Drosophila TRPAl Controls Thennotactic Behavior BY Mark Rosenzweig B. A., Biology Northwestern University, 2000 Submitted to the Department of Biology on May 25,2006, in partial Mfillment of the requirements for the Degree of Doctor of Philosophy in Biology at the Massachusetts Institute of Technology June 2006 8 2006 Massachusetts Institute of Technology All rights reserved ( i Signature of Author: 7 Certified By: I I v V - #I AA - v I JUN 0 2 2006 1 ( LIBRARIES l Department of Biology May 5,2006 \ e ., A Y r Paul A. Garrity Assistant Professor of Biology Thesis Advisor Accepted By: Stephen P. Bell Professor of Biology Chairman, Committee for Graduate students I Drosophila TRPAl Controls Thermotactic Behavior Mark Rosenzweig Submitted to the Department of Biology on May 25,2006, in partial fulfillment of the requirements for the Degree of Doctor of Philosophy in Biology ABSTRACT Temperature perception is an intricate process, essential for survival of many organisms. Temperatures far outside of the preferred range are usually harmful and are perceived as noxious (or painful), thus encouraging the animal to change its location or behavior. However, animals also have mechanisms to perceive more moderate, innocuous temperatures. Some animals exhibit clear directed movements in response to changes in temperature, a behavior known as thermotaxis. Thermotactic behavior has been most intensely studied in the nematode elegans, and a number of neurons and molecules that mediate thermotaxis in C. elegans have been identified. However, molecular mechanisms of temperature perception in general, and thermotaxis in particular, are still poorly understood. Drosophila melanogaster also exhibits strong temperature preferences and robust thermotactic behaviors, although the molecular mechanisms that mediate thermotaxis in Drosophila were unknown. We used a reverse genetic RNA interference (RNAi)based strategy to identify Drosophila TRPAl as a key regulator of thermotaxis. Larvae in which dTRPAl expression has been knocked down with RNAi fail to avoid regions of moderately elevated temperature (-31°C--350C). In heterologous cells, dTRPAl protein is activated by warming (Viswanath, et al., 2003, Nature, 423:6942), suggesting that the role dTRPAl plays in thermotaxis might be to sense the environmental temperature. We generated mutants in dTRPAl using homologous recombination-mediated insertional mutagenesis. As expected, dtrpAl mutants were defective for thermotaxis at elevated temperatures, and surprisingly were also defective for thermotaxis at cold temperatures. Interestingly, dtrpAl mutants were not defective for avoidance of all cold temperatures, but the thermotactic defects of dtrpAl mutants were more pronounced at colder temperatures than at more moderate temperatures, suggesting that other mechanisms might compensate for the loss of dTRPAl at more moderate temperatures. Temperature sensors were traditionally thought to be activated by and to mediate behavioral responses to restricted ranges of temperatures, either hot or cold temperatures, but not both. However, dTRPAl appears to be required for thermotaxis behavior at both elevated and cold temperatures, raising an exciting possibility that dTRPAl protein might mediate thermotaxis by sewing as a sensor of both hot and cold temperatures. We also identified dTRPAl-expressing cells and implicated a subset of these cells in mediating thermotactic response to elevated temperatures. Surprisingly, our results suggest that the cells required for thermotaxis at elevated temperatures (dtrpAl-promoter-Gal4-expressingcells) are located in the lawal central brain and not in the peripheral nervous system (PNS) where most thermal sensors are thought to function for mediating thermosensory behaviors. We further demonstrated that neither PAINLESS, a TRP channel required for proper lawal responses to a high-temperature (-38°C--520C) nociceptive stimulus, nor the peripherally located neurons that mediatepainless behavior, were required for thermotaxis at elevated temperatures. On the other hand, dTRPAl and the dtrpA1-promoter-Gal4-expressingcentral brain neurons that mediate thermotactic response to elevated temperatures appeared dispensable for lawal responses to a high-temperature nociceptive stimulus. These fiidings suggest that different TRP channels and different neurons are used to mediate different thermosensory behaviors in Drosophila. Thesis Advisor: Paul Garrity Title: Assistant Professor of Biology TABLE OF CONTENTS Abstract----------------------------------------------------------------------------------------------2 Chapter 1:Introduction: Mechanisms of temperature sensation------------------------------ 5 Chapter 2: The Drosophila ortholog of vertebrate