Ilimaquinone and 5-epi-Ilimaquinone: Beyond a Simple Diastereomeric Ratio, Biosynthetic Considerations from NMR-Based Analysis

Article abstract of DOI:10.1071/CH16455

Dactylospongia metachromia and Dactylospongia elegans collected from French Polynesia were studied with a particular focus on the variation of the diastereomeric ratio between ilimaquinone (4) and 5-epi-ilimaquinone (5). More than 100 samples, covering an area of 4100km2, were studied to try to clarify this intriguing issue. Nuclear magnetic resonance appeared as the non-destructive, straightforward technique of choice for a relative quantitative study. A random distribution, unique at that point in nature, is observed and leads to biosynthetic considerations. Biological evaluation of both compounds was also performed and showed moderate discrepancies in cytotoxicity and apoptosis induction.

Full text of DOI:10.1071/CH16455

CSIRO PUBLISHING
Aust. J. Chem.
Full Paper
http://dx.doi.org/10.1071/CH16455
Ilimaquinone and 5-epi-Ilimaquinone: Beyond a Simple
Diastereomeric Ratio, Biosynthetic Considerations from
NMR-Based Analysis
A
A
B
A
Asmaa Boufridi, David Lachkar, Dirk Erpenbeck, Mehdi A. Beniddir,
A
C D
,
C E
,
A E
,
Laurent Evanno, Sylvain Petek,
Ce´cile Debitus, and Erwan Poupon
A
BioCIS, Universite´ Paris-Sud, Centre National de la Recherche Scientifique (CNRS),
Universite´ Paris-Saclay, Chaˆtenay-Malabry 92290, France.
B
Department of Earth and Environmental Sciences and GeoBio-Center LMU,
Ludwig-Maximilians-Universitat Munich, Richard-Wagner-Str. 10, Munich 80333,
¨
Germany.
C
Institut de Recherche pour le De´veloppement (IRD), Ecosyste`mes Insulaires Oce´aniens
(UMR EIO), BP529, Papeete 98713, Polyne´sie-Franc¸aise, France.
D
Current address: Laboratoire des sciences de l’environnement marin (LEMAR),
Institut Universitaire Europe´en de la Mer (IUEM), Technopoˆle Brest-Iroise, Rue Dumont
d’Urville, Plouzane´ 29280, France.
E
Corresponding authors. Email: cecile.debitus@ird.fr; erwan.poupon@u-psud.fr
Dactylospongia metachromia and Dactylospongia elegans collected from French Polynesia were studied with a particular
focus on the variation of the diastereomeric ratio between ilimaquinone (4) and 5-epi-ilimaquinone (5). More than 100
samples, covering an area of 4100 km2, were studied to try to clarify this intriguing issue. Nuclear magnetic resonance
appeared as the non-destructive, straightforward technique of choice for a relative quantitative study. A random
distribution, unique at that point in nature, is observed and leads to biosynthetic considerations. Biological evaluation
of both compounds was also performed and showed moderate discrepancies in cytotoxicity and apoptosis induction.
Manuscript received: 2 August 2016.
Manuscript accepted: 17 November 2016.
Published online: 20 December 2016.
Introduction
paper the first study dedicated to better understanding the
co-existence of these sesquiterpenic quinones in two different
Dactylospongia species (D. metachromia[9] and D. elegans[10]).
Additionally, biosynthetic considerations as well as con-
sequences on the chemical and biological properties of 4 and 5
are provided in this work. The particular point of the ‘real’ origin
of these quinonic sesquiterpenes (i.e. sponges versus endo-
symbionts), even if not known concerning 4 and close con-
geners, was clarified earlier in Lamellodysidea herbacea (earlier
found as Dysidea herbacea, Dysideidae) and Dysidea avara
(Dysideidae) by Faulkner and Unson[11] and Uriz et al.,[12]
respectively.
Sesquiterpenic para-quinones constitute a homogeneous family
of natural substances of marine origin mainly found in sponges
from the order Dictyoceratida[1,2] that display valuable biolog-
ical properties.[3] Selected examples of such sesquiterpene
quinones (Fig. 1) bearing a clerodane decalin – differing from
the endo (D3,4) or exo (D4,11) position of the double bond –
include avarone (1, monosubstituted quinone), and C-18- and
C-19-methoxyavarone (2 and 3, disubstituted quinones).[1,2] Of
particular interest to us are trisubstituted quinones such as
ilimaquinone (4),[4] 5-epi-ilimaquinone (5),[5] recently
described 5,8-di-epi-ilimaquinone (6),[6] or isospongiaquinone
(7)[7] that differ from the configuration at carbons 5 (trans- or
cis-clerodane skeleton) and 8 and/or the position (endo versus
exo) of the decalin double bond. Because of a highly interesting
and unique biological profile (especially the ability of ilima-
quinone to reversibly dislocate the Golgi apparatus of living
cells)[8] and its abundance within specific marine organisms,
ilimaquinone (4) and congeners attracted our attention, partic-
ularly in relation with the above-mentioned stereochemical
particularities distinguishing compounds 4–7. As part of a pro-
gram aiming at the sustainable valorization of marine bio-
diversity from French Polynesia, we wish to disclose in this
Results and Discussion
Ilimaquinone versus 5-epi-Ilimaquinone
Diastereomeric Ratio
D. metachromia samples were collected from 10 different
locations in French Polynesia (Fig. 2) covering three archipel-
agos (Tetiaroa atoll from Socie´te´, Tikehau, Rangiroa, Toau,
Takaroa, Fakarava, Hereheretue, Makemo, and Anuanuraro
from Tuamotu, and finally three sites in Gambier archipe-
lagos), thus corresponding to a covered area of ,4100 km2.
