Introduction

Many insects rely on sex pheromones for mate location, recognition and acceptance1. The encoded chemical information needs to be reliable to avoid costly sexual interactions or even mismating with closely related species that may use similar chemical signals. Hence, speciation is often accompanied by a diversification of the chemical signals that enable exclusive channels for sexual communication and result in behavioural isolation of the involved species2,3. The genetic and biochemical mechanisms underlying pheromone diversification and pheromone perception, however, are only poorly understood, although some progress has been made in research on moths4,5,6,7,8,9 and fruitflies10,11,12,13. In addition to these taxa, the parasitic wasp genus Nasonia has become an important model system to study the molecular mechanisms underlying pheromone diversification in insects. The genus consists of four species, N. vitripennis (Nv), N. giraulti (Ng), N. longicornis (Nl) and N. oneida (No), all of which parasitize the pupae of fly species14,15,16. While Nv is a cosmopolitan species, the distribution of the other species is restricted to North America. Nv is sympatric with Ng and No in eastern North America and with Nl in the west14,15. Interspecific mating is possible although the likelihood that a female accepts a heterospecific male depends on the species combination and female’s age17,18,19. However, in most combinations, interspecific mating results in all male broods due to Wolbachia-mediated cytoplasmic incompatibility20. Hence, mechanisms enabling behavioural isolation to avoid costly interspecific sexual interactions of sympatric Nasonia species have evolved. One of these mechanisms is mate discrimination during courtship presumably mediated by species specific female cuticular hydrocarbons and male derived aphrodisiac pheromones of yet unknown chemical structure17,18,19,21. Additionally, a mechanism for mate discrimination by volatile sex pheromones has evolved in Nv. Nasonia males produce volatile sex pheromones to attract virgin females21,22,23,24. All species of the genus produce the two pheromone components consisting of the major component (4R,5S)-5-hydroxy-4-decanolide (RS) and the minor compound 4-methylquinazoline (MQ). The pheromone of Nv, however, contains significant amounts of a third component, (4R,5R)-5-hydroxy-4-decanolide (RR) (ca. 30% of the amount of RS)22,25. The presence of RR allows Nv females to discriminate between the pheromone of conspecific males and the less complex blend of other Nasonia species18,25. Quantitative trait locus (QTL) analyses using Nv and Ng as well as RNAi gene knockdown experiments revealed that an array of three very similar genes (NV10127, NV10128 and NV10129), coding for putative short-chain dehydrogenases/reductases (SDRs), accounts for the pheromone difference between the two species25. It remained unclear, however, which of the three SDRs are actually involved and what their exact biochemical functions are.

SDRs are NAD(P)+/NAD(P)H-dependent oxidoreductases catalysing a variety of biochemical redox reactions26,27. Many SDRs catalyse the oxidation of hydroxyl groups or the reduction of carbonyls to alcohols, but they may also function as epimerases26,27,28,29. Therefore, it has been suggested that the Nasonia SDRs might synthesise RR by inverting the stereochemistry at carbon atom five with (4R)-5-oxo-4-decanolide (ODL) occurring as an intermediate25. It is still unknown, however, whether the Nasonia SDRs actually use RS as a substrate and, if so, whether they act alone as epimerases or concertedly by catalysing the oxidation and the reduction successively. Orthologues of the three SDR-encoding genes also exist in the genome of Ng, a species that, according to previous reports, does not contain RR in its pheromone blend. This raises the question of whether the ability of Nv males to synthesise RR is due to sequence-related differences in the catalytic activity of the SDRs or differential SDR expression. Previous SDR gene expression analyses by qPCR in abdomens of Nv and Ng revealed higher expression of NV10127 in males and of NV10129 in females25 but no data are available for isolated pheromone glands at the protein level.

In the present study, we expressed the SDR genes of Nv and Ng heterologously in Escherichia coli to study the catalytic activities of the recombinant enzymes. Specifically, we monitored the activity of the SDRs on synthetic precursors at different NAD+/NADH ratios. Furthermore, we performed proteomic analyses of the male pheromone glands by liquid chromatography/tandem mass spectrometry (LC-MS/MS). Finally, we conducted in vivo labelling experiments using 13C-labelled pheromone precursors to study whether the epimerisation of RS via the intermediate ODL occurs in live Nasonia males.

