Introduction

Competition for resources such as territory, food, and mates is crucial for animal survival and reproduction [1-4]. Among diverse competitive strategies, physical combat is a prevalent form of fighting to repel rivals and regulated by both genetic factors and social environment [1, 5]. Understanding how social experience influences combat behaviors in animals to maximize their survival and reproduction remains a fundamentally important yet inadequately understood question.

In the fruit fly, Drosophila melanogaster, inter-male aggression includes low-intensity fights such as lunging, in which one fly rears up on its hind legs and snaps down onto its opponent, as well as high-intensity fights like tussling, where two males tumble over each other [6-8]. It is noteworthy that the majority of studies focus on the low-intensity lunging behavior, owing to its higher frequency and ease of observation. Substantial studies have identified regulatory genes and the neural circuit underlying the lunging-based aggressive behaviors [7, 9]. In particular, the male-specific volatile pheromone 11-cis-vaccenyl acetate (cVA) acutely promotes lunging behavior via its olfactory receptor Or67d [10]. Interestingly, chronic cVA exposure in socially-enriched males reduces the level of lunging via another olfactory receptor Or65a [11]. Consistently, it has been found that social isolation increases lunging frequency by altering the expression of genes such as Cyp6a20 [12], Drosulfakinin (Dsk)[13], and Tachykinin (Tk) [14]. Social isolation also increases male aggression in mice [15, 16] and even in humans [17]. In mice, the TK homologue Tachykinin 2 (Tac2) responds to social isolation and promotes inter-male aggressive behaviors [16]. These findings contribute to the consensus that social isolation escalates aggression, whereas social enrichment mitigates it in animal models.

Meanwhile, social enrichment increases male mating success by enhancing the sensitivity of Or47b-expressing olfactory receptor neurons (ORNs) to fly pheromones common to both sexes [18, 19]. Artificially activating Or47b ORNs also improves mating advantage in a competitive assay [20]. Group housing and age-dependent juvenile hormone (JH) jointly increase the expression of the male-specific proteins from fruitless (fru^M), which is a master gene for male courtship behaviors [21-23], in the Or47b-expressing ORNs [19, 24, 25]. Notably, courtship levels in single-housed and group-housed wild-type males are not significantly different [26]. In fact, group-housing decreases excitability of the courtship-promoting P1 neurons [27], which express both terminal genes in the sex determination hierarchy, fru and doublesex (dsx). These findings raise a question on how social enrichment, which suppresses lunging-based aggression and does not increase courtship intensity, would facilitate mating competition.

Here, we establish a behavioral paradigm for studying inter-male tussling in Drosophila and discover that this infrequent yet intense form of combat is enhanced by social enrichment and is crucial for territory dominance and mating competition. This process requires olfactory receptor Or47b, but not Or67d or Or65a. We further identify three pairs of central dsx-expressing pC1 neurons that specifically promote tussling. Moreover, we demonstrate that socially enriched males can overcome mating disadvantages associated with aging.

Results

A Paradigm for High-intensity Aggression in Drosophila

While studying the lunging-based aggression in Drosophila, we occasionally observed the high-intensity tussling behavior in male flies. We reasoned that the intensity, instead of the frequency, could be decisive for a combat. Thus, although tussling is a less frequent behavior, it could serve a vital function.

To induce a higher level of tussling for further manipulation, we first introduced a fixed live virgin female in the same behavioral chamber for assaying lunging. The female fly was immobilized in the middle of the food, with its head embedded in the food and its body exposed to enhance male competition (Figure 1A). We found that males that were group-housed for 7 days showed a significantly higher frequency of tussling behavior and reduced latency to tussle in the presence of an embedded virgin female (Figure 1B). We next compared tussling in male flies of different ages. We observed a significantly higher proportion of 14-day-old males raised in groups that showed tussling behavior during the 2-hour observation period, compared to the 7-day-old males (Figure 1B). Furthermore, 2-day-old males never displayed tussling behavior. In contrast, 14-day-old males showed lowed level of lunging behavior than 7-day-old males did (Figure 1—figure supplement 1A and 1B). Previous studies have shown that, in addition to mating resources, food quality is another crucial factor affecting aggressive behaviors [28]. Therefore, we next assayed tussling behavior in 14-day-old males in the presence of an embedded virgin female with different concentrations of sucrose or yeast in the food during the behavioral test. We found that these males displayed similar levels of tussling behavior with different concentrations of sucrose in the food (Figure 1C). On the contrary, male tussling intensity increased significantly with higher yeast concentrations such that more than 70% of males displayed this high-intensity fighting behavior (Figure 1D, see Video 1). Taken together, these results establish an efficient behavioral paradigm for the high-intensity tussling behavior and indicate that older males display more tussling but less lunging than younger ones.

