Total synthesis of ipomoeassin F

Article abstract of DOI:10.1021/ol900086b

The first total synthesis of ipomoeassin F was carried out using a convergent approach that relied upon the use of Schmidt glycosidation technology for the coupling of two suitably protected monosaccharide fragments. After two steps, ring-closing metathes

Full text of DOI:10.1021/ol900086b

ORGANIC
LETTERS
2009
Vol. 11, No. 6
1417-1420
Total Synthesis of Ipomoeassin F
Maarten H. D. Postema,*,† Karen TenDyke,‡ James Cutter,§ Galina Kuznetsov,|
and Qunli Xu
Departments of DiscoVery Chemistry, Integrated DiscoVery Technologies, Process
Chemistry, Functional Bioanalytic Technologies, and Molecular and Cellular
Pharmacology, Eisai Research Institute, 4 Corporate DriVe, AndoVer, Massachusetts 01810
maarten_postema@eri.eisai.com
Received January 15, 2009
ABSTRACT
The first total synthesis of ipomoeassin F was carried out using a convergent approach that relied upon the use of Schmidt glycosidation
technology for the coupling of two suitably protected monosaccharide fragments. After two steps, ring-closing metathesis was used to form
the macrocyclic ring, and seven more steps then furnished ipomoeassin F. In vitro inhibitory activity against a four-panel cell line showed low
nanomolar inhibitory activity.
The ipomoeassins are a structurally unique family of
compounds recently isolated from Ipomoea sp. in Suriname
by the Kingston group that showed cytotoxicity toward the
A2780 ovarian cancer cell line.1 Ipomoea sp., commonly
known as morning glory, has about 650 species distributed
worldwide with over 300 species found in the Americas
alone. Many of these plants contain glycoside resins that are
comprised of a few sugars with either single or multiple long-
chain fatty acid(s). Members of the morning glory family
have been long known to possess laxative and puragative
properties2 as well as displaying a range of other biological
effects. For example, these glycosidic resins have been used
as crop protectants in Mexico as they inhibit the growth of
invasive weeds.3 Ipomoea squamosa has shown activity
against Mycobacterium tuberculosis,4 while triclorin A,
isolated from Ipomoea tricolor displays activity against
cultured P-388 and human breast cancer cells3 as well as
antifungal activity.5
Figure 1. Ipomoeassin natural products.
† Discovery Chemistry.
‡ Integrated Discovery Technologies.
§ Process Chemistry.
| Functional Bioanalytic Technologies.
Molecular and Cellular Pharmacology.
(1) Cao, S.; Guza, R. C.; Wisse, J. H.; Miller, J. S.; Evans, R.; Kingston,
D. G. I. J. Nat. Prod. 2005, 68, 487–492.
(2) Pereda-Miranda, R.; Bah, M. Curr. Top. Med. Chem. 2003, 3, 111–
131.
The ipomoeassin natural products (Figure 1) possess
several interesting structural features. The presence of
oxygenation on the linking acid chain and the shorter tail at
(4) Barnes, C. C.; Smalley, M. K.; Manfredi, K. P.; Kindscher, K.;
Loring, H.; Sheeley, D. M. J. Nat. Prod. 2003, 66, 1457–1462.
(3) Pereda-Miranda, R.; Mata, R.; Anaya, A. L.; Wickramaratne,
D. B. M.; Pezzuto, J. M.; Kinghorn, A. D. J. Nat. Prod. 1993, 56, 571–
582.
(5) Castelli, M. V.; Cortes, J. C.; Escalante, A. M.; Bah, M.; Pereda-
Miranda, R.; Ribas, J. C.; Zacchino, S. A. Planta Med. 2002, 68, 739–742.
10.1021/ol900086b CCC: $40.75
Published on Web 02/19/2009
2009 American Chemical Society

