The biological significance of lipogenesis in Nasonia vitripennis

Abstract

Parasitic wasps have long been thought to be unable to synthesize fatty acids de novo, but recent ¹³C-labelling studies have challenged this view. It remained unclear, however, whether the reported biosynthesis rates are of biological relevance. Here, we show in Nasonia vitripennis that ageing females with partly depleted lipid reserves produce biologically relevant amounts of fatty acids de novo. Females with varying oviposition history (0–48 h) prior to feeding 20% ¹³C-labelled glucose solution showed ¹³C-incorporation rates of (mean ± SEM) 30 ± 2%, 50 ± 2%, 49 ± 3% and 21 ± 2% in palmitic, stearic, oleic and linoleic acid, respectively. The absolute amounts of fatty acids synthesized de novo across treatments corresponded to 28 ± 3 egg lipid equivalents. Females incorporated de novo synthesized fatty acids into their eggs, and glucose-fed females laid more eggs than water-fed control females. The number of eggs laid prior to glucose feeding did not correlate with the degree of lipogenesis, but the amounts of de novo synthesized fatty acids correlated with constitutive (not synthesized de novo) fatty acids. Hence, glucose feeding has a twofold effect on the fatty acid status of N. vitripennis females by decelerating the catabolism of existing fat reserves and partially replenishing ebbing fat reserves by lipogenesis.

1. Introduction

Lipogenesis, the conversion of carbohydrates via acetyl-CoA into fatty acids and their derivatives [1–4], is thought to be highly conserved in organisms [5]. It enables them to produce these multifunctional primary metabolites independently from nutritional lipids and use them, e.g. for the storage of energy, the formation of biomembranes or the production of essential signalling molecules [6,7]. Insects, like most animals, typically accumulate lipids when having unlimited access to carbohydrate sources [8]. In parasitic wasps, however, a hyperdiverse group of hymenopterans developing in or on other arthropods, many species were found to lack lipid accumulation despite ad libitum supply with carbohydrates, i.e. they did not show higher lipid mass after sugar feeding than prior to [8]. Although this was not true for all studied species, the absence of lipid accumulation in many species has often been erroneously equated to their general inability to synthesize fatty acids de novo from carbohydrates [8–10]. It has been argued that parasitic wasps receive sufficient fatty acid derivatives from their host having made lipogenesis redundant. In fact, parasitic wasps typically emerge from their host with plenty of lipids [8], and it is unlikely that they are motivated under these conditions to further accumulate lipid reserves by converting carbohydrates into fatty acid derivatives. This, however, does not necessarily mean that parasitic wasps are generally unable to synthesize fatty acids. It has been demonstrated, for instance, that parasitic wasps developing on fat-poor hosts accumulate lipids upon sugar feeding but do not gain lipid mass when developing on fat-rich hosts [11]. Furthermore, even those species emerging with rich lipid reserves might synthesize fatty acids later in their lives to replenish ebbing fat reserves due to energy-sapping activities such as foraging and egg laying, without resulting in measurable lipid mass gain. In agreement with this idea, recent studies using stable isotope labelling approaches and chemical analyses by gas chromatography-mass spectrometry (GC/MS) demonstrated that parasitic wasps do not generally lack lipogenesis but are able to convert glucose into fatty acid derivatives [12–14]. Wasps fed ¹³C-labelled α-d-glucose showed the ¹³C-label in fatty acid methyl esters (FAME) obtained by transesterification of their raw lipids. Furthermore, ageing adults of the braconid wasp Meteorus pulchricornis lost lipid mass despite having access to honey solution, but those wasps that had their fatty acid synthase (fas) genes knocked down by RNAi had lower lipid mass than comparable control wasps with intact fas [15]. While the general ability to synthesize fatty acids has now been established for a wide range of taxa, including those that were previously thought to lack this ability, it is unknown in many species whether lipogenesis occurs at fitness-relevant scales. ¹³C-incorporation rates less than 6% as found for many species in the ¹³C-labelling studies [12,13] might appear too low to be of biological relevance. However, ¹³C incorporation increased to much higher levels in wasps with partially depleted reserves of fatty acid derivatives suggesting that parasitic wasps synthesize fatty acids increasingly when they need to, i.e. when their fat reserves are ebbing [13]. Furthermore, it is unlikely that parasitic wasps maintain the costly enzymatic machinery of lipogenesis without benefitting from it at any time of their lives.