Figure 1.

Figure 1—figure supplement 1.

Social Enrichment Inhibits Low-intensity Lunging but Enhances High-intensity Tussling

Previous studies have shown that social enrichment inhibits aggressive behaviors, specifically the low-intensity lunging, in male flies. Therefore, we wondered whether social enrichment might also inhibit the high-intensity tussling behavior in males. To answer this question, we compared both fighting behaviors in single-housed (SH) and group-housed (GH) 14-day-old males (Figure 2A). Lunging and tussling events from each group of representative samples were exhibited in raster plots for comparison (Figure 2B). We scored lunging behavior in the first 10-min of test, due to its relatively higher occurrence, and tussling behavior during the whole 2-hour observational period since males take much longer time to initiate tussling (Figure 1D). Interestingly, we found that while SH males performed more frequent lunging behavior in 10 min than GH males did, which is consistent with previous findings, they rarely showed tussling behavior. In contrast, GH males displayed a large number of tussling events within 2 hours. Note that a tussling event may last up to several minutes, indicative of its high intensity, in contrast to lunging events that were often transient (Figure 2B). Statistical analysis of tussling and lunging behaviors showed that GH males exhibited reduced tussling latency and enhanced tussling frequency, in addition to increased lunging latency and decreased lunging frequency, compared to SH males (Figure 2C-F). These results indicate that social enrichment inhibits low-intensity lunging but promotes high-intensity tussling in male flies.

Figure 2.

Or47b ORNs Mediate Social Experience-induced Tussling Behavior

To further investigate the sensory basis underlying the opposite regulation of lunging and tussling behaviors by social experience, we tried to examine the role of cVA-sensitive neurons, as well as other fru^M-expressing ORNs, in mediating intermale aggression. Two cVA sensory neurons, Or67d and Or65a ORNs, play crucial roles in enhancing the lunging-based aggression in SH males, in which Or67d ORNs mediate acute cVA stimulation and promote male aggression[10], while Or65a ORNs mediate chronic cVA stimulation and inhibit male aggression in GH males [11]. In addition, the fru^M-expressing Or47b ORNs are sensitive to a class of pheromones common to both sexes, such as palmitoleic acid (PA), and their sensitivity is enhanced under group rearing conditions[19]. Ir84a ORNs, which are also fru^M-positive, are involved in the perception of mixed food odors and promote male courtship behavior[29]. We silenced Or67d, Or65a, Or47b or Ir84a ORNs by expressing the inwardly rectifying potassium channel Kir2.1 and assayed lunging and tussling behaviors in 14-day-old males. We found that silencing Or47b ORNs, but not Or67d, Or65a or Ir84a ORNs, significantly decreased male tussling, compared to the control (Figure 3A, B). In contrast, silencing Or67d ORNs, but not other ORNs, significantly reduced male lunging (Figure 3C, D). These loss-of-function results suggest that Or47b ORNs mediate male tussling behavior, whereas Or67d ORNs regulate male lunging behavior.

Figure 3.

Figure 3—figure supplement 1.

Figure 3—figure supplement 2.

To elucidate the potential role of the Or47b receptor in mediating tussling behavior, we knocked down its expression in Or47b ORNs. Our results showed a significant decrease in male tussling behavior when the Or47b receptor expression was reduced (Figure 3E, F), indicating that the functionality of Or47b ORNs in tussling is dependent on the Or47b receptor. As previous studies suggested that the male-specific fru^M sustains Or47b expression in an activity-dependent manner [19], we next knocked down fru^M expression in Or47b ORNs and assayed male tussling. The efficiency of Or47b and fru^M RNAi knockdown was both validated by quantitative PCR (Figure 3— figure supplement 1A and 1B). We observed a slight increase of tussling latency in fru^M knocked-down males, although not significantly different from the control (Figure 3E, F), suggesting that Or47b, rather than Fru^M, plays a more crucial role in mediating male tussling.

