Total synthesis of (+)-isatisine A

Article abstract of DOI:10.1002/anie.200906632

(Figure Presented) Isatis indigotica has long been used as a source of therapeutic compounds in traditional Chinese medicine (the roots are the source of Ben Lan Gen herbal tea). In 2007 the leaves of J. indigotica yielded isatisine A, an unusual oxidized bisindole. A concise, enantiospecific total synthesis of isatisine A is reported using a homochiral cyclopropane diester to construct the key tetrahydrofuran ring.

Full text of DOI:10.1002/anie.200906632

Angewandte
Chemie
DOI: 10.1002/anie.200906632
Total Synthesis
Total Synthesis of (+)-Isatisine A**
Avedis Karadeolian and Michael A. Kerr*
In memory of Keith Fagnou
Traditional Chinese medicine has yielded a plethora of
diverse naturally occurring bioactive compounds for many
years.[1] Isatis indigotica Fort. is a plant species cultivated in
China for its antiviral properties which include activity against
influenza, viral pneumonia, and hepatitis.[2] Recently Chen
and co-workers reinvestigated this organism in search of anti-
HIV compounds and reported the isolation of 64 mg of the
bis-indole 2 (Figure 1) from 50 kg of the dried leaves of I.
quantities of product for medicinal and biological investiga-
tion as well as clarifying the absolute configuration.
A retrosynthetic analysis for isatisine A is shown in
Scheme 1. The final step is an acid-catalyzed indole addition
to the hemiaminal 3 (which may have biomimetic implica-
Figure 1. Isatisine A (1) and its acetonide derivative (2).
indigotica.[3] Interestingly it became apparent that acetonide 2
was not in fact the native natural product but an artefact
arising from purification with an acetone eluent on silica gel.
Upon this realization and reinvestigation, compound 1,
named isatisine A was determined to be the native plant
constituent.
Acetonide 2, however, did provide a convenient com-
pound for full characterization which included unambiguous
structure determination by single crystal X-ray diffraction.
Although an optical rotation was obtained for 2, the absolute
stereochemical configuration was not reported. The only
reports of bioactivity in the isolation paper[3] are for the
acetonide 2, which showed an EC50 anti-HIV-1 activity of
37.8 mm. Isatisine A itself was not screened for bioactivity,
however, it is reasonable to assume that its activity may be
promising. Herein we report a concise synthesis of isatisine A
(1) (through its acetonide 2) capable of providing significant
Scheme 1. Retrosynthesis of isatisine A (1).
tions).[4] The masked 1,2-dicarbonyl in 3 would be installed
through oxidation of the 2,3-indole double bond in a substrate
such as 4.[5] The requisite dihydroxy moiety would arise from a
diastereoselective dihydroxylation of enoate 5, the olefin of
which coming from dehydrocarbonylation of one of the
geminal diesters in 6. The tetrahydrofuran ring in 6, which
represents a central structural feature of the natural product,
would be prepared through the elegant cyclopropane–alde-
hyde cycloaddition developed by Johnson and co-workers,[6]
between an indole-2-carboxaldehyde 7 and a suitable cyclo-
propane-1,1-diester 8. The S configuration of the cyclo-
propane as shown would set the C10 and C12 stereocenters as
R and S, respectively, by virtue of stereospecific inversion
upon ring-opening in the cycloaddition process and a cis
diastereoselectivity in the tetrahydrofuran synthesis. This
enantiomeric choice of starting material would prepare the
enantiomer of isatisine A depicted herein (and in the isolation
paper[3]).
[*] A. Karadeolian, Dr. M. A. Kerr
Department of Chemistry, The University of Western Ontario
London, ON, N6A 5B7 (Canada)
Fax: (+1)519-661-3022
E-mail: makerr@uwo.ca
Our synthesis commenced with the key Johnson tetrahy-
drofuran synthesis[6b] using the indole-2-carboxaldehyde 9 and
the homo-chiral (S)-vinylcyclopropane diester 10[7] under
catalytic influence of Sn(OTf)2 (Scheme 2). The tetrahydro-
furan 11 was isolated in 89% yield as an 11:1 mixture of the
2,5-cis:2,5-trans (furan numbering) isomers. There was mod-
erate selectivity at the geminal diester center, however this
[**] We thank the Natural Sciences and Engineering Research Council
(NSERC) and Merck Frosst for funding. We are grateful to Doug
Hairsine for performing MS analyses. A.K. is the recipient of an OGS
postgraduate scholarship. A special thanks to Dr. Jeffrey Manthorpe
for helpful discussion and suggestions.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200906632.
Angew. Chem. Int. Ed. 2010, 49, 1133 –1135
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1133

