Drosophila TRPA Channel Painless Inhibits Male–Male Courtship Behavior through Modulating Olfactory Sensation

The Drosophila melanogaster TRPA family member painless, expressed in a subset of multidendritic neurons embeding in the larval epidermis, is necessary for larval nociception of noxious heat or mechanical stimuli. However, the function of painless in adult flies remains largely unknown. Here we report that mutation of painless leads to a defect in male–male courtship behavior and alteration in olfaction sensitivity in adult flies. Specific downregulation of the expression of the Painless protein in the olfactory projection neurons (PNs) of the antennal lobes (ALs) resulted in a phenotype resembling that found in painless mutant flies, whereas overexpression of Painless in PNs of painless mutant males suppressed male–male courtship behavior. The downregulation of Painless exclusively during adulthood also resulted in male–male courtship behavior. In addition, mutation of the painless gene in flies caused changes in olfaction, suggesting a role for this gene in olfactory processing. These results indicate that functions of painless in the adult central nervous system of Drosophila include modulation of olfactory processing and inhibition of male–male courtship behavior.


Introduction
Transient receptor potential (TRP) channels play important roles in a variety of sensory systems [1]. In mammals, TRPA1 is critical for nociception and may contribute to cold sensation [2][3][4]. As a Drosophila TRPA channel [5], the painless gene was first identified as essential for the sensation of high temperature (above ,39uC) and for noxious mechanical stimulation in the larvae [6,7]. In addition, painless has been reported to be involved in sugar-stimulated neural excitation and avoidance in larvae [8], in wasabi sensation and noxious heat sensation in adult flies [9,10], as well as in negative gravity sensation [11]. The expression pattern of the Gal4 enhancer-trap allele painless-Gal4 (pain Gal4 ) suggests that painless may be endogenously expressed in the adult fly brain [9,10], but likely functional consequences of this expression are largely unknown. The Drosophila olfactory system is critical for its detection of volatile chemicals in the external environment [12][13][14]. These chemical cues are first detected and converted into electric signals by the primary olfactory receptor neurons (ORNs). After further processing by secondary projection neurons (PNs) and local interneurons (LNs) of the antennal lobes (ALs), this information is then conveyed to higher brain regions to direct appropriate behaviors, including foraging, fighting and courtship [15]. Pheromones act as important cues for gender identification in flies [16]. Thus, the Drosophila olfactory system, which detects volatile pheromone cues, plays an important role in courtship behavior. Dysfunction of the primary pheromone-sensing ORNs impairs pheromone detection and results in aberrant courtship behavior [17]. Additionally, activity of higher-order olfactory neurons, where information from ORNs is processed and modified, may also influence Drosophila courtship behavior.
In this study, we first examined the endogenous expression pattern of the painless gene in the adult Drosophila brain. Analysis of the painless mutant flies led to the observation that their olfactory sensitivity was decreased. In addition, we identified abnormal male-male courtship behavior. Manipulating of painless expression in specific brain regions revealed that its expression in the PNs of the ALs was critical for preventing the male-male courtship behavior. Further experiments using time-specific knockdown of painless under the regulation of the temperature-sensitive Gal80 protein (Gal80 ts ) showed that painless expression in the adult brain was essential for normal courtship behavior. Thus, painless may function in PNs to regulate male courtship behavior.

Drosophila Stocks and Culture
Flies were cultured on standard cornmeal-agar-molasses medium at 25uC with a relative humidity of 50-70%. The painless 1 , painless-Gal4, and UAS-painless AR9 lines generated in the w1118 background were a generous gift from Dr. D. Tracey (Duke University, Durham, NC, USA). The CS strain was used as a WT control. The painless 3 and painless 4 lines in a w1118 background were obtained from the Szeged Drosophila Stock Center (Szeged, Hungary). The GH146-Gal4, Poxn-Gal4-13-1, c507-Gal4, and UASshibire ts lines were kindly provided by Drs. R.F. Stocker

