Heterolignans: Stereoselective synthesis of an 11-amino analog of azaelliptitoxin

Article abstract of DOI:10.1002/ejoc.200800880

The 11-amino-azaelliptitoxin analog (11R,11aS,5R)-15 has been prepared stereoselectively in eight steps and 24% over-all yield from commercially available (S)-glycidol (7) via the original enantiopure chiral α-amino aldehyde 9, used as chiral precursor fo

Full text of DOI:10.1002/ejoc.200800880

SHORT COMMUNICATION
DOI: 10.1002/ejoc.200800880
Heterolignans: Stereoselective Synthesis of an 11-Amino Analog of
Azaelliptitoxin
Julie Routier,[a] Mihaela Calancea,[a] Marion David,[a] Yannick Vallée,[a] and
Jean-Noël Denis*[a]
Keywords: Stereoselective synthesis / α-Amino aldehydes / Bioactive heterolignans / Azaelliptitoxin / Pictet–Spengler
reaction
The 11-amino-azaelliptitoxin analog (11R,11aS,5R)-15 has
been prepared stereoselectively in eight steps and 24% over-
all yield from commercially available (S)-glycidol (7) via the
ral precursor for the creation of the two other stereogenic
centers of the target framework.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
original enantiopure chiral α-amino aldehyde 9, used as chi- Germany, 2009)
Introduction
Lignans represent a large class of natural products, which
possess a great variety of structures and a broad range of
biological activities.[1,2] For these reasons, these compounds
are attractive targets for biomedical and synthetic pur-
poses.[2]
Podophyllotoxin (1) is the most well-known of them,
studied for its inhibition activity on microtubule assembly
in the mitotic apparatus.[3] It has been used to develop semi-
synthetic analogs such as etoposide (VP-16, Figure 1) (2a)
and tenoposide (VM-26) (2b), which have no activity on
microtubule assemblies, but are potent inhibitors of DNA
topoisomerase II. The knowledge of those biologically
active compounds[3] led to design a new compound by mo-
lecular modelling, azaelliptitoxin (3) (or azatoxin, a new
heterolignan),[4,5] considered as a hybrid[5a] based on etopo-
side (2a) and ellipticine (4), another well-established topo-
isomerase II inhibitor (Figure 1).[6]
Modifications of the azaelliptitoxin skeleton have been
investigated (Figure 2).[5c–5g] The results revealed that 11-
alkoxy compounds 5a do not have any activity towards
topoisomerase II, but introduction of an 11-amino or -poly-
amino substituent, and in particular, introduction of a 4-
substituted anilino group at this C-11 position, led to 11-
amino compounds 5b, which have topoisomerase II inhibi-
tory activity.[5c]
Figure 1. Structures of bioactive lignans and ellipticine: podophyl-
lotoxin (1), etoposide (2a), teniposide (2b), azaelliptitoxin (3), ellip-
ticine (4).
To the best of our knowledge, only one preparation of
11-amino analogues 5b is reported.[5c] Moreover, this strat-
egy is based on the use of -tryptophanol as starting mate-
[a] Département de Chimie Moléculaire (SERCO), UMR-5250,
ICMG FR-2607, CNRS, Université Joseph Fourier
B. P. 53, 38041 Grenoble Cedex 9, France
Fax: +33-4-76635983
E-mail: jean-noel.denis@ujf-grenoble.fr
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Figure 2. Structures of some azaelliptitoxin derivatives: 5a (X =
OR), 5b (X = NHR1), 6 (benz[e, f and g]azaelliptitoxins).
Eur. J. Org. Chem. 2008, 5687–5691
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5687

J. Routier, M. Calancea, M. David, Y. Vallée, J.-N. Denis
SHORT COMMUNICATION
rial, limiting the scope of structural and stereochemical in CH2Cl2, according to the method described by Katsu-
modifications.
mura and co-workers (Scheme 1).[7] We obtained (S)-8 in
Later, the synthesis of benzannulated azaelliptitoxin de- 74% chemical yield with an optical rotation of [α]D = –31.4
rivatives has been described, affording benz[e, f and g]aza- (c = 2.2, CHCl3); this value is in good accordance with the
toxins 6 (Figure 2). It has been shown that these derivatives one described for the (R)-alcohol {[α]D = +29.8 (c = 1.03,
do not interact with topoisomerase II but are strong inhibi- CHCl3),[7a] [α]D = +32.3 (CHCl3)[7b]}, showing the enantio-
tors of tubulin polymerization. Biological data showed that purity of our (S) compound.
azaelliptitoxin derivatives could be either inhibitors of the
tubulin polymerization or inhibitors of the DNA topoiso-
merase II depending on their structure.