TRPAl regulates thennotaxis-------- 58 Chapter 3: TRPAl mediates both heat- and cold-avoidance behaviors in Drosophila--87 Conclusion and future directions------------------------------------------.. --------------------- 128 Acknowledgements-------------------------------- -----------------------------------------------135 Chapter 1. INTRODUCTION: Mechanisms of temperature sensation Abstract. Animals rely on thermal sensors to perceive temperature information within their environment. These thermal sensors are located throughout the body and contain mechanisms necessary to sense the temperature stimulus and to pass that information onto downstream target cells, providing the animal with data useful (and sometimes indispensable) for making decisions regarding the safety or attractiveness of its current environment. Over the past decade, members of Transient Receptor Potential (TRP) family of ion channels have emerged as excellent candidates for mediating the initial step of temperature sensation. In mammals, these "thermoTRPs" can be gated by temperatures perceived as noxious hot (TRPV1 and TRPV2), innocuous/warm (TRPV3, TRPV4, TRPM4 and TRPMS), cool (TRPM4, TRPMS and TRPM8) and noxious cold (TRPAI). However, the temperature ranges thermoTRPs can be gated by in cultured neurons and heterologous cells don't tell the whole story. The activity of thermoTRPs, including their gating by temperature, can be influenced by numerous cellular mechanisms, including other ligands, lipids, phosphorylation and de-phosphorylation, localization, oligomerization and membrane potential. In addition, other proteins such as background two-pore potassium channels and members of the DegenerinEpithelial Sodium Channel (DEGIENaC) family have been demonstrated to be temperatureresponsive. Thus, there can be quite a bit of overlap between the temperature ranges capable of being sensed by different thermosensory molecules. This overlap might explain why mouse knockouts lacking individual thermoTRPs exhibit fairly subtle and restricted thermosensory behavioral phenotypes. Here I review the mechanisms of temperature perception, including how modulation of thermoTRPs might affect temperature sensation. Much work remains to be done before we have a complete list of 6 molecules and a thorough map of the circuits that allow an animal to perceive temperature stimulus and to translate that perception into a behavioral response. All organisms must possess mechanisms for detecting temperature changes in their environment. Temperature dictates and regulates many physiological processes, from diffusion of molecules within and between cells, to changes in gene expression, developmental timing and sex determination [I]. From bacteria to humans, organisms function best within certain cognate temperature ranges, function less robustly at temperatures somewhat outside of their optimal ranges and cease functioning, often dying, at temperatures even farther outside of their preferred temperatures. Furthermore, local temperature fluctuations can also cause tissue damage. Organisms have developed a multitude of strategies to meet the challenges presented by different temperatures. Certain species of plants, bacteria, insects and vertebrates can survive while frozen, and others can supercool to survive the cold, adjusting their physiology to remain unfrozen in sub-zero temperatures [2]. Another common strategy used by many organisms, including mammals and some insects such as Drosophila melanogaster, is to attempt to avoid unfavorable environments, rather than succumbing to them. In order to avoid unfavorable temperatures, an animal must sense the temperature and then translate that sensation into a behavioral response. In organisms that try to avoid unfavorable temperatures, extreme heat and cold are often perceived as painful (noxious), or at least uncomfortable, leading to movement away from the noxious environment. Temperature sensing cells. Animals that prefer to avoid unfavorable thermal environments are able to sense and respond to changes in temperature. The periphery of such animals contains projections of the thermosensory and nociceptive (pain sensing) neurons, which possess mechanisms to sense temperature and to relay the sensory information to the CNS. In 8 mammals, the cell bodies of such neurons are located in the Dorsal Root Ganglia (DRG) or the Trigerminal Ganglia (TG) and their axons are either small-diameter unmyelinated C-fibers or larger thinly myelinated A6 fibers [3]. Subsets of C- and A6 fibers are activated at noxious thermal temperatures, leading to a perception of pain, while other subsets are activated by more innocuous temperatures, perhaps involved in the sensation of comfort [4,5]. The importance of proper