Total synthesis of (+)-madangamine D

Article abstract of DOI:10.1002/anie.201402263

Madangamines are a group of bioactive marine sponge alkaloids, embodying an unprecedented diazapentacyclic skeletal type. The enantioselective total synthesis of madangamine D has been accomplished, and represents the first total synthesis of an alkaloid of the madangamine group. It involves the stereoselective construction of the diazatricyclic ABC core using a phenylglycinol-derived lactam as the starting enantiomeric scaffold and the subsequent assembly of the peripheral macrocyclic rings. The synthesis provides, for the first time, a pure sample of madangamine D and confirms the absolute configuration of this alkaloid family.

Full text of DOI:10.1002/anie.201402263

Angewandte
Chemie
DOI: 10.1002/anie.201402263
Natural Product Synthesis
Total Synthesis of (+)-Madangamine D**
Roberto Ballette, Maria Pꢀrez, Stefano Proto, Mercedes Amat,* and Joan Bosch
Abstract: Madangamines are a group of bioactive marine
sponge alkaloids, embodying an unprecedented diazapenta-
cyclic skeletal type. The enantioselective total synthesis of
madangamine D has been accomplished, and represents the
first total synthesis of an alkaloid of the madangamine group. It
involves the stereoselective construction of the diazatricyclic
ABC core using a phenylglycinol-derived lactam as the starting
enantiomeric scaffold and the subsequent assembly of the
peripheral macrocyclic rings. The synthesis provides, for the
first time, a pure sample of madangamine D and confirms the
absolute configuration of this alkaloid family.
Structurally, madangamines are pentacyclic alkaloids with
an unprecedented skeletal type, characterized by a diazatri-
cyclic core (ABC rings) bearing three contiguous stereogenic
centers, one of them quaternary, and two linear carbon
bridges which connect N7 to C9 (D ring) and N1 to C3
(E ring). The peripheral macrocyclic ring D is different in
each madangamine, in size as well as in degree and position of
unsaturation, whereas ring E is identical in madangamines A–
E but different in madangamine F, which also incorporates
a C4 hydroxy group (Figure 1).
Sponges of the order Haplosclerida have proven to be a rich
source of structurally diverse but biogenetically related
alkaloids,[1] most of which display significant biological
activities. These marine alkaloids comprise a great variety
of unusual skeletal types, including an array of complex
polycyclic diamine structures bearing macrocyclic rings, such
as saraines, ingenamines, manzamines, nakadomarin A, and
madangamines.[2] Madangamines are one of the least studied
of these alkaloids from a synthetic standpoint,[3] and no total
synthesis on this series has been reported so far.[4] The first
isolation of an alkaloid of this group was madangamine A,
which was reported by Andersen and co-workers in 1994[5] to
have been found in the marine sponge Xestospongia ingens,
collected in Papua New Guinea. A few years later the same
team described[6] four new related alkaloids, madangami-
nes B–E,[7] from the same organism, and more recently,
Berlinck and co-workers reported the isolation of madanga-
mine F from the Brazilian sponge Pachychalina alcaloidi-
fera.[8] Madangamines A and F have shown significant in vitro
cytotoxicity against a number of tumor cell lines. However, no
bioactivity data have been reported for madangamines B–E,
and further pharmacological research on this alkaloid group
has been thwarted by the minute amounts of alkaloid samples
available from natural sources.
Figure 1. Alkaloids of the madangamine group.
We present herein the enantioselective synthesis of
(+)-madangamine D, and it provides, for the first time,
a pure sample[7] of this natural product and constitutes the
first total synthesis of an alkaloid of the madangamine group.
By using a phenylglycinol-derived bicyclic lactam[9] as the
starting enantiomeric scaffold,[10] our approach involves the
initial construction of the bridged diazatricyclic ABC core
common to all madangamines,[11] and the subsequent building
of the macrocyclic D and E rings (Figure 2).
