Blurring the boundary between bio- and geohopanoids: Plakohopanoid, a C32 biohopanoid ester from Plakortis cf. lita

Article abstract of DOI:10.1002/ejoc.201200676

Plakohopanoid (3a), a new type of hopanoid derivative composed of a C ₃₂ hopanoid acid ester linked to a mannosyl-myo-inositol, was isolated from the sponge Plakortis cf. lita as its peracetyl derivative 3b. The structure of 3b was determined by a combination of spectroscopic analysis and micro-scale chemical degradation. Even though plakohopanoid was isolated from a sponge, its component parts are clearly of bacterial origin, and its bacterial biosynthesis is very likely. Until now, C₃₂ hopanoic acids have been considered to be geohopanoids, i.e., diagenetic products that are formed through abiotic degradation of the biohopanoids present in bacteria. The presence of 3a in a marine living organism shows that there is a biosynthetic pathway to C₃₂ hopanoic acids, and these substances should therefore no longer with certainty be considered to be geohopanoids. Copyright

Full text of DOI:10.1002/ejoc.201200676

Job/Unit: O20676
/KAP1
Date: 13-08-12 16:55:13
Pages: 7
FULL PAPER
DOI: 10.1002/ejoc.201200676
Blurring the Boundary between Bio- and Geohopanoids: Plakohopanoid, a C32
Biohopanoid Ester from Plakortis cf. lita
Valeria Costantino,[a] Gerardo Della Sala,[a] Alfonso Mangoni,*[a] Cristina Perinu,[a][‡] and
Roberta Teta[a]
Dedicated to the memory of Ernesto Fattorusso
Keywords: Natural products / Biosynthesis / Terpenoids / Glycolipids / Hopanoids
Plakohopanoid (3a), a new type of hopanoid derivative com- Until now, C32 hopanoic acids have been considered to be
posed of a C32 hopanoid acid ester linked to a mannosyl-
myo-inositol, was isolated from the sponge Plakortis cf. lita
geohopanoids, i.e., diagenetic products that are formed
through abiotic degradation of the biohopanoids present in
as its peracetyl derivative 3b. The structure of 3b was deter- bacteria. The presence of 3a in a marine living organism
mined by a combination of spectroscopic analysis and micro-
scale chemical degradation. Even though plakohopanoid
was isolated from a sponge, its component parts are clearly
of bacterial origin, and its bacterial biosynthesis is very likely.
shows that there is a biosynthetic pathway to C32 hopanoic
acids, and these substances should therefore no longer with
certainty be considered to be geohopanoids.
lated in our laboratory from the Caribbean sponge
Plakortis simplex,[2] and was later shown to be present in
the cells of the bacterial symbionts of this sponge,[3] and
should therefore be considered to be a biohopanoid
(Scheme 1). Very recently, however, Schaeffer et al. demon-
strated that anhydro-BHT can be formed from BHT and
other hopanoids by geochemical processes,[4] so that the
anhydro-BHT that is present (sometimes predominantly) in
sediments may be of abiotic origin.
In this paper, we describe the isolation and the structural
elucidation of a new biohopanoid, plakohopanoid (3a)
from the Indonesian sponge Plakortis cf. lita. Plakoho-
panoid is an ester composed of a C32 hopanoic acid, a well-
known hopanoid that has, until now, been considered to be
a typical example of a geohopanoid, and mannosyl-myo-
inositol, a building block of phosphatidylinositol mannos-
ides, which are characteristic of Mycobacteria and related
species.
Introduction
Hopanoids (e.g., bacteriohopanetetrol, or BHT; 1a) are
bacterial metabolites composed of a pentacyclic C30 hopane
triterpene linked to a sugar-derived polyfunctionalized C5
chain (Scheme 1). They play a similar role in bacterial mem-
branes to that played by sterols in eukaryotes. The pentacy-
clic ring moiety of hopanoids is very stable and not readily
degraded, but after the death of the bacterial cell, the func-
tionalized side-chain can undergo various abiotic degrad-
ative processes, including reduction (as far as alkanes), loss
of carbon atoms, and epimerization. Degraded hopanoids
are ubiquitous in sediments, rocks, and crude oil,[1] and,
because the characteristic pentacyclic moiety is preserved,
they can easily be recognized and may be used as biomark-
ers in geochemical studies. Consequently, hopanoids are
generally classified into two classes: biohopanoids, pro-
duced in bacteria by biosynthetic processes, and geoho-
panoids, derived from abiotic degradation of biohopanoids.
