Search performance and octopamine neuronal signaling mediate parasitoid induced changes in Drosophila oviposition behavior

Diverse oviposition rates of Drosophila females after long exposure to wasps

To investigate whether D. melanogaster change oviposition behavior when they cohabit with Lb female wasps, we designed an experimental procedure and monitored egg laying for a much longer time than in previous experiments – approximately 20 days. Specifically, twenty 3-day-old female and five 3-day-old male D. melanogaster adults were placed in standard fly bottles containing fly food dishes. Flies were housed with twenty 2-day-old Lb female wasps (exposed) or without any female wasps (unexposed). The fly food dishes were replaced daily, and fly eggs were counted daily (Fig. 1a). Consistent with previous observations²⁴, the exposed Drosophila females had significantly reduced oviposition numbers compared to the unexposed flies (Fig. 1b). This response lasted approximately 6 days in the presence of Lb females. After that, we surprisingly found that the number of eggs laid by the exposed flies did not differ from the numbers laid by the unexposed controls (Fig. 1b). This variation led us to speculate that this decreased oviposition may have been induced by the diverse life-threatening pressure when D. melanogaster females encounter different aged wasps, as old ones present less danger to their offspring^28,29, or simply indicate that the flies become habituated to the constant presence of wasps.

To differentiate the two hypotheses, we performed another oviposition behavioral assay by adding the young and old wasps, respectively. Based on their effects on Drosophila egg-laying performance (Fig. 1b), the Lb females ranging from newly eclosed to 8 days old were considered as young wasps, while Lb females older than 8 days were considered as old wasps. In this assay, both flies and wasps were anesthetized by ice for a short time, and then, the Lb females were replaced by young (4-day-old) or old (12-day-old) female wasps on Day 7. The oviposition numbers were monitored for another 7 days. As expected, the exposed Drosophila females reduced their oviposition rate during approximately the first 6 or 7 days in the presence of female wasps (Fig. 1c, d). Interestingly, the Drosophila females in groups re-exposed to young wasps after Day 7 significantly responded with a reduction in the oviposition rate that lasted an extra 5 days (Fig. 1c). When flies were re-exposed to 12-day-old wasps, oviposition remained equivalent to that of the unexposed flies (Fig. 1d). Thus, these results indicate that D. melanogaster females are able to distinguish young and old wasps and reduce oviposition only in the presence of young parasitoids but not due to habituation.

To further check whether the housing experience with Lb females might contribute to the flies’ oviposition behavior, the 3-day-old naive flies were independently housed with different aged Lb females (Supplementary Fig. 1a). As expected, we found that the female flies, without any prior experience with Lb females, significantly decreased egg laying when they were exposed to young but not old wasps (Supplementary Fig. 1b). Moreover, the levels of oviposition reduction were much similar to flies that were continuously exposed to young Lb wasps (Supplementary Fig. 1c). These results suggest that the reduced oviposition behavior of exposed flies certainly requires no learning experiences.

Drosophila reduces egg laying in the presence of young parasitoids

We next investigated whether exposure to other species of Drosophila wasps at a young age also impaired the egg-laying rate. We first tested Lh, which is closely related to Lb, and subsequently Asobara japonica (Aj); all of these species have been shown previously to attack D. melanogaster 2nd instar larvae. Similar to previous processes, 2-day-old female wasps of Lh or Aj were placed in fly bottles with twenty 3-day-old females and five 3-day-old male D. melanogaster adults, and the numbers of Drosophila eggs laid were continuously monitored for 12 days. We found that the fly females laid fewer eggs during approximately the first 8 days in the presence of Lh or Aj female wasps, and the effects were similar to those following exposure to young female Lb (Supplementary Fig. 2a, b). However, when flies were exposed to old Lh or Aj wasps, oviposition numbers were equivalent to that of the unexposed flies. As a further test, when Lb males or an extra number of D. melanogaster males were present, the female flies displayed no such decrease in oviposition behavior (Supplementary Fig. 2c, d). To identify whether the presence of any other insects whatsoever reduces D. melanogaster egg laying, we used another species of Drosophila, D. suzukii, and the other two nonpredatory parasitoid wasps to D. melanogaster, Chouioia cunea and Scleroderma guani. We found that the presence of D. suzukii males did not have any effects on the egg laying of D. melanogaster females, nor did the presence of C. cunea or S. guani young females (Supplementary Fig. 2e−g). As such, these results suggest that D. melanogaster has specifically evolved to distinguish female from male Lb and parasitic parasitoids from nonparasitic insects.

