Pheromones and analogs from Neozeleboria wasps and the orchids that seduce them: a versatile synthesis of 2,5-dialkylated 1,3-cyclohexanediones

Article abstract of DOI:10.1016/j.tetlet.2008.02.037

Chiloglottone, a wasp pheromone and attractant of sexually deceptive Chiloglottis orchids, and several structural analogs were synthesized. The synthetic approach is facile, high yielding and versatile, enabling rapid divergence to generate dialkylated analogs of chiloglottone. The key transformation was an organocadmium-mediated desymmetrization of glutaric anhydride derivatives. This library of synthetic 2,5-dialkylated 1,3-cyclohexanediones may assist in future identification of natural products in further species.

Full text of DOI:10.1016/j.tetlet.2008.02.037

Available online at www.sciencedirect.com
Tetrahedron Letters 49 (2008) 2446–2449
Pheromones and analogs from Neozeleboria wasps and the orchids
that seduce them: a versatile synthesis of 2,5-dialkylated
1,3-cyclohexanediones
Jacqueline Poldy a,b, Rod Peakall b, Russell A. Barrow a,*
a Department of Chemistry, The Australian National University, Canberra, ACT 0200, Australia
b School of Botany and Zoology, The Australian National University, Canberra, ACT 0200, Australia
Received 13 December 2007; revised 24 January 2008; accepted 8 February 2008
Available online 13 February 2008
Abstract
Chiloglottone, a wasp pheromone and attractant of sexually deceptive Chiloglottis orchids, and several structural analogs were
synthesized. The synthetic approach is facile, high yielding and versatile, enabling rapid divergence to generate dialkylated analogs of
chiloglottone. The key transformation was an organocadmium-mediated desymmetrization of glutaric anhydride derivatives. This
library of synthetic 2,5-dialkylated 1,3-cyclohexanediones may assist in future identification of natural products in further species.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords: Cyclohexanedione; Pheromone; Organocadmium reagents; Sexual deception; Chiloglottis; Neozeleboria
The orchids (family Orchidaceae) exhibit a profusion of
non-rewarding species that utilize deceptive strategies to
exploit their insect pollinators. Of these pollination syn-
dromes, sexual deception is exclusive to the Orchidaceae,
where it has evolved independently in European, African,
Central and South American, and Australian genera.1
The pollination syndrome is characterized by exquisite
examples of floral mimicry used to attract male insects to
the flower where copulation attempts (‘pseudocopulation’)
can lead to pollination.2 Of primary importance in eliciting
mating behavior in pollinators are floral odor bouquets
that mimic the female sex pheromone.3 Endorsed with
other sensory cues, many insects use olfactory stimuli as
their principal means of discrimination in sex attraction
and recognition.4 In the Australian orchid genus Chiloglo-
ttis all species employ sexual deceit,5 suggesting that odor
might be important for floral isolation and maintaining
species integrity,6 on account of the necessary specificity
of the pollinators’ pheromone. Gas chromatographic anal-
ysis in conjunction with electroantennographic detection
(GC EAD) has detected attractive compounds from floral
extracts of Chiloglottis species. In C. trapeziformis, polli-
nated by the thynnine wasp Neozeleboria cryptoides, the
structure of the active constituent from orchid labella and
female wasp extracts was subsequently determined via
mass spectral analysis from both sources, and confirmed
through synthetic replication of the natural product 2-
ethyl-5-propyl-1,3-cyclohexanedione, ‘chiloglottone’ (5a).7
Likewise, the same compound has been identified as a sex-
ual attractant to N. monticola, the pollinator of C. valida,
with important ecological and evolutionary implications.8
The desire to investigate speciation and reproductive
isolation within this orchid genus and the pheromone sys-
tems of the wasp pollinators is substantially facilitated by
available synthetic compounds, including known attractive
orchid metabolites and their analogs. Furthermore, as
biological material is available in very low abundance
(< 0.1 lg/labellum), insufficient for extensive data acquisi-
tion,7 synthetic material is necessary for structural clarifica-
tion. We proposed to develop a versatile route that would
*
Corresponding author. Tel.: +61 2 6125 3419; fax: +61 2 6125 0760.
E-mail address: rab@anu.edu.au (R. A. Barrow).
0040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.02.037