There are a few hopanoids that escape this classification.
Among them is 32,35-anhydro-BHT (2), which was first iso-
Results and Discussion
[a] The NeaNat Group, Dipartimento di Chimica delle Sostanze
Naturali, Università di Napoli Federico II,
via D. Montesano 49, 80131 Napoli, Italy
Fax: +39-081-678-552
The MeOH extract of an Indonesian specimen of
Plakortis cf. lita was partitioned between BuOH and H2O,
and the organic phase was purified by reverse-phase
chromatography followed by normal-phase chromatog-
raphy to give a crude fraction mainly composed of glycolip-
ids, but also containing hopanoids. This fraction was acety-
lated with Ac2O in pyridine. The composition of the per-
acetylated glycolipid fraction was analyzed using normal-
E-mail: alfonso.mangoni@unina.it
Homepage: http://www.neanat.unina.it/
index.php?item=20&type=ricercatori
[‡] Current address: Telemark University College,
Kjølnes Ring 56, 3901 Porsgrunn, Norway
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201200676.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1

Job/Unit: O20676
/KAP1
Date: 13-08-12 16:55:13
Pages: 7
V. Costantino, G. D. Sala, A. Mangoni, C. Perinu, R. Teta
FULL PAPER
phase HPLC, and it turned out to be very similar to the doublet at δ = 0.92 ppm) and an acetylated carbohydrate
glycolipid fraction of P. simplex from the Caribbean Sea, moiety (13 carbinol protons between δ = 5.55 and
1
which has been extensively studied by our group. H NMR 4.05 ppm, and 8 acetyl singlets between δ = 2.14 and
spectroscopic analysis of the collected HPLC fractions 2.00 ppm).
showed the presence of the unusual metabolites previously
Examination of the correlation peaks of the methyl pro-
reported from P. simplex, i.e., plakoside A and B,[5] sim- tons in the HMBC spectrum allowed us to assign most of
plexides,[6] 12-methylbacteriohopanetetrol,[7] and crasser- the carbon atoms in the terpene skeleton of 3b and to
ides.[8] In addition, the HPLC separation provided the new sketch the partial structure depicted in Figure 1. This fitted
hopanoid derivative, plakohopanoid (3a) as its peracetyl de- well with a hopane skeleton, which, according to the 13C
rivative 3b (Scheme 1). Compound 3b could not be deacety- chemical shifts, should be devoid of any functional groups.
lated to give the natural hopanoid (i.e., 3a), because 3b is Indeed, a comparison of the 13C spectra of 3b and bac-
itself an ester. Nor were we able to isolate pure 3a from the teriohopanetetrol tetraacetate (1b) showed that for all the
crude glycolipid fraction, because of the well-known low signals of C-1–C-30 in the 13C spectrum of bacteriohopane-
solubility of biohopanoids in most organic solvents.[9] How- tetrol tetraacetate, there was a corresponding signal within
ever, the presence of natural 3a in the glycolipid fraction 0.2 ppm in the spectrum of 3b. This implied that the terpe-
could be demonstrated by an LC–HRMS experiment (see noid skeletons of 1b and 3b had the same structure and
below).
relative configuration, i.e., that of an extended hopane. The
configuration at C-22, which does not significantly affect
13C chemical shifts,[10] was determined on the basis of the
1H chemical shift of CH3-29 (δ = 0.92 ppm), which is diag-
nostic for the 22R configuration.[10]
Figure 1. Partial structure of plakohopanoid peracetate 3b as deter-
mined by long-range proton–carbon couplings of methyl protons
(bold lines).
The chemical shifts of the protons in the hopane part of
the molecule were then identified using an HSQC (hetero-
nuclear single-quantum correlation) experiment (Table 1).
All the correlation peaks that were present in the COSY,
TOCSY, and HMBC spectra were consistent with a hopane
skeleton and with the assignment reported in Table 1. The
structure of the remaining C2 fragment of the side-chain
that completed the structure of the hopanoid skeleton was
deduced from the correlations of the methylene protons at
C-31 (δ = 2.37 and 2.25 ppm) with the methylene protons
at C-30 in the COSY spectrum, and with the ester carbonyl
carbon atom at δ = 173.9 ppm in the HMBC spectrum.