Journal compilation ꢀ CSIRO 2016
www.publish.csiro.au/journals/ajc

B
A. Boufridi et al.
• disubstituted quinones:
• monosubstituted quinones:
terpene cyclase.[15] The successive formation of two carboca-
tionic intermediates 9 and 10 may then proceed, resulting from
peri-planar Wagner–Meerwein hydrogen and methyl shifts. At
the last stage from 10, two pathways (1 or 2) may finally explain
the outcome towards formation of ilimaquinone (4) versus
5-epi-ilimaquinone (5) before termination of the cascade via a
proton loss of carbocations 11 and 12, respectively. At this stage,
aside from a spontaneous epimerization at carbon 5 (never
observed experimentally), at least two distinct explanations may
arise: (i) a variable expression of two distinct terpene cyclases
affording 4 or 5; (ii) a lack of selectivity within the active site of
a single terpene cyclase that may be correlated to some extent to
a certain degree of liberty, especially of intermediate of type 10
that randomly affords both compounds.[16] To the best of our
knowledge, the present study is the first report of a chemical
variation at a non-epimerizable carbon centre (see chemical
reactivity studies below).
19
MeO
19
O
O
18
20
18
17
21
O
16
O
16
15
H
14
Me
Me
H
Me
13
Me
9
1
2
10
5
8
7
3
6
3
4
4
Me12
Me
11Me
Me
avarone (1)
C-18 methoxyavarone (2)
C-19methoxyavarone (3)
• trisubstituted quinones:
MeO
O
O
MeO
O
18
21
O
OH
16
OH
Furthermore, and very interestingly, the antipodal carboca-
tionic intermediate ent-8 was shown, in a sample of D. elegans
collected near Mooloolaba (Queensland, Australia), to be at the
origin of another complete series of structures such as
(þ)-hyatellaquinone (13, Scheme 2).[17] In our case, D. elegans
samples from Marquesas archipelago revealed the presence of
either ilimaquinone (4, Nuku Hiva island) or isospongiaquinone
(7, Tahuata and Fatu Hiva islands, Scheme 3) as the main
compounds, the latter 7 being described for the first time in this
species.
Me
Me
H
H
Me
Me
5
5
4
Me
Me
11
ilimaquinone (4)
5-epi-ilimaquinone (5)
O
MeO
O
MeO
O
OH
O
OH
Me
Chemical Transformations
Me
H
H
Me
Me
Both 4 and 5 could not be individually resolved on classic silica
gel TLC plate or by silica gel preparative flash chromatography.
Fortunately, they could be separated using silver nitrate-
impregnated silica gel with CH2Cl2 and methanol (MeOH) as
the eluents.[17] In order to experimentally counteract the argu-
ments of a possible spontaneous epimerization in the course of
extraction, the reactivity of ilimaquinone was investigated and
the latter was found to be stable under extraction conditions and
prolonged contact with silica gel. Under acidic conditions
(Scheme 4) and following protonation of the exocyclic
double bond, both 4 and 5 provided neomamanuthaquinone
(15)[18] after Wagner–Meerwein-type rearrangements follow-
ing pathway 1 and isospongiaquinone (7)[7] and 5-epi-
isospongiaquinone (14),[19] respectively, according to pathway
2. The formation of 15 in similar conversion yields and similar
ratios to those obtained for 7 and 14 demonstrates that both the
C-5 b-methyl of 4 and the C-5 a-methyl of 5 can undergo
Wagner–Meerwein migration without interconversion. These
results confirm that 4 and 5 are not interconverted during the
extraction process.
8
3
5
4
Me
Me
Me
5,8-di-epi-ilimaquinone (6)
isospongiaquinone (7)
Fig. 1. Selected examples of sesquiterpene quinones of marine origin. The
grey boxes highlight the differences in the sesquiterpene moiety.
Two specimen collected in Fiji Islands were also studied. We
were able to collect a unique set of over 100 samples of
D. metachromia that could be analyzed. Ilimaquinones 4 and 5
are the most abundant compounds in the total methanolic extract
on HPLC chromatograms in D. metachromia. Ilimaquinone
(4)/5-epi-ilimaquinone (5) diastereomeric ratio (dr) could be
1
evaluated rapidly by H NMR analysis of ethylic ether crude
extracts of the sponges (integration of the C-11 exo-methylene
singlets CH2-11 4: 4.40, 4.42 ppm; 5: 4.66, 4.69 ppm in CDCl3).
Data analysis clearly shows discrepancies between locations
and also small variations between specimens from the same
location at different depths. From 100 % ilimaquinone in spe-
cimens from Anuanuraro island to ,90 % 5-epi-ilimaquinone
in specimens from Gambier Islands (Tables 1 and 2), the 4/5
ratio varies in an unpredictably manner.[13,14]
Biological Evaluation
Despite a large amount of biological data available for ilima-
quinone, a lack of comparative data between 4 and 5 prompted
us to evaluate both compounds. To further characterize their
cytotoxicity profiles, we investigated their effects on the pro-
liferation of four human tumour cell lines: cervical adenocar-
cinoma (HeLa), prostate adenocarcinoma (PC3), and uterine
sarcoma (MES-SA) and its multidrug-resistant subline (MES-
SA/Dx5). The results presented in Table 3 revealed that both 4
and 5 displayed more potent cancer cell growth inhibitory
activities against MES-SA. Overall, ilimaquinone 4 was mod-
erately more cytotoxic than epi-ilimaquinone 5. Finally, they
Biosynthetic Considerations
Puzzled by these observations, we turned our attention to the
biosynthetic hypotheses (Scheme 1) of this family of natural
substances; the biosynthesis process may start with the farne-
sylation of an aromatic ring (the precursor of the quinone moi-
ety). A plausible biosynthetic pathway involves initial folding of
the suitable precursor 8 within the active site of a specific

Ilimaquinone and 5-epi-Ilimaquinone
C
Marquuesas
Fatu Hiva
Nuku Hiva
Tahuata
Rangiroa
Tikehau
Makemo
Tetiaroa
Tuamotu
Toau
Fakarava
Société
Hereheretue
Anuanuraro
Gambier
Tematangi
Gambier
500 km
Fig. 2. Map of sample collection sites in French Polynesia.
