3-Acetoxy-fatty acid isoprenyl esters from androconia of the ithomiine butterfly Ithomia salapia

  1. 1 ORCID Logo ,
  2. 1 ORCID Logo ,
  3. 2 ,
  4. 2,3 ORCID Logo ,
  5. 2 and
  6. 1 ORCID Logo
1Institute of Organic Chemistry, Technische Universität Braunschweig, Hagenring 30, 38106 Braunschweig, Germany
2Institut de Systématique Evolution Biodiversité, Centre National de la Recherche Scientifique, MNHN, Sorbonne Université, EPHE, Université des Antilles, 45 rue Buffon, CP 50, 75005 Paris, France
3Laboratoire Écologie, Évolution, Interactions des Systèmes Amazoniens (LEEISA), Université de Guyane, CNRS, IFREMER, 97300 Cayenne, France
  1. Corresponding author email
Associate Editor: J. S. Dickschat
Beilstein J. Org. Chem. 2020, 16, 2776–2787. https://doi.org/10.3762/bjoc.16.228
Received 31 Jul 2020, Accepted 27 Oct 2020, Published 16 Nov 2020
A non-peer-reviewed version of this article has been posted as a preprint https://doi.org/10.3762/bxiv.2020.90.v1
Full Research Paper
cc by logo


Male ithomiine butterflies (Nymphalidae: Danainae) have hairpencils on the forewings (i.e., androconia) that disseminate semiochemicals during courtship. While most ithomiines are known to contain derivatives of pyrrolizidine alkaloids, dihydropyrrolizines, or γ-lactones in these androconia, here we report on a new class of fatty acid esters identified in two subspecies, Ithomia salapia aquinia and I. s. derasa. The major components were identified as isoprenyl (3-methyl-3-butenyl) (Z)-3-acetoxy-11-octadecenoate, isoprenyl (Z)-3-acetoxy-13-octadecenoate (12) and isoprenyl 3-acetoxyoctadecanoate (11) by GC/MS and GC/IR analyses, microderivatizations, and synthesis of representative compounds. The absolute configuration of 12 was determined to be R. The two subspecies differed not only in the composition of the ester bouquet, but also in the composition of more volatile androconial constituents. While some individuals of I. s. aquinia contained ithomiolide A (3), a pyrrolizidine alkaloid derived γ-lactone, I. s. derasa carried the sesquiterpene α-elemol (8) in the androconia. These differences might be important for the reproductive isolation of the two subspecies, in line with previously reported low gene exchange between the two species in regions where they co-occur. Furthermore, the occurrence of positional isomers of unsaturated fatty acid derivatives indicates activity of two different desaturases within these butterflies, Δ9 and Δ11, which has not been reported before in male Lepidoptera.


The Neotropical butterfly tribe Ithomiini (Nymphalidae: Danainae) is very diverse and species-rich, with over 390 species and 50 genera [1,2] and extensively involved in Müllerian mimetic interactions [3]. Ithomiines are well suited for studies on speciation (species formation), as species often consist of multiple subspecies diverging for a number of adaptive traits, such as color pattern or host plants, which can then cause reproductive isolation. As such, they offer an excellent system to study the mechanisms underlying diversification and species recognition. Yet despite growing interests in this tribe, chemical differentiation between taxa has garnered surprisingly little attention until now.

Here we focus on the two closely related taxa, Ithomia salapia aquinia and I. s. derasa. The two subspecies have somewhat divergent wing color patterns (see Supporting Information File 1, Figure S1) [4], are widely distributed, and parapatric in north-eastern Peru [5]. Despite the geographic overlap in distribution, a recent genetic study showed limited gene flow [4]. Reproductive isolation in mimetic butterflies can be driven by multiple factors, notably non-random mating based on color pattern and/or sexual pheromones [6-8]. Determining whether the closely related subspecies of I. salapia differ in the chemical composition of volatiles is, therefore, of great interest.

All male ithomiine butterflies, including Ithomia, possess scent glands on their forewings, so-called androconia, covered with erectable hairpencils (Figure 1). They are used during courtship and are known to contain compounds acting as pheromones for the butterflies [2]. Adult ithomiines sequester pyrrolizidine alkaloids (PAs) pharmacophagously from various plants [9]. These alkaloids are transformed into the alkaloid and pheromone precursor lycopsamine (1, Scheme 1) [10-12] that can then be converted either into necine base derived compounds such as methyl hydroxydanaidoate (2), or into necic acids derived ones, e.g., ithomiolide A (3) [10-12]. While dihydropyrrolizines are also used by other Lepidoptera, e.g., danaines [13-15] or arctiines [16,17], γ-lactones derived from necic acids are specific to ithomiines of more derived taxa [11].


Figure 1: Extended hairs (arrow) of the androconia of a male Ithomia salapia aquinia (Photo: Melanie McClure).


Scheme 1: Pyrrolizidine alkaloid lycopsamine (1) and the putative pheromone compounds methyl hydroxydanaidoate (2) and ithomiolide A (3).

