Assessing synthetic strategies: Total syntheses of (±)- neodolabellane-type diterpenoids

Article abstract of DOI:10.1002/chem.200801161

Two strategies, namely a cross-metathesis/ring-closing metathesis and Pd-catalyzed Stille allylation/Nozaki-Hiyama-Kishi coupling, are examined for the preparation of neodolabellane-type diterpenoids 1 and 2. Whereas the first approach possessed synthetic

Full text of DOI:10.1002/chem.200801161

FULL PAPER
DOI: 10.1002/chem.200801161
Assessing Synthetic Strategies:
Total Syntheses of (Æ)-Neodolabellane-Type Diterpenoids
Cory Valente and Michael G. Organ*[a]
Abstract: Two strategies, namely a cross-metathesis/ring-closing metathesis and
Pd-catalyzed Stille allylation/Nozaki–Hiyama–Kishi coupling, are examined for the
Keywords: macrocyclization
·
·
preparation of neodolabellane-type diterpenoids 1 and 2. Whereas the first ap-
metathesis
proach possessed synthetic limitations, the latter was successfully employed to pro-
palladium
·
neodolabellane
vide compounds 1 and 2 in 8.8% (14 steps) and 8% (15 steps) overall yields, re-
spectively.
Introduction
Dolabellanes, and their relatively rare neodolabellane ana-
logues, are a class of bicyclic diterpenes comprised of an 11-
membered macrocycle fused to
a cyclopentyl moiety.
Whereas dolabellanes are isolated from marine and terres-
trial sources alike, neodolabellanes are to date exclusively of
marine origin, specifically corals.[1] Their proposed biosyn-
thesis is initiated with a metal-cation-based enzymatic ioni-
zation of geranylgeranyl diphosphate (Figure 1). A double-
cascade cyclization provides the dolabellane scaffold, fol-
lowed by a series of [1,2]-sigmatropic rearrangements to
give the corresponding neodolabellane. In addition to their
intriguing and varied structures, their cytotoxic, antibacteri-
al, antifungal, antiviral, antimalarial, molluscicidal, ichthyo-
toxic and phytotoxic activities have made them the aim of
many recent formal and total syntheses. Corey and co-work-
ers have recently established a program to develop efficient
strategies for the synthesis of these natural products.[2]
Mehta and co-workers reported the first total synthesis of
a dolabellane diterpenoid in 1990.[3] They prepared (À)-d-
araneosene using as their key step an oxy-Cope-induced
four-carbon annulative ring expansion to generate the dola-
bellane skeleton. Since then, most synthetic routes leading
to dolabellanes involve macrocyclization via intramolecular
Figure 1. Proposed biosynthesis of dolabellanes and neodolabellanes.
alkylations following in situ deprotonation of cyanohy-
drins,[4] b-ketoesters,[5] phosphonates (Horner–Wadsworth–
Emmons reaction)[6] and sulfones (Julia condensation).[7]
Using either d-mannitol or l-ascorbic acid as a chiral start-
ing material, Yamada has published five total syntheses rely-
ing on such alkylation methodology, although 35 steps or
more were required for each.[6,7a,c] Corey and co-workers
have developed several efficient routes relying on ring ex-
pansion and/or contractions to obtain the desired fused bicy-
clic framework. These include reductive pinacol couplings
followed by either a dianion-accelerated oxycope rearrange-
ment[8] or pinacol rearrangement.[2] An enantioselective
Claisen rearrangement[9] and a Diels–Alder macrobicycliza-
tion followed by a ring contraction[10] have also been used.
In total, their group has realized eight total syntheses of six
dolabellanes, all of which use derivatives of trans,trans-farne-
sol as building blocks that are preset with the two requisite
macrocycle E-trisubtituted olefins.
[a] C. Valente, Prof. Dr. M. G. Organ
Department of Chemistry, York University
4700 Keele Street, Toronto, ON, M3J1P3 (Canada)
Fax : (+1)416-736-5936
To date, 17 total syntheses of 11 dolabellanes have been
reported. However, presumably a consequence of their
rarity in nature, there exist only two total syntheses of neo-
dolabellanes, both of which were accomplished by Williams
and co-workers. They have employed both a Julia condensa-
E-mail: organ@yorku.ca
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200801161.
