Total synthesis of puerarin, an isoflavone C-glycoside

Article abstract of DOI:10.1016/S0040-4039(03)01715-5

We completed the first total synthesis of puerarin (1), an isoflavone C-glycoside. The key intermediate, β-D-glucopyranosyl-2,6-dimethoxybenzene (9), was obtained by coupling of a lithiated aromatic reagent (3) with pyranolactone (2) in 56% yield. Condensation of (16) with p-methoxybenzaldehyde gave the chalcone (17). The protected chalcone (18) was cyclized to (19) in the presence of Tl(NO₃)₃. Demethylation of (19) was accomplished by refluxing with TMSI in CH₃CN to give puerarin (1).

Full text of DOI:10.1016/S0040-4039(03)01715-5

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 6857–6859
Total synthesis of puerarin, an isoflavone C-glycosideꢀ
David Y. W. Lee,* Wu-Yan Zhang and Vishnu Vardhan R. Karnati
Bioorganic and Natural Products Laboratory, Harvard Medical School, McLean Hospital, 115 Mill Street, Belmont, MA 02478,
USA
Received 10 June 2003; revised 10 July 2003; accepted 12 July 2003
Abstract—We completed the first total synthesis of puerarin (1), an isoflavone C-glycoside. The key intermediate, b-D-glucopyran-
osyl-2,6-dimethoxybenzene (9), was obtained by coupling of a lithiated aromatic reagent (3) with pyranolactone (2) in 56% yield.
Condensation of (16) with p-methoxybenzaldehyde gave the chalcone (17). The protected chalcone (18) was cyclized to (19) in the
presence of Tl(NO3)3. Demethylation of (19) was accomplished by refluxing with TMSI in CH3CN to give puerarin (1).
© 2003 Published by Elsevier Ltd.
Puerarin (1), a naturally occurring isoflavone C-gly-
coside, was isolated from Pueraria lobota,1 one of the
most popular Chinese herbal medicine, traditionally
used to reduce febrile symptoms,2 and also used as an
anti-inebriation agent.3 It also has hypoglycemic activ-
ity,4 and increases coronary artery blood flow.5 Recent
studies showed that puerarin suppresses alcohol drink-
ing in alcohol preferring (P) rats, high alcohol drinking
(HAD) rats, and monkeys.6 Because the glycosidic CꢀC
bond is not cleavable under normal physiological con-
ditions, the pharmacological profile of C-glycoside of
isoflavone differs significantly from its corresponding
O-glycoside and aglycone. Although the 7,4%-di-O-
methyl analog of puerarin was synthesized previously,7
the total synthesis of puerarin has not been reported. In
order to fully explore its pharmacological properties
and to establish structure–activity relationships, we
developed a synthetic approach to puerarin (1).
ride was reported to gave b-C-glycoside (9) and its
corresponding a-isomer.7 However, we had to use 12
equiv. of Grignard reagent and the stereoselectivity was
poor. a-Isomer was the major product and the overall
yield was approximately 30%. In addition, isolation of
(9), a highly water soluble and magnesium complexed
product, was a nightmare. We tried to reduce the
equivalents of Grignard reagent, unfortunately it
resulted a complex mixture. In order to improve the
yield of the key intermediate (9), a b-isomer of
D-
glucopyranosyl-2,6-dimethoxybenzene, we employed
coupling reaction of a lithiated aromatic reagent (3)
with a benzyl protected glycopyranolactone (2) fol-
lowed by hydride reduction using TESH and BF3·Et2O
to give predominately the b-isomer (6).9a Interestingly,
it was reported in the literature that the same lithium
reagent (3) failed to react with (2) to give the corre-
sponding (6).9b It was assumed that the failure of the
reaction might result from enolization of the lactone. In
our hands, b-C-glycoside (6) was obtained stereoselec-
tively in 56% overall yield upon coupling of lactone (2)
with 1.6 equivalents of lithium reagent (3), and fol-
lowed by hydride reduction with TESH in the presence
of BF3·Et2O. Hydrogenolysis of (6) using Pd/C as
catalyst in methanol gave (9) in almost quantitative
yield. The spectroscopic and physical data were identi-
cal to the product (9) obtained from Grignard reaction
and no trace of a-isomer was found.
The key intermediate in our synthetic approach out-
lined in Scheme 1 is the b-D-glucopyranosyl-2,6-
dimethoxybenzene (9). In review of the literature, a vast
array of methods are available for carbonꢀcarbon bond
formation at the anomeric carbon, however, in many
cases the sugar ring is attached to the less-hindered
position.8 Grignard reaction of 2,6-dimethoxyphenyl-
magnesium bromide with tetraacetyl-a-D-glucosyl chlo-
Keywords: puerarin; total synthesis; natural product; isoflavone C-
glycoside.
We also tried to react pyranolactone (2) with various
sterically hindered lithium reagents. The mono-methoxy
analog (10) was reported by coupling of 2-methoxy
ꢀ Supplementary data associated with this article can be found at
doi:10.1016/S0040-4039(03)01715-5
phenylmagnesium bromide (12) and tetraacetyl-a-
D-
* Corresponding author. Fax: 617-855-2040; e-mail: dlee@
mclean.harvard.edu
glucosyl bromide (11).10 When we repeated the same
0040-4039/$ - see front matter © 2003 Published by Elsevier Ltd.
doi:10.1016/S0040-4039(03)01715-5