Decreased oviposition is relevant to the potential parasitic efficiency

Focusing on the main finding of this study that the presence of young female parasitic wasps triggered decreased oviposition, we further investigated the difference between young and old wasp females. Some studies have already revealed that the parasitic efficiency of female wasps declines as they age^28,29. We then performed an assay to detect whether parasitic efficiency was impaired in the young and old Lb female wasps in this system. We found that the parasitic rates were 92%, 88%, 78% and 67% for the 2-day-old, 4-day-old, 6-day-old, and 8-day-old young wasps, respectively, whereas the parasitic rate of old wasps (10-day-old and 12-day-old) dramatically decreased to 57% and 43%, respectively (Fig. 2a). To further identify the distinct life-threatening stimuli from the cohabiting Lb females, we generated two olfaction-defective wasp strains, including a strain with knockdown of Orco (a gene encoding an obligate coreceptor of all odorant receptor proteins) mediated by RNA interference (Fig. 2b) and an antenna-ablated strain. As expected, the Orco RNAi-treated and antenna-ablated female wasps presented low or zero parasitic ability (Fig. 2c), which is consistent with the fact that olfaction is important and necessary for host seeking of parasitoid wasps^30,31,32. We then placed Drosophila females with these parasitism-defective 4-day-old young female wasps and detected the egg-laying behavior. We found that decreased oviposition was not observed upon exposure to Orco RNAi-treated female wasps and antenna-ablated female wasps (Fig. 2d). Overall, these results suggest that wasp parasitism behaviors might cause the decreased oviposition of flies exposed to young parasitoids.

Drosophila oviposition depression is correlated with the host search performance of female wasps

We next sought to determine how the female wasps provided cues that elicit the changes in Drosophila oviposition because it is not possible for flies to directly obtain the parasitic rate of Lb. Once the Lb females were released into the fly food bottles, they began to search for hosts to parasitize on the food surface, and they also intermittently rested on the fly food or the fly bottle walls. The stereotyped search behavior of Lb female wasps includes rhythmically drumming with their antennal tips, continuous movement on food substrate and frequent stinging with their sharp ovipositor in fly food (Supplementary Movie 1). We then compared the search index (SI), which is defined as the percent of time a wasp presents search behavior in a certain period of observation time (10 min in this study). The results showed that the SI was significantly higher for 4-day-old than 12-day-old female wasps, although it was variable among different tested individuals (Fig. 3a). Correspondingly, locomotion trajectory analysis showed that the young females moved more on the food medium than did the old wasps (Fig. 3b, c), indicating that the old wasps spent much more time resting. Moreover, the resulting locomotion speed of the young female wasps was higher than that of the old wasps (Fig. 3d). To further investigate whether the impaired host search performance was due to a reduced complement of eggs, we examined the ovary pairs in both young (4-day-old) and old (12-day-old) female parasitoid wasps. After dissection, we found that the ovary sizes and the mature egg numbers of young and old Lb females were comparable (Supplementary Fig. 3a, b). Collectively, the differences in search performance, such as the SI and locomotion speed, between the young and old Lb females might account for the decline in parasitic efficacy, which in turn affects decreased Drosophila oviposition.

