Efficient total syntheses of resin glycosides and analogues by ring-closing olefin metathesis

Article abstract of DOI:10.1021/ja991361l

A highly efficient entry into the resin glycoside family of natural products is outlined which takes advantage of the inherently modular character of ring-closing metathesis (RCM) for the formation of their macrolactone substructures. Starting from only t

Full text of DOI:10.1021/ja991361l

7814
J. Am. Chem. Soc. 1999, 121, 7814-7821
Efficient Total Syntheses of Resin Glycosides and Analogues by
Ring-Closing Olefin Metathesis
Alois Fu1rstner* and Thomas Mu1ller
Contribution from the Max-Planck-Institut fu¨r Kohlenforschung, Kaiser-Wilhelm-Platz 1,
D-45470 Mu¨lheim/Ruhr, Germany
ReceiVed April 26, 1999
Abstract: A highly efficient entry into the resin glycoside family of natural products is outlined which takes
advantage of the inherently modular character of ring-closing metathesis (RCM) for the formation of their
macrolactone substructures. Starting from only three well accessible sugar building blocks and (6S)-undec-
1-en-6-ol (7) (prepared by enantioselective addition of dipentylzinc to hexenal in the presence of a catalyst
formed from Ti(OiPr)4 and bis-(R,R)-trifluoromethanesulfonamide (9)), it was possible to achieve total syntheses
of tricolorin A (1), tricolorin G (2), and jalapinolic acid (58) as well as the synthesis of the disaccharidic unit
48 which constitutes a common structural motif of all simonin, operculin, tuguajalapin, orizabin, mammoside,
quamoclin, and stoloniferin resin glycosides. Furthermore, various analogues of these naturally occurring
glycolipids have been obtained in a straightforward manner from the same set of substrates. This highlights
the flexibility of the chosen approach and opens the door for a synthesis-driven mapping of the structure/
activity profile of this structurally demanding class of oligosaccharides. The macrocyclization reactions via
RCM have been performed using carbene 22 introduced by Grubbs, the cationic allenylidene complex 23, or
the particularly convenient precatalyst [(p-cymene)RuCl2(PCy3)] 24. All these ruthenium-based systems catalyze
the ring closure of the highly functionalized diene substrates with comparable efficiency and turned out to be
compatible with an array of functional groups including unprotected secondary hydroxyl functions.
Introduction
characteristic structural motif, resin glycosides are rich in deoxy
sugars, particularly D-fucose, L-rhamnose, and D-quinovose, with
L-Rhap-(1f2)-D-Fuc, L-Rhap-L-Rha and D-Glcp-(1f2)-D-Fuc
representing highly conserved disaccharidic subunits.
Although the biological properties of resin glycosides have
not yet been fully assessed, a closer look into this class of natural
products seems highly promising in view of the existing data
on the use of glycolipids in general for the treatment of severe
immune disorders.12 The fact that many of them are isolated
from plants which are essential ingredients of traditional
Chemical investigations on resin glycosides were initiated
as early as the middle of the 19th century.1 Until the advent of
modern spectroscopy, however, it was impossible for chemists
to deduce the amazingly complex but aesthetically most
appealing structures of the individual members of this family
of glycolipids. All of them contain jalapinolic acid (11(S)-
hydroxyhexadecanoic acid) as the aglycon which is usually tied
back to form a macrolactone ring spanning two or more units
of their saccharide backbones.2-11 In addition to this very
(10) Quamoclins: Ono, M.; Kuwabata, K.; Kawasaki, T.; Miyahara, K.
Chem. Pharm. Bull. 1992, 40, 2674.
(1) (a) Johnston, J. F. W. Philos. Trans. 1840, 342. (b) Kayser, G. A.
Justus Liebigs Ann. Chem. 1844, 51, 81. (c) Mayer, W. Justus Liebigs Ann.
Chem. 1855, 95, 129. (d) Power, F. B.; Rogerson, H. J. Chem. Soc. Trans.
1912, 101, 1. (e) Mannich, C.; Schumann, P. Arch. Pharm. Ber. Dtsch.
Pharm. Ges. 1938, 276, 221.
(2) Tricolorin family: (a) Pereda-Miranda, R.; Mata, R.; Anaya, A. L.;
Wickramaratne, M.; Pezzuto, J. M.; Kinghorn, A. D. J. Nat. Prod. 1993,
56, 571. (b) Bah, M.; Pereda-Miranda, R. Tetrahedron 1996, 52, 13063.
(c) Bah, M.; Pereda-Miranda, R. Tetrahedron 1997, 53, 9007.
(3) Simonins: Noda, N.; Yoda, S.; Kawasaki, T.; Miyahara, Y. Chem.
Pharm. Bull. 1992, 40, 3163.
(4) Orizabins: Noda, N.; Ono, M.; Miyahara, K.; Kawasaki, T.; Okabe,
M. Tetrahedron 1987, 43, 3889.
(5) Multifidins: Ono, N.; Honda, F.; Karahashi, A.; Kawasaki, T.;
Miyahara, K. Chem. Pharm. Bull. 1997, 45, 1955.
(6) Stoloniferins: (a) Noda, N.; Takahashi, N.; Miyahara, K.; Yang, C.-
R. Phytochemistry 1998, 48, 837. (b) Noda, N.; Takahashi, N.; Kawasaki,
T.; Miyahara, K.; Yang, C.-R. Phytochemistry 1994, 36, 365.
(7) Operculins: (a) Ono, M.; Nishi, M.; Kawasaki, T.; Miyahara, K.
