Alleviation of thermal nociception depends on heat-sensitive neurons and a TRP channel in the brain

Acute avoidance of dangerous temperatures is critical for animals to prevent or minimize injury. Therefore, surface receptors have evolved to endow neurons with the capacity to detect noxious heat so that animals can initiate escape behaviors. Animals including humans have evolved intrinsic pain-suppressing systems to attenuate nociception under some circumstances. Here, using Drosophila melanogaster , we uncovered a new mechanism through which thermal nociception is suppressed. We identiﬁed a single descending neuron in each brain hemisphere, which is the center for suppression of thermal nociception. These Epi neurons, for Epione—the goddess of soothing of pain—express a nociception-suppressing neuropeptide Allatostatin C (AstC), which is related to a mammalian anti-nociceptive peptide, somatostatin. Epi neurons are direct sensors for noxious heat, and when activated they release AstC, which diminishes nociception. We found that Epi neurons also express the heat-activated TRP channel, Painless (Pain), and thermal activation of Epi neurons and the subsequent suppression of thermal nociception depend on Pain. Thus


In brief
Sensing noxious stimuli is critical for survival. Animals are also endowed with mechanisms to suppress nociception. Using Drosophila, Liu et al. identify a pair of neurons in the brain (Epi neurons) critical for suppressing thermal nociception. Epi neurons sense heat via a TRP channel and then release a neuropeptide that suppresses nociception.

INTRODUCTION
Endogenous pain inhibitory systems can temporarily provide relief. Millions of people suffer from chronic and debilitating pain, some of which might be induced by abnormalities in the descending pain modulatory system. [1][2][3][4][5][6][7][8][9][10][11][12] In mammals, neurotransmitters and neuromodulators, including endogenous opioids (b-endorphin, encephalin, and dynorphin) and endogenous cannabinoids, play important roles in nociception inhibition. 13,14 Brain imaging and electrophysiological studies indicate that the pain-suppressing descending modulatory circuit receives input from multiple brain regions including the rostral anterior cingulate cortex, the periaqueductal gray region, and the rostral ventromedial medulla. 12,13 However, the key neurons that are activated in the inhibitory pathway, and the target neurons that are silenced, have not been clearly delineated. A pain inhibitory system has also been documented in worms. In C. elegans avoidance responses that are mediated through the polymodal ASH neurons 15,16 are suppressed by complex signaling pathways initiated by octopamine and neuropeptides. 17 Drosophila has also been employed to study the inhibition of nociception, 18 in addition to the far more extensive studies focusing on the mechanisms for detecting noxious stimuli, such as excessive heat, to initiate escape responses. 19,20 Tracey et al. revealed that a Drosophila channel, Painless (Pain), 21 which is related to the TRP channel in the fly's compound eye, 22 is critical for sensing noxious heat. This work, which followed the seminal discovery of TRPV1 as a heat sensor in mammals 23 and the finding that a related TRPV channel (Osm-9) contributes to several other sensory modalities in C. elegans, 24 contributed significantly to the notion that TRP channels are evolutionarily conserved polymodal sensors. 25 In addition to Pain, two other Drosophila TRP channels also function in sensing high temperatures to promote escape behavior: Pyrexia (Pyx) and TRPA1. [26][27][28][29] However, it is unclear whether any TRP channel serves to detect noxious heat for the purpose of alleviating thermally induced nociception.
In this work, we used the fruit fly, Drosophila melanogaster, to investigate an intrinsic system for suppression of thermal nociception. We identified a pair of bilaterally symmetrical neurons in the brain that is required for decreasing the nociceptive response to hot temperatures. These Epi neurons respond directly to heat and release a neuropeptide, Allatostatin C (AstC), which is required for suppression of nociception. The ability of Epi neurons to sense noxious heat depends on Pain, demonstrating a role for a thermo-TRP in suppressing rather than enhancing the nociceptive response to high temperatures.