We further monitored the search performance for Orco RNAi-treated young Lb female wasps and antenna-ablated young Lb female wasps since they present low parasitic efficiency (Fig. 2c). We also tested another kind of nonparasitic Lb wasp, ovipositor-ablated young female wasps. The SI values were significantly lower for the Orco RNAi-treated female wasps than the young females and were negligible for the antenna-ablated female wasps (Fig. 3a). Correspondingly, the trajectory analysis showed that the Orco RNAi-treated female wasps and antenna-ablated female wasps moved less than the young female wasps on fly food medium (Fig. 3b, c). In addition, the relevant locomotion speed of the Orco RNAi-treated female wasps and the antenna-ablated female wasps was also lower than that of the young Lb females (Fig. 3d). Strikingly, the ovipositor-ablated female wasps showed a normal SI value, locomotion trajectory and speed, which were comparable to those of the young female wasps (Fig. 3a–d). However, although the ovipositor-ablated Lb female wasps failed to parasitize the Drosophila host larvae, there was a similar defect in oviposition as that with the young female wasps (Fig. 3e, f). These results indicate that the search performance of female wasps is possibly responsible for the defensive response of Drosophila.

Visualization of the parasitoid wasp is responsible for decreased oviposition

It has been reported that vision is necessary for flies to initiate the defensive response to accelerate sexual behavior when encountering parasitic wasps²⁵. We next tested whether visual inputs were also responsible for the alternation of oviposition rates. We found that GMR-grim flies, which express an apoptotic activator in the developing retina leading to blindness³³, exhibited no oviposition changes when exposed to Lb females (Fig. 4a). In contrast, two independent Orco mutants (Orco¹ and Orco²), which fail to respond to most olfactory stimuli³⁴, initially had reduced oviposition rates in the presence of female wasps, but rates gradually returned to normal, as was observed with the wild-type flies (Fig. 4b, Supplementary Fig. 4a). To further elucidate the role of vision in the decreased oviposition response, we found that a visually impaired mutant, ninaB¹, showed equivalent egg laying in the presence and absence of Lb females (Supplementary Fig. 4b). We also performed the oviposition experiments in darkness (Fig. 4c). Oviposition rate decreases were not observed when the flies cohabited in darkness with either young (4-day-old) or old (12-day-old) female parasitoid wasps (Fig. 4d), confirming the findings that vision is very important to the decreased egg laying. We next placed the flies and wasps in a special apparatus (see methods and Fig. 4e) to physically separate the two populations but allowed them to see each other through a transparent window. Oviposition was significantly suppressed in the wild-type flies that could see the young female wasps but not in the flies that could see the old wasps (Fig. 4f). We note that the effect of egg laying in this special apparatus (20% reduction) is smaller than that in regular fly bottles (44% reduction), leading us to propose the possible explanation that the visual cues from wasps in the transparent chamber were much weaker. However, it is also possible that some other sensory modalities (e.g., olfaction and audition) can strength the visual-induced egg-laying reduction, while they are not sufficient alone.

Lobular columnar (LC) neurons in the lobula are responsible for transmitting primary visual information to higher brain regions^35,36. Silencing of one class of LC neurons, LC4, has previously been shown to reduce the mating acceleration response of Drosophila to parasitoids²⁵. We then used LC4-specific split GAL4 lines driving the UAS-TNT transgene (tetanus toxin light chain, which cleaves synaptobrevin to block synaptic transmission) to block LC4 neuron activity. We found that parental control flies containing a UAS-TNT transgene alone or split-GAL4 constructs “LC4-GAL4” alone showed the expected reduction in egg laying when exposed to 4-day-old (young) Lb females (Fig. 4g). However, exposed flies in which LC4 neurons were blocked (LC4-GAL4 > UAS-TNT) did not show a reduction in egg laying. When thermally activated transient receptor potential channel A1 (UAS-TRPA1) was ectopically expressed under the control of LC4-GAL4 to conditionally increase the activity of LC4 neurons, we found that the TrpA1 activation of LC4 neurons did not induce the egg reduction behavior (Supplementary Fig. 5). These results show that these neurons are necessary but not sufficient to initiate the effect of young female wasps (Fig. 4g; Supplementary Fig. 5).

Taken together, these results indicate that D. melanogaster females depend on LC4 visual projection neurons to sense parasitoid female wasps in their environment and the visual cues from Lb females (i.e., the search performance) initiate changes in oviposition.

