New applications of the n-pentenyl glycoside method in the synthesis and immunoconjugation of fucosyl GM1: A highly tumor-specific antigen associated with small cell lung carcinoma

Article abstract of DOI:10.1021/ja992594f

The synthesis of fucosyl GM₁ pentenyl glycoside 1b, and its conjugation to carrier protein KLH to give 1c is related. Bioconjugation of 1b was realized using the pendant olefin contained in the reducing end n-pentenyl glycoside (NPG). The key s

Full text of DOI:10.1021/ja992594f

J. Am. Chem. Soc. 1999, 121, 10875-10882
10875
New Applications of the n-Pentenyl Glycoside Method in the
Synthesis and Immunoconjugation of Fucosyl GM1: A Highly
Tumor-Specific Antigen Associated with Small Cell Lung Carcinoma
Jennifer R. Allen† and Samuel J. Danishefsky*,†,‡
Contribution from the Laboratory of Bioorganic Chemistry, Sloan-Kettering Institute for Cancer Research,
1275 York AVenue, New York, New York 10021 (USA) and Department of Chemistry, Columbia UniVersity,
HaVemeyer Hall, New York, New York, 10027
ReceiVed July 21, 1999
Abstract: The synthesis of fucosyl GM1 pentenyl glycoside 1b, and its conjugation to carrier protein KLH to
give 1c is related. Bioconjugation of 1b was realized using the pendant olefin contained in the reducing end
n-pentenyl glycoside (NPG). The key step of the endeavor is a stereospecific [3 + 3] coupling reaction using
our sulfonamido glycosidation protocol (see 23 + 12 f 15). Pre-installation of the NPG was required for an
optimal [3 + 3] coupling yield and to allow for smooth global deprotection. The synthesis and subsequent
immunocharacterization served to confirm the assigned structure of the natural tumor antigen. Fully synthetic
conjugate 1c advances our program toward the goal of using a synthetic vaccine containing fucosyl GM1 as
a potential target for immune attack against small cell lung carcinoma.
Introduction
cells in a variety of cancers (breast, colon, prostate, ovarian,
liver, small cell lung, and adenocarcinomas). In addition, they
have been immunocharacterized by monoclonal antibodies and
therefore have relevant serological markers available for im-
munological studies. Such studies suggest that patients im-
munized in an adjuvant setting with carbohydrate-based vaccines
produce antibodies reactive with cancer cells and that the
production of such antibodies prohibits tumor reoccurrence and
correlates with a more favorable diagnosis.7
It has been known for some time that certain glycolipids and
glycoproteins, which are either not detectable or barely detect-
able in a particular normal tissue type by immunohistochemistry,
are more commonly encountered in tumors of that tissue.1 These
high levels of expression on tumor cells will often result in an
antibody response, apparently too weak or otherwise ineffective
to provide requisite immunoprotection or immunorejection.
Nonetheless, the idea that vaccination with noncell surface-
bound chemically equivalent antigens could trigger a clinically
useful active immune response is certainly very attractive. It
provides the basis for using carbohydrates in the development
of antitumor vaccines.2
A major drawback in using carbohydrate epitopes, however,
is that they are generally not readily available by isolation from
natural sources. It therefore falls to the organic chemist to
provide a workable program of synthesis such that the epitope
of interest can be systematically studied in a favorable molecular
context optimized to elicit an immunological response. The
present state of the art in the field primarily relies on incorpora-
tion of the epitope into a carrier protein.8 Several strategies for
fashioning the linkage between carbohydrate and protein have
been followed, although optimal spacer-linker combinations
have not been determined. Nonetheless, successful synthesis of
the entire construct provides a basis for beginning the im-
munological evaluation.
Cancer carbohydrate antigens such as TF, Tn,3 sTn, KH-1,4
Ley ,5 and Globo-H6 are suitable targets for both active and
passive immunotherapies because they have been carefully
characterized as being overexpressed at the surface of malignant
† Sloan-Kettering Institute for Cancer Research
‡ Columbia University
(1) (a) Hakomori, S. Cancer Res. 1985, 45, 2405. (b) Feizi, T. Cancer
SurV. 1985, 4, 245. (c) Lloyd, K. O. Cancer Biol. 1991, 2, 421. (d) P. O.
Livingston Immunol. ReV. 1995, 145, 147-166.
(2) (a) Hakomori, S.; Zhang, Y. Chem. Biol. 1997, 4, 97. (b) Livingston,
P. O. Curr. Opin. Immunol. 1992, 4, 2. (c) Dranoff, G.; Jaffee, E.; Lazenby,
A.; Golumbek, P.; Levitsky, H.; Brose, K.; Jackson, V.; Hamada, H.;
Paradoll, D.; Mulligan, R. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 3538.
(d) Tao, M.-H.; Levy, R. Nature 1993, 362, 755. (e) Boon, T. Int. J. Cancer
1993, 54, 177. (f) Lo-Man, R.; Bay, S.; Vicher-Guerre, S.; Deriaud, E.;
Cantacuzene, D.; Leclerc, C. Cancer Res. 1999, 59, 1520. (g) Ragupathi,
G. Cancer Immunol. Immunother. 1996, 43, 152-157.
(3) (a) Springer, G. F. Science 1984, 224, 1198. (b) Campbell, B. J.;
Finnie, E. F.; Hounsell, E. F.; Rhodes, J. M. J. Clin. InVest. 1995, 95, 571.
(4) (a) Nudelman, E.; Levery, S. B.; Kaizu, T.; Hakomori, S. J. Biol.
Chem. 1986, 261, 11247. (b) Kaizu, T.; Levery, S. B.; Nudelman, E.;
Stenkamp, R. E.; Hakomori, S. J. Biol. Chem. 1986, 261, 11254. (c) Kim,
S. Y.; Yuan, M.; Itzkowitz, S. H.; Sun, Q.; Kaizu, T.; Palekar, A.; Trump,
B. F.; Hakamori, S.-I. Cancer Res. 1986, 46, 5985.
(6) (a) Kannagi, R.; Levery, S. B.; Ishijamik, F.; Hakomori, S.;
Schevinsky, L. H.; Knowles, B. B.; Solter, D. J. Biol. Chem. 1983, 258,
8934; (b) Bremer, E. G.; Levery, S. B.; Sonnino, S.; Ghidoni, R.; Canevari,
S.; Kannagi, R.; Hakomori, S. J. Biol. Chem. 1984, 259, 14773. (c) S.
Menard, E. Tagliabue, S. Canevari, G. Fossati, M. I. Colnaghi, Cancer Res.
1983, 43, 1295. (d) Kudryashov, V.; Ragupathi, G.; Kim, I. J.; Breimer,
M. E.; Dansihefsky, S. J.; Livingston, P. O.; Lloyd, K. O. Glycoconjugate
J. 1998, 15, 243.
(7) Francis, M. J. In Vaccines: Recent Trends and Progress; Gregoriadis,
G., Allison, A. C., Poste, G., Eds.; Plenum Press: New York, 1991.
(8) (a) Stroud, M. R.; Levery, S. B.; Martensson, S.; Salyan, M. E. K.;
Clausen, H.; Hakomori, S. Biochemistry 1994, 33, 10672. (b) Yuen, C.-T.;
Bezouska, K.; O’Brien, J.; Stoll, M.; Lemoine, R.; Lubineau, A.; Kiso, M.;
Hasegawa, A.; Bockovich, N. J.; Nicolaou, K. C.; Feizi, T. J. Biochem.
