neurons to promote avoidance of UV during egg-laying

Transcription

1 Genetics: Early Online, published on December 7, 2016 as /genetics Title: H 2 O 2 -sensitive isoforms of Drosophila TRPA1 act in bitter-sensing gustatory neurons to promote avoidance of UV during egg-laying 3 Running title: UV-sensitive dtrpa1 promotes UV avoidance Authors: Ananya R. Guntur 1*, Bin Gou 1*, Pengyu Gu 2, Ruo He 1, Ulrich Stern 3, Yang Xiang 2, Chung-Hui Yang 1 1. Department of Neurobiology, Duke University Medical Center, Durham, NC Department of Neurobiology, University of Massachusetts Medical School, Worcester, MA Independent researcher * These two authors contributed equally Author of correspondence: Chung-Hui Yang, Room 427E Bryan Research Building, 311 Research Drive, Duke University Medical School, Durham, NC yang@neuro.duke.edu Key words: Drosophila egg-laying, UV-sensing, dtrpa1, bitter-sensing neurons 1 Copyright 2016.

2 Abstract The evolutionarily conserved TRPA1 channel can sense various stimuli including temperatures and chemical irritants. Recent results have suggested that specific isoforms of Drosophila TRPA1 (dtrpa1) are UV sensitive and that their UV sensitivity is due to H 2 O 2 sensitivity. However, whether such UV sensitivity serves any physiological purposes in behaving animals was unclear. Here, we demonstrate that H 2 O 2 -sensitive dtrpa1 isoforms promote avoidance of UV when adult Drosophila females are selecting sites for egg-laying. First, we show that blind/visionless females are still capable of sensing and avoiding UV during egg-laying when intensity of UV is high yet within the range of natural sunlight. Second, we show that such vision-independent UV avoidance is mediated by a group of bitter-sensing neurons on the proboscis that express H 2 O 2 - sensitive dtrpa1 isoforms. We show that these bitter-sensing neurons exhibit dtrpa1- dependent UV sensitivity. Importantly, inhibiting activities of these bitter-sensing neurons, reducing their dtrpa1 expression, or reducing their H 2 O 2 -sensitivity all significantly reduced blind females UV avoidance whereas selectively restoring a H 2 O 2 - sensitive isoform of dtrpa1 in these neurons restored UV avoidance. Lastly, we show that expressing the red-shifted channelrhodopsin CsChrimson specifically in these bittersensing neurons promotes egg-laying avoidance of red light, an otherwise neutral cue for egg-laying females. Together, these results demonstrate a physiological role of the UV-sensitive dtrpa1 isoforms, reveal that adult Drosophila possess at least two sensory systems for detecting UV, and uncover an unexpected role of bitter-sensing taste neurons in UV sensing. 38 2

3 Introduction TRPA1 is a member of the evolutionarily conserved TRP family of non-selective cation channels and has been shown to play important roles in sensory functions across species (Clapham, 2003; Jordt et al., 2004; Julius, 2013). For example, previous studies have shown that TRPA1 can directly sense temperatures as well as chemical irritants such as AITC, an active ingredient of mustard oil, in both vertebrates and invertebrates (Chatzigeorgiou et al., 2010; Cordero-Morales et al., 2011; Hamada et al., 2008; Kang et al., 2010; Rosenzweig et al., 2005; Story et al., 2003; Viswanath et al., 2003). In addition, TRPA1 can also act downstream of specific sensory receptors in some sensory neurons (Bellono et al., 2013; Kim et al., 2010; Kwon et al., 2010; Shen et al., 2011; Xiang et al., 2010). For example, TRPA1 has been suggested to transduce chronic itch signal sensed by other receptors in the rodent DRG neurons (Morita et al., 2015), as well as to transduce light signal downstream of the Gr28b receptor in Drosophila polymodal somatosensory C4da neurons (Xiang et al., 2010). We have recently expanded the repertoire of TRPA1 functions by demonstrating that some of the dtrpa1 isoforms in Drosophila are H 2 O 2 -sensitive and can sense blue and UV lights (Guntur et al., 2015). We found that ectopically expressing these dtrpa1 isoforms can confer robust UV-induced calcium responses to light-insensitive cultured HEK 293 cells (Guntur et al., 2015). More strikingly, ectopic expression of these isoforms in two different groups of light-insensitive motor neurons in adult Drosophila enabled muscle contraction in response to UV (Guntur et al., 2015). Further, dtrpa1 isoforms that lacked H 2 O 2 sensitivity cannot confer UV sensitivity and that overexpression of 3

4 catalase, an enzyme that degrades H 2 O 2, significantly reduced UV sensitivity conferred by H 2 O 2 -sensitive dtrpa1 (Guntur et al., 2015), suggesting that H 2 O 2 -sensitive dtrpa1 isoforms can sense UV due to their capacity to detect UV-induced reactive oxygen species (ROS) production (Bhatla and Horvitz, 2015; Guntur et al., 2015; Hockberger et al., 1999). However, despite these results, the physiological purpose of UV sensitivity of H 2 O 2 -sensitive dtrpa1 isoforms in behaving animals is not well understood. In this work, we provide evidence suggesting that UV sensitivity of the H 2 O 2 - sensitive dtrpa1 isoforms plays an important role in promoting avoidance of UV during female egg-laying. We first discovered that whereas blind (visionless) females were unable to lay eggs away from low intensity UV as previously reported (Zhu et al., 2014), they exhibited a clear avoidance of UV for egg-laying when the level of UV was elevated but still within the range of natural sunlight. Further, we found surprisingly that such vision-independent UV avoidance was primarily mediated by a group of bitter-sensing neurons on the proboscis. These gustatory neurons expressed the H 2 O 2 -sensitive dtrpa1 isoforms and exhibited dtrpa1-dependent UV sensitivity. Importantly, reducing dtrpa1 expression in these neurons, reducing their H 2 O 2 sensitivity, or inhibiting their neuronal activities all significantly reduced blind Drosophila females UV avoidance for egg-laying, whereas selectively restoring the H 2 O 2 -sensitive isoform in these neurons rescued UV avoidance. Lastly, we showed that optogenetic activation of these dtrpa1-expressing bitter-sensing neurons was sufficient to promote egg-laying avoidance: expressing the red-shifted channelrhodopsin CsChrimson (Klapoetke et al., 2014) in these neurons caused females to lay eggs away from red light an otherwise 4