Communications
This yield is more impressive than at first glance in that it
represents the removal of the trans-furan diastereomer as well
as some product resulting from the undesired regioisomeric
elimination to the 2,3-unsaturated furan. Dihydroxylation of
15 under standard conditions gave diol 16 as a single
diastereomer in 62% yield. Acetonide formation was con-
comitant with silyl ether removal giving 17 in 83% overall
yield. In order to set up indole oxidation, the N-tosyl group
was removed under the influence of magnesium metal in
methanol,[11] giving 18 in 77% yield. In what proved to be a
wonderfully efficient strategy, isatisine A acetonide (2) was
produced from 18 in a simple two-step procedure with no
intermediate purification. Treatment of 18 with mCPBA in
CH2Cl2 for 2 h at room temperature,[5a] gave a crude oxidation
product which, upon isolation was treated with indole and
camphorsulfonic acid in CH2Cl2 for 42 h. Standard workup
gave isatisine A acetonide (2) in 50% overall yield. Simple
hydrolysis of the acetonide in acidic methanol produced the
natural product 1 in 82% yield.
The key transformation of 18 to 2 deserves comment and
is outlined in Scheme 3. Although, in practice it was used
crude, the oxidation product of 18 with mCPBA was in fact
Scheme 2. Synthesis of (+)-isatisine A (1). a) Sn(OTf)2, CH2Cl2 (89%);
b) methanesulfonamide, NMO, OsO4, THF/H2O/acetone; c) NaIO4,
THF/H2O, 08C; d) NaBH4, THF/H2O/EtOH (87%, 3 steps); e) TBSCl,
imidazole, CH2Cl2 (94%); f) Pd/C, H2 (1 atm), THF; g) PPh3, allyl
alcohol, DIAD, THF (99%, 2 steps); h) [Pd2(dba)3], CH3CN, 808C (45–
55%); i) methanesulfonamide, NMO, OsO4, acetone/H2O (62%);
j) 2,2-dimethoxypropane, p-toluenesulfonic acid, acetone (83%, >98%
ee); k) Mg0, NH4Cl, MeOH (77%); l) mCPBA, CH2Cl2; m) indole, CSA,
CH2Cl2, 42 h (50%, >98% ee); n) 1n HCl, MeOH (82%). NMO=N-
methylmorpholine-N-oxide, TBS=tert-butyldimethylsilyl, DIAD=diiso-
propyl azodicarboxylate, dba=dibenzylideneacetone, mCPBA=m-
chloroperbenzoic acid, CSA=10-camphorsulfonic acid, Ts=toluene-4-
sulfonyl, Bn=benzyl.
would prove to be inconsequential since this carbon would
become sp2 hybridized in subsequent steps.
The pendant vinyl substituent in 11 was converted to a
primary alcohol through dihydroxylation, oxidative cleavage,
and reduction of the resulting aldehyde, producing 12 in 87%
overall yield. Protection of the primary alcohol as a silyl ether
was uneventful, providing 13 in 94% yield. It is worthy of note
at this juncture that the obvious use of a cyclopropane bearing
the required hydroxy substituent was not successful under a
variety of conditions. The cycloaddition apparently requires
the p-donor activity provided by the vinyl substituent.
In order to install the double bond necessary for
formation of the 9,13-dihydroxyl moiety (isatisine A number-
ing; Figure 1), we initially considered the pedestrian strategy
of decarboxylation, installation of a leaving group, and
elimination. After some unsuccessful attempts along this
line, we settled on a much more productive strategy—namely
to eliminate one of the esters directly. To this end, hydro-
genolysis of the benzyl ester and subsequent Mitsunobu
allylation,[8] gave a 99% yield of the allyl ester 14. Treatment
of 14 with [Pd2(dba)3] gave a 45–55% yield of enoate 15.[9,10]
Scheme 3. Oxidation of 18 with indole addition.
characterized and found to be a 2:1 epimeric mixture of
aminals 19. The diastereomers were separable and stable,
however, the stereochemical configuration at the aminal
carbon could not be unambiguously determined. This of
course would prove to be inconsequential since treatment of
19 with indole and camphorsulfonic acid resulted in C3 indole
alkylation through an N-acyliminium ion devoid of asymme-
try at this carbon. After 4.5 h, a 2.3:1 mixture of indole
addition products was formed in favor of the undesired
diastereomer 20. This indole addition is apparently reversible,
since after 14 h, the ratio had readjusted to a 3:1 ratio in favor
of the desired isomer 21. After 24 h that ratio was constant
and a small amount of the isatisine A acetonide (2) was
1134
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 1133 –1135