Chem. Eur. J. 2008, 14, 8239 – 8245
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8239

tion[7d] and a reductive pinacol coupling[11] to close the 11-
membered macrocycle, the latter synthesis being the most
highly convergent approach for the preparation of these
molecules to date. We became interested in not only in-
creasing the number of synthetically prepared neodolabel-
lanes but, more importantly, expanding the methodology for
the macrocyclization of this molecule class. As such, we
sought out to prepare neodolabellane-type diterpenoids 1
and 2 whose structures were reported in 2004 after being
isolated from the Okinawan soft coral C. koellikeri.[12] Al-
though compound 2 was not subjected to biological assays,
compound 1 was shown to possess in vitro growth inhibition
against various cancer cell lines.[13]
Results and Discussion
Synthesis from disconnection A: The cross-metathesis ap-
proach
In order to assess the feasibility of the cross-metathesis reac-
tion, we opted to first carry out a model study in which com-
pound 12, prepared in only two steps, would act as a surro-
gate to advanced intermediate 11 (Figure 3). The “metathe-
sis site” of both molecules is identical, with the assumption
that the distal cyclic moieties will have little effect on the
cross-metathesis.
We envisaged two possible routes for the preparation of 1
and 2 (Figure 2), both relying on the manipulation of the
macrocyclic olefins to achieve macrocyclization, either by
ring-closing metathesis (RCM)[14] or Nozaki–Hiyama–Kishi
(NHK) coupling.[15] Although in the latter case the olefin ge-
ometry is pre-defined, it has been shown possible to manip-
ulate the E/Z selectivity in the RCM of 11-membered
rings.[16] Following the retrosynthesis from disconnection A,
a cross-metathesis[17] of 5 with 4 would realize the E-trisub-
stituted olefin[18] as the first key step. The retrosynthesis
from disconnection B relies on the Pd-catalyzed allylation of
a vinyl stannane as the first key step. Both proposed routes
are relatively efficient, and stem from copper-catalyzed con-
jugate addition to commercially-available 2-methyl-2-cyclo-
penten-1-one (6) to set the necessary relative stereochemis-
try on the five-membered ring.[11,19]
Figure 3. Compound 12 as a surrogate for compound 11 in the cross-
metathesis model study.
Oxidation of commercially available 2-cyclohexylethanol
(13) with PDC followed by addition of isopropenylmagnesi-
um bromide to the intermediate aldehyde provided allylic
alcohol 12 (Scheme 1). Silylation with TBDMSCl or oxida-
tion with PCC provided compounds 14 and 15, respectively.
Unfortunately, cross-metathesis of 12, 14, or 15 with 5-
hexen-2-one using Grubbs second-generation N-heterocyclic
carbene–Ru catalyst provided homodimer 16 as the only
product.
Scheme 1. PDC = pyridinium dichromate, TBSCl = tert-butyldimethyl-
silyl chloride, PCC = pyridinium chlorochromate.
Given the predisposition of cross-metathesis toward less
sterically hindered olefins, we opted to join the two frag-
ments via a temporary silicon tether to give silaketal 18 via
addition of Ph2SiCl2 to an equimolar mixture of alcohols 12
and 17 (Scheme 2).[20] The nowintramolecular ring-closing
metathesis proceeded under dilute reaction conditions to
give diol 19 after fluoride-induced cleavage of the silicon
tether. Although establishing proof-of-concept, the requisite
E geometry of the trisubstituted olefin is energetically disfa-
vored when forming nine-membered rings. As such, we ex-
plored relocating the anchor site for the tether in compound
12.
Figure 2. Retrosynthetic analyses of 1 and 2.
8240
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8239 – 8245

Total Syntheses of Diterpenoids
FULL PAPER
retrosynthesis A prompted us to switch focus to retrosynthe-
sis B for the preparation of diterpenoids 1 and 2.
Synthesis from disconnection B: The p-allyl substitution/
NHK coupling approach
Scheme 2. TEA = Triethylamine.
Following Pierꢁs protocol,[19b] TMSCl/HMPA-accelerated
copper-catalyzed conjugate addition[24] of isopropylmagnesi-
um chloride to 2-methyl-2-cyclopenten-1-one (6) followed
by lithium enolate capture with allyl bromide provided race-
mic 27 in good overall yield (Scheme 4).[25] This procedure
was later expanded on by Williams and co-workers to give
ketoaldehyde 28 after ozonolysis.[11] Chemoselective protec-
tion of the aldehyde in the presence of the hindered neo-
pentyl ketone was achieved using Noyoriꢁs conditions to
provide 29;[26] exocyclic olefin 30 was prepared subsequently
via one-carbon homologation of 29.
The required vinyl methyl moiety in compound 18 preju-
diced our choices to three options. Allylsilane 20 (option 1)
and allyl ether 21 (option 2) possess synthetic challenges in
terms of tether detachment post-RCM, namely olefin migra-
tion and deoxygenation, respectively. Replacing the silicon
tether with a vinyl phosphate tether (22, option 3) offers the
intriguing possibility of installing the methyl moiety after
RCM via Ni-mediated cross-coupling of the vinyl phosphate,
a pseudo-halide, with MeMgBr.[21] We therefore explored
the preparation of phosphate 22 (Scheme 3).
