Enantiospecific Formal Total Syntheses of (-)-Salicylihalamides A and B from D-Glucose and L-Rhamnose

Article abstract of DOI:10.1021/ol035630v

(Matrix Presented) Two formal chiral pool syntheses of the (-)-salicylihalamides A and B were achieved from commercially available 1,2,5,6-diacetone-D-glucose and L-rhamnose.

Full text of DOI:10.1021/ol035630v

ORGANIC
LETTERS
2003
Vol. 5, No. 21
4007-4009
Enantiospecific Formal Total Syntheses
of (−)-Salicylihalamides A and B from
D-Glucose and L-Rhamnose
KyoungLang Yang, Torsten Haack, Burchelle Blackman, Wibke E. Diederich,
Subho Roy, Srinivas Pusuluri, and Gunda I. Georg*
Department of Medicinal Chemistry, Center for Cancer Experimental Therapeutics and
Drug DiscoVery Program, Higuchi Biosciences Center, UniVersity of Kansas,
1251 Wescoe Hall DriVe, Lawrence, Kansas 66045-7582
georg@ku.edu
Received August 27, 2003
ABSTRACT
Two formal chiral pool syntheses of the (−)-salicylihalamides A and B were achieved from commercially available 1,2,5,6-diacetone-D-glucose
and L-rhamnose.
The salicylihalamides A and B1 belong to a family of recently
discovered natural products that contain a salicylic acid
moiety, a macrolactone, and an unusual enamide side chain.
This class of molecules, which includes the oximidines I and
II,2 the lobatamides A-F,3 and apicularens A and B,4 shows
considerable potency and differential cytotoxicity in the 60-
cancer cell line assay of the NCI.5 The salicylihalamides 1
and 2, isolated from Haliclona sp. in 1997,1 have received
substantial attention, resulting in several partial and total
syntheses.6 In our initial approach toward the synthesis of
the macrocyclic core of the salicylihalamides, we chose a
chiral pool strategy using commercially available 1,2,5,6-
diacetone-D-glucose as the starting material.6e However, since
the absolute stereochemistry of the salicylihalamides was
revised by total synthesis,6a we needed to modify our strategy
in order to prepare the naturally occurring salicylihalamides.
Herein, we report two chiral pool approaches toward the
synthesis of the (-)-salicylihalamides A and B. One of them
is also based on using 1,2,5,6-diacetone-D-glucose6e as the
starting material, while the other one employs L-rhamnose,
both leading to the formal total synthesis of the salicylihala-
mides.
(6) (a) Wu, Y.; Esser, L.; De Brabander, J. K. Angew. Chem., Int. Ed.
2000, 39, 4308-4310. (b) Fu¨rstner, A.; Thiel, O. R.; Blanda, G. Org. Lett.
2000, 2, 3731-3734. (c) Wu, Y.; Seguil, O. R.; De Brabander, J. K. Org.
Lett. 2000, 2, 4241-4244. (d) Labrecque, D.; Charron, S.; Rej, R.; Blais,
C.; Lamothe, S. Tetrahedron Lett. 2001, 42, 2645-2648. (e) Georg, G. I.;
Ahn, Y. M.; Blackman, B.; Farokhi, F.; Flaherty, P. T.; Mossman, C. J.;
Roy, S.; Yang, K. L. J. Chem. Soc., Chem. Commun. 2001, 255-256. (f)
Smith, A. B.; Zheng, J. Synlett 2001, 1019-1023. (g) Snider, B. B.; Song,
F. Org. Lett. 2001, 3, 1817-1820. (h) Fu¨rstner, A.; Dierkes, T.; Thiel, O.
R.; Blanda, G. Chem. Eur. J. 2001, 7, 5286-5298. (i) Wu, Y.; Liao, X.;
Wang, R.; Xie, X.; De Brabander, J. K. J. Am. Chem. Soc. 2002, 124, 3245-
3253. (j) Smith, A. B.; Zheng, J. Tetrahedron 2002, 58, 6455-6471. (k)
Bauer, M.; Maier, M. M. Org. Lett. 2002, 4, 2205-2208. (l) Holloway, G.
A.; Hu¨gel, H.; Rizzacasa, M. A. J. Org. Chem. 2003, 68, 2200-2204.
(1) Erickson, K. L.; Beutler, J. A.; Cardellina, J. H., II; Boyd, M. R. J.
Org. Chem. 1997, 62, 8188-8192.
(2) Kim, J. W.; Shin-ya, K.; Furihata, K.; Hayakawa, Y.; Seto, H. J.
Org. Chem. 1999, 64, 153-155.
(3) McKee, T. C.; Galinis, D. L.; Pannell, L. K.; Cardellina, J. H., II;
Lakkso, J.; Ireland, C. M.; Murray, L.; Capon, R. J.; Boyd, M. R. J. Org.
Chem. 1998, 63, 7805-7810.
(4) Jansen, R.; Kunze, B.; Reichenbach, H.; Ho¨fle, G. Eur. J. Org. Chem.
2000, 6, 913-919.
(5) Boyd, M. R.; Farina, C.; Belfiore, P.; Gagliardi, S.; Kim, J. W.;
Hayakawa, Y.; Beutler, J. A.; McKee, T. C.; Bowman, B. J.; Bowman, E.
J. J. Pharmacol. Exp. Ther. 2001, 297, 114-120.
10.1021/ol035630v CCC: $25.00 © 2003 American Chemical Society
Published on Web 09/20/2003