To further confirm that the effect of social enrichment on tussling depends on the activity of Or47b ORNs, we artificially activated Or47b ORNs by expressing the bacterially derived sodium channel (NaChBac)[30] in 14-day-old SH males. We found that these SH males with enhanced activity of Or47b ORNs showed a significant increase in tussling behavior compared to control SH males. Interestingly, the level of tussling in these SH males with activated Or47b ORNs was indistinguishable from that in GH males (Figure 3G, H). These gain-of-function results indicate that social enrichment acts through Or47b ORNs to promote male tussling. Consistent with this, we observed substantially more downstream neurons of Or47b ORNs in GH males than those in SH males (Figure 3—figure supplement 2A and 2B), as visualized by the trans-Tango technique [31].

Distinct Central Brain Neurons for Lunging and Tussling

We next set out to identify the central neurons involved in regulating the high-intensity tussling. We found that silencing the mushroom body neurons via Kir2.1 did not affect male tussling, suggesting that the experience-dependent tussling behavior is not dependent on the classical learning and memory center. Surprisingly, inactivation of the previously-identified aggression-promoting P1^a and TK neurons did not affect tussling behavior (Figure 4A, B). These results imply that the central circuit for tussling behavior may be distinct from the previously reported aggression circuits. As dsx-expressing pC1 neurons, also termed P1 if co-expressing fru^M, are crucial for sexual and aggressive behaviors in both males and females, we also silenced two subsets of pC1 neurons (pC1^SS1 and pC1^SS2), whose functions were previously characterized in females but not in males. We found that inactivation of pC1^SS2 but not pC1^SS1 neurons significantly reduced the tussling behavior in male flies (Figure 4A, B). The cell bodies of pC1^SS2 and P1^a neurons are both located in the center of posterior brain and project to overlapping but distinct parts of the lateral protocerebral complex (Figure 4C-E). These results suggest that pC1^SS2 neurons may be key central neurons regulating tussling behavior.

Figure 4.

Figure 4—figure supplement 1.

To further compare the function of P1^a and pC1^SS2 neurons in regulating male behaviors, we performed optogenetic activation experiments on these neurons expressing CsChrimson. Compared to control male pairs that did not show any significant change upon red light stimulation (Figure 4F-H), optogenetic activation of P1^a neurons induced acute courtship behavior as indicated by unilateral wing extension (UWE), and upon red light removal, there was a significant increase in lunging but not tussling (Figure 4I-K), consistent with previous findings. In contrast, optogenetic activation of pC1^SS2 neurons induced strong and acute tussling behavior, with more than 50% of the time during red light in tussling, and the tussling behavior ceased immediately upon removal of red light (Figure 4L, see Video 2). Activation of pC1^SS2 neurons did not significantly affect either male lunging or courtship (Figure 4M, N). These results indicate that pC1^SS2 neurons acutely promote the high-intensity tussling behavior.

Previous studies have revealed a crucial role of P1^a neurons in promoting a persistent internal state that enhances male aggression. To test whether P1^a neurons can promote both lunging and tussling behaviors, we further performed thermal activation experiments. We found that activation of P1^a neurons via dTrpA1 at 30°C induced male tussling, just like activation of pC1^SS2 neurons did (Figure 4—figure supplement 1A and 1B). As thermal activation but not optogenetic activation of P1^a neurons induced tussling, we reasoned that vision might be a key factor for this discrepancy, since all optogenetic experiments were performed in dark or red light. Thus, we re-performed these thermal activation experiments in dark and found that activation of pC1^SS2 but not P1^a neurons induced tussling, although the level of tussling by pC1^SS2-activated males was lower compared to that in light, suggesting a crucial role of vision in male tussling. These optogenetic and themogenetic activation experiments suggest that P1^a neurons can promote both lunging and tussling behaviors, but their function on tussling requires visual stimulation, while pC1^SS2 neurons play a more direct role in acutely promoting male tussling.

Territory and Mating Advantage in Tussling-favored Experienced Males

As our results showed that GH males and SH males were more likely to perform the low-frequency, high-intensity tussling and the high-frequency, low intensity lunging, respectively, we wondered which fighting strategy was more practical in the competition for territory and mating resources.