1994, 269, 1595. (c) Stroud, M. R.; Levery, S. B.; Nudelman, E.; Salyan,
M. E. K.; Towell, J. A.; Roberts, C. E.; Watanabe, M.; Hakomori, S. J.
Biol. Chem. 1991, 266, 8439.
(5) (a) Lloyd, K. O. Am. J. Clin. Pathol. 1987, 87, 129. (b) Lloyd, K. O.
Cancer Biol. 1991, 2, 421. (c) Yin, B. W. T. et al. Int. J. Cancer 1996, 65,
406.
10.1021/ja992594f CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/11/1999

10876 J. Am. Chem. Soc., Vol. 121, No. 47, 1999
Allen and Danishefsky
An ongoing project in our laboratory involves the develop-
ment of anti-cancer vaccines that incorporate fully synthetic
carbohydrate moieties.9 Our strategy involves total synthesis of
the carbohydrate epitope and its subsequent covalent biocon-
jugation to carrier protein. The vaccine constructs are then
subjected to appropriate mouse immunization studies, with the
ultimate goal of advancing to human clinical trials. This strategy
has resulted in several fully synthetic tumor-associated carbo-
hydrate-based vaccines which are at various stages of advanced
preclinical and clinical processing. In fact, our Globo-H based
vaccine is undergoing clinical evaluation for the treatment of
prostate and breast carcinomas at the phase II level,10 while a
Lewisy antigen-based vaccine, already tested in ovarian cancer,
is awaiting more extensive follow-up evaluation.11
We became interested in the fucosylated GM1 ganglioside,
1a, for a number of reasons. Nilsson et al. identified fucosyl
GM1 as a specific marker associated with small cell lung cancer
(SCLC) cells.12 These workers isolated the glycosphingolipid
fucosyl GM1 (1a) as the major ganglioside component contained
in human SCLC tissue. Furthermore, monoclonal antibodies
(F12) to the antigen serve to detect fucosyl GM1 in tissues and
serum of SCLC patients.13 Immunohistochemistry studies have
suggested that, due to its highly restricted distribution in normal
tissues, fucosyl GM1 could be an excellent target for immune
attack against SCLC. Remarkably, fucosyl GM1 has thus far
not been found on any other human cancer cell lines, indicating
that it is Very SCLC-specific.14
Figure 1. Fucosyl GM1.
We report herein our synthesis of fucosyl GM1 pentenyl
glycoside, 1b. The pendant olefin-containing glycoside provided
three large advantages. It was used to facilitate [3 + 3] coupling,
to enable global deprotection and to function as an access point
for conjugation to the carrier protein keyhole limpet hemocyanin
(KLH). We also report the initial immunocharacterization of
1b (Figure 1).
Synthetic Strategy and Results
The structural assignment of the carbohydrate moiety of the
SCLC antigen was based on a combination of enzymatic and
chemical degradations.12a While there was no particular reason
to question this assignment, the development of a carbohydrate-
based attack on SCLC could benefit from a definitive assignment
of the linkage modes of the various monosaccharides, including
the stereochemistry at each glycosidic attachment. No syntheses
of this carbohydrate sector have appeared in the literature. Thus,
our program directed to the preparation of an immunoofunctional
antigen by chemical means (vide infra) would also encompass
the goal of securing the structure assignment. We note that our
synthetic program would allow for presentation of the hexa-
saccharide epitope independent of the ceramide to the F12 mAb
to ensure that all specificity is directed at the carbohydrate sector.
Furthermore, the construct must be so functionalized as to
anticipate the need for its conjugation to the carrier protein in
anticipation of building an effective antitumor vaccine.
Given that there were no reported total syntheses of the
hexasaccharide moiety, at the outset of this program we were
much influenced by our earlier ventures directed at Globo-H15
and the pentasaccharide GM1.16 Those syntheses were high-
lighted by a remarkable late-stage convergent sulfonamido gly-
cosidation reaction governed by a proximal hydroxyl-directing
effect.17 Because of the presence of potentially labile fucose
and sialic acid moieties required for fucosyl GM1, such a coup-
ling was viewed as a particularly difficult merger. It was not
without some concerns that we charted a course based on ex-
posing our sulfonamido glycosidation format18 to a new chal-
lenge. Our retrosynthetic analysis is shown in Scheme 1.
As seen, we dissected the hexasaccharide core of 2a into two
component trisaccharide structures, 3 and 4, each containing a
densely functionalized array. In the forward sense, the coupling
of DEF donor 3 and ABC acceptor 4 would be the most complex
union we have thus far attempted using the sulfonamido
glycosidation protocol. Our previous results had indicated that
the presence of a free hydroxyl at C4 of the reducing end
galactose (see asterisk) in donor 3 would be necessary to direct
the formation of the required â-linkage.15,17 Successful coupling
would lead to the hexasaccharide core containing both the fucose
and sialic acid moieties. We initially set out to synthesize 1 via
a hexasaccharide glycal of type 2a. In accordance with earlier
work in our laboratory, the flexible terminal glycal would,
hopefully, allow for a late-stage installation of the ceramide side
chain to ultimately generate glycolipid 1a. Moreover, the glycal
could be used to allow for production of allyl glycoside 1d,
whose terminal olefin could then serve as a linker group for
bioconjugation. A functionalized glucopyranoside 2b could also
serve our purposes (vide infra).
(9) For a recent review of efforts continuing in our laboratory, see:
Danishefsky, S. J.; Allen, J. R. Angew. Chem., Int. Ed. 1999, in press and
references therein.
(10) (a) Ragupathi, G.; Park, T. K.; Zhang, S.; Kim, I. J.; Graber, L.;
Adluri, S.; Lloyd, K. O.; Danihefsky, S. J.; Livingston, P. O. Angew. Chem.,
Int. Ed. Engl. 1997, 36, 125. (b) Ragupathi, G.; Slovin, S. F.; Adluri, S.;
Sames, D.; Kim, I. J.; Kim, H.; Spassova, M.; Bornmann, W. G.; Lloyd,
K. O.; Scher, H. I.; Livingston, P. O.; Danishefsky, S. J. Angew. Chem.,
Int. Ed. 1999, 38, 563. (c) Slovin, S. F.; Ragupathi, G.; Adluri, S.; Ungers,
G.; Terry, K.; Kim, S.; Spassova, M.; Bornmann, W. G.; Fazzari, M.; Dantis,
L.; Olkiewicz, K.; Lloyd, K. O.; Livingston, P. O.; Danishefsky, S. J.; Scher,
H. I. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 5710.
(11) Kudryashov, V.; Kim, H. M.; Ragupathi, G.; Danishefsky, S. J.;
Livingston, P. O.; Lloyd, K. O. Cancer Immunol. Immunother. 1998, 45,
281.
(12) (a) Nilsson, O.; Mansson, J.-E.; Brezicka, T.; Holmgren, J.;
Lindholm, L.; Sorenson, S.; Yngvason, F.; Svennerholm, L. Glycocojugate
J. 1984, 1, 43. (b) Brezicka, F. T.; Olling, S.; Nilsson, O.; Bergh, J.;
Holmgren, J.; Sorenson, S.; Yngvason, F. Cancer Res. 1989, 49, 1300.
