Studies on the synthesis of the ABC rings of (±)-hexacyclinic acid

Article abstract of DOI:10.1021/jo901547k

(Chemical Equation Presented) A synthesis of the ABC-rings of the polyketide natural product hexacyclinic acid has been achieved. The B-ring was formed first via an intramolecular ester-tethered Diels-Alder reaction, and the A-ring was annulated to this b

Full text of DOI:10.1021/jo901547k

pubs.acs.org/joc
Studies on the Synthesis of the ABC Rings of (()-Hexacyclinic Acid†
Paul A. Clarke,*,‡,§, Andrew P. Cridland,§ Gabriele A. Rolla,‡ Mudassar Iqbal,‡ Nigel
P. Bainbridge,‡ Adrian C. Whitwood,‡,^ and Claire Wilson§,^,#
‡Department of Chemistry, University of York, Heslington, York, North Yorks YO10 5DD, U.K., and
§School of Chemistry, University of Nottingham, University Park, Nottingham, Notts NG7 2RD, U.K.
Current address for correspondence: University of York. ^Corresponding authors for information of X-ray
crystallographic data. #Current address for correspondence: Rigaku Europe, Unit B6, Chaucer Business
Park, Watery Lane, Kemsing, Sevenoaks, Kent TN15 6QY, U.K.
pac507@york.ac.uk
Received July 18, 2009
A synthesis of the ABC-rings of the polyketide natural product hexacyclinic acid has been achieved.
The B-ring was formed first via an intramolecular ester-tethered Diels-Alder reaction, and the
A-ring was annulated to this by means of a SmI2 mediated reductive 5-exo-trig cyclization of a
samarium-ketyl radical onto a vinyl group. Finally, the C-ring was closed using olefin metathesis.
Interestingly, use of enyne metathesis resulted in the synthesis of a more functionalized 5-membered
C-ring in a model system, but an undesired 6-membered C-ring in the actual system. An investigation
of this change in selectivity is discussed.
Introduction
Hexacyclinic acid was isolated from a strain of Strepto-
myces using the OSMAC (one strain/many compounds
approach) (Figure 1). After extensive optimization, 56 mg
L-1 of a new metabolite was isolated from the organic
extracts after column chromatography using CHCl3/metha-
nol as the mobile phase. Extensive spectroscopic studies
indicated that the structure of this new metabolite was
hexacyclinic acid; this was subsequently confirmed by
X-ray single-crystal analysis.1 The absolute stereochemistry
of hexacyclinic acid was determined by performing advanced
Mosher’s ester methodology on the C16 (X-ray numbering)
hydroxyl on the A-ring. The polyketide nature of hexacycli-
nic acid was confirmed by 13C-labeled feeding experiments.
This showed that seven acetate and four propionate units
made up the carbon skeleton of hexacyclinic acid;1 however,
a definitive biosynthetic pathway was not elucidated. When
tested in three cell lines (HM02, HEPG2 and MCF7),
FIGURE 1. Structure of hexacyclinic acid and FR182877.
hexacyclinic acid showed some cytotoxicity with GI50 values
in the low μM range;1 further biological results have yet to be
reported. While this cytotoxicity is lower than that of the
related compound FR182877,2 since no structure-activity
studies have been reported that compare hexacyclinic acid
and FR182877, it is not known which features of FR182877
give it its increased cytotoxicity over hexacyclinic acid.
(2) (a) Sato, B.; Muramatsu, H.; Miyauchi, M.; Hori, Y.; Takase, S.;
Hino, M.; Hashimoto, S.; Terano, H. J. Antibiot. 2000, 53, 123. (b) Sato, B.;
Nakajima, H.; Hori, Y.; Hino, M.; Hashimoto, S.; Terano, H. J. Antibiot.
2000, 53, 204. (c) Yoshimura, S.; Sato, B.; Kinoshita, T.; Takase, S.; Terano,
H. J. Antibiot. 2000, 53, 615.
† Dedicated to Prof. Richard J. K. Taylor on the occasion of his 60th birthday.
*Corresponding author. Phone: þ44 1904 432614. Fax: þ44 1904 432516.
€
(1) Hofs, R.; Walker, M.; Zeeck, A. Angew. Chem., Int. Ed. 2000, 39,
3258.
7812 J. Org. Chem. 2009, 74, 7812–7821
Published on Web 09/24/2009
DOI: 10.1021/jo901547k
r
2009 American Chemical Society

Clarke et al.
JOCArticle
SCHEME 1. Retrosynthetic Analysis
While the related compound FR182877 has succumbed to
total synthesis by the groups of Sorensen,3 Evans,4 and
Nakada5 and has been the subject of further synthetic
activity,6 hexacyclinic acid has not. To date, only four groups
have published strategies toward fragments of hexacyclinic
acid. The group of Landais has published a free-radical-
mediated 5-exo-trig cyclization to yield a structure which has
similarities to the ABC-rings of hexacyclinic acid.7 The
group of Kalesse has published a route to the ABC-rings of
hexacyclinic acid which lack only the B-ring carboxylic acid
function,8 and the group of Prunet has published a ring-
closing metathesis approach to the A-ring of hexacyclinic
acid9 and a Michael addition/radical cyclization approach to
the functionalized ABC-rings of hexacyclinic acid.10 Our
own endeavors have resulted in the synthesis of the DEF-
rings of both hexacyclinic acid and FR182877 by means of a
transannular iodocyclisation reaction11 and the synthesis of
the AB-rings of hexacyclinic acid and the B-ring of
FR182877 via an ester-tethered Diels-Alder approach.12
In this paper, we disclose fully our studies toward and the
synthesis of the ABC-rings of hexacyclinic acid.
Results and Discussion
Our retrosynthetic strategy is outlined in Scheme 1. Dis-
connection of the lactone and hemiketal units in hexacyclinic
acid 1 provides a functionalized 9-membered ring 2, which
has the potential to cyclize in a transannular fashion to
furnish the DEF-rings of the natural product.11 Retrosyn-
thetic removal of six of the carbons in the 9-membered ring
reveals tricyclic ketone 3 as an ABC-ring subunit target.
Disconnection of the C-ring reveals a functionalized AB-ring
unit 4, the A-ring of which was envisaged to be formed by
means of a reductive 5-exo-trig cyclization of a ketyl radical
onto a double bond 5. The cyclcohexene ring 6 could then be
disconnected back to a simple tethered diene/dienophile 7.
As we wished to keep our options open, we considered the
use of either an alkenic or alkynic dienophile partner tethered
to the diene by means of either an ester or hydroxamate
linkage 7. This in turn could be disconnected back to readily
available simple Diels-Alder precursors 813 and 9.