One species which was long assumed to lack lipogenesis is Nasonia vitripennis [8,9,16] (figure 1h), a frequently studied model organism for all aspects of parasitic wasp biology [17–19]. In experiments offering N. vitripennis wasps unlimited access to sugar, they did not accumulate total lipids in sufficient amounts detectable by bulk methods such as colorimetry [20] or gravimetry [9]. Gene expression and deuterium-based isotope labelling studies have been interpreted to confirm the inability of N. vitripennis to synthesize fatty acids from sugar [9,16]. Recent studies using fully ¹³C-labelled α-d-glucose, however, revealed that N. vitripennis, like the congeneric species N. giraulti and N. longicornis, incorporate sugar-derived acetyl-CoA into saturated fatty acids, triacylglycerides, as well as into the fatty acid-derived male sex attractant [12,13]. Experiments in these studies were conducted with 10% glucose solutions although natural carbohydrate sources used by parasitoids such as floral and extrafloral nectar as well as honeydew have often higher concentrations [21–24] which might increase lipogenesis. Furthermore, ¹³C incorporation into palmitic acid methyl ester (PAME) and stearic acid methyl ester (SAME) in the previous studies was concluded from the increased abundance of the mass spectrometric diagnostic ion m/z 90. However, this calculation method, while suitable to reliably detect lipogenesis, is rather conservative, because it considers only those fatty acid molecules carrying labelled carbon atoms at positions 1–3 and tends to underestimate the actual degree of fatty acid biosynthesis [13]. Calculation based on the isotope clusters of the molecular ions can be expected to be both higher and more precise, because it considers both fully and partially labelled fatty acid molecules.

Figure 1.

Effect of variable host exposure times on de novo biosynthesis of fatty acids. (a) Scheme of the experimental design (for details see text). (b–e) Calculated ¹³C-incorporation rates into (b) palmitic acid methyl ester (PAME), (c) stearic acid methyl ester (SAME), (d) oleic acid methyl ester (OAME) and (e) linoleic acid methyl ester (LAME) obtained from the raw lipids of N. vitripennis females. Prior to lipid extraction, 2-day-old females were exposed to hosts for different time periods and had subsequently 2-day access to a 20% solution of fully ¹³C-labelled α-d-glucose. For control, females without access to hosts were fed water. (f) Total amounts of de novo synthesized fatty acid methyl esters (FAME) and resulting calculated number of egg equivalents. (g) Total amounts of constitutive FAME. (h) Image of a Nasonia female feeding on a drop of glucose solution. Box-and-whisker plots show median (horizontal line), 25–75% quartiles (box), maximum/minimum range (whiskers) and outliers (° greater than 1.5× upper quartile; * greater than 3× upper quartile). Different lowercase letters indicate significant differences at p < 0.001 within each panel (Kruskal–Wallis test and multiple Dunn's test with sequential Bonferroni correction (n = 10). (Online version in colour.)

De novo fatty acid biosynthesis is detectable by the ¹³C-labelling technique even in newly emerged N. vitripennis females [13] that typically emerge with plenty of lipid reserves [13,20]. ¹³C incorporation, however, increased to higher levels in ageing females that had depleted their lipid reserves suggesting that the degree of de novo fatty acid biosynthesis generally increases in N. vitripennis when the available reserves of fatty acid derivatives decrease [13]. Ageing, foraging and oviposition are factors that might lead to both a rapid fat depletion and an increase of de novo fatty acid biosynthesis in N. vitripennis. Ovipositing N. vitripennis females may lose more than 50% of their lipid reserves within the first 72 h after emergence [13]. However, the relation between the number of offspring and lipid availability, on the one hand, and the degree of lipogenesis, on the other, are unknown. The same is true for the question whether the amounts of fatty acids synthesized by N. vitripennis females de novo are sufficient to produce additional eggs and whether de novo synthesized fatty acids are effectively incorporated into eggs.

In the present study, we investigated the biological significance and factors influencing the degree of fatty acid biosynthesis in N. vitripennis. First, we investigated whether an increase of ¹³C-glucose concentration and the more exact calculation method result in increased ¹³C incorporation. Secondly, we studied whether the duration of egg laying prior to glucose feeding influences the degree of lipogenesis in N. vitripennis females. Thirdly, we investigated whether the amounts of fatty acids produced de novo by females are sufficient to produce additional eggs and whether de novo synthesized fatty acids are incorporated into eggs. Finally, we studied whether regular access to glucose increases both longevity and fecundity of N. vitripennis females.

2. Material and methods

(a) . Insects

Nasonia vitripennis is a cosmopolitan pupal parasitoid of various cyclorrhaphous flies occurring in nests of hole breeding birds and on rotting carcasses. The species is a gregarious ectoparasitoid, i.e. females lay several eggs onto the surface of the fly pupa inside the fly puparium, and hatching larvae feed on the fly pupa until pupation [19,25]. The 2–3 mm-long adult wasps emerge after approximately 14 days (at 25°C) from the host puparium with males occurring slightly before the females. Pheromone-mediated courtship and mating occur typically at the natal patch shortly after emergence of the females. After mating, females readily disperse from the natal patch to search for new oviposition sites [19,25,26]. Nasonia vitripennis is a synovigenic species, i.e. females emerge with no or only a few immature eggs and develop eggs successively throughout their lives [27]. Egg production in synovigenic species is supported by feeding on the haemolymph of their hosts or carbohydrate sources [20,28,29]. Lifetime of N. vitripennis depends strongly on size and the availability of food ranging between 4 and 16 days [20].

