Volatile amphibian pheromones: Macrolides from mantellid frogs from madagascar

Article abstract of DOI:10.1002/anie.201106592

Amphibians like water, but do they also notice volatile compounds in the air? Yes, they do. Macrolides, such as phoracantholide-J (see picture; upper right structure) or the newly discovered natural product gephyromantolide-A (left structure), are used for communication by mantelline frogs from Madagascar. Copyright

Full text of DOI:10.1002/anie.201106592

Angewandte
Chemie
DOI: 10.1002/anie.201106592
Pheromones
Volatile Amphibian Pheromones: Macrolides from Mantellid Frogs
from Madagascar
Dennis Poth, Katharina C. Wollenberg, Miguel Vences,* and Stefan Schulz*
Anuran amphibians (frogs) communicate mainly by acoustic,
visual, and tactile signals. However, frogs and other amphib-
ians can also use pheromones.[1,2] In some cases these have
been chemically characterized; in frogs, newts, and salaman-
ders they proved to be peptides or proteins, readily dissolving
in water or spreading on the water surface.[2,3] These
compounds are used by the males to attract females in
water. An example is the 25 amino acid peptide splendi-
pherin, which is used as a pheromone by males of the
Australian tree frog, Litoria splendida.[4] In salamanders,
proteins act as courtship pheromones, as shown for the red-
legged salamander Plethodon shermani and related species.[5]
Recently the sex pheromone of females of the Korean
salamander Hynobius leechii has been identified as prosta-
glandin F2a,[6] which is also released into water. Male newts of
the genus Cynops use 10 amino acid peptides such as sodefrin
and silefrin to attract females, again in water.[7] Apart from
these non-volatile cues, frogs have been shown to be able to
respond to volatile pheromonal cues, but their chemical
composition is still unknown.[8]
Figure 1. Ventral and dorsolateral view of a male Mantidactylus multi-
plicatus (left) and ventral view of a male Gephyromantis decaryi (right).
Arrows point at femoral glands in the two species, and at the gland
internal view of G. decaryi (bottom right).
The Mantellidae are a very species-rich family of small
frogs occurring in the rainforest of Madagascar.[9] A subfamily
of mantellids, the Mantellinae, is characterized by particular
glandular structures on the ventral side of the shanks of the
males (Figure 1).[10] The function of these femoral glands is
not known, but their sex-specific occurrence and their
location might point to a use in pheromonal communication.
Herein we present the surprising finding of nonpeptidic
volatile compounds, being structurally related to volatile
insect secretions and acting as pheromones, in these glands.
The femoral glands of five males of Mantidactylus multi-
plicatus were excised and extracted with dichloromethane.
The individual extracts were separately analyzed by gas
chromatography–mass spectrometry (GC-MS), and two
major volatile components A and B were detected from
each specimen (Figure 2). These compounds were accompa-
nied by a less volatile part of the secretion that comprised
diacetylated monoglycerides and to a lesser extent monoacy-
lated diglycerides.
[*] D. Poth, Prof. Dr. S. Schulz
Institut fꢀr Organische Chemie
Technische Universitꢁt Braunschweig
Hagenring 30, 38106 Braunschweig (Germany)
E-mail: stefan.schulz@tu-bs.de
Homepage: http://aks7.org-chem.nat.tu-bs.de
Dr. K. C. Wollenberg
Museum of Comparative Zoology, Department of Organismic and
Evolutionary Biology, Harvard University
26 Oxford St, Cambridge, MA 02138 (USA)
Prof. Dr. M. Vences
Zoologisches Institut, Technische Universitꢁt Braunschweig
Spielmannstrasse 8, 38106 Braunschweig (Germany)
Figure 2. Gas chromatogram of a dichloromethane extract of a femoral
gland of Mantidactylus multiplicatus. Two volatile major components A
and B are accompanied by a less volatile complex secretion of
glycerides. C: 3-(dodecanoyloxy)propane-1,2-diyl diacetate, D: 3-(palmi-
toyloxy)propane-1,2-diyl diacetate, X: artefact.
Supporting information for this article, including experimental
procedures, is available on the WWW under http://dx.doi.org/10.
1002/anie.201106592.
Angew. Chem. Int. Ed. 2012, 51, 2187 –2190
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2187