Results

To study the catalytic activities of the SDRs under controlled experimental conditions, we introduced the coding sequence of each of the three SDRs (NV10127, NV10128, and NV10129) from Nv and Ng into an E. coli-based expression system. After purification of the recombinant SDRs, which do not contain any signal peptides, prosthetic groups, or metal ions, we incubated equal amounts of each enzyme for 1, 5, and 22 h with enantiopure RS or RR in the presence of NAD+ or an equimolar mixture of NAD+/NADH as coenzymes. GC/MS analysis of the reaction products revealed that all six SDRs epimerised RS into RR and vice versa (Fig. 1). Epimerisation rates were dependent on the coenzyme availability (Supplementary Fig. S1) and differed between the various SDRs (i.e., identity of the SDR protein and species-specific origin) (Fig. 2). Although it is plausible that these differences are due to steric constraints at the active site, it is difficult to identify the responsible amino acid residues in the absence of a high-resolution crystal structure with bound substrate and coenzyme. NV10128 and NV10129 converted RS and RR with a higher epimerisation rate than NV10127, and Nv-derived SDRs were generally faster than the respective orthologues from Ng. The epimerisation rate of Nv-derived NV10128 with RS as substrate was striking as indicated by an abundance of 26±1.8% RR after a reaction time of only 1 h. Therefore, we studied the effect of coenzyme availability on the epimerisation rate in more detail with NV10128 of either species by increasing the NAD+/NADH ratio of the reaction mixture stepwise from 0 to 100%. Epimerisation rate of NV10128 from both species increased exponentially with increasing NAD+ availability (Fig. 3). The presence of small amounts of ODL in the reaction mixtures of the epimerisation experiments (Fig. 1a,b) showed that ODL is likely the intermediate formed during epimerisation. To confirm this conclusion, we incubated the SDRs for 1, 5 and 22 h with synthetic ODL in the presence of NADH, the reduced form of the coenzyme which is necessary to reduce ODL into the hydroxylactones. All SDRs reduced the keto group of ODL readily to RR and RS (Fig. 1c). Less than 5% of the added ODL was detectable after 1 h in all reactions and after 5 h, all enzymes had reduced more than 99% of the ODL. However, the stereoselectivity of the reduction differed between the SDRs (Fig. 2c). While NV10127 and NV10128 of either species produced RR biased ratios (>80% RR) of HDL, NV10129 produced higher proportions of RS and Nv-derived NV10129 produced a RS-biased HDL ratio.

Figure 1
figure 1

Epimerisation and reduction capacity of recombinant SDR NV10128 from Nasonia vitripennis.

Representative total ion current chromatograms of dichloromethane extracts obtained after incubating NV10128 for 5 h with (a) (4R,5S)-5-hydroxy-4-decanolide (RS) + 1.5 mM NAD+, (b) (4R,5R)-5-hydroxy-4-decanolide (RR) + 1.5 mM NAD+, or (c) (4R)-5-oxo-4-decanolide (ODL) and 1.5 mM NADH. The upper chromatograms in each panel show the control treatments in which the protein was omitted.

Full size image
Figure 2
figure 2

Functional characterisation of recombinant SDRs from Nasonia wasps.

Percentage of (a) RR and (b) RS formed (individual data points from n = 3 replicates are given) by NV10127, NV10128 and NV10129 from N. vitripennis and N. giraulti after addition of 0.5 mM RS and RR, respectively in the presence of 1.5 mM NAD+ as coenzyme; (c) percentage of RR formed by the same enzymes after addition of 0.5 mM (4R)-5-oxo-4-decanolide (ODL) in the presence of 1.5 mM NADH as coenzyme (summed peak areas of RS + RR = 100%).

Full size image
Figure 3
figure 3

Influence of the coenzyme availability on the epimerisation rate of NV10128.

Percentage of RR (mean + SEM, n = 3 for each coenzyme status) formed by NV10128 from N. vitripennis (orange) and N. giraulti (blue) after the addition of RS in the presence of different NAD+/NADH ratios (total coenzyme concentration 1.5 mM, summed peak areas of RS + RR = 100%, reaction time 5 h).