Nasonia vitripennis used in this study originated from an inbred strain that was originally collected from birds' nests in northern Germany. Wasps were reared on freeze-killed pupae of the green bottle fly, Lucilia caesar (Diptera, Calliphoridae), as reported earlier [30]. Virgin females were obtained by isolating female pupae from parasitized hosts 2 days before the expected emergence date. Insect rearing and all experiments were done at 25°C, 60% relative humidity and a 16 : 8 h light : dark regime.

(b) . Influence of glucose concentration on fatty acid biosynthesis

Two-day-old, mated females were exposed for 48 h to five hosts for oviposition thereby spending parts of their lipid reserves. Subsequently, females were put in groups of five into 1.5 ml microcentrifuge vials and fed for 48 h with 30 µl of 10% or 20% solutions of fully ¹³C-labelled α-d-glucose (Sigma-Aldrich, Taufkirchen, Germany) (n = 10). Control females were fed water for the same period (n = 10). The liquids were directly pipetted into the vials. After the feeding period, females were killed by freezing and used for fatty acid analysis as described below.

(c) . Influence of variable host exposure times on de novo biosynthesis of fatty acids

To investigate whether the egg-laying duration has an impact on the degree of de novo biosynthesis of fatty acids, we exposed 2-day-old, mated females in Petri dishes for 0, 4, 8, 24 or 48 h to five hosts to enable oviposition (figure 1a). To control for age, females that spent less than 48 h with hosts were kept the rest of the 48 h without hosts. Subsequently, females were fed in groups of five for 48 h with 30 µl of a 20% solution of fully ¹³C-labelled α-d-glucose as described above. After the feeding period, females were killed by freezing used for fatty acid analysis as described below. Parasitized hosts were kept under rearing conditions until the next generation emerged, and the offspring produced by each female was counted.

(d) . Analysis of de novo synthesized fatty acids in N. vitripennis females and eggs

Two-day-old, mated females that had 24 h access to five hosts and were fed for 48 h a 20% solution of ¹³C-labelled α-d-glucose (n = 16) or water (control, n = 16) were allowed to lay eggs a second time after glucose feeding (figure 1a). To facilitate the collection of laid eggs, hosts were put upright in 300 µl conical glass inserts (CZT, Kriftel Germany) making sure that the females had only access to the upper half of the host puparium (electronic supplementary material, figure S1 in the supplementary material). After adding the female, the glass insert was put into a 1.5 ml autosampler vial, the cap of which was pierced with a 2 cm glass capillary to ensure the supply of the female with air. A metal spring inside the vial helped to press the insert against the cap thus preventing the female from escaping from the insert. After the 24 h oviposition period, the females were killed by freezing and used for fatty acid analysis as described below. Parasitized host puparia were opened under a stereo microscope and eggs were carefully transferred to 1.5 ml autosampler vials containing 200 µl of dichloromethane. Eggs laid by each female were counted and pooled for fatty acid analyses as described below. Only those samples were considered for chemical analysis, from which at least five eggs were recovered (n = 11 and 10 for ¹³C treatment and water control, respectively).

(e) . Transesterification of raw lipids for GC/MS analysis

Individual N. vitripennis wasps were homogenized in microcentrifuge vials with a glass pestle after adding 100 µl dichloromethane containing 0.75 µg 2-methyloctadecanoic acid (Sigma-Aldrich) as an internal standard. Extracts were kept at room temperature for 30 min and subsequently transferred to 1.5 ml glass vials. The microcentrifuge vials were rinsed with another 100 µl dichloromethane and combined extracts were dried under a stream of nitrogen. The residues were re-suspended in 200 µl of methanol and 20 µl of acetyl chloride (10%, dissolved in methanol) and transesterified for 2 h at 75°C. Thereafter, 200 µl of an aqueous solution of sodium hydrogen carbonate (5%) was added and FAME were extracted with 200 µl hexane. The hexane phase was concentrated under nitrogen to a volume of 15 µl and used for GC/MS analysis. Nasonia vitripennis eggs were subjected to the same protocol, but only 0.2 µg (1–5 eggs), 0.4 µg (6–10 eggs) or 0.6 µg (more than 10 eggs) of the internal standard was used in this case, and the final volume of the hexane phase was adjusted to ca 3 µl.