For structural elucidation of the carbohydrate moiety of
3b, a second set of one- and two-dimensional NMR experi-
ments was recorded using C6D6 as the solvent, because a
better signal dispersion was observed in the mid-field region
of the 1H NMR spectrum using this solvent. The 13C NMR
spectrum showed the presence of only one anomeric carbon
(δ = 100.0 ppm), which was correlated in the HSQC spec-
Scheme 1.
1
trum with the H doublet at δ = 4.98 ppm (J = 2.0 Hz).
The ESI HRMS spectrum of compound 3b showed a Using this resonance as the starting point, analysis of
pseudomolecular [M + Na]+ ion peak at m/z = 1153.5885, COSY correlation peaks allowed us to identify a sequence
indicating the molecular formula C60H90O20. Analysis of of four methine protons and a pair of methylene protons at
the 1H NMR spectrum of 3b (CDCl3) showed the molecule δ = 5.65 (2Ј-H), 5.77 (3Ј-H), 5.85 (4Ј-H), 4.52 (5Ј-H), and
to be composed of a terpenoid moiety (6 methyl singlets at 4.57/4.44 ppm (6Ј-H2), accounting for a hexose. That this
δ = 0.94, 0.93, 0.84, 0.80, 0.78, and 0.69 ppm, and a methyl was in the pyranose form was deduced from the relatively
2
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20676
/KAP1
Date: 13-08-12 16:55:13
Pages: 7
Plakohopanoid, a C32 Biohopanoid Ester
Table 1. NMR spectroscopic data for compound 3b.
700 MHz, CDCl3
175 MHz, CDCl3
δC, type
500 MHz, C6D6
125 MHz, C6D6
δC, type
Pos.
δH [ppm] (J [Hz])[a]
δH [ppm] (J [Hz])[a]
1
2
3
4
1.64, 0.76
1.56, 1.34
1.34, 1.12
–
40.1, CH2
18.7, CH2
42.1, CH2
33.4, C
1.65, 0.77
1.60, 1.38
1.38, 1.16
–
40.6, CH2
19.1, CH2
42.4, CH2
33.4, C
5
6
7
8
0.70
1.48, 1.31
1.45, 1.21
–
56.1, CH
18.6, CH2
33.2, CH2
41.8, C
0.75
56.5, CH
19.1, CH2
33.7, CH2
42.0, C
1.53, 1.34
1.49, 1.23
–
9
1.24
–
50.4, CH
37.4, C
1.28
–
50.8, CH
37.6, C
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
1.52, 1.29
1.44, 1.39
1.31
20.9, CH2
23.9, CH2
49.2, CH
41.6, C
33.6, CH2
22.8, CH2
54.4, CH
44.3, C
41.5, CH2
27.5, CH2
45.8, CH
36.2, CH
33.3, CH3
21.6, CH3
15.9, CH3
16.5, CH3
16.6, CH3
15.8, CH3
19.7, CH3
30.5, CH2
30.9, CH2
1.53, 1.27
1.42, 1.42
1.30
21.3, CH2
24.3, CH2
49.6, CH
41.9, C
34.0, CH2
23.1, CH2
54.6, CH
44.5, C
41.8, CH2
27.8, CH2
46.3, CH
36.6, CH
33.6, CH3
21.8, CH3
16.2, CH3
16.9, CH3
16.7, CH3
16.0, CH3
19.9, CH3
31.1, CH2
31.3, CH2
–
–
1.34, 1.23
1.71, 1.52
1.28
1.31, 1.22
1.66, 1.50
1.18
–
–
1.52, 0.90
1.81, 1.54
1.73
1.51
0.84 (s)
0.78 (s)
0.80 (s)
0.93 (s)
0.94 (s)
1.54, 0.88
1.81, 1.64
1.68
1.56
0.912 (s)
0.863 (s)
0.857 (s)
0.976 (s)
0.960 (s)
0.69 (s)
0.720 (s)
0.92 (d, 7.6)
1.81, 1.29
2.37 (ddd, 15.9, 10.6, 5.0)
2.25 (ddd, 15.9, 10.0, 6.0)
–
4.95 (d, 1.7)
5.36 (dd, 3.1, 1.7)
5.39
0.954 (d, 6.5)
2.04, 1.43
2.45 (ddd, 15.9, 9.9, 5.5)
2.41 (ddd, 15.9, 9.1, 6.8)
–
a
b
32
1Ј
2Ј
3Ј
4Ј
5Ј
6Ј
173.9, C
99.4, CH
69.6, CH
65.5, CH
68.6, CH