[*] R. Ballette, Dr. M. Pꢀrez, Dr. S. Proto, Prof. Dr. M. Amat,
Prof. Dr. J. Bosch
Laboratory of Organic Chemistry, Faculty of Pharmacy and Institute
of Biomedicine (IBUB), University of Barcelona
08028-Barcelona (Spain)
E-mail: amat@ub.edu
Homepage: http://www.ub.edu/sintefarma/
The starting enantiopure lactam 2 was easily accessible[12]
by cyclocondensation of the oxoester 1 with (R)-phenyl-
glycinol, a process which installs the first stereocenter (C5 in
the madangamine numbering)[13] by dynamic kinetic resolu-
tion of the racemic substrate. The key functionalized diaza-
tricyclic intermediates would be prepared from an unsatu-
rated lactam, derived from 2, by successive construction of the
carbocyclic C and piperidine A rings. Crucial stereochemical
issues are the generation of the required B/C cis ring junction,
by a stereoselective conjugate addition reaction followed by
a ring-closing metathesis process, and the control of the C9
[**] Financial support from the Spanish Ministry of Economy and
Competitiveness (Project CTQ2012-35250) and the AGAUR, Gen-
eralitat de Catalunya (Grant 2009-SGR-1111) is gratefully acknowl-
edged. Thanks are also due to the Ministry of Education (Spain) for
a fellowship to R.B. and to PharmaMar S.A. (Madrid) for the
cytotoxicity assays.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201402263.
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
was accomplished by a stereocontrolled cascade amino-
hydroxylation. To this end, once 7 was converted into the
azide 8 via a mesylate, and the cyclohexene double bond was
epoxidized, a Staudinger reduction of 9 led to an intermediate
amino epoxide, which underwent a smooth in situ cyclization.
A subsequent protection of the resulting diazatricyclic alcohol
led to the N-tosyl derivative 10.
With the functionalized diazatricyclic derivative 10 in
hand, the next phase of the synthesis was the construction of
the western 14-membered D ring.[15] After benzylation of the
C3 hydroxy group, selective deprotection of N7 in the
resulting orthogonally protected diamino derivative, followed
by acylation with 7-octenoyl chloride, led to the tricyclic
amide 11. Hydrolysis of the acetal function and Wittig
methylenation of the resulting aldehyde gave the required
dialkene derivative 12. A ring-closing metathesis reaction of
12 under dilute conditions using the first-generation Grubbs
catalyst provided the expected tetracyclic alkene 13 (2:1
mixture of Z/E isomers). A subsequent catalytic hydrogena-
tion, which led to both the reduction of the carbon–carbon
double bond and the removal of the benzyl ether protecting
group, followed by Dess–Martin periodinane oxidation of the
resulting alcohol led to the ketone 14, which served as
a platform to construct the eastern 11-membered E ring.
The (Z,Z)-unsaturated eight-carbon fragment required to
complete the synthesis of madangamine D was incorporated
in a straightforward manner by a Wittig reaction using the
ylide generated from the phosphonium salt 15[16] under strictly
anhydrous conditions. Removal of the tosyl substituent in the
resulting diastereoisomeric mixture of alkenes 16 (2.2:1 Z/E
ratio),[17] followed by hydrolysis of the ester function and
macrolactamization, led to the pentacyclic dilactam 17. A
final LiAlH4 reduction provided madangamine D. The 1H and
13C NMR data of our synthetic madangamine were coincident
with those reported[6] for the natural product (see Tables in
the Supporting Information).
To date, the absolute configuration of madangamines has
only been inferred by correlation with that of their presu-
med[1,2b,d,3a,b] biosynthetic precursors, ingenamines.[18] Given
that our synthetic madangamine D, having unambiguous 2S,
5S, 9R, 12R absolute configuration, has a specific rotation
{[a] =+ 101.3 (c = 0.29, CHCl3)} with the same sign as in the
closely related madangamines A–C,[19] our synthesis confirms
the absolute configuration of this alkaloid family.
Madangamine D showed significant in vitro cytotoxic
activity against human colon HT29 (GI50 4.4 mgmLÀ1) and
pancreas PSN1 (GI50 7.4 mgmLÀ1) cancer cell lines, but was
inactive against lung NSCLC A549 and breast MDA-MB-231
cancer cell lines at the highest assayed concentration
(10 mgmLÀ1).
Figure 2. Synthetic strategy.
stereochemistry in the alkylation step. Finally, the assembly of
the macrocyclic rings would be accomplished by a ring-closing
metathesis reaction (ring D) and a Wittig olefination followed
by macrolactamization (ring E).