Full size image

Since the sequence-related catalytic characteristics of the SDRs alone did not explain the exclusive occurrence of RR in Nv, we asked whether species-related differences in SDR expression contribute to the unique pheromone composition of Nv males. The pheromones of Nasonia males are synthesised in the rectal vesicle30. Pheromone biosynthesis starts after emergence of the adult males and titres maximise in 2-day-old males22. Therefore, we studied the functional expression of SDR genes in the pheromone glands of 1–2 d old males of either species by a proteomic approach using LC-MS/MS. To this end, protein samples derived from 10 rectal vesicles were subjected to SDS-PAGE (n = 3 per species). Subsequent LC-MS/MS analysis of in-gel trypsin digested proteins revealed the presence of NV10127 and NV10128/NV10129 in the rectal vesicles of Nv males in all three replicates as depicted in Supplementary Fig. S2. Due to the very high sequence similarity of NV10128 and NV1012925, these two SDRs could not be distinguished based on the peptides detected. No peptide sequence uniquely present in either of the two SDRs was found. Peptides derived from NV10127 covered a greater proportion of its sequence compared to peptides from NV10128/NV10129 (Table 1). A quantitative difference in protein expression levels is reflected by the emPAI-values (exponentially modified protein abundance index), which can be used for an approximate relative quantitation of proteins in a mixture31. This index suggests that NV10127 is the most abundant of the three SDRs in the rectal vesicles of Nv males. According to the mass spectrometric analysis, the SDRs NV10127 and NV10128/NV10129 are also expressed in rectal vesicles of Ng males (Supplementary Fig. S3). However, numbers of detected peptides, sequence coverage and emPAI-values are lower compared to the SDRs of Nv, suggesting a lower expression level in Ng (Table 1). Moreover, NV10128/NV10129 could only be detected in one of the three replicates with Ng.

Table 1 Results of the mass spectrometric analysis of SDRs in the rectal vesicles of male N. vitripennis (Nv) and N. giraulti (Ng).
Full size table

Our results show that Ng, a species thought to produce no RR25, has the substrate RS and minor amounts of the SDR NV10127 and NV10128/NV10129 in the pheromone gland. This led us to hypothesise that, in contrast to previous reports25, at least traces of RR should also be detectable in Ng. A targeted search for RR in Ng-derived pheromone extracts revealed that traces of RR are indeed present (1.8 ng ± 0.2 ng/wasp (mean ± SEM), Supplementary Fig. S4).

To demonstrate that the epimerisation of RS to RR via the formation of ODL is a process occurring in live Nv males, we performed 13C in vivo labelling experiments. We used Ng males to produce partially 13C-labelled RS and ODL. For this purpose, we reared Ng males on hosts that had been experimentally enriched in 13C linoleic acid, a known precursor of RS and RR in Nasonia wasps32. Ng males reared on these hosts produced RS, which was 64.2 ± 2.6% 13C-labelled (Supplementary Fig. S5). Again, the Ng extracts contained traces of RR (0.14 ± 0.06%). We used an aliquot of the labelled RS to synthesise 13C-labelled ODL (Supplementary Fig. S6). Subsequently, we applied the 13C-labelled RS or ODL to the abdomen of Nv males and extracted their pheromone the following day. Males treated with labelled RS had significantly more labelled RR in their pheromone than the Ng-derived precursor (Mann-Whitney-U-test, P < 0.001, Fig. 4). This demonstrates that Nv males are able to epimerise RS into RR in vivo. Nv males treated with 13C-labelled ODL reduced the precursor to RS and RR, with a product ratio similar to the ratio found in untreated Nv males (Fig. 5a–e)22. Surprisingly, Ng males also produced significant amounts of RR when treated with 13C-labelled ODL. However, this was significantly less than the amounts typically found in Nv (Mann-Whitney-U-test, P < 0.001; Fig. 5a–e). No residue of the applied 13C-labelled ODL was detectable in the extracts of either species.

Figure 4
figure 4

In vivo epimerisation of 13C-labelled RS into RR by Nasonia vitripennis males (Nv).

Data show the relative abundance (horizontal line: median, box: 25–75 percent quartiles, whiskers: maximum/minimum range) of the diagnostic ion m/z 90 in the mass spectrum of RR calculated by integration of the mass trace m/z 90 at the retention times of RR and RS and relating the peak area of RR to the summed peak area of RR + RS. Left column: pheromone analyses (n = 9) of Nv males (orange) treated with a purified pheromone extract containing partially 13C-labelled RS as a precursor. Right column: multiple analysis (n = 11) of the Ng-derived (blue) precursor extract to determine the amounts of 13C-labelled RR present a priori and to evaluate the variability of the method (statistical analysis by Mann-Whitney-U-test).

Full size image
Figure 5
figure 5

In vivo reduction of partially 13C-labelled ODL to RR and RS by living N. vitripennis (Nv) and N. giraulti (Ng) males.