Chem. Pharm. Bull. 1990, 38, 2986. (b) Ono, M.; Fukunaga, T.; Kawasaki,
T.; Miyahara, K. Chem. Pharm. Bull. 1990, 38, 2650.
(11) For yet other families of resin glycosides see the following for
leading references: (a) MacLeod, J. K.; Ward, A.; Oelrichs, P. B. J. Nat.
Prod. 1997, 60, 467. (b) Kitagawa, I.; Baek, N. I.; Kawashima, K.;
Yokokawa, Y.; Yoshikawa, M.; Ohashi, K.; Shibuya, H. Chem. Pharm.
Bull. 1996, 44, 1680. (c) Noda, N.; Tsuji, K.; Kawasaki, T.; Miyahara, K.;
Hanazono, H.; Yang, C.-R. Chem. Pharm. Bull. 1995, 43, 1061. (d) Noda,
N.; Kobayashi, H.; Miyahara, K.; Kawasaki, T. Chem. Pharm. Bull. 1988,
36, 920. (e) Kitagawa, I.; Baek, N. I.; Yokokawa, Y.; Yoshikawa, M.;
Ohashi, K.; Shibuya, H. Chem. Pharm. Bull. 1996, 44, 1693. (f) Fang, Y.-
W.; Chai, W.-R.; Chen, S.-M.; He, Y.-Z.; Zhao, L.; Peng, J.-H.; Huang,
H.-W.; Xin, B. Carbohydr. Res. 1993, 245, 259.
(12) Although most resin glycosides have been isolated from plants
(ConVolVulaceae), scattered reports on the isolation of related glycolipids
from bacteria, fungi, and yeasts can be found in the literature, some of
which contain fatty acid components other than jalapinolic acid as the
aglycon. For an extensive compilation of relevant literature on the isolation,
structure, and biological properties of such glycolipids see the following
for leading references: (a) Bisht, K. S.; Gross, R. A.; Kaplan, D. L. J.
Org. Chem. 1999, 64, 780. (b) See also: Legler, G. Phytochemistry 1965,
4, 29. (c) Wagner, H.; Kazmaier, P. Tetrahedron Lett. 1971, 12, 3233. (d)
Kawasaki, T.; Okabe, H.; Nakatsuka, I. Chem. Pharm. Bull. 1971, 19, 1144
and literature cited therein. (e) Key, B. A.; Gray, G. W.; Wilkinson, S. G.
Biochem. J. 1970, 13, 593. (f) Hirayama, T.; Kato, I. FEBS Lett. 1982,
139, 81. (g) Weber, L.; Stach, J.; Haufe, G.; Hommel, R.; Kleber, H.-P.
Carbohydr. Res. 1990, 206, 13. (h) Asmer, H. J.; Lang, S.; Wagner, F.;
Wray, V. J. Am. Oil Soc. 1988, 65, 1460.
(8) Tuguajalapins: Noda, N.; Tsuji, K.; Miyahara, K.; Yang, C.-R. Chem.
Pharm. Bull. 1994, 42, 2011.
(9) Mammosides: (a) Kitagawa, I.; Ohashi, K.; Baek, N. I.; Sakagami,
M.; Yoshikawa, M.; Shibuya, H. Chem. Pharm. Bull. 1997, 45, 786. (b)
Kitagawa, I.; Baek, N. I.; Ohashi, K.; Sakagami, M.; Yoshikawa, M.;
Shibuya, H. Chem. Pharm. Bull. 1989, 37, 1131.
10.1021/ja991361l CCC: $18.00 © 1999 American Chemical Society
Published on Web 08/10/1999

Syntheses of Resin Glycosides and Analogues
J. Am. Chem. Soc., Vol. 121, No. 34, 1999 7815
medicine recipies provides additional support for this notion.
As a prototype example one may quote the use of Ipomoea
batatas in Brazilian folk medicine.3 This plant, which has
subsequently been introduced as a health food in Japan, is a
rich source of various resin glycosides such as simonin I (3)
and analogues.3 Even more interesting are recent reports on the
significant cytotoxic activity of tricolorin A (1) against cultured
P-388 and human breast cancer cell lines.2 This particular resin
glycoside and congeners (e.g., tricolorin G (2)) also constitute
the allelochemical principle of Ipomoea tricolor, a plant serving
in traditional agriculture in Mexico as a cover crop for the
protection of sugar cane.2
total synthesis of several naturally occurring resin glycosides
and provide at the same time access to a panel of advanced
structural analogues due to the inherent flexibility of the chosen
approach.
Results and Discussion
Total Synthesis of Tricolorin A. Our previous investigations
on RCM have uncovered the essential requirements for produc-
tive macrocyclizations and have revealed the remarkable scope
of this method.17-19 From these data we concluded that the
presence of polar “relay” substituents, their affinity toward the
catalytically active metal species, the proper distance of these
polar groups to the alkene moieties, and low steric hindrance
close to the double bonds constitute decisive parameters for the
productive formation of a large ring via RCM.18
We reasoned that all of these criteria can be met en route to
1, provided that the large ring is closed within the lipidic
disaccharide unit 4 at (or near) the site indicated in Scheme 1.