Scheme 4. HMPA = hexamethylphosphoric triamide.
We envisioned several routes to manipulate either 27, 29
or 30 to the corresponding allylic alcohol or a derivative
thereof (Figure 4).
Scheme 3. TMU = N,N,N,N-Tetramethylurea.
Figure 4. Possible intermediates en route to the corresponding allylic al-
cohols or derivatives thereof.
Ethynylmagnesium chloride addition to aldehyde 23 fol-
lowed by mercuric triflate/N,N,N,N-tetramethylurea-cata-
lyzed hydration[22] of the propargylic alcohol effectively pro-
vided a-hydroxy methyl ketone 24. Silylation of 24 with
TBDMSCl proceeded as expected to yield 25. The kinetic
potassium enolate of 25 was quenched chemoselectively
with ethyl dichlorophosphate, followed by addition of the
sodium alkoxide of racemic 5-hexen-2-ol (preformed) to
provide 22. Despite reports of successful RCM with vinyl
ethers,[23] cross-metathesis providing homodimer 26 was pre-
dominate in our case, the mass balance being comprised of
a complex mixture of unidentified compounds. The combi-
nation of results thus far obtained for the first key step of
*
Approach A: Elimination of the tertiary alcohol 31
would provide access the corresponding tert-butylallyl
ether, however, addition of tert-butoxymethyllithium[27]
or the less basic tert-butoxymethylcerium chloride[28] to
ketone 27 was unsuccessful, presumably a consequence
of preferred enolization.
*
Approach B: Formation of the vinyl anion is possible via
either the Shapiro reaction[29] or lithium–iodide exchange.
Treatment of the trisylhydrazone[30] 32 with sBuLi and
quenching the intermediate vinyl lithium species with
Chem. Eur. J. 2008, 14, 8239 – 8245
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8241

M. G. Organ and C. Valente
suitable electrophiles (e.g. DMF or methyl chlorofor-
mate) provided the corresponding products in a maxi-
mum 38% yield. Conversely, lithium–iodide exchange of
the vinyl iodide 33 and quenching with methyl chlorofor-
mate provided the a,b-unsaturated ester in 70% yield.[31]
Although Luche reduction[32] should provide the allylic
alcohol, the growing step-count prompted us to explore
alternative routes.
Approach C: Williams and co-workers successfully cou-
pled vinyl triflate 34 with Bu3SnCH2OH[33] to provide the
corresponding allylic alcohol in 67% yield, the mass bal-
ance being the protodetriflated species formed via b-hy-
dride elimination of Bu3SnCH2OH.[11]
the presence of mCPBA led us to propose that epoxide 35b
rearranges to give the transient enol 37 in situ[37b] that imme-
diately undergoes Rubottom-type oxidation[38] with a second
equivalent of mCPBA to provide both observed epimers of
36 via intermediate 38 (Scheme 5). Baeyer–Villiger oxida-
tion to the formate analogue 39 with a third equivalent of
mCPBA, followed by hydrolysis provides access to ketone
29.[39] It may be that the initial rearrangement of 35b to 37
is acid-catalyzed (mCPBA pKa ꢀ7.5, sold commercially as a
mixture with 3-chlorobenzoic acid pKa ꢀ3.8)[40] as a similar
pathway does not prevail in the presence of dimethyldioxir-
ane.
*
*
Approach D: An alternative two-step procedure would
be to form epoxide 35 from 29 via the Corey–Chaykov-
sky reaction.[34] Xu and Sun reported on the successful in-
sertion of methylene into a similar neopentyl ketone
À
using the in situ-generated Me2S+CH2 ylide.[35] Howev-
er, use of either dimethylsulfonium or dimethyloxosulfo-
nium methylide completely failed in our case. Perhaps
the presence of the isopropyl substituent in compound 29
(absent in Xu and Sunꢁs case) forces the five-membered
ring into a less reactive conformation. Although warming
the reaction could potentially overcome this barrier, the
thermal instability of sulfur ylides excludes such an ap-
proach. This conjecture is supported by the fact that exo-
cyclic olefin 30 was quantitatively prepared from 29 at
808C in benzene via the thermally stable phosphonium
ylide.
Scheme 5. Proposed pathway for reversion of 35b to 29.