Shown are representative total ion chromatograms (TIC) and mass traces (diagnostic ion m/z 90, in purple) of pheromone extracts of (a) Nv and (c) Ng males treated with partially 13C-labelled ODL and (b,d) for control with the pure solvent. (e) Relative abundance (horizontal line: median, box: 25–75 percent quartiles, whiskers: maximum/minimum range) of labelled RR as found in the pheromone extracts (difference to 100% =RS, statistical analysis by Mann-Whitney-U-test, n = 10).

Full size image

Discussion

The present investigation demonstrates that all of the studied SDRs from Nv and Ng are capable of catalysing the epimerisation of RS into RR, as well as the reverse reaction, using NAD+ as coenzyme with ODL as an intermediate. However, there are clear sequence-related differences in the epimerisation activity of NV10128 and NV10129 from either species compared to the less active NV10127. In general, the orthologues from Nv were more effective in epimerising HDL than the respective enzymes from Ng. Nevertheless, our assays suggest that species-specific allelic variations of the SDRs alone are not sufficient to explain the pheromone difference between Nv and Ng, because the Ng-derived SDRs are also capable of epimerising the hydroxylactones. However, the semiquantitative LC-MS/MS analyses of the pheromone gland proteins revealed that both NV10127 and NV10128/NV10129 were more abundant in Nv than in Ng as indicated by the much higher emPAI values (Table 1). These differences in SDR concentrations might also contribute to the species specific pheromone compositions. Our data furthermore demonstrate that the coenzyme status has a strong influence on the epimerisation efficiency of the SDRs. Thus, the differing availability of NAD+ of the two species might be another factor contributing to the high abundance of RR in Nv (for a summarising model explaining the differing pheromone composition of Nv and Ng males see Fig. 6). However, a differing redox status would presumably influence not only the pheromone biosynthesis but also several other cell functions. Thus, potential differences in the redox status might be restricted to particular compartments of the pheromone producing gland such as the rectal papillae where at least some steps of the pheromone biosynthesis take place30.

Figure 6
figure 6

Proposed biochemical model explaining the differing pheromone composition of N. vitripennis (Nv) and N. giraulti (Ng) males.

(4R,5S)-5-hydroxy-4-decanolide (RS) is the primary pheromone component in both species. A significant proportion (ca. 30% of the total pheromone22) of RS is epimerised in Nv to the diastereomer (4R,5R)-5-hydroxy-4-decanolide (RR) by the NAD+-dependent SDRs NV10127, NV10128 and/or NV10129 that are more abundant in Nv than in Ng. Additionally, the Nv-derived SDRs have higher epimerisation rates. In contrast, the pheromone glands of Ng contain only traces of RR. Different redox states (NAD+/NADH ratios) in the pheromone glands might additionally contribute to the higher abundance of RR in Nv. Differing epimerisation rate is indicated by the size and differing abundance by the number of icons.