The required cyclization precursor, i.e., diene 6, can be
assembled via established glycosidation strategies from D-
glucose, D-fucose, and (6S)-undec-1-en-6-ol (7); the latter should
be readily accessible via asymmetric synthesis. A convenient
approach to the L-Rhap-(1f3)-L-Rha building block 5 necessary
for the formation of tricolorin A as well as the assembly of 4
and 5 to the final target have already been described in the
literature.13,14
This retrosynthetic analysis was reduced to practice as shown
in Schemes 2 and 3. Thus, an enantioselective addition of
dipentylzinc to 5-hexenal (8) in the presence of a catalyst formed
in situ from Ti(O-iPr)4 and bis-(R,R)-trifluoromethanesulfon-
amide (9)20 provided (S)-7 in good yield and in excellent
enantiomeric purity (ee g 99%) on a multigram scale. Its
glycosylation with tri-O-acetyl-R-D-fucopyranosyl bromide (10)21
in the presence of AgNO3 on silica/alumina as the promotor 22
gave compound 11 in 69% yield, which was deacetylated (11
f 12) and subsequently converted into the isopropylidene acetal
13 under standard conditions.
(17) For recent reviews on RCM see the following for leading refer-
ences: (a) Grubbs, R. H.; Chang, S. Tetrahedron 1998, 54, 4413. (b)
Fu¨rstner, A. Top. Catal. 1997, 4, 285. (c) Schuster, M.; Blechert, S. Angew.
Chem., Int. Ed. Engl. 1997, 36, 2036. (d) Fu¨rstner, A. Top. Organomet.
Chem. 1998, 1, 37.
To probe these promising and diverse biological effects of
resin glycosides in more detail, a synthesis-driven mapping of
their structure/activity profile is called for. Because of the
difficulties posed by the intricate structures of these glycolipids,
however, very little work has been devoted to this area. The
major challenge resides in the regioselective formation of their
macrolide entities which have been cyclized in all of the
syntheses published so far by conventional macrolactonization
procedures.13-15 It is clear, however, that this approach will
hardly allow the formation of a wide range of structural
analogues, because each new compound ultimately requires an
independent multistep synthesis. Therefore, we have explored
an alternative entry into this challenging family of natural
products by taking advantage of the favorable profile of ring-
closing metathesis (RCM) for the formation of large rings.16
We now disclose our endeavors in this field, which led to the
(18) For RCM-based macrocycle syntheses from our laboratory see: (a)
Fu¨rstner, A.; Langemann, K. J. Org. Chem. 1996, 61, 3942. (b) Fu¨rstner,
A.; Langemann, K. Synthesis 1997, 792-803. (c) Fu¨rstner, A.; Kindler, N.
Tetrahedron Lett. 1996, 37, 7005. (d) Fu¨rstner, A.; Langemann, K. J. Org.
Chem. 1996, 61, 8746. (e) Fu¨rstner, A.; Mu¨ller, T. Synlett 1997, 1010. (f)
Fu¨rstner, A.; Langemann, K. J. Am. Chem. Soc. 1997, 119, 9130. (g)
Fu¨rstner, A.; Gastner, T.; Weintritt, H. J. Org. Chem. 1999, 64, 2361. (h)
Fu¨rstner, A.; Seidel, G.; Kindler, N. Tetrahedron 1999, 55, 8215.
(19) For RCM-based total syntheses of natural products from other groups
see the following for leading references: (a) Nicolaou, K. C.; King, N. B.;
He, Y. Top. Organomet. Chem. 1998, 1, 73. (b) Hoveyda, A. H. Top.
Organomet. Chem. 1998, 1, 105. (c) Miller, S. J.; Blackwell, H. E.; Grubbs,
R. H. J. Am. Chem. Soc. 1996, 118, 9606. (d) Bertinato, P.; Sorensen, E.
J.; Meng, D.; Danishefsky, S. J. J. Org. Chem. 1996, 61, 8000. (e) Martin,
S. F.; Humphrey, J. M.; Ali, A.; Hillier, M. C. J. Am. Chem. Soc. 1999,
121, 866. (f) Magnier, E.; Langlois, Y. Tetrahedron Lett. 1998, 837. (g)
Kim, S. H.; Figueroa, I.; Fuchs, P. L. Tetrahedron Lett. 1997, 38, 2601.
(h) Irie, O.; Samizu, K.; Henry, J. R.; Weinreb, S. M. J. Org. Chem. 1999,
64, 587. (i) Arakawa, K.; Eguchi, T.; Kakinuma, K. J. Org. Chem. 1998,
63, 4741. (j) May, S. A.; Grieco, P. A. Chem. Commun. 1998, 1597.
(20) (a) Takahashi, H.; Kawakita, T.; Ohno, M.; Yoshioka, M.; Koba-
yashi, S. Tetrahedron 1992, 48, 5691. (b) Knochel, P. In ActiVe Metals.
Preparation, Characterization, Applications; Fu¨rstner, A., Ed.; VCH:
Weinheim, 1996; p 191. (c) Knochel, P. Synlett 1995, 393. (d) Noyori, R.
Asymmetric Catalysis in Organic Synthesis; Wiley: New York, 1994; p 255.
(21) Prepared from D-fucose by acetylation (Ac2O, DMAP, pyridine,
98%) and subsequent treatment with HBr (33% in HOAc) in Ac2O/
CH2Cl2 at 0 °C f room temperature; cf. Nicolaou, K. C.; Hummel, C. W.;
Nakada, M.; Shibayama, K.; Pitsinos, E. N.; Saimoto, H.; Mizuno, Y.;
Baldenius, K.-U.; Smith, A. L. J. Am. Chem. Soc. 1993, 115, 7625.
