New cytotoxic 1,2,4-thiadiazole alkaloids from the ascidian Polycarpa aurata

Article abstract of DOI:10.1021/ol400791n

Two new alkaloids, polycarpathiamines A and B (1 and 2), were isolated from the ascidian Polycarpa aurata. Their structures were unambiguously determined by 1D, 2D NMR, and HRESIMS measurements and further confirmed by comparison with a closely related analogue, 3-dimethylamino-5-benzoyl-1,2,4-thiadiazole (4), that was prepared by chemical synthesis. Compounds 1 and 2 both feature an uncommon 1,2,4-thiadiazole ring whose biosynthetic origin is proposed. Compound 1 showed significant cytotoxic activity against L5178Y murine lymphoma cells (IC₅₀ 0.41 μM).

Full text of DOI:10.1021/ol400791n

ORGANIC
LETTERS
2013
Vol. 15, No. 9
2230–2233
New Cytotoxic 1,2,4-Thiadiazole Alkaloids
from the Ascidian Polycarpa aurata
Cong-Dat Pham,† Horst Weber,‡ Rudolf Hartmann,§ Victor Wray,^ Wenhan Lin,
Daowan Lai,*,† and Peter Proksch*,†
Institute of Pharmaceutical Biology and Biotechnology, and Institute of Pharmaceutical
and Medicinal Chemistry, Heinrich-Heine University, 40225 Du€sseldorf, Germany,
Institute of Complex Systems (ICS-6), Forschungszentrum Ju€lich GmbH, 52425 Ju€lich,
Germany, Helmholtz Centre for Infection Research, D-38124 Braunschweig, Germany,
and State Key Laboratory of Natural and Biomimetic Drugs, Peking University,
Beijing 100191, P. R. China
laidaowan123@gmail.com; proksch@uni-duesseldorf.de
Received March 24, 2013
ABSTRACT
Two new alkaloids, polycarpathiamines A and B (1 and 2), were isolated from the ascidian Polycarpa aurata. Their structures were unambiguously
determined by 1D, 2D NMR, and HRESIMS measurements and further confirmed by comparison with a closely related analogue, 3-dimethylamino-
5-benzoyl-1,2,4-thiadiazole (4), that was prepared by chemical synthesis. Compounds 1 and 2 both feature an uncommon 1,2,4-thiadiazole ring
whose biosynthetic origin is proposed. Compound 1 showed significant cytotoxic activity against L5178Y murine lymphoma cells (IC50 0.41 μM).
Marine invertebrates such as ascidians are prominent
sourcesofawide varietyofnatural products, many of them
containing nitrogen atoms.1There is a high probability of
discovering novel structures with unprecedented skeletons
from these organisms as exemplified by ecteinascidin
743 (trabectedin), eudistomin C, dendrodoine, and lisso-
clinotoxin A.2 Ascidians of the genus Polycarpa are known
for producing various sulfur-containing alkaloids such as
polycarpaurines AꢀC,3 polycarpamines AꢀE,4 the di-
meric disulfide alkaloid polycarpine,5 and derivatives6
thereof. These compounds exhibited diverse biological
activities, such as antifungal activity,4 cytotoxic3,6a,7 and
antitumor activities,7 inhibition ofinosine monophosphate
dehydrogenase,5 inhibition of inflammatory cytokine
production in lipopolysaccharide-stimulated RAW
264.7 cells,8 and induction of apoptosisinJB6cellsthrough
p53- and caspase 3-dependent pathways.9 Thus, members
† Institute of Pharmaceutical Biology and Biotechnology, Heinrich-Heine
University.
‡ Institute of Pharmaceutical and Medicinal Chemistry, Heinrich-Heine
University.
(3) Wang, W.; Oda, T.; Fujita, A.; Mangindaan, R. E. P.; Nakazawa,
T.; Ukai, K.; Kobayashi, H.; Namikoshi, M. Tetrahedron 2007, 63, 409.
(4) Lindquist, N.; Fenical, W. Tetrahedron Lett. 1990, 31, 2389.