Table 1. Diasteromeric ratioA between ilimaquinone (4) and 5-epi-
ilimaquinone (5) from samples of D. metachromia collected from
different areas of French Polynesia and Fiji
Table 2. Ilimaquinone (4)/5-epi-ilimaquinone (5) diastereomeric
ratioA of D. metachromia collected at different depth in two areas at
Rangiroa atoll (Tiputa pass and Avatoru pass)
Location
Number of samples
% of 4/5 ratio mean value
(standard deviation)B
Depth of collection (m)
Tiputa passB
Avatoru passC
7
76/24; 79/21D
75/25; 75/25D)
83/17; 64/36D
83/17; 41/59D
–
87/13; 76/24D
84/16; 75/25D
86/14; 79/21D
89/11; 89/11D
74/26E
Tetiaroa
15
18
10
8
61.5/38.5 (ꢀ6.2)
77.5/22.5 (ꢀ5.2)
76.5/23.5 (ꢀ2.1)
100/0C
17
27
37
47
57
Tikehau
Rangiroa
Toau
Fakarava
16
1
87.5/12.5 (ꢀ10.8)
–
89/11; 70/30
Makemo
85/15
100/0C
Hereheretue
Anuanuraro
Tematangi
Gambier
5
ADetermined by 1H NMR spectroscopy (CDCl3, 300 MHz) analysis of a
crude ethylic ether extract of lyophilized sponges and integration of C-11
methylene singlets.
4
100/0C
19
7
43.5/56.5 (ꢀ6)
10.5/89.5 (ꢀ1.7)
100/0C
B14858.209 S; 147837.375 W.
Ngele Levu (Fiji)
Vanua Levu (Fiji)
1
C14856.054 S; 147842.492 W.
1
100/0
DTwo samples were collected.
EOne sample was collected.
ADetermined by 1H NMR spectroscopy (CDCl3, 300 MHz) analysis of a
crude ethylic ether extract of frozen dried sponges and integration of C-11
methylene singlets.
most sensitive cell lines. Cleavage of pro-caspases to active
caspases is one of the hallmarks of apoptosis. MES-SA and
MES-SA/Dx5 cells were incubated with 50 and 5 mM of both 4
and 5 (concentrations ranging between ½IC50 to 2IC50 of the
two compounds) for 24 h, and the activities of caspase-3 and
caspase-7 were evaluated using the standard caspase assays. The
results depicted in Fig. 3 show a significant dose-dependent
increase in proteolytic activity of caspases in both MES-SA and
MES-SA/Dx5 cell lines.
BMean values of ratios and standard deviations are provided for information
for each collecting spot.
CNMR spectra were recorded on ,50–100 mg/0.5 mL CDCl3; 100/0 ratios
were obtained when 1H-signals of 5 could not be detected from the baseline.
were both found inactive against PC3 and HeLa cells (half-
maximal inhibitory concentration (IC50) . 50 mM). To verify a
presumed correlation between the cytotoxicity of 4 and 5 and
their ability to induce apoptosis, the activity of caspase-3 and
caspase-7 was measured by a specific apoptosis assay on the
These data suggested that 4 and 5, in addition to their
cytotoxic properties, also induced apoptosis in the above-
mentioned cell lines. Again, ilimaquinone was found to be more

D
A. Boufridi et al.
ent-terpene cyclase
• typical biosynthetic precursors:
terpene cyclase
active site
active site
Ar
RЈ
Ar
Me
Me
Me
Me
skimate-type
precursor of
quinone
Me
Me
R
Me
Me
Me
ϩ H
Me
H ϩ
Me
(Ar ϭ quinone)
farnesylation
Me
8
ent-folding of 8
(aromatic prenyltransferase)
• biosynthetic hypothesis to ilimaquinones:
see ref. [17]
terpene cyclase
active site
Ar
Ar
Ar
Ar
Me
Me
Me
Me
Me
Me
ϩ
Me
cationic cyclization
ϩ
ϩ
Me
Me
Me
Me
Me
Me
ϩ
H
Me
ent-9
Me
H
9
Me
H
H
H
H
H
8
Wagner–Meerwein
rearrangements
9
(Ar ϭ quinone)
further Wagner–Meerwein
no more Wagner–Meerwein shifts
Ar
Me
shifts towards 4 and 5
Me
1
example:
Me
[see Scheme 1]
O
MeO
O
O
ϩ
MeO
Me
10
2
O
OH
Me
OH
Me
H
H
Me
1
2
pathway
Me
pathway
5
Me
Me
Me
Ar
Me
Me
Me
Me
ϩ
ϩ
11
12
ilimaquinone (4)
(ϩ)-hyatellaquinone (13)
H
Ar
H
Me
Scheme 2. Antipodal cyclization processes observed in Dactylospongia
Me
sp.: ilimaquinone versus hyatellaquinone series.