Exposure to wasps inhibits fly ovulation and induces egg retention

To identify the underlying mechanisms that change the oviposition rate in the presence of a threat from wasps, we examined the ovary pairs in both wasp-exposed and unexposed female flies. As with the previous approaches, the flies in the group exposed to young female wasps laid fewer eggs on Day 2 than those in the unexposed group, but the oviposition rate of the exposed flies was normal on Day 10 (Fig. 1b). After dissection, we found that the ovaries of the exposed flies were much larger than those of the unexposed controls on Day 2 (Supplementary Fig. 6a–c). As expected, the ovary sizes were comparable on Day 10, since both the unexposed and exposed female flies laid similar numbers of eggs (Supplementary Fig. 6a–c). Moreover, ovary enlargement in the exposed flies on Day 2 was caused by increased ovarian retention of mature follicles, as these ovaries contained more mature eggs per ovary than those of the unexposed flies (Fig. 5a, b). These results indicate that the reduction in egg laying is probably due to an arrest in ovulation.

During ovulation, mature eggs (stage 14 egg chambers, also known as mature follicles or mature oocytes) are released from the ovary into the oviduct and subsequently into the uterus. This process requires active proteolytic degradation of the follicle wall and follicle rupture³⁷. Specifically, posterior follicle cells of a mature egg chamber are trimmed, breaking the follicle-cell layer and allowing the egg to be released into the oviduct^38,39. We further investigated whether this process was impaired in the exposed flies using two different methods (see details in the Methods). First, samples were DAPI stained, and the ratio of mature oocytes fully or partially covered by follicular cells was quantified (Fig. 5c). As expected, posterior trimmed follicles were readily observed in the ovaries of the controls, accounting for 20% and 19% of the total mature follicles in the unexposed control flies on Day 2 and Day 10, respectively (Fig. 5d). The percent of posterior trimmed follicles was reduced significantly in the females exposed to young Lb female wasps for 2 days, but the percent was similar in the exposed and unexposed flies on Day 10 (Fig. 5d). Second, we used R47A04-Gal4 to drive the expression of UAS-RFP specifically in follicle cells³⁷. By monitoring the RFP signal, we also observed a decrease in follicle trimming of mature oocytes in the flies exposed to wasps for 2 days but not for 10 days compared to the unexposed controls (Supplementary Fig. 7a, b).

Matrix metalloproteinase 2 (Mmp2) is an enzyme that is required for trimming posterior follicle cells and plays a crucial role in facilitating ovulation⁴⁰. We thus speculated that the activity of Mmp2 contributes to the egg-laying arrest in flies that occurs after exposure to wasps. We carried out in situ gelatinase assays to measure Mmp2 activity within follicles of the exposed and unexposed flies (Fig. 5e). Approximately 35% and 29% of the mature follicles had gelatinase activity at their posterior end in the unexposed flies on Day 2 and Day 10, respectively (Fig. 5f). In contrast, 18% of the mature follicles in the exposed flies had gelatinase activity on Day 2, a significant decrease compared to the unexposed controls (Fig. 5e, f). Mmp2 activity did not differ in the exposed flies and the unexposed flies on Day 10 (Fig. 5e, f). Together, our data indicate that Mmp2 activity is significantly decreased, which in turn impairs the rupture of follicular cells around mature oocytes with a subsequent reduction in the number of eggs laid when flies are exposed to young parasitic wasps.

Exposure to wasps stimulates the contractions of Drosophila oviduct muscle

The Drosophila oviduct contains circular striated muscle fibers but no longitudinal muscle fibers⁴¹. The relaxation of oviduct muscles facilitates the movement of eggs from the ovary into the oviduct⁴². We speculated that exposure to wasps also affects the relaxation of the oviduct in flies. To test this hypothesis, we exposed female flies that express the myosin heavy chain fused to GFP (MHC-GFP)⁴³ to Lb females. In these flies, the distance between each GFP band reflects the sarcomere length. As such, we can easily examine oviduct muscle tonus by measuring the average sarcomere length at the upper common oviduct (Fig. 5g). Sarcomere lengths were significantly shorter in the exposed flies than in the unexposed controls on Day 2 (Fig. 5h, i). In contrast, sarcomere lengths were indistinguishable in the exposed and unexposed flies on Day 10 (Fig. 5h, i). We next investigated whether the eggs became stuck in the oviduct due to abnormal muscle contraction. To our surprise, there was no obvious difference in the oviduct with the stuck eggs between the exposed flies and the unexposed controls (Supplementary Fig. 8). These results indicate that the presence of young wasp females causes dysfunctional contractions of oviduct muscles, but this change may not be the key factor to suppress egg laying in exposed flies.