Regardless of the dienophile, bromodiene 8 was required,
and this was initially prepared from (E)-3-hydroxyprop-1-
enylboronic acid and 3-(benzyloxy)-1,1-dibromoprop-1-ene
in 77% yield.12 However, the poor yields for the synthesis of
(E)-3-hydroxyprop-1-enylboronic acid (<25%) forced us to
resort to a longer route which could ultimately service us
with more material (an overall yield of 51% from cis-butene-
1,4-diol).13 With quantities of 8 in hand, we could now
investigate the tethered Diels-Alder reaction. As previous
studies with ester-tethered fumarates had shown that exo- vs
endo-selectivity and competing sigmatropic rearrangements
were a significant problem,14 we decided to investigate
the use of a hydroxamate tether as expounded by Ishikawa.15
(3) Vosburg, D. A; Vanderwal, C. D; Sorensen, E. J. J. Am. Chem. Soc.
2002, 124, 4552.
(4) Evans, D. A.; Starr, J. T. Angew. Chem., Int. Ed. 2002, 41, 1787.
(5) Tanaka, N.; Suzuki, T.; Matsumura, T.; Hosoya, Y.; Nakada, M.
Angew. Chem., Int. Ed. 2009, 48, 2580.
(6) (a) Armstrong, A.; Goldberg, F. W.; Sandham, D. A. Tetrahedron
Lett. 2001, 42, 4585. (b) Methot, J. L.; Roush, W. R. Org. Lett. 2003, 5, 4223.
(c) Funel, J.-A.; Prunet, J. J. Org. Chem. 2004, 69, 4555.
(7) James, P.; Felpin, F.-X.; Landais, Y.; Schenk, K. J. Org. Chem. 2005,
70, 7985.
(8) (a) Stellfeld, T.; Bhatt, U.; Kalesse, M. Org. Lett. 2004, 6, 3889. (b)
Stelmakh, A.; Stellfeld, T.; Kalesse, M. Org. Lett. 2006, 8, 3485.
(9) Toueg, J.; Prunet, J. Synlett 2006, 2807.
(10) Toueg, J.; Prunet, J. Org. Lett. 2008, 10, 45.
(11) (a) Clarke, P. A.; Grist, M.; Ebden, M.; Wilson, C. Chem. Commun.
2003, 1560. (b) Clarke, P. A.; Grist, M.; Ebden, M. Tetrahedron Lett. 2004,
45, 927. (c) Clarke, P. A.; Grist, M.; Ebden, M.; Wilson, C.; Blake, A. J.
Tetrahedron 2005, 61, 353.
(13) Clarke, P. A.; Rolla, G. A.; Cridland, A. P.; Gill, A. A. Tetrahedron
2007, 63, 9124.
(14) Clarke, P. A.; Davie, R. L.; Peace, S. Tetrahedron 2005, 61, 2335.
(15) Ishikawa, T.; Senzaki, M.; Kadoya, R.; Marimoto, T.; Miyake, M.;
Izawa, M.; Saito, S. J. Am. Chem. Soc. 2001, 123, 4607.
(12) Clarke, P. A.; Cridland, A. P. Org. Lett. 2005, 7, 4221.
J. Org. Chem. Vol. 74, No. 20, 2009 7813

Clarke et al.
SCHEME 3. Manipulation of the Hydroxamate-Tethered
JOCArticle
SCHEME 2. Hydroxamate-Tethered Diels-Alder Reaction
Diels-Alder Adducts
They showed that by the use of N-benzyl hydroxa-
mate tethers the diastereoselectivity of the Diels-Alder
reaction could be controlled to give exclusively the endo-
Diels-Alder adduct, which is required for the synthesis of
the B-ring of 1.
Monoethyl fumarate was treated with oxalyl chloride
in CH2Cl2 with catalytic DMF to generate the acid chloride
which was then reacted with N-benzyl hydroxylamine to
generate 10 in an unoptimized 60% yield. An attempt
was made to form the tethered Diels-Alder precursor via a
Mitsunobu coupling of 8 and 10; however, cyclization
proved too facile under these conditions and a 1:1 mixture
of exo- and endo-Diels-Alder adducts were isolated in a
combined yield of 78% (Scheme 2). While this was not
what was expected, it was decided to use this opportunity
to attempt the synthesis of the ABC-rings of both
hexacyclinic acid and FR182877, as 11 had the correct
configuration for hexacyclinic acid and 12 had the correct
configuration for FR182877, and they could both be
separated by flash column chromatography. Manipula-
tion of 11 and 12 in an effort to form the A-ring was
initially investigated (Scheme 3). This was attempted by
reductive cleavage of the hydroxamate N-O bond by
several methods such as Zn/Cu(OAc)2 in AcOH/H2O,
SmI2 in THF/H2O, and Fe/HCl; however, they all failed
to return anything other than starting material. Reagents
such as SnCl2/HCl, Zn/AcOH, and SmI2 with HMPA/
H2O caused decomposition of the material. The only condi-
tions to return an isolable product of N-O cleavage
were SmI2 in THF, which gave 13 and 14 in 47% and 26%
yields, respectively. It was suspected that the problems with
the reductive N-O cleavage could be traced to the vinyl
bromide present in both 11 and 12, and indeed when this was
replaced in 12 with a methyl group 16, as required for the
synthesis of FR182877, the reduction went smoothly gen-
erating the desired N-O cleaved product in 74% yield.
Reduction of the ethyl ester present in 11, 13, 16, and 17
was investigated in order to install the required vinyl group
of 6. A number of methods including Dibal-H, LiAlH4, and
LiBH4 were tried, but in no cases was any of the desired
alcohol or aldehyde product formed. Oxidation of the
alcohols in 13 and 17 was also attempted by Dess-Martin,
benzyl protecting group of 11 with BCl3 to yield 15 in 94%
yield.
As ester-tethered fumarates had also proved proble-
matic,14 the decision was taken to examine the use of
propiolates as the dienophile in the Diels-Alder cyclization,
as the desired vinyl group functionality (e.g., 6) could then be
added via conjugate addition in a later step (Scheme 4). To
this end, propiolic acid was joined to 8 under Mitsunobu
conditions to give 18 in 88% yield and thermally cyclized in
toluene under reflux to provide Diels-Alder adduct 19 in
80% yield. The vinyl group was introduced by a copper-
catalyzed addition of vinyl Grignard, and this proceeded to
give 20 in a 91% yield. The highly diastereoselective forma-
tion of the two new stereocenters is rationalized due to
conjugate addition and subsequent enolate protonation,
both occurring on the less hindered convex β-face of the
molecule. A cross-metathesis with methyl acrylate using
Grubbs II catalyst generated 21 in 100% yield. Removal of
the benzyl group with BCl3 SMe2 complex proceeded in
82% yield and gave 22, which was converted into 23 by the
action of Ph3P and CBr4 in 78% yield.