Past studies of the androconia of Ithomia have reported the presence of 3 in Ithomia iphianassa from Venezuela [10] and in I. salapia salapia from Ecuador [11], whereas no PA-derived compounds were found in any Ithomia spp. including I. agnosia agnosia [11]. Information on non-PA derived compounds in the androconia of ithomiines is mostly lacking, although we recently described (Z)-9-hydroxy-6-nonenoic acid and derivatives including dimers and fatty acid conjugates as major constituents of the androconia of Oleria onega [18].

Here we report on the chemical composition of the androconia of Ithomia salapia aquinia and I. s. derasa. A new type of butterfly scent gland constituents, acylated isoprenyl esters of fatty acids, is described, representing a combination of fatty acid and terpene biosynthesis. We also reveal small but reproducible differences between the two subspecies that could potentially be involved in species recognition and reproductive isolation.


Extracts from the wing androconia of I. s. derasa and I. s. aquinia were analyzed by GC/MS. The extracts consisted predominately of fatty acid esters with few other compounds (Table 1). While most ithomiines possess PA-derivatives in the androconia [10-12,18,19], only two of the five samples of I. s. aquinia contained small amounts of ithomiolide A (3), whereas PA derivatives were entirely absent in I. s. derasa.

Table 1: Compounds found in extracts of the androconia of Ithomia salapia derasa and I. salapia aquinia. Five individuals of each subspecies were analyzed. Only compounds occurring at least in two individuals of a subspecies are listed. The peak group refers to compounds eluting closely together. The number before the colon indicates the number of individuals carrying this compound, followed by the range of the relative amount.

No Compound Peak group Retention index I. salapia derasa I. salapia aquinia
1 ithomiolide A (3)   1219 2: 1.91–2.64
2 β-elemene   1388 4: 0.01–0.19
3 elemol/hedycaryol isomer   1517 3: 0.02–0.06
4 α-elemol (8)   1554 5: 0.11–2.88
5 elemol/hedycaryol isomer   1662 3: 0.01–0.02
6 hexadecenoic acid   1942 3: 0.55–5.76
7 hexadecanoic acid   1961 3: 0.02–0.25 3: 0.28–12.88
8 7-heneicosene   2081 3: 0.15–13.97
9 heneicosane   2100 3: 0.02–0.54
10 octadecenoic acid   2144 4: 0.62–3.69 2: 1.02–7.88
11 isoprenyl 9-hexadecenoate A 2233 3: 0.01–0.33
12 isoprenyl 11-hexadecenoate A 2244 5: 0.32–2.02
13 isoprenyl hexadecanoate A 2258 5: 0.11–2.24
14 tricosane   2300 5: 0.01–0.44 3: 001–0.11
15 11-methyltricosane   2335 4: 0.06–4.04 5: 0.02–0.89
16 eicosenoic acid   2360 3: 0.08–0.96
17 isoprenyl octadecadienoate B 2431 4: 0.01–0.30
18 isoprenyl 9-octadecenoate (10) B 2444 5: 0.36–8.27 5: 0.01–12.19
19 isoprenyl 11-octadecenoate B 2455 2: 0.01–0.02 4: 0.01–0.33
20 isoprenyl octadecanoate B 2463 5: 0.01–0.32 3: 0.01–0.07
21 isoprenyl 3-acetoxy-11-hexadecenoate B 2481 5: 0.10–0.40 5: 0.01–0.42
22 isoprenyl 3-acetoxyhexadecanoate B 2491 5: 0.30–1.32 5: 0.76–4.93
23 pentacosane B 2500 5: 0.01–0.13 3: 0.01–0.10
24 isoprenyl (2E,11Z)-2,11-octadecadienoate B 2506 4: 0.14–12.63 4: 0.16–0.89
25 isoprenyl (2E,13Z)-2,13-octadecadienoate B 2516 4: 0.01–0.92 4: 0.13–0.38
26 isoprenyl (E)-2-octadecenoate B 2523 5: 0.04–1.96 4: 0.18–0.56
27 11- and 13-methylpentacosane   2535 3: 0.02–0.05 2: 0.01–0.03
28 isoprenyl 3-hydroxy-11-octadecenoate C 2603 5: 1.10–5.02
29 isoprenyl 3-hydroxy-13-octadecenoate (24) C 2622 5: 0.07–0.40 2: 0.03–0.05
30 isoprenyl 3-hydroxyoctadecanoate C 2626 5: 0.98–2.41
31 isoprenyl (Z)-3-acetoxy-11-octadecenoate D 2678 5: 22.72–45.28 5: 14.58–41.42
32 isoprenyl (Z)-3-acetoxy-13-octadecenoate (12) D 2692 5: 3.87–14.67 5: 2.38–30.43
33 isoprenyl 3-acetoxyoctadecanoate (11) D 2698 5: 16.01–25.44 5: 26.20–43.73
34 isoprenyl 3-hydroxy-13-eicosenoate   2808 2: 0.01–0.45
35 isoprenyl 3-acetoxy-13-eicosenoate E 2874 5: 4.25–6.81 5: 0.01–1.20
36 isoprenyl 3-acetoxyeicosanoate E 2891 5: 0.02–0.35 3: 0.01–0.52