(5) Abbas, S. A.; Hossain, M. B.; van der Helm, D.; Schmitz, F. J.;
Laney, M.; Cabuslay, R.; Schatzman, R. C. J. Org. Chem. 1996, 61,
2709.
§ Institute of Complex Systems (ICS-6).
^ Helmholtz Centre for Infection Research.
Peking University.
(1) Blunt, J. W.; Copp, B. R.; Keyzers, R. A.; Munro, M. H. G.;
Prinsep, M. R. Nat. Prod. Rep. 2013, 30, 237.
(2) (a) Rinehart, K. L.; Holt, T. G.; Fregeau, N. L.; Keifer, P. A.;
Wilson, G. R.; Perun, T. J.; Sakai, R.; Thompson, A. G.; Stroh, J. G.;
Shield, L. S.; Seigler, D. S.; Li, L. H.; Martin, D. G.; Grimmelikhuijzen,
(6) (a) Kang, H.; Fenical, W. Tetrahedron Lett. 1996, 37, 2369. (b)
€
Wessels, M.; Konig, G. M.; Wright, A. D. J. Nat. Prod. 2001, 64, 1556.
(7) Popov, A. M.; Novikov, V. L.; Radchenko, O. S.; Elyakov, G. B.
Dokl. Biochem. Biophys. 2002, 385, 213.
(8) Oda, T.; Lee, J.-S.; Sato, Y.; Kabe, Y.; Sakamoto, S.; Handa, H.;
Mangindaan, R. E. P.; Namikoshi, M. Mar. Drugs 2009, 7, 589.
(9) Fedorov, S. N.; Bode, A. M.; Stonik, V. A.; Gorshkova, I. A.;
Schmid, P. C.; Radchenko, O. S.; Berdyshev, E. V.; Dong, Z. Pharm.
Res. 2004, 21, 2307.
€
C. J. P.; Gade, G. J. Nat. Prod. 1990, 53, 771. (b) Rinehart, K. L.;
Kobayashi, J.; Harbour, G. C.; Hughes, R. G.; Mizsak, S. A.; Scahill,
T. A. J. Am. Chem. Soc. 1984, 106, 1524. (c) Heitz, S.; Durgeat, M.;
Guyot, M.; Brassy, C.; Bachet, B. Tetrahedron Lett. 1980, 21, 1457. (d)
Litaudon, M.; Guyot, M. Tetrahedron Lett. 1991, 32, 911.
r
10.1021/ol400791n
Published on Web 04/12/2013
2013 American Chemical Society

of this genus are interesting sources of potentially new
bioactive metabolites.
as one sulfur, and two nitrogen atoms remained to be
assigned according to the molecular formula. Since the
benzoyl group accounts for 5 DB, the remaining 3DB have
to originate from the ꢀC2N2SꢀNH2 moiety. This moiety
(C2N2S) has to be cyclic, as otherwise a linear substructure
with an alkyne group or cumulative double bonds must be
formulated which is not supported by the carbon-13
chemical shifts. Theoretically, there are four possible
heterocyclic ring structures for the ꢀC2N2Sꢀ moiety of 1
(h-1ꢀh-4, Figure 2). A thorough literature search for the
possible ring structures was carried out, and the chemical
shifts for the carbons in these heterocyclic rings were either
extracted based on reported values, or assigned by NMR
measurements for those compounds that were commer-
cially available (see Table S1 in the Supporting Infor-
mation). The 1,2,4-thiadiazole substructure (h-3) was
found to be the only possible heterocyclic ring in which
one carbon resonates at around 170 ppm, and the second
carbon at 180ꢀ190 ppm, which is very similar to the
chemical shifts obtained for 1. However, it was difficult
to assign the positions of the NH2 and benzoyl groups in
this heterocyclic ring structure, since the sulfur bearing
carbon (C5) generally resonates at 180ꢀ190 ppm regard-
less of the substitution (Table S1). Thus, two possible
structures (i.e., 1a and 1b) remained (Figure 2). We
strongly favored 1a, a guanidine containing molecule,
whose biogenetic origin could be easily explained since
this substructure is also present in polycarpaurine C (3),
and N,N-didesmethylgrossularine-1 that were likewise
isolated from P. aurata. Moreover, an indole alkaloid,
dendrodoine, featuring a similar 3-dimethylamino-1,2,4-
thiadiazole moiety had previously been reported to be
from the ascidian Dendrodoa grossularia.2c A comparison
of the 13C NMR data of 1 with those of dendrodoine2c,11
revealed close similarities in the 1,2,4-thiadiazole unit if
one assumes that the original assignments for C-5 and the
carbonyl group of dendrodoine are wrong and should be
In our search for bioactive metabolites from marine
organisms,10 we observed that the ethanolic extract of
the ascidian Polycarpa aurata, collected in Indonesia,
completely inhibited the growth of the murine lymphoma
cell line L5178Y at a concentration of 10 μg/mL. Chro-
matographic separation afforded two new alkaloids (1ꢀ2)
(Figure 1), together with four known compounds, which
were identifiedaspolycarpaurineC (3),3 N,N-didesmethyl-
grossularine-1,5 4-methoxybenzoic acid,6b and N-methyl-
2-(4-methoxyphenyl)-2-oxoacetamide6b based on their
spectral data and on comparison with the literature. Here-
in, we describe the structure elucidation of the new compounds
1 and 2 and cytotoxicity of the isolated compounds.