Behavioral Assays
All behavioral assays were carried out at 2561uC with 40-60% relative humility at zeitgeber time ZT1-ZT8, except where otherwise indicated. Flies used in behavioral assays were collected by gentle aspiration or under light CO 2 anesthesia shortly after eclosion (within 6 h), and aged 4-6 days in all the experiment except that of RNAi experiment and Gal80 ts experiment. Males were reared in groups of 8-10 animals or individually [18,19], while females were kept in groups of 10 animals.
Chaining Assay. Eight males reared in groups were introduced into a 3.5 cm plastic dish by gentle aspiration or after immobilization by anesthetics. The courtship behavior between these males was video recorded and analyzed manually. The chaining behavior was defined as the display of courtship behavior among at least three males, with formation of a chain. The chaining index (ChI) is the percentage of time during which chaining behavior was observed [20]. For the antenna ablation experiment, the antenna (antennae) of males was (were) carefully removed by using fine forceps under light CO 2 anesthesia in the first day after eclosion. Males were left to recover for 4 days before the chaining assay.
Courtship Pairing Assay. One individually reared tester male and one target female were sequentially introduced into a courtship chamber (diameter, 1 cm; height, 0.3 cm) by gentle aspiration, and the courtship behavior between the pair was video recorded and analyzed manually. The courtship index (CI) is the percentage of time in a given observation period during which the tester exhibits courtship behavior towards the target, which includes chasing, wing vibration, tapping, and attempted copulation [20].
Courtship Preference Assay. One individually reared tester male was introduced into a courtship chamber by gentle aspiration or by light CO 2 anesthesia, and was left alone for 5 min. Two decapitated target flies, i.e., a WT male and a WT female, were then presented to the tester. The CI values towards the female and the male during a 10 min period were simultaneously measured and were subsequently compared [21].
Olfactory Sensitivity. Olfactory sensitivity was measured using a Tmaze test [22,23]. Briefly, flies were collected shortly after eclosion without anesthesia and were reared in groups until they were 2-3 days old. After being food deprived and kept in darkness for 0.5 h, ,100 flies were introduced into the elevator of the T-maze and were left to choose between the two tubes, one filled with air and the other with MCH (Fluka, USA), for 2 min. The flies in each tube were then counted and the PI was calculated. All flies were reared on food without propionic acid and yeast, and the experiments were performed at 2561uC with 75-90% relative humility in exclusive darkness with a red LED as a light source.

Blocking Neurotransmission using shibire ts
For the experiments that required the use of UAS-shibire ts , flies were kept at 19uC until the experiment. To shift the environment temperature, flies were housed in 3.5 cm dishes made of aluminum and placed on a custom-built heating plate that can shift between 19 and 30uC within 1 min.

Data analysis
Student's t-test was used to compare the intensity of courtship behaviour, after testing the normality of the data distribution with the Kolmogorov-Smirnov test. Otherwise, the Kruskal-Wallis test was used.

Painless Is Expressed in Adult Fly Brain
The Gal4 expression pattern in the Gal4 enhancer-trap line pain Gal4 , which contains Gal4 inserted within the painless transcriptional unit [6], revealed that this Gal4 is broadly expressed in the adult brain [10]. Using the same Gal4 line, we found that painless is expressed in several clusters of PNs of the ALs, in Kenyon cells (KCs) of mushroom bodies (MBs), and in the neurons that form the ellipsoid body ( Figure 1A). To confirm the endogenous expression of painless in the adult fly brain, we performed RT-PCR to examine the presence of the painless mRNA. As shown in Figure 1B, painless mRNA was detected in dissected brains of WT Canton-S (CS) adult flies, indicating the endogenous expression of the painless gene in the adult brain. We further confirmed the endogenous expression of painless by immunostaining. The specificity of the Painless antibody was first validated by specific staining of the Painless-myc fusion protein in cell lines overexpressing the protein, and detection of overexpression Painless protein in adult brains ( Figure S1). We next examined the expression pattern of endogenous painless in adult brains using GFP expressed in PNs of the ALs (pain Gal4 ; UAS-mGFP and GH146 Gal4 ; UAS-mGFP). As shown in Figures 1C 1 and 1C 2 , Painless signal was detected in the circumference of the ALs, where the somata of olfactory PNs and LNs are located, and co-staining experiments showed that Painless was expressed in certain PNs. Also, we observed that some glomeruli, which receive the dendrites of PNs, were labeled by the antibody (Figure S1), suggesting a function of Painless in the dendrites. Although strong Painless signal was not detected in other brain regions, such as MBs, we could not exclude the possibility that Painless was also expressed in KCs and other neurons at relatively low levels.