J. Poldy et al. / Tetrahedron Letters 49 (2008) 2446–2449
2447
deliver the natural product 5a as well as mono and 2,5-di-
alkylated cyclohexanedione analogs 5b–f, some of which
we anticipate may be represented in the natural environ-
ment. The selection of target compounds represents an
assortment of structural isomers and analogs, but is by
no means an exhaustive array (Fig. 1).
An earlier synthesis of 5a has been reported via a conju-
gate addition and Dieckmann-type condensation of two
asymmetric starting fragments, with subsequent decarb-
oxylation yielding the desired cyclohexanedione.7 How-
ever, the sequence would require de novo preparation of
the advanced precursors to generate a diversity of com-
pounds with alternative 2,5-dialkyl groups. Here, we
describe a general approach to 2,5-dialkylated cyclohexan-
ediones and report the first full characterization of chilo-
glottone (5a).
In our synthesis (Scheme 1), alkylation of diethyl malon-
ate delivered an assortment of derivatives 1,9 substituted at
what was to become the 5-position of the cyclohexanedione
products. The corresponding diols10 generated on LiAlH4
reduction were mesylated in good to excellent yield prior
to homologation through reaction with KCN in DMSO
to return the dinitriles 2.10a,11 These transformations were
conveniently monitored by IR spectroscopy, by means of
the conspicuous CN absorption at ca. 2240 cmꢀ1. Hydroly-
sis of the pure nitriles to glutaric acid derivatives12 was
followed by intramolecular condensation to deliver the
symmetric 4-alkyl substituted glutaric anhydride 310a,13 in
excellent yields, representing the key intermediates in the
synthesis.
zation, to deliver the appropriate 3-alkyl substituted d-keto
carboxylic acids. On treatment with diazomethane, the cor-
responding methyl esters14 4 were obtained in quantitative
yield. 13C NMR spectra showed resonances at d 210 and d
173 symptomatic of ketone and carboxyl carbon reso-
nances. Subsequent addition to the resulting ketone is
avoided due to the inherently inferior reactivity of alkyl-
cadmium species with respect to their magnesium
counterparts.15,16
Attempts to isolate crude keto acids in an aqueous
NaHCO3 extraction gave low returns, despite good to
excellent crude yields (79–100%). Somewhat unexpectedly,
the organic phase was found to retain the bulk of the mate-
rial. We propose that the biphasic distribution is attribut-
able to tautomerism between the keto acids and their
isomeric hydroxy lactones (Scheme 2). This elucidation is
consistent with accounts of similar equilibria, particularly
in structures capable of forming thermodynamically favor-
able 6-membered heterocycles.17,18
Ultimately, when the d-keto esters 4 were treated with
freshly prepared KOtBu the desired mono and 2,5-dialkyl-
ated 1,3-cyclohexanediones 5 were obtained. It is interest-
ing to note that these compounds exist exclusively as
their enol tautomers in CD3OD solution, as indicated from
their NMR data.19 Typical overall yields ranged from 26%
OH
R"
O
OH
O
O
R"
R'
O
R'
With the cyclic anhydrides in hand, the synthesis became
divergent through a dialkylcadmium mediated desymmetri-
Scheme 2. Proposed tautomerism of d-keto acids with hydroxy lactones.
O
O
O
O
O
O
O
O
O
O
O
O
5a
5b
5c
5d
5e
5f
Fig. 1. A selection of mono and 2,5-dialkylated 1,3-cyclohexanediones demonstrating the versatility of the reported method.
O
O
O
O
a
b, c, d
e, f
R"
EtO
OEt
EtO
OEt
NC
CN
R'
R'
1
2
R"
1
O
O
O
O
OH
3
O
5
O
O
CH2R"
3
g, h
1
1
3
5
i
5
5
3
CO2Me
R'
R'
R'
R'
3
4
5
Scheme 1. Preparation of mono and 2,5-dialkylated 1,3-cyclohexanediones exhibiting diverse alkyl substituents. Reagents and conditions: (a) NaOEt, R0X
(82–95%); (b) LiAlH4 (95–98%); (c) py, MsCl (73–93%); (d) KCN (80–98%); (e) aq NaOH (75–98%); (f) TFAA (97–100%); (g) Cd(CH2R00)2 (78–99%); (h)
CH2N2 (quant.); (i) KOtBu (73–98%).