.
Angewandte
Communications
Compound A showed a mass spectrum similar to that of 2-
decanol, but a variation in the gas chromatographic retention
index, indicating 8-methyl-2-nonanol (1) as possible structure.
The synthesis of 1 verified the proposal; both natural
compound and synthetic material showed identical mass
spectra and gas chromatographic retention indices (see the
Supporting Information). The alcohol 1 was synthesized by
coupling the tosylate of 5-hexen-1-ol with isopropylmagne-
sium bromide, followed by Wacker oxidation and LAH
reduction. The enantiomers of the resulting rac-8-methyl-2-
nonanol were separated by enzymatic resolution of the
corresponding acetate. Chiral-phase gas chromatography
revealed that the natural material has a R configuration (see
the Supporting Information).
The mass spectrum of component B was identical to that
of phoracantholide J (2) published in mass spectrometry
databases. Already known synthetic routes to enantiomeri-
cally pure 2 were either quite long or used ring-closing
metathesis (RCM) as the cyclization method,[11,12] which in
our hands gave lower yields and lower stereoisomeric purity
than reported. A new and short synthesis was therefore
designed using the Corey–Nicolaou macrolactonization[13]
method as a key step. The enantiomers of lactone 2 were
synthesized as shown in Scheme 1 for the S enantiomer. The
Subsequently, we used 109 live individuals of M. multi-
plicatus to evaluate whether the compounds 1 and 2 might
influence their behavior. Individual frogs were placed in
plastic containers with cotton cloths on either half, one of
which was impregnated with either of the two compounds.
Experiments were conducted during the breeding season with
freshly collected specimens and under natural temperature
conditions. Each frog was tested eight times with several
hours of rest between experiments. On the basis of 800 series
of seven photos automatically taken after 5 min intervals for
all trials (see the Supporting Information), we measured frog
activity as the number of body position changes. Relative to
cotton cloths impregnated with water only, activity was
reduced by exposure to the hexane solvent. However, relative
to the hexane control, males showed a higher activity when
exposed to compounds 1 and 2 (Figure 3). Furthermore,
Figure 3. Activity (number of body position changes in 7 intervals of
5 min) in males (M) and females (F) of M. multiplicatus if exposed to
water (C-I, control; males 100, females 216 series of observations),
hexane (C-II, control; male 73, female 42) and hexane-dissolved
compounds 1 (male 78, female 46) and 2 (male 36, female 123),
indicating an increase of activity relative to hexane exposure (asterisks
mark significant differences to C-II; P<0.05).
males in their first change of sides in the container showed
a significant preference for moving towards compound 1 (see
the Supporting Information). Despite the apparent confound-
ing effect of the solvent, these preliminary results support that
compounds 1 and 2 are able to release changes in behavior of
conspecific individuals. Therefore, they conform to the
general definition of pheromones and might play a role in
intraspecific communication, although their exact function is
currently unknown.
Tropical amphibian species communities can be extremely
species-rich. In Ranomafana National Park in Madagascar,
over 100 mantelline species of often very similar morphology
are known to co-occur.[9] Therefore, it was then tested
whether in this area mantelline species contain the same or
different compounds in their femoral glands. Analysis of
glands from different species indicated that volatile com-
pounds are widespread in mantellines, but occur in species-
specific mixtures. In males of the closely related Mantidacty-
lus betsileanus from three different localities, compound 2 was
detected as well but alcohol 1 was absent. M. femoralis
femoral glands contained the closely related macrolide, (S)-
phoracantholide I (7), the hydrogenated analogue of 2, that
again is known from Phoracantha beetles in opposite config-
uration.[11,15] This compound was synthesized by hydrogena-
Scheme 1. Synthesis of phoracantholides I (7) and J (2): a) 1. Mg,
Et2O, 2. CuCN, 3. (S)-propylene oxide, 08C, 12 h, 91%; b) K2OsO4,
NaIO4, dioxane, H2O, 08C, 2 h, 71%; c) nBuLi, NaHMDS,
[Ph3P(CH2)3COOH]Br, THF, À788C to RT, 2 h, 69%; d) 1. dipyridiyl-
disulfide, 2. AgClO4, toluene, reflux, 14 h, 26%; e) H2, 10% Pd/C,
MeOH, 5 h, 75%.
alcohol 4 was obtained by reaction of homoallylmagnesium
bromide with (S)-propylene oxide. Lemieux–Johnson oxida-
tion[14] of the double bond furnished the aldehyde 5 that gave
preferentially the Z-configured hydroxyacid 6 by Wittig
reaction with (3-carboxypropyl)triphenylphosphonium ylide
(Z/E = 95:5). Finally, cyclization according to the Corey–
Nicolaou method[13] yielded phoracantholide J (2), which is
identical to the natural compound B. Compound 2 is a well-
known defensive secretion constituent from the metasternal
gland of Australian Phoracantha synonyma beetles,[11,15]
where it occurs as R enantiomer. In contrast, the S enan-
tiomer is produced by the frog, as the investigation using GC
on a chiral phase revealed (see the Supporting Information).
None of the two identified compounds were detected in the
ventral skin of the frogs, which is devoid of the major gland
complexes typical for the thighs.
2188
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 2187 –2190