Me
Me
Me
Ar
Me
H
Ar
O
Me
2
Me
Me
Ar
Me
O
MeO
O
MeO
O
H
ϩ
1
H
OH
Me
OH
Me
H
H
1
2
pathway
pathway
O
Me
Me
5
5
MeO
O
O
MeO
Me
Me
OH
Me
O
OH
Me
ilimaquinone (4)
5-epi-ilimaquinone (5)
H
H
Me
Me
Scheme 1. Proposed biosynthetic pathways of ilimaquinone (4) and 5-epi-
ilimaquinone (5).
Me
Me
potent in inducing apoptosis than its epimer. Finally, it is
interesting to point that the potency of ilimaquinone and its
epimer to induce apoptosis at 50 mM was more important for the
multidrug-resistant cell line MES-SA/Dx5.
ilimaquinone (4)
isospongiaquinone(7)
Scheme 3. Pathways for the formation of ilimaquinone (4) and isospon-
giaquinone (7) in D. elegans.
Conclusions
A striking random fluctuation in the ratio between ilimaquinone
and its C-5 epimer has been studied in one of the largest col-
lections of samples (both in terms of number of specimens and
the geographical scope of the study) for a single species of
marine sponge. Slight differences in the biological properties of
both compounds have been revealed. Biosynthetic hypotheses
have been put forward and merit further investigations. From
one region to another, the 4/5 diastereomeric ratio varies sig-
nificantly in an unpredictable manner from 100/0 to 10/90. This
variation may originate from either the lack of selectivity of a
terpene cyclase or the presence of two different iso-enzymes.
The genomic study confirmed the identity of a single species,
and this empirical observation indicates the existence of dif-
ferent molecular phenotypes.

Ilimaquinone and 5-epi-Ilimaquinone
E
MeO
O
O
O
MeO
O
Experimental
General Experimental Procedures
OH
OH
Infrared spectra were recorded on a Vector 22 Bruker spec-
trometer. NMR spectra were recorded on a Bruker AM-300
(300 MHz) and a Bruker AM-400 (400 MHz) instruments using
[D4]methanol, [D]chloroform and [D6]DMSO as solvents.
The solvent signals were used as references. Multiplicities are
described by the following abbreviations: s ¼ singlet, d ¼
doublet, t ¼ triplet, q ¼ quartet, m ¼ multiplet, b ¼ broad. High-
resolution–electrospray ionization–mass spectrometry and
liquid chromatography–mass spectrometry were run on a
Thermoquest TLM LCQ Deca ion-trap spectrometer equipped
with a sunfire analytical C18 column (150 mm ꢁ 2.1 mm;
3.5 mm; Waters) and a UV-visible diode array detector (190–
600 nm; Waters 2996). Analytical thin layer chromatography
(TLC) was performed using Merck silica gel F254 (230–400
mesh) plates, and analysis was conducted either using UV light
or by staining upon heating with vanillin solution (2 g vanillin,
1 mL concentrated H2SO4, 100 mL ethanol). For silica gel
chromatography, the flash chromatography technique was used,
involving Merck silica gel 60 (230–400 mesh) and analytical
silver nitrate-impregnated silica gel TLC was obtained by
dipping p. a. grade TLC plates in a 2 wt-% solution of AgNO3 in
acetonitrile and drying at 808C. Silver nitrate-impregnated silica
gel (2 wt-%) was obtained by suspension of silica gel (100 g) in a
mixture of a 2 wt-% solution of AgNO3 in acetonitrile (100 mL)
and dichloromethane (300 mL), followed by further drying at
408C under reduced pressure. All the chemicals and solvents
were purchased from Aldrich and SDS (France) and required no
further purification.
Me
Me
H
H
Me
Me
H
Me
Me
1
4
2
5
ϩ
H
Me
H
2
1
1
2
O
MeO
O
OH
Me
Me
O
O
MeO
O
MeO
O
OH
OH
Me
Me
Me
Me
H
H
Me
Me
neomamanuthaquinone
(15)
Me
Me
Me
Me
isospongiaquinone (7)
5-epi-isospongiaquinone (14)
Scheme 4. Isomerization under acidic conditions. Reagents and condi-
tions: MeOH/acetic acid/HCl, room temperature, 15 min (7: 48 % yield, 14:
53 % yield, 15: 32 % yield, 35 % yield, respectively from 4 and 5).
Collection of Samples
Table 3. IC50 (lM) of compounds 4 and 5 in the presence of human
cancer and normal cell lines
The sponges were collected during SCUBA diving on field
trips led aboard the research vessel Alis in 2007 (Fiji,
BSMF),[20] 2009 (Marquesas, BSMPF-1),[21] 2011 (Tuamotu,
Tuam’2011),[22] and in Gambier archipelago under the project
‘Coralspot’ in 2010. Sampling at specific depths was carried
out in 2013 at Rangiroa (Tuamotu) using local facilities. See
Table S2 in the Supplementary Material.
Values are mean ꢀ standard errors of three independent experiments
HeLa
MES-SA
MES-SA/Dx5
PC3
4
5
.50
.50
2.44 ꢀ 0.36
2.56 ꢀ 0.70
10.48 ꢀ 1.45
12.54 ꢀ 1.45
.50
.50
Marine Sponge Specimens
Taxonomic identifications were performed by the team of John
N. A. Hooper at the Queensland Museum (Brisbane, Australia),
where voucher specimens of D. metachromia de Laubenfels,
1954 (class Demospongiae, order Dictyoceratida, family
Thorectidae) are deposited under the accessing numbers
G333275 (Gambier, French Polynesia), G333295 and G333722
(Tuamotu, French Polynesia), and G324634 (Vanua Levu, Fiji).