In contrast, the sesquiterpene α-elemol (8) was exclusively present in all tested individuals of I. s. derasa, together with some related minor sesquiterpenes. This sesquiterpene alcohol is likely formed from hedycaryol (7) during GC/MS analysis by a Cope-rearrangement [20,21], indicating that 7 might be originally present in the hairpencils. That said, we cannot disprove that this rearrangement could also occur in the androconia. Hedycaryol is an early product of sesquiterpene biosynthesis, formed by a 1,10-cyclization of the farnesyl cation 5 obtained from farnesyl pyrophosphate (4) (Scheme 2). Trapping the cation 6 with water leads to 7, which in turn might rearrange into 8 [22].


Scheme 2: Biosynthetic formation of hedycaryol (7) and α-elemol (8).

The fatty acid ester composition also differed between the two subspecies (Figure 2). Based on their elution order, five groups of compounds were detected, labelled A–E in Table 1. Groups A and B consisted of saturated and unsaturated C16 and C18 pentenyl esters. These compounds proved to be 3-methyl-3-butenyl esters, which were previously reported in bees [23,24]. Biosynthetically the alcohol part seems to originate from the terpene building block 3-methyl-3-butenyl (isoprenyl) pyrophosphate. Because isoprenyl pyrophosphate is partly converted to 3-methyl-2-butenyl (prenyl) pyrophosphate during terpene biosynthesis, the presence of prenyl esters could not be excluded. Nevertheless, the two ester types can be readily distinguished by EIMS. While 3-methyl-3-butenyl esters of saturated acids have a dominating ion at m/z 68, 3-methyl-2-butenyl esters show a peak pair m/z 68 and 69 of similar intensity (see Supporting Information File 1, Figure S2), as reported earlier [24]. This difference in the spectra can be explained by the different stabilization of the respective ions (Figure 3). The abundance of m/z 68 is higher in isoprenyl esters due to the more stable allyl radical cation (Figure 3A). In contrast, prenyl ester fragmentation produces a stabilized allyl cation m/z 69 (Figure 3B), while isoprenyl esters form a less stable homoallyl cation. This situation changes when a double bond is present in the acid part. In both isoprenyl (9) and prenyl esters (10) ion m/z 69 becomes the base peak, but the proportion of m/z 68 is higher in the former esters (Figure 4). Other significant differences can be found in the region around the acylium ions. Monounsaturated prenyl esters show the elimination of C5H10 (M − 70, m/z 280 in A), likely formed by rearrangement of an allylic H to the carbonyl group, followed by H-transfer (Figure 3C). Furthermore, the prenyl group can be lost (M − 69, m/z 281) and the acylium ion m/z 263 is formed. In contrast, isoprenyl esters lack the M − 69 ion, but additionally show acylium +1 and +2 ions (m/z 264 and 265 in A).


Figure 2: Total ion current chromatogram of androconial extracts of male butterflies of the two subspecies I. salapia derasa (A) and I. s. aquinia (B). The numbers refer to the entry numbers in Table 1.


Figure 3: Proposed mass spectrometric formation of characteristic ions in prenyl and isoprenyl esters. Formation of m/z 68 (A), m/z 69 (B), and m/z 280 (C).


Figure 4: Mass spectra and fragmentation of A: isoprenyl (3-methyl-3-butenyl) 9-octadenoate (9) and B: prenyl (3-methyl-2-butenyl) 9-octadecenoate (10). Red arrows show characteristics in the mass spectra differentiating prenyl and isoprenyl esters.

The location of the double bonds in the unsaturated esters was determined by dimethyl disulfide (DMDS) addition [25,26]. Because the double bond in the isoprenyl side chain would likely interfere, the esters were first transformed into the respective methyl esters via a microreaction with NaOMe [27]. The following DMDS derivatization revealed the presence of two isomers of each chain length, 9- and 11-hexadecenoate, as well as 9- and 11-octadecenoate (Supporting Information File 1, Table S1). Therefore, groups A and B consisted predominately of isoprenyl esters of saturated and unsaturated C16- and C18-acids.

Major components of both subspecies were group D compounds. The peak pair m/z 68/69 including the prominent base peak indicated again isoprenyl esters. The mass spectrum of the saturated compound showed a small putative M ion at m/z 410 and m/z 408 for the unsaturated analogs (Figure 5). A loss of 59/60 amu from M suggested an acetoxy group located somewhere along the chain. The position could not be derived from the mass spectrum. Nevertheless, the transesterified sample discussed before contained methyl hydroxyalkanoates, which allowed easy location of the hydroxy-group position by GC/MS [28]. The ion m/z 103 in the spectra of the three dominating acids confirmed the location of the acetoxy group at C-3 (see Supporting Information File 1, Figure S3). The positions of the double bonds in the methyl esters were determined by DMDS derivatization. The prominent ions present in these adducts allowed the localization of the double bonds in the natural products. Surprisingly, double bonds were found at C-11 and C-13, deducible by the ions m/z 145 ([CH3SC7H14]+), 261 ([CH3CO2C11H22SCH3]+), 243 (261 − H2O), as well as 213 (261 − H2O − CH2O), and m/z 117 ([CH3SC5H10]+), 289 ([CH3CO2C13H26SCH3]+), 271 (289 − H2O) as well as 241 (289 − H2O − CH2O), respectively (see Supporting Information File 1, Figure S4). An isomer with a C-9 double bond present in the simple isoprenyl esters was not detected. The configuration of the double bonds was confirmed to be (Z) as expected, because GC–DD-IR analyses showed a characteristic C–H stretch band at 3004 cm−1 (see Supporting Information File 1, Figure S9) [29,30]. As such, group D consisted of isoprenyl esters of 3-acetoxy-C18-fatty acids, a group of compounds not described before in nature. To confirm this, representative isomers were synthesized as outlined below.