(13) (a) Larson, D. P.; Heathcock, C. H. J. Org. Chem. 1996, 61, 5208.
(b) Larson, D. P.; Heathcock, C. H. J. Org. Chem. 1997, 62, 8406.
(14) (a) Lu, S.-F.; O’yang, Q.; Guo, Z.-W.; Yu, B.; Hui, Y.-Z. Angew.
Chem. 1997, 109, 2442; Angew. Chem., Int. Ed. Engl. 1997, 36, 2344. (b)
Lu, S.-F.; O’yang, Q.; Guo, Z.-W.; Yu, B.; Hui, Y.-Z. J. Org. Chem. 1997,
62, 8400.
(15) Jiang, Z.-H.; Geyer, A.; Schmidt, R. R. Angew. Chem. 1995, 107,
2730; Angew. Chem., Int. Ed. Engl. 1995, 34, 2520.
(16) For a preliminary communication see: Fu¨rstner, A.; Mu¨ller, T. J.
Org. Chem. 1998, 63, 424.

7816 J. Am. Chem. Soc., Vol. 121, No. 34, 1999
Fu¨rstner and Mu¨ller
Scheme 1
Scheme 3a
Scheme 2a
a Conditions: [a] Ac2O, DMAP (cat.), pyridine, 0 °C f rt; [b] (i)
BnNH2, THF, 0 °C f rt; (ii) 1 N HCl; [c] Cl3CCN, Cs2CO3, CH2Cl2,
rt; [d] 13, BF3‚Et2O (cat.), -20 °C, CH2Cl2/n-hexane (1/1); [e]
KOMe (cat.), MeOH, rt; [f] 6-heptenoic acid, DCC, DMAP, CH2Cl2,
rt; [g] 22 (5 mol %), CH2Cl2, reflux, high dilution; [h] H2 (1 atm),
Pd/C (5 mol %), EtOH, rt.
position of 15 using benzylamine gave compound 16,13 which
readily afforded trichloroacetimidate 17 on treatment with
Cl3CCN and Cs2CO3 in CH2Cl2. Reaction of this glycosyl donor
(R:â ≈ 2:1) with the fucose derivative 13 in CH2Cl2/hexane at
-20 °C in the presence of catalytic amounts of BF3‚Et2O cleanly
provided the desired â-configurated disaccharide 18 in 82%
yield.24 Deprotection followed by acylation of the resulting diol
19 with 6-heptenoic acid in the presence of DCC and DMAP
proceeded selectively at the O-3 position, thus affording ester
20 as the only product in 71% yield. The site of acylation was
unambiguously assigned by NMR, and the reactivity pattern
O-3 . O-2 in diol 19 is in full accordance with Heathcock’s
observations in his macrolactonization approach toward tricol-
orin A.13
a Conditions: [a] Reagents as indicated, toluene, rt; [b] AgNO3 on
silica/alumina,22 molecular sieves 3 Å, CH2Cl2, -10 °C; [c] KOMe
(cat.), MeOH, rt; [d] 2,2-dimethoxypropane, p-TsOH‚H2O (cat.),
acetone.
In line with our expectations, diene 20 cleanly cyclized to
the desired 19-membered ring on reaction with the ruthenium
carbene 22 (5 mol %)25 in refluxing CH2Cl2. The fact that neither
the free hydroxyl group nor any other functional group in the
The other monosaccharide unit was obtained from 4,6-O-
benzylidene-D-glucopyranose (14)23 as outlined in Scheme 3.
Peracetylation of 14 followed by deprotection of the anomeric
(24) For pertinent reviews on the trichloroacetimidate method see: (a)
Schmidt, R. R. Angew. Chem. 1986, 98, 213; Angew. Chem., Int. Ed. Engl.
1986, 25, 212. (b) Schmidt, R. R. In ComprehensiVe Organic Synthesis;
Trost, B. M., Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 6, p 33. (c)
Schmidt, R. R.; Kinzy, W. AdV. Carbohydr. Chem. Biochem. 1994, 50, 21.
(25) (a) Nguyen, S. T.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc.
1993, 115, 9858. (b) See also: Schwab, P.; France, M. B.; Ziller, J. W.;
Grubbs, R. H. Angew. Chem. 1995, 107, 2179.
(22) Van Boeckel, C. A. A.; Beetz, T.; Kock-van Dalen, A. C.; van
Bekkum, H. Recl. TraV. Chim. Pays-Bas 1987, 106, 596.
(23) Barili, P. L.; Berti, G.; Catelani, G.; Cini, C.; D’Andrea, F.;
Mastrorilli, E. Carbohydr. Res. 1995, 278, 43.

Syntheses of Resin Glycosides and Analogues
J. Am. Chem. Soc., Vol. 121, No. 34, 1999 7817
Scheme 4a
Table 1. Screening of Different Ruthenium-Based Metathesis
Precatalysts (5 mol % each) in the Cyclization of Diene 20 to
Macrocycle 21
entry
precatalyst
solvent
T (°C)
yield (%)a
1
2
3
22
CH2Cl2
toluene
CH2Cl2
40
80
40
77
85
70
23
24b,c
a After standard hydrogenation of the crude cycloalkene mixture over
Pd/C in EtOH. b Formed in situ from [(p-cymene)RuCl2]2 (2.5 mol %)
and PCy3 (5 mol %). c The reaction is carried out under neon light; cf.
ref 27a.