Olfaction Is Affected in painless Mutant Flies
As expression of endogenous Painless was detected in the PNs, we speculated that painless may function in the olfactory pathway, and contribute to the olfactory behavior. In order to test this possibility, we examined the olfactory sensitivity of painless mutant flies using the T-maze test [22,23]. As shown in Figure S2, when provided with the odorant 4-methyl-cyclohexanol (MCH), WT flies preferred the tube perfused with air to that perfused with MCH. And as the concentration of MCH increased, the MCH avoidance behavior became stronger. Interestingly, this MCH avoidance behavior was significantly weakened in painless mutant flies (pain Gal4 and pain 1 ), suggesting that olfactory sensitivity of these flies was affected. When MCH was provided at lower concentrations (5.0610 26 and 5610 27 , v/v dilution), painless mutant flies still displayed a preference behavior, suggesting an alteration in the responses towards this odor. To further test whether this olfactory defect was due to mutation of painless, we examined the avoidance behavior of painless mutant flies overexpressing Painless in pain Gal4 positive neurons. We found that the avoidance behavior could be partially restored to a level resembling that of WT flies at high concentrations of MCH, but not at low concentrations ( Figure S2). These results indicate that painless is involved in olfactory processing in flies.

Male-male Courtship Behavior in pain Gal4 Flies
In addition to the olfactory defects, we observed that male pain Gal4 flies, in which the expression of painless was interfered by the Gal4 insertion, spontaneously displayed a male-male courtship behavior in the morning (zeitgeber time: ZT0.5-ZT3) ( Figure 2A). We measured the chaining index (ChI) of the males, i.e., the percentage of time spent courting one another and forming chains. We found that the ChI of male pain Gal4 flies was ,10%, significantly higher than that of WT males (,0%). We fortuitously observed that, after recovery from a mild anesthesia (brief treatments with CO 2 , nitrogen, or 220uC chilling), the intensity of this male-male courtship behavior was greatly enhanced over a long period (up to ,3 h after CO 2 treatment, Figure 2B). Interestingly, this anesthesia-induced enhancement of the malemale courtship behavior was not specific to painless mutant flies, as a brief anesthesia with either CO 2 or nitrogen also induced a partial male-male courtship behavior lasting a very short period (0.3-0.5 s) in WT flies, i.e., an increase in the occurrence of courtship initiation ( Figure 2C), which included the orientation, chasing, and wing vibration steps, but not followed by other courtship steps as in the painless mutant males. These observations suggest that male-male courtship behavior could be triggered among flies after recovery from a treatment of mild anesthesia, and WT males are able to inhibit this abnormal courtship behavior whereas painless mutant males cannot. Thus, in the following experiments, we treated male flies with brief CO 2 anesthesia (10 s or 1 min) before the chaining assay to examine their ability to inhibit male-male courtship behavior.

Male-male Courtship Behavior Is Caused by painless Mutation
To confirm that the male-male courtship behavior of pain Gal4 flies was caused by mutation of the painless gene, we examined the behavior of other mutant alleles of painless. We found that painless 1 (pain 1 ) males exhibited long-lasting male-male courtship behavior after recovery from mild CO 2 anesthesia, with a ChI lower than that of pain Gal4 males, but significantly higher than that of WT males ( Figure 3A). In addition, we performed complementation experiments and found strong male-male courtship behavior in pain Gal4 /pain 1 double-heterozygous flies, with an average ChI as high as that observed in pain Gal4 homozygous males. In contrast, pain Gal4 /+ and pain 1 /+ heterozygous male flies did not display obvious male-male courtship behavior. In addition to pain Gal4 / pain 1 males, we also examined the behavior of pain 1 /pain 3 , pain Gal4 / pain 3 , and pain Gal4 /pain 4 heterozygous males and found that the intensity of their male-male courtship behavior was significantly higher than that of WT animals ( Figure S3), but shorter and weaker as compared with pain Gal4 /pain 1 males.
Further rescue experiments showed that overexpression of Painless suppressed the male-male courtship behavior of pain Gal4 ; UAS-Painless/+ flies ( Figure 3B). Moreover, the courting ability of these Painless-overexpressing males towards WT virgin females was similar to that of WT males ( Figure 3C). These results demonstrate that the abnormal courtship behavior found in pain Gal4 mutant males was due to the painless mutation.