In order to flesh out the available information on struc-
ture/activity relationships, we have developed an original
strategy, allowing the access to 11-aminoazaelliptitoxin de-
rivatives from three commercially available compounds: (S)-
glycidol, indole and p-methoxybenzaldehyde dimethylace-
tal.
Scheme 1. Enantioselective synthesis of the nitrone (S)-10. Rea-
gents and conditions: (i) BnNCO (1 equiv.), Et3N (1.8 equiv.),
CH2Cl2, 40 °C, 18 h (74%); (ii) DMP (2 equiv.), CH2Cl2, room
temp., 1.5 h (60%), or IBX (3 equiv.), EtOAc, 80 °C, 3 h (quant.);
(iii) BnNHOH (1.7 equiv.), MgSO4 (excess), CH2Cl2, room temp.,
3 h (98%).
Results and Discussion
In this paper, we describe the stereoselective synthesis of
the 11-(pivaloylamino) analog (11R,11aS,5R)-15 as summa-
rized in the retrosynthetic analysis depicted in Figure 3. All
parts of the azaelliptitoxin skeleton could be modified ac-
cording to the proposed disconnections shown in target
compound 15, with the possibility to prepare several deriva-
tives by using the same strategy.
The next step was the oxidation of the alcohol (S)-8 into
the chiral α-amino aldehyde (R)-9. To the best of our
knowledge, this aldehyde (as well as its enantiomer) has
never been isolated nor characterized. Katsumura and co-
workers reported that the oxidation of alcohol (R)-8 under
Swern conditions, or by using PCC, PDC, tetra-n-propyl-
ammonium perruthenate, or sulfur trioxide–pyridine com-
plex only gave unstable compounds, which decomposed
into a complex mixture.[7b]
We then chose to study the oxidation of the alcohol (S)-
8 by using two well-known mild oxidants, namely, Dess–
Martin periodinane and o-iodoxybenzoic acid. The alcohol
(S)-8, oxidized with Dess–Martin periodinane (DMP,
2 equiv.)[8] in CH2Cl2 (Scheme 2), led to the aldehyde (R)-
9 in a modest 60% isolated yield, whereas the use of o-
iodoxybenzoic acid (IBX, 3 equiv.)[9] in EtOAc at 80 °C led
to the same aldehyde in 99% isolated yield. It is a stable
solid that could be stored under argon at 0 °C for several
days and could be purified by flash column chromatog-
raphy. The 1H NMR spectrum in CDCl3 showed that it
exists in several forms, among which hydrates are predomi-
Figure 3. Retrosynthetic analysis for the preparation of the 11-
(pivaloylamino) analog (11R,11aS,5R)-15.
nant, whereas the aldehyde form predominates (Ն98%) in
1
The strategy is first based on the stereoselective synthesis the H NMR spectrum in [D8]toluene. The optical purity
1
of the indolic N-benzyl-N-hydroxylamine 11a by reaction of aldehyde (R)-9 was determined by comparison of the H
of indole with the highly functionalized nitrone 10, which NMR spectrum of the Mosher ester of its corresponding
could be prepared from the commercially available (S)-gly- alcohol (S)-8 (obtained by reduction with NaBH4 in etha-
1
cidol (7) via the chiral α-amino aldehyde 9, which has been nol)[10] and the H NMR spectra of Mosher esters derived
chosen as the chiral precursor to induce the stereoselective from the (R)- and racemic (Ϯ)-alcohols 8. This result
creation of the two other stereogenic centers (Figure 3).
showed that (R)-9 was enantiopure.[11]
Adequate transformation of the N-benzyl-N-hydroxyl-
It is expected that enantiopure α-oxazolidino aldehyde
amine 11a to an indolic free-oxazolidinone intermediate (R)-9 could become an important alternative to Garner’s
should afford the expected adduct 15 after a Pictet–Speng- aldehyde in organic synthesis because of its straightforward
ler reaction with p-methoxybenzaldehyde dimethylacetal.