3
With 23 in hand, it was now possible to explore the
formation of the C-ring of hexacyclinic acid (Scheme 5).
It was initially envisaged that the reduction of the R,β-
unsaturated ester with SmI2 would generate an enolate which
could be alkylated intramolecularly by the pendant alkyl
bromide to form the C-ring. While precedence for the
cyclization of an enolate, generated in this way, onto an
alkyl bromide was not found, it had been reported by Procter
that enolates of R,β-unsaturated esters generated by the
action of SmI2 could be cyclized onto a pendant ketone
Swern, PDC, SO3 py/Hunig’s base, MnO2, and Pfitzer-
3
Moffat conditions, but in all cases degradation resulted or
the starting material was reisolated. The only modification
which could be carried out successfully was removal of the
7814 J. Org. Chem. Vol. 74, No. 20, 2009

Clarke et al.
JOCArticle
SCHEME 4. Propiolate-Tethered Formation of the B-Ring of
Hexacyclinic Acid
SCHEME 6. Formation of the C-Ring via Ring-Closing Me-
tathesis
SCHEME 7. Formation of the C-Ring by Enyne Ring-Closing
Metathesis
SCHEME 5. Attempts at C-Ring Formation
metathesis to form the C-ring (Scheme 6). To this end, the
benzyl group of 20 was removed using BCl3 SMe2 complex
carbonyl to form cyclopentanols in high yields.16 However,
use of the conditions reported by Procter resulted in the
isolation of the saturated ester 24, with none of the cyclized
product being detected.
A report by Little showed that a cyclopentane ring could
be formed by the cyclization of the enolate of an ester,
formed by the conjugate addition of cyanide anion to an
R,β-unsaturated ester, on to a primary alkyl bromide.17
However, application of these conditions to 23 resulted in
the product of simple conjugate addition being isolated as a
mixture of diastereomers 25.
3
and the resulting alcohol oxidized with Dess-Martin period-
inane to aldehyde 26. Interestingly, there was no sign of
migration of the double bond into conjugation with the
carbonyl group under these conditions. Aldehyde 26 was
reacted with vinylmagnesium bromide to generate 27 as a
mixture of diastereomeric alcohols, which in turn were
cyclized to 28 in 69% yield by ring-closing metathesis using
Grubbs’ II catalyst. The mixture of diastereomers was
inconsequential at this stage as the hydroxyl group would
need to be removed for the completion of the synthesis of 1.
However, at this time it was felt that there was insufficient
functionality on the cyclopentenol C-ring in 28 to allow for
easy elaboration of the DEF-rings, and so an alternative
enyne metathesis was investigated (Scheme 7). Aldehyde 26
was treated with propynylmagnesium bromide to generate
propargylic alcohol 29 as a 2.5: 1 mixture of diastereomers in
76% yield, which was then protected with TESCl and
With these methods failing to secure the formation of the
C-ring it was decided to investigate the use of ring closing
(16) (a) Hutton, T. K.; Muir, K.; Procter, D. J. Org. Lett. 2002, 4, 2345.
(b) Hutton, T. K.; Muir, K.; Procter, D. J. Org. Lett. 2003, 5, 4811.
ꢀ
(17) Little, R. D.; Verhe, R.; Monte, W. T.; Nugent, S.; Dawson, R. J. J.
Org. Chem. 1982, 47, 362.
J. Org. Chem. Vol. 74, No. 20, 2009 7815

Clarke et al.
JOCArticle
SCHEME 8. Removal of the Hydroxyl and Formation of the
C-Ring
SCHEME 9. Possible Mechanistic Pathways for the Formation
of the Metathesis Products
imidazole to give silyl ether 30 quantitatively. Enyne metath-
esis of 30 went smoothly in toluene under reflux with
Grubbs’ II catalyst to yield the BC-ring system 31 in 93%
yield.
catalyst in a regioselective fashion, placing the ruthenium
center closest to the methyl group 38, before then reacting
with the vinyl group 46.
As the C-ring in hexacyclinic acid 1 does not possess a
secondary hydroxyl group, the formation of a C-ring devoid
of this functionality was attempted. It was decided not to
investigate the removal of the hydroxyl group from 31 as it
was anticipated that both ionic and radical-mediated re-
moval could interfere with the double bonds and either
promote reduction of one or both of them or even migration
of the endocyclic double bond around the C-ring. Instead
removal of the hydroxyl from propargylic alcohol 29 was
investigated by use of a Nicholas reaction (Scheme 8). The
acetylene of 29 was complexed with Co2(CO)8 and then
treated with TFA to form the stabilized cation and then with
NaBH4 to reduce it to the methylene group. The Co2(CO)6
It is possible that the single product 31 which arises from
the reaction of 30 does so as the steric bulk of the silyl ether
forces the catalyst to react via the “ene-then-yne” pathway 2.
When the silyl ether is no longer present, 32 then the “yne-
then-ene” pathways can also compete which leads to the
formation of 34 as well as 33. An alternative explanation may
be that the catalyst can coordinate to the silyl ether oxygen19
in 30, which leads to regioselective formation of the alkyne
addition intermediate 37 where the metal center is closest to
the silyl ether, which would give rise to 31. With the silyl ether
removed 32, addition of the catalyst to the alkyne in an
uncontrolled fashion occurs giving rise to both products 33
and 34.
was then decomplexed by oxidation with Fe(NO3)3 9H2O to
3
Due to the low-yielding removal of the hydroxyl group in
the Nicholas reaction and the mixture of enyne metathesis
products arising from the cyclization of 32, it was instead
decided to investigate formation of the A-ring from silyl
ether 31 (Scheme 10). This was to be achieved via reduction
of the lactone and opening of the resulting lactol with 1,2-
ethanedithiol in order to form the dithiolane. However, the
TES ether proved to be very labile under the reaction
conditions, and so the more robust TBS ether 48 was
employed. In this route, the lactone reduction yielded the
lactol 49 in good yield, but opening of the lactol with 1,2-
ethanedithiol and TiCl4 only caused decomposition of 49. As
the initial B-ring to BC-ring to ABC-ring strategy had failed
it was decided to investigate an alternative B-ring to AB-ring
to ABC-ring route, which it was hoped would result in
greater success.
Investigation of the B- to AB- to ABC-ring strategy began
from lactone 20 (Scheme 11). Reduction of lactone 20 with
Dibal-H proceeded smoothly to give 50 as a single diaster-
eomer, the stereochemistry of the lactol stereogenic center
was determined by X-ray crystallography, which also con-
firmed the stereochemical assignment around the cyclohex-
ene ring (see Figure 1 in the Supporting Information). The
lactol 50 was opened to form dithiolane 51 by the action of
1,2-ethanedithiol and TiCl4. If the reaction was worked up
give 32 in 37% yield. The use of BF3 OEt2 and Et3SiH were
3
also investigated in various combinations for the reduction
step, as well as other methods for oxidative decomplexation
such as NMO, (NH4)2Ce(NO3)6 and even O2. Ultimately,
however, the initial conditions proved to be the best.
