We asked whether the mating of male and female fruit flies would be affected by the presence of parasitoid wasps. We placed a pair of D. melanogaster flies in a small Petri dish, either with or without parasitoid wasps (Fig. 1a). In an initial experiment we used the wasp Leptopilina boulardi, which specializes on D. melanogaster and on closely related fly species14.
a Courtship arena containing a male and virgin female fly with (left) and without (right) two wasps, one male and one female. b Copulation latency of D. melanogaster. p < 0.0001, two-tailed Mann–Whitney test, n = 105 for +Lb, i.e., exposed to wasps, n = 102 for unexposed, error bars = SEM. The experiment was allowed to run for 1 h. c Copulation latency of other Drosophila species while in arena. Four wasps, two males and two females, were used for exposure. Two-tailed Mann–Whitney test. D. simulans, p = 0.0035, n = 67 (exposed), n = 63 (unexposed); D. yakuba, p = 0.0001, n = 62, 63; D. biarmipes, p < 0.0001, n = 62, 62; D. willistoni, p = 0.0009, n = 93, 92. Error bars = SEM. d Competition paradigm. In the left vial, fluorescently labeled flies (green) are incubated with wasps; in the right vial, unlabeled flies are incubated without wasps; in the arena below, a labeled exposed fly and an unlabeled unexposed fly of the same sex are allowed to compete for a fly of the opposite sex. d′ Competition paradigm. In the left vial, unlabeled flies are incubated with wasps; in the right vial, labeled flies are incubated without wasps; in the arena below, an unlabeled exposed fly and a labeled unexposed fly of the same sex are allowed to compete for a fly of the opposite sex. e Left, competition between a male exposed to L. boulardi and an unexposed male for an unexposed female; dark shading indicates the percentage of competitions won by the exposed male (p = 0.843, chi-squared test, n = 102). Right, a female exposed to L. boulardi and an unexposed female are placed in an arena with an unexposed male; dark shading indicates the percentage of pairings in which the exposed female copulated (p = 0.0038, chi-squared test, n = 115). f Experiments comparing the mating success of an exposed and unexposed female of other species, placed with a male of the same species. Chi-squared test. D. simulans (p = 0.0129, n = 102); D. yakuba (p = 0.0350, n = 119); D. biarmipes (p = 0.0001, n = 84); D. willistoni (p = 0.0057, n = 119).
We expected that the presence of the parasitoids, a lethal threat to reproductive success, would delay mating. We were surprised to find exactly the opposite. In the presence of the wasps, the mean time to copulation was 16 min, compared to 25 min in the absence of the wasp, i.e. a 36% reduction in copulation latency (Fig. 1b, ****p < 0.0001, two-tailed Mann–Whitney test, n = 105, 102). The exposure to wasps did not reduce the fraction of pairs that copulated in this experiment: copulation occurred in 80% of pairs that were not exposed to wasps and in 86% that were exposed (p > 0.05, chi-squared test). We then tested four additional Drosophila species closely related to D. melanogaster and found the same acceleration to copulation in all four (Fig. 1c).
Exposure to wasps could affect the behavior of either the male fly or the female fly. We tested these possibilities through competition experiments. If exposure increased male motivation, we might expect an exposed male to outcompete an unexposed male when both males were placed in a chamber with a single unexposed female. We placed fluorescently labeled male flies in a chamber with wasps for 2 h, and in parallel placed unlabeled male flies in a separate chamber without wasps (Fig. 1d). We then allowed a fluorescent exposed male to compete with an unexposed male for the same female in a small arena. To control for any effect of the fluorescent label, we also performed the reciprocal experiment, i.e. we exposed unlabeled males to wasps and allowed them to compete with unexposed, labeled males (Fig. 1d′). The results of the two reciprocal experiments were scored blind and averaged. We found that exposed males and unexposed males were equally successful in mating competitions (Fig. 1e, left).
By contrast, when we placed an exposed female and an unexposed female together with an unexposed male, in 63% of tests the exposed female was the one that copulated (Fig. 1e, right; **p < 0.01, chi-squared test, n = 102, 115). These results suggested the hypothesis that exposure of females to wasps increases their receptivity to males. We tested this hypothesis by pre-exposing females of the other four Drosophila species to wasps. In all four cases, pre-exposed females were more likely to mate than non-exposed females (Fig. 1f).