Full size image

The enzymes investigated in this study are characterised by very high amino acid sequence similarities both within and between species (Supplementary Fig. S7, for an analysis of the phylogenetic relationships of the three SDRs within and between both Nasonia species see Niehuis et al.25), with NV10128 and NV10129 sharing more than 97% amino acid residues while NV10127 has only 79% or 83% of the amino sequence in common with NV10128 and NV10129, respectively. The lower amino acid sequence similarity of NV10127 compared to the other two SDRs is reflected by its lower epimerisation activity. However, the activity of NV10127 to reduce ODL is comparable to that of NV10128 and NV10129. The enzymes studied here are “classical” NAD+/NADH dependent SDRs sensu Persson et al.27, thus exhibiting the typical highly conserved amino acid sequence motifs found in this class of biocatalysts (Supplementary Fig. S7). The binding site of the coenzyme is characterised by a TGxxxGxG motif and the preference for NAD+/NADH rather than NADP+/NADPH as a coenzyme is predicted by a negatively charged amino acid residue at the end of the second β-strand (here: Glu 45)27,33. The active site of most SDRs includes a S–Y–K triad (here: Ser 146, Tyr 158 and Lys 162) with an anionic Tyr acting as the catalytic base, Ser stabilising the substrate and Lys interacting with the nicotinamide ribose, thus lowering the pKa value of the Tyr-OH27,34. Mutational and structural analyses using 3β/17β-hydroxysteroid dehydrogenase as a model have revealed that a conserved Asn residue (here: Asn 115) is also involved in the catalytic mechanism35. SDRs are oxidoreductases with a wide range of biological functions including the functions described here, i.e., carbonyl and alcohol oxidoreduction26. The Nasonia SDRs catalyse both steps and thus function as epimerases. The best studied SDR with epimerase activity is UDP-galactose 4-epimerase (GALE), catalysing the interconversion between UDP-glucose and UDP-galactose within the Leloir pathway28,29,36. GALE contains a tightly bound NAD+ molecule, which stays attached to the enzyme during the reaction cycle and undergoes different redox state changes. After oxidation of galactose at carbon atom four and concomitant reduction of NAD+ to NADH, the resulting 4-ketopyranose rotates within the active site by about 180°, thus presenting the opposite side of the substrate to the NADH. Subsequently, the keto group of the sugar is reduced by hydride transfer from NADH, eventually resulting in a stereochemical inversion of the substrate’s C4 hydroxyl group26,28,29. We propose a similar mechanism for the stereochemical inversion at carbon atom five of the hydroxylactones catalysed by the Nasonia SDRs (Fig. 7). The differing epimerisation rates of the Nasonia SDRs might be explained by steric constraints that influence the mobility of the substrate molecule within the active site37,38. The proposed mechanism implies that most of the intermediate ODL is reduced to the epimer without leaving the enzyme-substrate complex. This and the fact that the reduction of ODL occurs very quickly (Fig. 2c) might explain why ODL has not been detected in the pheromone glands of Nasonia wasps so far. The application of 13C-labeled ODL to the abdomen of Ng males resulted in the formation of significant amounts of RR by these males while only minute amounts of RR occur in untreated wasps. We therefore conclude that ODL does not occur in significant amounts as a free intermediate in the pheromone glands of Nasonia males and that the traces of RR detected in Ng result from the epimerisation of RS catalysed by the diluted SDRs in the pheromone glands of Ng males (Table 1).

Figure 7
figure 7

Proposed mechanism for the epimerisation of RS to RR via ODL by the Nasonia SDRs in analogy to the mechanism known from UDP-galactose 4-epimerase.

As a first step of the epimerisation, the catalytic base (Tyr 158, B:) deprotonates the hydroxyl group and carbon 5 is concomitantly oxidised by the bound coenzyme NAD+ forming ODL and NADH. ODL then rotates 180°, thus presenting the opposite side of the substrate to the NADH. Subsequently, the ketone resident in ODL is reduced by hydride transfer from NADH ultimately resulting in a stereochemical inversion of the hydroxyl group.

Full size image

The fact that all SDR genes are also expressed in female wasps of both species25 suggests that the SDRs might have additional functions in Nasonia. Strikingly, the Nasonia SDR sequences have high sequence similarity with 15-hydroxyprostaglandin dehydrogenases (15-PGDH), enzymes that deactivate prostaglandins by oxidation of the 15(S) hydroxyl group39. Given the numerous hormonal functions prostaglandins have in insects40, the SDRs studied here might have evolved secondarily from 15-PGDHs. Characterisation of these putative ancestral functions, identification of possible alternative substrates and clarification of the structure-activity relationships underlying epimerase activity of the SDRs will help to further disentangle the mechanisms underlying sex pheromone evolution in Nasonia.

Methods

Insects

Nv originated from the inbred strain Phero01, which was also used in previous pheromone studies 22,24,25,30. Ng originated from the inbred strain NGVA2 and were kindly provided by Thomas Schmitt (University of Würzburg, Germany). Both species were reared on puparia of the green bottle fly, Lucilia caesar, as described elsewhere41. Wasps of defined age and mating status were obtained by dissecting hosts 1–2 days prior to emergence of the wasps and isolating single all-black parasitoid pupae in 1.5 ml microcentrifuge tubes until eclosion.

Chemical syntheses

(4R,5R)-5-hydroxy-4-decanolide

Enantiopure RR was synthesised by Sharpless Asymmetric Dihydroxylation from ethyl (4E)-dec-4-enoate (Molekula, Gillingham, UK) as described by Garbe & Tressl42.