6858
D. Y. W. Lee et al. / Tetrahedron Letters 44 (2003) 6857–6859
Scheme 1. Reagents and conditions: (a) THF, −78 to −10°C; 2. TESH, BF3·Et2O, CH2Cl2, −78°C to rt; (b) Pd/C, H2; (c) Ac2O,
Py; (d) AlCl3, AcCl, Et2O; (e) 1. Na, MeOH; 2. H+ resin; (f) p-MeOC6H4CHO, NaOH, EtOH; (g) 1. Tl(NO3)3, MeOH/
CH(OCH3)3; 2. 10% HCl, MeOH, reflux; (h) TMSI, CH3CN, reflux, 5 days.
Grignard reaction, the major product showed the same
1H and 13C NMR data as reported in the literature. But
a close examination of its DEPT spectrum, we found
two methylene carbons in the sugar region and proved
to be a C-glucofuranoside (13) as revealed by COSY,
HMQC and HMBC analyses (Scheme 2). Apparently,
it was derived from sugar ring contraction under reflux-
ing Grignard reaction condition.
firmed unambiguously by DEPT, COSY, HMQC and
HMBC analyses and no trace of a-isomer was present.
In the case of MOM protected aryl lithium (5), it
reacted smoothly with pyranolactone (2) to yield b-C-
glycoside (8). MOM group was cleaved by BF3·Et2O
during the reduction.
Following the literature procedure,7 acetylation of gly-
coside (9) gave tetraacetate (14). Reaction of tetra-
acetate (14) in anhydrous ether with acetyl chloride in
the presence of anhydrous AlCl3 gave the acyl com-
pound (15). During the acylation the methoxyl group
ortho to the acyl group was also demethylated. After
removal of acetyl groups, compound (16) was con-
densed with p-methoxybenzaldehyde to give chalcone
(17). Subsequent acetylation provided protected chal-
cone (18). Oxidative rearrangement of (18) with
Tl(NO3)3 in a mixed solvent of methanol and trimethyl
orthoformate (1:1 v/v), followed by refluxing in
methanol with 10% HCl gave desired 7,4%-di-O-
methylpuerarin (19) in 84% yield.7 Subsequent
demethylation encountered many difficulties. Several
conditions of demethylation of (19) were attempted and
Scheme 2.
Following the previous coupling procedure, using
lithium reagent (4), the corresponding b-C-glycoside (7)
was obtained in 65% yield after reduction with TESH
and BF3·Et2O. Hydrogenolysis of (7) proceeded
smoothly to give (10). The structure of (10) was con-