substrate interferes with RCM illustrates the excellent compat-
ibility and selectivity of the Grubbs catalyst. Hydrogenation of
the crude cycloalkene (E/Z-mixture) thus obtained over Pd/C
afforded the desired disaccharide 21 in 77% yield over both
steps. Since 21 has already been used as a key building block
in previous studies on tricolorin A,13,14 our RCM-based approach
completes a formal total synthesis of this particular target. The
molecular structure of product 21 was confirmed by X-ray
analysis (for details see the Supporting Information).26
Evaluation of Other Metathesis Catalysts. Parallel to our
studies on the application of RCM to natural product synthesis,
we are engaged in a program on the development of new
metathesis precatalysts. In this context, we identified compounds
23 and 24 (which can be formed in situ from commercially
available [(p-cymene)RuCl2]2 and PCy3) as very promising
candidates which exhibit similar application profiles as the
standard ruthenium carbene complex 22 introduced by Grubbs
in a series of model studies.27 They bear, however, the additional
advantage of being particularly easy to make starting from
commercially available reagents only.
a Conditions: [a] (25a): 4-pentenoic acid, DCC, DMAP (cat.),
CH2Cl2, rt, 80%; (25b): 10-undecenoic acid, DCC, DMAP (cat.),
CH2Cl2, rt, 67%; [b] 22 (20 mol %), CH2Cl2, reflux, 3 d; [c] H2 (1
atm), Pd/C, EtOH; [d] 22 (10 mol %), CH2Cl2, reflux, 24 h.
tenoic acid followed by RCM of the resulting dienes allow a
convenient entry into tricolorin A analogues differing from the
parent compound in their lipophilic character. The biological
response to such compounds may unravel as to which extent
the proper balance between the hydrophilicity of the oligosac-
charide segment and the lipophilicity of the macrolide entity
accounts for the physiological activity of resin glycosides in
general.
Therefore, we have prepared compound 26 missing two CH2
units in the lactone ring as well as the significantly more
lipophilic, ring-expanded analogue 27 (Scheme 4). Esterification
of 19 with 4-pentenoic acid or 10-undecenoic acid delivers
dienes 25a,b which are converted into the 17-membered ring
26 and into the 23-membered macrolide 27, respectively, upon
exposure to catalyst 22 and subsequent hydrogenation of the
resulting cycloalkenes (E/Z-mixtures) over Pd on charcoal.
Both macrocyclizations proceed well and deliver the desired
products in respectable yields; it is, however, interesting to note
that the rate of ring closure is significantly slower in the case
of the pentenoic acid derivative 25a and that a higher catalyst
loading is required to achieve quantitative conversion. This
behavior is very much in line with our rationale for RCM-based
macrocyclizations outlined previously,18 which emphasizes the
importance of the distance between the polar groups and the
alkenes to be metathesized. According to this model 4-pen-
tenoates constitute borderline cases, because the reactivity of
the emerging carbenes will be attenuated by the formation ofsix-
membered chelate structures such as A (Scheme 5). It is the
We were tempted to probe the performance of these new
precatalysts in the cyclization of the highly functionalized diene
20. Gratifyingly, both of them afford the desired macrocycle
21 (after hydrogenation of the crude cycloalkene mixture) in
good to excellent yield (Table 1). This includes the most user-
friendly, photochemically driven procedure for RCM outlined
recently (entry 3) which uses only the commercially available
ruthenium dimer [(p-cymene)RuCl2]2 as the precatalyst.27a,28
These results corroborate our previous conclusions that the new
ruthenium-based precatalysts compare well to the Grubbs
carbene 22 in terms of yield, rate, compatibility with functional
groups, and overall selectivity profile.
Tricolorin A Analogues. It should be emphasized that regio-
selective esterifications of diol 19 with acids other than 6-hep-
(28) The precise mode of action of this photochemically driven catalyst
system is yet unknown. Although precatalyst 24 is also effective in toluene,
it is best employed in CH2Cl2; cf. ref 27a. As pointed out by one reviewer,
one can speculate if CH2Cl2 is activated by the ruthenium complex, thus
serving as the actual carbene source. See the following for a leading
reference on the formation of ruthenium carbene complexes by oxidative
insertion into gem-dihaloalkanes: Belderrain, T. R.; Grubbs, R. H.
Organometallics 1997, 16, 4001.
(26) A full report on the X-ray structure of compound 21 can be found
in Lehmann, C. W.; Rust, J.; Fu¨rstner, A., Mu¨ller, T. Z. Kristallogr.,
submitted for publication.
(27) (a) Fu¨rstner, A.; Ackermann, L. Chem. Commun. 1999, 95. (b)
Fu¨rstner, A.; Picquet, M.; Bruneau, C.; Dixneuf, P. H. Chem. Commun.
1998, 1315. (c) See also: Fu¨rstner, A.; Hill, A. F.; Liebl, M.; Winton-Ely,
J. D. E. T. Chem. Commun. 1999, 601.

7818 J. Am. Chem. Soc., Vol. 121, No. 34, 1999
Fu¨rstner and Mu¨ller
Scheme 5
Scheme 7a
Scheme 6a
a Conditions: [a] (i) (Bu3Sn)2O, toluene, reflux, 6 h; (ii) Bu4NI, BnBr,
2.5 h, 34 (57%) + 36 (18%); [b] Ac2O, DMAP, CH2Cl2, rt, 92%; [c]
Ac2O, Et3N, DMAP(cat.), CH2Cl2, 80%.