(4R,5S)-5-hydroxy-4-decanolide

Enantiopure RS was prepared from RR by a Mitsunobu43 inversion of the C5 hydroxyl group. To a solution of pure RR (820 mg, 4.4 mmol) in THF (80 mL) at 0 °C was added p-nitrobenzoic acid (1.5 g, 8.8 mmol, Acros) and triphenylphosphine (2.5 g, 9.7 mmol, Acros). The resulting mixture was stirred until the solid reagents dissolved (5 min) and a solution of diethyl azodicarboxylate (40% in toluene, 4.4 mL, 9.7 mmol, Sigma-Aldrich) was added dropwise over 5 min. The cooling bath was removed and the reaction mixture was let warm to room temperature and stirred overnight. The volatiles were removed by rotary evaporation and the residue was purified by silica gel column chromatography (eluent: 20% ethyl acetate in hexanes) to give the ester intermediate as a mixture with triphenylphosphine. This mixture was dissolved in dry methanol (20 mL) and potassium carbonate (20 mg, 0.14 mmol) was added. The reaction mixture was let stir at room temperature for 2 h, the volatiles were removed, and the residue was purified by silica gel column chromatography (eluent: 20% ethyl acetate in hexanes) to give pure RS (140 mg, 17%). The purity of the product was verified by enantioselective GC/MS as described elsewhere22. The structure of purified RS was determined by NMR (1H, 13C, COSY, HSQC, HMBC) and comparison with the RR isomer using standard and enantioselective GC/MS.

(4R)-5-oxo-4-decanolide

ODL was synthesised by oxidation of RR using Dess-Martin-Periodinane as described by Garbe & Tressl42.

In vivo production of 13C-labelled precursors

We reared ca. 120 Ng males on hosts, which had been experimentally enriched in 13C-labelled linoleic acid as described elsewhere32. RS was extracted from two batches of 50 males each with 500 μl dichloromethane and purified by adsorption chromatography as described elsewhere22. Purified RS from one batch was dissolved in acetone and adjusted to a concentration of 10 μg/μl. After determination of the 13C incorporation rate into RS by GC/MS32, this solution was used for in vivo13C labelling (see below). The RS isolated from the second batch of male wasps was used to synthesise partially 13C-labelled ODL. For this purpose, the purified RS was dissolved in 500 μl dichloromethane and a spatula tip of Dess-Martin periodinane (Sigma-Aldrich) was added. The solution was shaken for 2 h at room temperature and excess periodinane was destroyed by washing the reaction mixture with 1 ml of 1 M sodium thiosulphate and 1 ml of saturated sodium hydrogen carbonate solution. The dichloromethane phase was removed, dried over sodium sulphate and analysed by GC/MS to determine the purity of ODL and the percentage of 13C incorporation. Finally, the solvent was removed under a stream of nitrogen and the ODL was re-dissolved in acetone (10 μg/μl) for in vivo13C labelling.

Gene Cloning

The coding-sequences of the target genes25 were amplified from cDNA-containing plasmid constructs (synthesised by GeneArt AG, Regensburg, Germany) using the oligonucleotide PCR primers specified in Table 2. Amplicons were inserted into the expression plasmid pET28a(+) using the introduced BamHI/XhoI restriction sites.

Table 2 cDNA-containing plasmid constructs and oligonucleotide primers used to PCR-amplify the nucleotide sequences of SDR-encoding target genes.
Full size table

Protein Expression and Purification

For expression of the SDR genes, E. coli BL21 CodonPlus (DE3) RIPL (Stratagene) was transformed with pET28a(+)-NV10127, pET28a(+)-NV10128, pET28a(+)-NV10129 (either from Nv or Ng), respectively. Transformed cells were grown at 37 °C in lysogenic broth (LB) with 50 μg/ml kanamycin and 30 μg/ml chloramphenicol to OD600 = 0.5. Gene expression was induced by addition of 0.5 mM IPTG. After growth overnight at 20 °C, cells were harvested by centrifugation and disrupted by ultrasonication. The recombinant proteins carrying an N-terminal His6-tag, a thrombin cleavage site and a 14 amino acid long linker were purified from the soluble fraction of the cell extract by Ni2+-affinity chromatography using a His Spin Trap column (GE Healthcare). For this purpose, proteins dissolved in a dilute imidazole buffer (100 mM potassium phosphate, pH 7.5, 300 mM potassium chloride and 10 mM imidazole) were loaded onto the column and subsequently eluted with a concentrated imidazole buffer (100 mM potassium phosphate, pH 7.5, 300 mM potassium chloride and 1 M imidazole). Elution fractions containing pure protein (>95%) as determined by SDS-PAGE were pooled, dissolved in 100 mM potassium phosphate, pH 7.5 using a NAPTM-5 column (GE Healthcare), dropped into liquid nitrogen, and stored at −80 °C. Protein concentrations were determined by measuring the absorbance at 280 nm using molar and specific extinctions coefficients (ε280nm for Nv/Ng are NV10127: 23045/23170 M−1 cm−1; NV10128: 20525/19035 M−1 cm−1; NV10129: 20525/20525 M−1 cm−1) that were calculated from the amino acid sequence (http://web.expasy.org/protparam/).

