Convergent C-glycolipid synthesis via the Ramberg-Baecklund reaction: active antiproliferative glycolipids.

Article abstract of DOI:10.1021/ol991211f

[formula: see text] A novel methodology has been developed, employing the Ramberg-Baecklund rearrangement and ionic hydrogenation to synthesize C-glycosides with high stereoselectivity at the anomeric center. The C-glycolipid 14b exhibits antiproliferativ

Full text of DOI:10.1021/ol991211f

ORGANIC
LETTERS
1999
Vol. 1, No. 13
2149-2151
Convergent C-Glycolipid Synthesis via
the Ramberg−Ba1cklund Reaction:
Active Antiproliferative Glycolipids
,†
Guangli Yang,† Richard W. Franck,* Hoe-Sup Byun,‡ Robert Bittman,‡
Pranati Samadder,§ and Gilbert Arthur§
Department of Chemistry, Hunter College/CUNY, 695 Park AVenue,
New York, New York 10021, Queens College/CUNY, Flushing, New York 11367, and
Department of Biochemistry and Molecular Biology, UniVersity of Manitoba,
Winnipeg, Manitoba, Canada R3E 0W3
rfranck@shiVa.hunter.cuny.edu
Received November 2, 1999
ABSTRACT
A novel methodology has been developed, employing the Ramberg−Ba1cklund rearrangement and ionic hydrogenation to synthesize C-glycosides
with high stereoselectivity at the anomeric center. The C-glycolipid 14b exhibits antiproliferative properties similar to those of O-glycoside
analogue 14a.
In the field of C-glycoside synthesis,1 C-glycolipids are
becoming interesting targets. Recent papers report linear
syntheses of C-glycolipids from simple C-glycofragments.2
Two groups, ours3 and Taylor’s,4 simultaneously described
a new convergent procedure for preparing C-glycosides via
exo-glycals, which are derived from sulfonyl glycosides using
the Ramberg-Ba¨cklund (RB) rearrangement. The key step
for both groups was the formation of an exo-glycal using a
one-pot procedure originally developed by Meyers,5 which
uses a halocarbon and basesin our examples, CF2Br2 and
KOH/Al2O3 at room temperature.6 We found that the RB
conditions using CF2Br2 (bp 23 °C) failed to afford good
yields of the RB product, if the sulfones did not have one
benzylic group. The simple expedient of using CF2BrCF2Br
(bp 47 °C) at reflux solved this problem,7 as we now
describe.
† Hunter College/CUNY.
‡ Queens College/CUNY.
§ University of Manitoba.
(1) (a) Postema, M. H. D. C-Glycoside Synthesis; CRC Press: Boca
Raton, FL, 1995. (b) Levy, D. E.; Tang, C. The Chemistry of C-Glycosides;
Pergamon: Oxford, U.K., 1995. (c) Beau, J.-M.; Gallagher, T. Nucleophilic
C-Glycosyl Donors for C-Glycoside Synthesis. Top. Curr. Chem. 1997,
187, 1-54. (d) Nicotra, F. Synthesis of C-Glycosides of Biological Interest.
Top. in Curr. Chem. 1997, 187, 55-83. (e) Du, Y.; Linhardt, R. J.; Vlahov,
I. R. Tetrahedron 1998, 54, 9913-9959.
(2) (a) Cipolla, L.; Nicotra, F.; Vismara, E.; Guerrini, M. Tetrahedron
1997, 53, 6163-6170. (b) Gurjar, M. K.; Reddy, R. Carbohydr. Lett. 1997,
2, 293-298. (c) Dondoni, A.; Perrone, D.; Turturici, E. J. Org. Chem. 1999,
64, 5557-5564.
(3) Belica, P. S.; Franck, R. W. Tetrahedron Lett. 1998, 39, 8225-
8228.
We first turned our attention to the use of exo-glycals for
the convergent preparation of the model C-glycolipid 5. In
(5) Meyers, C. Y.; Malte, A. M.; Matthews, W. S. J. Am. Chem. Soc.