(f) . GC/MS analysis and calculation of ¹³C incorporation

Chemical analyses were performed using a Shimadzu QP2010 Plus GC/MS system equipped with a 60 m × 0.25 mm inner diameter BPX5 capillary column (film thickness 0.25 µm, SGE Analytical Science Europe, Milton Keynes, UK). Samples of the wasp extracts (2 µl) were injected splitless at 300°C using a Shimadzu AOC 20i auto sampler. The egg samples were injected completely by manual injection. The MS was operated in the electron impact ionization mode at 70 eV; the mass range was m/z 35–500. Helium was used as carrier gas at a constant velocity of 30 cm s⁻¹. The temperature program started at 150°C, increased at 3°C min⁻¹ to 300°C and was kept at this temperature for 15 min. Identification of FAME was done by comparison of retention times and mass spectra with those of authentic reference chemicals (reference mixture of 37 FAME, Sigma-Aldrich).

For the quantification of individual and total amounts of FAME in the samples, we integrated the peak areas of PAME, SAME, oleic acid methyl ester (OAME) and linoleic acid methyl ester (LAME) in the total ion current chromatograms and related them to the peak area of the internal standard 2-methyloctadecanoic acid methyl ester. These four major compounds made more than 90% of the total FAME in the samples and were thus considered as a proxy for the general fatty acid status of the wasps. We calculated the incorporation of glucose-derived ¹³C using the cluster of fully and partially labelled molecular ions resulting from the varying number of glucose-derived, labelled acetyl-CoA units being incorporated into the fatty acid chain (PAME: unlabelled m/z 270, labelled m/z 274, 276, 278, 280, 282, 284, 286; SAME: unlabelled m/z 298, labelled m/z 302, 304, 306, 308, 310, 312, 314, 316); OAME: unlabelled: m/z 296, labelled m/z 300, 302, 304, 306, 308, 310, 312, 314; LAME: unlabelled: m/z 294, labelled m/z 298, 300, 302, 304, 306, 308, 310, 312 [13]. We did not consider the M + 2 satellites (PAME m/z 272, SAME m/z 300, OAME: m/z 298, LAME: m/z 296) because of the M+2 ion resulting from the naturally occurring ¹⁸O in the ester function interfering with the M + 2 signal caused by the incorporation of one labelled acetyl-CoA unit. For comparison, we calculated ¹³C-incorporation rates of ¹³C-glucose-fed wasps also by using the diagnostic ion pair m/z 87/90 [13]. In the samples of the egg-laying experiment, a weak, non-specific signal at m/z 276 was detected in PAME in control samples due to a co-eluting contaminant (7,9-Di-tert-butyl-1-oxaspiro(4,5)deca-6,9-diene-2,8-dione, molecular mass: 276 amu, identification based on mass spectrum and retention time of an authentic reference compound, Sigma-Aldrich). This led to apparent ¹³C-incorporation rates of 1.3 ± 0.4% (mean ± SE) in the control samples. Therefore, m/z 276 was excluded from the calculation of ¹³C-incorporation rates of PAME in samples of the oviposition experiment leading to slightly underestimated values.

We integrated the ion chromatograms of the diagnostic ions at the retention times of the four FAME and calculated ¹³C-incorporation rates (in %) by relating the sum of peak areas of the isotope cluster ions to the sum of peak areas of isotope cluster ions + unlabelled molecular ion. By applying the estimated individual ¹³C-incorporation rates of PAME, SAME, OAME and LAME to the absolute amounts of the compounds (in µg) determined in the respective female, we calculated the absolute amounts of de novo synthesized fatty acids. By dividing these amounts by the amounts of the total FAME found in individual N. vitripennis eggs (0.13 µg), we calculated the number of FAME egg equivalents synthesized de novo by each female. Amounts of constitutive FAME (i.e. the amounts of FAME not having been synthesized de novo by the wasps) were calculated by subtracting the amounts of de novo FAME from the amounts of total FAME.

(g) . Impact of glucose availability on longevity and fecundity of N. vitripennis females

To study the impact of regular access to glucose on longevity and fecundity, newly emerged, mated N. vitripennis females were exposed for 16 days alternately to a 20% unlabelled glucose solution and to five hosts for oviposition. Females (n = 34) were kept singly in Petri dishes (9 cm diameter) and 20 µl of the glucose solution were applied in four droplets to the lid of the Petri dish on days 1–2, 5–6, 9–10 and 13–14. Control females (n = 34) were provided with water instead. The sugar solution and water, respectively, were checked daily and renewed when necessary. After each glucose feeding phase, females were provided with five hosts on days 3–4, 7–8, 11–12 and 15–16. Petri dishes were washed with water after each feeding phase. Parasitized hosts were removed after each oviposition phase and kept under rearing conditions until the offspring had emerged. Surviving females were counted daily and offspring produced by each wasp in each oviposition phase was determined 18 days after the end of the experiment. Hosts were additionally checked for diapausing larvae which were counted and added to the offspring of the respective females.