In vitro epimerisation and reduction assay with purified SDRs

For the assays with recombinant SDRs, a solution of 125 μg of each protein dissolved in 1.8 ml Tris-HCl buffer (pH 7.0) was added with 100 μl of either NAD+, NADH or NAD+ + NADH (dissolved in 1 M Tris-HCl buffer, total concentration of the coenzymes: 1.5 mM) and 19 μl of either RR, RS or ODL (dissolved in ethanol). The final concentration of the precursors was 0.5 mM which is below the concentration found in the pheromone gland (ca. 0.3 M, assuming a gland diameter of 100 μm, a spherical structure of the gland and a mean RS/RR amount of 240 ng/gland30. When ODL was used as a precursor (reduction assay), solely NADH was employed as coenzyme. When RR or RS was used as precursor (epimerisation assay) either NAD+ or the 1:1 mixture of NAD+/NADH was employed as coenzyme to investigate whether the coenzyme availability influences the epimerisation rate of the enzymes. Additionally, we investigated the coenzyme availability in more detail by incubating Nv-derived NV10128 for 5 h with RS in the presence of varying NAD+/NADH ratios (0–100% NAD+ increased by 10% steps, total coenzyme concentration 1.5 mM).

All reactions were kept at 30 °C, shaken at a rate 300 rpm, and samples of 500 μl each were taken after 1 h, 5 h and 22 h. All samples were extracted twice in 250 μl dichloromethane and analysed by GC/MS. Each combination of protein/precursor/coenzyme/time was tested in triplicate (n = 3), resulting in a total of 336 assays. Control assays were conducted for each treatment as described above without adding the SDRs.

GC/MS analysis

Chemical analyses were performed using the conditions and instrumentation described elsewhere32. For the detection of incorporated 13C in insect-derived HDL and ODL, the mass spectra at the expected retention times of RR, RS and ODL were scrutinised for the appearance of diagnostic ions (HDL: m/z 90, 107, 120, 196 Supplementary Fig. S5; ODL: m/z 46, 76, 89, 105, 194, Supplementary Fig. S6). Incorporation rates of the in vivo produced precursors were calculated by relating the peak areas of labelled diagnostic ions (HDL: m/z 90; ODL: m/z 105) to the total peak area of the respective labelled plus unlabelled (HDL: m/z 86; ODL: m/z 99) ions. To increase the sensitivity of the method and to detect minute amounts of 13C-labelled RR/RS in the in vivo labelling experiments, pheromone extracts were analysed in the selective ion monitoring (SIM) mode focusing on the diagnostic ion pairs m/z 86/90, 101/107, 115/120, and 186/196 (Supplementary Fig. S5). In vivo epimerisation of 13C-labelled RS into RR by Nv males was calculated by integration of the diagnostic ion trace at m/z 90 and relating the peak area of RR to the summed peak areas of RR and RS. These data were compared with the relative abundance of RR in the 13C-labelled precursor solution. In vivo reduction of 13C-labelled ODL to RR and RS by Nv and Ng was likewise monitored by the diagnostic ion trace at m/z 90. For analysis of the epimerisation assays with the recombinant SDRs, we related the peak area of the respective epimerisation product to the summed peak areas of the unreacted precursor and the product (RR + RS = 100%). For analysis of the reduction assay with ODL as precursor, we related the peak area of RR to the total peak area of both products (RR + RS = 100%).