To address this question, we first designed an experimental paradigm for territory competition (Figure 5A). A circular territory, which is provided with food and a fixed virgin female and separated by a wall with a small door open to the outer area, was competed by two tester males. We manually analyzed the first 10 winning events; each was defined as one male successfully displacing the other from the inner resource area. The winning index was calculated as the proportional difference in the number of wins between each pair of males. In control groups such as two GH males of the same age, we observed some ‘winner takes all’ phenomenon (one male won nearly all of the 10 encounters), but overall, the winning index was close to zero. We further found that while GH 7-day-old males (G7) won slightly, but not significantly, more than SH 7-day-old males (S7) (Figure 5B), older G14 males had a significant advantage over the S14 males (Figure 5C and Figure 5—figure supplement 1, see Video 3).

Figure 5.

Figure 5—figure supplement 1.

Next, we conducted further investigations into the role of social experience and the two types of fighting behaviors in mating competition (Figure 5D). Two males and one virgin female were loaded into the behavioral chamber at the same time and recorded for courtship behaviors. We found that G7 males had no significant advantage over S7 males (Figure 5E); however, G14 males displayed a significantly increased mating advantage over S14 males (Figure 5F, see Video 4). Together, these results suggest that the tussling-favored GH males are more competitive in territory control and mating compared to the lunging-favored SH males.

The above results showed that GH males were more competitive compare to SH males, especially in older males, suggesting that social experience may play a potential role in altering age-related mating competition. To test this hypothesis, we first tested mating competition between males of different ages under the same rearing conditions. The results showed that younger males, whether reared in group or isolation, had a greater mating advantage compared to older males (G7 > G14, S7 > S14, G14 > G21, S14 > S21) (Figure 5G). These results indicate that there is generally a disadvantage in mating competition associated with aging. We next investigated whether the mating advantage induced by social experience could compensate for the mating disadvantage caused by aging. We tested pairs of males with different ages and rearing conditions (G7 vs. S14, S7 vs. G14, G14 vs. S21, S14 vs. G21) for mating competition. We found that while 14-day-old males were less competitive than 7-day-old males regardless of the rearing condition, indicative of the aging-related disadvantage, 21-day-old GH males were significantly more competitive than 14-day-old SH males (G21 > S14) (Figure 5H). These results indicate that long-term social enrichment in older males increases their territorial and mating competition, thus overcoming their mating disadvantage associated with aging (Figure 6).

Figure 6.

Discussion

Regulation of physiology and behavior by social experience is an important feature across animal species. Here, we found that social experience shapes the fighting strategies of male flies, in which GH males tend to reduce the low-intensity lunging and increase the high-intensity tussling, which enhances their territorial and mating competition. We further showed that the two forms of fighting are mediated by distinct sensory and central neurons. The complex implications of these findings are discussed below.

Previous studies have found that social isolation generally enhances aggression but decreases mating competition in animal models, presenting a paradox between social experience, aggression and reproductive success. Our results resolved this paradox by dissecting aggressive behaviors into the well-studied, frequently observed lunging behavior and the less frequent but high-intensity tussling behavior. We established a new behavioral paradigm for studying tussling behavior and revealed that social experience oppositely regulates lunging and tussling. It is important to note that the frequency of lunging decreases while the frequency of tussling increases from 7-day-old to 14-day-old males, suggesting a shift in fighting strategy with age. Our results further suggest that social experience may enhance male mating competition by shifting their fighting strategy to the low-frequency, high-intensity tussling, which may be more effective in combat, although direct evidence is still lacking. Interestingly, we found that the mating advantage gained through social enrichment can even offset the mating disadvantage associated with aging, further supporting the vital role of shifting fighting strategies in experienced, aged males. This dissection of the two fighting forms rectifies our understanding of how social experiences regulates aggression and reproductive success.