(h) . Statistical analyses

Statistical analyses were performed using Past 4.05 scientific software [31]. Data not meeting the assumptions of parametric testing (test for normality by Shapiro–Wilks test) were analysed by non-parametric methods. ¹³C-incorporation rates into FAME from ¹³C and control treatments as well as total amounts of FAME in the samples were compared by Mann–Whitney U-test (two treatments) or by Kruskal–Wallis test followed by multiple Dunn's tests with sequential Bonferroni correction (more than two treatments). ¹³C-incorporation rates into the different FAME of one treatment were compared by a Friedman test followed by multiple Wilcoxon matched pair tests with sequential Bonferroni correction. Egg numbers laid by ¹³C-glucose and water fed females were analysed by a t-test. The correlation between de novo synthesized FAME and offspring number/constitutive FAME was analysed using a Pearson correlation test. Longevity of glucose and water-fed wasps were analysed with a Kaplan–Meier survival analysis and log rank test. Offspring numbers of glucose and water-fed wasps were compared by Mann–Whitney U-test. ¹³C-incorporation rates estimated by the two calculation methods were compared using a Wilcoxon paired test.

3. Results

(a) . Influence of glucose concentration on fatty acid biosynthesis

Females having spent parts of their lipid reserves by oviposition and having had access for 2 days to a 20% ¹³C-labelled α-d-glucose solution showed significantly higher ¹³C-incorporation rates into PAME, SAME, LAME and OAME (mean ± SE: 21.8 ± 3.0%, 31.7 ± 4.0%, 38.6 ± 6.4% and 14.8 ± 2.5%, respectively) than those having fed a 10% solution (mean ± SE: 3.7 ± 1.1%, 3.7 ± 1.5%, 3.4 ± 2.1% and 1.3 ± 0.7%, respectively). As expected, no ¹³C incorporation was detectable in the water-fed control females (Kruskal–Wallis test p < 0.001, Bonferroni-corrected Dunn's test: p < 0.05 for all compounds, figure 2a, table S2a in the electronic supplementary material). In terms of absolute amounts, all females fed ¹³C-glucose synthesized more FAME de novo than the water-fed control females, in which, due to the experimental design, ¹³C-labelled FAME were not detectable. However, females fed 20% ¹³C-glucose synthesized more FAME de novo (mean ± SE: 3.1 ± 0.7 µg representing 24.0 ± 5.5 calculated egg equivalents) than females of the 10% glucose treatment (0.09 ± 0.29 µg FAME, 0.7 ± 0.3 calculated egg equivalents) (Kruskal–Wallis: p < 0.001, Bonferroni-corrected Dunn's test: p < 0.05 for all comparisons, figure 2b, table S2b in the electronic supplementary material).

Figure 2.

Influence of glucose concentration on de novo biosynthesis of fatty acids. (a) Calculated ¹³C-incorporation rates into palmitic acid methyl ester (PAME), stearic acid methyl ester (SAME), oleic acid methyl ester (OAME) and linoleic acid methyl ester (LAME) obtained from the raw lipids of N. vitripennis females. Prior to lipid extraction, females had 2-day access to water (control), or 10%- or 20% solutions of fully ¹³C-labelled α-d-glucose. (b) Total amounts of de novo synthesized fatty acid methyl esters (FAME) and resulting calculated number of egg equivalents. Box-and-whisker plots show median (horizontal line), 25–75% quartiles (box), maximum/minimum range (whiskers), and outliers (° greater than 1.5× upper quartile; * greater than 3× upper quartile). Different lowercase letters indicate significant differences at p < 0.001 within each triple comparison (Kruskal–Wallis test and Bonferroni-corrected multiple Dunn's test (n = 10). (Online version in colour.)

(b) . Influence of variable host exposure times on de novo biosynthesis of fatty acids