The voucher specimen of D. elegans Thiele, 1899 (class
Demospongiae, order Dictyoceratida, family Thorectidae) is
deposited under the accessing number G331081 (Marquesas,
French Polynesia).
900
MES-SA
MES-SA/Dx5
800
700
600
500
400
300
200
100
0
Genetic Analysis of Dactylospongia Species
Genetic differences of the Polynesian D. metachromia were
tested by comparing sequence signatures of the internal tran-
scribed spacers (ITSs) 1 and 2 of the nuclear ribosomal cistron.
The ITS of multiple samples per collection site, following the
biochemical analyses, was sequenced: Hereheretue, Anuanur-
aro, Toau, Tikehau (1 site with 6 samples each), Fakarava
(2 sites with 6 samples each), Rangiroa (3 sites with 6 samples
0.1 %
DMSO
4 at 50
µM
5 at 50
µM
4 at 5 µM 5 at 5 µM
Concentration
Fig. 3. Apoptotic effects of 4 and 5 on MES-SA and MES-SA/Dx5 cells.
Results are expressed as the percentage of apoptotic cells detected following
24 h of treatment with 4 and 5 at two concentrations.

F
A. Boufridi et al.
each), Tetiaroa (2 sites with 2 samples each), and one sample of
Makemo. DNA was extracted following standard methods,[23]
and amplification of ITS fragments was performed with primers
ITS-RA2-fwd and ITS2.2-rvse,[24] or ITS1-fwd and ITS4-
rev.[25] Forward and reverse strands were sequenced with the
ABI BigDye Terminator 3.1 chemistry following the manu-
facturers protocol on an ABI3730 capillary sequenced in the
Genomic Sequencing Unit of the LMU. Sequencing processing
and assembly was facilitated with CodonCodeAligner (www.
codoncode.com) before analysis with MacClade version
4.08[26] and Geneious version 8.1.[27] Sequences are deposited
in NCBI GenBank and the Sponge Barcoding Database (www.
spongebarcoding.org). Sites suspected of displaying intrage-
nomic polymorphisms were coded with IUPAC codes and
removed from the analysis. The resulting alignment consisted of
807 characters spanning over partial 18S and 28S and complete
ITS1, 5.8S, and ITS2. All samples from all locations assessed
has identical ITS signatures with the exception of two specimens
from Hereheretue (1.7 % difference) and the one Makemo
sample (1.5 % difference). Due to the fact that ITS is a highly
variable marker and the genetic differences can be regarded as
low when compared with other dictyoceratid taxa,[28] there is no
basis for interpretation as different species.
dried over Na2SO4 and concentrated under reduce pressure. The
obtained residue was purified by flash chromatography over
silver nitrate-impregnated silica gel (2 % w/w, elution gradient:
petroleum ether/dichloromethane 1 : 1 - 0 : 1), affording 7
(48 mg, 48 %) and then 15 (32 mg, 32 %).
Acidic Isomerization of 5
To a solution of 5 (200 mg, 0.56 mmol) in MeOH (20 mL) was
added a mixture of a solution of aqueous 35 % HCl (10 mL) and
acetic acid (10 mL). After 15 min of stirring, the reaction mix-
ture was quenched by addition of water (20 mL) and extraction
with CH2Cl2 (3 ꢁ 20 mL). The combined organic layers were
dried over Na2SO4 and concentrated under reduce pressure. The
obtained residue was purified by flash chromatography over
silver nitrate-impregnated silica gel (2 wt-%) (elution gradient:
petroleum ether/dichloromethane 1 : 1 - 0 : 1), affording 14
(106 mg, 53 %) and then 15 (71 mg, 35 %). Compounds 7, 14,
and 15 displayed spectroscopic data identical to those previ-
ously reported (see Supplementary Material).
Biological Evaluation
HeLa cell line was originally purchased from the American
Type Culture Collection (ATCC) and maintained in Roswell
Park Memorial Institute (RPMI) medium supplemented with
10 % fetal bovine serum and 1 % penicillin/streptomycin. MES-
SA cells and MES-SADX5, from ATCC, were grown in
McCoy’s 5A (ATCC) with 10 % fetal bovine serum and 1 %
penicillin/streptomycin. PC-3 cell line was kindly given by O.
Filhol (BCI laboratory, iRTSV, Grenoble, France) and main-
tained in RPMI medium supplemented with 10 % fetal bovine
serum and 1 % penicillin/streptomycin. The cell viability assay
was performed in 96-wells microplates. Depending of the
cell type and the rate of cell growth, cells were seeded at 8000
cells per well (HeLa, PC-3) or 12 000 cells per well (MES-SA,
MES-SA DX5) in 90 mL culture medium and allowed to grow
for 24 h. Compounds at different concentrations were then
added to the cell-containing wells. Cells were allowed to grow
for additional 24 h. Cell viability was evaluated using the
CellTiter-Gloꢁ luminescent cell viability assay (Promega
France) according to supplier’s protocol. The results were
expressed as the percentage of viable cells relative to parallel,
control cell seeding (0 %, 25 %, 50 %, 75 %, and 100 %). Apo-
ptosis induction assay was performed in 96-wells microplates.
Depending on the cell type and the rate of cell growth, cells were
seeded at 8000 cells per well (HeLa, PC3) or 12 000 cells per
well (MES-SA, MES-SA DX5) in 90 mL culture medium and
allowed to grow for 24 h. Compounds at different concentrations
or extracts at different dilutions were then added to the cell-
containing wells. Cells were allowed to grow for additional 24 h.
Apoptosis induction was evaluated using the Caspase-Gloꢁ 3/7
assay (Promega France) according to supplier’s protocol. The
results were expressed as fold of apoptosis induction relative to
basal apoptosis induction measured on untreated, control cells.