We next asked whether exposure to other species of parasitoid wasps also accelerated mating behavior. We first tested Leptopilina heterotoma, which is closely related to L. boulardi, and subsequently Trichopria drosophilae and Asobara tabida; all of these species have been shown previously to attack Drosophila melanogaster larvae or pupae15. When wasps of any of these three species were placed in an arena with a pair of D. melanogaster flies, copulation began sooner (Fig. 2a–c). As a further test, a female fly that was pre-exposed to each wasp species was more likely to copulate than a non-exposed female fly when both were placed in an arena with a male (Fig. 2d–f).
a Leptopilina heterotoma, n = 121 (exposed), 100 (unexposed). p = 0.0044, two-tailed Mann–Whitney test. Error bars = SEM. b Trichopria drosophilae, n = 67 (exposed), 67 (unexposed). p < 0.0001. c Asobara tabida, n = 81 (exposed), 85 (unexposed), p = 0.0051. g Drosophila suzukii, n = 121 (exposed), 119 (unexposed), p = 0.9592. h Muscidiflurax zaraptor, n = 101 (exposed), 101 (unexposed), p = 0.9568. Error bars in b, c, g, h are SEM. The statistical test used in b, c, g, h is a two-tailed Mann–Whitney test. In each case a male D. melanogaster and a virgin female D. melanogaster were placed in an arena with one male and one female of the indicated species. d–f, i, j A female D. melanogaster exposed to the indicated species and an unexposed female D. melanogaster were both placed in an arena with a male D. melanogaster as in Fig. 1d, d′. Dark shading indicates the percentage of cases in which the exposed female mated. d L. heterotoma p = 0.0352, chi-squared test, n = 100. e T. drosophilae p = 0.0018, n = 100. f A. tabida p = 0.0464, n = 91. i D. suzukii p = 0.9196, n = 98. j M. zaraptor p = 0.5546, n = 103. The statistical test used in e, f, i, j is a chi-squared test.
Does the presence of any other insects whatsoever accelerate Drosophila mating? The presence of another species of Drosophila, D. suzukii16, did not have any of these effects on the mating of a pair of D. melanogaster, and neither did the presence of Muscidiflurax zaraptor17, a parasitoid that deposits eggs in larger flies (Fig. 2g–j).
We hypothesized that the wasps that affected mating produced an odor that was not emitted by the other tested species, and that this odor activated a circuit that accelerated sexual behavior. Such an odor might be detected by Or49a, Or85f, or perhaps another receptor of the Or family18. If an Or were essential to the acceleration in copulation onset, then the copulation latency of flies mutant for the obligate Or co-receptor Orco19 would be expected to be unaffected by exposure to wasps. However, we found that Orco mutant flies showed an acceleration in mating onset comparable to that of control flies following exposure to L. boulardi; the copulation latency was reduced by 38% (Fig. 3a; compare to Fig. 1b). Moreover, competition experiments again provided evidence of an effect on females but not males, as in control flies (Fig. 3b). Similar results were obtained when L. heterotoma was used as the wasp (Supplementary Fig. 1a, b).
a Mating of an Orco1 mutant is affected by exposure to L. boulardi. p = 0.0001, two-tailed Mann–Whitney test, n = 79 (exposed), 76 (unexposed). Error bars = SEM. b Left, competition experiments between an exposed Orco1 male and an unexposed Orco1 male, for an Orco1 female (p = 0.9042, chi-squared test, n = 69). Right, an exposed Orco1 female and an unexposed Orco1 female are placed in an arena with an Orco1 male (p = 0.0389, chi-squared test, n = 60). Dark shading indicates the percentage of cases in which the exposed animal mated. c An Ir8a; Orco1 double mutant is affected by exposure to L. boulardi. n = 66 (exposed), 67 (unexposed), p = 0.0085. d Left, competition experiments between an exposed and unexposed IR8a; Orco1 male, for an IR8a; Orco1 female (p = 0.8974, chi-squared test, n = 60). Right, an exposed and unexposed IR8a; Orco1 female are placed in an arena with an IR8a; Orco1 male (p = 0.0223, chi-squared test, n = 62). e A ninaB1 mutant is not affected by exposure to L. boulardi under room light, n = 67 (exposed), 66 (unexposed), p = 0.7904. We note that this mutant is slow to mate. f ninaB1 is not affected by exposure to L. heterotoma under room light, n = 69 (exposed), 68 (unexposed), p = 0.9598. 2 male and 2 female wasps were used in e–j. g Our Canton-S wild type strain is not affected by exposure to L. boulardi under dim red light n = 74 (exposed), 74 (unexposed), p = 0.195. h Canton-S is not affected by exposure to L. heterotoma under dim red light n = 74 (exposed), 74 (unexposed), p = 0.6537. i Canton-S is affected by exposure to L. boulardi under green light, n = 68 (exposed), 68 (unexposed), p < 0.0001. j Our Canton–S wild-type strain is affected by exposure to L. heterotoma under green light, n = 64 (exposed), n = 65 (unexposed), p = 0.0005. k Exposure to L. boulardi affects copulation latency of parental control TNT: n = 63 (exposed), 63 (unexposed), p = 0.0001. l Exposure to L. boulardi affects copulation latency of parental control LC4: n = 82 (exposed), 79 (unexposed), p = 0.0076. m–p Flies in which LC4 neurons are blocked with TNT are not affected by exposure to the indicated four species of wasps: in m (L. boulardi), n = 84 (exposed), 78 (unexposed), p = 0.6211; in n (L. heterotoma), n = 80 (exposed), 78 (unexposed), p = 0.6289; in o (Trichopria drosophilae), n = 90 (exposed), 81 (unexposed), p = 0.4496; in p (Asobara tabida), n = 85 (exposed), 81 (unexposed), p = 0.643. Error bars in c, e–p are SEM. The statistical test used in c, e–p is a two-tailed Mann–Whitney test.