Epi neurons function in suppression of thermal nociception
To identify neurons that may play roles in the suppression of thermal nociception, we devised a thermal nociception assay    (Figures 1A, S1A, and S1B). We placed flies on a hot plate and assayed the percentage that jumped within 10 s. To prevent the insects from flying away during the assay, we first amputated their wings and allowed them to recover for 24 h. To establish the relationship between temperature and the jump response, we exposed control flies (w 1118 ) to a series of temperatures ranging from 29 C to 50 C. At 29 C-30 C, few flies exhibited jump responses (17.2%), while at temperatures between 33 C and 34 C more than half (54.2%) of the flies jumped ( Figure 1B). The percentage of flies that responded continued to increase with temperature. Between 38 C and 44 C, nearly all the files jumped (92.6%-100%; Figure 1B). Once the temperature exceeded 44 C, all of the flies reacted to the noxious stimuli ( Figure 1B).
To address whether acute activation of Epi-Gal4-positive neurons is sufficient to alleviate nociception, we used an optogenetic approach. We expressed the red-shifted channelrhodopsin, ReaChR (UAS-ReaChR), under control of the Epi-Gal4-1 or Epi-Gal4-2; exposed the animals to red lights (655 nm peak) for 30 s; and assayed their jump responses when placed on a 45 C-46 C surface. We found that acute activation of the Epi-Gal4 neurons significantly reduced the percentages of flies that jumped ( Figures 1H and S2I) and increased the jump latency ( Figures 1I and S2J).
To determine how long the nociception-alleviation effects continued following acute activation of the Epi-Gal4 neurons, we exposed the UAS-ReaChR flies to light for 30 s and then maintained them in the dark for 0.5-5 min before testing their jump responses. We found that the alleviation of nociception gradually diminished over time. After only 30-60 s in the dark, the jump percentages and the average jump latencies were not statistically different from flies that did not have their Epi neurons light activated ( Figures 1H, 1I, S2I, and S2J). However, 3 min was required before the decreased jump percentages, and latencies were significantly different from flies immediately after activation ( Figures 1I, 1J, S2I, and S2J). Because several minutes were required for full restoration of the nociceptive response, a neuromodulator such as AstC, which is longer lasting than a neurotransmitter, might underlie this relatively slow recovery.
To confirm the role of Epi-Gal4 neurons in suppressing nociception, we tested the effects of inhibiting synaptic transmission from these neurons and on the reaction to a 35 C-36 C heat stimulus. We expressed tetanus toxin (UAS-TNTG) 31 under control of the Epi-Gal4-2 driver and found that disrupting signaling from these neurons significantly increased the jump percentage ( Figure 1J) and decreased the average jump latency ( Figure 1K).
To identify the neurons expressing the Epi-Gal4-1 and the Epi-Gal4-2, we used these lines to drive expression of UAS-mCD8::GFP and performed staining with anti-GFP. The Epi-Gal4-1 reporter stained multiple brain regions, including neurons in a region near the optic lobe (OL), neurons in the primary taste center (the subesophageal zone [SEZ]), and a cluster of neurons projecting to the ellipsoid body (EB) of the central complex, which functions in multisensory integration (Figures 2A and  2B). 24,32,33 The Epi-Gal4-2 displayed a much more restricted pattern-labeling one pair of neurons in the brain with large cell bodies ( Figures 2E and 2F). We expressed a nuclear GFP (UAS-Stinger 2), which confirmed that there were only two cells labeled in the brain using the Epi-Gal4-2 ( Figure S3A). We did not detect Epi-Gal4-2 expression in other organs, including the legs, wings, antenna, proboscis, and digestion system ( Figures S3B-S3H). The Epi-Gal4-2-positive neurons arborized a portion of the dorsal region in the brain ( Figures 2E and 2F) and projected to multiple segments in the ventral nerve cord (VNC) (Figures 2G and 2H). The arborization was more extensive with the Epi-Gal4-1 (Figures 2A-2D). Based on the position and arborization pattern, the neurons labeled by the Epi-Gal4-2 (arrows, Figures 2E and 2F) also appeared to be labeled by the Epi-Gal4-1 (arrows, Figures 2A and 2B). We refer to this pair of neurons as Epione (Epi) neurons after the Greek goddess of soothing of pain, 34 since chronic or acute activation of these neurons is sufficient to alleviate thermal nociception.