(13) (a) Nilsson, O.; Brezicka, T. F.; Holmgren, J.; Sorenson, S.;
Svennerholm, L.; Yngvason, F.; Lindholm, L. Cancer Res. 1986, 46, 1403.
(b) Vangsted, A. J.; Clausen, H.; Kjeldsen, T. B.; White, T.; Sweeney, B.;
Hakomori, S.; Drivsholm, L.; Zeuthen, J. Cancer Res. 1991, 51, 2879.
(14) Zhang, S.; Cordon-Cardo, C.; Zhang, H. S.; Reutter, V. E.; Adluri,
S.; Hamilton, W. B.; Lloyd, K. O.; Livingston, P. O. Int. J. Cancer 1997,
73, 42.
(15) Park, T. K.; Kim, I. J.; Hu, S.; Bilodeau, M. T.; Randolph, J. T.;
Kwon, O.; Danishefsky, S. J. J. Am. Chem. Soc. 1996, 118, 11488.
(16) Bhattacharya, S.; Dansihefsky, S. J. J. Org. Chem., in press.
(17) Kwon, O.; Danishefsky, S. J. J. Am. Chem. Soc. 1998, 120, 1588.
(18) (a) Griffith, D. A.; Danishefsky, S. J. J. Am. Chem. Soc. 1990, 112,
2, 5811. (b) Griffith, D. A.; Danishefsky, S. J. J. Am. Chem. Soc. 1991,
113, 5863.

Synthesis and Immunoconjugation of Fucosyl GM1
J. Am. Chem. Soc., Vol. 121, No. 47, 1999 10877
Scheme 1
Scheme 2
In anticipation of the decisive [3 + 3] coupling, the
trisaccharide acceptor 4 required a differentiated C4 acceptor
site in the B ring. Thus, aside from the uncertainties associated
with carrying potentially labile fucoside and sialoside linkages
in the coupling step, there was another concern. As seen, both
the acceptor and donor moieties would be carrying potentially
equally reactiVe C4 axial galactose hydroxyl groups. Hence,
there was a significant question as to the controllable sequencing
of the donor and acceptor roles. It was conceivable that careful
selection of protecting groups could serve to “de-activate” the
C4 axial hydroxyl in the donor trisaccharide (3) toward
nucleophilic attack, thus allowing for 4 to act as the requisite
acceptor.19 Such a strategy would presumably rely on steric and
electronic arguments. In addition, inherent in the designed
merger is that glycosidic activation of the donor 3 should also
serve to de-activate its C4 hydroxyl group toward self-coupling
and consequently permit the desired coupling. Since acceptor
4 also eventually joins to the sialic acid moiety, the ABC
trisaccharide can further be dissected into bifunctionalized
lactose derivative 7 and an appropriate sialic acid donor 8. The
synthesis of the glycal corresponding to 4 was worked out in
the course of our synthesis of GM1.16 We hoped to prepare the
donor corresponding to 3 by using methodology in our synthesis
of the MBr1 antigen, Globo-H.15 The DEF trisaccharide is
constructed by selective fucosylation using fucosyl donor 6 on
disaccharide 5, itself available from two glycal building blocks.
Thus, in short order we hoped to be in a position to examine
the exciting possibility of constructing the core of fucosyl GM1
by a highly convergent, but challenging, [3 + 3] coupling.
We begin our account (as shown in Scheme 2) with the
synthesis of the ABC trisaccharide starting from the known
protected lactal derivative, 9.17 Selective sialylation of the C3′
equatorial hydroxyl in 9 (see bold) proceeded smoothly with
phosphite donor 1020 to yield the glycal 11 as the only ob-
servable isomer in 75% yield. We note that use of the chloro
donor21 and the thiomethyl donor22 corresponding to 10 in
coupling reactions with 9 proceeded in poor yields and with
reduced anomeric selectivity. In addition, we employed propi-
onitrile as the solvent because of the necessity to perform the
reaction at low temperatures. Use of elevated reaction temper-
atures in acetonitrile as the solvent resulted in diminished
anomeric selectivity, regioselectivity, and lower chemical yields.
(20) (a) Sim, M. M.; Kondo, H.; Wong, C.-H. J. Am. Chem. Soc. 1993,
115, 2260. (b) Chappell, M. D.; Halcomb, R. L. Tetrahedron 1997, 53,
11109.
(21) Martichonok, V.; Whitesides, G. M. J. Org. Chem. 1996, 61, 1702.
(22) (a) Hasegawa, A.; Ohki, H.; Nagahama, T.; Ishida, H.; Kiso, M.
Carbohydr. Res. 1991, 212, 277. (b) Kanie, O.; Kiso, M.; Hasegawa, A. J.
Carbohydr. Chem. 1988, 7, 501.
(19) Our synthesis of asialo GM1 addresses a similar situation. See ref
17.

10878 J. Am. Chem. Soc., Vol. 121, No. 47, 1999
Allen and Danishefsky
Scheme 3a
a Reagents: (a) MeOTf, CH2Cl2:Et2O (2:1), 0 °C, 23%; (b) (i) DMDO, CH2Cl2; (ii) PnOH, ZnCl2, -78 °C, 65%; (c) TBAF, AcOH, THF; (d)
NaOMe, MeOH; (e) NaOH, THF; (f) Na/NH3, THF -78 °C, then MeOH; (g) Ac2O, pyridine, DMAP, CH2Cl2, 46% five steps.
The key DEF trisaccharide was synthesized as previously
described in our Globo-H synthesis.15 The requisite thioethyl
donor 12 is shown in Scheme 2. On the basis of previous
experience, it was expected that this specific donor would favor
â-glycosidation via sulfonamido participation under the close
guidance of the “proximal hydroxyl” directing effect (see
asterisk).15,17 This presumption was in fact reduced to prac-
tice. In an experiment directed at “proof of principle”, reaction
protecting groups cleanly produced the desired hexasaccharide
containing nine hydroxyl groups. Saponification of the methyl
ester could be effected with either NaOH/H2O/MeOH, LiOH/
H2O/THF, or K2CO3/H2O/MeOH. However, upon subjecting
the resulting material to reduction with sodium in liquid
ammonia and re-acetylation, the peracetate 14 was not ob-
tained.26 While the decomposition products were not character-
ized in detail, they seemed to lack the terminal glycal linkage.
Deletion of the saponification step in the deprotection sequence,
resulted in elimination of the sialic acid residue during the
dissolving metal reduction phase. Thus, following this protocol,
we were able to obtain a peracetylated pentasaccharide corre-
sponding to essentially fucosyl asialo GM1. Apparently, there
are effects operating in the substrate for the dissolving metal
reduction which precludes clean conversion to 14.
In an effort to find a hexasaccharide which was suitable for
global deprotection, we considered replacing the reducing end
glycal. For vaccine development we required, in any case, a
linker that could be modified to allow for conjugation to KLH.
Previous syntheses in our carbohydrate-based vaccine program
utilized the terminal olefin contained in an allyl glycoside
(analogous to structure 1d) as a linker for bioconjugation. As
depicted in Figure 2, the allyl glycoside in those constructs was
installed in a late-stage glycosidation by solvolysis of a glycal
epoxide. Typically, the glycal was maintained through the global
deprotection to the peracetate stage. In those investigations, it
was demonstrated that epoxidation of the glycal, followed by
treatment with allyl alcohol gave the corresponding allyl
glycoside. Finally, removal of the ester-protecting groups yielded
fully deprotected allyl glycoside poised for bioconjugation.27
23
of 12 with 5.0 equivalents of MeOTf in the presence of 11
gave the desired hexasaccharide 13, albeit in low yield (23%,
Scheme 3).