Ring opening of the mixture of epoxides 35a and 35b
using Lewis acids and/or amine bases (e.g. Al
(OiPr)3,[41]
G
LDA, TMSI[42] or TMSOTf[43]) proved ineffective. However,
in situ generated diethylaluminum 2,2,6,6-tetramethylpiperi-
dide (DATMP)[44] cleanly converted a 1:2.2 mixture of epox-
ides to a ꢀ1:2.3 mixture of allylic alcohol 40 and aldehyde
41 at both room temperature and À788C (Table 2, entries 1
and 2). The ratios suggest that different reaction pathways
exist for each diastereoisomer. We postulate that the six-
membered transition state accessible to the 35a·DATMP co-
ordination complex leading to allylic alcohol 40 is not avail-
able to the 35b·DATMP complex (Scheme 6).[44,45] As such,
the latter undergoes Lewis acid-induced epoxide ring-open-
ing to generate the tertiary carbocation leading to 41 via the
diethylaluminum enol ether intermediate. To help substanti-
ate this theory, isolated epoxide 35a was subjected to the
same reaction conditions. At room temperature, a 3.8:1 se-
lectivity in favor of 40 is realized (Table 2, entry 3), suggest-
ing carbocation formation via epoxide ring opening is still a
competing pathway, however cooling the reaction to À788C
disfavors this pathway exclusively (Table 2, entry 4).[46]
Epoxidation of 30 with 2.5 equivalents of mCPBA provid-
ed the single diastereoisomer 35a in addition to a-hydrox-
yaldehydes a-36 and b-36 and ketone 29 (Table 1). Addition
of mCPBA to purified a,b-36 gave clean conversion to
ketone 29, substantiating it as an intermediate in the forma-
tion of the latter. Conversely, epoxidation with in situ gener-
ated dimethyldioxirane[36] provided a 1:2.2 ratio of diastereo-
isomers 35a and 35b, respectively, however, 35b proved un-
stable.[37a] In addition, epoxide 35a was inert to the addition
of excess mCPBA, whereas 35b underwent full conversion
to ketone 29.
The observed instability of epoxide 35b in conjunction
with the formation of ketone 29 via intermediate a,b-36 in
Table 1. Product distribution following the epoxidation of 30 using either
mCPBA or dimethyldioxirane as the oxidant.
Given the intricacies in the formation, stability and open-
ing of epoxides 35a and 35b, we opted to explore another
route in preparing the allylic alcohol. Both the endocyclic
allylic primary carbonate 9 and the exocyclic allylic secon-
dary carbonate 43 would provide the same Pd–p-allyl com-
plex for cross-coupling. As such, we turned our focus to the
preparation of 42 via allylic oxidation (Scheme 7). SeO2-cat-
alyzed allylic oxidation using tBuOOH as a co-oxidant was
superior to using stoichiometric amounts of SeO2 that led to
Conditions
35a/35b/a,b-36/29
mCPBA (82% combined yield)[a]
Oxone, NaHCO3, acetone/water
(96% combined yield)[b]
4.1:0:2.5:1
1:2.2:0:0
[a] Ratio determined from isolated yields. [b] Ratio determined by
1H NMR spectroscopy.
8242
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8239 – 8245

Total Syntheses of Diterpenoids
FULL PAPER
Table 2. Conditions and ratios of starting materials and product for the
epoxide ring opening mediated by DATMP.
Entry
T
35a/35b[a]
40/41[a]
1
2
3
4
RT
1:2.2
1:2.2
1:0
1:2.3
1:2.3
3.8:1
1:0
À788C
RT
788C
1
1:0
[a] Ratio determined by H NMR spectroscopy.
Scheme 7. IBX = 2-Iodoxybenzoic acid.
Scheme 8. Preparation of coupling partner 10.
Scheme 6. Proposed mechanism for epoxide 35a and 35b ring opening
with diethylaluminum 2,2,6,6-tetramethylpiperidide (DATMP).
Conversion of the 1,3-dioxolane 47 to aldehyde 7 proved
over-oxidation to the a,b-unsaturated enone.[47] Importantly,
this approach provided the required allylic alcohol in a
single synthetic step, a considerable advantage over all the
routes thus far attempted; subsequent treatment with
methyl chloroformate provided 43.
problematic. An assortment of Brønsted acids (i.e., PPTS,
[56]
TsOH, acetic acid, HCl),[54] Lewis acids (InCl3,[55] BiCl3
)
and other reagents (CAN,[57] water/microwave irradiation,[58]
thiourea[59]) were examined in a variety of reaction condi-
tions. In all cases, either complete retention or extensive de-
composition of 47 was observed. Similar reactivity was ob-
served when subjecting 46 to identical reaction conditions,
eliminating the vinyl iodide as the source of the problem.