perception of noxious temperatures and pain is underscored by the pathologies of people unable to properly sense temperature or pain. Patients suffering from certain types of Hereditary Sensory and Autonomic Neuropathies (HSAN) exhibit a reduction or deficiency in, and/or progressive degeneration (depending on the type and expressivity of the disease) of sensory neurons, including C- and A6 fibers [6]. These patients are usually impaired for sensing (and in some cases are entirely insensitive to) temperature and pain, in some cases leading to self-mutilationlamputation; painless fractures that, due to repeated trauma, heal slowly and incorrectly (when undetected andlor untreated); and decreased lifespan [6]. Interestingly, in addition to temperature sensing neurons, keratinocytes (a type of skin cell) appear to possess thermosensory mechanisms in mice [7,8] although how keratinocytes communicate the perceived thermal information to the CNS is unclear. In addition to peripheral temperature sensors, mammals possess internal, more central thermal sensors. The preoptic-anterior hypothalamus, a structure involved in regulation of body temperature, contains many neurons that are responsive to temperature [9]. In cats and dogs, localized heating of the preoptic region of the hypothalamus can initiate panting and suppresses shivering, while cooling of this region initiates shivering [9]. The brainstem and the spinal cord are also capable of sensing temperature [101. Furthermore, the activity of many neurons in the sensorimotor cortex of a cat were found 9 to be significantly sensitive to changes in local brain temperature [l 11, although the significance of this activity for the animals' temperature perception or thermal regulation is unclear. Non-mammals are also sensitive and responsive to temperature. Nematode C. elegans exhibits a fascinating thermosensory behavior: the worms don't seem to have an innate temperature preference per se, but they are attracted to temperatures at which they were raised with food, and avoid temperatures at which they were starved [12]. In C. elegans, peripherally located AFD neurons are responsive to warming [13] and mediate the worm's thermosensory behavior through AIY and AIZ interneurons [12, 141. Insects also exhibit behavioral responses to temperature. For instance, Drosophila adults prefer temperatures around 24OC when placed onto a thermal gradient [15]. Heat can also be used as a negative reinforcer in Drosophila learning assays [I 61 and can be used to entrain the circadian clock [17]. In the adult fly, at least a fraction of thermal sensors are located peripherally, in a section of the antennae (the third antenna1 segment) [15, 181. Drosophila larvae also exhibit a variety of temperature-dependent behaviors focused on avoiding noxious thermal environments and stimuli [19-211. Certain neurons in the peripheral nervous system (PNS) of the Drosophila larvae are heat- and cold-responsive [19]. Some of these neurons (multi-dendritic (md) neurons located within the body wall) have been implicated in avoidance of a high-temperature nociceptive stimulus (38 OC52OC) [20] and others (the terminal organ in the larval nose) in avoidance of cold temperatures [19]. The terminal organ appears dispensable for larval heat avoidance ([I91 and M.R. and P.G. unpublished data). Like mammals, insects also possess central neurons that can respond to changes in temperature. For instance, neurons in the prothoracic ganglion of the cockroach are temperature-sensitive [22]. Also, our data 10 implicate neurons within the central brain of the Drosophila larvae in mediating themotactic behavior [211. In our thermotaxis assays, larvae are challenged with lower temperatures (but still well above the fly's preferred temperature) than during the hightemperature nociception assays. Wild-type larvae, but not larvae in which a subset of central brain neurons are killed or inhibited, robustly avoid the aversive temperatures. The peripherally located md neurons that mediate larval responses to a high-temperature nociceptive stimulation, however, appear dispensable for themotaxis. Inhibition of the themotaxis-mediating subset of central brain neurons, on the other hand, does not impair the response to a high-temperature nociceptive stimulus. Furthermore, the molecules that mediate these two different thermosensory behaviors appear to be distinct ([20,21] and see below). Thus, similar to mammals, Drosophila appears to possess distinct mechanisms for sensing different temperatures. Temperature sensing molecules. Identifxation of TRPVl as a candidate temperature sensor in mammals. The presence of specialized temperature-sensing neurons that rapidly (on a scale of milliseconds to seconds) change