Extraction and Purification of 4 and 5
Freeze-dried sponge material (300 g) was extracted with a
mixture of CH2Cl2 and MeOH (1 : 1, v/v, 1 L; 3 ꢁ 1 L) using an
ultrasound cleaner bath for 1 h. Combined extracts were con-
centrated under reduced pressure yielding a dark purple crude
extract (,100 g). The crude extract was desalted by partition
between ethyl acetate (1 L) and water (3 ꢁ 250 mL). The organic
layer was dried over Na2SO4 and concentrated under reduced
pressure yielding an organic extract (30 g). The organic extract
was subjected to purification by flash chromatography over
silica gel (elution gradient: petroleum ether/ethyl acetate/
methanol 1 : 0 : 0 - 0 : 1 : 0 - 0 : 9 : 1), giving four fractions
named F1, F2, F3, and F4. F3 (11 g, 3.6 %), which was eluted
with a gradient of petroleum ether/ethyl acetate 7 : 1 - 1 : 1,
corresponds to a mixture of ilimaquinone 4 and 5-epi-
ilimaquinone 5. F3 was resolved by flash chromatography
over silver nitrate-impregnated silica gel (2 wt-%) (elution
gradient:
petroleum
ether/dichloromethane/methanol
2 : 8 : 0 - 0 : 1 : 0 - 0 : 9 : 1), affording 5-epi-ilimaquinone 5
(5 g, 1.6 %) followed by ilimaquinone 4 (3.3 g, 1.1 %). Com-
pounds 4 and 5 displayed spectroscopic data identical to those
previously reported (see Supplementary Material).
Determination of Ilimaquinone/5-epi-Ilimaquinone Ratio
Freeze-dried sponge material (500 mg) was extracted with die-
thyl ether (3 ꢁ 10 mL) using an ultrasound cleaner bath for
15 min. After filtration, the extracts were concentrated under
reduced pressure. The diastereomeric ratio between 4 and 5 was
determined by 1H NMR spectroscopy analysis conducted in [D]
chloroform. Signals of the methylene function of ilimaquinone
(4) (4.40/4.42 ppm) and 5-epi-ilimaquinone (5) (4.66/4.69 ppm)
were used as references.
Supplementary Material
1
Full analytical description, and H, 13C NMR spectra of com-
pounds 4, 5, 7, and 14 and sample collection complementary
data are available on the Journal’s website.
Acidic Isomerization of 4
To a solution of 4 (100 mg, 0.28 mmol) in MeOH (10 mL) was
added a mixture of a solution of aqueous 35 % HCl (10 mL) and
acetic acid (10 mL). After 15 min of stirring, the reaction mix-
ture was quenched by addition of water (15 mL) and extraction
with CH2Cl2 (3 ꢁ 10 mL). The combined organic layers were
Acknowledgements
This project was supported by the ANR 11-EBIM-006–01 (Netbiome
project POMARE). We thank Blandine Se´on-Me´niel for technical assistance

Ilimaquinone and 5-epi-Ilimaquinone
G
and Jean-Christophe Jullian for NMR assistance. We acknowledge IRD for
financing the field trips aboard the research vessel ALIS, the diving team of
IRD-Noume´a, and the R/V crew for their on-field help. We thank the Fiji
Islands Government for allowing us to collect our samples, and the Fisheries
Department and Klaus Feussner (USP-IAS) for their help and assistance.
The French Polynesia research department is also greatly acknowledged for
supporting the sponge survey. DE acknowledges Nicole Enghuber for
assistance in the laboratory, Gert Wo¨rheide for providing access to analytical
facilities, and support from the Deutsche Forschungsgemeinschaft (DFG,
grant no. ER611/3–1).
C. Schoenberg, D. Janussen, K. R. Tabachnick, M. Klautau, B. Picton,
M. Kelly, J. Vacelet, M. Dohrmann, M.-C. D´ıaz, P. Ca´rdenas) 2015.
Available at http://www.marinespecies.org/porifera/porifera.php?
p=taxdetails&id=165313 and http://www.marinespecies.org/pori-
fera/porifera.php?p=taxdetails&id=397152 (accessed 25 June 2015).
[10] Dactylospongia elegans Thiele 1899 (Thorectidae), also formerly
named Luffarella elegans Thiele 1899 (Thorectidae), see: R. van
Soest, in World Porifera Database (Eds R. W. M. Van Soest,
N. Boury-Esnault, J. N. A. Hooper, K. Ru¨tzler, N. J. de Voogd, B.
Alvarez de Glasby, E. Hajdu, A. B. Pisera, R. Manconi, C. Schoenberg,
D. Janussen, K. R. Tabachnick, M. Klautau, B. Picton, M. Kelly,
J. Vacelet, M. Dohrmann, M.-C. D´ıaz, P. Ca´rdenas) 2015. Available at
http://www.marinespecies.org/porifera/porifera.php?p=taxde-
tails&id=165312 (accessed 25 June 2015).
References
[1] See p. 899 in: J.-M. Kornprobst, Encyclopedia of Marine Natural
Products 2014 (Wiley-Blackwell: Weinheim).
[11] Cyanobacterial symbionts were shown to produce chlorinated meta-
bolites, whereas sesquiterpenoids were found only in the sponge cells,
see: M. D. Unson, D. J. Faulkner, Experientia 1993, 49, 349. doi:10.