The overall synthetic sequence is shown in Scheme 1. The
lactam 2 was converted in excellent overall yield to the
unsaturated lactam 3, through an epimeric mixture of
intermediate seleno derivatives. A stereoselective, stereo-
electronically controlled,[14] conjugate addition of an allyl
residue led to the cis-diallyl-substituted lactam 4, from which
the carbocyclic C ring was constructed by a ring-closing
metathesis reaction to give the cis-octahydroisoquinolone
derivative 5. A stereoselective alkylation from the most
accessible face of the b-ketoester moiety of 5 generated the
quaternary C9 stereocenter in 6 and installed a C9 function-
alized carbon chain. At this point, the removal of the
phenylethanol moiety from the chiral auxiliary was achieved
by successive treatment of 6 with Na in liquid NH3, which
À
caused the cleavage of the benzylic C N bond, and LiAlH4,
which brought about the reduction of the resulting unstable a-
oxylactam. Under the latter conditions, the lactam and ester
carbonyl functions were also reduced to give an N-unsub-
stituted piperidine-3-methanol derivative, which was imme-
diately protected as the N-Boc piperidine 7. In this way, the
tert-butoxycarbonyl group not only provided activation
towards the conjugate addition step and allowed the stereo-
selective alkylation of the 1,3-dicarbonyl intermediate 5, but
also serves as the precursor of the aminomethyl chain
required for the closure of the piperidine A ring. The latter
By using appropriately C9-substituted diazatricyclic
derivatives, the strategy we have developed could be applied
to the synthesis of other members of the madangamine
group.[20]
Received: February 18, 2014
Published online: && &&, &&&&
2
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Scheme 1. Enantioselective synthesis of madangamine D. Reagents and conditions: a) LiHMDS, (Boc)2O; then C6H5SeCl, THF, À788C, 93%;
=
b) H2O2, CH2Cl2, RT, 2 h; c) CH2 CHCH2MgBr, CuI, LiCl, TMSCl, THF, À788C, 20 h, 82% (from the seleno derivative); d) Grubbs second
generation catalyst, CH2Cl2, RT, 18 h, 80%; e) NaH, (CH2O)2CH(CH2)3Br, TBAI, DMF, RT, 18 h, 90%; f) Na/liq. NH3, À338C, 2 min; then LiAlH4,
1,4-dioxane, reflux, 20 h; then (Boc)2O, CH2Cl2, RT, 4 h, 45%; g) Et3N, MsCl, CH2Cl2, RT, 4 h; h) NaN3, DMF, 908C, 48 h, 79% (from 7); i) m-
CPBA, CH2Cl2, RT, 5 h; j) Me3P, THF, 1 h; then H2O, RT, 20 h; k) p-TsCl, Et3N, CH2Cl2, 08C, 2.5 h, 60% (from 8); l) NaH, BnBr, TBAI, DMF, 0.05m,
=
RT, 20 h, 80%; m) TFA, CH2Cl2, RT, 30 min; then ClCO(CH2)5CH CH2, Et3N, CH2Cl2, 08C, 3 h; then RT, 18 h, 92%; n) HCl, THF, RT, 2 h; then
KOtBu, BrÀ Ph3P+CH3, THF, RT, 20 h, 80%; o) Grubbs first generation catalyst, CH2Cl2, 0.2 mm, reflux, 12 h; p) H2, Pd/C, EtOH, RT, 24 h; then
+
Dess–Martin, CH2Cl2, RT, 4 h, 75% (from 12); q) NaHMDS, BrÀ (Z)-Ph3P CH2CH2CH CH(CH2)3CO2Me (15), THF, 0 to 608C, 70%; r) Na,
=
naphthalene, THF, À788C; s) aq. LiOH, THF; then EDCI, HOBt, DMF/CH2Cl2 (1:9), 0.02m, RT, syringe pump, 75% (from 16); t) LiAlH4, THF, RT,
3 h, 68%. Boc=tert-butoxycarbonyl, DMF=N,N-dimethylformamide, EDCI=1-(3-dimethylaminopropyl)-3-ethyl-carbodiimide hydrochloride,
HMDS=hexamethyldisilazide, HOBt=1-hydroxybenzotriazole, Ms=methanesulfonyl, TBAI=tetra-n-butylammonium iodide, TFA=trifluoroacetic
acid, THF=tetrahydrofuan, TMS=trimethylsilyl, Ts=4-toluenesulfonyl.