their activity in response to changes in temperature suggests the existence of specific, likely post-translational, mechanisms in these neurons for sensing temperature. A breakthrough in this field came with the expression cloning of the capsaicin receptor, vanilloid receptor 1 (VR-I), now known as TRPVl ('V' for vanilloid) (reviewed in [5,23]). When TRPVI, a member of a large family of Transient Receptor Potential (TRP) family of ion channels, was expressed in heterologous cells, TRPVl allowed influx of cations into the cell following stimulation with capsaicin or with heating (P-42*C) [24]. TRPV 1 temperature- and capsaicin-evoked currents were 11 detected even in excised membrane patches, leading to a proposal that the TRPVl channel is able to sense these stimuli directly, rather than being activated downstream of a receptor-mediated second messenger cascade [25]. TRPV 1 is expressed in many tissues, including a subset of DRG neurons that are capsaicin and heat responsive [26291. TRPVl knockout mice were reported to exhibit significant defects in neuronal responsiveness to heat. These mice exhibited profound defects in the number of cultured DRG and skin-nerve preparation neurons responding to 43OC heat [26,27] and the neurons that did respond to 43OC heat exhibited 45% reduction in heat-evoked discharge [26]. Activation of down-stream spinal chord neurons was also impaired in TRPV 1 knockout mice [26]. Furthermore, TRPVl knockout mice were impaired for the latency of response to noxious heat stimuli, (P-50°C) and failed to develop thermal hyperalgesia (hypersensitivity to temperature) in response to multiple inflammatory agents [26,27]. However, TRPVl knockout mice still retained significant behavioral responses to noxious heat stimuli and contained some heat-responsive sensory neurons, in particular, the neurons with higher thresholds of temperature-dependent activation (P-50°C) than the temperatures typically activating TRPVl in heterologous cells [26, 271. These findings are consistent with the observed presence of multiple populations of heat-responsive cultured DRG neurons, which exhibit distinct temperature activation thresholds [28,30]. The largely intact abilities of TRPVl knockout mice to respond to noxious heat stimuli, as well as the continued presence of heat-responsive neurons in those animals, indicated that while TRPVl is important for temperature perception and for mediating temperature-dependent behaviors, other, TRPV1-independent mechanisms for temperature sensation must also exist [26,27]. Overview of TRP channels. Members of the TRP family of proteins are diverse in terms of sequence, activating stimuli, expression patterns and cellular processes mediated by its members [3 11. There are 27-28 members in mammals, 16 in Drosophila, and 17 in C. elegans [32]. Based on homology, TRP proteins cluster into seven subfamilies; A (Anktml), C (Classical), M (Melastatin), ML (Mycolipin), N (NompC), P (Polycystin) and V (Vanilloid), named after the founding member of each subfamily [32]. All TRP channels are predicted to feature six transmembrane domains (TM), with a putative ion pore region between TM five and six, cytoplasmic N- and C-termini and are thought to function as tetramers [33]. Most TRPs are non-selective or poorly selective cation channels, although some TRPs do exhibit ion preferences and a few display marked selectivity in their ionic permeability [34]. ThermoTRPs as candidatesfor mediating the initial step of temperature sensation. Following the identification of TRPVl as a heat-activated ion channel ( M 3 O C ) [24,25,35], other TRPs were examined for the ability to be activated by temperature. Of the mammalian TRPs, TRPV2 can be activated by heat (W53OC) [35], TRPV3 and TRPV4 by innocuous warm, as well as by higher temperatures (P-33OC) [36-391, TRPM4 and TRPM5 by cool to innocuous temperatures (15OC<T<-35OC) [40], TRPM8 by still cooler temperatures between -8OC and -28OC [41,42], and TRPAl by noxious cold temperatures (T<-15OC) [43,44]. Considered together, the temperatures capable of activating known thermoTRPs account for almost the whole temperature spectnun that mammals are capable of sensing [5]. Figure 1. Alignment of mammalian and Drosophila TRP-family proteins. TRPs that are known to be temperature-responsive (thermoTRPs) are boxed. Drosophila TRPs are highlighted in blue. Dm - Drosophila melanogaster; Hu - Human; A - TRPA; V TRPV; ML - TRPML; PKD - TRPP; M - TRPM; C - TRPC. For the Drosophila proteins, CG designation or a name of the protein are provided. Alignment courtesy of Paul Garrity. FIGURE 1 TRP family proteins - r Drosophlla melanogaster i Tamperature-activated In culture In addition to temperature, a number of thermoTRPs can be activated