¹³C-incorporation rates into PAME, SAME, LAME and OAME from ¹³C-glucose-fed females of equal age but differing previous oviposition history (host exposure time) were significantly higher when compared to FAME from water-fed control females (Kruskal–Wallis: p < 0.001, Bonferroni-corrected Dunn's test: p < 0.05 for all comparisons, figure 1b–e, table S1a–d in the electronic supplementary material). However, the length of the host exposure time had no significant influence on the incorporation rates in any of the four FAME (Bonferroni-corrected Dunn's test: p > 0.05 for all comparisons). Likewise, the absolute amounts of de novo synthesized, ¹³C-labelled FAME were higher in ¹³C-fed females of any treatment than in water fed control females (0 h: 5.0 ± 1.1 µg, 38.7 ± 8.1 calculated egg equivalents; 4 h: 2.9 ± 0.4 µg, 22.4 ± 2.9 calculated egg equivalents; 8 h: 2.1 ± 0.3 µg, 16.1 ± 2.6 calculated egg equivalents, 24 h: 5.2 ± 1.2 µg, 39.6 calculated egg equivalents; 48 h: 2.8 ± 0.5 µg, 21.7 ± 4.1 calculated egg equivalents; Kruskal–Wallis test: p < 0.001, Bonferroni-corrected Dunn's test: p < 0.05 for all comparisons, table S1e in the electronic supplementary material) but did not differ significantly between the different host exposure times (Bonferroni-corrected Dunn's test: p > 0.05 for all comparisons). As for the constitutive FAME, females that had access to ¹³C-glucose but no previous opportunity to lay eggs had higher amounts of constitutive FAME than water-fed control females and ¹³C-glucose-fed females with previous oviposition experience (Kruskal–Wallis test: p < 0.001, figure 1g, table S1f in the electronic supplementary material); however, due to the sequential Bonferroni-correction, these differences became non-significant (Dunn's test: p > 0.05) except for the 8-h host exposure treatment (p = 0.0012) and the water control (p = 0.0266). The number of eggs laid by individual females did not correlate with the amounts of de novo synthesized FAME (figure 3a), but there was a significant correlation between de novo synthesized and constitutive FAME (figure 3b).

Figure 3.

Relationship between the total amount of fatty acid methyl ester (FAME) synthesized de novo by N. vitripennis females and (a) offspring number and (b) the total amount of constitutive (FAME) (Pearson correlation analysis). (Online version in colour.)

(c) . Analysis of de novo synthesized fatty acids in N. vitripennis females and eggs

¹³C-incorporation rates into PAME, SAME, LAME and OAME from ¹³C-glucose-fed females of the oviposition experiment were significantly higher when compared to FAME from water-fed control females (Mann–Whitney U-test: p < 0.001 for all comparisons, figure 4a). GC/MS analysis revealed furthermore that FAME from the eggs laid by ¹³C-glucose-fed females showed the ¹³C-label (figures S2a–b and S3a–b in the electronic supplementary material). ¹³C-incorporation rates into PAME, SAME, LAME and OAME from the eggs of the ¹³C-glucose-fed females were significantly higher when compared to FAME from the eggs of water-fed control females (Mann–Whitney U-test test: p < 0.001 for all comparisons). However, ¹³C-incorporation rates of the egg-derived FAME were significantly lower than in the respective female-derived FAME (Mann–Whitney U-test: p < 0.01) except for SAME (Mann–Whitney U-test: p = 0.0527, figure 4a). ¹³C-incorporation rates into egg-derived SAME (mean ± s.e.: 26.5 ± 1.8%) was significantly higher than for PAME (4.6 ± 0.6%), OAME (6.1 ± 0.8%) and LAME (1.8 ± 0.4%) (Friedman test: p < 0.0001, Bonferroni-corrected Wilcoxon matched pairs test: p < 0.01 for all comparisons). In terms of absolute amounts, ¹³C-glucose-fed females synthesized significantly more ¹³C-labelled FAME de novo than the water-fed control females (Mann–Whitney U-test: p < 0.001), in which, due to the experimental design, ¹³C-labelled FAME were not detectable also in this experiment, while the total amounts of constitutive FAME were not significantly different (Mann–Whitney U-test: p = 0.8503, figure 4b). ¹³C-glucose-fed females laid significantly more eggs during the 24 h oviposition period (mean ± s.e.: 10.2 ± 1.7) than water-fed control females (6.1 ± 0.7, t-test: p = 0.0324).

Figure 4.

Nasonia vitripennis eggs contain de novo synthesized fatty acids. (a) Calculated ¹³C-incorporation rates into palmitic acid methyl ester (PAME), stearic acid methyl ester (SAME), oleic acid methyl ester (OAME) and linoleic acid methyl ester (LAME) obtained from the raw lipids of N. vitripennis females and eggs laid by the respective females. Prior to lipid extraction, 2-day-old females were exposed to hosts for 24 h and had subsequently 2-day access to a 20% solution of fully ¹³C-labelled α-d-glucose (n = 11). Control females were fed water instead (n = 10). (b) Total amounts of de novo synthesized and constitutive fatty acid methyl esters (FAME) and egg number laid within 24 h by the ¹³C-glucose-fed and water-fed females (n = 16). Box-and-whisker plots show median (horizontal line), 25–75% quartiles (box), maximum/minimum range (whiskers) and outliers (° greater than 1.5× upper quartile; * greater than 3× upper quartile). Statistical analysis by Mann–Whitney U-test and t-test (egg number), respectively. (Online version in colour.)