Figure 5: Mass spectra and fragmentation of A: isoprenyl 3-acetoxyoctadecanoate (11); B: isoprenyl (Z)-3-acetoxy-13-octadecenoate (12).

Isoprenyl 3-acetoxyoctadecanoate (11) was synthesized according to Scheme 3. Hexadecanol (13) was oxidized to hexadecanal (14) using o-iodoxybenzoic acid (IBX) [31]. The resulting aldehyde was transformed into β-ketoacid 16 with ethyl diaazoacetate and SnCl2 [32], which upon reduction with NaBH4 in methanol delivered methyl 3-hydroxyoctadecanoate (17). Transesterification was performed with 3-methyl-3-buten-1-ol using distannoxan catalysis [33]. Final acetylation of the hydroxy esters delivered the target compound isoprenyl 3-acetoxyoctadecanoate (11). Comparison of mass spectra and retention index confirmed the identity of the naturally occurring compound and 11.


Scheme 3: Synthesis of isoprenyl 3-acetoxyoctadecanoate (11). a) IBX, EtOAc, 60 °C, 3.15 h, 99%; b) SnCl2, CH2Cl2, rt, 70%; c) NaBH4, 12 h, 98%; d) SnOBu2, 140°C, 36 h, 78%; e) Ac2O, pyridine, DMAP, CH2Cl2, 12 h rt, 67%.

An enantioselective synthesis of isoprenyl (Z)-3-acetoxyoctadec-13-enoate (12) was performed to verify the structural proposal and to determine the absolute configuration of the natural product (Scheme 4). The commercially available epoxide (S)-22 served as chiral starting material. 1,9-Nonanediol (19) was monobrominated and oxidized with IBX to yield 9-bromononanal (20). A Wittig reaction with pentylphosphonium bromide resulted in bromoalkene 21 in a 9:1 Z/E-mixture. In the following step, the Grignard reagent of 21 was converted into the respective Gilman cuprate with Cu(I)I for the selective reaction with the epoxide function of (S)-22 [34]. The hydroxyester 23 was obtained in good yield. The following stannoxane induced transesterification and the final acetylation procedure delivered 12. The two isomeric natural 3-acetoxyoctadecenyl esters had retention indices of 2678 and 2692, respectively, while synthetic 12 showed an I of 2688. Therefore, the second eluting ester is isoprenyl (Z)-3-acetoxy-13-octadecenoate, while the earlier eluting one is the 11-isomer.


Scheme 4: a) 48% HBraq, toluene, 24 h, 110 °C, 79%; b) IBX, EtOAc, 60 °C, 3.15 h, 90%; c) C5H11PPh3Br, LDA, THF, −78 °C, 12 h, 84%; d) i) Mg, 21, THF, ii) (S)-22, Cu(I)I, THF, –30 °C, 12 h, 79%; e) SnOBu2, 140°C, 36 h, 65%; f) Ac2O, pyridine, DMAP, CH2Cl2, 12 h rt, 74%.

With optically active material in hand, the absolute configuration of 12 was determined by enantioselective gas chromatography. Because direct separation of the large esters seemed to be difficult because of the high elution temperatures needed, we reasoned that the respective methyl 3-hydroxy esters would be much better suited, given the well-known separability of these compounds by chiral GC [35]. Therefore, a natural extract of the androconia and synthetic (R)-12 were transesterified with NaOMe as described above to yield methyl 3-hydroxyoctadecenoates. A synthetic sample of rac-12 obtained from rac-22 was also at hand. The analysis showed that only (R)-12 occurs naturally (Figure 6). Furthermore, the (Z)-configuration of the double bond was confirmed, because the minor amount of the (E)-isomer, present in the synthetic sample, did not coelute with the natural sample. Although only the configuration of natural 12 was determined to be exclusively (R), it seems likely that the other 3-acetoxy esters also show this configuration.


Figure 6: Separation of the enantiomers of methyl (Z)-3-hydroxy-13-octadecenoate (25) on a β-6-TBDMS hydrodex gas chromatographic phase. A) Natural extract; B) synthetic rac-25; C) synthetic (R)-25; X: methyl 3-hydroxy-11-octadecenoate; Y: methyl 3-hydroxyoctadecanoate; E: (E)-isomer of (R)-25. The enantiomer (S,E)-25 elutes together with (R,Z)-25, indicated by the broader base of this peak in B compared to C.