the fucoside anomeric center are both uncommon relative
to other glycoside resins. The cinnamate and tiglate functions
are somewhat rare in related compounds since the reduced
or hydrated forms of the tiglate (2-methylbutyrate or 3-hy-
droxy-2-methylbutyric acid) are the more commonly en-
countered motifs.6
Scheme 1. Retrosynthetic Analysis for Ipomoeassin F
The ipomoeassin natural products consist of a fucose-
and glucose-derived (1f2)-ꢀ-disaccharide joined at O-1′ and
O-6′′ by a 14-carbon seco acid chain. Ipomoeassins A and
B are differentiated from ipomoeassin C, D, and E by the
presence of a free or acetylated hydroxyl group at C-5 on
the lipid chain. Ipomoeassins A, C, and D differ from B and
E in the sense that they have O-4′ acetylated.
The published cell growth inhibitory activity for these
compounds against the A2780 OVCAR cell line was in the
range of 35-1900 nM with ipomoeassin D being the most
potent of the series.1 Recently, a new member of the family,
ipomoeassin F,7 bearing striking similarity to ipomoeassin
A was isolated by the same group. Ipomoeassin F (1f)
boasted superior potency in our in-house cell growth inhibi-
tory assays compared to the other members of the ipomoessin
family (1a-e) (Table 1).8
shielding O-4′′ from unwanted derivatization thus reducing
the number of required protecting groups. On the fucoside
sugar, the O-4′ hydroxyl would be acetylated thereby
necessitating preservation of this function for the entire
synthesis. Upon completion of this work, a communication
outlining the total synthesis of total ipomoeassin B and E,
the least potent compounds of the group, appeared.11 These
workers relied upon an elegant ring-closing metathesis
(RCM) approach12 in their synthetic strategy.
Table 1. Cell Growth Inhibitory Data for Natural Ipomoeassin
A, B, D, E, and F (nM)a
HT-29
MDA-MB-435
H522-T1
U937
Our synthesis begins with the preparation of the two
required monosaccharide fragments, gluco donor 7 and
Ipomoeassin A
Ipomoeassin B
Ipomoeassin D
Ipomoeassin E
Ipomoeassin F
46.1
396
11.8
393
4.2
42.6
2700
19.9
1633
9.4
108.9
1070
23.2
967
12.9
20.2
134
7.9
163
2.6
fucoside acceptor 15. The acetate on the known glucal13
2
was exchanged for a TBS group in excellent yield, and
subsequent dihydroxylation (OsO4, NMNO, 94%)14 gave a
mixture of anomers 4 followed by dichloroacetylation to give
5 (Scheme 2). The ꢀ-isomer, which was fully characterized,15
a Values averaged over two experiments.
Given the number of varied acyl-based groups contained
within ipomoeassin F, a strategy based upon ring closing
metathesis (RCM) with the O-6′′ acyl chain and the O-4′
acetate in place followed by a late stage O-3′′ and O-4′′
acylation was chosen as a possible approach to the target
molecule (Scheme 1). This is contrasted by an approach
based upon macrolactonization9 that would necessitate a
higher degree of orthogonal hydroxyl protection. We chose
to employ the R-chloroacetate group (ClCH2C(O)) to protect
the sole two free hydroxyl groups of ipomoeassin F that
would be liberated in the last step of the synthesis under
suitably mild conditions.10 We anticipated that a TBS group
would serve well to block O-3′′ while simultaneously
Scheme 2. Synthesis of Glucosyl Donor 7
(6) See for example: Herna´ndez-Carlos, B.; Bye, R.; Pereda-Miranda,
R. J. Nat. Prod. 1999, 62, 1096–1100.
was found to be the major product (2.3:1, 1H NMR).
Selective removal of the anomeric acyl group with hydrazine
acetate16 gave compound 6, and this was followed by a high-
yielding installation of the trichloroacetimidate group to
furnish the Schmidt donor17 7 in 86% yield as a yellow
crystalline solid.
(7) Cao, S.; Norris, A.; Wisse, J. H.; Miller, J. S.; Evans, R.; Kingston,
D. G. I. Nat. Prod Res. 2007, 21, 872–876.
(8) We thank Professor Kingston for supplying these compounds to
us.
(9) (a) Larson, D. P.; Heathcock, C. H. J. Org. Chem. 1997, 62, 8406–
8418. (b) Lu, S.-F.; O’yang, Q.; Guo, Z.-W.; Yu, B.; Hui, Y.-Z. J. Org.
Chem. 1997, 62, 8400–8405. (c) Brito-Arias, M.; Pereda-Miranda, R.;
Heathcock, C. H. J. Org. Chem. 2004, 69, 4567–4570.
(10) Greene, T. W.; Wuts, P. G. W. ProtectiVe Groups in Organic
Synthesis, 3rd ed.; Wiley: New York, 1999.
The required fucoside was prepared by condensation of
the known fucosyl bromide 818 with chiral alcohol 919 to
1418
Org. Lett., Vol. 11, No. 6, 2009