We postulated that either the product aldehyde was not
stable in the reaction conditions or that the intermediate
oxonium ion was undergoing intramolecular rearrangement
to give the tertiary carbocation concomitant with six-mem-
bered ring formation, which then further reacts to give the
complex mixture of products (Scheme 9). The associated dif-
ficulty in the deprotection of unactivated acetals relative to
that of ketals and activated acetals (i.e., aryl, benzyl) exacer-
bates the problem.[54] The product aldehyde 7 was deemed
stable to the reaction conditions as there was no significant
decomposition observed when further treated to identical
reaction conditions (up to 16 h). Thus, the presumed intra-
molecular rearrangement might be disfavored by either the
addition of water as an intermolecular source of oxonium
ion quench and/or by lowering the reaction temperature
En route to coupling partner 10, vinyl iodide 49 was pre-
pared via zirconium-mediated methylalumination/iodination
of 5-pentyn-1-ol (48) followed by TBDMS protection of the
primary alcohol (Scheme 8).[48] Lithium–iodine exchange
and quenching with tri-n-butyltin chloride provided non-
polar vinyl stannane 50 as an inseparable mixture with un-
identified organostannane impurities. Silyl cleavage with
fluoride provided cross-coupling partner 10 quantitatively.[49]
Isolation of the pure compound was not possible due to its
decomposition on silica gel, thus the mixture was used di-
rectly in the next step without consequence. Regiospecific
Pd-catalyzed allylation of 43 with 10 provided 8 followed by
oxidation with IBX to give 44 (Scheme 7).[50] One carbon
homologation of aldehyde 44 to the terminal alkyne 45 was
achieved using the Bestmann–Ohira reagent (Scheme 8).[51]
Methylation provided 46 and regioselective hydrozircona-
tion provided the trisubstituted olefin 47 after iodine
quench.[52,53]
Chem. Eur. J. 2008, 14, 8239 – 8245
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8243

M. G. Organ and C. Valente
[1] For reviews on dolabellane diterpenes and their synthesis, see: a) M.
Hiersemann, H. Helmboldt, Top. Curr. Chem. 2005, 243, 73–136;
b) A. D. Rodríguez, E. Gonzµlez, C. Ramírez, Tetrahedron 1998, 54,
11683–11729.
[2] E. J. Corey, J. S. Kingsbury, J. Am. Chem. Soc. 2005, 127, 13813–
13815.
[3] a) G. Mehta, N. Krishnamurthy, S. R. Karra, J. Am. Chem. Soc.
1991, 113, 5764–5775; b) G. Mehta, Pure Appl. Chem. 1990, 62,
1263–1268.
Scheme 9. Proposed acid-catalyzed decomposition of 47.
[4] a) N. Kato, A. Higo, X. Wu, H. Takeshita, Heterocycles 1997, 46,
123–127; b) D. R. Williams, P. J. Coleman, C. R. Nevill, L. A. Robin-
son, Tetrahedron Lett. 1993, 34, 7895–7898.
[5] L. Jenny, H.-J. Borschberg, Helv. Chim. Acta 1995, 78, 715–731.
[6] H. Miyaoka, T. Baba, H. Mitome, Y. Yamada, Tetrahedron Lett.
2001, 42, 9233–9236.
[7] a) H. Miyaoka, Y. Isaji, H. Mitome, Y. Yamada, Tetrahedron 2003,
59, 61–75; b) I. R. B. Baldwin, R. J. Whitby, Chem. Commun. 2003,
2786–2787; c) H. Miyaoka, Y. Isaji, Y. Kajiwara, I. Kunimune, Y.
Yamada, Tetrahedron Lett. 1998, 39, 6503–6506; d) D. R. Williams,
P. J. Coleman, Tetrahedron Lett. 1995, 36, 35–38.
[8] E. J. Corey, R. S. Kania, Tetrahedron Lett. 1998, 39, 741–744.
[9] E. J. Corey, R. S. Kania, J. Am. Chem. Soc. 1996, 118, 1229–1230.
[10] S. A. Snyder, E. J. Corey, J. Am. Chem. Soc. 2006, 128, 740–742.
[11] D. R. Williams, R. W. Heidebrecht, J. Am. Chem. Soc. 2003, 125,
1843–1850.
(i.e., room temperature). Indeed, using a 1:1 mixture of
THF and 1m HCl at room temperature provided the target
compound.[60] However, the reaction was slow and typically
halted at around 55% conversion after 72 h. Adding more
acid, water or letting the reaction stir longer was not suc-
cessful, and typically led to further decomposition. Howev-
er, diluting the reaction from the outset greatly accelerated
the deprotection, reducing the reaction to just 10 h
(Scheme 10). As aldehyde 7 could not be isolated from un-
[12] K. Iguchi, T. Fukaya, A. Yasumoto, K. Watanabe, J. Nat. Prod. 2004,
67, 577–583.
[13] Taken from ref. [12]: In vitro growth inhibition against lung cancer
(NCI-H522, GI50 5.2 mgmLÀ1), melanoma (LOX-IMVI, GI50
4.9 mgmLÀ1), stomach cancer (MKN74, GI50 5.2 mgmLÀ1), and cen-
tral nervous system cancer (SF-539 and SNB75, GI50 each
4.9 mgmLÀ1) cells.