To better visualize the distribution of the dendrites and axons of the Epi neurons, we used the Epi-Ga4-2 to drive expression of DenMark and synaptotagmin::GFP (Syt::GFP), which label dendrites and axons, respectively. 35,36 Both DenMark and Syt::GFP also label the cell bodies. We found that the dendrites of the Epi neurons innervated multiple regions in the brain, including the OL, the lateral horn (LH), and areas close to the mushroom body (MB) (Figures 2I, 2K, and 2O). However, there was very limited DenMark staining in the VNC ( Figure 2L). The axonal signals, which were marked by Syt::GFP, branched extensively in the brain, including regions in the OL, MBs, and the SEZ ( Figures 2J, 2K, and 2O). The axons also projected to multiple segments in the VNC, including the prothoracic, metathoracic, mesothoracic, and abdominal ganglion ( Figures 2M-2O).
AstC is required in Epi neurons to control nociception AstC receptors have been proposed to be expressed in nociceptive neurons, 37 but the neurons secreting AstC to inhibit nociception have not been identified. To determine whether Epi neurons express the AstC neuropeptide, we performed double labeling. We found that the two most prominent neurons that were labeled with anti-AstC also expressed the Epi-Gal4-1 and the Epi-Gal4-2, which drove the expression of UAS-mCD8::GFP (Figures 2P-2U). Thus, Epi neurons express AstC.
AstC is related to the mammalian neuropeptide, somatostatin, which plays a role in suppressing thermally induced pain. 38 Expression of AstC in Epi neurons raised the possibility that it is a nociceptive-suppressing neuromodulator produced by Epi neurons. To address this idea, we used the Epi-Gal4-2 to knock down AstC (UAS-AstC RNAi ). This approach was effective since the anti-AstC staining was virtually eliminated in the Epi neurons ( Figures S4A and S4B). To determine whether the reduction in AstC increased the nociceptive response, we tested whether the flies exhibited hypersensitivity to 35 C-36 C. We found that knockdown of AstC in Epi neurons (Epi-Gal4-2>UAS-AstC RNAi ) greatly increased the percentage of flies that jumped ( Figure 3A) and decreased the jump latency ( Figure 3B).
To confirm a role for AstC as a nociception modulator, we created a mutant that deleted most of the region coding for the 122 amino acid protein that is the precursor for the 15 residue AstC peptide. However, the line was lethal. Therefore, we generated an allele (AstC 1 ), which changed the C-terminal two residues from cysteine and phenylalanine to leucine and lysine (Figure S4D). The mutation reduced expression of the AstC peptide below the level of detection and was therefore a strong allele ( Figure S4C). However, AstC 1 was not a null since it did not cause lethality. Consistent with the RNAi knockdown phenotype, AstC 1 flies showed an increased propensity to jump upon contacting a 35 C-36 C surface and a decreased jump latency ( Figures 3C and 3D). Because knockdown of AstC in Epi neurons increased nociception, we tested the idea that increased expression of AstC would render flies less sensitive to nociception. Therefore, we expressed UAS-AstC under control of the Epi-Gal4-2 and tested the response to 45 C-46 C. We found that the percentage of flies that jumped declined, while the jump latency increased (Figures 3E and 3F). We then tested the idea that artificially activating Epi neurons would decrease their sensitivity to nociception. We expressed ReaChR in Epi neurons to acutely activate them with red lights (655 nm) and tested the response of the flies to 45 C-46 C. Again, the percentage of flies that jumped declined and the jump latency increased ( Figures 3I and 3J). The reduction in thermal nociception (45 C-46 C) induced by optogenetic activation of Epi neurons was suppressed by RNAi knockdown of AstC ( Figures 3I and 3J), while overexpression of AstC in combination with optogenetic stimulation of these neurons resulted in a greater alleviation of thermal nociception ( Figures 3I and 3J).