1007/BF01923420
[12] (a) Avarol (dihydro-1) and congeners were found within sponge
choanocytes, see: M.-J. Uriz, X. Turon, J. Galera, J. M. Tur, Cell
Tissue Res. 1996, 285, 519. doi:10.1007/S004410050668
[2] (a) Selected review articles: M. Gordaliza, Mar. Drugs 2012, 10, 358.
doi:10.3390/MD10020358
(b) M. Gordaliza, Mar. Drugs 2010, 8, 2849. doi:10.3390/MD8122849
(c) R. J. Capon, in Studies in Natural Products Chemistry: Structure
and Chemistry (Part C) (Ed. Atta-ur Rahman) 1995, Vol. 15,
pp. 289–326 (Elsevier: Amsterdam).
[3] For selected examples on antitumoral activities, see: (a) H.-Y. Lee, K.
J. Chung, I. H. Hwang, J. Gwak, S. Park, B. G. Ju, E. Yun, D.-E. Kim,
Y.-H. Chung, M. Na, G.-Y. Song, S. Oh, Mar. Drugs 2015, 13, 543.
doi:10.3390/MD13010543
(b) Avarol was produced from cultures of primmorphs from Dysidea
avara, see: W. E. G. Mu¨ller, M. Bo¨hm, R. Batel, S. De Rosa, G.
Tommonaro, I. M. Mu¨ller, H. C. Schro¨der, J. Nat. Prod. 2000, 63,
1077. doi:10.1021/NP000003P
(b) S. Park, E. Yun, I. H. Hwang, S. Yoon, D.-E. Kim, J. S. Kim, M. Na,
G.-Y. Song, S. Oh, Mar. Drugs 2014, 12, 3231. doi:10.3390/
MD12063231
(c) M. T. Do, M. Na, H. G. Kim, T. Khanal, J. H. Choi, S. W. Jin, S. H.
Oh, I. H. Hwang, Y. C. Chung, H. S. Kim, T. C. Jeong, H. G. Jeong,
Food Chem. Toxicol. 2014, 71, 51 and references cited therein.
doi:10.1016/J.FCT.2014.06.001
(c) Very interestingly, the variation of concentrations of avarol
depending on time period and biotope-dependent factors was evaluat-
ed, see: S. De Caralt, D. Bry, N. Bontemps, X. Turon, M.-J. Uriz, B.
Banaigs, Mar. Drugs 2013, 11, 489. doi:10.3390/MD11020489
[13] The gathering and cross-checking of data from literature dealing with
ilimaquinones shows a variation of 4/5 ratio between publications but
strangely this statement is never discussed as such based on multi-
sample studies. Reports of co-isolation of 4 and 5 include D. elegans
from Okinawa, Japan and 4/5 (96/4) ratio based on amounts presented
in the experimental section after purification and not a ratio based on
crude extract analysis, see: (a) H. Mitome, T. Nagasawa, H. Miyaoka,
Y. Yamada, R. W. M. van Soest, J. Nat. Prod. 2001, 64, 1506.
doi:10.1021/NP010299E
For antiviral activities, see: (d) S. Loya, A. Hizi, J. Biol. Chem. 1993,
268, 9323.
(e) S. Loya, R. Tal, Y. Kashman, A. Hizi, Antimicrob. Agents Che-
mother. 1990, 34, 2009. doi:10.1128/AAC.34.10.2009
[4] (a) Ilimaquinone (4) was first isolated from Hippiospongia metachro-
mia (see ref. [9] for taxonomy details). See: R. T. Luibrand, T. R.
Erdman, J. J. Vollmer, P. J. Scheuer, J. Finer, J. Clardy, Tetrahedron
1979, 35, 609. doi:10.1016/0040-4020(79)87004-0
(b) Dactylospongia elegans from West Flores, Indonesia: 4/5 (3/5 ratio
based on amounts presented in experimental section and not a ratio
based on crude extract analysis), see: S. Aoki, D. Kong, K. Matsui,
R. Rachmat, M. Kobayashi, Chem. Pharm. Bull. 2004, 52, 935.
doi:10.1248/CPB.52.935
(b) The absolute configuration was later revised, see: R. J. Capon,
J. K. MacLeod, J. Org. Chem. 1987, 52, 5059. doi:10.1021/
JO00231A051
[5] 5-epi-Ilimaquinone (5) was first isolated from a Fenestraspongia sp.
specimen (which was ‘in an unusually poor condition for identifica-
tion’); see also ref. [4b] for stereochemical revision: B. Carte´, C. B.
Rose, D. J. Faulkner, J. Org. Chem. 1985, 50, 2785. doi:10.1021/
JO00215A039
[6] 5,8-epi-Ilimaquinone (6) was recently isolated from D. elegans from
the coast of Palau, see: L. Du, Y.-D. Zhou, D. G. Nagle, J. Nat. Prod.
2013, 76, 1175. doi:10.1021/NP400320R
(c) Dactylospongia elegans from Truant Island, Australia: 4/5 (1/1
mixture, not separated), see: S. P. B. Ovenden, J. L. Nielson, C. H.
Liptrot, R. H. Willis, D. M. Tapiolas, A. D. Wright, C. A. Motti, J. Nat.
Prod. 2011, 74, 65. doi:10.1021/NP100669P
(d) Dactylospongia elegans from Fiji: only the presence of 5 is
reported, see: J. Rodr´ıguez, E. Quin˜oa, R. Riguera, B. M. Peters, L.
M. Abrell, P. Crews, Tetrahedron 1992, 48, 6667. doi:10.1016/S0040-
4020(01)80012-0
[7] (a) Isospongiaquinone (7) was isolated from a sponge ‘tentatively
classified as Stelospongia conulata’; no stereochemistry was provided
at that time, see: R. Kazlauskas, P. T. Murphy, R. G. Warren, R. J.