Group E compounds represented bishomologs of 12, isoprenyl eicosanoate and isoprenyl 13-eicosenoate, determined by DMDS derivatization. Next to these major esters, minor amounts of related esters occurred in some samples. These include deacylated 3-hydroxy esters, isoprenyl 3-hydroxyoctadecenoates and 3-hydroxyoctadecanoate, occurring in group C. Finally, respective elimination products, e.g., isoprenyl 2,11-octadecadienoate and isoprenyl 2-octadecenoate occurred in group B. The location of the C-2 double bond was verified by DMDS derivatization. Because deactivated bonds such as α,β-unsaturated double bonds are too unreactive for DMDS addition, their location can be verified by mass spectrometric fragments missing 2 amu compared to saturated analogs (see Supporting Information File 1, Figure S5) [36]. Hydroxy and α,β-unsaturated esters were not observed during GC of synthetic samples, making an artificial formation during chromatography unlikely.

In addition to the esters, minor amounts of alkanes and alkenes were present, as well as fatty acids. The latter occurred in varying amounts in the samples, maybe depending on variation in quality of the individual sample (see Tables S3 and S4, Supporting Information File 1). Acids are also present in other tissues of the butterflies, unlike the esters. Esters were also not detected in wings of female I. salapia.

The major components of the androconia were identical in both subspecies (Table 1). Variations were observed between individuals, and no defined proportion between saturated and unsaturated esters or between different double bond isomers were detected. Nevertheless, differences were present in the minor components and more volatile compounds (Table 1). Sesquiterpene 8 is restricted to I. s. derasa, whereas ithomiolide A (3) was present exclusively in some of the I. s. aquinia samples. C16-isoprenyl esters are only found in I. s. aquinia, while isoprenyl 3-acetoxy-13-eicosenoate, abundant in I. s. derasa, occurs only in trace amounts in I. s. aquinia.


The occasional occurrence of PA derivative 3 in two samples of I. s. aquinia may depend on the availability of the PA precursor in the wild. It might be that all individuals devoid of 3 simply had no access to PAs and/or that its absence is a specific trait of I. s. derasa. In contrast, elemol/hedycaryol (8) is specific to the latter subspecies. Although sesquiterpenes are common in plants, the occurrence of a single sesquiterpene might indicate individual biosynthesis in this subspecies or specific take-up, because plant sesquiterpenes usually occur in mixtures. Furthermore, hedycaryol is a quite simple sesquiterpene, needing only one biosynthetic cyclization step from the universal sesquiterpene precursor farnesyl pyrophosphate (Scheme 2) [22]. The differences in the isoprenyl esters reported are present in all individuals tested, pointing to distinct differences in activity of biosynthetic enzymes between the two subspecies.

Fatty acid esters, which were repeatedly reported to occur in androconia and male scent glands of butterflies [37-41], have been proposed to function e.g., as fixatives for more volatile pheromones [37], but their exact function remains mostly unknown. Because of the quite low volatility of the isoprenyl esters, especially of the major acetoxy esters, olfactory activity seems likely only in close vicinity of the male wings, although the evaporation rate might be increased by erection of the androconia hairpencils (Figure 1). Alternatively, direct or close contact might be needed for detection, probably taking place during contact of the female antennae with the male wings. What this potential signal might indicate remains speculative.

Nevertheless, the unusual location of the double bonds suggests an active function of the isoprenyl ester as signaling compounds [42]. Unsaturated fatty acid derivatives in pheromone glands are typically introduced by desaturases acting on saturated precursors. In the biosynthetically well-known butterfly genus, Bicyclus, a Δ11-desaturase, also often involved in moth pheromone biosynthesis, leads to derivatives of Δ11-C16 and C18-acids [43]. The fatty acid derivatives present in the male hairpencils of the danaine butterfly Lycorea ceres ceres also indicate the presence of a Δ11-desaturase [44]. In contrast, products consistent with Δ9-desaturase activity are present in androconia of the genus Heliconius [45]. Unlike these species, Ithomia seems to use at least two different desaturases, Δ9 and Δ11, leading to regioisomeric mixtures of isoprenyl esters. The position of the double bonds in the acyl chain of the esters can be explained by a biosynthetic pathway described in detail in Scheme 5. The double bond distribution is consistent with both desaturases acting on palmitic acid, leading to the respective hexadecenoic acids. These acids are the starting material for an additional elongation cycle of the fatty acid biosynthesis, leading en route to 3-hydroxy- and 2-alkenoic acids and finally to 11- and 13-octadecenoic acids. While the latter free acid was not observed, 9-octadecenoic acid was also present, formed likely by action of the Δ9-desaturase on stearic acid.