2448
J. Poldy et al. / Tetrahedron Letters 49 (2008) 2446–2449
We are grateful to J. Allen and G. Lockhart of the ANU
mass spectrometry facility.
Supplementary data
Experimental procedures and characterization data for
1
compounds 1–5; H, 13C NMR spectra and GC EI-MS
of chiloglottone (5a) are available. Supplementary data
associated with this article can be found, in the online
version, at doi:10.1016/j.tetlet.2008.02.037.
19.00
20.00
21.00
22.00
23.00
Fig. 2. GC FID (top) and GC EAD (bottom) of chiloglottone (5a).
References and notes
1. Jersa´kova´, J.; Johnson, S. D.; Kindlmann, P. Biol. Rev. 2006, 81, 219–
235.
2. Schiestl, F. P. Naturwissenschaften 2005, 92, 255–264.
3. Schiestl, F. P.; Ayasse, M. Plant Syst. Evol. 2002, 234, 111–119.
4. Ayasse, M.; Paxton, R. J.; Tengo¨, J. Annu. Rev. Entomol. 2001, 46,
31–78.
5. Bower, C. C. Aust. J. Bot. 1996, 44, 15–33.
6. Mant, J. G.; Schiestl, F. P.; Peakall, R.; Weston, P. H. Evolution 2002,
56, 888–898.
to 48% over nine steps. Reaction mixtures were assessed by
GC MS to determine the extent and efficiency of the cycli-
zations. Two compounds for which cyclization proved
most challenging exhibited a 2-isopropyl substituent (5e,
5f). Despite more forcing conditions and longer reaction
times low yields of the desired products were returned
along with starting material.
The mono and 2,5-dialkylated 1,3-cyclohexanediones
generated using this divergent methodology represent struc-
tural isomers (5c, 5d, 5f) and smaller homologs (5b, 5e) of
the confirmed natural product 5a. Chiloglottone (5a) pro-
duced the expected electroantennographic response
(EAD) on freshly mounted N. cryptoides antennae (Fig. 2)
and field studies confirmed its attractive nature toward male
wasps. Interestingly, when 5b–f were evaluated employing
the EAD assay we also observed electroantennographic
responses comparable to those achieved with the naturally
occurring pheromone 5a. Although this in vitro activity is
not necessarily sufficient to confer a behavioral effect in field
trials, it supports the notion that diverse 2,5-dialkylated 1,3-
cyclohexanediones could represent sex attractants, which
may yet be detected in related taxa.
In summary, we have demonstrated a facile and versatile
method for synthesizing mono and 2,5-dialkylated 1,3-
cyclohexanediones that proceeded in good to excellent
yields. The protocol is applicable to a range of alkyl halides
to impart various 2- and 5-alkyl substituents. Moreover,
the glutaric acid derivatives represent stable and conve-
nient intermediates that enable rapid divergence to dialky-
lated analogs. The availability of a synthetic library of
analogs provides a practical tool with which to probe the
behavioral activity of the pheromones, and anticipates
potential natural products which may yet be encountered.
Efforts are currently directed at adapting the synthesis to
further broaden the existing library and encapsulate puta-
tive pheromones and will be reported as part of a full
paper.
7. Schiestl, F. P.; Peakall, R.; Mant, J. G.; Ibarra, F.; Schulz, C.;
Franke, S.; Francke, W. Science 2003, 302, 437–438.
8. Schiestl, F. P.; Peakall, R. Funct. Ecol. 2005, 19, 674–680.
9. (a) Chilmonczyk, Z.; Sienicki, Ł.; Łozowicka, B.; Lisowska-Kuz´micz,
´
M.; Jonczyk, A.; Aboul-Enein, H. Il Farmaco 2005, 60, 439–443 (Pr);
(b) Cahiez, G.; Alami, M. Tetrahedron 1989, 13, 4163–4176 (iPr, Et);
(c) Necˇas, D.; Tursky´, M.; Kotora, M. J. Am. Chem. Soc. 2004, 126,
10222–10223 (Me).
10. (a) Shintani, R.; Fu, G. C. Angew. Chem., Int. Ed. 2002, 41, 1057–
1059 (Pr); (b) Zhang, F.; Peng, L.; Zhang, T.; Mei, M.; Liu, H.; Li, Y.
Synth. Commun. 2003, 33, 3761–3770 (iPr); (c) Gaucher, A.; Ollivier,
J.; Marguerite, J.; Paugam, R.; Salaun, J. Can. J. Chem. 1994, 72,
¨
1312–1327 (Et).
11. Becker, J. Y.; Koch, T. A. Electrochim. Acta 1994, 39, 2067–2071 (Et).
12. Irwin, A. J.; Jones, J. B. J. Am. Chem. Soc. 1977, 99, 556–561 (iPr).
13. Theisen, P. D.; Heathcock, C. H. J. Org. Chem. 1993, 58, 142–146 (Et,
iPr).
14. Blackburne, I. D.; Park, R. J.; Sutherland, M. D. Aust. J. Chem. 1971,
24, 995–1007.
15. Cason, J. J. Org. Chem. 1948, 13, 227–238.
16. Cason, J. J. Am. Chem. Soc. 1946, 68, 2078–2081.
17. Saeed, A. Monatsh. Chem. 2003, 134, 457–463.
18. Hendrickson, J. B.; DeCapite, P. M. J. Org. Chem. 1985, 50, 2112–
2115.
19. Characterization data for selected compounds: Compound 3 (R0 = Pr):
1H NMR (300 MHz, CDCl3) d 2.87 (2H, dd, 2J = 17.1, 3J = 4.5,
H-3a, 5a), 2.41 (2H, dd, 2J = 17.1, 3J = 10.3, H-3b, 5b), 2.16 (1H, m,
H-4), 1.39–1.37 (4H, m, CH2-10, CH2-20), 0.92 (3H, m, CH3-30); 13C
NMR (75 MHz, CDCl3) d 166.5 (C, C-2, 6), 36.6 (CH2, C-10), 36.0
(CH2, C-3, 5), 28.4 (CH, C-4), 19.5 (CH2, C-20), 13.7 (CH3, C-30); EI-
MS m/z (%): 157 (52, [M+H]+), 97 (40), 84 (79), 70 (67), 69 (66), 56
(94), 55 (81), 43 (99), 42 (100). 3-Propyl-5-oxooctanoic acid: IR mmax
3050, 2960, 2934, 2875, 1700 cmꢀ1 1H NMR (300 MHz, CDCl3) d
;
11.1 (1H, br, s, 1-CO2H) 2.46–2.24 (7H, m, CH2-2, CH2-4, CH2-6,
H-3), 1.57 (2H, tq, 3J = 7.4, 3J = 7.4, CH2-7), 1.39–1.25 (4H, m, CH2-
10, CH2-20), 0.88 (6H, m, CH3-8, CH3-30); 13C NMR (75 MHz,
CDCl3) d 210.6 (C, C-5), 178.8 (C, C-1), 46.6, 45.1 (CH2, C-4, C-6),
38.2 (CH2, C-2), 36.3 (CH2, C-10), 30.4 (CH, C-3), 19.8 (CH2, C-20),
17.1 (CH2, C-7), 14.0, 13.7 (CH3, C-8, C-30); GC EI-MS m/z (%): 200
(<1, [M]+ꢁ), 182 (2), 157 (26), 129 (14), 111 (19), 97 (20), 87 (37), 83
(37), 71 (85), 69 (43), 55 (49), 43 (100); HREI-MS: found 200.1413
(calculated for C11H20O3 200.1412). Compound 4 (R0 = Pr, R00 = Et):
Acknowledgments
This work was supported by an ARC Discovery Project
(DP0451374) from the Australian Research Council. J.P.
thanks the ANU for an Australian Postgraduate Award.
IR mmax 2959, 2934, 2874, 1738, 1713 cmꢀ1 1H NMR (300 MHz,
;