5 neutral stimulus for egg-laying females. Taken together, our results show that UV sensitivity of the H 2 O 2 -sensitive dtrpa1 isoforms plays an important role in ensuring that Drosophila females can continue to deposit theirs eggs away from an aversive and damaging stimulus for their progenies when their vision fails or is temporarily blocked, and that specific dtrpa1-expressing bitter-sensing neurons on the proboscis are recruited to accomplish this task. 89 5

6 Materials and Methods Fly stocks Animals were raised in standard molasses/cornmeal fly food and kept in a Darwin Chamber with temperature typically set at 25 C and humidity level at 60-65%. The following stocks were used in this work: w 1118, norpa 36 (BL-9048), dtrpa1 KO (Hamada et al., 2008), Hdc JK910 (Burg et al., 1993), UAS-catalase (BL-24621), UAS-dTRPA1(A)10a (Guntur et al., 2015), UAS-dTRPA1-RNAi (Hamada et al., 2008), Gr66a-Gal4 (Wang et al., 2004), Gr66a-lexA (Thistle et al., 2012), dtrpa1-gal4 (Petersen and Stowers, 2011), lexa-op2-flp (BL-55819), tub>gal80> (Gordon and Scott, 2009), UAS-GCaMP6s (BL ), Gr5a-lexA (Gordon and Scott, 2009), lexa-op2-gcamp6(s) (BL-44589), UAS-Kir 2.1 (Baines et al., 2001), UAS-CsChrimson (BL-55134). Reverse transcription PCR RNA was extracted from whole adults or from dissected labellum using TRIzol and RNeasy plus micro kit (Qiagen). First-strand cdna was synthesized by using SuperScript III RT kit (Invitrogen) with oligo dt primer. EmeraldAmp Max HS PCR Master Mix (TAKARA) was used for amplifying the dtrpa1 isoform specific products. PCR Program was set as below: 94 C for 5 min, 94 C for 20 s, 55 C for 20 s, 72 C for 2 min, repeat steps 2 4 for 29 times (30 cycles in total), followed by 72 C, 7min. For discriminating different dtrpa1 isoforms, the primer pairs used were as followed: (i) dtrpa1(a)10a: A- F/10a-R, (ii) dtrpa1(a)10b: A-F/10b-R. Sequences of the primers described are as follows (5 to 3 ): A-F: GCCGGAACAGCAAGTATT 6

7 a-R: CCATGTGTTACCATGGTATTCAAAG 10b-R: TGTTACCATGGTGTTCACCA Egg-laying assay UV vs. Dark assay. To assess how females respond to UV when selecting for egg-laying site, we assayed their preference in our custom egg-laying chambers that were fitted with LED holders. The general design of the egg-laying chambers and UV light control were as previously described (Zhu et al., 2014) but with two modifications. First, the UV (380 nm) intensity was set at 6 µw/mm 2 instead of < 0.5 µw/mm 2. (Note that we chose 6 µw/mm 2 because it is the highest level of UV our controller can deliver with the LEDs we have purchased). Second, to maintain adequate amount of egg-laying, we directly added 150 mm sucrose (a food source) into the 1% agarose as opposed to placing a drop of grape juice into hole in the middle of the chamber. Red LED vs. Dark assay. To assess how females respond to red light when expressing CsChrimson in specific groups of neurons, we replaced the UV LED with red light LED (Mouser Electronics, part # C503B-RAN-CA0B0AA1) in the light holder. The intensity of red light was set at 2.3 or 4.5 µw/mm 2. Animal preparation. To maintain adequate levels of egg-laying rate, we typically prepared the females by depriving them of egg-laying using the previously described method (Gou et al., 2016; Zhu et al., 2014). Briefly, for every ~ 30 females, we paired them with ~ 15 males and put them in a regular food vial that was supplemented with some active yeast paste (made with mixing 3 grams of active yeast with 5 ml of 0.3 % of propionic acid), and we let the females mate and lay eggs in the vial for about 4-5 days. Females would lay many eggs initially due to the presence of yeast paste and 7

8 thus caused the surface of the food vial to become very mushy, a texture that deters egg-laying and thus causes females to start withholding eggs. After withholding eggs for several days, these females would readily lay eggs when placed in our egg-laying chambers. For females to be assayed in red LED vs. dark assay, we supplemented the active yeast paste and the fly food with 500 µm all-trans retinal (Sigma Aldrich), the chromophore for activating CsChrimson. We typically let the females lay eggs overnight in our egg-laying chambers and calculate the preference index (PI) according to the following formula: (N LED -N Dark )/(N LED +N Dark ). N LED and N Dark represent the number of egg on the LED and Dark sides, respectively. Behavioral tracking Females to be tracked were fed active yeast paste and deprived of egg-laying as described previously so that they were in egg-laying state (Gou et al., 2014; Gou et al., 2016). They were then placed in egg-laying chambers that contained two sucrose substrates (prepared as described earlier) but UV-LEDs were mounted below the egglaying chamber so that cameras can be mounted above to videotape females behaviors (see also Zhu et al., 2014). We typically recorded the females for two hours and then analyzed the periods in which they did not lay eggs. This is because we wanted to determine whether females in egg-laying state showed signs of positional avoidance of UV. Selected videos were then tracked using Ctrax (Branson et al., 2009), and the tracked trajectories analyzed using custom MATLAB code (available upon request). Positional preference index (PI) was calculated according to the following formula: (N LED - 8