Attempts to increase the efficiency in the [3 + 3] coupling
met with only limited success. A significant complication was
that the promoter (MeOTf), after prolonged reaction times or
when used in a large excess, reacted with the N-acetamide
contained on the sialic acid ring to give an imidate derivative,
13i.24 Besides 13i, byproducts formed in the [3 + 3] coupling
included degraded donor corresponding to 12 and the imidate
version of recovered acceptor 11, also formed by exposure to
MeOTf. Not surprisingly, the imidates could be hydrolyzed back
to the desired acetamides.25 However, given the extra steps
involved in such a protocol, we chose to attempt to avoid imidate
formation by using a minimal amount of promoter (1.5-2.5
equiv). This attempted fine-tuning also adversely affected the
coupling yield. Similarly, use of either a large excess of donor
12 or acceptor 11 did not significantly improve this step.
Nonetheless, with 13 in hand, we next attempted the removal
of the protecting groups. Unfortunately, under all conditions
attempted, we were unable to accomplish this goal. Desilylation
of 13 with buffered TBAF followed by transesterification with
sodium methoxide to remove the cyclic carbonate and ester-
(26) We note that this is not a general result in using highly advanced
glycals. For example, analogous deprotection steps have been accomplished
with the hexasaccharide glycal corresponding to Globo-H and the non-
asaccharide glycal corresponding to the KH-1 antigen (see ref 9 for a
review). This appears to be the only case thus far where we have been
unable to affect such conversions, pointing to perhaps the sensitivity of the
sialic acid and glycal units in this particular structure.
(23) (a) Lonn, H. Carbohydr. Res. 1985, 134, 105. (b) Lonn, H. J.
Carbohydr. Chem. 1987, 6, 301.
(24) See Experimental Section for characterization of this compound.
(25) Conditions used to effect this transformation were either Bu4NI/
Amberlyst H+/acetone or H2O2/AcOH/CH2Cl2.

Synthesis and Immunoconjugation of Fucosyl GM1
J. Am. Chem. Soc., Vol. 121, No. 47, 1999 10879
Figure 3. n-Pentenyl glycosides (NPG) as glycosidic donors.
In our earliest attempts at reduction to practice, glycal 13
was subjected to epoxidation under standard procedures with
3,3-dimethyldioxirane (Scheme 3). Reaction with pentenyl
alcohol and anhydrous zinc chloride29 afforded the glycoside
15 in 65% yield. Indeed, with the pentenyl glycoside in place,
global deprotection of 15 was possible. The sequence shown
in Scheme 3 furnished the peracetylated hexasaccharide lactone
16 in 46% yield ( five steps). Removal of the acetates with
sodium methoxide followed by saponification of the resulting
methyl ester yielded the target, fucosyl GM1 pentenyl glycoside,
1b. Our assignment of structure 1b is based on 1H and 13C NMR
analysis of 1b, in conjunction with characterization of inter-
mediates en route to the final structure, and is supported by
high-resolution mass spectrometry.
As noted at the outset, our goal was not only to carry out a
total synthesis of a bioconjugatable analogue of fucosyl GM1
but also to produce significant quantities of this epitope for
potential preclinical, and eventually clinical evaluation. Clearly,
the low [3 + 3] coupling yield would undercut this goal and
would thereby hamper the progress on biological evaluation.
Accordingly we explored the possibility that the pivotal [3
+ 3] coupling step might be more effective with a glycoside
rather than with a glycal at the “reducing end” of the acceptor.
Ideally, this type of modification might allow for a more efficient
total synthesis in terms of convergency, while still allowing for
smooth global deprotection. As shown in Scheme 4, we first
focused on pentenyl lactoside. For this purpose, lactose octa-
acetate was converted to the known bromide 17.30 Reaction of
this compound with pentenyl alcohol under promotion by silver
carbonate delivered the desired pentenyl glycoside, 18, on 100
g scale.31 An analogous coupling to produce 17 using silver
triflate as promoter resulted in only a 17% yield of the desired
product. Removal of the acetates yielded lactoside 19. Again
we engaged the C3′ and C4′ hydroxyl groups, this time as the
dimethyl ketal 20. This reaction, as currently conducted, is
accompanied by formation of minor amounts of 4, 6-ac-
etonide.32,33 Presumably the 4,6-isomer can be recycled into the
scheme, although at this stage we have not done so. However,
in light of the subsequent results (vide infra), the NPG method
Figure 2. Solvolysis of glycal epoxides with allyl alcohol en route to
KLH conjugation.
Through these previous studies, it was also determined that pre-
functionalization to the allylic ether at the stage of original
protection was not possible due to its instability under the
necessary dissolving metal reduction conditions in the global
deprotection sequence. The inability to affect the global depro-
tection of hexasaccharide 13 to peracetate 14 undermined
application of this methodology to the synthesis of fucosyl GM1
allyl glycoside, 1d.
Following the difficulties we encountered, it was recognized
that that the well-known pentenyl glycoside linkage, thoroughly
and elegantly developed by Fraser-Reid and associates,28 might
be used to advantage in this context. n-Pentenyl glycosides
(NPG) have been widely employed as substrates for a variety
of reactions occurring at the anomeric center of oligosaccharides.
NPGs are stable to a range of reaction conditions and reagents,
but are readily activated for glycosidation reactions by treatment
with a halogen oxidant (Figure 3). As a result of their stability
and the neutral conditions required for their activation, pentenyl
glycosides have been demonstrated to be valuable linkages for
mechanistic and synthetic studies. However, a terminal pentenyl
group could also provide a valuable handle for bioconjugation.
This linkage, in contrast to the aforementioned allyl glycoside,
should be stable under the deprotection conditions we planned
to apply, in particular the reductiVe phase of the global
deprotection sequence. Although pentenyl glycosides have not
previously been used for immunoconjugation, it seemed reason-
able to suppose that the four-carbon aldehyde derived from 1b
might be suitable for this purpose.
(27) The methodology described was applied in the synthesis of allyl
glycosides corresponding to the Globo-H, KH-I, Lewisy, and N3 antigens.
See ref 9 for a review of this methodology as applied to carbohydrate-
based vaccines.
(28) For a review of NPGs, see: (a) Fraser-Reid, B. O.; Udodong, U.
E.; Zufan, W.; Ottosson, H.; Merritt, R.; Rao, S.; Roberts, C.; Madsen, R.
Synlett 1992, 927. See also: (b) Udodong, U. E.; Madsen, R.; Roberts, C.;
Fraser-Reid, B. O. J. Am. Chem. Soc. 1993, 115, 7886. (c) Merritt, J. R.;
Fraser-Reid, B. O. J. Am. Chem. Soc. 1994, 116, 8334. (d) Fraser-Reid, B.;
Wu, Z.; Udodong, U. E.; Ottosson, H. J. Org. Chem. 1990, 55, 6068. (e)
Mootoo, D. R.; Date, V.; Fraser-Reid, B. O. J. Am. Chem. Soc. 1988, 110,
2662. (f) Mootoo, D. R.; Konradsson, P.; Fraser-Reid, B. O. J. Am. Chem.
Soc. 1989, 111, 8540 and references therein.
(29) Gordon, D.; Danishefsky, S. J. Carbohydrate Res. 1990, 206, 361.