J. Poldy et al. / Tetrahedron Letters 49 (2008) 2446–2449
2449
CDCl3) d 3.60 (3H, s, CO2CH3) 2.45–2.20 (7H, m, CH2-2, CH2-4,
(500 MHz, CD3OD) d 2.46 (2H, dd, 2J = 16.5, 3J = 4.3, H-4eq, 6eq),
3
3
3
2
3
CH2-6, H-3), 1.54 (2H, tq, J = 7.4, J = 7.4, CH2-7), 1.39–1.22 (4H,
m, CH2-10, CH2-20), 1.86 (6H, m, CH3-8, CH3-30); 13C NMR
(75 MHz, CDCl3) d 210.2 (C, C-5), 173.2 (C, C-1), 51.3 (CO2CH3),
46.8 (CH2, C-4), 45.0 (CH2, C-6), 38.2, 36.4 (CH2, C-2, C-10), 30.6
(CH, C-3), 19.8 (CH2, C-20), 17.1 (CH2, C-7), 14.0, 13.6 (CH3, C-8,
C-30); EI-MS m/z (%): 214 (6, [M]+ꢁ), 183 (22), 171 (79), 143 (25), 129
(93), 128 (33), 111 (25), 97 (77), 83 (47), 71 (78), 69 (64), 55 (63), 43
(100); HREI-MS: found 214.1573 (calculated for C12H22O3 214.1569).
Compound 5a (R0 = Pr, R00 = Et, chiloglottone): 1H NMR
2.25 (2H, q, J = 7.5, CH2-100), 2.14 (2H, dd, J = 16.5, J = 11.3, H-
4
ax, 6ax), 2.04 (1H, m, H-5), 1.38 (4H, m, CH2-10, CH2-20), 0.94 (3H, t,
3
3J = 6.5, CH3-30), 0.90 (3H, t, J = 7.5, CH3-200); 13C-APT NMR
(125 MHz, CD3OD) d 176.5 (C, C-1, 3), 118.1 (C, C-2), 39.3 (CH2, C-
4, 6), 38.8 (CH2, C-10), 34.4 (CH, C-5), 20.7 (CH2, C-20), 16.0 (CH2,
C-100), 14.4, 13.6 (CH3, C-30, C-200); EI-MS m/z (%): 182 (60, [M]+ꢁ),
167 (12), 154 (6), 139 (23), 125 (27), 112 (66), 97 (100), 84 (64), 69 (95),
55 (89), 43 (48), 41 (60); HREI-MS: found 182.1307 (calculated for
C11H18O2 182.1307).