Could the effect be mediated by an odor that activates an ionotropic receptor (IR) odor receptor20? We tested flies doubly mutant for the Orco and IR8a co-receptors; IR8a is essential for many IR-mediated responses21. These double mutants again showed the same mating acceleration phenotypes, for both L. boulardi and L. heterotoma, and again pre-exposure of females but not males had an effect in competition experiments (Fig. 3c, d and Supplementary Fig. 1c, d). Taken together, these results suggested that olfactory cues do not drive the acceleration of mating.
We next tested the hypothesis that the mating effect depended on visual cues. Consistent with this hypothesis, we found that a visually impaired mutant that is defective in the synthesis of visual pigments, ninaB1 (ref. 22), showed equivalent copulation latency in the presence and absence of L. boulardi (Fig. 3e); the same result was found with L. heterotoma (Fig. 3f). Likewise, when the assay was conducted in dim red light, to which Drosophila has very low sensitivity, copulation latency of wild type flies was not affected by the presence of either species of wasps (Fig. 3g, h). As a control, we conducted the assay in green light (~500–565 nm), to which sensitivity is much higher, and we observed effects on mating (Fig. 3i, j). The simplest interpretation of these results is that copulation dynamics is affected by visual cues emanating from wasps. Color cues are unlikely to be essential, however, since they should be minimal under green light.
To test further the hypothesis that the effect of wasps on fly mating behavior is mediated by vision, we asked whether blockage of certain VPNs affected mating behavior. VPNs transmit information from the primary visual center of the brain, the optic lobe, to higher brain regions. The most numerous VPNs in the lobula are lobular columnar (LC) neurons, which fall into multiple types based on their anatomy23,24. Activation of different LC types with split-GAL4 lines evokes different behaviors25. Silencing of one class of LC neurons, LC4, has previously been shown to reduce the escape response of Drosophila to predators26,27.
We used split-GAL4 lines to block LC4, and found that this blockage eliminated the effect of wasps on copulation latency. Parental control flies expressing a UAS-TNT transgene alone (“TNT”, Tetanus Toxin Light Chain, which cleaves synaptobrevin to block synaptic transmission), or split-GAL4 constructs alone (“LC4”), showed the expected reduction in copulation latency when exposed to L. boulardi (Fig. 3k, l). However, flies in which LC4 neurons were blocked (“LC4, TNT”) did not show a reduction in copulation latency, when exposed to L. boulardi or any of three other tested species, L. heterotoma, Trichopria drosophilae or Asobara tabida (Fig. 3m–p).
Visual cues were found in an earlier study to drive the depression in oviposition that follows exposure of flies to wasps28. In that study, visual cues from wasps in one transparent chamber were shown to affect the behavior of flies in an adjacent chamber. We found that cues from wasps could also affect sexual behavior of flies that were in visual contact with them from another chamber; copulation latency was accelerated (Fig. 4a). In this initial experiment male flies and female flies were exposed to wasps separately and then allowed to mate. We also found that female flies exposed in this manner showed accelerated copulation latency when allowed to mate with unexposed male flies (Fig. 4b), consistent with our earlier findings that wasps affected female behavior (Fig. 1e, f).