In our first approach, 1,2,5,6-diacetone-D-glucose was
transformed into alcohol 3 in five steps as described
previously.7 To obtain the correct stereochemistry at C12,
we first converted the alcohol into the corresponding bromide
and then introduced the allyl group, giving rise to intermedi-
ate 4 (Scheme 1).
Scheme 2a
Scheme 1a
a Reagents and conditions: (a) DIBAL-H, Et2O, -78 °C, 30 min,
93%. (b) PPh3dCHCN, PhMe, 80 °C, 2 h, 97%. (c) PNBA, DEAD,
PPh3, PhMe, -30 °C to rt, 24 h, 74%.
10, which affords 12, proceeded in very good yield (89%).
The same reaction failed with the unprotected alcohol
because the hydroxy group tends to add into the acryl nitrile
in a Michael fashion under the given reaction conditions.
a Reagents and conditions: (a) CBr4, PPh3, THF, 0 °C to rt, 6 h,
90%. (b) (allyl)2Cu(CN)Li2, THF, Et2O, -78 °C, 30 min, 83%. (c)
AcOH, H2O, 70 °C, 3 h, 85%. (d) AgCO3/Celite, PhH, reflux, 1 h,
81%. (e) 6, DCC, CH2Cl2, rt, 24 h, 95%.
Scheme 3a
Deprotection of the ketal and subsequent oxidation led to
building block 5. Esterification of 5 with the aromatic
fragment 68 using DCC provided compound 7 with the
correct stereochemistry at C15. Reduction of lactone 7 to
the corresponding lactol and subsequent Wittig olefination
with (triphenylphosphoranylidene)acetonitrile resulted in
acryl nitrile 8 as a separable 3.6:1 mixture of E:Z isomers
(Scheme 2). The conversion of (E)-8 into 9 under Mitsunobu
conditions using p-nitrobenzoic acid (PNBA) proceeded in
good yield (74%). Interestingly, (Z)-8 did not react in
satisfactory yields under the same conditions, presumably
due to steric hindrance. Reduction of acryl nitrile 9, leading
to the saturated nitrile 11, was carried out with hydrido-
(triphenylphosphine)copper(I) hexamer9 (Scheme 3). The
moderate yield of this reaction (45%) can be attributed to
the presence of the nitro functionality of the PNB group,
interfering with the reducing agent. This problem could be
avoided by replacing the PNB group with an acetate (10).
Although this procedure adds two steps to the synthesis, it
saves material because the double-bond reduction of acetate
a Reagents and conditions: (a) K2CO3, MeOH, rt, 3 h, 95%. (b)
Ac2O, TEA, CH2Cl2, 2 h, rt, 97%. (c) [CuHPPh3]6, PhH, H2O,
reflux, 30 min, 45% for 11, 89% for 12. (d) (PCy3)2Cl2RudCHPh,
CH2Cl2, reflux, 3 h, 89% (E/Z ) 18:1) for 13, 87% (E/Z ) 26:1)
for 14. (e) BBr3, CH2Cl2, -78 °C, 30 min, 85% from 13, 89%
from 14. (f) K2CO3, MeOH, RT, 5 h, 95% (both). (g) TBSOTf,
2,6-lutidine, CH2Cl2, rt, 1 day, 82%.
(7) (a) Lee, W. M.; Chan, W. H.; Weng, H. C.; Wong, M. S. Synth.
Commun. 1989, 19, 547-552. (b) Iacono, S.; Rasmussen, R.; Card, P. J.
Smart, B. E. Organic Syntheses; Wiley: New York, 1990; Collect. Vol.
VII, pp 139-141. (c) Patroni, J. J.; Stick, R. V. Aus. J. Chem. 1978, 31,
445-446.
(8) Yang, K. L.; Ahn, Y.; Blackman, B.; Diederich, W. E.; Flaherty, P.
T.; Mossman, C. J.; Roy, S.; Georg, G. I. J. Org. Chem 2003, in press.
(9) Mahoney, W. S.; Brestensky, D. M.; Stryker, J. M. J. Am. Chem.
Soc. 1988, 110, 291-293.
4008
Org. Lett., Vol. 5, No. 21, 2003