a Conditions: [a] sym-collidine, EtOH, Bu4NI, 80 °C, 8 h, 66%; [b]
BnBr, KOH, THF, reflux, 8 h, 78%; [c] HOAc/H2O (4:1), rt, 16 h; [d]
Cs2CO3, Cl3CCN, CH2Cl2, rt, 14 h, 46%.
optimize this step any further. Reaction of aliquots of 34 and
36 with Ac2O in CH2Cl2/DMAP gave acetates 35 and 37; the
detailed analysis of their 2D NMR spectra allowed an unequivo-
cal differentiation between the regioisomers and confirmed that
the major product formed in the benzylation reaction is the
desired compound 34 necessary for the total synthesis of
tricolorin G.
stability of such intermediates which decides if the macrocy-
clization reaction will proceed or if the catalyst is merely
sequestered in an unproductive form.18
Total Synthesis of Tricolorin G. Not only is compound 19
the key intermediate en route to tricolorin A 1 and analogues
thereof, but it also opens access to tricolorin G 2, in which
jalapinolic acid spans a trisaccharidic core. For this purpose, a
properly functionalized rhamnose unit must be attached to the
less reactive C-2′′ hydroxyl group of diol 19 prior to the crucial
RCM reaction.
The required rhamnosyl donor 33 is prepared as outlined in
Scheme 6. Thus, reaction of 2,3,4-tri-O-acetyl-R-L-rhamnopyrano-
syl bromide (28)29 with sym-collidine and EtOH delivers product
29 on a multigram scale, the acetyl groups of which can then
be replaced by benzyl ethers. Hydrolysis of the ortho ester group
in 30 with aqueous acetic acid leads to the formation of a
mixture of acetates 31 and 32.30 These individual compounds,
however, must not be separated because they converge into a
single trichloroacetimidate, 33, on exposure to Cl3CCN and
Cs2CO3 in CH2Cl2 as previously outlined for a related case in
the glucose series.31 This outcome reflects a rapid base-induced
migration of the acetyl group, converting 32 into 31 prior to
the formation of the trichloroacetimidate function.
In view of the reactivity pattern of diol 19 (vide supra), its
more nucleophilic 3′′-OH group must be blocked prior to the
crucial glycosidation step. This was achieved by reacting it with
(n-Bu3Sn)2O in toluene, followed by exposure of the resulting
stannyl ether to (n-Bu)4NI and benzyl bromide (Scheme 7).32
Surprisingly, however, this benzylation turned out to be less
selective than the acylations carried out within the syntheses of
tricolorin A and analogues, affording a ∼3.2:1 mixture of
compounds 34 and 36. Since these regioisomers can be readily
separated by conventional flash chromatography, we did not
Reaction of alcohol 34 with trichloroacetimidate 33 catalyzed
by BF3‚Et2O in CH2Cl2/hexane affords trisaccharide 38 in
reasonable yield (Scheme 8).24 Replacement of the acetyl group
by a 6-heptenoic acid ester delivers compound 40 without
incident and sets the stage for the crucial ring closure reaction.
Exposure of diene 40 to catalytic amounts of ruthenium carbene
22 in refluxing CH2Cl2 and subsequent hydrogenation of the
resulting mixture of cycloalkenes in the presence of Wilkinson’s
catalyst afford fully protected tricolorin G (41) in remarkable
93% yield over both steps. Due to the lability of the glycosidic
linkages in this compound, its deprotection turned out to be
somewhat delicate and must be carried out under carefully con-
trolled conditions. For this purpose, the acetal functions are
cleaved by means of dilute trifluoroacetic acid in aqueous CH2-
Cl2 prior to removal of the residual benzyl groups with H2 in
the presence of Pd/C. Quite unexpectedly, we noticed that this
final hydrogenolysis only goes to completion if trifluoroacetic
acid is added to the methanolic medium. Because of the polarity
of the resulting product 2, the final purification has to be carried
out by preparative HPLC, providing tricolorin G in 49% yield
and completing the first total synthesis of this complex
glycolipid. The analytical properties and in particular the 600
MHz NMR data of the synthetic material are in full agreement
with those reported in the literature for the natural product.2c
Synthesis of Tricolorin G Analogues. Upon submitting the
regioisomeric product 36 obtained in the monobenzylation step
of diol 19 described above to the same sequence of reactions,
we were able to obtain compound 46 which constitutes an
analogue of tricolorin G differing from the natural product only
in the site of attachment of the rhamnopyranose ring to the
residual sugar backbone.
(29) Prepared from L-rhamnose by acetylation (Ac2O, DMAP, pyridine,
98%) and subsequent treatment with HBr (33% in HOAc) in Ac2O/CH2-
Cl2 at 0 °C f room temperature; cf. Experimental Section.
(30) The same sequence of reactions using D-glucose instead of L-
rhamnose as the starting material was used during a recent total synthesis
of caloporoside; cf. Fu¨rstner, A.; Konetzki, I. J. Org. Chem. 1998, 63, 3072.
(31) Fu¨rstner, A.; Konetzki, I. Tetrahedron Lett. 1998, 39, 5721.
(32) David, S. In PreparatiVe Carbohydrate Chemistry; Hanessian, S.,
Ed.; Marcel Dekker: New York, 1997; p 69.