9 N Dark )/(N LED +N Dark ). N LED and N Dark represent the number of frames females spent on the LED and Dark sides, respectively. Ca 2+ imaging Fly preparation. To image calcium responses of gustatory neurons on the labellum, we severed the proboscis and placed them in a custom-built imaging chamber (Gou et al., 2014) that was filled with imaging buffer. The imaging buffer contains the following chemicals: NaCl (120 mm), KCl (3 mm), MgCl 2 (4 mm), CaCl 2 (2 mm), NaHCO 3 (10 mm), trehalose (10 mm), glucose (10 mm), TES (5 mm), sucrose (10 mm), HEPES (10mM), and we adjust the final ph to 7.25). UV stimulation. We used X-cite lamp as the light source and we used filter set # 49 from Zeiss ( nm, peak at 365 nm) to produce UV. Light intensity was measured with a Thorlabs PM100USB optical power meter. Signal acquisition. To acquire GCaMP fluorescence change, we used a Zeiss LSM700 confocal microscope equipped with a 40X/0.8 NA water immersion objective and the live-series ZEN image acquisition software. Briefly, baseline GCaMP fluorescence was acquired by scanning the cells with 488 nm laser at 128 X 128 pixels at 8-bit dynamic range. The GCaMP6 signal acquired before light stimulation (baseline) was used as F 0 and the signal after 5 light stimulation was used as F. ΔF/F was then calculated as (F F 0 ) /F 0 to reflect the changes in GCaMP signals before and after stimulation. For each experiment, we examined GCaMP responses from 4 to 10 animals. Because each animal has multiple GCaMP positive cells (~7-18 cells), we calculated the GCaMP response for each experiment by first obtaining an averaged ΔF/F for each animal. For analyzing caffeine and H 2 O 2 induced responses ImageJ software (NIH) was used to register images and 9

10 calculate fluorescence in assigned regions of interest. The GCaMP6 signal acquired before addition of caffeine and H 2 O 2 was used as F 0 and the peak change in fluorescence after addition of H 2 O 2 was used as F. ΔF/F was then calculated as (F F 0 )/F 0 to reflect the change in GCaMP6 after H 2 O 2 stimulation. Data availability Drosophila lines are available at Bloomington Drosophila Stock Center or upon request

11 Results Blind females avoid laying eggs on UV site when the level of UV is close to that of sunlight We have previously shown that Drosophila females robustly avoided laying eggs on the UV-illuminated site when given a choice between an unilluminated (dark) option and a UV-illuminated one in our two-choice egg-laying paradigm (Zhu et al., 2014). This avoidance of UV for egg-laying critically depended on phototransduction in the compound eyes: animals that lacked the phototransduction molecule Phospholipase C (norpa mutants) no longer avoided UV whereas selective rescue of norpa function in the UV-sensing R7 photoreceptors restored UV avoidance (Zhu et al., 2014). However, the intensity of UV we used in that study (< 0.5 μw/mm 2 ) was significantly lower than that of the UV component of sunlight received on the Earth s surface (~ 25 μw/mm 2 ), as most studies on UV-induced behaviors in adult flies employ very low intensity of UV (Gao et al., 2008; Karuppudurai et al., 2014). We therefore wondered whether Drosophila females continue to rely exclusively on their visual system to sense and avoid UV during egg-laying when the level of UV intensity is closer to that of sunlight. Notably, a previous study has suggested that Drosophila larvae use Bolwig organ (the primitive larval eye) for detecting low levels of light but they recruit additional sensors the polymodal somatosensory C4da neurons on the body wall for detecting high levels of blue and UV lights (Xiang et al., 2010). To test the idea that adult Drosophila may recruit extraocular UV sensors to avoid higher but physiologically relevant levels of UV during egg-laying, we first assessed 11

12 egg-laying preference of blind (norpa 36 ) females in a new UV vs. dark two-choice task where we raised the UV intensity to 6 μw/mm 2 (Figure 1A). Interestingly, we found that whereas blind (norpa 36 ) females failed to avoid very low levels of UV (< 0.5 μw/mm 2 ) as reported previously (Zhu et al., 2014), they consistently showed a moderate but significant avoidance of a higher level of UV (6 μw/mm 2 ) for egg-laying (Figure 1B-D). To further confirm this result, we examined egg-laying preference of another blind mutant Hdc JK910 that cannot produce histamine, the critical neurotransmitter by which photoreceptors in the eyes signal their targets (Burg et al., 1993). Again, Hdc mutants showed a moderate but significant avoidance of UV for egg-laying (Figure 1D). These results suggest that adult Drosophila indeed possess extraocular UV sensors that enable them to avoid laying eggs on UV sites (where the level of UV is higher than what we had used in our earlier study but is still physiologically relevant) Blind females preference to lay eggs away from UV is the result of positional avoidance of UV We next assessed the behavioral mechanism that underlies blind females avoidance of laying eggs on site illuminated with high UV. There are at least two possible explanations. First, egg-laying females avoid approaching a site illuminated with UV. Second, egg-laying females readily approach the UV site but selectively withhold depositing eggs while visiting it. To distinguish between these two possibilities, we tracked the behaviors of egg-laying females as they explored the UV vs. dark chamber. Plotting and analyzing the trajectories of egg-laying wild-type (w 1118 ) females revealed that they tended to spend times away from the UV site (Figure 2A, C). 12

13 Further, analysis of trajectories of blind (norpa) egg-laying females revealed that they also exhibited a moderate but significant positional avoidance of UV despite their lack of visual input (Figure 2B, C). Importantly, because positional avoidance of UV of these animals was evident even during periods when there were no egg-laying events (Figure 2A, C) despite these animals were in egg-laying state in general (Gou et al., 2014) such avoidance cannot be trivially explained because 1) the physical act of egg-laying takes time, and 2) females tend to stay immobile for a while on the site where they had just deposited an egg. Instead, this result suggests that females in egg-laying state tend to avoid spending time on sites exposed to higher levels of UV and that extraocular UV sensors play a role in promoting such UV avoidance Proboscis of adult flies houses neurons that express H 2 O 2 -sensitive isoforms of dtrpa1 We next set out to determine the identity of the extraocular UV-sensitive neurons that promote egg-laying avoidance of UV in blind females. Because H 2 O 2 - sensitive isoforms of dtrpa1 are UV sensitive, we hypothesized that the extraocular UVsensitive neurons may rely on these isoforms to sense UV. Previous findings have suggested that one of the H 2 O 2 -sensitive TRPA1 isoforms is expressed in the Gr66aexpressing bitter-sensing neurons on the proboscis (Kang et al., 2012). To further confirm this result, we performed RT-PCT experiment and found that H 2 O 2 /UV-sensitive isoforms were indeed present on the proboscis (Figure 3A, B). These results suggest that Gr66a-expressing bitter-sensing neurons may exhibit dtrpa1-depedent UV sensitivity. 13

