Concise synthesis of norrisolide

Article abstract of DOI:10.1002/ejoc.201200134

The marine natural product norrisolide has been synthesized in a convergent manner with a longest linear sequence of 14 steps. The hydrindane portion of the molecules is prepared through conjugate addition of a functionalized allyl group into cyclopentenone, followed by stereoselective trapping of the enolate generated from the resulting enol silyl ether with an allyl electrophile. Ring-closing metathesis then establishes the 6-5 ring system. Likewise, selective preparation of the side chain features an enantioselective cyclopropanation of furan-2-one, followed by rearrangement and hydrogenation. Coupling of the two major fragments through a Shapiro reaction, followed by reduction and olefination then completes the carbon framework of the natural product. The final steps of the synthesis involve adjustments to the oxidation state of the side chain. New strategies to generate both the hydrindane core and the oxidized side chain has allowed for the more concise and efficient preparation of the natural product. A new strategy to prepare Norrisolide's hydrindane core, as well as the oxidized side chain, has allowed for a more concise and efficient preparation of the natural product. The significantly abbreviated synthetic sequences to these fragments will allow for more facile preparations of related natural product targets.

Full text of DOI:10.1002/ejoc.201200134

SHORT COMMUNICATION
DOI: 10.1002/ejoc.201200134
Concise Synthesis of Norrisolide
Krista Granger[a] and Marc L. Snapper*[a]
Keywords: Natural products / Total synthesis / Asymmetric synthesis / Asymmetric catalysis / Vesicles
The marine natural product norrisolide has been synthesized
in a convergent manner with a longest linear sequence of 14
steps. The hydrindane portion of the molecules is prepared
through conjugate addition of a functionalized allyl group
into cyclopentenone, followed by stereoselective trapping of
the enolate generated from the resulting enol silyl ether with
an allyl electrophile. Ring-closing metathesis then estab-
lishes the 6–5 ring system. Likewise, selective preparation of
the side chain features an enantioselective cyclopropanation
of furan-2-one, followed by rearrangement and hydrogena-
tion. Coupling of the two major fragments through a Shapiro
reaction, followed by reduction and olefination then com-
pletes the carbon framework of the natural product. The final
steps of the synthesis involve adjustments to the oxidation
state of the side chain. New strategies to generate both the
hydrindane core and the oxidized side chain has allowed for
the more concise and efficient preparation of the natural
product.
Introduction
Norrisolide is a marine natural product that was first iso-
lated from the nudibranch mollusc Chromodoris norrisi in
1983 (Figure 1).[1] Norrisolide has also been found in minor
amounts in sponges from the same area. It has been sug-
gested that the molluscs acquire the precursor to norrisolide
and related compounds through feeding on the sponges.[2]
As shown in Figure 1, there are several natural products
that share some of norrisolide’s structural features. For ex-
ample, macfarlandin C,[3] spongionellin,[4] and dendrillol-
ides A and E[5] all possess the same oxygenated side chain.
Likewise, chelonoplysin, norrlandin, and cheviolenes C and
E have norrisolide’s hydrophobic hydrindane ring system.[6]
Chromodorolides A[7] and B[8] can be found in the same
aplysillid sponge as the cheviolenes. The chromodorolides
have shown significant cytotoxicity against the P388 mouse
leukemia cell line. In addition, chromodorolide A possesses
antimicrobial and nematocidal activity. To the best of our
knowledge, no total synthesis of any of these structurally
related natural products, other than norrisolide, has been
reported.
Biological activities reported for norrisolide include sec-
retary phospholipase A2 (SPLA2) inhibition,[9] ichthyotox-
icity,[10] and cytotoxicity.[11] Given our work with ilimaqui-
none,[12] we were particularly interested in Theodorakis and
co-workers’ disclosure concerning norrisolide’s irreversible
influence on the structure of the Golgi apparatus.[13] To bet-
Figure 1. Norrisolide and structurally related natural products.
[a] Department of Chemistry,
Merkert Chemistry Center, Boston College,
Chestnut Hill, MA 02467 USA
ter understand norrisolide’s biological activities, the Theo-
dorakis group described the first total synthesis of the natu-
ral product in 2004.[14] A second-generation approach
shortened their longest linear sequence to norrisolide from
Fax: +1-617-552-1442
E-mail: marc.snapper@bc.edu
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201200134.