Figure 1. Structures of compounds 1ꢀ4 and dendrodoine.
Compound 1 was obtained as a dark yellow amorphous
solid with a faint sulfur odor. The HRESIMS of 1 showed
pseudomolecular ion peaks at m/z 222.0331 [M þ H]þ,
244.0148 [M þ Na]þ, and 465.0406 [2M þ Na]þ indicating
the molecular formula C9H7N3O2S, bearing 8 double bond
15
interchanged.12 A 1Hꢀ N HMBC measurement of 1 was
performed to distinguish between the two possible struc-
tures 1a and 1b (Figure 2). The observed two three-bond
correlations from the NH2 group to two nitrogen atoms
(δN 171.3, 202.2) clearly indicate 1a to be the correct
structure for compound 1.
In order to further confirm the existence of this unusual
heterocyclic ring structure in 1, we synthesized an analo-
gous structure, 3-dimethylamino-5-benzoyl-1,2,4-thiadi-
azole (4), which closely resembles 1 and dendrodoine,
following a method reported in the literature.11 The ana-
logue was synthesized according to a two-step reaction: (1)
the formation of 5-dimethylamino-1,3,4-oxathiazol-2-one,
which was used to produce N,N-dimethylaminonitrile sulfide
in the following step via thermolysis; (2) a 1,3-dipolar
1
equivalents (8 DB). The H NMR spectrum (DMSO-d6)
exhibited signals at δH 8.37 (2H, d, J = 8.8 Hz, H-20/60),
and 6.94 (2H, d, J = 8.8 Hz, H-30/50), suggesting a typical
AA0BB0 spin system of a parasubstitutedbenzenering, and
further signals at δH 10.79 (1H, br.s) for a hydroxy group
and at δH 7.08 (2H, s) for an NH2 group. The 13C NMR
data for C-10ꢀC-70 (DMSO-d6, Table 1) together with the
HMBC correlations from H-20/60 to C-40 (δC 163.8),
C-30/50 (δC 115.6), C-60/20 (δC 133.6), C-10 (δC 125.1), and
C-70 (δC 180.6), and from H-30/50 to C-10, C-40, and C-50/30,
revealed the presence of a 4-hydroxybenzoyl moiety, which
is consistent with the observation of a base peak at m/z 121
(C7H5O2þ) in EI-MS. Apart from this unit, two carbon
atoms (δC 185.7, 171.7), one NH2 group (δH 7.08), as well
(11) Hogan, I. T.; Sainsbury, M. Tetrahedron 1984, 40, 681.
(12) The original assignments (see ref 11) for C-5 at δC 175.7 and the
carbonyl conjugated with an indole group at δC 187.8 were not con-
sistent with the reported 1,2,4-thiadiazole derivatives (usually C-5 at
180ꢀ190 ppm; see Table S1), and the indole-3-carbonyl containing
derivatives (e.g., Rhopaladin D, δCdO 177.4 ( Sato, H.; Tsuda, M.;
Watanabe, K.; Kobayashi, J. Tetrahedron 1998, 54, 8687), and such a
carbonyl usually resonates at lower than 180 ppm).