1969, 91, 7510-7512.
(6) Chan, T.-L.; Fong, S.; Man, T.-O.; Poon, C. D. J. Chem. Soc., Chem.
Commun. 1994, 1771-1772.
(7) Griffin, F. K.; Paterson, D. E.; Taylor, R. J. K. Angew. Chem., Int.
Ed. 1999, 38, 2939-2942. This article reports the use of CCl4 at 60 °C to
solve the reactivity problem. However, earlier works report contamination
of the RB alkenes with dichlorocyclopropanes formed from the decomposi-
tion of CCl4.
(4) Griffin, F. K.; Murphy, P. V.; Paterson, D. E.; Taylor, R. J. K.
Tetrahedron Lett. 1998, 39, 8179-8182.
10.1021/ol991211f CCC: $18.00 © 1999 American Chemical Society
Published on Web 12/04/1999

our synthesis (Scheme 1), the benzylated thioglycoside 2 was
made from the benzylated 2-deoxyglucose derivative 1 via
in an axial location so that it could be eliminated. Therefore,
we prepared the bicyclic galactosyl sulfone 9 as shown in
Scheme 2 on the basis of the assumption that any twisting
Scheme 1a
Scheme 2a
a Reagents and conditions: (a) (1) CH3OCH2CO2H, DCC, ether,
87%, (2) C18H37SH, Yb(OTf)3, CH3CN, 60 °C, 89%; (b) MMPP,
EtOH/THF/H2O, 55 °C, 95%; (c) C2F4Br2, KOH/Al2O3, reflux; (d)
Et3SiH, TFA, CH2Cl2, -70 °C.
a Reagents and conditions: (a) C2F4Br2, KOH/Al2O3, reflux, 83%;
(b) (1) TMSCl, MeOH, 64%, (2) NaH, 1 equiv of BnBr, THF,
65%; (c) i-Pr2SiHCl, CH2Cl2, imidazole, 80%; (d) (1) Ac2O,
pyridine, DMAP, 93%, (2) Et3SiH, BF3‚Et2O, CH2Cl2, 20 °C, 95%,
(3) K2CO3, MeOH, 95%; (e) CH2Cl2, TFA/BF3‚Et2O (4:1), 64%.
the Inanaga-Yb(OTf)3 method.8 The sulfide 2 was oxidized
using MMPP to form sulfone 3 in good yield. Initial attempts
with Chan’s variation6 of the one-pot RB conditions, using
CF2Br2 and KOH/Al2O3, failed. When CF2BrCF2Br was used
at reflux in place of CF2Br2, the reaction worked quite well
with â-sulfone to afford the RB alkene 4 in 85% yield (Z:E
) 1:1). For the anomeric R-sulfone 3, there was very little
RB reaction, with the recovered material being unchanged.
The product enol ether 4 was a fairly sensitive material;
therefore, it was reduced by ionic hydrogenation using
Et3SiH/TFA at -70 °C9 to afford 5 in 60% yield and
overreduced 6 in 7% yield. It was observed that the product
6, due to the Ferrier rearrangement10 before reduction, was
the major product when the reaction temperature was
increased to -20 °C.
or flipping motion necessary to force the 2-benzyloxy group
axial would be difficult. With this modification, the RB
reaction took place in good yield to form 10. This material
could then be processed to form either R- or â-C-glycoside
stereoselectively. Thus, intermolecular ionic hydrogenation
conditions cleanly produced â-C-glycoside 13â. Alterna-
tively, intramolecular hydride transfer from 12, the hydro-
silylation product of 11, afforded R-C-glycoside 13R.9
Application of our methodology to the preparation of the
C-glycolipid 14b of 2-deoxyglucosyl glyceride was an
important goal of this work. O-Glycolipid 14a is an active
A challenging test of the CF2BrCF2Br reflux modification
was the galactose-derived sulfone 7. Here our conditions led
to elimination, forming endocyclic sulfonylglycal 8 with no
trace of the desired RB product. It should be noted that, in
analogue of the antiproliferative lead compound ET-18-OCH3
(edelfosine),11 which itself is a platelet-activating factor
analogue. O-Glycolipid 14a was found to specifically inhibit
the growth of tumor cells, tumor cell invasion, and metasta-
sis.12 Our rationale for preparing C-glycoside 14b was to
enhance the metabolic stability compared to 14a, and our
hope was that there would not be a significant decrease of
antiproliferative activity.