2308
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 2308–2311

Concise Synthesis of Norrisolide
28 to 23 steps.[15] In addition to preparing the natural prod- steps as a single diastereomer. With the diene in hand, ring-
uct, their routes also provided access to structural analogs closing metathesis with Grubbs second generation catalyst
designed to probe norrisolide’s biological activities.[16]
10 (0.5 mol-%) gave the ring-closed product in 80% yield,
thus constructing the carbon framework for the hydrindane
portion of the molecule in only three steps. Hydrogenation
of the olefin with Pd/C gave (Ϯ)-3 quantitatively without
the need for further purification.
Results and Discussion
Our approach to norrisolide features some significant
differences in the preparation of the key fragments. As sum-
marized in Scheme 1, construction of hydrindane fragment
3 was envisioned through ring-closing metathesis/reduction
of appropriately functionalized cyclopentanone 5, which in
turn could be prepared by selective conjugate addition/allyl-
ation of methylcyclopentenone 6. Oxidized side chain 4
could be prepared through the thermal rearrangement/re-
duction of cyclopropane 7, which is generated through
asymmetric cyclopropanation of lactone 8. The overall plan
allowed for a concise preparation of the natural product
with a longest linear sequence of 14 steps.
Scheme 2. Synthesis of hydrindane 3. Conditions: (a) CuBr·DMS,
TMSCl, HMPA, THF, –78 °C; Et3N, hexanes, –78 °C. (b) MeLi,
HMPA, THF, –15 °C to r.t.; allyl bromide, THF, –78 °C to r.t., 12 h
(60% yield, 2 steps). (c) Grubbs 2nd gen. cat. 10, CH2Cl2, r.t., 12 h,
(80% yield). (d) H2 (10 atm), Pd/C (5 mol-%), EtOAc, r.t., 12 h
(99% yield). (e) (S)-(–)-2-Methyl-CBS-oxazaborolidine (10 mol-%),
BH3·DMS, THF, 0 °C, 90 s (36% yield recovered 3, Ͼ95%ee).
To access enantiomerically enriched 3, it was necessary
to perform a kinetic resolution. A Corey–Bakshi–Shibata
(CBS) reduction of (Ϯ)-3 was used to selectively reduce the
undesired enantiomer while leaving the desired enantiomer
at the ketone oxidation state. Specifically, treatment of (Ϯ)-
3 with (S)-(–)-2-methyl-CBS-oxazaborolidine (10 mol-%)
and BH3·DMS (0.5 equiv.) yielded a 36% recovery of (+)-
3. The enantiomeric excess of the product typically ranged
from 95 to 98%ee, indicating an S value of approximately
12 (Ϯ3).[21] This five-step sequence can be used to provide
the hydrindane portion of norrisolide in an enantiomer-
ically enriched fashion on a multigram scale.
The concise preparation of norrisolide’s tetrahydrofurof-
uranone side chain should be possible through a formal 1,3-
dipolar cycloaddition between a diazo diester compound
and the appropriate furanyl derivative, as demonstrated by
Pirrung and co-workers.[22] In this regard, furan-2(3H)-one
Scheme 1. Retrosynthetic analysis of norrisolide.
At the start of the project, the state of the art for the 8 was cyclopropanated with dimethyl 2-diazomalonate to
preparation of the hydrindane portion of norrisolide (i.e., generate cycloadduct 7 (Scheme 3).[23] Unlike earlier exam-
3) was a 12-step route developed by the Paquette group for ples, this intermediate did not rearrange in situ to the de-
their synthesis of ent-grindelic acid.[17] In light of the work sired ring system. We surmised that the electron-with-
by Lipshutz and co-workers demonstrating a selective con- drawing nature of the lactone (relative to furan) hindered
jugate addition of various allyl groups to cyclic enones,[18] the cleavage of the cyclopropane, suppressing the usually
we envisioned combining this transformation with an allyl- observed rearrangement. Fortunately, thermolysis of 7 in
ation and ring-closing metathesis to provide a more rapid benzene at 185 °C provided adduct 11. Using this strategy,
access to the desired hydrindane fragment (i.e., 6 Ǟ 5 Ǟ the framework of the norrisane side chain was prepared in
3).[19]
Along these lines, enone 6 can be alkylated with the cupr-
just two steps.