material) with CSA in methanol. Selective DCC-mediated
acylation with 4-nonone-8-enoic acid22 of the primary O-6′′
alcohol (17f18, 85%, Scheme 4) gave the precursor to ring-
closing metathesis. RCM proceeded best23 with 15 mol %
of Hoveyda-Grubbs catalyst24 added portion-wise in 1,2-
dichloroethane (45 °C for 3 h at 0.05 M dilution) to provide
19 in 84% yield as a mixture of isomers that were not
separated but were directly subjected to hydrogenation (H2,
Pd/C, EtOAc-EtOH) to first saturate the mixture of olefins
(as shown by TLC, silica, 30% EtOAc-hexanes) and then
reductively cleave the O-3′ benzyl group to deliver 20.
Selective chloroacetylation of the O-3′ hydroxyl group then
gave 21 (Scheme 4).
Scheme 3. Synthesis of Fucoside Acceptor 15
The TBS group at O-3′′ hindered the O-4″ hydroxyl to a
greater extent than initially anticipated, and the cinnamoyl
group could only be installed by heating a DCC-mediated
coupling reaction in 1,2-dichloroethane under evaporative
conditions. After several iterations of solvent evaporation
and replacement, a 62% yield of 22 was obtained (75% based
on recovered starting material). A high-yielding removal of
the TBS group was effected with SiF425 to give an 84% yield
of alcohol 23 (Scheme 5). Tigloylation (tiglic acid, DCC,
4-DMAP, 4-DMAP·HCl) gave the fully blocked precursor
to ipomoeassin F 24, and removal of the R-chloroacetates
by employing an excess of DABCO in hot ethanol26
furnished the natural product ipomoeassin F (1f) in 39% yield
after purification by flash chromatography and HPLC.
The synthetic material obtained was identical to a natural
sample of ipomoeassin F kindly provided by Professor
give 10 as a single isomer in 64% yield (Scheme 3). Global
deacetylation furnished triol 11, and stannylidene-mediated
equatorial benzylation20 brought the sequence as far as 12.
Selective silylation (TBSCl, imidazole, DMF) of the equato-
rial hydroxyl group21 gave 13 in 78% yield. Acetylation then
delivered 14, and the sequence was completed by acid-
mediated deprotection of the TBS group to produce 15 in
85% yield over two steps along with 8% of diol 12.
Optimized coupling conditions called for the use of a 1.5:1
ratio of 7 and 15 and 0.5 equiv of boron trifluoride etherate to
furnish a 52% yield (87% based on recovered starting material)
of the target ꢀ-disaccharide 16 along with 17, the product in
which the acetonide was lost in 21% yield (Scheme 4).
(11) Fu¨rstner, A.; Nagano, T. J. Am. Chem. Soc. 2007, 129, 1906–1907.
(12) In fact, Fu¨rstner was the first to employ a RCM approach to these
types of compounds, see: (a) Fu¨rstner, A.; Muller, T. J. Org. Chem. 1998,
63, 424–425. (b) Fu¨rstner, A.; Muller, T. J. Am. Chem. Soc. 1999, 121,
7814–7821. (c) Fu¨rstner, A.; Jeanjean, F.; Razon, P. Angew. Chem., Int.
Ed. 2002, 41, 2097–2101.
Scheme 4. Disaccharide Coupling and Macrocycle Formation
(13) RajanBabu, T. V. J. Org. Chem. 1985, 50, 3642–3644.
(14) Crich, D.; Lim, L. B. J. Chem. Soc., Perkin Trans. 1 1991, 9, 2209–
2214.
(15) All new compounds were fully characterized by 1H, 13C, and COSY
NMR, high-resolution mass spectroscopy, FT-IR, and optical rotation. HPLC
and GC methods were employed when deemed appropriate.
(16) Fu¨rstner, A.; Jeanjean, F.; Razon, P.; Wirtz, C.; Mynott, R.
Chem.-Eur. J. 2003, 9, 307–319.
(17) Schmidt, R. R.; Kinzy, W. AdV. Carbohydr. Chem. Biochem. 1994,
50, 21–123.
(18) Adelhorst, K.; Whitesides, G. M. Carbohydr. Res. 1993, 242, 69–
76.
(19) Chiral alcohol 9 was prepared in >99% ee by addition of a vinyl
cuprate to optically pure 1,2-epoxyheptane. The latter was readily available
through the use of Jacobsen chiral epoxidation technology: Schaus, S. E.;
Brandes, B. D.; Larrow, J. F.; Tokunaga, M.; Hansen, K. B.; Gould, A. E.;
Furrow, M. E.; Jacobsen, E. N. J. Am. Chem. Soc. 2002, 124, 1307–1315.
(20) See, for example: Kanie, O.; Takeda, T.; Ogihara, Y. Carbohydr.
Res. 1989, 190, 53–64.
(21) Corey, E. J.; Venkateswarlu, A. J. Am. Chem. Soc. 1972, 94, 6190–
6191.
(22) The γ-keto-acid was prepared by copper-mediated addition of the
Grignard reagent derived from 5-bromo-1-pentene to succinic anhydride,
see: L’hommet, G.; Freville, S.; Thuy, V.; Petit, H.; Celerier, J. P. Synth.
Commun. 1996, 26, 3897–3901.
(23) In a closely related system, it was found that both the first- and
second-generation Grubbs catalysts mediated the ring closure, but catalyst
loadings of up to 50 and 35 mol %, respectively, were needed to push the
reactions to completion making purification somewhat problematic.
(24) Kingsbury, J. S.; Harrity, J. P. A.; Bonitatebus, P. J.; Hoveyda,
A. H. J. Am. Chem. Soc. 1999, 121, 791–799.
(25) Corey, E. J.; Yi, K. Y. Tetrahedron Lett. 1992, 33, 2289–2290.
(26) Lefeber, D. J.; Kamerling, J. P.; Vliegenthart, J. F. G. Org. Lett.
2000, 2, 701–703.
To complete the conversion of 16 to 17, the acetonide was
removed in 43% yield (67% based on recovered starting
Org. Lett., Vol. 11, No. 6, 2009
1419