Exposure to wasps decreases Drosophila egg laying through OA neuronal signaling

Approximately 70-100 octopaminergic neurons are dispersed throughout the Drosophila nervous system and produce the octopamine (OA), which is an important neuromodulator⁴⁴. In the ventral nerve cord (VNC) region, there are five different octopaminergic neuron clusters based on their position, which include PTS (single cell in the midline of the prothoracic neuromere), PTC (cell cluster in the midline of the prothoracic neuromere), MSC (cell cluster in the midline of the mesothoracic neuromere), MTC (cell cluster in the midline of the metathoracic neuromere) and AC (cell cluster in the thoracic abdominal ganglia)^45,46 (Fig. 6a). Importantly, AC neurons in the VNC region innervate female reproductive tissues such as the ovaries, oviducts, and uterus⁴⁷ and modulate OA-dependent egg-laying behaviors, including mature follicle trimming, rupture and oviduct muscle relaxation^39,41,48 (Fig. 6a). Therefore, our above results have suggested that OA is possibly involved in the decreased oviposition response in the presence of young Lb females (Fig. 5; Supplementary Fig. 7). OA is synthesized from tyrosine by the sequential actions of tyrosine decarboxylase 2 (Tdc2) and tyramine beta-hydroxylase (Tβh)^49,50. We then analyzed the expression of the mRNAs encoding Tdc2 and Tβh in the brain and VNC of the exposed and unexposed Drosophila females by qRT-PCR. Strikingly, no significant difference was observed in the brains of the exposed flies compared to the unexposed female flies after exposure to young Lb wasps on Day 2 or on Day 10 (Fig. 6b, c). However, the levels of Tdc2 and Tβh were significantly decreased in VNC when the female flies were exposed to wasps for 2 days but not for 10 days (Fig. 6d, e). We also examined the expression levels of Tdc2 and Tβh in the GMR-grim flies. There was no difference between the exposed and unexposed GMR-grim flies in VNC after a 2-day exposure to young female wasps (Supplementary Fig. 9). These results indicate that the expression of two key genes (Tdc2 and Tβh) is downregulated specifically in the VNC region of Drosophila females when they see active natural enemies.

Decreased levels of Tdc2 and Tβh upon threat exposure are expected to reduce both OA and tyramine levels, and OA is widely reported to regulate insect egg-laying behaviors^{37.47,48,50,51}. We then used whole-mount immunohistochemistry to examine the locations and levels of OA in the VNC of the exposed and unexposed Drosophila females. Antibody staining supports that OA in VNC is produced specifically in different octopaminergic neuron clusters, including PTS, PTC, MSC, MTC and AC (Fig. 6f). Consistent with the qRT-PCR results, we found that the levels of OA were significantly decreased in all octopaminergic neurons of the VNC after a 2-day exposure to young female wasps (Fig. 6f, g). We next examined whether OA levels were impaired in the axons of octopaminergic neurons on the reproductive tracts of exposed Drosophila females. Immunohistochemistry analyses showed that OA-immunoreactive nerve termini were found in all regions of the reproductive tract, including the ovary (OV), lateral oviducts (LO), upper common oviduct (COU), and lower common oviduct (COD) (Fig. 7a−c). The highest fluorescence intensity for OA was in nerve termini of the LO, followed by COU and COD (Fig. 7c). The results further showed that OA intensity was significantly decreased in all different regions in the reproductive tracts of exposed flies compared to unexposed female flies after exposure to young Lb wasps on Day 2 (Fig. 7b, c).