69.4, CH
61.7, CH2
173.5, C
4.98 (d, 2.0)
5.65 (dd, 3.2, 2.0)
5.77 (dd, 10.0, 3.2)
5.85 (t, 10.0)
4.52 (ddd, 10.0, 3.9, 2.0)
4.57 (dd, 12.3, 3.9)
4.44 (dd, 12.3, 2.0)
4.88(dd, 10.6, 2.4)
4.20 (t, 2.4)
5.01 (dd, 10.4, 2.4)
5.902(dd, 10.4, 9.7)
5.33 (t, 9.7)
5.897 (dd, 10.4, 9.7)
100.0, CH
70.4, CH
69.3, CH
66.0, CH
70.5, CH
62.0, CH2
5.40
4.18 (m)
a
b
4.26 (dd, 12.5, 4.3)
4.09 (dd, 12.5, 2.2)
5.00 (dd, 10.5, 2.5)
4.29 (t, 2.5)
5.08 (dd, 10.5, 2.5)
5.53 (dd, 10.5, 9.9)
5.19 (t, 9.9)
5.49 (dd, 10.5, 9.9)
1ЈЈ
2ЈЈ
3ЈЈ
4ЈЈ
5ЈЈ
6ЈЈ
Ac
69.6, CH
76.4, CH
70.3, CH
69.5, CH
70.6, CH
69.3, CH
70.3, CH
77.3, CH
70.7, CH
70.6, CH
71.3, CH
70.0, CH
2.14, 2.09, 2.07, 2.06, 2.04, 2.03, 2.01, 20.9–20.5, CH3
2.00 (8 singlets)
169.9–169.4, C
1.935, 1.880, 1.672, 1.672, 1.665, 20.4–20.1, CH3
1.625, 1.599, 1.599 (8 singlets)
169.6–169.1, C
[a] Multiplicity not shown for overlapping signals; chemical shifts for these signals were determined by 2D NMR spectroscopy.
shielded chemical shift of 5Ј-H and the coupling between around 158–162 Hz for axial anomeric protons, and around
5Ј-H and C-1Ј in the HMBC spectrum. The large couplings 169–171 Hz for equatorial anomeric protons. The remain-
(10.0 Hz) of 4Ј-H with both 3Ј-H and 5Ј-H were indicative ing six oxymethine protons observed in the mid-field region
of the axial orientation of these three protons, while the of the proton NMR spectrum were shown from COSY data
small coupling (3.3 Hz) of 3Ј-H with 2Ј-H showed that the to be cyclically arranged, and therefore were indicative of
latter proton was equatorial. Therefore, the hexopyranose an inositol unit. The myo configuration of the inositol was
unit was a mannopyranose unit. The α anomeric configura- deduced by coupling constant analysis (Table 1), which
tion of the mannose unit was deduced from the 173 Hz cou- showed all the inositol protons to be axial except for 2ЈЈ-H
pling constant between 1Ј-H and C-1Ј (which we could (δ = 4.20 ppm). The shielded chemical shift of 2ЈЈ-H and
measure from the residual 1Ј-H,C-1Ј one-bond correlation the HMBC correlation between 2ЈЈ-H and C-1Ј showed that
observed in the HMBC spectrum), because it is known the myo-inositol was glycosylated at O-2ЈЈ by the α-manno-
1
from the literature[11] that JC,H coupling constants are pyranose unit. Finally, the linkage between the terpene and
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3

Job/Unit: O20676
/KAP1
Date: 13-08-12 16:55:13
Pages: 7
V. Costantino, G. D. Sala, A. Mangoni, C. Perinu, R. Teta
FULL PAPER
carbohydrate moieties of plakohopanoid was identified as (i.e., 3c) was isolated, and it showed a proton NMR spec-
an ester bond between the carbonyl C-32 and the mannose trum identical to that of 3b, except for the absence of the
6Ј-oxymethylene group, on the basis of the correlation eight acetyl singlets.