Protein analysis by LC-MS/MS

To isolated rectal vesicles from 10 males of either N. vitripennis or N. giraulti (n = 3 replicates per species), 40 μl NuPAGE® LDS sample buffer (Invitrogen) containing 50 mM DTT were added. Samples were incubated for 20 min at 70 °C with two times vortexing in-between. After briefly spinning down cellular debris, the supernatant was subjected to SDS-PAGE on a precast 10% Bis-Tris gel (Invitrogen). Proteins were visualised by Coomassie-staining using SimplyBlue™ SafeStain (Invitrogen). For proteomic analysis, a gel lane was cut into 20 slices. The gel slices were washed consecutively with 50 mM NH4HCO3, 50 mM NH4HCO3/acetonitrile (3/1) and 50 mM NH4HCO3/acetonitrile (1/1), shrunk by adding 100% acetonitrile and lyophilised. Cysteines were blocked by reduction with DTT for 30 min at 57 °C followed by an alkylation step with iodoacetamide for 30 min at RT in the dark. Gel slices were washed again and lyophilised as described above. Subsequently, proteins were in gel-digested with trypsin (Trypsin Gold, mass spectrometry grade, Promega) overnight at 37 °C. Approximately 2 μg trypsin in 50 mM NH4HCO3 was used per 100 μl gel volume. Peptides were eluted twice with 100 mM NH4HCO3 followed by an additional extraction with 50 mM NH4HCO3 in 50% acetonitrile. Prior to LC-MS/MS analysis combined eluates were lyophilised and reconstituted in 20 μl of 1% formic acid. Peptides were separated on an UltiMate 3000 RSLCnano System (Thermo Scientific, Dreieich, Germany) by reversed-phase chromatography using a preconcentration column (C18 Acclaim Pepmap 100, 100 μm i.d. x 20 mm, Thermo Fisher) followed by a Reprosil-Pur Basic C18 nano column (75 μm i.d. x 250 mm, Dr. Maisch GmbH, Ammerbuch, Germany) in a linear gradient of 4% to 40% acetonitrile in 0.1% formic acid for 60 min at 300 nl/min. The LC-system was coupled to a maXis plus UHR-QTOF System (Bruker Daltonics, Bremen, Germany) via a CaptiveSpray nanoflow electrospray source (Bruker Daltonics). The mass spectrometer was operated in DDA mode at resolution of minimum 60000 for MS and MS/MS scans. MS/MS spectra were acquired with collision induced dissociation (CID) fragmentation. Compass 1.7 acquisition and processing software (Bruker Daltonics) allowed the use of a dynamic method with a fixed cycle time of 3 s and a m/z dependent collision energy adjustment between 34 and 55 eV. The scan rate of MS spectra acquisition was 2 Hz, the mass range of the precursor scan was set from m/z 175 to m/z 2000.

Raw data were processed in Data Analysis 4.2 (Bruker Daltonics) and processed by the Mascot database search engine using Protein Scape 3.1.3 (Bruker Daltonics). Mascot 2.5.1 (Matrix Science) was used to search the NCBI nr protein database. Furthermore, customised databases were searched comprising N. vitripennis and N. giraulti entries from NCBI supplemented with the three N. giraulti SDRs that were derived from cDNA sequencing and alignment25. Search parameters were as follows: enzyme specificity trypsin with 1 missed cleavage allowed, precursor tolerance 0.02 Da, MS/MS tolerance 0.04 Da, carbamidomethylation or propionamide modification of cysteine, oxidation of methionine, deamidation of asparagine and glutamine were set as variable modifications. The Protein Extractor function of ProteinScape facilitated protein list compilation. Finally, MS/MS spectra of Nasonia spp. peptides were subjected to manual validation.

In vivo13C labelling experiments

1–2-d-old Nv males were isolated in microcentrifuge tubes and cold-sedated on an ice bath. Subsequently, 0.1 μl of the acetone solutions containing partially 13C-labelled RS (n = 9) and ODL (n = 10), respectively, were applied to the abdominal tip of the wasps using a 5 μl microsyringe designed for GC on-column injection (Hamilton, Bonaduz, Switzerland). The possible reduction of ODL to HDL was also tested with Ng males (n = 10). Control wasps (n = 10 for each species) were treated with pure acetone. After 20 h, wasps were frozen at −20 °C and extracted for 30 min with 10 μl dichloromethane. These extracts were used for GC/MS analysis. Additionally, we analysed the 13C-RS solution that was used (ten times) to quantify the minute amounts of 13C-labelled RR present in the extract before application to the wasps and to evaluate the variability of the method.

Statistical analysis

The relative abundance of 13C-labelled RR in the in vivo labelling experiment and the RR/RS ratios resulting from the in vivo reduction of ODL by Nv and Ng males were compared by a two-sided Mann-Whitney-U-test using Past 3.0 scientific software.

Additional Information

How to cite this article: Ruther, J. et al. Epimerisation of chiral hydroxylactones by short-chain dehydrogenases/reductases accounts for sex pheromone evolution in Nasonia. Sci. Rep. 6, 34697; doi: 10.1038/srep34697 (2016).