Courtship Preference of pain Gal4 Males towards Females
As the olfactory sensitivity to MCH was decreased in painless mutant flies, we next assessed whether the sensitivity of these flies to specific pheromones was also altered. If so, this could affect the ability of males to distinguish females from males and, thus be a cause for abnormal courtship behavior. We performed a courtship preference assay by measuring the preference index (PI) of painless mutant males towards decapitated male and female flies. As shown in Figure 4A, when placed with two decapitated WT flies of opposite genders, pain Gal4 males and WT males displayed a similar delay in courtship initiation, suggesting that pain Gal4 males were capable of sensing the targets and of initiating courtship behavior.
We then examined the time spent by the pain Gal4 male and WT males courting the two decapitated targets and found that pain Gal4 males were able to distinguish female from male flies, although the PI was slightly lower as compared with that of WT flies ( Figure 4B and 4C). These results suggest that the male-male courtship behavior of painless mutant flies was unlikely to be caused exclusively by their inability to differentiate between female and male flies. In addition, we noticed that the total time spent by pain Gal4 males on courting was significantly longer than that observed for WT males ( Figure 4B).
We also performed the experiment under dim red light to remove effects of the visual cues. We found that under such condition, both WT males and pain Gal4 mutant males showed a shorter courtship delay ( Figure 4D). In addition, WT males spent more time in courting the male target ( Figure 4E), suggesting that visual input is important for the WT males to distinguish males from females. In contrast, pain Gal4 mutant males showed similar sex-discriminating ability as that under white light ( Figure 4B and 4F). Thus, it is possible that pain Gal4 males are incapable of sensing inhibitory visual cues, or pain Gal4 males have additional deficiencies which mask the vision-deprivation effect.

Painless Expression in PNs Inhibits Male-male Courtship Behavior
The above-described results showed that disruption of the painless gene led to decreased olfactory sensitivity and male-male courtship behavior. However, the brain regions in which Painless expression was required for the prevention of this aberrant behavior are unknown. The expression pattern of painless suggested a role for painless in olfactory processing. Therefore, we manipulated Painless expression in the olfactory PNs of the ALs. We first used two RNAi transgenic lines targeting painless [24] (UAS-pain-RNAi-1 and UAS-pain-RNAi-2) to downregulate Painless expression globally under the control of pain Gal4 . As shown in Figures 5A and Figure S4, global downregulation of Painless expression in pain Gal4 -positive neurons using either RNAi line induced male-male courtship behavior, while pain Gal4 /+ heterozygous flies and the two UAS-pain-RNAi-alone flies did not exhibit this behavior. These results suggest that these two RNAi lines effectively downregulated Painless expression and induced the male-male courtship behavior.
We next used different Gal4 lines to knockdown Painless expression preferentially in selective subsets of neurons in an attempt to localize the site of action of Painless during the regulation of courtship behavior. We found that preferential knockdown of Painless expression in about two thirds of the olfactory PNs [25] using a GH146 Gal4 -driven RNAi caused severe male-male courtship behavior ( Figure 5A). Downregulation of Painless expression in certain gustatory receptor neurons (GRNs) using RNAi driven by Gr66a Gal4 , Poxn Gal4 , or MD Gal4 resulted in weak male-male courtship behavior ( Figure 5B). As KCs in MBs and the neurons in the central complex were pain Gal4 -positive, we further downregulated Painless expression in these two clusters of neurons by driving RNAi expression under the control of OK107 Gal4 and C507 Gal4 , respectively. No significant male-male courtship behavior was observed in these two fly genotypes ( Figure 5B).
In parallel, we conducted rescue experiments to examine whether preferential overexpression of Painless in PNs or GRNs prevented the male-male courtship behavior caused by the painless mutation. As shown in Figures 5C and 5D, the intensity and duration of the male-male courtship behavior in pain 1 flies were significantly reduced following preferential expression of Painless in the GH146 Gal4 labeled PNs. However, overexpression of Painless in Gr66a Gal4 -positive neurons did not prevent male-male courtship behavior. These results of the cell-type-specific knockdown and rescue experiments suggest that the expression of Painless in PNs is necessary and sufficient for the suppression of male-male courtship behavior.