peaks in the HMBC spectrum between both C-6Ј protons
A further confirmation of the presence of natural plako-
and C-32.
hopanoid 3a in P. cf. lita came from high-resolution HPLC–
The carbohydrate moiety of plakohopanoid peracetate ESI-MS analysis of the crude glycolipid mixture (before
(3b) is very similar to that of discoside peracetate (4b), pre-
viously isolated by our group from Discodermia disso-
luta,[12] but also present in Caribbean specimens of
Plakortis (unpublished data) as well as in the Indonesian
specimen studied in this paper. In 4a, there are two fatty
acid residues ester-linked to the mannosylinositol unit at O-
4Ј and O-6Ј. In plakohopanoid (3a) they are replaced by a
single C32 hopanoyl unit linked at O-6Ј. A comparison be-
tween the proton NMR spectra of 3b and 4b revealed that
all the carbohydrate signals show almost identical chemical
shifts and coupling constants in the two compounds, fur-
ther confirming structure 3b. The absolute configuration of
plakohopanoid was established by micro-scale chemical de-
gradation (Scheme 2). Acidic methanolysis of 3b with 1 m
HCl in 92% MeOH produced a mixture of hopanoic acid
methyl ester (5) and methyl mannopyranoside (6). After
partitioning between CHCl3 and H2O, these two com-
pounds were found, respectively, in the CHCl3 and H2O
layers.
Scheme 2. Degradation procedure of plakohopanoid peracetate
(3b).
Hopanoic ester 5 was purified by HPLC (SiO2, n-hexane/
1
EtOAc, 95:5), and its H NMR spectrum matched that re-
ported in the literature.[13] In addition, the positive optical
rotation of 5 ([α]D = +83) confirmed the absolute configura-
tion of the hopane skeleton. Methyl mannopyranoside (6)
was benzoylated and purified by HPLC as described[12] to
give tetrabenzoate 7. The 1H NMR and CD spectra of com-
Figure 2. HPLC–ESI-MS analysis of the crude glycolipid fraction
pound 7 were identical to those reported for synthetic of P. cf. lita and extracted-ion chromatograms of the [M + Na]+
methyl α-d-mannopyranoside perbenzoate.[12] This com-
pleted the elucidation of the structure and stereochemistry
of plakohopanoid peracetate 3b.
ions of several metabolites of the sponge. From top to bottom:
(a) total ion chromatogram; (b) extracted ion chromatogram of
mass m/z = 817.51 for plakohopanoid (3a); (c) extracted ion chro-
matogram of mass m/z = 569.45 for bacteriohopanetetrol (1a);
As reported above, we were not able to obtain natural 3a
in pure form. The possible pre-occurrence of an acetyl
group in the natural compound was therefore excluded by
acetylating a small portion of the crude glycolipid mixture
from P. simplex with deuterated acetic anhydride. The re-
(d) extracted ion chromatogram of mass m/z = 583.47 for 12-meth-
ylbacteriohopanetetrol (1c); (e) extracted-ion chromatogram of
mass m/z = 925.66 for the major homologue of discoside (4a); (f–
i) mass spectra at retention times 19.69, 20.38, 20.76, and
23.21 min, showing peaks for 3a (calcd. 817.5072), 1a (calcd.
569.4540), 1c (calcd. 583.4697), and 4a (calcd. 925.6587), respec-
sulting octakis(trideuteroacetate)-protected plakohopanoid tively.
4
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20676
/KAP1
Date: 13-08-12 16:55:13
Pages: 7
Plakohopanoid, a C32 Biohopanoid Ester
acetylation). Extracted-ion chromatograms were produced is further proof of the biosynthetic diversity and biotechno-
at mass values corresponding to plakohopanoid (3a) and to logical potential of sponge-associated bacteria.
three compounds previously isolated from P. simplex,
namely bacteriohopanetetrol (1a), 12-methylbacterioho-
panetetrol (1c),[7] and discoside (4a).[12] The chromatograms Experimental Section
showed peaks for all these compounds, with an elution or-
General Methods: Optical rotations were measured at 589 nm with
der that matched their expected lipophilicity (Figure 2). In
a Jasco P-2000 polarimeter using a 10 cm microcell. High Resolu-
addition, each of the full mass spectra recorded at the elu-
tion ESI-MS and ESI-MS/MS spectra were performed with a
tion time of the individual compounds showed a prominent
peak, whose exact mass value was in very good agreement
with the calculated mass.