(30) (a) Reithal, Y. J. Am. Chem. Soc. 1952, 74, 4210. (b) Dasgupta, F.,
Anderson, L. Carbohydr. Res. 1994, 264, 155.
(31) Rodriguez, E. B.; Stick, R. V. Aust. J. Chem. 1990, 43, 665.

10880 J. Am. Chem. Soc., Vol. 121, No. 47, 1999
Allen and Danishefsky
Scheme 4a
Figure 4. Inhibition of F12 antibody with synthetic 1b.
a Reagents: (a) Ag2CO3, cat. I2, PnOH, CH2Cl2, 75%; (b) NaOMe,
MeOH; (c) acetone, cat. PPTS, 44% two steps; (d) BnBr, NaH, DMF;
84%; (e) 80% AcOH: H2O, 90%; (f) 10, TMSOTf, EtCN, molecular
sieves, -40 °C, 77%.
Scheme 6
Scheme 5a
to reach 15 (R ) H). Formation of compound 15 (R ) Bn) by
the direct [3 + 3] coupling route presented in Scheme 5
represents a significant improvement in overall chemical yield
and processing. Thus, these results demonstrate that in this
particular merger, replacement of the glycal functionality by a
more stable glycosidic protecting group does indeed improve
the efficiency of the coupling step. Global deprotection under
conditions identical to those in Scheme 3 yielded the character-
ized hexasaccharide, 1b.
Since we were confident of the structure of our fully synthetic
1b, we could use the construct assembled in the laboratory to
verify the assignment of the carbohydrate sector of the natural
tumor antigen. At the immune characterization level, synthetic
1b was shown to bind to monoclonal antibody F12 in ELISA
and immune thin-layer chromatography assays. Inhibition studies
revealed that preincubation of F12 with 1b completely inhibits
reactivity of natural fucosyl GM1, 1a, with the antibody (Figure
4).34 Clearly the synthetic fucosyl GM1 pentenyl glycoside
provides the antigenic epitope with which F12 reacts on SCLC
cells.
a Reagents: (a) MeOTf, CH2Cl2:Et2O, 0 °C, 70%; (b) TBAF, AcOH,
THF; (c) NaOMe, MeOH; (d) NaOH, THF; (e) Na/NH3, THF -78
°C, then MeOH; (f) Ac2O, pyridine, DMAP, CH2Cl2, 45% five steps,
(g) steps c-d, 96%.
Attention was then directed to the final of our outlined goals.
We subjected synthetic 1b to conjugation to carrier protein KLH,
as shown in Scheme 6. The protocol35 started with ozonolysis,
thereby producing the uncharacterized aldehyde derivative. This
step was followed by coupling to KLH by using reductive
amination under the agency of sodium cyanoborohydride.
Presumably coupling of the carbohydrate had occurred with the
ꢀ-amino group of lysine residues in the KLH. Hydrolytic
carbohydrate analysis revealed approximately 331 carbohydrate
residues per molecule of KLH.36 Mouse immunization studies
with this fully synthetic vaccine, 1c, will be reported separately.
is well-justified even with this inconvenience. Perbenzylation
of 20 give 21 followed by actetonide removal with aqueous
acetic acid yielded the desired AB acceptor 22. Sialylation using
phosphite donor 10 (see Scheme 2) proceeded in comparable
yield to give the trisaccharide acceptor, 23.
In the event, coupling of donor 12 with a 2.0 molar excess
of the acceptor 23 containing the pentenyl linker proceeded with
MeOTf promotion (1.5 equiv × 2) in excellent (70%) yield (see
Scheme 5). It will be recalled that the previous [3 + 3] coupling
using the glycal acceptor 11 proceeded in low yield (23%,
Scheme 3). Moreover, subsequent transformations were required
(34) Inhibition studies were performed with increasing concentrations
of 1b and were analyzed by ELISA.
(32) Catelani, G.; Colonna, F.; Marra, A. Carbohydr. Res. 1988, 182,
297.
(35) (a) Bernstein, M. A.; Hall, L. D. Carbohydr. Res. 1980, 78, C1. (b)
Lemieux, R. U. Chem. Soc. ReV. 1978, 7, 423.
(33) See Supporting Information for details.

Synthesis and Immunoconjugation of Fucosyl GM1
J. Am. Chem. Soc., Vol. 121, No. 47, 1999 10881
103.48, 102.29, 98.32, 82.90, 81.80, 78.37, 76.50, 76.30, 75.31, 75.01,
74.89, 74.82, 73.23, 72.97, 72.66, 72.37, 69.16, 69.03, 68.69, 68.43,
68.36, 67.81, 67.08, 62.21, 52.99, 49.17, 36.41, 30.17, 28.89, 23.11,
21.08, 20.77, 60.67, 60.47; HRMS (FAB) calcd for C72H87NO23Na (M
+ Na+) 1356.5566, found 1356.5557.
Conclusions
The goals of chemical synthesis, bioconjugation, and immu-
nocharacterization of the SCLC-specific fucosyl GM1 have been
realized. The approach to this hexasaccharide relies on a highly
convergent [3 + 3] coupling reaction between the known donor
12 and acceptor 23 which carries with it a pentenyl linker for
bioconjugation. The use of a pentenyl glycoside-protecting group
at the reducing end of 1 offers clear advantages. We observed
an increase in chemical yield in the crucial coupling using
acceptor 23, with the more stable glycosidic linkage, as
compared to trisaccharide glycal 11. The NPG modification
allowed for a rather more efficient synthesis of potential
conjugation precursors. Happily, it was demonstrated that the
pentenyl linker serves as well as the allyl linker for conjugation
purposes.37 We are actively pursuing the efficacy of pentenyl
glycosides in an analogous synthesis of the MBr1 antigen Globo-
H. Reports on this work will be reported in due course.38
We have conducted the studies reported herein with the
expectation that patients immunized in an adjuvant setting with
fully synthetic vaccine conjugate 1c might produce antibodies
reactive with SCLC cells and that the production of such
antibodies might mitigate against tumor spread which might
have with it a more favorable prognosis. The preclinical results
toward this goal are underway and will be disclosed when they
become available.
Hexasaccharide 15 (R ) Bn). The thioethyl donor 12 (311 mg,
0.243 mmol) and acceptor 23 (627 mg, 0.487 mmol) were combined,
azeotroped with anhydrous benzene (5 × 5 mL), and placed under high
vacuum for 5 h. The mixture was then dissolved in 1.6 mL of CH2Cl2
and 3.2 mL of Et2O (0.05 M total), treated with freshly prepared 4 Å
molecular sieves, and cooled to 0 °C. Methyl triflate (1.5 equiv, 41
µL) was added in one portion, and the reaction stirred at 0 °C overnight.