by compounds perceived by humans as hot (vanilloid compounds activate TRPVl [24]), burning (cinnamaldehyde, main constituent of cinnamon oil; isothiocyanates, pungent ingredients in wasabi, horseradish and mustard oils; allicin and diallyl disulfide, pungent ingredients in garlic, all activate TRPAl [44-48]), warming (oregano and thyme, which elicit warming sensation, and camphor, which sensitizes skin to warmth, activate TRPV3 [8,49]; camphor also activates TRPV 1, although with lower potency than TRPV3 [50] and oregano, which is also perceived as pungent, can activate TRPAl [49]), or cooling (cooling agent menthol and synthetic cooling agent icilin activate TRPM8 [41,42, 5 11 and icilin also activates TRPAl with lower potency [43]). The ability of these chemical activators of thermoTRPs to elicit powerful temperature-involving sensations in humans is consistent with the proposed roles of thermoTRPs as thermal sensors. In mammals, these "ThermoTRPs" are expressed in temperature responsive tissues such as the brain, the tongue and the skin, including subsets of DRG and TG neurons that send projections to temperature-responsive tissues, in addition to thermoTRPs' expression in many other tissues not thought to participate in temperature sensation or thermal regulation [5]. For instance, TRPM4 and TRPM5 are expressed in the taste receptor neurons in the tongue, where the activation of TRPM5 by temperature potentiates G-protein coupled receptor (GPCR)-mediated sweet sensation [40, 52,531. Furthermore, in addition to their expression in DRG neurons, TRPV3 and TRPV4 are also expressed in non-neuronal keratinocytes which are known to be temperatureresponsive [7, 81. Further supporting the critical role for thermoTRPs in temperature sensation, mouse knockouts lacking individual therrnoTRPs exhibit defects in some thermosensory 16 behaviors [8,26,40,54,55]. The findings reporting temperature activation of thermoTRPs, their expression in temperature-responsive tissues, and their requirement for thermosensory behaviors lead to the development of models in which the combination of thermoTRPs provides a mechanism for the initial step of mammalian temperature perception [5]. However, temperature sensation appears to be more complicated than could be construed simply from the model above. One complication is the controversy regarding the mechanisms of noxious cold perception. The crux of the issue is that the sensory neurons responsive to strong cooling exhibit different pharmacology than would be expected if TRPAl were mediating the cold sensation in those neurons, leading some to question whether TRPA1 is the molecule that mediates the sensation of noxious cold ([45,56]; reviewed in [57,58]). There are clearly multiple populations of coldresponsive neurons, with some activated by moderate cooling (and most of these neurons are also responsive to menthol, which activates TRPM8) and others by stronger cooling (and most of these neurons are not sensitive to menthol) [56,59]. However, Jordt, et al., [45] and Nagata, et al., [60] could not reproduce earlier finding by Ardem Patapoutian's group [43] that TRPAl was activated by noxious cold in heterologous cells. And while at least two other groups have since reproduced TRPAl activation by cold in heterologous cells (in [57], cited as unpublished data; in [61], cited as personal communication), if TRPAl is a cold sensor in vivo, why the pharmacological properties of TRPAl don't match the pharmacological properties of noxious cold-sensing sensory neurons [45, 56-58] still remains to be explained. In particular, TRPAl agonist mustard oil does not stimulate most cold-sensitive neurons and mustard oil sensitive neurons are mostly cold-insensitive [45,56-581. Furthermore, TRPAl knockout mice were reported 17 to possess normal numbers and fractions of cold sensitive (both menthol-sensitive and menthol-insensitive) sensory neurons and to respond normally to challenges with cold stimuli despite being completely insensitive to mustard oil [48]. However, other studies have suggested a role for TRPAl in cold sensation. Independently derived TRPAl knockout mice have been reported to be defective in responding to a noxious cold stimulus (withdrawal from a cold plate set at O°C) and to exhibit a less distressed behavior compared to wild-type mice following evaporative cooling caused by application of a drop of acetone, while being very defective for, but not insensitive to, the responses to mustard oil [55]. Bautista, et al., [48] examined both acetone evaporative-cooling responsiveness and the responsiveness to O°C cold plate during the characterization of their TRPAl knockout mice without detecting any defects. The reason for the discrepancy between the findings of