Access to the enantiomerically enriched norrisane side
ate of isoprenyl Grignard in the presence of HMPA[20] to chain is possible by carrying out the cyclopropanation with
generate corresponding silyl enol ether 9 (Scheme 2). Be- Müller’s catalyst 17.[24] Using slow addition of the diazo
cause attempts to allylate the enolate directly were unsuc- compound, cycloadduct 7 was obtained with 85:15er. Inter-
cessful, treatment of TMS enol ether 9 with methyllithium estingly, relative to other catalysts tested, complex 17 dis-
and allyl bromide provided diene 5 in 60% yield over two played an increase in the rate of the reaction, reducing the
Eur. J. Org. Chem. 2012, 2308–2311
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2309

K. Granger, M. L. Snapper
SHORT COMMUNICATION
Scheme 4. Completion of norrisolide. Conditions: (a) 2,4,6-Triiso-
propylphenylsulfonyl hydrazide, HBF4, CH3CN, r.t., 12 h (72%
yield). (b) nBuLi, THF, 2 h, –78 to 0 °C; 13, THF, –78 to –30 °C,
12 h, (92% yield). (c) H2 (10 atm), Rh/Al2O3 (10 mol-%), EtOAc,
r.t., 12 h (65% yield). (d) TMSCH2MgCl, Et2O, r.t. (99% yield).
(e) KHMDS, THF, r.t.; Tf2O, pyr. THF, r.t. (60% yield) (f) TBAF,
AcOH, THF, r.t. (g) PCC, NaOAc, CH2Cl2, 4 Å MS, (70% yield,
2 steps) (h) TFA/H2O (3.3:1). (i) Ac2O, DMAP, Et3N, CH2Cl2, r.t.
(40% yield, 2 steps).
Scheme 3. Preparation of the norrisane side chain. Conditions:
(a) Müller’s catalyst 17, dimethyl 2-diazomalonate, PhF (70% yield,
60–70%ee). (b) Benzene, 185 °C (82% yield). (c) H2 (60 atm), Pd/
C (5 mol-%), EtOAc, r.t., 72 h (65% yield). (d) DIBAL-H, THF,
–65 to –50 °C. (e) TBSCl, imid., DMF (50%, 2 steps). (f) MeN(-
OMe)·HCl, iPrMgCl, THF, –15 °C, 4 h (65% yield).
cyclopropanation time from 12 h for the racemic reaction
with Rh2(OAc)4 to 2 h. This rate enhancement also allowed
for a reduction in the number of equivalents of 8, which
tended to isomerize into conjugation during the course of
the reaction. After cyclopropanation, catalyst 17 can be
precipitated from acetonitrile with 85% recovery and used
one more time for the cyclopropanation with the same effi-
ciency.
Further elaboration of the side chain was continued by
diastereoselective hydrogenation of intermediate 11 by
using Pd/C under 60 atm H2 with triethylamine (1 equiv.,
Scheme 3). The resulting lactone was then reduced with DI-
BAL-H and protected as the TBS lactol to provide com-
pound 12 with a syn relationship between all stereogenic
ring system hydrogen atoms, establishing the cup-shaped
ring system present in norrisolide. Finally, the methyl ester
of intermediate 12 was transformed into Weinreb amide
13,[25] which was found to be a suitable electrophile for the
fragment coupling step.
The completion of the synthesis is shown in Scheme 4. A
Shapiro reaction was used to join the indane fragment to
the oxidized side chain. Condensation of ketone 3 with
2,4,6-triisopropylphenylsulfonyl hydrazide followed by
treatment of the resulting hydrazone with nBuLi (2 equiv.)
gave a vinyl anion, which was treated with Weinreb amide
13. This sequence provided enone 14 as a mixture of dia-
stereomers stemming from the minor enantiomer from the
cyclopropanation. At this stage, the diastereomers were not
readily separable and were taken on as a mixture. Hydro-
genation of 14 by using rhodium on alumina in ethyl acet-
ate provided ketone 15 along with the minor diastereomer,
which at this point could be separated.
tuted olefin, through addition of methyl Grignard and then
elimination with thionyl chloride. In our case, methyl Grig-
nard added cleanly to the ketone; however, attempts to
eliminate water to form the 1,1-disubstituted olefin were
unproductive. After significant optimization, promising re-
sults were identified by using a stepwise Peterson ole-
fination procedure.[30] Addition of TMSCH2MgCl into
ketone 15 occurred in quantitative yield. Unfortunately, di-
rect elimination with the use of either basic or acidic condi-
tions, with and without additives (e.g., HMPA, TMEDA,
etc.) failed to generate desired methenylated product 16.