Although far from satisfactory, this route did provide some
material to study the enyne metathesis reaction. Enyne
metathesis of 32 with Grubbs’ II catalyst generated two
products in a 3:1 ratio which were identified as 33 and 34
respectively. It was surprising and disappointing that con-
stitutional isomer 34 was now being formed under the
reaction conditions.
The formation of 33 arises from one of two possible
reaction pathways (Scheme 9): Pathway 2: the vinyl group
reacts first with the catalyst to give 36 followed by the alkyne
40 in the so-called “ene-then-yne” mechanism. Alternatively,
pathway 3: the alkyne reacts preferentially with the catalyst
in a regioselective fashion, placing the ruthenium center
closest to the 6-membered B-ring 41 before reacting with
the vinyl group 45. The formation of 34 can also arise from
two different reaction pathways. In pathway 1, the vinyl
group in reacts first with the catalyst 36 followed by the
alkyne to give a highly strained metallocycle intermediate 39,
and for this reason this pathway was discounted.18 More
likely, in pathway 4, the alkyne reacts preferentially with the
(19) For co-ordination to a hydroxyl group directing enyne metathesis,
see: Niethe A.; Fischer, D.; Blechert, S. J. Org. Chem., 2008, 73, 3088.
(18) Sashuk, V.; Grela, K. J. Mol. Catal. A Chem. 2006, 257, 59.
7816 J. Org. Chem. Vol. 74, No. 20, 2009

Clarke et al.
JOCArticle
SCHEME 10. Investigation of A-Ring Formation from BC-
Ring 48
FIGURE 2. Felkin-Anh rationalization for the formation of 54
and diagnostic NOE data for lactol 55.
SCHEME 12. Formation of the AB-Rings 59
SCHEME 11. En Route to the AB-Rings
alcohol 54 was formed as the major product in a ratio of 30:1
and in an isolated yield of 80%.
Proof of the configuration of the newly formed stereogenic
center was obtained by removing the dithiolane protecting
group with EtI and CaCO3 in MeCN/H2O under reflux and
allowing lactol formation (Scheme 11). This generated lactol
55, where the relative configurations of the stereogenic
centers was determined by diagnostic NOE data (Figure 2).
With the stereochemistry of the vinyl addition confirmed
as correct it was now possible to continue with the synthesis
(Scheme 12). Allylic alcohol 54 was protected as the TBS
ether 56 to avoid lactone formation upon removal of the
dithiolane. Removal of the ditholane from 56 proved pro-
blematic at first. The conditions employed previously for the
deprotection of 54 had provided enough material for diag-
nostic purposes only and were unsuitable for a multistep
synthesis and so were not initially investigated in the removal
of the dithiolane from 56. Alternative reagents like HgCl2
and NCS/AgNO3 left the dithiolane unit intact. Oxidizing
agents have been reported to remove dithiolanes; however,
use of either DMP,20 IBX, or CAN21 resulted in decomposi-
tion as did Tl(NO3)3. Use of bis(trifluoroacetoxy)iodo-
benzene22 did furnish some of the desired aldehyde 57, but
in less than 20% yield. With this lack of success it was decided
to return to examine the original deprotection conditions.
Reaction was slow with EtI and CaCO3 in MeCN/H2O, and
too soon, hemithioacetal 52 could also be isolated; however,
resubmission of this to the reaction conditions ensured its
conversion into 51. Oxidation with SO3 py complex gave
3
aldehyde 53; interestingly, these were the only conditions
that did not promote isomerization of the double bond into
conjugation with the aldehyde carbonyl. The stage was now
set for the introduction of the vinyl group required for
reductive closure of the A-ring. It was rationalized, using a
Felkin-Anh analysis, that treatment of aldehyde 53 with
vinylmagnesium bromide would furnish a product where the
major diastereomer was the one required for the continua-
tion of the synthesis 54 (Figure 2). Indeed when 53 was
treated with vinylmagnesium bromide at -100 ꢀC, allylic
(20) Langille, N. F.; Dakin, L. A.; Panek, J. S. Org. Lett. 2003, 5, 575.
(21) Ho, T.-L.; Ho, H. C.; Wong, C. M. J. Chem. Soc., Chem. Commun.
1972, 791.
(22) Stork, G.; Zhao, K. Tetrahedron Lett. 1989, 30, 287.
J. Org. Chem. Vol. 74, No. 20, 2009 7817

Clarke et al.
JOCArticle
SCHEME 13. Holzapfel’s SmI2-Promoted Cyclizations to Cy-
clopentanols
while warming increased the rate of the reaction, it also led to
some decomposition of the material. This necessitated pur-
ification via flash column chromatography, which was un-
desirable as the aldehyde 57 was prone to decomposition on
silica gel. Changing EtI for MeI did speed the reaction
slightly,23 but 24 h was still excessive. As heating caused
decomposition it was necessary to find other ways to increase
the rate of reaction and the role of the base was next
investigated. It was decided to employ a base with a greater
affinity for iodine and the use of Tl2CO3 meant the reaction
was over in 3 days and 57 could be isolated in 62% yield with
no need for purification. However, the breakthrough came
with the use of Ag2CO3, which increased the rate of the
reaction to such an extent that it was over in 24 h and
aldehyde 57 could be isolated in 93% yield without the need
for purification.
FIGURE 3. Key NOE data supporting the structure of 58.
With aldehyde 57 in hand, the key reductive A-ring
annulation reaction could be investigated. The first condi-
tions tried were those reported by Holzapfel24 (SmI2,
HMPA, BuOH in THF at -78 ꢀC) for the cyclization of 60
to 61 and 62 to 63 (Scheme 13) which showed that the desired
anti-anti stereochemistry between the already present benzyl
ether functionality and the newly formed methyl group and
the newly formed alcohol could be obtained. It is this relative
stereochemistry which is required for the synthesis of the A-
ring of 1. However, when 57 was subjected to these reaction
conditions the major product 58 was isolated from the
reaction but only in a 23% yield. The structure and stereo-
chemistry of 58 were determined by 2D NMR experiments
and NOE interactions (Figure 3).