Thermo LTQ Orbitrap XL mass spectrometer. The spectra were
recorded by infusion into the ESI source using MeOH as the sol-
vent. EI-MS spectra were performed using an Agilent 6850 II/5973
MSD GC–MS instrument and a HP-5MS capillary column (Ag-
ilent, 5% phenyl methyl siloxane; 30 mϫ0.25 mmϫ0.25 μm). He-
lium was used as a carrier gas, injection was in split mode, the
program was as follows: hold at 150 °C for 15 min, heat to 300 °C
at 5 °C min–1, hold at 300 °C for 10 min. NMR spectra were deter-
mined with Varian Unity Inova spectrometers at 500 MHz and
700 MHz. Chemical shifts were referenced to the residual solvent
signals (CDCl3: δH = 7.26, δC = 77.0; C6D6: δH = 7.15, δC = 128.15).
For an accurate measurement of the coupling constants, the one-
Conclusions
The structure of plakohopanoid (3a) was determined by
a combination of spectroscopic analysis and micro-scale
chemical degradation. Plakohopanoid represents a new
type of natural hopanoid derivative, composed of a C32 ho-
panoid acid ester linked to a carbohydrate part. Even
dimensional 1H NMR spectra were transformed at 64 K points
1
though 3a was isolated from a sponge, its constituent parts, (digital resolution: 0.09 Hz). Homonuclear H connectivities were
determined by COSY experiments. The HSQC and HMBC experi-
hopanoic acid and mannosyl-myo-inositol, are clearly of
bacterial origin. Therefore, its bacterial biosynthesis is very
likely, especially if one considers that sponge-associated
bacteria are extremely abundant in Plakortis species (they
account for over 50% of the sponge weight), and that there
is evidence[3] that several of the many metabolites isolated
from this sponge are actually produced by the bacterial
symbionts.
Until now, C32 hopanoic acids have been found in sedi-
ments and other geological formations, and have been con-
sidered to be diagenetic products, formed through abiotic
degradation of the biohopanoids present in bacteria. Diage-
netic transformation from biohopanoids to geohopanoids
can take place quickly after the death of bacteria, and is
well documented. However, the presence of plakohopanoid
in a marine living organism shows that there is a biosyn-
thetic pathway to C32 hopanoic acids, and these substances
should therefore no longer with certainty be considered to
be geohopanoids.
The occurrence of 12-methylbacteriohopanetetrol (1c) in
the Indonesian specimen of Plakortis also deserves a com-
ment. This compound was isolated by our research group
in 2000 from a Caribbean specimen of P. simplex,[7] and it
was the first example of a hopanoid methylated at C-12.
Since then, no further 12-methylhopanoids have been
found, which may be compared to the 19 naturally occur-
ring 2-methylhopanoids and the 25 naturally occurring 3-
methylhopanoids that have been described to date. Bacterial
communities associated with many species of sponges are
highly specific, phylogenetically very different from free-liv-
ing bacteria in the surrounding water, and consistently con-
served in specimens collected at different times and in dif-
ferent geographical areas,[14] suggesting vertical trans-
mission within their hosts[15] and an independent evolution-
ary history. Therefore, it is very possible that a unique bio-
synthetic pathway to 12-methylhopanoids has been devel-
1
ments were adjusted, respectively, for an average JCH of 142 Hz
2,3
and a
J
CH
of 8.3 Hz. HPLC was performed with a Varian Prostar
210 apparatus equipped with a Varian 350 refractive index detector.
Animal Material: A specimen of Plakortis cf. lita (order Homo-
sclerophorida, family Plakinidae) was collected in January 2008
along the coasts of the Bunaken island in the Bunaken Marine
Park of Manado. A voucher sample (No. MAN08–02) has been
deposited at the Dipartimento di Chimica delle Sostanze Naturali,
Università di Napoli Federico II.