Scheme 10.
[14] For reviews on ruthenium-catalyzed metathesis, see: a) A. H. Hov-
eyda, A. R. Zhugralin, Nature 2007, 450, 243–251; b) Y. Schrodi,
R. L. Pederson, Aldrichimica Acta 2007, 40, 45–52; c) R. H. Grubbs,
Tetrahedron 2004, 60, 7117–7140.
[15] a) T. Luker, R. J. Whitby, Tetrahedron Lett. 1996, 37, 7661–7664;
b) During the preparation of neodolabellanes 1 and 2 in our labora-
tory, an NKH reaction to close the macrocycle of a dolabellane was
achieved: J. Brian, M. V. Wingerden, J. M. Ready, J. Am. Chem. Soc.
2006, 128, 7428–7429.
[16] a) K. C. Nicolaou, G. Vassilikogiannakis, T. Montagnon, Angew.
Chem. 2002, 114, 3410–3415; Angew. Chem. Int. Ed. 2002, 41, 3276–
3281; b) G. Vassilikogiannakis, I. Margaros, M. Tofi, Org. Lett. 2004,
6, 205–208.
reacted 47 using standard techniques, the mixture was sub-
jected directly to the intramolecular NHK coupling to pro-
vide neodolabellane 1. Noteworthy, the NHK reaction pro-
vided the desired diastereoisomer, a consequence of the es-
tablished strong conformational preference of the 11-mem-
bered macrocycle in neodolabellanes.[15b,61] IBX-mediated
oxidation of the allylic alcohol (1) cleanly provided neodola-
bellane 2. All spectral data were consistent with those re-
ported for the nature-derived compounds.[12]
[17] For a reviewon cross-metathesis see: S. J. Connon, S. Blechert,
Angew. Chem. 2003, 115, 1944–1968; Angew. Chem. Int. Ed. 2003,
42, 1900–1923.
Conclusion
[18] A. K. Chatterjee, R. H. Grubbs, Org. Lett. 1999, 1, 1751–1753.
[19] a) G. B. Dudley, S. J. Danishefsky, Org. Lett. 2001, 3, 2399–2402;
b) E. Piers, J. Renaud, S. J. Rettig, Synthesis 1998, 590–602; c) An
asymmetric variant has been reported for conjugate addition of iso-
propenyl cuprate to 2-methyl-2-cyclopenten-1-one, however this re-
quires two equivalents of a chiral auxiliary. See: M. Mandal, H.
Yun, G. B. Dudley, S. Lin, D. S. Tan, S. J. Danishefsky, J. Org. Chem.
2005, 70, 10619–10637, and references therein.
[20] For a reviewon disposable tethers, see: D. R. Gauthier, K. S. Zandi,
K. J. Shea, Tetrahedron 1998, 54, 2289–2338.
[21] a) D. L. J. Clive, L. Ou, Tetrahedron Lett. 2002, 43, 4559–4563;
b) A. S. E. Karlstrçm, K. Itami, J.-E. Bäckvall, J. Org. Chem. 1999,
64, 1745–1749; c) T. Hayashi, T. Fujiwa, Y. Okamoto, Y. Katsuro,
M. Kumada, Synthesis 1981, 1001–1003.
In summary, an assessment of synthetic strategies has been
conducted for the preparation of neodolabellanes. Whereas
the CM/RCM sequence (retrosynthesis A) proved unfruit-
ful, the alternate Pd-catalyzed allylation and diastereoselec-
tive NHK coupling (retrosynthesis B) culminated in the
preparation of neodolabellanes 1 and 2 in 8.8% (14 steps)
and 8% (15 steps) overall yields, respectively. This report
constitutes the first reported total syntheses of these com-
pounds possessing this interesting molecular architecture.
[22] M. Nishizawa, M. Skwarczynski, H. Imagawa, T. Sugihara, Chem.
Lett. 2002, 12–13.
Acknowledgement
[23] a) C. F. Sturino, J. C. Y. Wong, Tetrahedron Lett. 1998, 39, 9623–
9626; b) O. Fujimura, G. C. Fu, R. H. Grubbs, J. Org. Chem. 1994,
59, 4029–4031.
We thank ORDCF and NSERC Canada for financial support and Dr.
Howard Hunter, Dr. Debasis Mallik and Prof. Arturo Orellana for their
valued collaborations.
8244
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 8239 – 8245

Total Syntheses of Diterpenoids
FULL PAPER
[24] Y. Horiguchi, S. Matsuzawa, E. Nakamura, I. Kuwajima, Tetrahe-
dron Lett. 1986, 27, 4025–4028.
[44] A. Yasuda, S. Tanaka, K. Osima, H. Yamamoto, H. Nozaki, J. Am.