our earlier work,3 neither mannose- nor glucose-derived
sulfones gave elimination when treated with the CF2Br2 (bp
23 °C) reagent. We reasoned that a distortion of the galactose
ring must have occurred so as to put the 2-benzyloxy group
(8) Inanaga, J.; Tokoyama, Y.; Hanamoto, T. Tetrahedron Lett. 1993,
34, 2791-2794.
(11) Bittman, R.; Arthur, G. In Liposomes: Rational Design; Janoff, A.
S., Ed.; Marcel Dekker: New York, 1999; pp 125-144.
(12) Marino-Albernas, J. R.; Bittman, R.; Peters, A.; Mayhew, E. J. Med.
Chem. 1996, 39, 3241-3247.
(9) McCombie, S. W.; Ortiz, C.; Cox, B.; Ganguly, A. K. Synlett 1992,
541-547.
(10) Ferrier, R. J.; Prasad, N. J. Chem. Soc. C 1969, 570-575.
2150
Org. Lett., Vol. 1, No. 13, 1999

The synthesis of the lipid (S)-4-O-hexadecyl-3-O-benzoyl-
1-bromobutane (17) was easily accomplished (Scheme 3)
products and debenzylation proceeded to form both 2-deoxy
C-glycolipid 14b and 2,3-dideoxy C-glycolipid 23.18
Table 1 presents the data for the in vitro biological
evaluation of the growth inhibitory properties of O-glycoside
Scheme 3
Table 1. Growth Inhibitory Properties of 14a,b and 23: IC50
Values for Inhibition of Cell Proliferationa
IC50 (µM)
breast cancer cell line
14a
14b
23
MCF-7
6.9312
25.6
12.2
34.4
40.0
21.2
21.9
24.2
27.8
MDA-MB-435
MDA-MB-468
MDA-MB-231
a The IC50 values for 14b and 23 were determined as described in ref
19. Briefly, exponentially growing cells were incubated with the drugs (0-
60 µM), and the increase in cell numbers after 48 h was determined and
expressed as a percentage of the controls, which had no drug. The IC50
value for 14a was determined after a 72 h incubation period using the
sulforhodamine assay described in ref 12.
with (S)-(-)-1,2,4-butanetriol (15) as starting material. This
procedure is based on selective protection of 15 followed
by alkylation. Ether 16 was treated with NBS and BaCO3 at
reflux to afford bromide 17 in good yield.13 The thioacetate
19 can be made easily in two steps14,15 from glucal 18. The
bromide 17 was treated with thioacetate 19 in cysteamine
and dithioerythritol (DTE)16 to afford a separable R and â
mixture (R:â ) 30:70) of thioglycoside 20 in 64% yield
(Scheme 4). Deprotection using NaOMe, followed by O-
14a, C-glycoside 14b, and C-glycoside 23 on human tumor
cells. The IC50 values19 (drug concentrations required to
inhibit growth by 50%) indicate that the C-glycoside
analogues show antiproliferative properties similar to those
of O-glycoside analogue 14a. Thus, the activities are not
dependent on the glycosidic oxygen in the drug.
Scheme 4a
Acknowledgment. Support for the research at Hunter has
come from NIH Grant GM 51216, PSC-CUNY grants, and
RCMI Grant NIH RR 03037, which supports the instrumen-
tation infrastructure of the Chemistry Department at Hunter.
OL991211F
(13) Hanessian, S. Org. Synth. 1987, 65, 243-247.
(14) Lesimple, P.; Beau, J. M.; Sinay, P. Carbohydr. Res. 1987, 171,
289-300.
(15) Blanc-Musser, M.; Defaye, J.; Driguez, H. Carbohydr. Res. 1978,
67, 305-328.