K. V. Rao, Y.-M. Choo, M. T. Hamann in Modern Alkaloids:
Structure, Isolation, Synthesis and Biology, (Eds.: E. Fattorusso,
O. Taglialatela-Scafati), Wiley-VCH, Weinheim, 2008, chap. 8,
pp. 189 – 232.
Keywords: alkaloids · asymmetric synthesis · macrocycles ·
nitrogen heterocycles · total synthesis
.
[3] For reviews, see: a) N. Matzanke, R. J. Gregg, S. M. Weinreb,
Org. Prep. Proced. Int. 1998, 30, 1 – 51; b) E. Magnier, Y.
Langlois, Tetrahedron 1998, 54, 6201 – 6258; c) A. Nishida, T.
Nagata, M. Nakagawa, Top. Heterocycl. Chem. 2006, 5, 255 –
280; d) R. Duval, E. Poupon in Biomimetic Organic Synthesis
(Eds.: E. Poupon, B. Nay), Wiley-VCH, Weinheim, 2011, chap. 6,
pp. 181 – 224.
[4] Previous to our work in this field, only four synthetic approaches
to the diazatricyclic ABC core of the madangamines,[4a–d] and the
construction of the 11-membered E macrocycle[4e] common to
madangamines A–E from a model azabicyclic derivative had
[1] R. J. Andersen, R. W. M. Van Soest, F. Kong in Alkaloids:
Chemical and Biological Perspectives, Vol. 10 (Ed.: S. W. Pellet-
ier), Pergamon, New York, 1996, pp. 301 – 355.
[2] a) A. M. P. Almeida, R. G. S. Berlinck, E. Hajdu, Quim. Nova
1997, 20, 170 – 185; b) J. Rodrꢀguez in Studies in Natural Products
Chemistry, Vol. 24 (Ed.: A.-U. Rahman), Elsevier, Oxford, 2000,
pp. 573 – 681; c) J.-F. Hu, M. T. Hamann, R. Hill, M. Kelly in The
Alkaloids: Chemistry and Biology, Vol. 60 (Ed.: G. A. Cordell),
Elsevier, San Diego, CA, 2003, pp. 207 – 285; d) R. G. S. Ber-
linck, Top. Heterocycl. Chem. 2007, 10, 211 – 238; e) J. Peng,
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
been reported: a) N. Matzanke, R. J. Gregg, S. M. Weinreb, M.
Parvez, J. Org. Chem. 1997, 62, 1920 – 1921; b) N. Yamazaki, T.
Kusanagi, C. Kibayashi, Tetrahedron Lett. 2004, 45, 6509 – 6512;
c) H. M. Tong, M.-T. Martin, A. Chiaroni, M. Benechie, C.
Marazano, Org. Lett. 2005, 7, 2437 – 2440; d) J. Quirante, L.
Paloma, F. Diaba, X. Vila, J. Bonjoch, J. Org. Chem. 2008, 73,
768 – 771; e) Y. Yoshimura, J. Inoue, N. Yamazaki, S. Aoyagi, C.
Kibayashi, Tetrahedron Lett. 2006, 47, 3489 – 3492.
b) H. Wong, E. C. Garnier-Amblard, L. S. Liebeskind, J. Am.
Chem. Soc. 2011, 133, 7517 – 7527.
[11] For our initial studies on the enantioselective construction of the
diazatricyclic core of madangamines, see: M. Amat, M. Pꢁrez, S.
Proto, T. Gatti, J. Bosch, Chem. Eur. J. 2010, 16, 9438 – 9441.
[12] a) M. Amat, M. Pꢁrez, A. T. Minaglia, N. Casamitjana, J. Bosch,
Org. Lett. 2005, 7, 3653 – 3656; b) M. Amat, M. Pꢁrez, A. T.
Minaglia, B. Peretto, J. Bosch, Tetrahedron 2007, 63, 5839 – 5848.