Chem. Soc. 1974, 96, 6513–6514.
[25] Compounds 1, 2, 8, 27–30, 32, 35, 36, 40–44 and 48–50 were isolated
as racemic mixtures.
[26] T. Tsunoda, M. Suzuki, R. Noyori, Tetrahedron Lett. 1980, 21, 1357–
1358.
[27] E. J. Corey, T. M. Eckrich, Tetrahedron Lett. 1983, 24, 3165–3168.
[28] a) T. Imamoto, Y. Sugiura, N. Takiyama, Tetrahedron Lett. 1984, 25,
4233–4236; b) H.-J. Liu, K.-S. Shia, X. Shang, B.-Y. Zhu, Tetrahe-
dron 1999, 55, 3803–3830.
[29] a) R. H. Shapiro, M. F. Lipton, K. J. Kolonko, R. L. Buswell, L. A.
Capuano, Tetrahedron Lett. 1975, 16, 1811–1814; b) A. R. Chamber-
lin, J. E. Stemke, F. T. Bond, J. Org. Chem. 1978, 43, 147–154.
[30] The trisylhydrazone was prepared from ketone 27 by treatment with
trisylhydazide and 12n HCl in acetonitrile (71% yield). Full experi-
mental details can be found in the Supporting Information.
[31] The vinyl iodide was prepared in four steps from ketoaldehyde 28
following a protocol by Williams and co-workers (ref. [11]).
[32] J. L. Luche, J. Am. Chem. Soc. 1978, 100, 2226–2227.
[33] R. L. Danheiser, K. R. Romines, H. Koyama, S. K. Gee, C. R. John-
son, J. R. Medich, Org. Synth. 1993, 71, 133.
[45] For cyclic syn-eliminations where epoxide opening has substantial
stereoelectronic requirements, see: a) B. M. Trost, M. T. Bogdano-
wicz, J. Am. Chem. Soc. 1973, 95, 5311–5321; b) B. M. Trost, S. Kur-
ozumi, Tetrahedron Lett. 1974, 15, 1929–1932; c) A. C. Cope, J. K.
Heeren, J. Am. Chem. Soc. 1965, 87, 3125–3129.
[46] Whereas aldehyde 41 was not formed on small scale at À788C,
scale-up of DATMP-mediated epoxide 35a ring opening (i.e.,
3.7 mmol of 35a) resulted in ꢀ8% of aldehyde 41 being formed.
Full details are provided in the Supporting Information.
[47] M. A. Umbreit, K. B. Sharpless, J. Am. Chem. Soc. 1977, 99, 5526–
5528.
[48] a) D. E. Van Horn, E.-i. Negishi, J. Am. Chem. Soc. 1978, 100,
2252–2254; b) Procedure adapted from: U. Groth, N. Richter, A.
Kalogerakis, Eur. J. Org. Chem. 2003, 23, 4634–4639.
[49] For an alternate synthesis, see: T. Ichige, S. Kamimura, K. Mayumi,
Y. Sakamoto, S. Terashita, E. Ohteki, N. Kanoh, M. Nakata, Tetrahe-
dron Lett. 2005, 46, 1263–1267.
[50] Conditions for the palladium-catalyzed Stille reaction were adapted
from ref. [10].
[34] a) E. J. Corey, M. Chaykovsky, J. Am. Chem. Soc. 1962, 84, 867–
868; b) E. J. Corey, M. Chaykovsky, J. Am. Chem. Soc. 1965, 87,
1353–1364; c) E. J. Corey, M. Chaykovsky, Org. Synth. 1973, 5, 751.
[35] B. Sun, X. Zu, Tetrahedron Lett. 2005, 46, 8431–8434.
[36] S. E. Denmark, D. C. Forbes, D. S. Hays, J. S. DePue, R. G. Wilde, J.
Org. Chem. 1995, 60, 1391–1407.
[37] a) Upon standing for one week under atmospheric conditions,
1H NMR spectroscopy revealed the absence of epoxide 35b with the
appearance of an array of unidentified decomposition products. Ep-
oxide 35a appeared unaffected. b) It was found that epoxide 35b
slightly decomposes during column chromatography on silica gel.
The 1H NMR spectrum of the isolated product contained a peak cor-
responding to the presence of aldehyde 41. This further demon-
strates the susceptibility of 35b to undergo acid mediated ring open-
ing.
[51] a) S. Ohira, Synth. Commun. 1989, 19, 561–564; b) S. Müller, B. Lie-
pold, G. J. Roth, H.-J. Bestmann, Synlett 1996, 521–522; c) G. J.
Roth, B. Liepold, S. G. Müller, H.-J. Bestmann, Synthesis 2004, 59–
62; d) for the preparation of the Bestmann–Ohira reagent see: J.