14 Gr66a-expressing neurons on the proboscis exhibit dtrpa1-dependent UV sensitivity We next tested whether functional H 2 O 2 -sensitive isoforms of dtrpa1 are indeed present in the Gr66a-expressing bitter-sensing neurons on the proboscis and confer these neurons with UV sensitivity. We first used Gr66a-Gal4 to express in these neurons the genetically encoded calcium indicator GCaMP6 (Chen et al., 2013) and found that these neurons showed a clear response to both UV and H 2 O 2 (Figure 4A-C). Importantly, illuminating even very strong UV onto the Gr5a-expressing sweet-sensing neurons on the proboscis did not induce any responses, suggesting that the UV-induced calcium response of Gr66a neurons we observed was not a non-specific injury-induced response (Figure. 4C). Two additional experiments also support that H 2 O 2 -sensitive isoforms of dtrpa1 are present in the Gr66a neurons on the proboscis. First, in addition to exhibiting a clear sensitivity to UV and H 2 O 2, we found that proboscis neurons labeled by dtrpa1-gal4 (Petersen and Stowers, 2011), a reporter for dtrpa1 expression, showed a clear sensitivity to caffeine (Figure. 4G, black bar), a known bitter agonist of Gr66a neurons (Lee et al., 2009; Moon et al., 2006). Second, co-labeling experiments showed that most if not all of the neurons labeled by the dtrpa1-gal4 on the proboscis expressed Gr66a also (Figure 4D, E, F and Figure S1). Does the UV sensitivity exhibited by dtrpa1/gr66a-expressing neurons depend on H 2 O 2 -sensitive dtrpa1? We did two experiments to address this. First, we eliminated dtrpa1 function from these neurons (by examining them in the dtrpa1 KO mutant background) and found that they no longer responded to UV or H 2 O 2 but still responded to caffeine (Figure. 4G). Second, we overexpressed the H 2 O 2 degrading 14

15 enzyme catalase (cat) in the dtrpa1-expressing neurons and found their UV response reduced significantly (Figure. 4G). Taken together, our collective results so far suggest that the bitter-sensing dtrpa1/gr66a-gal4-expressing gustatory neurons on the proboscis are UV sensitive and that their UV sensitivity critically depends on lightinduced H 2 O 2 production and H 2 O 2 -sensitive dtrpa1 isoforms Gr66a-expressing neurons on the proboscis are critical for UV avoidance of blind females We next addressed whether the bitter-sensing Gr66a neurons on the proboscis play a role in promoting blind females egg-laying avoidance of UV. We first surgically removed the proboscis from two different blind mutants (norpa 36 and Hdc JK910 mutants) and found that these blind and proboscis-less animals no longer avoided UV for egglaying in the UV vs. dark assay (Figure. 5A). We next directly assessed the role of the Gr66a neurons in promoting UV avoidance in blind females. Because there is no Gal4 as far as we know that labels only the Gr66a bitter-sensing neurons on the proboscis (and nowhere else in the nervous system), we used two different Gal4s dtrpa1-gal4 (Petersen and Stowers, 2011) and Gr66a-Gal4 (Wang et al., 2004) to silence the activities of these neurons. These two Gal4s both labeled H 2 O 2 /UV sensitive neurons on the proboscis (Figure. 4D, E, F and Figure. S1), and more importantly, outside of the proboscis, their expression patterns were sparse and rather distinct (Figure S2 and Figure 5B, C). Expressing the hyperpolarizing Kir2.1 channel using either Gal4 caused the blind females to reduce their egg-laying avoidance of UV (Figure 5D). Thus, both the surgery experiment and the neuron-silencing experiment support the idea that 15

16 Gr66a/dTRPA1-expressing neurons on the proboscis play a critical role in promoting vision-independent egg-laying avoidance of UV Gr66a-expressing bitter-sensing neurons rely on H 2 O 2 -sensitive dtrpa1 to promote UV avoidance in blind females Do the Gr66a neurons on the proboscis rely on H 2 O 2 -sensitive dtrpa1 to promote avoidance of UV? To address this, we first reduced dtrpa1 expression in these neurons by employing a commonly used UAS-dTRPA1-RNAi tool (that was designed against all dtrpa1 isoforms) (Hamada et al., 2008) and found that this manipulation significantly reduced blind females egg-laying avoidance of UV (Figure 6A). Second, we reduced these neurons UV sensitivity by overexpressing catalase (cat) in them and found that this significantly reduced UV avoidance also (Figure 6B). Lastly, we reintroduced the H 2 O 2 -sensitive isoform of dtrpa1 into the Gr66a neurons in the norpa; dtrpa1 double mutants (that were completely incapable of avoiding UV) and found it restored animals UV avoidance (Figure 6C). Importantly, the ability of these rescued animals to avoid UV disappeared if we surgically removed their proboscis (Figure 6C), suggesting that a major source of the rescue likely resides on the proboscis. Taken together, these results suggest strongly that the Gr66a neurons on the proboscis rely on H 2 O 2 -sensitive dtrpa1 to promote vision-independent egg-laying avoidance of UV Optogenetic activation of the Gr66a-expressing neurons on the proboscis is sufficient to promote egg-laying avoidance 16