Drosophila encodes two AstC receptors, AstC-R1 and AstC-R2. 39,40 To address which of these two receptors might function in the suppression of thermal nociception, we examined the jump reactions of AstC-R1 MI04794 and AstC-R2 f01336 mutants in response to a 35 C-36 C surface. We found that mutation of AstC-R1 increased the percentage of flies that jumped and decreased the jump latency, while disruption of AstC-R2 had no effect (Figures 3G and 3H).
Epi neurons are sensors in the brain for noxious heat To test whether Epi neurons respond to noxious heat, we expressed the genetically encoded Ca 2+ sensor, GCaMP6f, 41 in Epi neurons. We dissected out the brains and monitored fluorescence changes (DF/F 0 ) in the cell bodies as we applied an 18 C-44 C temperature ramp and then decreased the temperature back to 18 C ( Figure 4A). We found the signals in Epi neurons were not increased by the lower temperatures. Between 18 C and 40 C, the DF/F 0 dipped slightly ( Figure 4D). Once the temperature reached 39 C-40 C, the Ca 2+ signals increased (Figures 4D and 4G; quantification was limited to the cells bodies, indicated by the dashed circles). Then during the ll OPEN ACCESS 44 C-18 C ramp, the fluorescence declined ( Figures 4D and  4G). When we limited the temperature rise to 32 C ( Figure 4B), there was no increase in Ca 2+ (Figures 4E and 4H).
To address whether the Epi neurons directly sense noxious heat, we applied the voltage-gated Na + channel blocker tetrodotoxin (TTX) to the brain to inhibit action potential firing. Epi neurons responded to noxious heat even in the presence of TTX ( Figures 4C, 4F, 4I, and 4J; dashed circles indicate cell bodies), supporting the conclusion that Epi neurons are direct internal sensors for noxious heat. In some samples, we were able to detect GCaMP6f signals in arborizations from the Epi neurons ( Figure 4I, panel 2). However, for consistency, these Ca 2+ signals were not included in the quantification. The absence of effect of TTX on the responses was not due to technical difficulties in applying the TTX to the brain since TTX inhibited GCaMP signals induced by ATP in neurons expression the ATP-gated P2X2 cation channel (Figures S5A-S5E).
We also compared AstC signals within Epi neurons before and after a 5-min heat shock. We examined the anti-AstC signals from flies at 25 C and after heating to 39 C (near the threshold for detecting changes in GCaMP6f fluorescence) or to 44 C. We found that either the 39 C or 44 C treatment significantly decreased the anti-AstC signals in Epi neurons ( Figures 4K-4N), suggesting that Epi neurons that are exposed to heat release AstC.

Painless required in Epi neurons to sense noxious heat
The TRP channel, Pain, is a prime candidate temperature sensor in Epi neurons since it has an activation threshold in vivo in the mid 30 C range 21 and is widely expressed in the brain. 42 To test whether pain is expressed in Epi neurons, we used a Gal4 reporter to drive membrane GFP (UAS-mCD8::GFP) and performed double labeling using anti-AstC and anti-GPF. The pain reporter was widely expressed in the brain, including AstC-positive Epi neurons (Figures 5A-5C).
To test whether the heat responsiveness of Epi neurons depends on the pain gene, we generated a pain-null allele by inserting mini-white in the coding region, thereby creating a deletion that removed +1,877-2,660 base pairs, including part of transmembrane domain 4, and all of the transmembrane domains 5 and 6 (pain 4 ; Figure 5D). The pain 4 mutation abolished the response of the Epi neurons to temperatures R40 C ( Figures 5E-5G). The small, initial dip in DF/F 0 during the early phase of the temperature ramp still occurred ( Figure 5F), demonstrating that this phase is pain independent.