As 58 was only formed in 23% yield an improvement was
sought. Attempts to remove the use of HMPA resorted in the
formation of alcohol 65 with no product of cyclization seen25
(Figure 4). When n-BuOH was replaced with bulky non-
coordinating t-BuOH,26 58 was again formed in 23% yield
but dimer 66 was also isolated in 43% yield (Figure 4). As 57
was racemic, 66 was formed as a mixture of meso- and d/l-
isomers, and while they were separable individual assign-
ment was not possible. Mechanistically, the dimers 66 form
FIGURE 4. Alternative products of reductive cyclization.
when 2 equiv of the radical anion react together to form a
new carbon-carbon bond. Formation of dimers in SmI2-
promoted reactions is known25 and is due to the relative
concentrations of the radical anion and SmI2. On this
occasion, a solution of SmI2 was added slowly to a solution
of 57, and the blue coloration of the organometallic reagent
dissipated rapidly, suggesting the SmI2 was being consumed
instantly. The resulting radical anion could not react with a
second equivalent of SmI2 because its concentration was very
low, and instead, the radical anions self-condensed to form a
dimer 66. Inverse addition suppressed the formation of the
dimer, but it was always isolated in small quantities even
when the addition of the aldehyde was gradual. Changing the
proton source to H2O also suppressed the formation of this
undesired dimer 66 (15%) and delivered 58 in 46% yield.
In the t-BuOH-containing reductive cyclization of alde-
hyde 57, the protonation is an ionic intermolecular event
between the organosamarium and t-BuOH. As some alco-
hols and water are known to displace bulk solvent (THF) and
ligate the samarium center in the absence of HMPA27 (where
t-BuOH is not), the use of water as proton donor introduces
the possibility the water is activating the samarium so that
after cyclization the resulting radical is reduced more rapidly
to the anion. This anion can then be protonated in an
(23) Fetizon, M.; Jurion, M. J. Chem. Soc., Chem. Commun. 1972, 382.
ꢀ
(24) Grove, J. J.; Holzapfel, C. W.; Williams, D. B. G. Tetrahedron Lett.,
1996, 37, 1305 and 5817.
(25) Molander, G. A.; McKie, J. A. J. Org. Chem. 1992, 57, 3132.
(26) Chopade, P. R.; Prasad, E.; Flowers, R. A. II. J. Am. Chem. Soc.
2004, 126, 44.
(27) Hasegawa, E.; Curran, D. P. J. Org. Chem. 1993, 58, 5008.
7818 J. Org. Chem. Vol. 74, No. 20, 2009

Clarke et al.
JOCArticle
intramolecular fashion by the water coordinated to the
samarium.
SCHEME 14. Preparation of C-Ring Precursor Substrates
With it established that use of HMPA and H2O led to
improvements in the yield of 58, the next set of conditions
combined both of these additives and led to the formation of
58 in an improved 59% yield with only 6% of dimer 66 being
formed. Despite this improvement in yield it was sought to
increase the yield of 58 further and to this end the use of
lithium halide additives as reported by Flowers were inves-
tigated.28 Flowers had shown that premixing SmI2 and LiBr
produced SmBr2 in situ, which was shown to be a more
powerful reductant than SmI2 alone and similar in strength
to SmI2-HMPA. It was hoped that good yields of 58 could
achieved without the need for the carcinogenic additive
HMPA. Preforming SmBr2 followed by addition of this
material to a solution of 57 and water led to the rapid
consumption of starting material and the isolation of 58 in
only 26% yield from a complex reaction mixture; however,
as there was no trace of dimer formation with this additive
system it was investigated further. Alternatively, if LiBr was
used as an additive, i.e., a solution of 57, H2O, and LiBr was
treated with a solution of SmI2, the desired 58 was isolated as
the major product in 64% yield with a trace of another
diastereomer, which proved too unstable to characterize
unambiguously. It had previously been postulated that water
competed with HMPA to coordinate to the samarium center,
which activated the samarium species and promoted reduc-
tion of the primary radical, which consequently suppressed
dimer formation. When LiBr was the additive this effect was
magnified as no dimer was formed. To coordinate to the
metal center, water was competing with THF, iodide, and
bromide ions, all of which have a much lower affinity for
lanthanides than HMPA. Therefore, it is postulated water
was able to coordinate more effectively with the samarium
atom, thus promoting reduction of the radical to the anion,
which subsequently suppressed dimer 66 formation comple-
tely. With the optimized formation of 58 complete the free
hydroxyl group was silylated quantitatively with TBSOTf to
give 59, the structure of which was confirmed by single-
crystal X-ray analysis (see Figure 2 in the Supporting
Information).
SCHEME 15. Synthesis of the ABC-Rings of Hexacyclinic Acid 1
The annulation of the C-ring onto the AB-ring system
could now be attempted (Scheme 14). The benzyl group was
removed with BCl3 SMe2 in 92% yield, and the resultant
3
alcohol 67 was oxidized quantitatively with Dess-Martin
periodinane to the aldehyde 68. Aldehyde 68 proved to be
relatively unstable so upon isolation it was reacted immedi-
ately with propynylmagnesium bromide to furnish a 4:1
mixture of diastereomeric alcohols 69 and 70, which could
be separated by flash column chromatography to give 69 in
6% yield and 70 in 54% yield. The major diastereomer was
protected as the TMS ether and then subjected to enyne
metathesis under our previous conditions. However, we were
gravely disappointed to find that the product of enyne
metathesis of 71 was 72 which contained the undesired 6-
membered C-ring we had seen previously in the enyne
metathesis of 32. We had hoped that the presence of a silyl
ether would promote formation of the desired 5-membered
ring, as we had observed with enyne 30. Sadly, this was not
the case, and we are currently unable to explain this change in
reactivity, although we postulate that it may be due to
differences in the conformations between 30 and 71, or the
possible co-ordinating effect of the lactone carbonyl in 30
directing the catalyst to the olefin first.29
With enyne metathesis failing to give the desired ring
system it was decided to resort to a simple ring closing olefin
metathesis which had been used in the model compound 27.
To this end 68 was treated with vinylmagnesium bromide to
give 73 (Scheme 15) as a mixture of diastereomers, which
were not separated, in 78% yield. The diastereomeric mix-
ture 73 was subjected to Grubbs’ II catalyst which generated
(28) Miller, R. S.; Sealy, J. M.; Shabangi, M.; Kulman, M. L.; Fuchs, J.
R.; Flowers, R. A. II. J. Am. Chem. Soc. 2000, 122, 7718.
(29) Grubbs, R. H. Personal communication.
J. Org. Chem. Vol. 74, No. 20, 2009 7819

Clarke et al.
JOCArticle
the desired ABC-ring system 74 of hexacyclinic acid in a
respectrable 63% yield.
Calcd for C26H28BrNO5: C, 60.71; H, 5.49; N, 2.72. Found: C,
60.46; H, 5.50; N, 2.90.