To further investigate whether ectopically increasing the activity of OA neurons could rescue the egg-laying defect of the exposed Drosophila females, we carried out two independent experiments with the help of Tdc2-GAL4. This GAL4 line is specifically expressed in OA- and tyramine-producing neurons and has been widely used to manipulate OA neuronal activity in vivo^49,52,53. First, we ectopically expressed UAS-TRPA1 under the control of Tdc2-GAL4 to conditionally increase the activity of OA neurons^54,55 (Fig. 7d). At the nonactivating temperature of 23 °C, egg-laying defects were observed in both the Tdc2-GAL4 and UAS-TRPA1 control fly females and in the Tdc2-GAL4 > UAS-TRPA1 females after 2 days of exposure to young Lb females (Fig. 7d). However, at the TRPA1-activating temperature of 29 °C, the wasp-exposed Tdc2-GAL4 > UAS-TRPA1 females, which have high levels of OA neuron activity, laid significantly more eggs than the wasp-exposed control females (Fig. 7d). Next, we used the Tdc2-GAL4 driver to express a gene (UAS-eag^DN) encoding a dominant-negative ether-a-gogo potassium channel subunit in OA neurons⁵⁶. Loss of function of eag results in increased neuronal activity⁴⁸. As expected, the control Tdc2-GAL4 and UAS-eag^DN females laid significantly fewer eggs after exposure to wasps for 2 days than the unexposed flies (Fig. 7e). However, the Tdc2-GAL4 > UAS-eag^DN females, which have elevated activity of OA neurons, showed a significant increase in oviposition rate relative to the controls that were exposed to young Lb females (Fig. 7e). These results indicate that increasing OA neuronal activity in Drosophila females compensates for the oviposition deficiency induced by the presence of young female wasps.

We then examined whether injection of OA would stimulate egg laying in flies exposed to wasps. As expected, the OA-injected individuals had an increased oviposition rate compared to the controls injected with doubly distilled H₂O when they were exposed to young Lb female wasps (Fig. 7f). These results further support our conclusion that the reduced oviposition rate of wasp-exposed Drosophila females can be rescued by elevating OA neuronal activity.

Taken together, the results indicate that a reduction in OA neuronal signaling in D. melanogaster females triggered by visual cues in the presence of young Lb wasps results in depression of oviposition.

Two neuronal signaling pathways underlie the diverse defensive behavior responses to the same parasitoid threats

In addition to depression of oviposition, Drosophila females also switch to laying their eggs on ethanol-laden food when they encounter deadly parasitic wasps²³. This strategy is thought to be a behavioral immune response that protects hatched offspring from infection. The ethanol preference in D. melanogaster is linked to a decrease in neuropeptide F (NPF) in the brain in the presence of parasitic wasps²³. We therefore sought to determine whether NPF also plays a role in depressed oviposition responses and whether NPF and OA neuronal signaling have direct causal connections. We inhibited the expression of NPF using a shRNA expressed from the NPF-GAL4 driver, which is specifically expressed in all NPF-producing neurons. As expected, a significant decrease in NPF levels was found when fly brains were immunostained with NPF antiserum, indicating that NPF was successfully targeted for degradation (Supplementary Fig. 10a, b). The proportion of eggs laid by the NPF-GAL4 > UAS-NPF RNAi females was marginally reduced compared to that of the NPF-GAL4 and UAS-NPF RNAi control females on several testing days (Supplementary Fig. 10c). We next evaluated follicle rupture and Mmp2 activity in flies with decreased NPF levels and observed no significant differences between the NPF-Gal4 > UAS-NPF RNAi and control females (Supplementary Fig. 10d, e). Similarly, when NPF receptor (NPFR) was specifically knocked down in OA neurons of Drosophila females, the oviposition rate, the percent of posterior trimmed follicles, and Mmp2 activity were not affected (Supplementary Fig. 11a−c). These data indicate that decreased NPF does not alter OA-mediated neuronal signaling to prevent Drosophila ovulation. Correspondingly, to address whether OA-mediated neuronal signaling influences NPF levels, we inhibited the expression of the four OA receptors in NPF-producing neurons^57,58,59. There were no changes in NPF levels in OA signaling-ablated NPF neurons (Supplementary Fig. 12a−d).

Taken together, these observations indicate that the two neuronal signaling pathways act independently to elicit two different behavioral outputs in the presence of female wasps. A reduction in NPF signaling is mainly responsible for the ethanol preference response²³, whereas a reduction in OA signaling leads to decreased oviposition.

Source: Ecology - nature.com