Wells, J. F. Blount, Aust. J. Chem. 1978, 31, 2685. doi:10.1071/
CH9782685
(e) Fenestraspongia sp., Thorectidae from Urukthapel Island, Palau:
4/5 (6/4 mixture), see ref. [5].
(f) Polyfibrospongia australis, Thorectidae (now Fasciospongia tur-
gida, Thorectidae) from Taiwan: mixture of 4 and 5 (no further
precision). See: Y.-C. Shen, P.-W. Hsieh, J. Nat. Prod. 1997, 60, 93.
doi:10.1021/NP9605302
(b) The absolute stereochemistry of 7 was later determined by chemi-
cal correlations, see: R. J. Capon, J. Nat. Prod. 1990, 53, 753.
doi:10.1021/NP50069A042
[14] Having the same decalin ring system but with, apparently, opposite
absolute configuration, asmarines A–F (isolated from a specimen of
Raspailia sp.) bear diazepino-purines a in place of the quinone. They
also exists as pairs of epimers at C-5 (asmarines A-B, C-D, E-F
respectively), see: T. Yosief, A. Rudi, Y. Kashman, J. Nat. Prod. 2000,
63, 299. In the latter paper, the term ‘epi’ is misleading and refers to a
C-5 configuration (S), (R), (S) for asmarines B, D, F, respectively.
No indication of the different ratios is given.
[8] (a) P. A. Takizawa, J. K. Yucel, B. Veit, D. J. Faulkner, T. Deerinck,
G. Soto, M. Ellisman, V. Malhotra, Cell 1993, 73, 1079. doi:10.1016/
0092-8674(93)90638-7
(b) B. Veit, J. K. Yucel, V. Malhotra, J. Cell Biol. 1993, 122, 1197.
doi:10.1083/JCB.122.6.1197
[9] Dactylospongia metachromia de Laubenfels 1954 (Thorectidae), also
formerly named Hippospongia metachromia de Laubenfels 1954
(Spongidae), see: R. van Soest, in World Porifera Database (Eds R.
W. M. Van Soest, N. Boury-Esnault, J. N. A Hooper, K. Ru¨tzler, N. J.
de Voogd, B. Alvarez de Glasby, E. Hajdu, A. B. Pisera, R. Manconi,
[15] The case of 5,8-di-epi-ilimaquinone (6) described recently by Nagle
and coworkers is striking in terms of plausible biosynthetic pathway
but is not discussed by the authors (see ref. [6]). It would apparently
necessitate an unusual precursor i.e. Z,E-(8).

H
A. Boufridi et al.
[16] (a) The lack of selectivity of terpene cyclases in terms of end products
is now well documented. See for example: M. Ko¨ksal, Y. Jin, R. M.
Coates, R. Croteau, D. W. Christianson, Nature 2011, 469, 116.
doi:10.1038/NATURE09628
[21] C. Debitus, French Oceanographic Cruises – BSMPF-1 2009. Avail-
able at: doi:10.17600/9100030
[22] C. Debitus, French Oceanographic Cruises – TUAM’2011 2011.
Available at: doi:10.17600/11100010
(b) See also, among others: V. Gonzalez, S. Touchet, D. J. Grundy,
J. A. Faraldos, R. K. Allemann, J. Am. Chem. Soc. 2014, 136, 14505.
doi:10.1021/JA5066366
[23] E. Sambrook, F. Fritsch, T. Maniatis, Molecular Cloning 1989 (Cold
Spring Harbor Press: New York, NY).
[24] G. Wo¨rheide, Facies 1998, 38, 1. doi:10.1007/BF02537358
[25] C. Chombard, N. Boury-Esnault, S. Tillier, Syst. Biol. 1998, 47, 351.
doi:10.1080/106351598260761
[26] W. P. Maddison, D. R. Maddison, MacClade 3: Analysis of
Phylogeny and Character Evolution 1992 (Sinauer Associates:
Sunderland, MA).
[17] K. W. L. Yong, A. Jankam, J. N. A. Hooper, A. Suksamrarn, M. J.
Garson, Tetrahedron 2008, 64, 6341. doi:10.1016/J.TET.2008.04.091
[18] (a) Neomamanuthaquinone 15 was isolated from a Dactylospongia sp.
as a natural substance (see ref. [17]) but was described before as an
semi-synthetic compound. See: (a) ref. [5],
(b) J. C. Swersey, L. R. Barrows, C. M. Ireland, Tetrahedron Lett.
[27] M. Kearse, R. Moir, A. Wilson, S. Stones-Havas, M. Cheung,
S. Sturrock, S. Buxton, A. Cooper, S. Markowitz, C. Duran, T. Thierer,
B. Ashton, P. Meintjes, A. Drummond, Bioinformatics 2012, 28, 1647.
doi:10.1093/BIOINFORMATICS/BTS199
1991, 32, 6687. doi:10.1016/S0040-4039(00)93575-5
(c) N. K. Utkina, V. A. Denisenko, O. V. Scholokova, A. E.
Makarchenko, J. Nat. Prod. 2003, 66, 1263. doi:10.1021/NP030115R
[19] S. Urban, R. J. Capon, J. Nat. Prod. 1992, 55, 1638. doi:10.1021/
NP50089A012
[28] J. Po¨ppe, P. Sutcliffe, J. N. A. Hooper, G. Wo¨rheide, D. Erpenbeck,
PLoS One 2010, 5, e9950. doi:10.1371/JOURNAL.PONE.0009950
[20] C. Payri, French Oceanographic Cruises – BSM-FIDJI 2007. Avail-
able at: doi:10.17600/7100030