Scheme 5: Proposed biosynthetic pathway of fatty acids leading to the observed regioisomers of the isoprenyl esters. All acids shown were found in form of their isoprenyl esters. (Z)-9-Hexadecenoic acid is obtained from palmitic acid by a Δ9-desaturase (ds). Malonate elongation and reduction (red) leads to (Z)-3-hydroxy-11-octadecenoic acid, an intermediate of the fatty acid elongation cycle. The following elimination by a dehydratase (dh) leads to (2E,11Z)-2,11-octadecadienoic acid and after hydrogenation (hy) to (Z)-11-octadecenoic acid, completing the C2-elongation. A second elongation furnishes (Z)-3-hydroxy-13-eicosenoic acid. Similarly, a Δ11-desaturase gives (Z)-11-hexadecenoic acid, (Z)-3-hydroxy-13-octadecenoic acid and (2E,13Z)-2,13-octadecadienoic acid. Both desaturases might also act on octadecanoic acid, but only the elongation of (Z)-11-octadecenoic acid can be observed, leading to (Z)-3-hydroxy-13-eicosenoic acid. The saturated 3-hydroxyoctadecanoic and stearic acids as well as (E)-2-octadecenoic acids are obtained similarly directly from palmitic acid. The proposed biosynthesis likely takes place in form of the conjugated acids, e.g., coenzyme A esters or acyl carrier proteins. Finally, the acids are converted into the isoprenyl esters and the hydroxy acids are acylated.

3-Acetoxylated fatty acid esters are rarely found as natural products. Ethyl (S)-3-acetoxyeicosanoate and longer analogs are produced by the plant Schizolaena hystrix [46], but similar compounds from insects are unknown. Related are cactoblastins used as trail-following pheromones by Cactoblastis cactorum [47], which represent methyl esters of 3-hydroxy fatty acids acylated at O-3 with another fatty acid (structures see Supporting Information File 1, Figure S8).

The isoprenyl fatty acid esters are not restricted to the genus Ithomia within the Ithomiini. Preliminary analysis also revealed that these esters are also constituents of the androconia of e.g. Hypothyris anastasia, Hyposcada illinissa, H. anchiala, or Melinaea menophilus. In contrast, 9-hydroxynonanoic acid derived acids and esters are currently only reported from Oleria [18].


In summary, we here describe a group of esters, never before reported in nature, 3-acetoxyacyl isoprenyl esters from Ithomia salapia. The large amounts of these esters in the androconia and the specialized enzymes needed to produce them seem to indicate a pheromonal function of them, especially at close range. Differences in composition between the two subspecies suggest a possible role of the chemical bouquet in reproductive isolation, although other factors, such as wing color pattern, can also act as a reproductive barrier.

Supporting Information

Supporting Information File 1: Butterfly photos, mass, IR and NMR spectra, experimental procedures and analysis of individuals.
Format: PDF Size: 3.9 MB Download


We thank the Peruvian authorities and Dr. Gerardo Lamas (Museo de Historia Natural, Universidad Mayor de San Marcos) for research permits (236-2012-AG-DGFFS-DGEFFS). We also thank Armando Silva-Vásquez and Stéphanie Galusser for their precious help in the field. Melanie McClure allowed usage of her Ithomia-photo in the graphical abstract.


We thank the Deutsche Forschungsgemeinschaft (DFG, Schu/12-1) and the French National Agency for Research (ANR) for financial support. This work was funded by the Deutsche Forschungsgemeinschaft (DFG) (Schu 984/12-1) and the French National Agency for Research (ANR) through the project SPECREP (ANR-14-CE02-0011).