Fortunately, activation of the tertiary hydroxy group with
triflic anhydride and the use of an excess amount of pyr-
idine cleanly generated desired 1,1-disubstituted olefin 16
as the only product.
With 16 in hand, we were able to complete the synthesis,
with the remaining steps consisting of deprotection of the
lactol and oxidation to the lactone and manipulation of the
methyl acetal by epimerization and acetylation. The TBS
acetal of 16 was deprotected with acetic acid buffered
TBAF. The resulting lactol was not purified, but oxidized
directly with PCC in 70% yield for the two steps. After TFA
hydrolysis, intermediate 17 was obtained as a 1:1 mixture
of lactol epimers. Upon acetylation, only the desired dia-
stereomer, which had spectroscopic data that matched that
of natural norrisolide, was isolated in 40% yield over the
two steps.
Conclusions
The next step was to olefinate the relatively hindered
In summary, the total synthesis of norrisolide was com-
carbonyl group in 15, which proved to be more difficult pleted with a longest linear sequence of 14 steps in an over-
than initially expected.[26] Unfortunately, direct methenyla- all yield of 1.7%. This synthesis is concise and offers asym-
tions of this ketone were unsuccessful.[27] For example, metric entry to the hydrindane core and the catalytic asym-
Takai olefination[28] and methenylations with the use of the metric preparation of the oxidized norrisane side chain, ad-
Nysted reagent[29] failed to give the desired product. Theo- vancements that can facilitate the preparation of additional
dorakis used a two-step approach to access the 1,1-disubsti- members of this class of marine natural products.
2310
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 2308–2311

Concise Synthesis of Norrisolide
[16] a) G. Guizzunti, T. P. Brady, V. Malhotra, E. A. Theodorakis,
J. Am. Chem. Soc. 2006, 128, 4190–4191; b) G. Guizzunti, T. P.
Brady, D. Fischer, V. Malhotra, E. A. Theodorakis, Bioorg.
Med. Chem. 2010, 18, 2115–2122.
Supporting Information (see footnote on the first page of this arti-
cle): Experimental procedures and 1H NMR and 13C NMR spectra
for all key intermediates and final products.
[17] L. A. Paquette, H.-L. Wang, J. Org. Chem. 1996, 61, 5352–
5357.
Acknowledgments
[18] B. H. Lipshutz, C. Hackmann, J. Org. Chem. 1994, 59, 7437–
7444.
[19] For an example of this strategy, see: A. Fürstner, K. Langem-
ann, J. Org. Chem. 1996, 61, 8746–8749.
We thank the National Institutes of Health (R01-CA066617) for
financial support.
[20] D. R. Williams, R. W. Heidebrecht Jr., J. Am. Chem. Soc. 2003,
125, 1843–1850.
[1] a) D. J. Faulkner, J. E. Hochlowski, J. Org. Chem. 1983, 48,
1141–114; b) S. C. Bozbin, D. J. Faulkner, J. Org. Chem. 1989,
54, 3902–3907.
[21] H. B. Kagan, J. C. Fiaud, Top. Stereochem. 1998, 18, 249–325.
[22] M. C. Pirrung, J. Zhang, K. Lackey, D. D. Sternbach, F.
Brown, J. Org. Chem. 1995, 60, 2112–2124.
[23] For a summary, see: R. L. Casaubon, “The Golgi-Disrupting
Agents Ilimaquinone and Norrisolide: Determination of Bio-
logical Interactions Through Synthesis”, Ph.D. Dissertation,
Boston College, Chestnut Hill, MA, 2004.
[24] P. Müller, Y. Allenbach, E. Robert, Tetrahedron: Asymmetry
2003, 14, 779–785.
[25] J. M. Williams, R. B. Jobson, N. Yasuda, G. Marchesini, U.-H.