(3aR*,5aR*,8aR*,8bR*)-5-Bromo-6-[(triethylsilyl)oxy]-7-(prop-
10-en-20-yl)-3,3a,5a,6-tetrahydro-8aH-indeno[4,5-c]furan-1(8bH)-
one (31). A solution of 30 (8.6 mg, 20.2 μmol) and Grubbs’ II
generation catalyst (5.1 mg, 6.0 μmol) in benzene (2 mL) was
heated under reflux under an ethene atmosphere for 1 h. The
reaction mixture was cooled to room temperature, concentrated
in vacuo, and purified by flash column chromatography (6:1
pentate-Et2O) to give a colorless oil (6.1 mg, 71%). Data
reported for the major diastereoisomer: νmax/cm-1 (film) 2956,
2912, 2876, 1784, 1640, 1376, 1265, 1239, 1170, 1143, 1016, 739;
Conclusion
We have executed a synthesis of the ABC-rings of hex-
acylinic acid, with suitable functionality on the C-ring to
enable the annelation of a DEF-ring precursor unit. En route
to the ABC-rings we have uncovered some subtle effects
governing the selectivities of both the SmI2 mediated cycliza-
tion reactions we used to form the AB-ring and in the ring-
closing enyne metathesis we used to close the C-ring. In the
case of the SmI2 cyclization, we believe that the change in
selectivity is due to the co-ordination of the proton donor to
the Sm metal center. In the case of the metathesis reaction it
would seem that subtle conformational changes, or again co-
ordination of the catalyst to nearby functionality, affects the
course of the reaction. Annulation of this unit and studies
aimed at completing the synthesis of hexacyclinic acid are
underway and will be reported in due course.
δH (500 MHz; CDCl3) 5.99 (1H, s), 5.90 (1H, t, J = 2.5 Hz), 5.29
(1H, s), 5.12(1H, ddd, J = 8.8, 2.2, 1.2Hz), 5.06 (1H, s), 4.55(1H,
t, J = 9.1 Hz), 4.01 (1H, t, J = 9.1, Hz), 3.37 (1H, qt, J = 9.1, 2.5
Hz), 2.78 (1H, dd, J = 12.7, 9.1 Hz), 2.76 (1H, m), 2.58 (1H, dd,
J = 12.7, 10.5Hz), 1.89 (3H, s), 0.94(9H, t, J =7.9 Hz), 0.64 (6H,
q, J = 7.9 Hz) ppm; δC (125 MHz; CDCl3) 176.7, 149.8, 137.4,
126.2, 124.3, 124.3, 115.1, 79.3, 70.8, 57.4, 44.1, 42.8, 39.3, 21.1,
7.2, 6.4 ppm; m/z (ESþ) 447, 449 (1:1, M þ Naþ); found
447.0962, M þ Naþ C20H2979BrNaO3Si requires 447.0967.
(1R*,2R*,3S*,3aR*,4R*,5R*,7aR*)-5-[(Benzyloxy)methyl]-6-
bromo-1-[(tert butyldimethylsilyl)oxy]-2-methyl-4-vinyl-2,3,3a,4,5,
7a-hexahydro-1H-indene-3-ol (58). A mixture of samarium pow-
der (225 mg, 1.50 mmol) and diiodoethane (283 mg, 1.00 mmol)
was evacuated, purged with N2 (ꢀ3), and then treated with dry
degassed THF (1 mL). The mixture was stirred vigorously at rt
(CARE: exotherm and gas evolution) until the solution turned
dark blue. The solution was stirred for 30 min, and then dry
degassed THF (9 mL) was added to form a 0.1 M solution of
SmI2. Nitrogen was bubbled through a solution of HMPA (555
μL, 3.19 mmol) and water (239 μL, 13.3 mmol) in THF (5 mL)
for 1 h. Separately, N2 was bubbled through a solution of
aldehyde 57 in dry THF (15 mL) for 1 h. To the first solution
was added the THF solution of SmI2 via cannula, and the
resulting purple solution was cooled to -78 ꢀC under N2. To
this was added the second THF solution at a rate of 0.5 mL
min-1. The solution was stirred for 5 min then it was opened to
the atmosphere and allowed to warm to rt over 30 min. The
solution was diluted with Et2O (50 mL), then washed with water
(6 ꢀ 20 mL), brine (20 mL), dried (MgSO4), filtered, and
concentrated in vacuo to leave a crude yellow oil (72.7 mg,
overweight). Flash chromatography (gradient [20-50%] Et2O
in pentane) afforded a pale yellow oil 58 (39.5 mg, 59%) and a
yellow oil 66a and 66b (4.1 mg, 6%). Data for 58: νmax/cm-1
(solution; CHCl3) 3607, 3031, 2956, 2928, 2857, 1640, 1602,
1258, 1122, 1050, 1028, 835; δH (500 MHz, CDCl3) 7.36-7.28
(5H, m), 6.28 (1H, dd, J = 5.0, 2.5 Hz), 5.58 (1H, ddd, J = 17.0,
9.5, 9.5 Hz), 5.15 (1H, dd, J = 9.5, 1.0 Hz), 5.09 (1H, dd, J =
17.0, 1.0 Hz), 4.55 (1H, d, J = 12.0 Hz), 4.47 (1H, d, J = 12.0
Hz), 3.84 (1H, dd, J = 9.5, 2.5 Hz), 3.60 (1H, dd, J = 9.5, 2.5
Hz), 3.41 (1H, dd, J = 9.0, 4.0 Hz), 3.35 (1H, dd, J = 9.0, 9.0
Hz), 2.57 (1H, dddd, J = 9.0, 9.0, 5.0, 2.5 Hz), 2.42 (1H, ddd, J
= 10.5, 9.5, 9.5 Hz), 2.28 (1H, ddddd, J = 9.5, 2.5, 2.5, 2.5, 2.5
Hz), 1.85 (1H, ddd, J = 10.5, 9.0, 4.0 Hz), 1.75 (1H, ddq, J =
9.0, 9.0, 7.0 Hz), 1.07 (3H, d, J = 7.0 Hz), 0.91 (9H, s), 0.08 (3H,
s), 0.04 (3H, s) ppm; δC (125 MHz, CDCl3) 140.1, 138.6, 131.6,
128.2, 127.6, 127.5, 125.1, 117.7, 81.3, 80.6, 73.1, 69.1, 50.4, 47.5,
46.3, 46.0, 44.6, 25.8, 17.9, 15.6, -4.1, -4.2 ppm; m/z (CI)
524.2187 (M þ NH4þ, C26H4379BrNO3Si requires 524.2190),
531 and 529 (1:1, M þ Naþ), 491 and 489 (1:1, Mþ - OH), 387
and 385 (1:1, Mþ - CH2OCH2C6H5).