17 Finally, we set out to ask whether artificially activating the Gr66a-expressing neurons on the proboscis is sufficient to drive egg-laying avoidance. If so, this would lend further support to the idea that detection of UV by these neurons promotes egglaying avoidance. We addressed this by taking advantage of the recently made available red-shifted channelrhodopsin CsChrimson (Klapoetke et al., 2014) (Figure 7A). We found that whereas control females showed no innate avoidance of red light for egglaying in our red LED vs. dark assays (Figure 7B), animals that overexpressed CsChrimson in their dtrpa1-gal4- or the Gr66a-Gal4-expressing neurons showed a clear preference to lay eggs away from red light when given a choice between a red light-illuminated option and an unilluminated one, likely because egg-laying females avoided red light positionally (Figure 7B). To further assess whether the avoidance of red light for egg-laying we observed was mediated specifically by the activation of Gr66a-expressing neurons on the proboscis (as opposed to by ones residing on the leg, for example), we attempted to restrict CsChrimson expression to just the proboscis neurons by using an intersectional approach. We typically found that about five of the dtrpa1-gal4-expressing neurons on the proboscis were co-labeled by the Gr66a-lexA. (It is worth pointing out that the lack of substantial overlap between these two drivers is consistent with the lack of significant overlap between Gr66a-lexA and Gr66a-Gal4 expression on the proboscis (Figure 7C)). By using a combination of transgenes (dtrpa1-gal4, Gr66a-lexA, tub>gal80>, lexa-flp, and UAS-CsChrimson), we indeed limited CsChrimson expression to just a few neurons on the proboscis (Figure 7D). These animals showed a clear 17

18 avoidance of red light for egg-laying (Figure 7E). Their level of avoidance was reduced as compared to the non-intersected animals, however, most likely because they had fewer neurons that expressed CsChrimson. Importantly, these animals no longer showed red light avoidance if we did not supplement their diet with retinal a critical chromophore for CsChrimson activation (Figure 7E), suggesting that the red light avoidance we observed was not an artifact. Taken together, these results suggest that ectopically expressing CsChrimson in the Gr66a-expressing gustatory neurons on the proboscis is sufficient to promote egg-laying avoidance of red light, further supporting the notion that these neurons play a role in promoting egg-laying avoidance of UV, a natural stimulus for them

19 Discussion In this work we addressed the question of whether the H 2 O 2 -mediated UV sensitivity of specific Drosophila TRPA1 isoforms (Guntur et al., 2015) has a physiological role in behaving animals. By using a combination of behavioral assays, calcium imaging, and neuronal activity manipulation experiments, we found that H 2 O 2 /UV-sensitive dtrpa1 isoforms act in a group of bitter-sensing gustatory neurons on the proboscis to promote egg-laying avoidance of strong UV, an aversive and damaging stimulus for young larvae. We found that animals that lacked vision (norpa or Hdc mutants) were still capable of sensing and avoiding UV for egg-laying. But these blind females no longer avoided UV if they lacked dtrpa1 or if their bitter- and UV-sensitive gustatory neurons on the proboscis were rendered UV insensitive (by catalase overexpression or removal of dtrpa1 function) or hyperpolarized (by Kir2.1 overexpression). Additionally, ectopic expression of the red-shifted channelrhodopsin CsChrimson in these bittersensing gustatory neurons enabled females to lay eggs away from red light, an otherwise neutral cue for egg-laying females, suggesting that these neurons are capable of promoting egg-laying avoidance when activated. Taken together, these results support the hypothesis that UV sensitivity of the H 2 O 2 -sensitive dtrpa1 isoforms is physiologically relevant and that these isoforms act in the bitter-sensing gustatory neurons as an extraocular backup to promote UV avoidance during egg-laying. What is the advantage of having an extraocular system for UV sensing? UV is a known aversive stimulus for young larvae and may incur damage to them (Xiang et al., 2010). Moreover, laying eggs on UV-illuminated substrates as opposed to laying eggs 19

20 in the dark may expose the egg-laying females to increased predation risk also (Yang et al., 2008; Yang et al., 2015). Thus having a second UV-sensing system should confer some advantage to the species in case animals vision fails due to genetic defects, injuries, or temporary blockage or desensitization. It is worth noting that such dual UV sensors design is not unprecedented. As we have mentioned earlier, a previous report has shown that Drosophila larvae employ two different systems, the Bolwig organ and the C4da neurons, for light sensing (Xiang et al., 2010). Interestingly, reminiscent of the dual light-sensing systems we describe here, the Bolwig organ has a high light sensitivity and uses the conventional Rhodopsin-PLC pathway to sense light (Busto et al., 1999; Hassan et al., 2000; Malpel et al., 2002) whereas the C4da neurons have a low light sensitivity and use the Gr28b-dTRPA1 pathway to specifically sense strong blue and UV lights (Xiang et al., 2010). And for both larvae and adults, the extraocular UV sensors are recruited from sensory neurons that are known to promote behavioral avoidance (Xiang et al., 2010). It is also worth pointing out that different isoforms of dtrpa1 appear to act in different subsets of Gr66a neurons in adult flies. For example, our results suggest H 2 O 2 - sensitive isoforms of dtrpa1 are present in the Gr66a neurons on the labellum (at the tip of the proboscis) and can act to drive positional avoidance of UV in egg-laying females. On the other hand, a recent report has suggested that Gr66a neurons that reside on the esophagus express both the H 2 O 2 -sensitive and insensitive dtrpa1 isoform and can act to suppress ingestion of bacterial toxin LPS another candidate agonist of dtrpa1 (Soldano et al., 2016). These results again highlight the notion that, 20

21 in Drosophila, dtrpa1 appears to primarily act as a sensor for aversive stimuli and are often housed in different avoidance-promoting sensory neurons. Lastly, while our data suggest the dtrpa1-expressing gustatory neurons play a critical role in enabling females to lay eggs away from UV in the absence of vision, it is likely that they serve additional roles. First, these neurons are present on the proboscis of males also and norpa males did show weaker but significant positional avoidance of UV (PI ~ , n = 22). Moreover, one important function of bitter-sensing gustatory neurons is to suppress feeding. It is conceivable that expression of H 2 O 2 - sensitive dtrpa1 in these neurons may act to suppress feeding when levels of UV are stronger and closer to that of sunlight. While the specific advantage of feeding suppression in response to high levels of UV is unclear, Bhatla et al. have recently shown that, in C. elegans, short-wavelength blue and UV lights can indeed inhibit feeding via activating a specific group of pharyngeal neurons (Bhatla and Horvitz, 2015). Curiously, blue and UV lights activate these neurons via light-induced H 2 O 2 production also, but G protein-coupled receptors LITE-1 and GUR3, not TRPA1, are suggested to be the molecular sensors for H 2 O 2 in this case (Bhatla and Horvitz, 2015). UV-induced feeding suppression aside, presence of the H 2 O 2 -sensitive dtrpa1 isoforms in the gustatory neurons may also serve to prevent feeding of ROS-containing chemicals that can activate the H 2 O 2 -sensitive dtrpa1 in these neurons (Du et al., 2016; Kang et al., 2012), as ingestion some of the ROS-containing chemicals may cause harms to the animals. Future work is needed to fully elucidate the roles of UV/H 2 O 2 -sensitive dtrpa1 in behaving animals. 21