  1. Chazot, N.; Willmott, K. R.; Lamas, G.; Freitas, A. V. L.; Piron‐Prunier, F.; Arias, C. F.; Mallet, J.; De‐Silva, D. L.; Elias, M. Global Ecol. Biogeogr. 2019, 28, 1118–1132. doi:10.1111/geb.12919
    Return to citation in text: [1]
  2. Beccaloni, G. W. Trop. Lepid. 1997, 8, 103–124.
    Return to citation in text: [1] [2]
  3. Müller, F. Trans. Entomol. Soc. London 1879, xx–xxix.
    Return to citation in text: [1]
  4. Gauthier, J.; de Silva, D. L.; Gompert, Z.; Whibley, A.; Houssin, C.; Le Poul, Y.; McClure, M.; Lemaitre, C.; Legeai, F.; Mallet, J.; Elias, M. Mol. Ecol. 2020, 29, 1328–1343. doi:10.1111/mec.15403
    Return to citation in text: [1] [2]
  5. Dasmahapatra, K. K.; Lamas, G.; Simpson, F.; Mallet, J. Mol. Ecol. 2010, 19, 4283–4301. doi:10.1111/j.1365-294x.2010.04802.x
    Return to citation in text: [1]
  6. McClure, M.; Mahrouche, L.; Houssin, C.; Monllor, M.; Le Poul, Y.; Frérot, B.; Furtos, A.; Elias, M. J. Anim. Ecol. 2019, 88, 940–952. doi:10.1111/1365-2656.12975
    Return to citation in text: [1]
  7. Darragh, K.; Montejo-Kovacevich, G.; Kozak, K. M.; Morrison, C. R.; Figueiredo, C. M. E.; Ready, J. S.; Salazar, C.; Linares, M.; Byers, K. J. R. P.; Merrill, R. M.; McMillan, W. O.; Schulz, S.; Jiggins, C. D. Ecol. Evol. 2020, 10, 3895–3918. doi:10.1002/ece3.6079
    Return to citation in text: [1]
  8. González-Rojas, M. F.; Darragh, K.; Robles, J.; Linares, M.; Schulz, S.; McMillan, W. O.; Jiggins, C. D.; Pardo-Diaz, C.; Salazar, C. Proc. R. Soc. B 2020, 287, 20200587. doi:10.1098/rspb.2020.0587
    Return to citation in text: [1]
  9. Boppré, M. Naturwissenschaften 1986, 73, 17–26. doi:10.1007/bf01168801
    Return to citation in text: [1]
  10. Edgar, J. A.; Culvenor, C. C. J.; Pliske, T. E. J. Chem. Ecol. 1976, 2, 263–270. doi:10.1007/bf00988276
    Return to citation in text: [1] [2] [3] [4]
  11. Schulz, S.; Beccaloni, G.; Brown, K. S., Jr.; Boppré, M.; Freitas, A. V. L.; Ockenfels, P.; Trigo, J. R. Biochem. Syst. Ecol. 2004, 32, 699–713. doi:10.1016/j.bse.2003.12.004
    Return to citation in text: [1] [2] [3] [4] [5] [6]
  12. Schulz, S. Liebigs Ann. Chem. 1992, 829–834. doi:10.1002/jlac.1992199201137
    Return to citation in text: [1] [2] [3]
  13. Edgar, J. A. Philos. Trans. R. Soc., B 1975, 272, 467–476. doi:10.1098/rstb.1975.0097
    Return to citation in text: [1]
  14. Schulz, S. Eur. J. Org. Chem. 1998, 13–20. doi:10.1002/(sici)1099-0690(199801)1998:1<13::aid-ejoc13>3.0.co;2-r
    Return to citation in text: [1]
  15. Honda, Y.; Honda, K.; Ômura, H. J. Insect Physiol. 2006, 52, 1043–1053. doi:10.1016/j.jinsphys.2006.07.003
    Return to citation in text: [1]
  16. Schneider, D.; Boppré, M.; Zweig, J.; Horsley, S. B.; Bell, T. W.; Meinwald, J.; Hansen, K.; Diehl, E. W. Science 1982, 215, 1264–1265. doi:10.1126/science.215.4537.1264
    Return to citation in text: [1]
  17. Culvenor, C. C. J.; Edgar, J. A. Experientia 1972, 28, 627–628. doi:10.1007/bf01944939
    Return to citation in text: [1]
  18. Stamm, P.; Mann, F.; McClure, M.; Elias, M.; Schulz, S. J. Chem. Ecol. 2019, 45, 768–778. doi:10.1007/s10886-019-01100-5
    Return to citation in text: [1] [2] [3]
  19. Schulz, S.; Francke, W.; Edgar, J.; Schneider, D. Z. Naturforsch., C: J. Biosci. 1988, 43, 99–104. doi:10.1515/znc-1988-1-219
    Return to citation in text: [1]
  20. Jones, R. V. H.; Sutherland, M. D. Chem. Commun. 1968, 1229–1230. doi:10.1039/c19680001229
    Return to citation in text: [1]
  21. Kodama, M.; Yokoo, S.; Matsuki, Y.; Itô, S. Tetrahedron Lett. 1979, 20, 1687–1690. doi:10.1016/s0040-4039(01)93624-x
    Return to citation in text: [1]
  22. Baer, P.; Rabe, P.; Citron, C. A.; de Oliveira Mann, C. C.; Kaufmann, N.; Groll, M.; Dickschat, J. S. ChemBioChem 2014, 15, 213–216. doi:10.1002/cbic.201300708
    Return to citation in text: [1] [2]
  23. Ayasse, M.; Engels, W.; Hefetz, A.; Lübke, G.; Francke, W. Z. Naturforsch., C: J. Biosci. 1990, 45, 709–714. doi:10.1515/znc-1990-0623
    Return to citation in text: [1]
  24. Duffield, R. M.; Fernandes, A.; Lamb, C.; Wheeler, J. W.; Eickwort, G. C. J. Chem. Ecol. 1981, 7, 319–331. doi:10.1007/bf00995755