€
(10) (a) Pham, C.-D.; Hartmann, R.; Muller, W. E. G.; de Voogd, N.;
Lai, D.; Proksch, P. J. Nat. Prod. 2013, 76, 103. (b) Niemann, H.; Lin,
€
W.; Muller, W. E. G.; Kubbutat, M.; Lai, D.; Proksch, P. J. Nat. Prod.
2013, 76, 121. (c) Lai, D.; Liu, D.; Deng, Z.; van Ofwegen, L.; Proksch,
P.; Lin, W. J. Nat. Prod. 2012, 75, 1595.
Org. Lett., Vol. 15, No. 9, 2013
2231

Compound 2 was obtained as a dark yellow amorphous
solid with a slight sulfur odor similar to compound 1.
HRESIMS of 2 showed signals at m/z 236.0488 [M þ H]þ
and 258.0307 [M þ Na]þ, indicatingthe molecular formula
C10H9N3O2S, which differs from that of 1 by a methyl
substituent. The NMR data of 2 closely resemble those of
1, except for a methoxy group (δH 3.90, δC 55.8) in 2
instead of a hydroxyl substituent in 1. The methoxy group
is located at C-40, since it shows an HMBC correlation to
C-40. Thus, compound 2 was identified as 3-amino-5-(4-
methoxybenzoyl)-1,2,4-thiadiazole and named polycar-
pathiamine B.
Scheme 1. Synthesis of 3-Dimethylamino-5-benzoyl-1,2,4-thia-
diazole (4)
Figure 2. Assignment of the heterocyclic unit in 1 by analysis of
the 13C NMR chemical shifts and 1Hꢀ N HMBC correlations.
15
cycloaddition between the nitrile sulfide generated in situ
and benzoylcyanide (Scheme 1). This synthetic analogue
(4) hastwoN-methyl groups, which are helpful inassigning
the chemical shifts of the thiadiazole unit. The HMBC
correlations from both N-methyl groups to C-3 (δC 172.5)
and from H-20/60 to C-70 (δC 182.9) allow the unambiguous
assignments for C-3 and C-70; thus the signal at δC 184.9
can only be assigned to C-5, which supports our
assumption that the previously reported assignment
for C-5 of dendrodoine was wrong. The strong similar-
ity between the carbon chemical shifts of 1 and the
synthetic analogue (4) (Table 1) in the thiadiazole unit
independently supports the proposed structure for 1.
Therefore, compound 1 was determined as the new
3-amino-5-(4-hydroxybenzoyl)-1,2,4-thiadiazole and
named polycarpathiamine A.
1,2,4-Thiadiazole alkaloids are rarely encountered in
natural products. So far, only three natural products with
this structural unit have been reported, including dendro-
doine from the ascidian Dendrodoa grossularia,2c and a
pair of enantiomers of an indole alkaloid from the root of
Isatis indigotica in a recent paper.13 The present study
reports two new members of this class. It is worth mention-
ing that compounds 1 and 2 are the second and third
example of 3-amino substituted 1,2,4-thiadiazole alkaloids
from nature. Thus, their biosynthetic pathway should
be different from that of the 1,2,4-thiadiazole alkaloid
without a 3-amino substitution.13 A plausible biosynthetic
pathway for 1 and 2 is proposed (Scheme 2).
In light of the frequent occurrence of the dimeric dis-
ulfide alkaloid, polycarpine, in several ascidians from the
genus Polycarpa,3,5,6a though not in the present study, we
hypothesize a monomeric precursor (m1) being present in
P. aurata, which would undergo methylation and oxida-
tion to form a S-radical intermediate. A dimerization of
this intermediate could afford polycarpine, in which the
cleavage of the SꢀS or CꢀS bond accompanied by oxida-
tion would give polycarpaurine B3 or C (3), respectively.