(16) Moreau, V.; Norrild, J. C.; Driguez, H. Carbohydr. Res. 1997, 300,
271-277.
(17) Representative Ramberg-Ba¨cklund rearrangement: preparation of
exo-glycal 22. A sample of sulfone 21 (0.135 g, 0.16 mmol, â-isomer) was
dissolved in 3 mL of C2F4Br2 and 1 mL of t-BuOH at 47 °C. Then KOH/
Al2O3 (0.25 g, 50 wt %) was added. The reaction mixture was heated at
reflux for 6 h. The mixture was filtered through a pad of Celite, and the
solids were washed with CH2Cl2. The residue was purified by column
chromatography on silica gel, with 10% EtOAc-PE as eluent, to afford 74
mg (60%) of alkene 22 (Z:E ) 1:1).
(18) C-Glycolipid 14b: white solid, mp 62-64 °C. [R]D ) -3.47° (c
) 1.5, CHCl3). Anal. Calcd for C27H54O6: C, 68.35; H, 11.39. Found: C,
68.12; H, 11.12. 1H NMR (500 MHz, CD2Cl2): δ 3.80 (dd, H, J ) 3.2,
11.4 Hz, H6), 3.69 (dd, 1H, J ) 5.1, 11.4 Hz, H6), 3.61 (m, 1H, H3), 3.45
(m, 1H, H1), 3.42-3.38 (m, 4H), 3.35 (s, 3H, -OCH3), 3.29 (m, 2H), 3.22
(m, 1H, H5), 2.47 (broad peak), 2.00 (dd, 1H, J ) 4.8, 12.5 Hz, H2), 1.63-
1.48 (m, 9H), 1.34-1.29 (m, 26H), 0.88 (t, 3H, J ) 7.0 Hz, CH3). 13C
NMR (75 MHz, CD2Cl2): δ 80.53, 79.81, 76.29, 73.60, 73.49, 73.36 (C-
1), 72.20, 63.18, 57.91, 39.84, 32.61, 31.99, 30.39, 30.20, 30.04, 28.39,
26.83, 23.38, 14.56. C-Glycolipid 23: white solid, mp 47-48 °C. Anal.
a Reagents and conditions: (a) (1) HCl, toluene, (2) KSAc,
HMPA, 74%; (b) 17, cysteamine, DTE, HMPA, 64%; (c) (1)
NaOMe, MeOH, 96%, (2) NaH, MeI, THF, 90%, (3) MMPP, 55
°C, 94%; (d) C2F4Br2, KOH/Al2O3, reflux, 60%; (e) (1) Et3SiH,
TFA, -70 °C, (2) H2, 10% Pd/C, MeOH.
1
Calcd for C27H54O5: C, 70.69; H, 11.86. Found: C, 70.48; H, 11.65. H
NMR (500 MHz, C6D6):δ 3.77 (m, 1H), 3.70 (m, 1H), 3.42 (m., 2H), 3.34-
3.30 (m, 4H), 3.27 (s, 3H, -OCH3), 3.23 (m, 1H), 3.14 (m, 1H), 1.93 (broad
peak), 1.73-1.54 (m, 8H), 1.41-1.05 (m, 30H), 0.87 (t, 3H, -CH3). 13C
NMR (75 MHz, CDCl3): δ 81.99, 80.64, 77.97, 73.44, 72.42, 68.56, 64.33,
58.26, 33.43, 32.69, 32.07, 31.69, 30.40, 30.25, 30.12, 28.44, 26.90, 23.46,
14.89.
methylation and oxidation, afforded sulfone 21. The Ram-
berg-Ba¨cklund rearrangement of 21 proceeded smoothly
with â-sulfone to afford alkene 22 (Z:E ) 1:1) in 60%
yield.17 Ionic hydrogenation followed by separation of the
(19) Samadder, P.; Byun, H.-S.; Bittman, R.; Arthur, G. Anticancer Res.
1998, 18, 465-470.
Org. Lett., Vol. 1, No. 13, 1999
2151