Angewandte
Chemie
[3] J.-F. Liu, Z.-Y. Jiang, R.-R. Wang, Y.-T. Zeng, J.-J. Chen, X.-M.
observed. A 42 h reaction time resulted in a 6.3:1:0.44 ratio of
2, 21, and 20. While 20 and 21 could, in principle form more of
the desired product, the yield of 2 diminished due to various
acid mediated decomposition pathways involving the indole
moiety. Thus the overall yield of 2 from 18 was 50% without
purification of intermediate 19.
The specific rotation of the synthetic acetonide 2 and the
synthetic isatisine A (1) were ½aꢀ2D5 = + 271 and ½aꢀD25 = + 274,
respectively. Our value for acetonide 2 is essentially equal and
opposite to that reported in the isolation paper (½aꢀ1D4 = ꢁ283
(c = 0.46, MeOH)).[3] Since our product is C2(R), C9(S),
C10(R), C12(S), C13(S), the natural material must be C2(S),
C9(R), C10(S), C12(R), C13(R) which is antipodal to the
structural depictions in the isolation paper. Clearly, the route
described herein starting with the R-cyclopropane 10 would
yield the natural enantiomer.
Zhang, Y.-B. Ma, Org. Lett. 2007, 9, 4127 – 4129.
[4] a) P. Astolfi, L. Creci, C. Rizzoli, P. Sgarabotto, G. Marrosu, J.
Chem. Soc. Perkin Trans. 2 2001, 1634 – 1640; b) C. I. Altinis
Kiraz, T. J. Emge, L. S. Jimenez, J. Org. Chem. 2004, 69, 2200 –
2202.
[5] a) M. J. Buller, T. G. Cook, Y. Kobayashi, Heterocycles 2007, 72,
163 – 166; b) Z. Wang, L. S. Jimenez, Tetrahedron Lett. 1996, 37,
6049 – 6052; c) V. J. Colandrea, S. Rajaraman, L. S. Jimenez,
Org. Lett. 1996, 37, 6049 – 6052; d) C.-S. Chien, T. Suzuki, T.
Kawasaki, M. Sakamoto, Chem. Pharm. Bull. 1984, 32, 3945.
[6] a) P. D. Pohlhaus, J. S. Johnson, J. Am. Chem. Soc. 2005, 127,
16014 – 16015; b) P. D. Pohlhaus, S. D. Sanders, A. T. Parsons, W.
Li, J. S. Johnson, J. Am. Chem. Soc. 2008, 130, 8642 – 8650;
c) A. T. Parsons, M. J. Campbell, J. S. Johnson, Org. Lett. 2008,
10, 2541 – 2544; d) A. T. Parsons, J. S. Johnson, J. Am. Chem. Soc.
2009, 131, 3122 – 3123; e) S. D. Sanders, A. Ruiz-Olalla, J. S.
Johnson, Chem. Commun. 2009, 5135 – 5137; f) M. J. Campbell,
J. S. Johnson, J. Am. Chem. Soc. 2009, 131, 10370 – 10371.
[7] Prepared from the dimethyl ester (see Supporting Information).
For the synthesis of the (S)-vinylcyclopropane-1,1-dimethyl ester
see C. A. Carson, M. A. Kerr, Angew. Chem. 2006, 118, 6710 –
6713; Angew. Chem. Int. Ed. 2006, 45, 6560 – 6563.
In summary, we have successfully completed the total
synthesis of (+)-isatisine A in 14 steps from homochiral
cyclopropane 10 in an overall yield of 5.8%. The synthesis
and biological investigation of isatisine A analogues is under-
way and will be reported in due course.
[8] O. Mitsunobu, M. Yamada, Bull. Chem. Soc. Jpn. 1967, 40, 2380 –
2382.
[9] I. Shimizu, T. Tsuji, J. Am. Chem. Soc. 1982, 104, 5844 – 5846.
[10] We thank Professor Jeffrey Manthorpe for this helpful sugges-
tion.
Received: November 24, 2009
Published online: January 5, 2010
[11] a) Y. Yokoyama, T. Matsumoto, Y. Murakami, J. Org. Chem.
1995, 60, 1486 – 1487; b) A. M. Elder, D. H. Rich, Org. Lett. 1999,
1, 1443 – 1446.
Keywords: alkaloids · asymmetric synthesis · heterocycles ·
isatisine A · total synthesis
.
Please note: Minor changes have been made to this Manuscript since
its publication in Angewandte Chemie Early View. The Editor.
[1] J. Zhou, G. Xie, X. Yan in Traditional Chinese Medicines, 2nd
ed., Ashgate, England, 2003.
[2] H. Z. Zheng, Z. H. Dong, Q. Yu in Modern Study of Traditional
Chinese Medicine, Vol. 1, Xueyuan, Beijing, 1997, p. 328 – 334.
Angew. Chem. Int. Ed. 2010, 49, 1133 –1135
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1135