and efficient access in only two steps from commercially
The first stage of the strategy consisted of the synthesis available enantiopure (S)-glycidol (7). Furthermore, the ox-
of the alcohol 8 from commercial (S)-glycidol (7) by reac- azolidin-2-one moiety is present in many bioactive com-
tion with benzyl isocyanate in the presence of triethylamine pounds and is widely used as pattern in chiral auxiliaries.[12]
5688
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 5687–5691

Heterolignans
The choice of the adequate protecting group for the pri-
mary amine 12 was then essential for the development of
our strategy: it has to be electron-withdrawing and bulky[14]
to avoid its implication in the final Pictet–Spengler cycliza-
tion. This protecting group must also be stable under both
acidic and basic conditions; the pivaloyl group meets all
these criteria.
Treatment of the primary amine 12 with pivaloyl chloride
(2.6 equiv.) in CH2Cl2 in the presence of Et3N gave the
amide 13 in 96% isolated yield. The debenzylation reaction
of the oxazolidinone moiety of 13 with lithium (7 equiv.) in
liquid NH3/THF at –78 °C gave compound 14 in a nearly
quantitative yield (Scheme 3).[15] Overall, this key interme-
diate was obtained in three steps from N-hydroxylamine 11a
in 96% yield and in seven steps from the commercial (S)-
glycidol (7) in 29% overall yield.
The next goal of the multi-step synthesis of the
(11R,11aS,5R)-11-(pivaloylamino) analog 15 consisted of
the implementation of the Pictet–Spengler reaction by treat-
ment of compound 14 with p-methoxybenzaldehyde di-
methylacetal under acidic conditions.[16] We found that the
best results were obtained in an anhydrous CH2Cl2/
CF3COOH (3:1) mixture at 40 °C. Under these conditions,
adduct 15 was obtained in 82% yield as a single dia-
stereoisomer (Scheme 4).
Scheme 2. Preparation of the N-hydroxylamine 11a. Reagents and
conditions: Method A: (i) indole (1 equiv.), AcCl (2 equiv.), MeOH,
0 °C, 4 h (52%), dr = 3:2. 11a (R,4S) (31.5%), 11b (S,4S), (20.5%);
method B: (i) indole (1 equiv.), HCl (2 equiv., 2  in Et2O), MeOH/
CH2Cl2 (4:1), 0 °C to room temp., 5 h, dr = 5:1, 11a (R,4S) (42%),
11b (S,4S) (8%).
The reaction of aldehyde (R)-9 with N-benzylhydroxyl-
amine in CH2Cl2 in the presence of an excess of anhydrous
magnesium sulfate[13] gave the α-chiral nitrone (S)-10 in a
nearly quantitative yield (Scheme 1).
When nitrone (S)-10 was added to indole under acidic
conditions in methanol at 0 °C (Method A, Scheme 2),[13]
a
mixture of diastereoisomers 11a (R,4S) and 11b (S,4S)[13c]
1
in a 3:2 ratio (determined by H NMR of crude mixture)
was obtained. Compounds 11a and 11b were isolated in
32% and 20% yield, respectively. Optimization of experi-
mental conditions (Method B, Scheme 2) led us to obtain
the two diastereoisomers 11a and 11b in a 5:1 ratio (deter-
1
mined by H NMR spectroscopy of crude product), which
were separated to obtain compounds 11a and 11b in 42%
and 8% yield, respectively.
We then prepared the α-(pivaloylamino)-β-oxazolidinone
13 from the N-hydroxylamine 11a. In the first step, N-hy-
droxylamine 11a was transformed into the corresponding
free primary amine 12 in quantitative yield by hydro-
genolysis in the presence of Pearlman’s catalyst in a MeOH/
AcOH mixture (85:15) (Scheme 3).
Scheme 4. Pictet–Spengler reaction. Reagents and conditions: (i) p-
methoxybenzaldehyde dimethylacetal (3 equiv.), CH2Cl2/TFA (3:1),
40 °C, 7 h (82%).