The ring-closing metathesis of 11 and 12 using the first-
generation Grubbs catalyst afforded the cyclized products
13 and 14 in good yields (87-89%) and excellent E:Z ratios
(18:1 and 26:1, respectively). Deprotection of each com-
pound, 13 and 14, provided compound 15, which in
comparison with previously synthesized enantiomer ent-15,8
exhibited identical physical properties except opposite optical
rotation. TBS protection generates the fully protected
compound 16 that can be used for the enamide side chain
introduction. In our previous report, we converted the nitrile
functionality of TIPS- and TBDPS-protected ent-15 to the
corresponding aldehyde, which matches the intermediate
synthesized by Labrecque et al.,6d and thus constitutes a
formal synthesis of the (-)-salicylihalamides.
Scheme 5a
Our second approach toward the synthesis of the salicyli-
halamides employs L-rhamnose as the chiral source, which
was first converted to known intermediate 17 in four steps
using a modified literature procedure (Scheme 4).10 The
Scheme 4a
a Reagents and conditions: (a) 6, EDCI, DMAP, CH2Cl2, rt, 1
day, 90%. (b) (i) DIBAL-H, toluene, -78 °C, 15 min; (ii) HCl,
dioxane, MeOH, rt, 4 h, 70%. (c) Tf2O, pyridine, 0 °C, 15 min. (d)
(allyl)2Cu(CN)Li2, -78 °C, 30% (two steps). (e) BCl3, CH2Cl2,
-78 °C, 30 min, 72%. (f) PPh3dCHCN, PhMe, 75 °C, 1 h, 98%.
(g) [CuHPPh3]6, PhH, 0 °C to rt, 15 min, 80%. (h) TBSOTf, 2,6-
lutidine, CH2Cl2, 0 °C to rt, 4 h, 91%. (i) (PCy3)2Cl2RudCHPh,
CH2Cl2, reflux, 3 h, 81%.
a Reagents and conditions: (a) BaCO3, Br2, H2O, rt, 1 day, 99%.
(b) BzCl, pyridine, 0 °C to rt, 99%. (c) H2 (50 psi), 10% Pd/C,
TEA, EtOAc, rt, 1 day, 99%. (d) (i) NaOMe, MeOH, rt, 3 h; (ii)
1,4-dioxane, H2SO4, rt, 1 h, 87%.
coupling reaction between 17 and the aromatic building block
6 (Scheme 5) revealed that the secondary hydroxy function
attached to the ring showed significantly higher reactivity
than the secondary hydroxy group of the side chain.
Therefore, the reaction gave rise to the desired intermediate
18 in 90% yield (Scheme 5). Reduction of lactone 18 and
methyl acetal formation in a one-pot procedure resulted in
19, which was obtained as a mixture of anomers. Triflate
activation of the secondary alcohol in 19 and its reaction
with a higher order allylcuprate formed the corresponding
allylated compound in moderate yield (30% over two steps).
Restricted reactivity, which might be due to the presence of
the anomeric methoxy group, and competing ester cleavage
seem to be the limiting factors in this reaction. Deprotection
of the allylated product gave compound 20, which was then
subjected to a Wittig two-carbon elongation using (triphen-
ylphosphoranylidene) acetonitrile leading to the unsaturated
nitrile. Subsequent reduction and double TBS protection gave
compound 21, which on treatment with the first generation
Grubbs catalyst resulted in the formation of compound 16,
as well as the undesired (Z)-isomer in a 4.7:1 ratio of (E)-
and (Z)-isomers.
In summary, we have achieved two efficient formal chiral-
pool syntheses of the salicylihalamides A and B. In the first
approach, the synthesis of macrocycle 16 from known
alcohol 3 was completed in 13 steps and 7% overall yield
or in 15 steps in 12% overall yield. The second approach
allows the synthesis of 16 from known diol 17 in 10 steps
and 8% overall yield.
Acknowledgment. We are grateful for financial support
from the NIH (CA84173). This work was also supported by
a fellowship (T.H.) within the Postdoctoral Program of the
German Academic Exchange Service (DAAD), a fellowship
from an NIH training grant (GM07775), a fellowship from
the American Foundation for Pharmaceutical Education
(B.B.), and a postdoctoral fellowship from the Deutsche
Forschungsgemeinschaft (W.E.D.).
1
Supporting Information Available: Physical data, H
NMR spectra, and 13C NMR spectra of compounds 15 and
16. This material is available free of charge via the Internet
at http://pubs.acs.org.
(10) (a) Bols, M.; Lundt, I.; Ottosen, E. R. Carbohydr. Res. 1991, 222,
141-149. (b) Han, O.; Liu, W. Tetrahedron Lett. 1987, 28, 1073-1076.
(c) Varela, O.; Cirelli, A. F.; Lederkremer, R. M. Carbohydr. Res. 1980,
85, 130-135.
OL035630V
Org. Lett., Vol. 5, No. 21, 2003
4009