Dolling, E. J. J. Grabowski, Tetrahedron Lett. 1995, 36, 5461–
5464.
[2] S. Carmely, Y. L. Cojocaru, Y. Kashman, J. Org. Chem. 1988,
53, 4801–4807.
[3] T. F. Molinki, D. J. Faulkner, H. Cun-Heng, G. D. Van Duyne,
J. Clardy, J. Org. Chem. 1986, 51, 4564–4567.
[4] L. Mayol, V. Piccialli, D. Sica, Tetrahedron 1986, 42, 5369–
5376.
[5] a) B. Sullivan, D. J. Faulkner, J. Org. Chem. 1984, 49, 3204–
3206; b) S. C. Bobzin, D. J. Faulkner, J. Org. Chem. 1989, 54,
3902–2907.
[6] P. R. Bergquist, B. F. Bowden, R. C. Cambie, P. A. Craw, P.
Karuso, A. Poiner, W. C. Taylor, Aust. J. Chem. 1993, 46, 623–
663.
[26] For a review on methenylations, see: I. Beadham, J. Mick-
lefield, Curr. Org. Synth. 2005, 2, 231–259.
[7] E. J. Dumdei, E. Dilip de Silva, R. J. Andersen, J. Am. Chem.
Soc. 1989, 111, 2712–2713.
[8] a) S. A. Morris, E. Delip de Silva, R. J. Andersen, Can. J.
Chem. 1991, 69, 768–771; b) W. Rungprom, W. Charvasiri, U.
Kokpol, A. Kotze, M. J. Garson, Mar. Drugs 2004, 2, 101–107.
[9] B. C. M. Potts, D. J. Faulkner, J. Nat. Prod. 1992, 55, 1701–
1717.
[10] M. Gavagnin, R. R. Vardaro, C. Avila, G. Cimino, J. Nat.
Prod. 1992, 55, 368–371.
[11] A. Rudi, Y. Kashman, Tetrahedron 1990, 46, 4019–4022.
[12] H. S. Radeke, C. A. Digits, R. L. Casaubon, M. L. Snapper,
Chem. Biol. 1999, 6, 639–647.
[13] C. Kim, R. Hoang, E. A. Theodorakis, Org. Lett. 1999, 1,
1295–1297.
[14] T. P. Brady, S. H. Kim, K. Wen, E. A. Theodorakis, Angew.
Chem. 2004, 116, 757; Angew. Chem. Int. Ed. 2004, 43, 739–
774.
[15] T. P. Brady, S. H. Kim, K. Wen, C. Kim, E. A. Theodorakis,
Chem. Eur. J. 2005, 11, 7175–7190.
[27] For examples, see: a) N. A. Petasis, E. I. Bzowej, J. Am. Chem.
Soc. 1990, 112, 6392–6394; b) H. Lebel, D. Guay, V. Paquet, K.
Huard, Org. Lett. 2004, 6, 3047–3050; c) H. Lebel, V. Paquet, J.
Am. Chem. Soc. 2004, 126, 320–328; d) F. Bertini, P. Grasselli,
G. Zubiani, Tetrahedron 1970, 26, 1281–1290.
[28] a) K. Takai, T. Kakiuchi, Y. Kataoka, K. Utimoto, J. Org.
Chem. 1994, 59, 2668–2670; b) S. Matsubara, T. Mizuno, T.
Otake, M. Kobata, K. Utimoto, K. Takai, Synlett 1998, 1369–
1371; c) L. Lombardo, Tetrahedron Lett. 1982, 23, 4293–4296;
d) L. Lombardo, Org. Synth. 1987, 65, 81–85.
[29] a) S. Matsubara, M. Sugihara, K. Utimoto, Synlett 1998, 313–
315; b) C. Aïssa, R. Fiveiros, J. Ragot, A. Fürstner, J. Am.
Chem. Soc. 2003, 125, 15512–15520.
[30] a) P. F. Hudrlik, D. Peterson, Tetrahedron Lett. 1972, 13, 1785–
1787; for a more in-depth discussion of this reaction, see: b)
L. F. van Staden, D. Gravestock, D. J. Ager, Chem. Soc. Rev.
2002, 31, 195–200.
Received: February 6, 2012
Published Online: March 13, 2012
Eur. J. Org. Chem. 2012, 2308–2311
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2311