Experimental Section
Ethyl (4aS*,5R*,6R*,8aR*)-3-Benzyl-6-[(benzyloxy)methyl]-
7-bromo-3,4,4a,5,6,8a-hexahydro-4-oxo-1H-benzo[1,2]oxazine-
5-carboxylate (11) and Ethyl (4aS*,5R*,6S*,8aS*)-3-Benzyl-6-
[(benzyloxy)methyl]-7-bromo-3,4,4a,5, 6,8a-hexahydro-4-oxo-
1H-benzo[1,2]oxazine-5-carboxylate (12). To a solution of hy-
droxamic acid 10 (238 mg, 0.956 mmol), bromodiene 8 (226 mg,
0.797 mmol), and PPh3 (313 mg, 1.20 mmol) in dry THF (20 mL)
at 0 ꢀC under N2 was added TBAD (367 mg, 1.59 mmol), and the
resulting orange-brown solution was allowed to warm to rt over
1 h and then stirred at rt for 72 h. Concentration in vacuo left a
viscous brown oil that was purified immediately. Flash chro-
matography (hexane-Et2O, 1:1) gave a pale yellow oil 12 (173
mg, 42%) and a yellow oil 11 (153 mg, 37%). Data for 12: νmax
/
cm-1 (film) 3063, 3031, 2979, 2927, 2869, 1727, 1645, 1453, 1367,
1269, 1160, 1107; δH (400 MHz, CDCl3) 7.37-7.27 (10H, m),
5.96 (1H, dd, J = 2.0, 1.0 Hz), 5.02 (1H, d, J = 15.0 Hz), 4.49
(1H, d, J = 12.0 Hz), 4.43 (1H, d, J = 15.0 Hz), 4.42 (1H, d, J =
12.0 Hz), 4.18 (1H, dq, J = 10.5, 7.0 Hz), 4.06 (1H, dq, J = 10.5,
7.0 Hz), 3.97 (1H, dd, J = 9.5, 9.5 Hz), 3.73 (1H, dd, J = 10.0,
6.0 Hz), 3.58 (1H, dd, J = 10.0, 2.0 Hz), 3.55 (1H, dd, J = 9.5,
9.5 Hz), 3.41 (1H, dd, J = 12.0, 12.0 Hz), 3.10 (1H, dd, J = 12.0,
6.0 Hz), 3.02 (1H, ddddd, J = 6.0, 6.0, 2.0, 2.0, 1.0 Hz), 2.54 (1H,
ddddd, J = 12.0, 9.5, 9.5, 2.0, 2.0 Hz), 1.24 (3H, t, J = 7.0 Hz)
ppm; δC (100 MHz, CDCl3) 172.8, 171.7, 138.0, 136.1, 128.7,
128.4, 128.2, 128.2, 127.9, 127.5, 127.5, 123.1, 73.1, 72.6,
68.3, 60.8, 49.7, 45.9, 42.7, 40.3, 37.0, 14.0 ppm; m/z (CI) 485
and 483 (1:1, Mþ - CH2O), 470 and 468 (1:1, Mþ - OC2H5),
434 (Mþ - Br), 424 and 422 (1:1, Mþ - CH2C6H5), 394 and 392
(1:1, Mþ - CH2OCH2C6H5), 378 and 376 (1:1, Mþ - CH2O -
OCH2C6H5); found 514.1228, M þ Hþ, C26H2979BrNO5 re-
quires 514.1224. Data for 11: νmax/cm-1 2982, 2924, 2870, 1724,
1643, 1454, 1368, 1305, 1105; δH (500 MHz, CDCl3) 7.38-7.27
(10H, m), 6.01 (1H, dd, J = 4.0, 1.5 Hz), 4.91 (1H, d, J = 15.0
Hz), 4.47 (1H, d, J = 11.5 Hz), 4.47 (1H, d, J = 15.0 Hz), 4.42
(1H, d, J = 11.5 Hz), 4.16 (2H, q, J = 7.0 Hz), 4.05 (1H, dd,
J = 11.5, 4.0 Hz), 3.92 (1H, dd, J = 3.5, 3.5 Hz), 3.91 (1H, dd,
J = 11.5, 4.0 Hz), 3.68 (1H, dd, J = 9.5, 3.5 Hz), 3.34 (1H, dd,
J = 8.0, 3.5 Hz), 3.33 (1H, dd, J = 9.5, 8.5 Hz), 3.22 (1H, ddddd,
J = 8.5, 3.5, 3.5, 1.5, 1.5 Hz), 3.01 (1H, ddddd, J = 8.0, 4.0, 4.0,
4.0, 1.5 Hz), 1.25 (1H, t, J = 7.0 Hz) ppm; δC (100 MHz, CDCl3)
173.1, 167.4, 138.1, 135.8, 129.5, 128.5, 128.3, 128.3, 127.8,
127.7, 127.5, 124.2, 73.0, 72.7, 69.9, 61.4, 50.6, 44.3, 42.6, 37.3,
36.2, 14.1 ppm; m/z (ES) 538 and 536 (1:1, M þ Naþ), 516 and
514 (1:1, M þ Hþ), 408 and 406 (1:1, Mþ - OCH2C6H5). Anal.
A mixture of dimer isomers 66 (9.1 mg, 9.0 μmol) was purified
by preparative TLC (CH2Cl2-Et2O, 9:1) to deliver a pale yellow
oil 66a (4.1 mg, 45%) and a pale yellow oil 66b (2.3 mg, 25%).
Data for 66a: νmax/cm-1 (film) 3394 br, 2953, 2928, 2856, 1641,
1458, 1255, 1121, 1029, 1003, 837; δH (500 MHz, CDCl3)
7.33-7.27 (5H, m), 6.27 (1H, dd, J = 4.5, 2.0 Hz), 5.54 (1H,
7820 J. Org. Chem. Vol. 74, No. 20, 2009

Clarke et al.