As shown in Scheme 9, this synthesis proceeds smoothly,
with the formation of the macrocycle via RCM (44 f 45) again
being a highly productive transformation. Very much in line
with the experiences encountered during the total synthesis of
2, the deprotection sequence requires carefully controlled reac-
tion conditions due to the inherent lability of the deoxysugar

Syntheses of Resin Glycosides and Analogues
J. Am. Chem. Soc., Vol. 121, No. 34, 1999 7819
Scheme 8a
Scheme 9a
a Conditions: [a] BF3‚Et2O (cat.), CH2Cl2/n-hexane, -20 °C, 1 h,
59%; [b] KOMe, MeOH, rt, 4 h; [c] 6-heptenoic Acid, DCC, DMAP
(cat.), CH2Cl2, rt, 82% (over both steps); [d] 22 (10 mol %), CH2Cl2,
reflux, 22 h; [e] H2 (1 atm), RhCl(PPh3)3 (cat.), EtOH, 83% (over both
steps); [f] (i) F3CCOOH, CH2Cl2, rt, 3 h; (ii) H2 (1 atm), Pd/C, MeOH,
F3CCOOH (cat.), 24 h, 46 (56%) + 47 (13%).
a Conditions: [a] BF3‚Et2O (cat.), CH2Cl2/n-hexane, -20 °C, 1 h,
53%; [b] KOMe, MeOH, rt, 4 h; [c] 6-heptenoic Acid, DCC, DMAP
(cat.), CH2Cl2, rt, 84% (over both steps); [d] 22 (10 mol %), CH2Cl2,
reflux, 22 h; [e] H2 (1 atm), RhCl(PPh3)3 (cat.), EtOH, 93% (over both
steps); [f] (i) F3CCOOH, CH2Cl2, rt, 3 h; (ii) H2 (1 atm), Pd/C, MeOH,
F3CCOOH (cat.), 49%.
members of the simonin,3 operculin,7 tuguajalapin,8 orizabin,4
mammoside,9 quamoclin,10 and stoloniferin families.6 Two
prototype examples of these groups of natural products (i.e.,
49 and 50) are displayed in Scheme 10 in addition to simonin
I (3) already mentioned in the Introduction. Since both building
blocks necessary for the construction of this highly conserved
structural motif 48 have already been used for the total syntheses
of tricolorin A and G outlined above, we were prompted to
extend our RCM approach to this particularly important
structural subunit.
As shown in Scheme 11, glycosidation of D-fucopyranoside
13 with trichloroacetimidate 33 under standard conditions
provides product 51 in 69% yield. Saponification of the acetyl
residue and attachment of a 6-heptenoate moiety deliver
substrate 53 required for the subsequent ring-closing metathesis
reaction. Once again, this step proceeds cleanly in the presence
of ruthenium carbene 22; subsequent hydrogenation of the
resulting E/Z-mixture of cycloalkenes in the presence of catalytic
amounts of RhCl(PPh3)3 delivers product 54 in 78% yield over
two steps. Gratifyingly, we noticed that hydrogenolysis of this
product (1 atm of H2, Pd/C) in HOAc/MeOH proceeds in a
stepwise manner, with the benzyl group at O-4′′ being signifi-
cantly more labile. Therefore, it is possible to obtain the
selectively deprotected compound 55 in good yield which is
ideally suited for incorporation into all types of resin glycosides
mentioned above.
glycosidic linkages in compound 45. Once again, traces of tri-
fluoroacetic acid turned out to be necessary to guarantee com-
plete debenzylation on exposure to H2 and Pd/C in MeOH as
the solvent. Under these conditions, however, a partial metha-
nolysis of the lactone ring could not be avoided, providing meth-
yl ester 47 as a minor byproduct in addition to the desired
macrolide 46.
The Common Disaccharide Core of the Simonin, Oper-
culin, Tuguajalapin, Orizabin, Mammoside, Quamoclin, and
Stoloniferin Family of Resin Glycosides. Despite the great
overall diversity of naturally occurring resin glycosides, many
of them share common structural motifs. As already mentioned
in the Introduction, the disaccharide entity L-Rhap-(1f2)-D-
Fuc (48) carrying a macrolactone ring derived from jalapinolic
Synthesis of Jalapinolic Acid. In view of the efficiency by
which jalapinolic acid lactones can be formed on various sugar
templates (vide supra), it is not surprising to find that the parent
jalapinolic acid (58) itself can also be obtained by a simple RCM
approach (Scheme 12).
acid is prominently featured in a large subset comprising several

7820 J. Am. Chem. Soc., Vol. 121, No. 34, 1999
Fu¨rstner and Mu¨ller
Scheme 10
Scheme 11a
a Conditions: [a] BF3‚Et2O (cat.), CH2Cl2/n-hexane, -20 °C, 1.5 h,
69%; [b] KOMe, MeOH, rt, 3 h; [c] 6-heptenoic acid, DCC, DMAP
(cat.), CH2Cl2, rt, 77% (over both steps); [d] 22 (10 mol %), CH2Cl2,
reflux, 16 h; [e] H2 (1 atm), RhCl(PPh3)3 (cat.), EtOH, rt, 78% (over
both steps); [f] H2 (1 atm), Pd/C, MeOH/HOAc (250:1), 2.5 d, 59%.
Thus, esterification of enantiomerically pure alcohol 7 with
6-heptenoic acid provides ester 56 which cyclizes under high
dilution conditions on exposure to catalytic amounts of 22 or
by means of the neon light driven protocol using 24 as the
precatalyst (formed in situ from commercially available [(p-
cymene)RuCl2]2 and PCy3). Hydrogenation of the resulting
macrocycle affords (11S)-jalapinolic acid lactone (57). The
somewhat lower yield obtained in this case is mainly caused
by difficulties in separating the cycloalkene mixture formed by
RCM from traces of unreacted starting material. Saponification
of 57 under fairly harsh conditions provides enantiomerically
pure jalapinolic acid (58) in 87% yield.