Scheme 5. Completion of the Synthesis of Ipomoeassin F
Kingston7 as shown by HPLC, 1H and 13C NMR, and optical
rotation.27 The in vitro cell growth inhibitory activity (Table
2) also correlated well with that from the natural material.
Figure 2. Phase contrast microscopy of Ipomoeassin F treated
H522-T1 cells.
The described synthetic route to ipomoeassin F is flexible
and should allow access to a variety of analogues as other
members of the ipomoeassin family.29
Table 2. Comparison of Cell Growth Inhibitory Data for
Natural and Synthetic Ipomoeassin F (nM)a
Acknowledgment. The authors would like to thank Yuan
(John) Wang (ERI) for supplying copious amounts of glycal
2, Jian DelVecchio and Michael Alaimo (ERI) for analytical
support, Ted Suh, Keith Wilcoxen, and Brian Gallagher
(ERI) for helpful chemistry discussions, Lynn Hawkins and
Keith Wilcoxen (ERI) for help in proof reading the mansu-
cript, Melvin Yu and Bruce Littlefield (ERI) for their support
of this project, and Professor David Kingston (Virginia
Polytechnic Institute) for supplying several natural ipomoe-
assins.
HT-29
MDA-MB-435
H522-T1
U937
Ipomoeassin F
Ipomoeassin F
(synthetic)
4.2
1.4
9.4
4.3
12.9
2.7
2.6
1.1
a Values averaged over two experiments.
The data show that the synthetic ipomoeassin F was
slightly more potent than the natural material, exhibiting
single digit nanomolar potency or less against the four cell
lines tested.
Supporting Information Available: Experimental pro-
cedures for 1f, 3-7, and 10-24. Copies of 1H and 13C NMR
spectra for compounds 1f, 3-7, 10-18, and 20-24 along
with HPLC data for 1f (synthetic and natural) along with
copies of COSY NMR spectra for compounds 7-18 and
20-24. This material is available free of charge via the
Internet at http://pubs.acs.org.
The effect of ipomoeassin F on tumor cells was further
examined by phase contrast microscopy. As shown in Figure
2, ipomoeassin F induced cell death of H522-T1 lung cancer
cells which was preceded by rounding up of the cells.28
(27) See Supporting Information.
OL900086B
(28) Rounding up of cells is commonly associated with loss of
cell-matrix adhesions and detachment of cells from the matrix. See, for
example: Hajdo-Milasinovic, A.; Ellenbroek, S. I. J.; van Es, S.; van der
Vaart, B.; Collard, J. G. J. Cell. Sci. 2007, 120, 555–566.
(29) By making the appropriate changes to the top chain and the fucoside
fragment, access to all current members of the ipomoeassin family should
be possible using the current synthetic strategy.
1420
Org. Lett., Vol. 11, No. 6, 2009