Extraction and Isolation: The sponge (380 mL by volume before
extraction, and 50 g by dry weight after extraction) was cut into
pieces and extracted with MeOH (3ϫ 1.5 L), MeOH/CHCl3 (4ϫ
1.5 L) and CHCl3 (2ϫ 1.5 L). The MeOH extracts were partitioned
between H2O and BuOH, and the organic layer was added to the
CHCl3 extracts to give 8.5 g of a dark brown oil, which was sub-
jected to chromatography on a column packed with R-18 silica-gel.
The fraction that was eluted with 100% CHCl3 (1.8 g) was sub-
jected to further chromatographic separation on a SiO2 column
with solvent of increasing polarity. The fraction that was eluted
with EtOAc/MeOH (7:3) was mainly composed of glycolipids. This
fraction (0.4 g) was acetylated by treatment with Ac2O in pyridine
overnight, and then subjected to HPLC separation on a SiO2 col-
umn [10 μm, 250ϫ10 mm; eluent: n-hexane/EtOAc (6:4); flow:
4 mLmin–1], affording crasserides (44 mg),[8] bacteriohopanetetrol
1a and 12-methylbacteriohopanetetrol 1c (as a mixture, 112 mg),[7]
plakoside A and B and simplexides (as a mixture, 16 mg),[5,6] and
discoside (5 mg),[12] all as their respective peracetylated derivatives,
which were identified from their 1H NMR spectra. In addition, the
HPLC separation gave pure plakohopanoid as its peracetyl deriva-
tive 3b (2.1 mg).
Plakohopanoid Peracetate (3b): Colorless amorphous solid. [α]2D5
=
+57 (c = 0.1 in CHCl3). 1H NMR (700 MHz, CDCl3 and 500 MHz,
C6D6) and 13C NMR (175 MHz, CDCl3 and 125 MHz, C6D6): see
Table 1. HRMS (ESI+, MeOH): calcd. for C60H90O20Na
[M + Na]+ 1153.5918; found 1153.5885.
HPLC–ESI-MS Analysis: For LC-MS, the Orbitrap MS spectrom-
eter was coupled to an Agilent model 1100 LC, which included a
oped in some symbiotic bacteria of Plakortis sponges. This solvent reservoir, in-line degasser, binary pump, and refrigerated
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5

Job/Unit: O20676
/KAP1
Date: 13-08-12 16:55:13
Pages: 7
V. Costantino, G. D. Sala, A. Mangoni, C. Perinu, R. Teta
FULL PAPER
autosampler. A 2.6 μm Kinetex C18 column (50ϫ2.1 mm), main-
, and by the Italian Ministero dell’Università e della Ricerca
(MIUR) (PRIN2009: Sostanze naturali ed analoghi sintetici in
tained at room temperature, was used. It was eluted at 0.2 mLmin–1
with H2O and MeOH, using as a gradient elution 60–100% MeOH grado di interferire con target biologici coinvolti nel controllo della
over 15 min and hold 15 min. A 2 mgmL–1 MeOH solution of the
crude glycolipid fraction was prepared, and 5 μL of this solution
was injected. The results are shown in Figure 2.
crescita tumorale). We thank Prof. G. Bavestrello (Università di
Genova) and Dr. B. Calcinai (Università Politecnica delle Marche)
for their help in identifying the sponge. Mass and NMR spectra
were recorded at the “Centro di Servizio Interdipartimentale di An-
alisi Strumentale”, Università di Napoli “Federico II”. The assist-
ance of the staff is gratefully acknowledged.
Degradation Analysis of Compound 3b: Compound 3b (0.5 mg) was
dissolved in 1 n HCl in 92% MeOH (500 μL), and the solution
obtained was kept at 80 °C for about 12 h. The reaction mixture
was dried under nitrogen and then partitioned between H2O and
CHCl3. The H2O layer contained methyl glycoside 6, which was
benzoylated with benzoyl chloride (50 μL) in pyridine (500 μL) at
25 °C for 16 h. The reaction was quenched with MeOH, and after
30 min, the mixture was dried under nitrogen. Methyl benzoate was
removed by keeping the residue under vacuum with an oil pump for
24 h, and the residue was purified by HPLC [SiO2 column, 5 μm,
250ϫ4.6 mm; eluent: n-hexane/iPrOH (99:1); flow: 1 mLmin–1].
The chromatogram contained a peak (tR = 9.5 min), which was
collected and identified as methyl 2,3,4,6-tetra-O-benzoyl-α-d-man-
[1] M. Rohmer, Lipids 2008, 43, 1095–1107.