a Copulation latency of a male fly and a female fly that were each exposed separately to the sight of L. boulardi wasps for 2 h and then allowed to mate (+Lb, n = 83), compared to mating of an unexposed male and an unexposed female (−Lb, n = 82). p < 0.0001, two-tailed Mann–Whitney test. Error bars = SEM. b Mating of a female fly that was exposed to wasps and an unexposed male (+Lb, n = 71) compared to mating of unexposed flies (−Lb, n = 69), p = 0.0003. c Mating of a female fly that was exposed to female L. boulardi wasps exclusively, and an unexposed male fly (+Lb, n = 67) compared to mating of unexposed flies (−Lb, n = 67), p = 0.0094. d Mating of a female fly that was in visual contact with male L. boulardi wasps exclusively, and an unexposed male fly (+Lb, n = 63) compared to mating of unexposed flies (−Lb, n = 63), p = 0.0017. e Mating of a female fly that was in visual contact with female L. heterotoma wasps exclusively, and an unexposed male fly (+Lh, n = 65) compared to mating of unexposed flies (−Lh, n = 63), p < 0.0001. f Mating of a female fly that was exposed to male L. heterotoma wasps exclusively, and an unexposed male fly (+Lh n = 63) compared to mating of unexposed flies (−Lh, n = 62), p < 0.0001. g Mating of a female fly that was in visual contact with female T. drosophilae wasps exclusively, and an unexposed male fly (+Td, n = 63) compared to mating of unexposed flies (−Td, n = 62), p = 0.0058. h Mating of a female fly that was in visual contact with male T. drosophilae wasps exclusively, and an unexposed male fly (+Td, n = 67) compared to mating of unexposed flies (−Td, n = 65), p = 0.0014. i Mating of a female fly that had been in visual contact with wasps (male and female) for 2 h and then isolated for 2 h before pairing with a male fly (n = 69) compared with mating of an unexposed female (n = 69), p = 0.7669. j Mating of a female that had been in visual contact with wasps for 24 h and allowed to mate with an unexposed male (n = 68) compared with mating of an unexposed female (n = 68), p = 0.0038. Error bars in b–j are SEM. The statistical test used in b–j is a two-tailed Mann–Whitney test.
Since only female wasps are a threat to flies, we hypothesized that only female wasps would affect the mating behavior of a female fly. Female L. boulardi have much shorter antennae than males. However, we found that visual contact with either female or male L. boulardi wasps affected copulation latency (Fig. 4c, d). We also tested two other wasp species, L. heterotoma and Trichopria drosophilae, in which the antennae are also shorter in females than males, and found the same results (Fig. 4e–h).
Is the effect on the female flies permanent? We allowed female flies to be in visual contact with wasps for 2 h, and then kept them apart from wasps for 2 h before testing them with males. We found that the effect was not permanent: there was no difference between the mating behavior of exposed female flies that had been away from wasps for 2 h and unexposed female flies (Fig. 4i).
We hypothesized that if we increased the time of exposure by an order of magnitude, that female flies would adapt to the cues and their mating would no longer be affected. However, we found that if female flies were in visual contact with L. boulardi for 24 h and then tested, mating was accelerated. In fact, the copulation latency was reduced severely, by 46% (Fig. 4j).
We wondered if gene expression was affected by exposure to wasps in this paradigm. As an initial screen, we carried out a differential RNA-seq analysis of the heads of female flies that had been exposed to L. boulardi for 2 h and heads of unexposed females. We identified 10 genes whose expression level was increased significantly (adjusted p value < 0.05) in each of two differential expression platforms (Fig. 5a and Supplementary Fig. 2). Among these 10 genes, the one that showed the greatest increase was IBIN (Induced by Infection, a gene that encodes a micropeptide); its expression increased by ~20× (the log2-fold change was 4.2 by CuffDiff and 4.6 by DESeq2). IBIN was not associated with a gene ontology (GO) term, but the other nine genes are associated with the GO terms “immune system” and/or “response to stress.” Six of these other genes encode canonical antimicrobial peptides. No gene showed a decrease in expression level by our statistical criteria.
