. 1998, 63, 1244.
30. Fish, P. V. Synth. Commun. 1996, 26, 433.
31. Xiao, L.; Kitazume, T. Tetrahedron: Asymmetry 1997, 8,
3597. See also: Corey: E. J.; Kirst, H.A. Tetrahedron Lett.
1968, 48, 5041.
32. Orita, A.; Watanabe, A.; Otera, J. Chem. Lett. 1997, 10,
1025.
33. Mohri, M.; Kinoshita, H.; Inomata, K.; Kotake, H.;
Takagaki, H.; Yamazaki, K. Chem. Lett. 1986, 1177.
34. Cravotto, G.; Giovenzana, G. B.; Sisti, M.; Palmisano, G.
Tetrahedron 1996, 52, 13007.
7. (a) Hasserodt, J.; Janda, K. D.; Lerner, R. A. J. Am. Chem.
Soc. 2000, 122, 40. (b) Hasserodt, J.; Janda, K. D.; Lerner,
R. A. Bioorg. Med. Chem. 2000, 8, 995.
8. Jencks, W. P. Catalysis in Chemistry and Enzymology;
McGraw-Hill: New York, 1969.
35. Padgett, H. C.; Csendes, I. G.; Rapoport, H. J. Org.
Chem. 1979, 44, 3492.
9. Janda, K. D.; Weinhouse, M. I.; Danon, T.; Pacelli, K. A.;
Schloeder, D. M. J. Am. Chem. Soc. 1991, 113, 5427.
10. (a) Umbreit, M. A.; Sharpless, K. B. J. Am. Chem. Soc.
1977, 99, 5526. (b) Bhalerao, U. T.; Rapoport, H. J. Am.
Chem. Soc. 1971, 93, 4835.
11. (a) Grieco, P. A.; Masaki, Y. J. Org. Chem. 1974, 39,
2135. (b) See also for the use of geranyl phenyl sulfone under
Mitsonubo conditions: Uemoto, K.; Kawahito, A.; Matsu-
shita, N.; Sakamoto, I.; Kaku, H.; Tsunoda, T. Tetrahedron
Lett. 2001, 42, 905.
36. Corey, E. J.; Jautelat, M.; Oppolzer, W. Tetrahedron Lett.
1967, 2325.
37. Fersht, A. Structure and Mechanism in Protein Science;
Freeman: New York, 1999.
38. (a) Clayton, R. B.; van Tamelen, E. E.; Nadeau, R. G. J.
Am. Chem. Soc. 1968, 90, 820. (b) Corey, E. J.; Lin, K.; Jau-
telat, M. J. Am. Chem. Soc. 1968, 90, 2724. (c) Kyler, K.;
Novak, M. J. NATO ASI Ser., Ser. C 1992, 381, 3.
39. Barbas, C. F., III; Heine, A.; Zhong, G.; Hoffmann, T.;
Gramatikova, S.; Bjornestedt, R.; List, B.; Anderson, J.; Stura,
¨
12. Yanagisawa, A.; Hibino, H.; Habaue, S.; Hisada, Y.;
Yamamoto, H. J. Org. Chem. 1992, 57, 6386.
E. A.; Wilson, I. A.; Lerner, R. A. Science 1997, 278, 2085.
40. Tipton, K. F.; Dixon, H. B. F. In Methods in Enzymology;
Academic: San Diego, 1979; Vol. 63, p 183.
13. Biellman, J. F.; Ducep, J. B. Tetrahedron Lett. 1969, 3707.
14. (a) See also other allylic cross-coupling techniques based
on phosphates, phosphonim salts and mesitoates, respectively:
Yanagisawa, A.; Hibino, H.; Nobuyoshi, N.; Hisada, Y.;
Yamamoto, H. J. Am. Chem. Soc. 1993, 115, 5879. (b) Axel-
rod, E. H.; Milne, G. M.; van Tamelen, E. E. J. Am. Chem.
Soc. 1970, 92, 2140. (c) Katzenellenbogen, J. A.; Lenox, R. S.
J. Org. Chem. 1973, 38, 326.
41. (a) Ochs, D.; Tappe, C. H.; Gartner, P.; Kellner, R.; Por-
¨
alla, K. Eur. J. Biochem. 1990, 194, 75. (b) Corey, E. J.;
Cheng, H.; Baker, C. H.; Matsuda, S. P. T.; Li, D.; Song, X. J.
Am. Chem. Soc. 1997, 119, 1277.
42. Wendt, K. U.; Schulz, G. E.; Corey, E. J.; Liu, D. R.
Angew. Chem., Int. Ed. 2000, 39, 2812.
43. Knowles, J. Crit. Rev. Biochem. 1976, 4, 165.
44. Housecroft, C. E.; Shayth, B. A.; Rheingold, A. L.;
Haggerty, B. S. Inorg. Chem. 1991, 30, 125.
15. Liu, H.-J.; Yip, J.; Shia, K.-S. Tetrahedron Lett. 1997, 38,
2253.