Pietruszka, A. Witt, Synthesis 2006, 4266–4268.
[52] For the regioselectivity of hydrozirconation of internal alkynes, see:
a) J. S. Panek, T. Hu, J. Org. Chem. 1997, 62, 4912–4913; b) Z.
Huang, E.-i. Negishi, Org. Lett. 2006, 8, 3675–3678; c) Procedure
adapted from A. Takizawa, K. Fujiwara, E. Doi, A. Murai, H.
Kawai, T. Suzuki, Tetrahedron 2006, 62, 7408–7435.
[53] Silylcupration of 49 and quenching the vinyl cuprate with MeI pro-
vided the corresponding vinyl silane (51) in 96% yield (9:1 regiose-
lectivity). Subsequent conversion to the vinyl iodide was unsuccess-
ful using standard techniques. Full experimental details can be
found in the Supporting Information.
[38] G. M. Rubottom, M. A. Vazquez, D. R. Pelegrina, Tetrahedron Lett.
1974, 15, 4319–4322.
[54] For the deprotection of 1,3-dioxolanes, see: P. G. M. Wuts, T. W.
Greene, Greeneꢀs Protective Groups in Organic Synthesis, 4th ed.,
Wiley, Hoboken, 2007, pp. 455–466, and references therein.
[55] B. C. Ranu, R. Jana, S. Samanta, Adv. Synth. Catal. 2004, 346, 446–
450.
[56] G. Sabitha, R. S. Babu, E. V. Reddy, J. S. Yadav, Chem. Lett. 2000,
1074–1075.
[57] a) A. Ates, A. Gautier, B. Leroy, J.-M. Plancher, Y. Quesnel, J.-C.
Vanherck, I. E. Markꢂ, Tetrahedron 2003, 59, 8989–8999; b) N.
Maulide, J.-C. Vanherck, A. Gautier, I. E. Markꢂ, Acc. Chem. Res.
2007, 40, 381–392.
[39] For the preparation of formates from aldehydes using mCPBA, see:
a) E. Alvarez-Manzaneda, R. Chahboun, F. Bentaleb, E. Alvarez,
M. A. Escobar, S. Sad-Diki, M. J. Cano, I. Messouri, Tetrahedron
2007, 63, 11204–11212; b) C. Margot, D. P. Simmons, D. Reichlin,
D. Skuy, Helv. Chim. Acta 2004, 87, 2662–2684; c) A. Watanabe, H.
Toshima, H. Nagase, T. Nagaoka, T. Yoshihara, Biosci. Biotechnol.
Biochem. 2001, 65, 1805–1811; d) B. Alcaide, M. F. Aly, M. A.
Sierra, J. Org. Chem. 1996, 61, 8819–8825; e) H. Hagiwara, T.
Akama, H. Uda, Chem. Lett. 1989, 2067–2068.
[40] 3-Chloroperbenzoic acid (mCPBA) was purchased from Aldrich
Chemical Co. (<77% purity). The remainder is a mixture of 3-
chlorobenzoic acid and water.
[58] A. Procopino, M. Gaspari, M. Nardi, M. Oliverio, A. Tagarelli, G.
Sindona, Tetrahedron Lett. 2007, 48, 8623–8627.
[59] S. Majumdar, A. Bhattacharjya, J. Org. Chem. 1999, 64, 5682–5685.
[60] Vinyl iodide 50 is light sensitive and as a result the acid-catalyzed
removal of the 1,3-dioxolane was performed in the dark.
[61] a) A. Matsu, K. Yoshida, K. Uohama, S. Hayashi, J. D. Connolly,
G. A. Sim, Chem. Lett. 1985, 935–938; b) A. Matsu, K. Yoshida, Y.
Fukazawa, M. Nakayama, K. Juriyama, Chem. Lett. 1987, 369–372.
[41] a) T. G. Waddell, P. A. Ross, J. Org. Chem. 1987, 52, 4802–4804;
b) for the application of Al(OiPr)3-mediated epoxide ring opening
A
in the preparation of Taxol and Baccatin III, see: D. J. Danishelfsky,
J. J. Masters, W. B. Young, J. T. Link, L. B. Snyder, T. V Magee,
D. K. Jung, R. C. A. Isaacs, W. G. Bornmann, C. A. Alaimo, C. A.
Coburn, M. J. Di Grandi, J. Am. Chem. Soc. 1996, 118, 2843–2859.
[42] G. A. Kraus, K. Frazier, J. Org. Chem. 1980, 45, 2579–2581.
[43] S. Murata, M. Suzuki, R. Noyori, J. Am. Chem. Soc. 1979, 101,
2738–2739.
Received: June 13, 2008
Published online: July 30, 2008
Chem. Eur. J. 2008, 14, 8239 – 8245
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8245