Scheme 12a
Conclusions. As few as three well accessible sugar building
blocks are necessary for the first total synthesis of the glycolipid
tricolorin G (2), a formal total synthesis of its congener tricolorin
A (1), and the preparation of the common core segment 55 found
in the simonin, operculin, tuguajalapin, orizabin, mammoside,
quamoclin, and stoloniferin families of resin glycosides. More-
over, various analogues of these bioactive natural products such
as 26, 27, 46, and 47 have been obtained in excellent yields
from the very same starting materials. Finally, a short synthesis
of jalapinolic acid (58) as the common lipophilic component
of all resin glycosides is outlined based upon two metal-
catalyzed C-C bond formations. This flexibility in synthetic
planning results from the exceptional performance and the
inherently modular character of ring-closing metathesis (RCM)
for the formation of macrocyclic rings which constitutes the
key step in all syntheses outlined in this paper. It is worth
mentioning that the classical ruthenium carbene complex 22
introduced by Grubbs25 and the more readily accessible “second-
generation” precatalysts 23 and 2427 perform similarly well. The
a Conditions: [a] 6-heptenoic acid, DCC, DMAP (cat.), CH2Cl2, rt,
89%; [d] 22 (5 mol %), CH2Cl2, reflux, 2.5 d; [e] H2 (1 atm), Pd/C,
EtOH, 8 h, rt, 43% (over both steps); [d] KOH, MeOH, reflux, 1.5 d,
87%.
ease by which these exceedingly useful reagents allow specific
modifications to be made even on polyfunctional arrays highly
recommends RCM as a strategic tool for the total synthesis of
complex natural products as well as for applications to medicinal
chemistry, where detailed structure/activity correlations require
systematic variations of given lead structures. The results

Syntheses of Resin Glycosides and Analogues
J. Am. Chem. Soc., Vol. 121, No. 34, 1999 7821
summarized above confirm this notion,33 because the syntheses
of the same panel of compounds by more conventional method-
ology would require unreasonable efforts. Current work in our
laboratory intends to corroborate this aspect and to expand the
scope of metathetic transformations even further.34
Acknowledgment. Generous financial support by the Max-
Planck-Gesellschaft and the Deutsche Forschungsgesellschaft
(Leibniz program) is acknowledged with gratitude. Th.M. thanks
the Fonds der Chemischen Industrie for a Kekule´ Stipendium.
We thank C. Wirtz for her assistance in recording and analyzing
some of the 2D NMR spectra.
Experimental Section
Supporting Information Available: Full experimental
section, compilation of the instrumentation used, details con-
cerning the X-ray structure of compound 21, list of IR data of
all new compounds, and copies of the NMR spectra of
keyintermediates and products (PDF). This material is available
free of charge via the Internet at http://pubs.acs.org.
See the Supporting Information.
(33) The epothilone case constitutes a most notable precedent for this
conclusion. For extensive reviews on how the SAR of this promising
anticancer agent has been established using RCM-based synthesis strategies
see ref 19a,d,j and the following for leading references: (a) Nicolaou; K.
C.; Roschangar, F.; Vourloumis, D. Angew. Chem., Int. Ed. 1998, 37, 2014.
(b) Nicolaou, K. C.; He, Y.; Roschanger, F.; King, N. P.; Vourloumis, D.;
Li, T. Angew. Chem., Int. Ed. Engl. 1998, 37, 84. (c) Nicolaou, K. C.,
Vallberg, H.; King, N. P.; Roschanger, F.; He, Y.; Vourloumis, D.; Nicolaou,
C. G. Chem. Eur. J. 1997, 3, 1957. (d) Nicolaou, K. C.; Vourloumis, D.;
Li, T.; Pastor, J.; Wissinger, N.; He, Y.; Ninkovic, S.; Sarabia, F.; Vallberg,
H.; Roschanger, F.; King, N. P., Finlay, M. R. V.; Giannakakou, P.; Verdier-
Pinard, P.; Hamel, E. Angew. Chem., Int. Ed. Engl. 1997, 36, 2097. (e)
Meng, D.; Bertinato, P.; Balog, A.; Su, D.-S.; Kamenecka, T.; Sorensen,
E. J.; Danishefsky, S. J. J. Am. Chem. Soc. 1997, 119, 10073. (f) Su, D.-S.;
Balog, A.; Meng, D.; Bertinato, P.; Danishefsky, S. J.; Zheng, Y.-H.; Chou,
T.-C.; He, L.; Horwitz, S. B. Angew. Chem., Int. Ed. Engl. 1997, 36, 2093
and literature cited therein.
JA991361L
(34) (a) For the first examples of ring-closing diyne metathesis see:
Fu¨rstner, A.; Seidel, G. Angew. Chem. 1998, 110, 1758; Angew. Chem.,
Int. Ed. Engl. 1998, 37, 1734. (b) For Pt(II)-catalyzed enyne metathesis
reactions see: Fu¨rstner, A.; Szillat, H.; Gabor, B.; Mynott, R. J. Am. Chem.
Soc. 1998, 120, 8305. (c) For metathesis in supercritical carbon dioxide
see: Fu¨rstner, A.; Koch, D.; Langemann, K.; Leitner, W.; Six, C. Angew.
Chem. 1997, 109, 2562; Angew. Chem., Int. Ed. Engl. 1997, 36, 2466.