To determine whether mutation of pain increases thermal nociception, we employed the jump assay. We tested 35 C-36 C since 45 C-46 C causes 100% of control flies to jump (Figure 1B). In response to 35 C-36 C, 49.6% of control flies jump, and they do so with a latency of 5.1 ± 0.2 s (Figures 5H  and 5I). We found that null pain 4 mutants exhibited a large increase in the percentage of flies that jumped ( Figure 5H; 88.4%) and a 4.1-fold decrease in the jump latency ( Figure 5I; control, 5.1 ± 0.2 s; pain 4 , 1.2 ± 0.3 s). Another Drosophila TRP channel, Pyx, is heat activated with a threshold of 40 C, 29 which is in a similar range as Pain. 21,43 Opposite to the pain mutant phenotype, we found that pyx ex44 mutants exhibited lower jump percentages and increased jump latencies at both 35 C-36 C and 45 C-46 C (Figures S5F-S5I). These results indicate that Pyx contributes to the thermal nociceptive response, rather than suppressing the response.
To address whether the pain mutant phenotype reflects a requirement for pain in Epi neurons, we used two effective RNAi lines 45 to knock down pain under control of the Epi-Gal4-2. We found that knockdown of pain specifically in Epi neurons increased the jump percentage ( Figure 5J) and reduced the average jump latency ( Figure 5K). Thus, Pain is required in Epi neurons for suppressing the nociceptive response to heat.
To address whether expression of Pain affects expression of AstC in Epi neurons, we stained both control and pain 4 mutant brains with anti-AstC. We found that pain is also required for   Figures S4E and S4F). However, AstC expression was still detected in other neurons in the pain 4 brain, such as the pair of small neurons proximal to the OLs (Figures S4E and  S4F, arrowheads).

DISCUSSION
We found that a single pair of bilaterally symmetrical Epi neurons in the fly brain is critical for suppressing thermal nociception. The importance of Epi neurons is underscored by the observation that artificial activation of these neurons is sufficient to suppress the aversive jump response to hot temperatures and that inhibition of signaling from these neurons increases the jump responses to moderate heat. The profound effect of a single pair of neurons in reducing thermal nociception is surprising given that multiple brain regions appear to function in pain suppression in mammals. 46 The dendrites of Epi neurons arborize to multiple regions of the brain, such as the OLs, the LH, and a region near the mushroom bodies, indicating that Epi neurons receive multiple signal inputs. The LH is a higher-order processing center that receives input from the antennal (olfactory) lobes and then sends relays to other ll OPEN ACCESS brain regions such as the mushroom bodies. 47 Therefore, it is intriguing to speculate that the Epi neurons may be activated by noxious odorants and aversive visual cues, which attenuate the avoidance behavioral responses to these stimuli. Epi neurons might also receive input from attractive olfactory and visual cues, which in turn diminish the escape responses to noxious stimuli such as high temperatures. In addition, the axons of Epi neurons project to the VNC, consistent with a role in descending control of motor output.
A key question is the mechanism through which Epi neurons respond to hot temperatures and alleviate thermal nociception. We found that Epi neurons are directly activated by hot temperatures and do so through activation of the thermo-TRP channel, Pain, 21 which is expressed in Epi neurons. The Pain channel is critical for suppressing nociception since mutation of the pain gene causes an increase in thermal pain sensitivity (hyperalgesia). While Epi neurons respond directly to heat and are anti-nociceptors, other neurons in the fly brain, the so-called anterior cell neurons, respond directly to suboptimal warm temperatures. 48 In contrast to the anti-nociceptive Epi neurons, the AC neurons function in thermal avoidance, which is mediated through thermal activation of TRPA1. 48 The next question is the mechanism through which activation of Epi neurons suppresses thermal pain. Epi neurons express a neuropeptide, AstC, which binds to receptors that have sequence homology (39.0% identity for AstC-R1; 38.5% identity for AstC-R2) to human opioid receptors, 49 which function in the suppression of nociception in mammals. 50,51 Moreover, mutation of AstC or knockdown of AstC in Epi neurons causes thermal hyperalgesia, and mutation of AstC-R1 elicits a similar phenotype. Heat stimulation diminishes the level of AstC in Epi neurons, indicating that activation of these neurons promotes release of AstC. We conclude that Epi neurons alleviate thermal nociception through a mechanism that depends on heat sensing by the Pain channel, leading to release of AstC.