Similar Pictet-Spengler reactions have already been de-
scribed in the literature for the synthesis of indole alka-
loids,[17] azaelliptitoxin and analogs[5] as well as azapodo-
phyllotoxin derivatives.[18] It is reported that the C-5 ste-
reogenic center created during the reaction has a trans rela-
tionship with the C-11a center of the oxazolidinone moiety
to afford a 1,3-trans stereochemistry.
The relative and absolute stereochemistries of the ste-
reogenic centers C-5, C-11 and C-11a in 15 were determined
by NOE experiments (1H NMR, 500 MHz, C6D6) (Fig-
ure 4). It was observed that irradiation of 11a-H induced a
positive NOE on 11-H, showing the expected syn relation-
ship.
Moreover, the coupling constant between these two hy-
drogen atoms ranged from 3.1 Hz (1H NMR, 400 MHz,
CDCl3) to 3.0 Hz (1H NMR, 500 MHz, C6D6), which is in
accordance with the values given in the literature for a sim-
ilar structure.[19] On the other hand, irradiation of 11a-H
did not induce an NOE on 5-H, demonstrating a trans rela-
tionship between them. Finally, an NOE was observed on
Scheme 3. Synthesis of the key intermediate 14. Reagents and con-
ditions: (i) H2, cat. Pd(OH)2, MeOH/AcOH (85:15), room temp.,
48 h (quant.); (ii) PivCl (2.6 equiv.), Et3N (1.6 equiv.), THF, room
temp., 5 h (96%); (iii) Li (7 equiv.), NH3/THF, –78 °C, 30 min
(quant.).
Eur. J. Org. Chem. 2008, 5687–5691
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5689

J. Routier, M. Calancea, M. David, Y. Vallée, J.-N. Denis
SHORT COMMUNICATION
Acknowledgments
Financial support from the Centre National de la Recherche Sci-
entifique (CNRS), the Université Joseph Fourier (UMR-5250,
ICMG FR-2607) and a fellowship (to J. R.) from the Research
Ministry (MESR) are gratefully acknowledged.
[1] a) D. A. Whiting, Nat. Prod. Rep. 1990, 7, 349–364; b) M. Gor-
daliza, M. A. Castro, J. M. Miguel del Corral, A. San Felici-
ano, Curr. Pharm. Des. 2000, 6, 1811–1839; c) S. Apers, A.
Vlietinck, L. Pieters, Phytochem. Rev. 2004, 2, 201–207; d) G. P.
Moss, Pure Appl. Chem. 2000, 72, 1493–1523; e) R. S. Ward,
Nat. Prod. Rep. 1999, 16, 75–96; f) R. S. Ward, Nat. Prod. Rep.
1995, 12, 183–205; g) R. S. Ward, Phytochem. Rev. 2004, 2,
391–400.
[2] a) R. S. Ward, Synthesis 1992, 719–730; b) J. Chang, J. Reiner,
J. Xie, Chem. Rev. 2005, 105, 4581–4609; c) R. S. Ward, Phyto-
chem. Rev. 2003, 2, 391–400.
[3] a) T. F. Imbert, Biochimie 1998, 80, 207–222; b) Z.-Q. Wang,
H. Hu, H.-X. Chen, Y.-C. Cheng, K.-H. Lee, J. Med. Chem.
1992, 35, 871–877; c) Y.-L. Zhang, X. Guo, Y.-C. Cheng, K.-
H. Lee, J. Med. Chem. 1994, 37, 446–452; d) J. D. Sellars, P. G.
Steel, Eur. J. Org. Chem. 2007, 3815–3828.
Figure 4. Stereoselectivity model for the Pictet–Spengler reaction.
the aromatic 2Ј-H and 6Ј-H protons after irradiation of
11a-H, confirming that the aromatic pendant group was on
the α side of the molecule (Figure 4). These results corre-
spond well with those described in the literature.[19] In our
reaction, the corresponding cis 5-H/11a-H stereoisomer was
not observed. It is possible to conclude that in compound
15 the aromatic pendant group adopts a pseudo-axial orien-
tation. For steric reasons, the plane of this aromatic pen-
dant ring is approximately orthogonal to the rest of the
molecule.
This result is in good agreement with the conclusions de-
scribed by Lemaire and co-workers[12d] who reported that,
during the synthesis of tetrahydroisoquinoline derivatives,
the stereochemical outcome of the Pictet–Spengler reaction
could be rationalized by a cyclization step involving the at-
tack of the aromatic core on the (Z)-iminium moiety from
the less hindered side, leading to the 1,3-trans derivatives
(Figure 4).