Angewandte
Chemie
tion of lactone 2 (Scheme 1) and its absolute configuration
was established by chiral-phase GC.
a difficult-to-separate mixture of compounds, including the E
and Z isomers as well as smaller ring products, were formed
by isomerization of the double bonds prior to ring closure.
The other three enantiomers were obtained in an identical
manner by using appropriate methylated precursors 12 and
16. Pure 18 was obtained by isolation with semi-preparative
HPLC. Chiral-phase GC revealed that the enantiomer
(2S,6E,10R)-18 is indeed compound E. This compound
proved to be a new natural product for which we propose
the name gephyromantolide A. While compounds 1, 2, and 7
are metabolites of the fatty acid biosynthetic pathway, on first
sight compound 18 shows a typical terpenoid branching.
Nevertheless, it lacks one carbon compared to sesquiterpenes
and it therefore might be as well derived from fatty acid
metabolism. The methyl branches are probably introduced by
methylmalonate building blocks replacing usual malonate
building blocks during biosynthesis.
In several other species of Mantidactylus, and of other
mantellid genera such as Gephyromantis, volatile compounds
occurred in species-specific mixtures, in agreement with the
hypothesized function of these compounds in intraspecific
communication and recognition. Often these compounds
have a macrolide structure, and they are currently under
structural investigation. The secretion of Gephyromantis
boulengeri contained essentially only one major component
E. The glandular extract was directly analyzed by several
NMR spectroscopy experiments (see the Supporting Infor-
mation). From these data and the HR-GC-MS data, which are
consistent with the molecular formula C14H24O2, the structure
18 was proposed. Compound (2S,6E,10R)-18 was then
synthesized in enantiomerically enriched form to confirm
the proposal and allow determination of the absolute config-
uration of the natural compound E (Scheme 2). A key step in
In conclusion, we have established the occurrence of
volatile compounds to be linked
to male-specific glands and that
frogs alter their behavior upon
experimental exposure to them.
This strongly suggests that
amphibians use volatile phero-
mones,
a trait not reported
before from this vertebrate
group. Furthermore, mantelline
frogs produce species specific
mixtures of compounds, often
including macrolides, but also
a variety of other components.
In most frogs, acoustic commu-
nication is certainly the predom-
inant mechanism to attract
females over wider distances.
However, in highly diverse spe-
cies assemblages, chemical com-
munication with volatile com-
pounds may constitute a hitherto
underrated means to distinguish
conspecifics in closer vicinity.
This also emphasizes the poten-
tial of species-specific phero-
mones in species formation and
evolution of these amphibians.
Scheme 2. Synthesis of gephyromantolide A (18) from Gephyromantis boulengeri: a) H2SO4, EtOH, reflux,
6 h, 84%; b) Li2CuCl4, isopropenylmagnesium bromide, THF, 08C, 12 h, 91%; c) KOH, EtOH, H2O,
reflux, 4 h, 84%; d) SOCl2, Et2O, 08C, 12 h, 84%; e) (S)-4-phenyloxazolidin-2-one, nBuLi, THF, À788C to
RT, 12 h, 91%; f) NaHMDS, MeI, THF, À788C to RT, 12 h, 95%; g) KOH, MeOH, H2O, reflux, 4 h, 82%;
h) SOCl2, Et2O, 08C, 12 h, 91%; i) (R)-4-phenyloxazolidin-2-one, nBuLi, THF, À788C to RT, 12 h, 88%;
j) NaHMDS, MeI, THF, À788C to RT, 12 h, 93%; k) LiAlH4, Et2O, RT, 86%; l) 16+12, DMAP, EDC·HCl,
CH2Cl2, 08C, 2 h, 92%; m) Stewart–Grubbs catalyst II, C6F6, toluene, 808C, 3 h, 15%.
the synthesis was a ring-closing-metathesis to form the
macrocyclic ring. 5-Bromopentanoic acid (8) was transformed
into the ester and coupled with 2-propenylmagnesium
Received: September 16, 2011
Revised: November 18, 2011
Published online: January 20, 2012
bromide by Li2CuCl4 catalysis.[16] The resulting ester 9 was
transformed into the Evans amide 10. Alkylation with
NaHMDS/MeI yielded 11 in excellent yield and 99%
diastereoselectivity. Cleavage of the auxiliary furnished the
chiral acid 12. Again, an Evans approach furnished the
alkylated Evans amide 15 starting from 5-hexenoic acid 13.
Reduction with LAH gave the alcohol 16 that was coupled
with 12 to form the RCM precursor 17. The RCM proved to
be difficult, and several conditions and catalysts were tried.
Best results were obtained with the Stewart–Grubbs catalyst
in refluxing toluene in presence of C6F6.[17] Nevertheless,
Keywords: amphibians · chemical communication ·
chiral-phase gas chromatography · macrolides · pheromones
.
[1] L. D. Houck, Annu. Rev. Physiol. 2009, 71, 161 – 176; S. K.
Woodley, J. Comp. Physiol. A 2010, 196, 713 – 727.
[2] R. M. Belanger, L. D. Corkum, J. Herpetol. 2009, 43, 184 – 191.
[3] A. W. Perriman, M. A. Apponyi, M. A. Buntine, J. R. Jackway,
M. W. Rutland, J. W. White, J. H. Bowie, FEBS J. 2008, 275,
3362 – 3374.
Angew. Chem. Int. Ed. 2012, 51, 2187 –2190
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2189