22 References Baines, R.A., Uhler, J.P., Thompson, A., Sweeney, S.T., and Bate, M. (2001). Altered electrical properties in drosophilaneurons developing without synaptic transmission. The Journal of neuroscience 21, Bellono, N.W., Kammel, L.G., Zimmerman, A.L., and Oancea, E. (2013). UV light phototransduction activates transient receptor potential A1 ion channels in human melanocytes. Proceedings of the National Academy of Sciences 110, Bhatla, N., and Horvitz, H.R. (2015). Light and hydrogen peroxide inhibit C. elegans feeding through gustatory receptor orthologs and pharyngeal neurons. Neuron 85, Branson, K., Robie, A.A., Bender, J., Perona, P., and Dickinson, M.H. (2009). Highthroughput ethomics in large groups of Drosophila. Nature methods 6, Burg, M.G., Sarthy, P., Koliantz, G., and Pak, W. (1993). Genetic and molecular identification of a Drosophila histidine decarboxylase gene required in photoreceptor transmitter synthesis. The EMBO journal 12, 911. Busto, M., Iyengar, B., and Campos, A.R. (1999). Genetic dissection of behavior: modulation of locomotion by light in the Drosophila melanogaster larva requires genetically distinct visual system functions. The Journal of neuroscience 19, Chatzigeorgiou, M., Yoo, S., Watson, J.D., Lee, W.-H., Spencer, W.C., Kindt, K.S., Hwang, S.W., Miller III, D.M., Treinin, M., Driscoll, M., et al. (2010). Specific roles for DEG/ENaC and TRP channels in touch and thermosensation in C. elegans nociceptors. Nature neuroscience 13, Chen, T.-W., Wardill, T.J., Sun, Y., Pulver, S.R., Renninger, S.L., Baohan, A., Schreiter, E.R., Kerr, R.A., Orger, M.B., Jayaraman, V., et al. (2013). Ultrasensitive fluorescent proteins for imaging neuronal activity. Nature 499, Clapham, D.E. (2003). TRP channels as cellular sensors. Nature 426, Cordero-Morales, J.F., Gracheva, E.O., and Julius, D. (2011). Cytoplasmic ankyrin repeats of transient receptor potential A1 (TRPA1) dictate sensitivity to thermal and chemical stimuli. Proceedings of the National Academy of Sciences 108, E1184-E1191. Du, E.J., Ahn, T.J., Wen, X., Seo, D.-W., Na, D.L., Kwon, J.Y., Choi, M., Kim, H.-W., Cho, H., and Kang, K. (2016). Nucleophile sensitivity of Drosophila TRPA1 underlies light-induced feeding deterrence. elife 5, e

23 Gao, S., Takemura, S.-y., Ting, C.-Y., Huang, S., Lu, Z., Luan, H., Rister, J., Thum, A.S., Yang, M., Hong, S.-T., et al. (2008). The neural substrate of spectral preference in Drosophila. Neuron 60, Gordon, M.D., and Scott, K. (2009). Motor control in a Drosophila taste circuit. Neuron 61, Gou, B., Liu, Y., Guntur, A.R., Stern, U., and Yang, C.-H. (2014). Mechanosensitive neurons on the internal reproductive tract contribute to egg-laying-induced acetic acid attraction in Drosophila. Cell reports 9, Gou, B., Zhu, E., He, R., Stern, U., and Yang, C.-H. (2016). High Throughput Assay to Examine Egg-Laying Preferences of Individual Drosophila melanogaster. JoVE (Journal of Visualized Experiments), e53716-e Guntur, A.R., Gu, P., Takle, K., Chen, J., Xiang, Y., and Yang, C.-H. (2015). Drosophila TRPA1 isoforms detect UV light via photochemical production of H2O2. Proceedings of the National Academy of Sciences 112, E5753-E5761. Hamada, F.N., Rosenzweig, M., Kang, K., Pulver, S.R., Ghezzi, A., Jegla, T.J., and Garrity, P.A. (2008). An internal thermal sensor controlling temperature preference in Drosophila. Nature 454, Hassan, J., Busto, M., Iyengar, B., and Campos, A.R. (2000). Behavioral characterization and genetic analysis of the Drosophila melanogaster larval response to light as revealed by a novel individual assay. Behavior genetics 30, Hockberger, P.E., Skimina, T.A., Centonze, V.E., Lavin, C., Chu, S., Dadras, S., Reddy, J.K., and White, J.G. (1999). Activation of flavin-containing oxidases underlies light-induced production of H2O2 in mammalian cells. Proceedings of the National Academy of Sciences 96, Jordt, S.-E., Bautista, D.M., Chuang, H.-h., McKemy, D.D., Zygmunt, P.M., Högestätt, E.D., Meng, I.D., and Julius, D. (2004). Mustard oils and cannabinoids excite sensory nerve fibres through the TRP channel ANKTM1. Nature 427, Julius, D. (2013). TRP channels and pain. Annual review of cell and developmental biology 29, Kang, K., Panzano, V.C., Chang, E.C., Ni, L., Dainis, A.M., Jenkins, A.M., Regna, K., Muskavitch, M.A., and Garrity, P.A. (2012). Modulation of TRPA1 thermal sensitivity enables sensory discrimination in Drosophila. Nature 481,