In the morning, another 20 µL of MeOTf was added, and the reaction
was allowed to stir for an additional 2 h at 5 °C. The reaction was
quenched by the addition of solid NaHCO3, filtered through Celite with
EtOAc, concentrated, and purified by flash column chromatography
(gradient elution 25% EtOAc/hexanes f 50% f 75% EtOAc/hexanes)
to give 437 mg (70%) of hexasaccharide as a white foam and 300 mg
of recovered trisaccharide acceptor: [R]22 -29.4° (c 3.25, CHCl3);
D
IR (film CHCl3) 3285, 3028, 2940, 2865, 1794, 1749, 1690, 1220, 1090
cm-1; 1H NMR (CDCl3, 400 MHz) δ 7.74 (d, 2H, J ) 7.5 Hz), 7.34-
7.08 (m, 43H), 5.75 (m, 1H), 5.52 (d, 1H, J ) 4.7 Hz), 5.29 (app s,
1H), 5.23 (dd, 1H, J ) 9.5, 1.4 Hz), 5.15 (m, 1H), 5.02 (d, 1H, J )
9.8 Hz)4.97-4.87 (m, 5H), 4.84 (d, 1H, J ) 10.9 Hz), 4.81-4.70 (m,
5H), 4.63 (d, 1H, J ) 11.6 Hz), 4.57 (m, 3H), 4.44 (d, 1H, J ) 7.2
Hz), 4.40 (d, 1H, J ) 12.2 Hz), 4.30 (d, 1H, J ) 7.8 Hz), 4.10 (m,
2H), 3.98-3.81 (m, 12H), 3.82 (s, 3H), 3.78-3.68 (m, 7H), 3.64-
3.45 (m, 8H), 3.27 (m, 3H), 3.17 (dd, 1H), 2.80 (d, OH, 1H, J ) 2.1
Hz), 2.19 (dd, 1H, J ) 13.0, 4.5 Hz), 2.10 (m, 3H), 2.01 (s, 3H), 1.92
(s, 3H), 1.88 (s, 3H), 1.82 (s, 3H), 1.81 (s, 3H), 1.68 (m, 2H), 1.08 (d,
3H, J ) 5.4 Hz), 1.00-0.92 (m, 42H); 13C NMR (CDCl3, 100 MHz)
δ 170.61, 170.34, 170.26, 169.66, 167.78, 155.48, 138.95, 138.65,
138.63, 138.56, 138.42, 138.38, 138.27, 138.05, 132.17, 129.02, 128.59,
128.46, 128.18, 128.05, 127.91, 127.63, 127.51, 127.24, 127.09, 114.80,
103.42, 102.76, 102.45, 100.16, 99.58, 98.76, 82.87, 81.53, 79.06, 77.32,
77.24, 77.16, 75.12, 75.07, 74.95, 74.80, 73.92, 73.27, 73.04, 72.93,
72.19, 69.23, 69.14, 69.09, 67.89, 67.53, 61.76, 61.58, 61.12, 56.39,
53.60, 49.19, 35.36, 30.17, 28.89, 23.13, 20.97, 20.75, 20.62, 20.53,
17.85, 17.53, 17.33, 16.72, 11.80, 11.74; HRMS (FAB) calcd for
Experimental Section39
Trisaccharide 23. The phosphite donor 10 (1.0 g, 1.35 mmmol)
and lactosyl acceptor 22 (2.5 g, 2.90 mmol) were combined, azeotroped
with anhydrous benzene, and placed under high vacuum for 2 h. The
mixture was dissolved in anhydrous CH3CH2CN (distilled from CaH2),
freshly activated 4 Å molecular sieves were added, and the reaction
was cooled to -40 °C. A portion of TMSOTf (0.1 equiv, 27 µL) was
added, and the reaction was allowed to stir at -40 °C overnight. The
reaction was then warmed to -30 °C, and another 0.1 equiv of TMSOTf
was added. Upon stirring for an additional 2 h at -30 °C, the reaction
was quenched by the addition of solid NaHCO3 and was filtered through
a plug of Celite with the aid of EtOAc. The organic layer was washed
with saturated NaHCO3 (2 × 400 mL) and dried over MgSO4.
Evaporation of the organic layer gave a cloudy oil which was subjected
to flash column chromatography using careful gradient elution in order
to recover acceptor and product trisaccharide (20% EtOAc/hexanes f
75% EtOAc/hexanes). The product (1.35 g, 75%) was obtained as a
white foam, and 0.95 g of starting acceptor was recovered: [R]22D 2.38°
(c 1.30, CHCl3); IR (film CHCl3) 3106, 2866, 1744, 1689, 1368, 1222,
1055 cm-1; 1H NMR (CDCl3, 400 MHz) δ 7.40-7.17 (m, 25H), 5.79
(m, 1H), 5.38 (m, 1H), 5.27 (dd, 1H, J ) 8.0, 2.0 Hz), 5.08 (d, 1H, J
) 10.0 Hz), 4.95 (m, 3H), 4.86 (d, 1H, J ) 10.9 Hz), 4.75 (d, 1H, J
) 5.7 Hz), 4.72 (d, 1H, J ) 10.8 Hz), 4.68 (d, 1H, J ) 11.0 Hz), 4.56
(d, 1H, J ) 11.9 Hz), 4.54 (d, 1H, J ) 7.6 Hz), 4.44 (d, 1H, J ) 12.2
Hz), 4.39 (m, 1H), 4.32-4.25 (m, 3H), 4.06-3.88 (m, 6H), 3.79 (m,
2H), 3.72 (s, 3H), 3.65 (m, 3H), 3.54-3.44 (m, 5H), 3.35 (m, 2H),
2.66 (d, OH, 1H, J ) 3.3 Hz), 2.47 (dd, 1H, J ) 13.0, 4.7 Hz), 2.12
(m, 2H), 2.06 (s, 3H), 2.02 (m, 1H), 1.98 (s, 3H), 1.95 (s, 3H),1.85 (s,
3H), 1.83 (s, 3H), 1.71 (s, 3H) ; 13C NMR (CDCl3, 100 MHz) δ 170.77,
170.53, 170.23, 169.92, 169.87, 168.32, 139.09, 138.90, 138.61, 138.45,
138.34, 138.05, 128.27, 128.21, 137.99, 127.51, 127.42, 127.11, 114.81,
C
136H178N2O39SSi2 (M + Na+) 2574.1163, found 2574.1130.
Compound 1b. To a solution of the hexasaccharide (130 mg, 0.0509
mmol) in THF (2.0 mL) was added glacial AcOH (10.0 equiv, 29 µL)
and TBAF (1.0 M THF, 10.0 equiv, 0.509 mL). The reaction stirred at
room temperature overnight and was poured into ice water and extracted
with EtOAc (3 × 50 mL). The organic extracts were washed with
saturated NaHCO3 (50 mL) and brine (50 mL), dried over MgSO4,
and concentrated to an oil which was purified through a short plug of
silica gel with EtOAc. The resulting triol was dissolved in anhydrous
MeOH (2.5 mL), and sodium methoxide was added (0.250 mL of a
25% solution in MeOH). The reaction stirred at room temperature for
18 h, and then 0.5 mL of THF and 0.5 mL of H2O were added. Stirring
at room temperature for an additional 24 h was followed by neutraliza-
tion with Dowex-H+, filtration with MeOH washings, and concentration.
The crude material was allowed to dry under high vacuum for 1 day.