[2] V. Costantino, E. Fattorusso, C. Imperatore, A. Mangoni, Tet-
rahedron 2001, 57, 4045–4048.
[3] M. Laroche, C. Imperatore, L. Grozdanov, V. Costantino, A.
Mangoni, U. Hentschel, E. Fattorusso, Marine Biol. 2007, 151,
1365–1373.
[4] P. Schaeffer, G. Schmit, P. Adam, M. Rohmer, Org. Geochem.
2010, 41, 1005–1008.
[5] V. Costantino, E. Fattorusso, A. Mangoni, M. Di Rosa, A. Ian-
aro, J. Am. Chem. Soc. 1997, 119, 12465–12470.
[6] V. Costantino, E. Fattorusso, A. Mangoni, M. Di Rosa, A. Ian-
aro, Bioorg. Med. Chem. Lett. 1999, 9, 271–276.
[7] V. Costantino, E. Fattorusso, C. Imperatore, A. Mangoni, Tet-
rahedron 2000, 56, 3781–3784.
1
nopyranoside (7) by comparison of its H NMR and CD spectra
with those reported.[12] The CHCl3 layer from the methanolysis re-
action was concentrated and purified by HPLC [SiO2 column,
5 μm, 250ϫ4.6 mm; eluent: n-hexane/EtOAc (95:5); flow:
1 mLmin–1] to give pure compound 5 (0.2 mg).
[8] V. Costantino, E. Fattorusso, A. Mangoni, J. Org. Chem. 1993,
58, 186–191.
[9] G. Ourisson, M. Rohmer, Acc. Chem. Res. 1992, 25, 403–408.
[10] S. Neunlist, P. Bisseret, M. Rohmer, Eur. J. Biochem. 1988, 171,
245–252.
[11] P. K. Agrawal, D. C. Jain, R. K. Gupta, R. S. Thakur, Phyto-
chemistry 1985, 24, 2479–2496.
Methyl (22R)-33,34,35-Trinorbacteriohopan-32-oate (5): [α]2D5 = +83
(c = 0.015 in CHCl3); ref.[13] +58. 1H NMR (CDCl3): data matched
that reported.[13] MS (EI): m/z (%) = 484 (4.6) [M+], 469 (9.2), 369
(20.0), 263 (100), 231 (6.4), 191 (32.2).
[12] L. Barbieri, V. Costantino, E. Fattorusso, A. Mangoni, J. Nat.
Prod. 2005, 68, 1527–1530.
Supporting Information (see footnote on the first page of this arti-
cle): 1D and 2D NMR spectra of compound 3b. EI-MS and 1H
NMR spectra of compound 5.
[13] B. Peiseler, M. Rohmer, J. Chem. Soc. Perkin Trans. 1 1991,
2449–2453.
[14] U. Hentschel, K. M. Usher, M. W. Taylor, FEMS Microbiol.
Ecol. 2006, 55, 167–177.
[15] O. O. Lee, P. Y. Chui, Y. H. Wong, J. R. Pawlik, P. Y. Qian,
Appl. Environ. Microbiol. 2009, 75, 6147–6156.
Received: May 18, 2012
Acknowledgments
This work was supported by the European Commission through
the FP7 Projects no. 229893 (NatPharma) and 311848 (BlueGenics)
Published Online: ᭿
6
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20676
/KAP1
Date: 13-08-12 16:55:13
Pages: 7
Plakohopanoid, a C32 Biohopanoid Ester
Natural Products
Plakohopanoid is a new type of hopanoid
composed of a C32 hopanoic acid ester
linked to a mannosyl-myo-inositol. It is
probably produced by bacterial symbionts
of Plakortis cf. lita. The existence of a bio-
synthetic pathway to the C32 hopanoic acid
shows that this compound should not be
V. Costantino, G. D. Sala, A. Mangoni,*
C. Perinu, R. Teta ............................. 1–7
Blurring the Boundary between Bio- and
Geohopanoids: Plakohopanoid, a C32 Bio-
hopanoid Ester from Plakortis cf. lita
necessarily considered to be
panoid.
a geoho-
Keywords: Natural products / Biosynthesis /
Terpenoids / Glycolipids / Hopanoids
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7