During reaction of 14 with p-methoxybenzaldehyde di-
methylacetal, the nucleophilic attack of the indole ring on
the less hindered side of the (Z)-iminium moiety led to the
expected compound 15. The stereochemistry of the newly
created stereogenic center C-5 is governed by the stereo-
chemistry of the oxazolidinone core, leading to a 1,3-trans
relationship between the two implicated substituents, not by
the presence of the 11-amino substituent.
[4] Azaelliptitoxin is the best well-known heterolignan for its bio-
logical activities. The generic name “heterolignan” was first in-
troduced by Medarde et al., see: M. Medarde, R. Peláez-Lam-
amié de Clairac, F. Tomé, J. L. Lopez, A. San Feliciano, Arch.
Pharm. 1995, 328, 403–407.
[5] a) F. Leleurtre, J. Madalengoitia, A. Orr, T. J. Cuzy, E. Lehnert,
T. Macdonald, Y. Pommier, Cancer Res. 1992, 52, 4478–4483;
b) E. Solary, F. Leteurtre, K. D. Paull, D. Scudiero, E. Hamel,
Y. Pommier, Biochem. Pharmacol. 1993, 45, 2449–2456; c) J. J.
Tepe, J. S. Madalengoitia, K. M. Slunt, K. W. Werbovetz, P. G.
Spoors, T. L. Macdonald, J. Med. Chem. 1996, 39, 2188–2196;
d) T. A. Miller, P. R. Vachaspati, M. A. Labroli, C. D. Thomp-
son, A. L. Bulman, T. L. Macdonald, Bioorg. Med. Chem. Lett.
1998, 8, 1065–1070; e) B. Eymin, E. Solary, S. Chevillard, L.
Dubrez, F. Goldwasser, O. Duchamp, P. Genne, F. Leteurtre,
Y. Pommier, Int. J. Cancer 1995, 63, 268–275; f) F. Leleurtre,
D. L. Sackett, J. Madalengoitia, G. Kohlhagen, T. Macdonald,
E. Hamel, K. D. Paull, Y. Pommier, Biochem. Pharmacol. 1995,
49, 1283–1290; g) J. S. Madalengoitia, J. J. Tepe, K. A. Werbov-
etz, E. K. Lehnert, T. L. Macdonald, Bioorg. Med. Chem. 1997,
5, 1807–1815.
[6] L. F. Liu, Annu. Rev. Biochem. 1989, 58, 351–375.
[7] a) S. Katsumura, A. Kondo, Q. Han, Chem. Lett. 1991, 1245–
1248; b) S. Katsumura, N. Yamamoto, M. Morita, Q. Han,
Tetrahedron: Asymmetry 1994, 5, 161–164.
[8] D. B. Dess, J. C. Martin, J. Org. Chem. 1983, 48, 4155–4156.
[9] D. More, N. S. Finney, Org. Lett. 2002, 4, 3001–3003. For a
recent review, see: U. Ladziata, V. V. Zhdankin, ARKIVOC
2006, 10, 26–58.
Conclusions
Enantiopure 11-(pivaloylamino) analog 15 of azaellipti-
toxin 3 has been prepared in eight steps from (S)-glycidol
(7) in 24% overall yield. The strategy is based on the prepa-
ration of the enantiopure α-chiral aldehyde (R)-9 and its
use as a building block in the stereoselective synthesis of 15,
which has the same absolute stereochemistry (11R,11aS,5R)
than azaelliptitoxin (3) and the corresponding 11-amino de-
rivatives 5b. We described the first enantioselective synthesis
of the key intermediate (R)-9 in a two-step sequence from
(S)-glycidol (7). Application of this stereoselective multi-
step sequence to the preparation of more complex analogs
of azaelliptitoxin is currently under investigation in our
group.
[10] The (R)- and racemic (Ϯ)-alcohols 8 were prepared from their
corresponding (R)- and racemic glycidols according to ref.[7]
[11] (+)-(R)-Mosher’s acid [(+)-(R)-MTPA] was used as chiral deri-
vating agent.
[12] Selected references: a) G. Wang, Anti-Infective Agents Med.