Our results revealed distinct neural circuits underlying the two forms of fighting, spanning from primary sensory neurons to central regulatory neurons. Specifically, silencing the cVA-sensing Or67d neurons reduced male lunging but not tussling, whereas inactivating Or47b neurons impaired male tussling but not lunging. Notably, Or47b neurons respond to pheromones common to both sexes, and the sensitivity of these sensory neurons is enhanced by social enrichment [18, 19, 21]. Unlike the Or67d-dependent lunging, which is primarily male-induced and socially inhibited, the Or47b-dependent tussling is promoted by social enrichment with either male or female flies. Recently, it has been found that female flies increase aggression towards mating pairs or mated females, depending on Or47b and Or67d, respectively [32, 33]. These results suggest a common function of Or47b and Or67d neurons in both sexes for regulating aggression. However, whether these neurons mediate different fighting patterns in females and how they respond to social experience in females remain unknown. Nevertheless, our results on these two sensory pathways in males provide crucial evidence for dissecting their functions.

Regarding central regulatory neurons, we identified a novel class of pC1 neurons, pC1^SS2, that specifically promote male tussling. In male flies, there are ∼60 pairs of dsx-expressing pC1 neurons, including the well-characterized P1^a neurons that also express fru^M [34, 35]; however, there are only three pairs of pC1^SS2 neurons whose function is both necessary and sufficient for inducing tussling in males. The three pairs of pC1^SS2 neurons may have distinct function from P1^a neurons in several ways. (1) P1^a neurons promotes both male courtship and aggression [36, 37], whereas pC1^SS2 neurons specifically promote aggression but not courtship. A recent study also identified pC1^SS2 neurons as aggression-promoting, but their roles in different fighting modes are not depicted [38]. (2) P1^a neurons promotes both lunging and tussling, although their function on tussling requires additional visual stimulation. In contrast, pC1^SS2 neurons specifically promote tussling, and the presence of light enhances but is not absolutely necessary for tussling. These results align with previous findings that visual stimuli amplify P1 activation and related behaviors [39-41], and visual stimulation is important for male aggression [42, 43]. (3) Male lunging is induced after P1^a activation; whereas male tussling is induced during pC1^SS2 activation. These results suggest that P1^a neurons may not directly regulate aggression but instead induce a persistent state that promote aggression in general, while pC1^SS2 neurons may play a more direct role in mediating tussling. Future studies are needed to further dissect the seemingly parallel sensory and central neuronal pathways involved in lunging and tussling, and to explore how these pathways interact during social experience and aging to optimize reproductive success.

Materials and methods

Fly Stocks

Flies were maintained at 22 or 25 °C in a 12 h:12 h light: dark cycle. Canton-S flies were used as the wild-type strain. Or67d-GAL4 (BDSC_9998), Or65a-GAL4 (BDSC_9993), Or47b-GAL4 (BDSC_9984), IR84a-GAL4 (BDSC_41734), TK-GAL4 (BDSC_51795), R19B03-GAL4 (BDSC_49830), UAS-dTrpA1 (BDSC_26263), UAS-Kir2.1 (BDSC_6595), UAS-CsChrimson (BDSC_82181), UAS-myrGFP (BDSC_32198) and trans-Tango (BDSC_77124) flies were obtained from Bloomington Drosophila Stock Center (BDSC). UAS-NaChBac [30], fru^GAL4 [44], P1^a-spGAL4 [45], pC1^SS1-spGAL4 [46], pC1^SS2-spGAL4 [46] and UAS-fruMi [47] were used as previously described. UAS-Or47b RNAi (THU_2599) was obtained from TsingHua Fly Center.

Male aggression assay

For the aggression assays presented in Figure 1, the prepared food was evenly spread into the chambers of two paradigms. There was no female in the traditional paradigm, while a 3-5 day old virgin female was gently fixed in the center of the food with its abdomen outside in the new paradigm. Then two male testers were placed into the upper chambers separated from the food (with or without a fixed female) by a film barrier. All testers were loaded by cold anesthesia. After a 30-minute adaptation, the film was removed, and the aggressive behavior was recorded for 2 hours. The aggression assays in Figure 2 and 3 were performed by using the new paradigm. The aggression assays in Figure 1—figure supplement 1 was performed by using the traditional paradigm.