24 Kang, K., Pulver, S.R., Panzano, V.C., Chang, E.C., Griffith, L.C., Theobald, D.L., and Garrity, P.A. (2010). Analysis of Drosophila TRPA1 reveals an ancient origin for human chemical nociception. Nature 464, Karuppudurai, T., Lin, T.-Y., Ting, C.-Y., Pursley, R., Melnattur, K.V., Diao, F., White, B.H., Macpherson, L.J., Gallio, M., Pohida, T., et al. (2014). A hard-wired glutamatergic circuit pools and relays UV signals to mediate spectral preference in Drosophila. Neuron 81, Kim, S.H., Lee, Y., Akitake, B., Woodward, O.M., Guggino, W.B., and Montell, C. (2010). Drosophila TRPA1 channel mediates chemical avoidance in gustatory receptor neurons. Proceedings of the National Academy of Sciences 107, Klapoetke, N.C., Murata, Y., Kim, S.S., Pulver, S.R., Birdsey-Benson, A., Cho, Y.K., Morimoto, T.K., Chuong, A.S., Carpenter, E.J., Tian, Z., et al. (2014). Independent optical excitation of distinct neural populations. Nature methods 11, Kwon, Y., Kim, S.H., Ronderos, D.S., Lee, Y., Akitake, B., Woodward, O.M., Guggino, W.B., Smith, D.P., and Montell, C. (2010). Drosophila TRPA1 channel is required to avoid the naturally occurring insect repellent citronellal. Current Biology 20, Lee, Y., Moon, S.J., and Montell, C. (2009). Multiple gustatory receptors required for the caffeine response in Drosophila. Proceedings of the National Academy of Sciences 106, Malpel, S., Klarsfeld, A., and Rouyer, F. (2002). Larval optic nerve and adult extra-retinal photoreceptors sequentially associate with clock neurons during Drosophila brain development. Development 129, Moon, S.J., Köttgen, M., Jiao, Y., Xu, H., and Montell, C. (2006). A taste receptor required for the caffeine response in vivo. Current biology 16, Morita, T., McClain, S.P., Batia, L.M., Pellegrino, M., Wilson, S.R., Kienzler, M.A., Lyman, K., Olsen, A.S.B., Wong, J.F., Stucky, C.L., et al. (2015). HTR7 mediates serotonergic acute and chronic itch. Neuron 87, Petersen, L.K., and Stowers, R.S. (2011). A Gateway MultiSite recombination cloning toolkit. PLoS ONE 6, e Rosenzweig, M., Brennan, K.M., Tayler, T.D., Phelps, P.O., Patapoutian, A., and Garrity, P.A. (2005). The Drosophila ortholog of vertebrate TRPA1 regulates thermotaxis. Genes & development 19,

25 Shen, W.L., Kwon, Y., Adegbola, A.A., Luo, J., Chess, A., and Montell, C. (2011). Function of rhodopsin in temperature discrimination in Drosophila. Science 331, Soldano, A., Alpizar, Y.A., Boonen, B., Franco, L., López-Requena, A., Liu, G., Mora, N., Yaksi, E., Voets, T., Vennekens, R., et al. (2016). Gustatory-mediated avoidance of bacterial lipopolysaccharides via TRPA1 activation in Drosophila. Elife 5, e Story, G.M., Peier, A.M., Reeve, A.J., Eid, S.R., Mosbacher, J., Hricik, T.R., Earley, T.J., Hergarden, A.C., Andersson, D.A., Hwang, S.W., et al. (2003). ANKTM1, a TRP-like channel expressed in nociceptive neurons, is activated by cold temperatures. Cell 112, Thistle, R., Cameron, P., Ghorayshi, A., Dennison, L., and Scott, K. (2012). Contact chemoreceptors mediate male-male repulsion and male-female attraction during Drosophila courtship. Cell 149, Viswanath, V., Story, G.M., Peier, A.M., Petrus, M.J., Lee, V.M., Hwang, S.W., Patapoutian, A., and Jegla, T. (2003). Ion channels: opposite thermosensor in fruitfly and mouse. Nature 423, Wang, Z., Singhvi, A., Kong, P., and Scott, K. (2004). Taste representations in the Drosophila brain. Cell 117, Xiang, Y., Yuan, Q., Vogt, N., Looger, L.L., Jan, L.Y., and Jan, Y.N. (2010). Light-avoidancemediating photoreceptors tile the Drosophila larval body wall. Nature 468, Yang, C.-h., Belawat, P., Hafen, E., Jan, L.Y., and Jan, Y.-N. (2008). Drosophila egg-laying site selection as a system to study simple decision-making processes. Science 319, Yang, C.-H., He, R., and Stern, U. (2015). Behavioral and circuit basis of sucrose rejection by Drosophila females in a simple decision-making task. The Journal of Neuroscience 35, Zhu, E.Y., Guntur, A.R., He, R., Stern, U., and Yang, C.-H. (2014). Egg-laying demand induces aversion of UV light in Drosophila females. Current Biology 24,

26 Acknowledgements: We like to thank Drs. S. Stowers, K. Scott, P. Garrity and the Bloomington Stock Center for providing us with fly stocks. We also thank members of the Yang and Xiang labs and Dr. A. Bellemer for comments and suggestions during the course of this work. This work was supported by NIH grant R01GM to C.-H. Yang and NIH grant R01NS to Y. Xiang. 26

27 Figure Legends Figure 1. Blind females still prefer to not lay eggs on UV site (A) A schematic diagram of the two-choice behavioral assay we used to test egg-laying preferences of Drosophila. Each fly was given two agarose-containing egg-laying options (that were separated by hard plastic in between). One of the options was illuminated with UV. (B) A representative picture of egg-laying result of a single w 1118 female. (C) A representative picture of egg-laying result of a single norpa 36 female. Note that there are fewer eggs deposited on the UV site. (D) Egg-laying preference index (PI) of w 1118, norpa 36, and Hdc JK910. Note that both vision mutants are null alleles. **: p < 0.01, ***: p<0.001, one-sample t test from 0. The number of flies used for each experiment performed in this work is labeled directly on each graph. Figure 2. Blind females exhibit positional avoidance of UV (A) Trajectory of a single w 1118 female as it explored the UV vs. Dark chamber for 2 hours. x-axis represents time. y-axis represents the y position of the animal in the chamber. It denotes the relative distance to the edges of the substrates. Further, each black circle denotes an aversive return towards UV. It indicates an event where the fly had changed its heading direction from moving towards UV to moving away from UV. The red square denotes an attractive return towards UV. It indicates that the fly had changed its heading direction from moving away from UV to moving towards UV. 27