Chem. 2008, 7, 32–49 and references cited therein; b) G. Zap-
pia, P. Menendez, G. D. Monache, D. Misiti, L. Nevola, B.
Botta, Mini-Rev. Med. Chem. 2007, 7, 389–409; c) J.-F. Li, R.
Yang, L. Nie, L.-J. Yang, R. Huang, J. Lin, J. Pept. Res. 2005,
66, 319–323; d) S. Aubry, S. Pellet-Rostaing, M. Lemaire, Eur.
J. Org. Chem. 2007, 5212–5225; e) D. A. Evans, Aldrichim. Acta
1982, 15, 23–32; f) D. J. Ager, I. Prakash, D. R. Schaad, Ald-
richim. Acta 1997, 30, 3–12; g) D. J. Ager, I. Prakash, D. R.
Schaad, Chem. Rev. 1996, 96, 835–875; h) J. L. Charlton, G.-
L. Chee, Can. J. Chem. 1997, 75, 1076–1083; i) J. T. Randolph,
P. Waid, C. Nichols, D. Sauer, F. Haviv, G. Diaz, G. Bammert,
L. M. Besecke, J. A. Segreti, K. M. Mohning, E. N. Bush, C. D.
Supporting Information (see footnote on the first page of this arti-
cle): Experimental procedures, characterization data and NMR
spectra for compounds 8–15.
5690
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 5687–5691

Heterolignans
Wegner, J. Greer, J. Med. Chem. 2004, 47, 1085–1097; j) C.
D. K. Pyun, C. H. Lee, H.-J. Ha, C. S. Park, J.-W. Chang,
W. K. Lee, Org. Lett. 2001, 3, 4197–4199.
Murruzzu, M. Alonso, A. Canales, J. Jimenez-Barbero, A. Ri-
era, Carbohydr. Res. 2007, 342, 1805–1812; k) M. S. M. Pear-
son, M. Evain, M. Mathé-Allainmat, J. Lebreton, Eur. J. Org.
Chem. 2007, 4888–4894; l) M. S. M. Pearson, R. O. Saad, T.
Dintinger, H. Amri, M. Mathé-Allainmat, J. Lebreton, Bioorg.
Med. Chem. Lett. 2006, 16, 3262–3267.
[16]
[17]
Masked carbonyl groups are known to give cleaner reactions
in several cases than the corresponding carbonyl groups, see:
C. C. Silveira, A. S. Vieira, T. S. Kaufman, Tetrahedron Lett.
2006, 47, 7545–7549 and references cited therein.
a) E. D. Cox, J. M. Cook, Chem. Rev. 1995, 95, 1797–1842; b)
E. L. Larghi, M. Amongero, A. B. J. Bracca, T. S. Kaufman,
ARKIVOC 2005, 12, 98–153; c) B. E. Maryanoff, H.-C. Zhang,
J. H. Cohen, I. J. Turchi, C. A. Maryanoff, Chem. Rev. 2004,
104, 1431–1628.
[13] a) J.-N. Denis, H. Mauger, Y. Vallée, Tetrahedron Lett. 1997,
38, 8515–8518; b) H. Chalaye-Mauger, J.-N. Denis, M.-T. Aver-
buch-Pouchot, Y. Vallée, Tetrahedron 2000, 56, 791–804; c) M.
David, J.-N. Denis, C. Philouze, A. Durif, Y. Vallée, C. R.
Chim. 2004, 7, 41–44.
[14] In this reaction, the use of electron-donating alkyl protecting
groups, such as the methyl group to give the dimethylamino
moiety, led to the degradation of the reaction mixture probably
by initial formation of the corresponding azafulvenium salt by
removal of dimethylamine.
[18] K. Tomioka, Y. Kubota, K. Koga, Tetrahedron 1993, 49, 1891–
1900, and references therein.
[19] The coupling constant between the same hydrogen atoms in
the 2-aza-4-methylepipodophyllotoxin is 1.8 Hz for the syn re-
lationship, see ref.[18]
Received: September 10, 2008
Published Online: October 23, 2008
[15] a) D. H. R. Barton, J. Camara, P. Dalko, S. D. Géro, B.
Quiclet-Sire, P. Stütz, J. Org. Chem. 1989, 54, 3764–3766; b)
Eur. J. Org. Chem. 2008, 5687–5691
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5691