To the resulting white solid was added THF (0.5 mL) and condensed
liquid NH3 (∼10 mL) at -78 °C. Sodium (∼50 mg) was added and
the resulting blue solution stirred at -78 °C for 1.5 h. The reaction
was quenched with anhydrous MeOH (∼5 mL), brought to room
temperature, and concentrated with a stream of dry N2 to a volume of
∼2 mL. The reaction was neutralized with Dowex-H+, filtered with
MeOH washings, and concentrated to a white solid. The white solid
was dissolved in 1.0 mL of pyridine and 1.0 mL of CH2Cl2 and cooled
to 0 °C. A crystal of DMAP was added followed by addition of acetic
anhydride (1.0 mL). The ice bath was removed, and the reaction stirred
at room temperature overnight. Concentration followed by purification
by flash column chromatography (gradient elution 75% EtOAc/hexanes
f 100% EtOAc f 5% MeOH/EtOAc) gave 44 mg (46%) of 16 as a
white solid: 1H NMR (MeOH, 400 MHz) δ 8.02 (d, 1H, J ) 9.9 Hz),
7.87 (d, 1H, J ) 9.2 Hz), 5.76 (m, 1H), 5.49 (m, 1H), 5.39 (d, 1H, J
) 2.9 Hz), 5.34-5.31 (m, 2H), 5.22 (d, 1H, J ) 3.4 Hz), 5.19 (d, 1H,
J ) 4.1 Hz), 5.17 (d, 1H, J ) 3.5 Hz), 5.12-5.05 (m, 3H), 4.97 (dd,
(36) For a typical procedure, see: (a) Lloyd, K. O.; Savage, A.
Glycoconjugate J. 1991, 8, 439. (b) Hardy, M. R.; Townsend, R. R. Proc.
Natl. Acad. Sci. U.S.A. 1988, 85, 3289.
(37) The yields of carbohydrate conjugates are similar for both the allyl
linker and the pentenyl linker. The incorporation of ∼300 carbohydrates
per molecule of carrier protein is what was expected based on results using
the allyl glycoside of Globo-H.
(38) Allen, J. R.; Allen, J. G.; Zhang, X.-F.; Williams, L. J.; Zatorski,
A.; Danishefsky, S. J. Eur. J. Chem., in press.
(39) Please see attached Supporting Information for general experimental
procedures.

10882 J. Am. Chem. Soc., Vol. 121, No. 47, 1999
Allen and Danishefsky
1H, J ) 16.8, 1.7 Hz), 4.91 (dd, 1H, J ) 10.0, 1.7 Hz), 4.81-4.75 (m,
3H), 4.65-4.60 (m, 2H), 4.52 (d, 1H, J ) 7.9 Hz), 4.48-4.44 (m,
2H), 4.37 (dd, 1H, J ) 10.0, 2.5 Hz), 4.28 (dd, 1H, J ) 12.5, 2.4 Hz),
4.22-4.18 (m, 2H), 4.14-3.99 (m, 9H), 3.96-3.92 (m, 2H), 3.89 (d,
1H, J ) 2.9 Hz), 3.88-3.77 (m, 4H), 3.72-3.62 (m, 3H), 3.51-3.45
(m, 1H), 2.74 (dd, 1H, J ) 11.3, 4.5 Hz), 2.19 (s, 3H), 2.13 (s, 3H),
2.11 (s, 3H), 2.10 (s, 3H), 2.09 (s, 3H), 2.08 (s, 3H), 2.06 (s, 3H), 2.05
(s, 3H), 2.02 (s, 3H), 2.01 (s, 3H), 2.00 (s, 3H), 1.99 (s, 3H), 1.98 (s,
3H), 1.97 (s, 3H), 1.95 (s, 3H), 1.94 (s, 3H), 0.91 (s, 3H), 0.180 (s,
3H), 1.61 (m, 2H), 1.14 (d, 3H, J ) 6.4 Hz), 3 protons buried beneath
acetates (2 Pn, 1 C3ax); 13C NMR (MeOH, 100 MHz) δ 174.64, 173.64,
172.98, 172.89, 172.63, 172.56, 172.48, 172.44, 172.34, 172.27, 172.04,
171.99, 171.76, 171.73, 171.62, 171.35, 171.25, 139.23, 115.47, 104.62,
103.26, 101.86, 101.63, 100.78, 97.31, 78.22, 76.53, 75.08, 74.69, 74.29,
73.91, 73.53, 72.94, 72.71, 72.56, 72.16, 72.06, 71.89, 71.74, 70.19,
69.87, 69.33, 69.11, 68.92, 65.96, 65.65, 63.68, 63.52, 62.69, 54.01,
53.09, 50.60, 40.19, 31.09, 29.96, 24.17, 24.06, 22.73, 21.76, 21.59,
21.46, 21.20, 21.06, 20.89, 20.75, 20.63, 20.55, 16.52.
10H), 4.38-4.32 (m, 5H), 4.30 (d, 1H), 4.18 (s, 3H), 4.21-4.12 (m,
6H), 4.06 (m, 2H), 3.99 (m, 4H), 3.85 (d, 1H), 3.74 (dd, 1H), 3.61 (m,
2H), 3.52 (t, 1H), 2.63 (dd, 1H, J ) 13.9, 5.0), 2.48 (dd, 1H, J ) 13.4
Hz), 2.35 (s, 3H), 2.01 (s, 3H), 1.98 (s, 3H), 1.72 (s, 3H), 1.64 (s, 3H),
1.57 (d, 3H, J ) 6.3), 1.31-1.20 (m, 42H); 13C NMR (100 MHz,
CDCl3) δ 169.71, 169.39, 169.18, 168.70, 168.12, 166.99, 154.75,
143.47, 137.81, 137.71, 137.51, 137.42, 137.07, 131.65, 128.25, 127.52,
128.32, 127.26, 127.23, 127.19, 127.10, 126.98, 126.91, 126.83, 126.73,
126.62, 126.53, 126.36, 126.29, 101.67, 101.35, 98.69, 98.32, 98.26,
97.33, 80.48, 78.05, 77.06, 76.20, 75.50, 74.64, 74.22, 73.87, 73.49,
72.90, 72.38, 72.26, 71.93, 71.47, 71.20, 70.34, 70.17, 69.99, 69.13,
68.62, 68.10, 67.92, 67.01, 66.88, 66.68, 65.52, 60.92, 60.61, 55.51,
52.59, 48.31, 34.87, 28.68, 22.19, 19.95, 19.77, 19.68, 19.59, 16.93,
16.88, 15.79, 10.86, 10.78; HRMS (FAB) calcd for C124H162N2O37Si2-
SNa [M + Na]+ 2382.0013, found 2382.0001.
Imido-hexasaccharide 13i. Performing the above reaction with 10
equiv of MeOTf added in one portion, under otherwise identical
conditions, yields 28% of the following compound, which is much less
polar than the parent N-acetylated hexasaccharide 13. Rf 0.35 (25%
The peracetate (40 mg) was dissolved in anhydrous MeOH (2.0 mL),
and 150 µL of sodium methoxide was added (25% solution in MeOH).