Male-male Courtship Behavior in pain Gal4 Flies Is Not Caused by Developmental Defects
To determine whether the male-male courtship behavior is caused by developmental defects associated with painless mutation, we used the temporal and regional gene-expression targeting (TARGET) system to switch on RNAi expression in adult flies in a time-specific manner [26]. We used the temperature-sensitive Gal80 protein (Gal80 ts ) to manipulate the activity of Gal4. The effectiveness of Gal80 ts was shown by the following experiments. The maintenance of flies carrying a copy of Gal80 ts at a permissive temperature (19uC) led to negligible detection of GFP signal in the brain of painGal4;UAS-mGFP/TubP-Gal80 ts animals. However, the transfer of the flies to the restrictive temperature (30uC) for 5-6 additional days restored GFP expression ( Figure S5). We then used the TARGET system to switch on RNAi expression in adult flies. As shown schematically in Figure 6A, we inhibited the expression of UAS-pain-RNAi during early development by maintaining the flies at 19uC until 5-6 days after eclosion to ensure their proper development, and then relieved the inhibiting activity of Gal80 to restore the RNAi expression by elevating the temperature to 30uC for additional 5-6 days. Male-male courtship behavior was observed in these flies, with intensity and duration similar to those observed for the RNAi flies not expressing Gal80 ts ( Figure 6B). This result indicates that downregulation of painless in adult flies was sufficient to induce male-male courtship behavior.

Discussion
To study the physiological role of Painless in the central nervous system of adult fruit flies, we examined the behavior of painless mutant flies. We found that the painless mutation significantly increased male-male courtship behavior (Figures 2 and 3). By altering the expression level of Painless in a subpopulation of PNs in ALs, we were able to induce this male-male courtship behavior, suggesting that Painless expression in the olfactory system was necessary for inhibition of this abnormal courtship behavior ( Figure 5). Furthermore, temporally specific Painless knockdown in the adult fly was sufficient to induce the male-male courtship phenotype ( Figure 6). Moreover, mutation of the painless gene reduced olfactory sensitivity ( Figure S2). Together, our results suggest that the expression of Drosophila Painless in a subset of PNs is essential for normal male courtship behavior. Because PNs are the principal relay neurons of the Drosophila olfactory system, Painless may play an important role in gating olfactory information; thus, mutation of this protein may lead to deficits in odor/pheromone perception and aberrant courtship behaviors.

Effects of Manipulating Painless Expression
In this study, the behaviors of several painless mutants were examined, and the expression level and patterns of painless were manipulated by using UAS-pain-RNAi and UAS-Painless driven by different Gal4 strains. We found that painless mutant males exhibited male-male courtship behavior (Figure 2 and Figure S3), at various intensities depending on the genotype. The painless mutants we used (pain 1 , pain 3 and pain Gal4 ) have been reported to have P-element insertions immediate upstream of the first noncoding exon of painless gene, resulting in the expression of mutant Painless proteins [6]. Thus, it is possible that the male-male courtship behavior observed in painless mutants is due to loss of WT Painless protein function, or the expression of mutant Painless proteins, or both. Our results showing that overexpression of WT Painless in the mutant background significantly inhibited malemale courtship behavior ( Figure 3B and Figure 5C), and that knockdown of WT Painless expression induced male-male courtship behavior ( Figure 5 and Figure S4) indicate that the loss of WT Painelss function is likely the major cause of this abnormal courtship behavior. However, we cannot exclude the possibility that the mutant Painless proteins resulting from the P-element insertion also contributes to the male-male courtship behavior, as the intensity of this behavior differed between painless mutant alleles.
By preferentially down-regulating Painless expression in different subsets of neurons, we found that the expression of painless in PNs was essential for the inhibiting courtship behavior among males. However, the contribution of painless expression in other neurons cannot be excluded. As shown in Figure 5C, preferential overexpression of painless in PNs of painless mutants did not fully suppress the abnormal courtship behavior. In addition, downregulation of painless expression in neurons (Gr66a Gal4 , Poxn Gal4 , and MD Gal4 ) other than PNs resulted in a mild male-male courtship behavior ( Figure 5B). As Painless is expressed in a subset of GRNs that overlaps with the Gr66a-expressing neurons [10], involved in the detection of bitter substances and certain cuticular pheromones [27][28][29], we speculate that painless may also contribute to the sensation of these non-volatile pheromones by the Drosophila gustatory system [30][31][32].