Further oxidation of m1 and subsequent nucleophilic
attack of water could produce an intermediate (m2), as
similar derivatives with a 2-carbonyl group were already
reported from ascidians.5,6a A follow-up ring-opening
Table 1. 13C NMR Data of 1-2, and 4
position
1a
1b
2c
4d
3
171.7 C
185.7 C
125.1 C
133.6 CH
115.6 CH
163.8 C
180.6 C
172.7 C
187.6 C
127.2 C
134.7 CH
116.5 CH
164.5 C
181.7 C
171.8 C
185.4 C
126.6 C
133.3 CH
114.3 CH
164.6 C
180.9 C
55.8 CH3
172.5 C
5
10
184.9 C
134.0 C
20 /60
30 /50
40
130.6 CH
128.9 CH
134.7 CH
182.9 C
70
40-OMe
3-NMe2
38.6 CH3
a Recorded at 100 MHz in DMSO-d6. b Recorded at 150 MHz in
acetone-d6. c Recorded at 75 MHz in DMSO-d6. d Recorded at 150 MHz
in DMSO-d6.
(13) Chen, M.; Lin, S.; Li, L.; Zhu, C.; Wang, X.; Wang, Y.; Jiang, B.;
Wang, S.; Li, Y.; Jiang, J.; Shi, J. Org. Lett. 2012, 14, 5668.
2232
Org. Lett., Vol. 15, No. 9, 2013

P. aurata.5 Oxidative ring closure of m3 would give rise to
compounds 1 and 2. Even though dendrodoine was iso-
lated from an unrelated ascidian,2c it seems plausible that a
similar biosynthetic pathway was also adopted here by
using a precursor that features an indole group instead of
the benzene group in m1, as suggested by the occurrence of
several indole analogues14 that resemble those hypotheti-
cal intermediates shown in Scheme 2.
Scheme 2. Plausible Biosynthetic Pathway for 1 and 2
All isolated compounds were examined for their effects on
the growth of L5178Y mouse lymphoma cells in vitro using the
MTT assay.10a,15 Compound 1 and N,N-didesmethylgrossu-
larine-1 inhibited the growth of L5178Y cells with IC50 values
(μM) of 0.41 and 1.86, respectively, while the other compounds
were not active (IC50 >10 μM). Compound 1 showed potent
inhibitory activity in a submicromolar concentration, which is
approximately 10-fold more active than the positive control
kahalalide F (IC50 4.3 μM). Apparently, the hydroxyl func-
tional group at C-40 of compound 1should be unsubstituted in
order to retain bioactivity since compound 2 was inactive. The
unique structure and notable cytotoxicity of 1 make it an
interesting starting point for cytotoxic drug development.
Further SAR and biological investigations are ongoing.
Acknowledgment. The financial support by a grant of
the BMBF to P.P. is gratefully acknowledged. We are
€
indebted to Prof. W. E. G. Muller (Johannes Gutenberg
University, Mainz, Germany) for cytotoxicity assays and
Dr. Nicole de Voogd (Naturalis Biodiversity Center, Lei-
den, The Netherlands) for identifying the ascidian. We
want to thank Prof. Phillip Crews (University of Califor-
nia, Santa Cruz, United States) for helpful discussions.
Supporting Information Available. Experimental pro-
cedures, ascidian material, extraction and isolation, char-
acterization data of 1ꢀ2, synthetic procedures and spectral
data of 4, cytotoxic assay, and copies of 1D and 2D NMR
and HRMS spectra of 1, 2, and 4. This material is available
free of charge via the Internet at http://pubs.acs.org.
reaction of m2 would give the guanidine amide of 2-aryl-2-
oxothioacetic acid (m3). A similar compound N-methyl-
(4-methoxyphenyl)-2-oxothioacetamide was isolated from
(15) Ashour, M.; Edrada, R.; Ebel, R.; Wray, V.; Watjen, W.;
€
Padmakumar, K.; Muller, W. E. G.; Lin, W. H.; Proksch, P. J. Nat.
Prod. 2006, 69, 1547.
(14) Mancini, I.; Guella, G.; Pietra, F.; Debitus, C.; Duhet, D. Helv.
Chim. Acta 1994, 77, 1886.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 9, 2013
2233