(d) . Impact of glucose availability on longevity and fecundity of N. vitripennis females

Females that had alternating access to a 20% glucose solution and hosts for oviposition had a significantly higher survival probability than control females having access to water and hosts (log rank test: χ² = 77.621, p < 0.0001, figure 5a). Of the females able to consume glucose between oviposition periods, 87.5% survived the 16-day observation period, while all control females died. Offspring number did not differ during the first oviposition period (glucose treatment: 43.4 ± 2.5, water treatment: 45.8 ± 2.4; Mann–Whitney U-test p = 0.5101, however, glucose-fed females laid significantly more eggs than the surviving water-fed control females in oviposition period 2 (glucose treatment: 54.5 ± 2.5, water treatment: 31.9 ± 3.2; Mann–Whitney U-test p < 0.0001) and during the total experiment (glucose treatment: 190.4 ± 8.3, water treatment: 56.1 ± 4.8; Mann–Whitney U-test p < 0.0001, figure 5b).

Figure 5.

The availability of glucose increases longevity and fecundity in N. vitripennis females. (a) Kaplan–Meier survival plot and (b) offspring number of N. vitripennis females having alternating access to a 20% solution of glucose or water (H₂O, control) (on days 1–2, 5–6, 9–10 and 13–14) or to five hosts for oviposition (on days 3–4, 7–8, 11–12 and 15–16) during a 16-day observation period. Box-and-whisker plots show median (horizontal line), 25–75% quartiles (box), maximum/minimum range (whiskers) and outliers (° greater than 1.5× upper quartile; * greater than 3× upper quartile). Statistical analyses in (a) by log-rank test and in (b) by Mann–Whitney U-test (number of replicates is given below each box). Data for days 11–16 were not compared statistically due to the low sample number in the control treatment. (Online version in colour.)

(e) . Comparison of the calculation methods for the estimation of ¹³C incorporation

Calculated ¹³C-incorporation rates into PAME and SAME based on the molecular ion cluster were significantly higher than those based on the diagnostic ion m/z 90 (Wilcoxon paired test: p < 0.0001 for both PAME and SAME, figure S4a,b in the electronic supplementary material). Mean incorporation rates for PAME and SAME across all female-derived samples increased from 20.0 ± 1.6% and 34.1 ± 2.4% to 23.0 ± 1.7% and 38.1 ± 2.5%, respectively. All data of this paper are available from the Dryad Digital Repository [32].

4. Discussion

Nasonia vitripennis has long been considered a prime example for the purported absence of lipogenesis in parasitic wasps because wasps emerging with full lipid stores did not gain lipid mass despite ad libitum access to sugar [9,16,20]. By contrast in the present study, we demonstrate that ageing N. vitripennis females with partially depleted lipid reserves due to egg laying accumulate biologically relevant amounts of de novo synthesized fatty acids (figure 1f). Nasonia vitripennis females typically mate immediately after emergence from the host at the natal patch [26]. Mating comes along with an increased locomotor activity [33], and mated females typically disperse within 2 h from their natal patch to search for oviposition sites [26,34]. Ageing and oviposition causes a decline of lipids [12,20] which may be partially compensated by host feeding (the consumption of host-derived haemolymph [28]), but also by synthesizing fatty acids de novo when feeding natural carbohydrate sources such as floral and extrafloral nectar or honeydew [35–37]. We mimicked this scenario in our experiments by enabling N. vitripennis females to lay eggs for different time periods prior to offering them ¹³C-labelled glucose. We demonstrate that females synthesize significant amounts of PAME, SAME, OAME and LAME within a feeding period of 48 h. We found ¹³C-incorporation rates of 30 ± 2%, 50 ± 2%, 49 ± 3% and 21 ± 2% (mean ± SE across treatments) for PAME, SAME, OAME and LAME, respectively, in the remaining fatty acid derivatives of individual N. vitripennis females. The total amounts of fatty acids synthesized de novo by N. vitripennis females were in the same range as the remaining constitutive fatty acids in water-fed control females (figure 1f–g). These de novo synthesized fatty acids can be used by N. vitripennis females as energy source in the future and/or to supply a considerable number of additional eggs with lipids. In fact, our labelling experiments revealed that de novo synthesized fatty acids are incorporated into the eggs of ¹³C-glucose-fed females even though the incorporation rates, except for SAME, were lower than in the body fat of the respective females. Incorporation rates in body fat and eggs, however, are likely to converge when lipid reserves decline further with increasing age and egg-laying. Glucose-fed females produced more eggs than water-fed control females in our ¹³C-labelling experiment (figure 4b) and produced more offspring in the longevity–fecundity experiment (figure 5b). Increased longevity and fecundity in sugar-fed females are often correlated in studies with parasitic wasps [38,39]. The increased longevity widely explains the increased total offspring number of glucose-fed females also in our study. Glucose-fed females lived much longer than water-fed control females thus confirming previous results [20]. A more detailed analysis of the offspring numbers produced within the different oviposition periods of the longevity-fecundity experiment, however, suggests that de novo synthesized fatty acids contributed also to the higher reproductive success of glucose-fed females. While glucose-fed and water-fed females produced equal offspring numbers in oviposition phase 1 (day 3–4), glucose-fed females produced almost twice as many offspring than the surviving water-fed control females in oviposition phase 2 (day 7–8, figure 5b). The reduced offspring number in starved females in oviposition phase 2 can be partly explained by an increased oosorption, i.e. the resorption of developing oocytes under unfavourable environmental conditions [38]. Lipids make up to 40% of the dry mass of insect eggs, being the most important energy source for the developing embryo [40]. The capacity of oocytes to synthesize lipids themselves is thought to be limited. Hence, the bulk of oocyte lipids has to be imported from the female's lipid reserves by circulating lipoproteins [40]. Considering our findings that glucose-fed females synthesize microgram amounts of fatty acids de novo and incorporate them into their eggs, it is reasonable to assume that also de novo synthesis of fatty acids contributed to higher offspring number in glucose-fed N. vitripennis females.