Surprisingly, mutation of pain also reduced expression of AstC in Epi neurons below the level of detection. This effect was not due to elimination of Epi neurons since pain mutant brains express UAS-GCaMP6f under control of the Epi-Gal4. Expression of neuropeptides has been linked to neuronal activity. 52 Moreover, there is an example in which a thermosensory TRPV channel affects expression of a neuropeptide receptor. 53,54 Pain is activated by thermal heat, with the most pronounced activation in the noxious heat range. 21,43 However, even at temperatures Article significantly below the flex point in which a given temperature rapidly opens the gate of a thermosensory TRP, such as Pain, there is some channel activity. We suggest that low levels of Pain and Epi neuron activities are necessary for expression of AstC, while high levels of activities that are induced by noxious heat are required for release of the AstC. A feature of activation of Epi neurons is that the pain suppression due to an acute 30-s activation of Epi neurons is sustained for several minutes. We suggest that the slow termination of the pain suppression following stimulation of these neurons is mediated by release of the neuromodulator AstC, which persists for several minutes.
Epi neurons appear to be non-adapting, as chronic activation of these neurons with the NaChBac channel leads to similar levels of pain suppression as acute stimulation with channelrhodopsin. This non-adapting feature of Epi neurons may be beneficial because it allows for pain suppression under conditions in which the aversive response to heat needs to be suppressed sufficiently long enough to allow activities that promote survival. Given that fruit flies are poikilothermic, and their body temperature equilibrates with the environment, direct activation of Epi neurons would allow the flies to suppress nociception and enter excessively warm environments to feed or avoid predators.
In conclusion, this study unveils a molecular and cellular basis for pain suppression in Drosophila. The observation that Pain is essential for suppressing nociception is surprising given that all other thermal-TRP channels function in avoidance of suboptimal or noxious temperatures. Mutation of pain in fly larvae eliminates the sensitivity to hot temperatures (hypoalgesia). 21 Thus, it is remarkable that the same TRP channel has opposite functions in nociception and anti-nociception in larvae and adults.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:

ACKNOWLEDGMENTS
This work was supported by grants from the National Institute on Deafness and Other Communication Disorders (DC007864) and National Eye Institute (EY008117) to C.M. We would like to thank Dr. Zhefeng Gong (Zhejiang University, China), who generously allowed J.L. to carry out experiments in his laboratory during 2020 when he was restricted from returning to the USA due to the pandemic. We thank Dr. Jan Veenstra (Universit e de Bordeaux, France) for sharing the AstC antibodies and Drs. Junjie Luo, Hsiang-Chin Chen, and Yijin Wang for discussions and technical advice. We also thank the Physics Machine Shop (UC Santa Barbara) for fabricating the behavioral plates.

AUTHOR CONTRIBUTIONS
J.L. designed and conducted most of the experiments, analyzed the data, and assisted in writing the original draft. W.L. generated the pain 4 mutant, D.T. constructed the heating setup for the Ca 2+ imaging, J.M. conducted some hot plate assays and Ca 2+ imaging experiments, A.S. assisted with the hot plate assays, and C.M. supervised the project, designed experiments, analyzed data, and wrote the original draft. Article RESOURCE AVAILABILITY

Lead contact
The lead contact is Craig Montell (cmontell@ucsb.edu).

Materials availability
All unique/stable reagents generated in this study are available from the Lead Contact without restriction. The fly stocks generated in this study will be deposited with the Bloomington Drosophila Stock Center for public distribution (http://flystocks.bio.indiana.edu/). Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact.
Data and code availability d This study did not generate any standardized datatypes. d This study did not generate any original code. d Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.  -1 (v39477), UAS-pain RNAi -2 (v39478). We described pyx ex previously. 44 UAS-GCaMP3;LexAop-P2X2 was from Dr. Orie Shafer (University of Michigan). 56 The following flies were outcrossed to w 1118 (BL5905) for 6 generations: Epi-Gal4-1