The reaction stirred at room temperature for 18 h, and then 0.5 mL of
THF and 0.5 mL of H2O were added. The reaction stirred for another
24 h at room temperature. Neutralization with Dowex-H+ (pH ∼6-7)
was followed by filtration with MeOH washings, concentration, and
purification using P-2 Gel (H2O elutent) to yield 24 mg (96%) of a
1
EtOAc/hexanes); H NMR (500 MHz, C6D6) δ 8.31 (d, 2H), 7.66 (d,
2H), 7.53 (t, 4H), 7.48 (d, 2H), 7.42-7.16 (m, 31H), 6.46 (d, 1H),
6.21 (app s, 1H), 6.15 (d, 1H, J ) 4.3 Hz), 5.81 (d, 1H, J ) 9.2 Hz),
5.72 (dt, 1H, J ) 12.8, 2.4 Hz), 5.40 (m, 1H), 5.38 (d, 1H, J ) 3.5
Hz), 5.20 (d, 1H, J ) 10.2 Hz), 5.12 (t, 2H), 5.00 (m, 3H), 4.84 (d,
1H, J-6.2 Hz), 4.81 (d, 1H, J ) 4.5 Hz), 4.73 (m, 2H), 4.70 (m, 2H),
4.67 (d, 1H, J ) 2.6 Hz), 4.65 (m, 1H), 4.59 (m, 3H), 4.53-4.46 (m,
6H), 4.40 (m, 5H), 4.36 (d, 1H, J ) 3.1 Hz), 4.30 (d, 1H, J ) 3.4 Hz),
4.26 (m, 3H), 4.23 (app s, 1H), 4.20 (m, 3H), 4.11 (m, 2H), 4.04 (d,
1H, J ) 5.9 Hz), 3.99 (s, 3H), 3.92 (d, 1H, J ) 3.2 Hz), 3.87 (d, 1H,
J ) 2.9 Hz), 3.82 (d, 1H, J ) 6.5 Hz), 3.70 (m, 1H), 3.64 (s, 3H), 3.60
(d, 1H), 3.28 (t, 1H), 2.94 (dd, 1H, J ) 13.7, 4.5 Hz), 2.36 (t, 1H, J )
13.3 Hz), 2.14 (s, 3H), 1.91 (s, 3H), 1.83 (s, 3H), 1.81 (s, 3H), 1.60 (s,
3H), 1.53 (d, 3H, J ) 6.5 Hz), 1.32-1.23 (m, 42H); 13C NMR (100
MHz, CHCl3) δ 170.43, 169.30, 169.20, 168.98, 168.03, 164.74, 155.82,
144.74, 139.09, 138.75, 138.52, 138.48, 138.40, 138.39, 138.25, 138.17,
132.56, 129.22, 128.85, 128.39, 128.35, 128.30, 128.25, 128.01, 127.79,
127.71, 127.60, 127.55, 127.50, 127.48, 127.34, 102.57, 102.24, 99.69,
99.11, 98.25, 81.35, 79.09, 87.22, 75.64, 75.40, 74.90, 74.60, 74.15,
73.95, 73.50, 73.33, 72.94, 72.84, 72.52, 71.37, 71.17, 70.47, 70.17,
69.66, 69.05, 68.47, 68.11, 67.96, 67.71, 67.55, 61.91, 61.54, 61.05,
57.70, 56.50, 53.65, 52.75, 31.94, 29.71, 21.70, 20.97, 20.89, 20.64,
20.46, 20.44, 17.57, 16.81, 15.38, 14.13, 11.89, 11.80; LRMS (FAB)
1
white solid: IR 3346, 2940, 2882, 1657, 1620, 1376, 1069 cm-1; H
NMR (D2O, 400 MHz) δ 5.86 (m, 1H), 5.18 (d, 1H, J ) 4.0 Hz), 5.04
(dd, 1H, J ) 17.22, 1.7 Hz), 4.97 (dd, 1H, J ) 10.6 Hz), 4.63 (d, 1H,
J ) 7.6 Hz), 4.57 (d, 1H, J ) 7.7), 4.46 (d, 1H, J ) 7.9 Hz), 4.43 (d,
1H, J ) 8.1 Hz), 4.15 (m, 1H), 4.09-4.02 (m, 3H), 3.94-3.84 (m,
5H), 3.80-3.63 (m, 18H), 3.60-3.53 (m, 6H), 3.47 (dd, 1H, J ) 10.3,
1.8), 3.32 (t, 1H), 3.26 (t, 2H), 2.62 (dd, 1H, J ) 13.4, 4.3 Hz), 2.09
(m, 2H), 1.98 (s, 6H), 1.86 (m, 1H), 1.67 (m, 2H), 1.15 (d, 3H, J )
6.5 Hz); 13C NMR (D2O, 100 MHz) δ 176.29, 175.43, 175.16, 139.97,
115.99, 104.38, 103.77, 103.30, 103.22, 102.25, 100.35, 79.67, 78.12,
77.65, 77.03, 76.06, 75.94, 75.62, 75.44, 75.24, 74.85, 74.19, 74.01,
73.45, 73.01, 71.15, 70.72, 70.32, 69.87, 69.64, 69.25, 67.93, 64.01,
62.29, 62.07, 61.63, 61.29, 52.79, 52.70, 50.04, 38.45, 30.53, 29.17,
23.89, 23.23, 16.53; HRMS (FAB) calcd for C48H79N2O33Na2 [M - H
+ 2Na]+ 1257.4360, found 1257.4337.
Glycal Hexasaccharide 13. The thioethyl donor 12 (120 mg, 0.0938
mmol) and acceptor 11 (122 mg, 0.108 mmol) were combined,
azeotroped with anhydrous benzene (5 × 5 mL), and placed under high
vacuum overnight. The mixture was dissolved in a 2:1 mixture of Et2O:
CH2Cl2 (2.7 mL total), 4 Å molecular sieves were added and the mixture
stirred at room temperature for 1 h. The reaction was cooled to 0 °C,
and 1.0 equiv of MeOTf (0.020 mL) was added. After 4 h at 0 °C
another equivalent of MeOTf was added (0.020 mL), and the reaction
continued to stir for another 4 h at 10 °C. The reaction was quenched
with solid NaHCO3, filtered through Celite with additional EtOAc (100
mL), and concentrated. The resulting mixture was purified by flash
column chromatography to give 50 mg (23%) of the hexasaccharide
glycal 13 and 85 mg of starting acceptor, 11: Rf 0.35 (66% EtOAc/
C
125H164N2O37SSi2Na 2373 [M + Na]+.
Acknowledgment. This research was supported by the
National Institutes of Health (Grant numbers: AI16943 and
CA28824). Postdoctoral fellowship support is gratefully ac-
knowledged by J.R.A. (NIH GM19578). We gratefully ac-
knowledge Dr. George Sukenick (S.K.I. Core Grant: CA-08748)
of the NMR Core Facility for mass spectral and NMR analyses.
This paper is dedicated to Dr. Bert Fraser-Reid on the occasion
of his 65th birthday.
1
Supporting Information Available: Experimental proce-
dures for compounds 18, 20, 21, and 22, 1H and 13C spectra for
hexanes); H NMR (500 MHz, C6D6) δ 8.31 (d, 2H), 7.62 (d, 2H),
7.52 (m, 4H), 7.45 (d, 2H), 7.40-7.15 (m, 31H), 6.47 (d, 1H, J ) 6.3
Hz), 6.28 (apparent s, 1H), 6.09 (d, 1H, J ) 3.8 Hz), 5.72 (m, 1H),
5.55 (dd, 1H, J ) 9.3, 1.2 Hz), 5.51 (d, 1H, J ) 3.5 Hz), 5.22 (d, 1H,
J ) 10.8 Hz), 5.15 (s, 1H), 5.13-5.06 (m, 3H), 5.05 (d, 1H, J ) 8.1
Hz), 5.02 (m, 1H), 4.98 (d, 1H, J ) 10.8 Hz), 4.85 (d, 1H, J ) 10.6
Hz), 4.82 (d, 1H, J ) 9.4 Hz), 4.73-4.66 (m, 8H), 4.55-4.34 (m,
1
compounds 13, 15, 16, 18, 20-23, 1b and H spectra for 13i
and 19 are available (PDF). This information is available free
of charge via the Internet at http://pubs.acs.org.
JA992594F