The offspring number produced by individual females did not correlate with the amounts of de novo produced fatty acids, i.e. females with partially depleted fat reserves generally converted glucose into fatty acids, but the degree of lipogenesis did not depend on the offspring number produced prior to sugar feeding (figure 3a). This suggests that it is predominantly the amount of sugar consumed by individual wasps that controls the degree of lipogenesis and that there is a widely stable rate of consumed glucose being converted to lipids by N. vitripennis females. Females with access to ¹³C- labelled glucose had significantly more constitutive fat than water-fed control females (figure 1g), and there was a significant correlation between the amounts of de novo synthesized and constitutive fatty acids (figure 3b). This demonstrates that glucose feeding influences the lipid status of N. vitripennis females in two ways: First, it enables partial replenishment of fat reserves by lipogenesis. Secondly, consumption of glucose provides energy that would otherwise have to come from fatty acid catabolism, thus leading to a decelerated depletion of constitutive fatty acid reserves.

The use of a realistic glucose concentration of 20% for the ¹³C-feeding experiments instead of 10% used in our previous studies [12,13] resulted in much higher ¹³C-incorporation rates into FAME than previously reported for N. vitripennis [12,13]. This increase was much higher than twofold and thus cannot be explained by the higher glucose concentration alone (figure 2). It is likely that the higher sugar concentration stimulated the females to consume higher amounts, as has been previously demonstrated in other parasitic wasps [41,42] as well as in bumblebees [43], orchid bees [44], blowflies [45] and ants [46]. Additionally, there might be physical constraints preventing the wasps from compensating concentration effects by ingesting larger volumes of the consumed sugar solution. Given that total sugar concentrations in floral and extrafloral nectar can reach values higher than 20% [21–24], experiments involving such concentrations might result in even higher degrees of lipogenesis than reported here.

Recent studies involving 16 additional parasitic wasp species have demonstrated the general ability to produce fatty acids de novo from sugar disproving the claimed ‘lack of lipogenesis’ in these insects [12,13]. These studies were intended as a proof of principle and used 10% sugar solutions and a conservative calculation method for determining ¹³C-incorporation rates into FAME. Thus, the reported ¹³C-incorporation rates were relatively low (≤6%) raising the question of their biological relevance. In the present study, we demonstrate that the use of higher sugar concentration and the molecular ion cluster for the mass spectrometric calculation of ¹³C-incorporation rates leads to significantly higher values. The differences between the two calculation methods, however, were relatively moderate, likely due to the high incorporation rates found in this study. Natural carbohydrate sources such as nectar and honeydew contain typically not only glucose but also other mono-, di- and oligosaccharides which can have differing impact on parasitoid life-history parameters [39,47]. Whether these sugars can be equally well converted to fatty acids is unknown. Hence, future experiments are needed involving partially lipid-depleted specimens of additional parasitic wasp species, field-relevant sugar concentrations and compositions of different naturally occurring saccharides, as well as applying the precise calculation of ¹³C incorporation reported here. Such studies are likely to show that parasitic wasps are not so much different from other insects with respect to lipogenesis than previously thought [8,9,16,48]: They simply make use of de novo biosynthesis of fatty acids when they need to.

Data accessibility

All relevant data of this paper are available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.9kd51c5kf [49].

Authors' contributions

M.-T.M.: formal analysis, investigation, writing—review and editing; M.W.: formal analysis, investigation, writing—review and editing; J.R.: conceptualization, formal analysis, investigation, methodology, supervision, visualization, writing—original draft.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

We received no funding for this study.