.
Angewandte
Communications
[4] P. A. Wabnitz, J. H. Bowie, M. J. Tyler, J. C. Wallace, B. P. Smith,
Eur. J. Biochem. 2000, 267, 269 – 275; S. Kikuyama, K. Yama-
moto, T. Iwata, F. Toyoda, Comp. Biochem. Physiol. Part B 2002,
132, 69 – 74.
[10] M. Vences, G. Wahl-Boos, S. Hoegg, F. Glaw, E. Spinelli Oli-
veira, A. Meyer, S. Perry, Biol. J. Linn. Soc. 2007, 92, 529 – 539.
[11] T. Kitahara, K. Koseki, K. Mori, Agric. Biol. Chem. 1983, 47,
389 – 393.
[5] R. A. Watts, C. A. Palmer, R. C. Feldhoff, P. W. Feldhoff, L. D.
Houck, A. G. Jones, M. E. Pfrender, S. M. Rollmann, S. J.
Arnold, Mol. Biol. Evol. 2004, 21, 1032 – 1041; C. A. Palmer,
R. A. Watts, R. G. Gregg, M. A. McCall, L. D. Houck, R.
Highton, S. J. Arnold, Mol. Biol. Evol. 2005, 22, 2243 – 2256.
[6] J. Eom, Y. R. Jung, D. Park, Comp. Biochem. Physiol. Part A
2009, 154, 61 – 69.
[7] S. Kikuyama, F. Toyoda, T. Ohmiya, K. Matsuda, S. Tanaka, H.
Hayashi, Science 1995, 267, 1643 – 1645; K. Yamamoto, Y.
Kawai, T. Hayashi, Y. Ohe, H. Hayashi, F. Toyoda, G. Kawahara,
T. Iwata, S. Kikuyama, FEBS Lett. 2000, 472, 267 – 270.
[8] P. G. Byrne, J. S. Keogh, Anim. Behav. 2007, 74, 1155 – 1162.
[9] D. R. Vieites, K. C. Wollenberg, F. Andreone, J. Kçhler, F. Glaw,
M. Vences, Proc. Natl. Acad. Sci. USA 2009, 106, 8267 – 8272.
[12] G. H. Posner, M. A. Hatcher, W. A. Maio, Org. Lett. 2005, 7,
4301 – 4303; R. ChÞnevert, N. Pelchat, P. Morin, Tetrahedron:
Asymmetry 2009, 20, 1191 – 1196.
[13] E. J. Corey, K. C. Nicolaou, J. Am. Chem. Soc. 1974, 96, 5614 –
5616; H. Gerlach, A. Thalmann, Helv. Chim. Acta 1974, 57,
2661 – 2663.
[14] R. Pappo, D. S. Allen, R. U. Lemieux, W. S. Johnson, J. Org.
Chem. 1956, 21, 478 – 479.
[15] B. P. Moore, W. V. Brown, Aust. J. Chem. 1976, 29, 1365 – 1374.
[16] J. Klimentovꢀ, P. Kosꢀk, K. Vꢀvrovꢀ, T. Holas, A. Hrabꢀlek,
Bioorg. Med. Chem. 2006, 14, 7681 – 7687.
[17] I. C. Stewart, T. Ung, A. A. Pletnev, J. M. Berlin, R. H. Grubbs,
Y. Schrodi, Org. Lett. 2007, 9, 1589 – 1592; D. Rost, M. Porta, S.
Gessler, S. Blechert, Tetrahedron Lett. 2008, 49, 5968 – 5971.
2190 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 2187 –2190