D. Y. W. Lee et al. / Tetrahedron Letters 44 (2003) 6857–6859
6859
unsuccessful. For instance, treatment of (19) with
BBr3,11 BCl3,12 Ph2PLi in THF,13 pyridinium tribromide
in xylene,14 EtSNa in DMF,15 refluxing in concentrated
HBr,16 48% HBr with NaI at 110°C,17 NaCN in DMSO
at 125°C,18 and AlBr3 in refluxing EtSH.19 We noticed
that upon refluxing of (19) with Me3SiI in CHCl3 for
4 days gave a small amount of 7-O-methylpuerarin
6. Lin, R. C.; Li, T. K. Am. J. Clin. Nutrition 1998, 68 (6
Suppl), 1512S.
7. (a) Eade, R. A.; McDonald, F. J.; Simes, J. J. H. Aust. J.
Chem. 1975, 28, 2011; (b) Eade, R. A.; McDonald, F. J.;
Pham, H. P. Aust. J. Chem. 1978, 31, 2699.
8. Du, Y.; Linhardt, R. J. Tetrahedron 1998, 54, 9913.
9. (a) Calzada, E.; Clarke, C. A.; Roussin-Bouchard, C.;
Wightman, R. H. J. Chem. Soc., Perkin Trans. 1 1995,
517; (b) Kraus, G. A.; Molina, M. T. J. Org. Chem. 1988,
53, 752.
10. Qiu, D.; Wang, Y.; Cai, M. Chem. J. Chinese Univ. 1989,
5, 158.
11. McOmie, J. F. W.; West, D. E. Org. Synth. Coll. Vol. V.
1973, 412.
12. Carvalho, C. F.; Sargent, M. V. J. Chem. Soc., Chem.
Commun. 1984, 4, 227.
13. Ireland, R. E.; Walba, D. M. Org. Synth. 1977, 56, 44.
14. Ijaz, A. S.; Alam, M.; Ahmad, B. Indian J. Chem. Sect. B
1994, 33B, 288.
15. Feutrill, G. I.; Mirrington, R. N. Tetrahedron Lett. 1970,
16, 1327; Aust. J. Chem. 1972, 25, 1719, 1731.
16. Kawasaki, I.; Matsuda, K.; Kaneko, T. Bull. Chem. Soc.
Jpn. 1971, 44, 1986.
20
(21), unfortunately,
a prolonged reflux failed to
improve the yield and further demethylation at C-7
position never happened. Finally, when CH3CN was
used as solvent, refluxing with Me3SiI for 5 days,
starting material 19 was consumed and converted to the
desired target compound (1)21 in 35% yield along with
partially demethylated 7-O-methylpuerarin (21) in 39%
yield after column chromatography. Further prolong-
ing the reaction time did not improve the yield. The
overall yield of the synthesis of (1) was 10%. The
spectroscopic and physical data were identical to the
natural sample of puerarin (1). In order to improve the
yield at the final step, other protecting groups are being
examined.
Acknowledgements
17. Li, G.; Patel, D.; Hruby, V. J. Tetrahedron Lett. 1993, 34,
5393.
18. McCarthy, J. R.; Moore, J. L.; Cregge, R. J. Tetrahedron
Lett. 1978, 52, 5183.
19. Node, M.; Nishide, K.; Fuji, K.; Fujita, E. J. Org. Chem.
1980, 45, 4275.
Financial support from National Institute of Alcohol
Abuse and Alcoholism (N44AA02006) is greatly
acknowledged.
20. Jung, M. E.; Lyster, M. A. J. Org. Chem. 1977, 42, 3761.
21. Spectral data for compound 1: solid, mp: 184–186°C;
HRMS m/e for C21H20O9+H calcd 417.1185, found
417.1174, 1H NMR (CD3OD, 300 MHz, ppm), l 3.40–
3.60 (3H, m), 3.72–3.87 (2H, m), 4.12 (1H, s, br), 5.09
(1H, d, J=9.9 Hz, H-1¦), 6.85 (2H, d, J=8.7 Hz, H-2%
and H-6%), 6.99 (1H, d, J=9 Hz, H-6), 7.37 (2H, d, J=8.4
Hz, H-3% and H-5%), 8.06 (1H, d, J=8.7 Hz, H-5), 8.20
(1H, s, H-2); 13C NMR (CD3OD, 75.4 MHz, ppm), l
62.74, 71.68, 72.98, 75.63, 79.98, 82.72, 113.1, 116.23,
116.58, 118.45, 124.16, 125.48, 128.07, 131.34, 154.46,
158.64, 162.97, 178.25.
References
1. Cao, X.; Tian, Y.; Zhang, T.; Li, X.; Ito, Y. J. Chro-
matogr. A 1999, 855, 709.
2. Fang, Q. J. Ethnopharmacol. 1980, 2, 57.
3. Keung, W. M.; Vallee, B. L. Proc. Natl. Acad. Sci. USA
1993, 90, 10008.
4. Shen, Z.-F.; Xie, M.-Z. Acta Pharm. Sin. 1985, 20, 863.
5. Fan, L.-L.; O’Keefe, D. D.; Powell, W. J. Acta Pharm.
Sin. 1984, 19, 801.