    Return to citation in text: [1] [2]
  25. Scribe, P.; Guezennec, J.; Dagaut, J.; Pepe, C.; Saliot, A. Anal. Chem. (Washington, DC, U. S.) 1988, 60, 928–931. doi:10.1021/ac00160a019
    Return to citation in text: [1]
  26. Schulze, T.; Weldon, P. J.; Schulz, S. Z. Naturforsch., C: J. Biosci. 2017, 72, 265–275. doi:10.1515/znc-2017-0006
    Return to citation in text: [1]
  27. Flachsbarth, B.; Fritzsche, M.; Weldon, P. J.; Schulz, S. Chem. Biodiversity 2009, 6, 1–37. doi:10.1002/cbdv.200800265
    Return to citation in text: [1]
  28. Ryhage, R.; Stenhagen, E. Ark. Kemi 1960, 15, 545–574.
    Return to citation in text: [1]
  29. Sémon, E.; Ferary, S.; Auger, J.; Le Quéré, J. L. J. Am. Oil Chem. Soc. 1998, 75, 101–105. doi:10.1007/s11746-998-0018-7
    Return to citation in text: [1]
  30. Schlawis, C.; Schulz, S. Nat. Prod. Rep. 2020. doi:10.1039/d0np00013b
    Return to citation in text: [1]
  31. More, J. D.; Finney, N. S. Org. Lett. 2002, 4, 3001–3003. doi:10.1021/ol026427n
    Return to citation in text: [1]
  32. Holmquist, C. R.; Roskamp, E. J. J. Org. Chem. 1989, 54, 3258–3260. doi:10.1021/jo00275a006
    Return to citation in text: [1]
  33. Otera, J.; Yano, T.; Kawabata, A.; Nozaki, H. Tetrahedron Lett. 1986, 27, 2383–2386. doi:10.1016/s0040-4039(00)84535-9
    Return to citation in text: [1]
  34. Huang, G.; Hollingsworth, R. I. Tetrahedron: Asymmetry 1998, 9, 4113–4115. doi:10.1016/s0957-4166(98)00441-8
    Return to citation in text: [1]
  35. König, W. A.; Gehrcke, B.; Pete, M. G.; Prestwich, G. D. Tetrahedron: Asymmetry 1993, 4, 165–168. doi:10.1016/s0957-4166(00)82328-9
    Return to citation in text: [1]
  36. Neumann, A.; Patzelt, D.; Wagner-Döbler, I.; Schulz, S. ChemBioChem 2013, 14, 2355–2361. doi:10.1002/cbic.201300424
    Return to citation in text: [1]
  37. Schulz, S.; Estrada, C.; Yildizhan, S.; Boppré, M.; Gilbert, L. E. J. Chem. Ecol. 2008, 34, 82–93. doi:10.1007/s10886-007-9393-z
    Return to citation in text: [1] [2]
  38. Yildizhan, S.; van Loon, J.; Sramkova, A.; Ayasse, M.; Arsene, C.; ten Broeke, C.; Schulz, S. ChemBioChem 2009, 10, 1666–1677. doi:10.1002/cbic.200900183
    Return to citation in text: [1]
  39. Estrada, C.; Schulz, S.; Yildizhan, S.; Gilbert, L. E. Evolution 2011, 65, 2843–2854. doi:10.1111/j.1558-5646.2011.01352.x
    Return to citation in text: [1]
  40. Wang, H.-L.; Brattström, O.; Brakefield, P. M.; Francke, W.; Löfstedt, C. J. Chem. Ecol. 2014, 40, 549–559. doi:10.1007/s10886-014-0452-y
    Return to citation in text: [1]
  41. Bacquet, P. M. B.; Brattström, O.; Wang, H.-L.; Allen, C. E.; Löfstedt, C.; Brakefield, P. M.; Nieberding, C. M. Proc. R. Soc. B 2015, 282, 20142734. doi:10.1098/rspb.2014.2734
    Return to citation in text: [1]
  42. Jurenka, R. Insect pheromone biosynthesis. In The Chemistry of Pheromones and Other Semiochemicals I; Schulz, S., Ed.; Topics in Current Chemistry, Vol. 239; Springer: Berlin, Heidelberg, 2004; pp 97–132. doi:10.1007/b95450
    Return to citation in text: [1]
  43. Liénard, M. A.; Hagström, Å. K.; Lassance, J.-M.; Löfstedt, C. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 10955–10960. doi:10.1073/pnas.1000823107
    Return to citation in text: [1]
  44. Meinwald, J.; Meinwald, Y. C. J. Am. Chem. Soc. 1966, 88, 1305–1310. doi:10.1021/ja00958a039
    Return to citation in text: [1]
  45. Mann, F.; Vanjari, S.; Rosser, N.; Mann, S.; Dasmahapatra, K. K.; Corbin, C.; Linares, M.; Pardo-Diaz, C.; Salazar, C.; Jiggins, C.; Schulz, S. J. Chem. Ecol. 2017, 43, 843–857. doi:10.1007/s10886-017-0867-3
    Return to citation in text: [1]
  46. Murphy, B. T.; Cao, S.; Norris, A.; Miller, J. S.; Ratovoson, F.; Andriantsiferana, R.; Rasamison, V. E.; Kingston, D. G. I. Planta Med. 2006, 72, 1235–1238. doi:10.1055/s-2006-947192
    Return to citation in text: [1]
  47. Rossi, F. M.; Rojas, D.; Cervasio, D. A.; Posillico, J.; Parella, K.; Fitzgerald, T. D. Chemoecology 2020, 30, 245–253. doi:10.1007/s00049-020-00314-4
    Return to citation in text: [1]
Other Beilstein-Institut Open Science Activities