JOCArticle
ddd, J = 17.0, 10.0, 10.0 Hz), 5.14 (1H, d, J = 10.0 Hz), 5.04
(1H, d, J = 17.0 Hz), 4.53 (1H, d, J = 12.0 Hz), 4.46 (1H, d, J =
12.0 Hz), 3.84 (1H, dd, J = 9.5, 2.0 Hz), 3.60-3.56 (2H, m), 3.43
(1H, dd, J = 9.0, 9.0 Hz), 2.60 (1H, dddd, J = 9.0, 9.0, 4.5, 2.0
Hz), 2.39 (1H, ddd, J = 10.0, 9.0, 9.0 Hz), 2.25 (1H, ddddd, J =
9.0, 2.0, 2.0, 2.0, 2.0 Hz), 1.91-1.79 (2H, m), 1.71 (1H, br ddd,
J = 7.0, 7.0, 7.0 Hz), 1.28 (1H, m), 0.89 (9H, s), 0.06 (3H, s), 0.01
(3H, s) ppm; δC (125 MHz, CDCl3) 139.9, 138.6, 131.6, 128.2,
127.6, 127.4, 125.3, 118.0, 81.4, 79.2, 73.1, 68.9, 55.5, 47.2, 46.6,
46.6, 44.0, 30.2, 25.9, 17.9, -4.1, -4.1 ppm; m/z (ES) 1033 (M þ
Naþ), 1037, 1035, and 1033 (1:2:1, M þ Naþ), 1032, 1030, and
1028 (1:2:1, M þ NH4þ), 1015, 1013, and 1011 (1:2:1, M þ Hþ);
found 1033.3455, M þ Naþ, C52H7679Br2NaO6Si2 requires
1033.3439. Data for 66b: νmax/cm-1 (film) 3369 br, 2953, 2928,
2857, 1640, 1456, 1255, 1119, 1028, 1003, 837; δH (500 MHz,
CDCl3) 7.34-7.25 (5H, m), 6.27 (1H, dd, J = 5.0, 2.5 Hz), 5.56
(1H, ddd, J = 17.0, 9.5, 9.5 Hz), 5.16 (1H, dd, J = 9.5, 1.0 Hz),
5.07 (1H, dd, J = 17.0, 1.0 Hz), 4.54 (1H, d, J = 12.0 Hz), 4.47
(1H, d, J = 12.0 Hz), 3.85 (1H, dd, J = 9.5, 2.5 Hz), 3.61-3.59
(2H, m), 3.43 (1H, dd, J = 9.0, 9.0 Hz), 2.59 (1H, dddd, J = 9.0,
9.0, 5.0, 2.0 Hz), 2.41 (1H, ddd, J = 11.0, 9.5, 9.5 Hz), 2.26 (1H,
ddddd, J = 9.5, 2.5, 2.5, 2.5, 2.5 Hz), 1.83 (1H, ddd, J = 11.0,
9.0, 3.0 Hz), 1.78-1.72 (2H, m), 1.40 (1H, m), 0.89 (9H, s), 0.06
(3H, s), 0.00 (3H, s) ppm; δC (125 MHz, CDCl3) 140.0, 138.7,
131.6, 128.2, 127.5, 127.4, 125.3, 118.0, 81.0, 78.7, 73.1, 69.0,
55.9, 47.2, 46.6, 46.6, 43.9, 29.3, 25.9, 17.9, -4.1, -4.1 ppm; m/z
(ES) 1037, 1035, and 1033 (1:2:1, M þ Naþ), 1032, 1030, and
1028 (1:2:1, M þ NH4þ), 1015, 1013, and 1011 (1:2:1, M þ Hþ);
found 1033.3440, M þ Naþ, C52H7679Br2NaO6Si2 requires
1033.3439.
afforded a clear oil 73 (54.2 mg, 78%) as an inseparable mixture
of diastereoisomers, which were used without further purifica-
tion. Nitrogen gas was bubbled through a solution of precursor
73 (26.7 mg, 0.048 mmol) in dry CH2Cl2 (2.5 mL) for 1 h. After
this time, Grubbs’ II catalyst (4 mg, 0.0048 mmol) was added,
and reaction was stirred for 2 h at room temperature. Once the
reaction had gone to completion (TLC), the solution was filtered
through a plug of silica eluted with diethyl ether and concen-
trated in vacuo to yield a dark brown oil. Flash column
chromatography (pentane/diethyl ether 15:1) gave a product
as a brown oil 74 (16.2 mg, 63%): νmax/cm-1 (film) 2957, 2930,
3887, 2857, 1472, 1463, 1361, 1259, 1006; δH (500 MHz, CDCl3)
6.04-6.03 (2H, m), 5.86 (1H, ddd, J = 4.5, 2.5, 1.5 Hz), 4.75
(1H, brd, J = 8.8 Hz), 3.51 (1H, dd, J = 8.0, 3.5 Hz), 3.32 (1H,
dd, J = 9.5, 9.5 Hz), 2.70 (1H, dddd, J = 9.5, 9.5, 2.8, 2.8 Hz),
2.55 (1H, dddd, J = 8.8, 5.5, 2.8, 2.5 Hz), 2.29 (1H, m), 2.09
(1H, m), 1.82 (1H, ddq, J = 9.5, 8.0, 6.5 Hz), 1.26 (1H, s), 1.04
(3H, d, J = 6.5 Hz), 0.91 (9H, s), 0.88 (9H, s), 0.11(3H, s), 0.08
(6H, s), 0.03 (3H, s) ppm; δC (100 MHz, CDCl3) 135.8, 131.9,
130.6, 120.2, 81.2, 81.0, 78.8, 59.4, 51.1, 50.4, 47.9, 25.9, 25.8,
18.0, 17.9, 15.8, -3.7, -3.8, -3.9, -4.1 ppm; m/z (ESIþ) 531
and 529 (1:1, Mþ þ H), 515 and 513 (1:1, Mþ - OH), 401 and
79
399 (1:1, Mþ - OSi(CH3)2C(CH3)3), found 529.2163, C25H46
BrO3Si2 requires 529.2169.
-
Acknowledgment. We thank the EPSRC (award GR/
S77301/01), the Higher Education Commission of Pakistan,
Merck Sharp and DohmeLtd., the University of Nottingham,
the University of York, and the University of York Depart-
ment of Chemistry’s Wild Fund for funding, Prof. D. J.
Procter (University of Manchester) for useful discussions,
and the National Mass Spectrometry Service, Swansea, for
accurate mass determinations.
(1S*,2S*,3R*,3aR*,5aR*,8aR*,8bR*)-5-Bromo-1,3, di(tert-
butyldimethyl silyl)oxy-1,2,3,3a,5a,6,8a,8b-octahydro-2-methy-
lasindacen-13-ol (74). To a solution of aldehyde 68 (66 mg,
0.125 mmol) in dry THF was added a 1 M solution of vinyl-
magnesium bromide in THF (0.37 mL, 0.37 mmol) at -78 ꢀC,
and the mixture was stirred for 2 h. After this the reaction was
quenched with saturated aqueous NH4Cl solution (6 mL) and
the reaction mixture was extracted with ethyl acetate (3 ꢀ 40 mL),
combined organics were washed with brine (50 mL), dried (MgSO4),
filtered and concentrated in vacuo give a brown oil. Flash
column chromatography (petroleum ether/ethyl acetate (9:1))
Supporting Information Available: Experimental proce-
dures and spectroscopic data for all new compounds not
included in the Experimental Section; copies of 1H and 13C
NMR for all new compounds; X-ray crystallographic CIF files
for compounds 50 and 59. This material is available free of
charge via the Internet at http://pubs.acs.org.
J. Org. Chem. Vol. 74, No. 20, 2009 7821