[13] For the sake of clarity, the madangamine numbering is used
throughout this manuscript, even for the synthetic intermediates.
[14] a) P. Deslongchamps in Stereoelectronic Effects in Organic
Chemistry, Pergamon, Oxford, 1983, p. 221; b) P. Perlmutter in
Conjugate Addition Reactions in Organic Synthesis, Pergamon,
Oxford, 1992, p. 25.
[15] For a preliminary account on the construction of the macrocyclic
D ring, see: M. Amat, R. Ballette, S. Proto, M. Pꢁrez, J. Bosch,
Chem. Commun. 2013, 49, 3149 – 3151.
[16] J. Sandri, J. Viala, J. Org. Chem. 1995, 60, 6627 – 6630.
[17] For a model study on the construction of the (Z,Z)-unsaturated
11-membered E ring using 15, see: S. Proto, M. Amat, M. Pꢁrez,
R. Ballette, F. Romagnoli, A. Mancinelli, J. Bosch, Org. Lett.
2012, 14, 3916 – 3919. Starting from a simplified bicyclic AC ring
analogue of 14, the Wittig olefination took place with a high
stereoselectivity (Z/E 10:1).
[5] F. Kong, R. J. Andersen, T. M. Allen, J. Am. Chem. Soc. 1994,
116, 6007 – 6008.
[6] F. Kong, E. I. Graziani, R. J. Andersen, J. Nat. Prod. 1998, 61,
267 – 271.
[7] Madangamines D and E were isolated as an inseparable mixture.
[8] J. H. H. L. De Oliveira, A. M. Nascimento, M. H. Kossuga, B. C.
Cavalcanti, C. O. Pessoa, M. O. Moraes, M. L. Macedo, A. G.
Ferreira, E. Hajdu, U. S. Pinheiro, R. G. S. Berlinck, J. Nat. Prod.
2007, 70, 538 – 543.
[9] For reviews on the use of phenylglycinol-derived bicyclic d-
lactams in the enantioselective synthesis of piperidine-contain-
ing alkaloids, see: a) D. Romo, A. I. Meyers, Tetrahedron 1991,
47, 9503 – 9569; b) A. I. Meyers, G. P. Brengel, Chem. Commun.
1997, 1 – 8; c) M. D. Groaning, A. I. Meyers, Tetrahedron 2000,
56, 9843 – 9873; d) C. Escolano, M. Amat, J. Bosch, Chem. Eur. J.
2006, 12, 8198 – 8207; e) M. Amat, M. Pꢁrez, J. Bosch, Synlett
2011, 143 – 160; f) M. Amat, M. Pꢁrez, J. Bosch, Chem. Eur. J.
2011, 17, 7724 – 7732.
[18] F. Kong, R. J. Andersen, Tetrahedron 1995, 51, 2895 – 2906.
[19] Madangamine A: [a] =+ 319 (c = 1.0, EtOAc);[5] madangami-
ne B: [a] =+ 150.7 (c = 0.067, EtOAc)[6] madangamine C: [a] =
+ 140.8 (c = 0.09, EtOAc).[6]
[20] For model studies on the construction of the western D ring of
other madangamines see Ref. [17].
[10] Enantiomeric scaffolding strategy: a) T. C. Coombs, M. D.
Lee IV, H. Wong, M. Armstrong, B. Cheng, W. Chen, A. F.
Moretto, L. S. Liebeskind, J. Org. Chem. 2008, 73, 882 – 888;
4
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Natural Product Synthesis
R. Ballette, M. Pꢁrez, S. Proto, M. Amat,*
J. Bosch
&&&& — &&&&
Total Synthesis of (+)-Madangamine D
Mad about madangamines: The first total
synthesis of an alkaloid of the madang-
amine group has been accomplished.
Using a phenylglycinol-derived lactam as
the starting enantiomeric scaffold, the
synthesis of (+)-madangamine D
the six-membered carbocyclic C and
heterocyclic A rings to generate the
functionalized diazatricyclic ABC inter-
mediates and the subsequent assembly
of the peripheral macrocyclic D and
E rings. Ts=4-toluenesulfonyl.
involves the successive construction of
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!