The food recipes for the male aggression assay were listed below. A mixture of 2.5% yeast, 2.5% sucrose and 1% agar dissolved in apple juice was used for Figure 1B; a mixture of 2.5% yeast, different concentrations of sucrose (0, 2.5%, 5% or 10%) and 1% agar dissolved in apple juice was used for Figure 1C; a mixture of different concentrations of yeast (0, 2.5%, 5% or 10%), 2.5% sucrose and 1% agar dissolved in apple juice was used for Figure 1D; A mixture of 10% yeast and 1% agar dissolved in apple juice was used for other experiments.

dTrpA1-mediated activation experiments

For thermogenetic activation assay showed in Figure 4—figure supplement 1, two 5-7 days old males were loaded into the circular chamber (d=15mm) under cold anesthesia and separated with a transparent film. Before the recording start, the chamber was transferred to the recording platform at a specific temperature (22°C or 30°C) for 30 minutes adaptation. After that, the film was removed, and the behavior was recorded for 1 hour. The 10-minute video after film extraction was used for behavioral analysis.

CsChrimson-mediated optogenetic activation experiments

For optogenetic activation assay showed in Figure 4, males were raised at 25°C on standard fly food, collected after eclosion and housed in the dark for 5-7 days on food containing 0.4 mM all trans-Retinal (MFCD00001550, Sigma-Aldrich, St. Louis, MO). The specific stimulation protocols are showed in the schematic (Figure 4F). In brief, Two 5-7 days old males were loaded into the circular chamber (d=10mm) under freeze anesthesia and separated with a transparent film in the dark. Then the chamber was moved on the plane light source (IR channel: 865nm and red-light channel: 630nm) for 15 mins adaptation. Before video recording, the IR light channel was switched on and adjusted to the appropriate brightness for recording. The film was removed before the recording start, with the red-light channel switched on (constant red light, 0.02 mW/mm²) during the 2nd and 5th minutes of the experiment. The statistics for the tussling, lunging, and unilateral wing extension (UWE) are as follows. For tussling time and UWE time, we used LifesongX software to calculate the percentage of the total time these behaviors occurred per minute. For lunges, we counted the number of lunging occurrences per minute.

Territorial competition experiment

For the territorial competition experiment in Figures 5A-C, the behavioral setup was produced using 3D printing technology, as illustrated in the schematic, with a transparent cover plate placed on the top. The preparation of food and female were the same as the new aggression paradigm. The female was fixed in the center of the food and two male testers were placed separately in the two small chambers on both sides. Male wings were marked 24 hours before the experiment. Before initiating the experiment, the two barriers were closed. After a 30-minute adaptation, the barriers were opened to allow the males access to the central large chamber, and were closed once the males went inside, and the territorial behavior was recorded for 2 hours. The winning index stand for territorial control ability was calculated by the following example formula: (num. of wins by GH males – num. of wins by SH males)/10 encounters * 100%. The first 10 encounters occurring after 1 hour of recording were used for statistical analysis. The winning event was defined as a male driving the other male out of the food area.

Mating competition experiment

For the mating competition experiment shown in Figure 5D-H, the wings of males were marked 24 hours prior to the experiment using the same method described above. During the experiment, a 3-5 days old virgin female was placed into the chamber under freeze anesthesia and covered with a transparent film. Then the unmarked males were loaded into the chamber in the same manner and also covered with a transparent film. Finally, the marked males were loaded. After 30 minutes adaptation, two films were removed, and the behavior was recorded for 1 hour. The chambers without successful copulation during the 1-hour period were excluded from further analysis. Each assay included 120 efficient chambers at least, and 10 random chambers were measured as one sample. The copulation advance index was calculated by the following example formula: (num. of GH winners – num. of SH winners)/10 chambers * 100%). The winner was defined as the male that completed the first copulation with female.

Quantitative Real-time PCR

Adult fly samples were frozen in liquid nitrogen and decapitated by vortex. The heads were then separated from the bodies using metal sieves. Each sample, consisting of 40 frozen heads, was used for total RNA preparation using TRIzol reagent (15596026, Thermo Fisher Scientific, Waltham, MA). We purified total RNA using a DNA-freeTM Kit (AM1906, Thermo Fisher Scientific, Waltham, MA) and performed reverse transcription using SuperScriptTM IV (18091050, Thermo Fisher Scientific, Waltham, MA) to obtain cDNA used for templates. Quantitative PCR was performed on the Roche LightCycler® 96 Real-Time PCR machine using AceQ qPCR SYBR Green Master Mix (Q121-02, Vazyme, Nanjing). Transcript levels were analyzed by the 2^^-ΔCT method using actin as control. Each sample was run in triplicate. Each experiment was repeated three times using independent sets of genetic crosses. Primers used for RT-qPCR quantification were:

actin forward: 5′-GTCGCGATTTAACCGACTACCTGA-3′

actin reverse: 5′-TCTTGCTT CGAGATCCACATCTGC-3′

fru-P1 forward: 5′-GTGTGCGTACGTTTGAGTGT-3′

fru-P1 reverse: 5′-TAATCCTGTGACGTCGCCAT-3′

Or47b forward: 5′-CAAATCTCAGCCTTCTGCGG-3′

Or47b reverse: 5′-GATACTGGCACAGCAAACTCA-3′

Tissue Dissection, Staining, and Imaging

Brains were dissected in Schneider’s insect medium (S2) and fixed in 4% paraformaldehyde in 0.5% Triton X-100 and 0.5% bovine serum albumin in phosphate-buffered saline (PAT) for 30 min at room temperature. After 4 × 10-min washes, tissues were blocked in 3% normal goat serum (NGS) for 60 min, then incubated in primary antibodies diluted in 3% NGS for 4 h at room temperature and 1 day at 4°C, then washed in PAT, and incubated in secondary antibodies diluted in 3% NGS for 4 h at room temperature and 1 day at 4°C. The tissues were then washed thoroughly in PAT and mounted for imaging. The antibodies used were Rabbit anti-HA (1:1000), Mouse anti-GFP (1:1000), nc82(1:50) and Donkey anti-Mouse, Alexa Fluor 488 (1:500), Donkey anti-Rabbit, Alexa Fluor 555 (1:500). Samples were imaged on Zeiss 900 confocal microscopes using ZEN and processed with Fiji software.

Brain image registration

The protocol for standard brain registration is described previously [48]. Confocal images for brains containing P1^a neurons and pC1^SS2 neurons were registered onto a standard reference brain, which was generated by the images of nc82 stained brains.

Statistics

Experimental flies and genetic controls were tested under the same conditions, and data were collected from at least two independent experiments. Statistical analysis was performed using GraphPad Prism 9 (https://www.graphpad.com/scientific-software/prism/) as indicated in each figure legend. Data were first verified for normal distribution by the D’Agostino–Pearson normality test. For normally-distributed data, Student’s t test was used for pairwise comparisons, and one-way ANOVA was used for comparisons among multiple groups, followed by Tukey’s multiple comparisons. For data not normally-distributed, the Mann-Whitney test was used for pairwise comparisons, and the Kruskal–Wallis test was used for comparison among multiple groups, followed by Dunn’s multiple comparisons. For group matched data that were normally-distributed, the Paired t test signed rank test was used for pairwise comparisons, and RM one-way ANOVA was used for comparison among multiple groups, followed by RM one-way ANOVA multiple comparisons. For group matched data that were not normally-distributed, the Wilcoxon matched-pairs signed rank test was used for pairwise comparisons, and the Friedman test was used for comparison among multiple groups, followed by Dunn’s multiple comparisons. For qPCR experiments, the average relative expression of three independent experiments was analyzed using Mann-Whitney test. For one group compared with zero, we used one sample t test. The Chi-square test was used to compare male tussling percentage between two groups.

Data availability

All data generated or analyzed during this study are included in the article and its supporting information files. Fly stocks and reagents used in this study are available from the corresponding author upon reasonable request.

Additional information

Author contributions

C. Gao, Q. Peng and Y. Pan conceived the study; C. Gao and M. Ma performed all experiments and analyzed the data with help from J. Chen and X. Ji; Q. Peng and Y. Pan supervised the study; C. Gao and Y. Pan wrote the manuscript with input from all authors.

Competing interests

The authors declare no competing interests.

Acknowledgements

We thank the Bloomington Drosophila Stock Center, Tsinghua Fly Center, and Dr. Kaiyu Wang and Fei Wang for fly stocks. This work was supported by grants from National Key R&D Program of China (2019YFA0802400), and the National Natural Science Foundation of China (32371067), and the Fundamental Research Funds for the Central Universities (2242023R40054).