a Head transcriptomes of flies that have been in visual contact with L. boulardi for 2 h, and of control flies. The transcriptome is based on two biological replicates. FPKM fragments per kilobase of transcript per million mapped reads. b RT-qPCR analysis of IBIN. Exposure to L. boulardi or M. zaraptor was for 2 h. p = 0.046, 0.07, 0.0008, 0.019 (from left to right), **p < 0.01, Mann–Whitney U test, two-tailed, n = 10. Error bars = SEM. c RT-PCR amplification from the indicated tissues. The eye tissue includes both retina and lamina. RT reverse transcriptase. The control gene is eIF3c. RT(−) refers to a negative control where no reverse transcriptase is added to ensure that genomic DNA is not being amplified. n = 2. d RT-qPCR analysis of IBIN from the indicated dissected tissues of flies that had been in visual contact with L. boulardi for 2 hv. from unexposed control flies. n = 3. Error bars = SEM. e RNA-seq reads (y-axis = 0–8000 reads), n = 2. SignalP (version 5.0) was used to predict the presence of signaling peptide (SP in figure) and the cleavage site, and DeepLoc (version 1.0) was used to predict its extracellular localization. f Generation of a mutant IBIN1 allele. Green indicates core promoter GAL4 sequences; red indicates 3XP 3-DsRed sequences. g, h Copulation latency of IBIN1 mutant flies and wCS control flies that had been paired with wasps (L. boulardi). Two-tailed Mann–Whitney test, n = 67 in all cases. Error bars = SEM. In g, p < 0.0001 (left and right). In h, p = 0.377 (left), p = 0.1718 (right).
We attempted to confirm and extend this result with quantitative real-time reverse transcriptase PCR (RT-qPCR). We again found an increase of ~20× in the expression of IBIN in the heads of females exposed for 2 h to L. boulardi relative to heads of unexposed females (Fig. 5b, left bar). Exposure to the parasitoid wasp M. zaraptor did not increase the level of IBIN (Fig. 5b), just as exposure to this wasp did not affect mating behavior (Fig. 2h, j).
Blockage of LC4 neurons with UAS-TNT (“LC4 > TNT”) also eliminated the increase of IBIN induced by exposure to L. boulardi (Fig. 5b), consistent with its inhibition of the behavioral effect (Fig. 3m). As expected, parental control lines carrying UAS-TNT or split-GAL4 constructs alone showed increases following L. boulardi exposure, but not M. zaraptor exposure (Fig. 5b).
We confirmed these results in an independent experiment with different sources of tissue and a different method. We dissected eyes, optic lobes, and brain tissue from which optic lobes had been removed, and performed RT-PCR. An IBIN amplification product was easily visible in all three tissue preparations in flies that had been exposed to L. boulardi (Fig. 5c and Supplementary Fig. 3). The amplification product appeared upregulated compared to unexposed tissues in this initial PCR experiment. Upregulation of IBIN was not observed in LC4 > TNT flies (Supplementary Fig. 3). We then confirmed by RT-qPCR that IBIN was upregulated in all three of these tissues (Fig. 5d; note log scale). The simplest interpretation of these results is that the micropeptide is upregulated in nervous tissue following exposure to L. boulardi.
Remarkably, IBIN was recently shown to be induced in second-instar Drosophila larvae 48 h after infection by L. boulardi29. IBIN was also the gene most highly induced by infection of Drosophila adults with the bacterium Micrococcus luteus29. The gene is induced in immune-responsive tissues, and it can be activated by either the Toll or immune deficiency (Imd) pathways, which is unusual. Its overexpression leads to increased survival after bacterial infection, although the mechanism is unclear. Adult overexpression of IBIN also leads to elevation of sugar levels in the hemolymph, and of Hsp70Bb, a stress-responsive gene.
An environmental stressor was also recently found to upregulate IBIN in males29. Four days of social isolation, a stressor that modulates behavior widely across animal phylogeny, upregulated IBIN in male heads.
The IBIN gene was originally annotated as a long non-coding RNA and referred to as CR44404 or lincRNA-IBIN. However, it contains an open reading frame that is predicted to encode a micropeptide of 41 amino acids (Fig. 5e). This micropeptide is conserved among a variety of Drosophila species (Supplementary Fig. 4). It is predicted to have a signal peptide that is cleaved so as to generate an extracellular micropeptide of ~23 amino acids.
To test directly whether IBIN is required for the effect of wasp exposure on fly mating behavior, we generated a deletion mutation of IBIN by CRISPR/Cas9 genome editing (Fig. 5f). We backcrossed the IBIN deletion allele for five generations against our control line to minimize the possibility of genetic background effects. We found that while the control line shows the expected acceleration in copulation latency with L. boulardi and L. heterotoma, the IBIN mutant did not (Fig. 5g, h). The simplest interpretation of these results is that the micropeptide is required for the acceleration of mating behavior.
Source: Ecology - nature.com