Role of Painless in Gender Preference in flies
Male flies use multiple sensory modalities to discriminate female from males. When confronting with two targets of opposite genders, WT males spent almost all their time courting the female target, while painless mutant males also courted the male target ( Figure 4). This observation suggests that painless mutant males have deficency in sensing the inhibitory cues on the male targets. Combined with results obtained from down-or up-regulating Painless expression in subsets of neurons ( Figure 5), we surmise that painless expression in the olfactory system is involved in sensing cues. Meanwhile, the gender-preference assay revealed that painless mutants exhibited a longer courtship period towards both females and males as compared with WT flies (Figure 4), suggesting a general enhancement of courtship activity. This alteration is also suggested by the finding that painless mutant females exhibit enhanced sexual receptivity [33]. Thus, painless may play a role both in perception of inhibitory cues from males and in gating the intensity of the courtship behavior.

Effects of Anesthesia on Courtship Behavior
It is known that anesthesia affects the courtship behavior of individual males towards females [34]. The copulation latency of WT flies becomes longer after recovery from CO 2 or chilling anesthesia [34]. Consistently, we found that brief anesthesia led to a significant decrease in the intensity of the courtship behavior of individual males (both WT males and painless mutants) towards decapitated females (Figure 4). However, the courtship behavior of an individual male towards another decapitated male was unchanged, while this behavior among a group of males was markedly enhanced. One explanation for these differential effects of anesthesia on male-male and male-female courtship behaviors is that anesthesia could weaken the sensation of attractive or repulsive pheromonal cues from females or males, respectively.
Distinct parts of the nervous system have differential sensitivities to anesthetics [35]. We speculate that in males there are neural circuits for sensing cues from other males to inhibit the male-male courtship behavior, and these circuits are highly sensitive to anesthetics. In addition, there are neural components which are responsible for the exhibition of courtship behavior, which are less sensitive than the inhibitory circuits. In the presence of mild anesthetics, e.g., brief CO 2 , nitrogen, and chilling used in the present study, the inhibitory neural circuits may loss their function more easily than the neural circuits responsible for courting, causing the animals to exhibit male-male courting. As the dose of anesthetics increased, the neural circuits responsible for courting were also disabled, resulting in decreases and eventually abolishment of courtship behavior. ( Figure 2B).
We also noticed that after recovery from an anesthesia of same dose, two painless mutants with the same P-element insertion (pain 1 and pain Gal4 ) exhibited male-male courtship behavior of distinct intensities and temporal patterns ( Figure 3). This is probably due to differences between the natures of these two painless mutant alleles (e.g. motility, vision). Furthermore, this might be due to the different sensitivity to anesthesia in pain 1 and pain Gal4 mutants, as the recovery times from anesthesia in pain 1 was longer than that in pain Gal4 males ( Figure S6).
We also tested the idea whether anesthesia-sensitive neural circuits have some overlap with the painless circuits. By timespecific blockade of the neural transmission of pain Gal4 -expressing neurons with a temperature-sensitive Shibire protein [36], a malemale courtship behavior could be triggered in the absence of anesthetics ( Figure S7), suggesting that Painless might function in some courtship-inhibiting neural circuits which are also anesthesia-sensitive. Therefore, the enhanced courtship behavior observed among the painless mutant males after recovery from anesthesia should comprise integrated effects of both the impairment of painless function and anesthesia.
The mating behavior of painless mutant male flies towards other male flies may be caused by removal of the inhibition mechanism that prevents a male fly from pursuing other male flies. Removal of the inhibition mechanism in painless mutant flies is probably caused by their inability to sense the signal from other male flies correctly. Our results showing that the loss of function of the painless gene in PNs was essential for male-male courtship behavior, and that olfactory sensitivity was decreased in painless mutant flies, suggest that olfactory perception in painless mutant flies is impaired and may result in their inability to perceive inhibitory chemical cues (e.g. cis-Vaccenyl acetate) from other male flies effectively [30,37,38]. As one of the members of the TrpA family of ion channels, the expression of Painless in PNs may contribute to the electrophysiological properties of these neurons (e.g., the excitability and firing pattern of PNs). Thus, we speculate that painless may be involved in the coding of the olfactory/pheromone signals in PNs, which are important for the inhibition of male-male mating.   Figure S7 Blockade of the neurotransmission of the pain Gal4 -positive neurons resulted in the male-male courtship behavior. The temperature was firstly shifted from 19uC to 30uC, and after maintaining for a period, was shifted back to 19uC. The behavior between eight males of indicated genotypes at either 19uC or 30uC were analyzed. Histograms show the average ChI, and error bars mean SEM. For each genotype, more than eight groups were analyzed. (TIF)