28 (B) Trajectory of a single norpa 36 female as it explored the UV vs. Dark chamber for two hours. (C) Positional preference index (PI) of w 1118 and norpa 36 flies. **: p < 0.01, ***: p<0.001, one-sample t test from 0. Figure 3. H 2 O 2 /UV-sensitive dtrpa1 isoforms are present on the proboscis (A-B) RT-PCR showing that the H 2 O 2 /UV-sensitive isoforms dtrpa1(a)10a and 10b are present on the labellum of proboscis. L: labellum. A: whole adult. Note that the nomenclature of dtrpa1 isoforms we adopted in this work and in Guntur et al., 2015 was conceived, and proposed to us by Dr. Paul Garrity and his student Vincent Panzano. Figure 4. H 2 O 2 /UV-sensitive dtrpa1 isoforms are expressed in the Gr66a-expressing neurons and confer them with UV sensitivity (A-B) Representative calcium responses of Gr66a neurons illuminated with UV. (C) Quantification of calcium responses of Gr66a neurons to UV and H 2 O 2. In contrast, Gr5a neurons did not respond to UV even its intensity was elevated to 35 mw/mm 2. (Note that response of Gr66a neurons to 35 mw/mm 2 UV is significantly higher than their response at 6 mw/mm 2, not shown.) *: p < 0.05, **: p < 0.01, one-sample t test from 0 (D-F) Representative expression pattern of Gr66a-Gal4 alone (D), Gr66a-Gal4 plus dtrpa1-gal4 (E), and dtrpa1-gal4 alone (F) on the proboscis. Note that the numbers of neurons labeled on the proboscis by Gr66a-Gal4 alone vs. Gr66a-Gal4 plus dtrpa1- Gal4 were comparable. 28

29 (G) Quantification of calcium responses of dtrpa1-gal4-labeled neurons to caffeine, UV, and H 2 O 2. Note that UV and H 2 O 2 responses both reduced significantly when these neurons lacked dtrpa1 or when they overexpressed catalase (cat). For a-b: *= p < 0.05, **= p < 0.01, one-sample t test from 0. For c: *= p<0.05, ***= p<0.001, one-way ANOVA followed by Tukey s multiple comparison test. For d: **= p<0.01, unpaired t test. Figure 5. Gr66a-expressing neurons are critical for UV avoidance in blind females (A) Egg-laying PI of proboscis-less norpa 36 and proboscis-less Hdc JK910 females. Neither is significantly different from 0, one-sample t-test. Note that proboscis-less females laid fewer eggs. For example, the average numbers of eggs laid overnight by individual proboscis-less norpa and Hdc females were 18.1 ± 1.5 and 18.3 ± 1.8 respectively. While the egg-laying numbers were low, they were still sufficient for us to calculate PI our typical cutoff for PI is 10 eggs. (B-C) Expression pattern of dtrpa1-gal4 and Gr66a-Gal4 in adult CNS. Arrows point to the axonal termini (in the SEG) of the gustatory neurons on the proboscis. (D) Egg-laying PIs of blind (norpa) flies whose dtrpa1/gr66a neurons were silenced by Kir2.1 overexpression. *: p<0.05, **: p<0.01, and ***: p<0.001, one-way ANOVA followed by Tukey s multiple comparison test. Figure 6. Expression of H 2 O 2 /UV-sensitive dtrpa1 isoforms in Gr66a-expressing neurons promotes egg-laying avoidance of UV in blind females 29

30 (A) Egg-laying PIs of blind (norpa) flies whose Gr66a neurons overexpressed dtrpa1- RNAi. *: p<0.05, **: p<0.01, and ***: p<0.001, one-way ANOVA followed by Tukey s multiple comparison test. (B) Egg-laying PIs of blind (norpa) flies whose Gr66a neurons overexpressed catalase (cat). *: p<0.05, **: p<0.01, and ***: p<0.001, one-way ANOVA followed by Tukey s multiple comparison test. (C) Egg-laying PIs of norpa; dtrpa1 KO double mutant (left bar), norpa 36 ; dtrpa1 KO double mutant with isoform dtrpa1(a)10a rescued in their Gr66a neurons (middle bar), and the rescued flies with proboscis severed (right bar). *: p<0.05, **: p<0.01, and ***: p<0.001, one-way ANOVA followed by Tukey s multiple comparison test. Figure 7. Optogenetic activation of Gr66a-expressing neurons on the proboscis promotes egg-laying avoidance of red light (A) A schematic diagram of the Red LED vs. dark egg-laying assay. (B) Egg-laying PIs of flies whose Gr66a neurons overexpressed the red-shifted channelrhodopsin CsChrimson (CsC). *: p<0.05, **: p<0.01, and ***: p<0.001, one-way ANOVA followed by Tukey s multiple comparison test. (C) Expression pattern of Gr66a-Gal4 and Gr66a-lexA driving GFP and nls-cherry, respectively, on the proboscis. Note that while Gr66a-Gal4 typically labels ~ 22 neurons on the proboscis, only 7 of them were co-labeled by Gr66a-lexA. This was the reason why only a few neurons were labeled when we used Gr66a-lexA to intersect with dtrpa1-gal4. 30

31 (D) Representative images of CsChrimson labeling in the brain and VNC of the intersected females. Note that the only processes that were labeled in these animals were the axons of bitter-sensing neurons in the SEG. > : FRT site. (E) Egg-laying PIs of the intersected females that were or were not fed with retinalsupplemented food. ***: p<0.001, unpaired t test. 31

32

33

34

35

36

37

38