these two studies is not clear (discussed in [62]) but might be due to the differences in scoring the behavioral responses and to the use of only male mice by Bautista, et al., [48]. The later point might be especially significant because while Kwan, et al., [55] observed significant defects in cold-evoked behavioral responses in both male and female mice, the defects were more pronounced in female mice. Another factor potentially contributing to the differences between the two studies above [48,55] is difference in genetic backgrounds of the knockout mice analyzed in these studies. Additional evidence for the involvement of TRPAl in cold sensation includes a study reporting that teeth-innervating dental afferent neurons express combinations of TRPV 1, TRPM8 and TRPAl mRNAs, and that icilin-responsive menthol-insensitive dental afferent neurons (that presumably therefore express TRPAl but not TRPM8) are activated by strong cold (threshold for activation -1 7°C) [63], although this study did not 18 examine whether the observed neuronal responses to strong cold were dependent on TRPAl . Furthermore, short interfering RNA (siRNA)-mediated knockdown of TRPA 1 alleviates cold hyperalgesia in rats after inflammation or spinal nerve injury [64,65]. These findings argue that TRPAl may play a role in thermosensory processes. So how can these findings be reconciled with the studies arguing that TRPAl is not required for cold sensation? It is possible that TRPAl mediates cold sensation in a very restricted context. It is also possible that TRPAl activity and responsiveness to mustard oil andlor cinnamon oil (the primary agents used to assess TRPAl presence/responsiveness in a given cell) are regulated in vivo, perhaps by oligomerization with another protein andlor a different TRPAl splice isoform, or by post-translational modifications. Thus, responsiveness to mustard oil and to cinnamon oil might not represent all TRPAl expressing neuronal populations. Finally, keeping in mind the possibility that some TRPAl -expressing cells may not be responsive to mustard oil, TRPAl could be acting partially redundantly with another cold-sensing molecule to mediate cold perception. While the mechanisms for sensing noxious cold temperatures are not yet clear, sensation of noxious heat might also not be as simple as suggested by a parsimonious model predicting TRPVl mediating responses to heat (42OC<T<-50°C) and TRPV2 mediating responses to higher temperatures. In a recent re-examination of heat sensation in TRPVl knockout mice, it was found that contrary to previous reports [26,27], TRPVl knockout mice do not exhibit either the reduction in numbers of heat-responsive neurons or alterations in the responses of these neurons to heat [66]. The difference between this study and previous studies may be due to the use by Woodbury, et al., [66] of a more "in vivo" preparation involving explanting the spinal cord along with intact afferents rather than employing skin preparations and cultured (dissociated) DRG neuron preparations for 19 analysis [26,27]. By using cell labeling followed by electrophysiologicalrecordings and irnrnunostaining analysis, Woodbury, et al., [66] further found that most neurons responding to high heat (T=52OC) in TRPVl knockout mice did not express TRPV2. The mechanisms that mediate responses to heat in cells lacking TRPVl or TRPV2 are unknown. Behavioral studies in TRPVl [26], TRPV3 [8], TRPV4 [54], TRPM5 [40] and TRPAl [55] knockout mice support the involvement of these thermoTRPs in some types of temperature-dependent behaviors. With the exception of TRPM5's requirement for temperature modulation of taste perception, none of the thermosensory behavioral assays performed on thermoTRP knockout mice revealed complete loss of responsiveness to any given temperature or temperature-dependent stimuli. These findings may reflect overlap or redundancy in temperature responsiveness andlor expression among thermoTRPs. Such redundancies and overlaps may be quite complex because the temperature ranges thermoTRPs are sensitive to can be expanded in response to modulation by endogenous or exogenous factors (see below). In addition, mechanisms that do not involve TRP channels might also contribute to temperature sensation. Non-TRP ion channels that mayfunction in temperature sensation. In addition to thermoTRPs, other ion channels have been implicated in temperature sensation andlor mediating responses to temperature in vivo. Reid and Flonta discovered that cooling a subset of DRG neurons from 32OC to 20°C resulted in their depolarization due to inhibition of background K+ conductance (which is involved in setting the membrane potential) and proposed that a member of a two-pore-domain K+ channel family could be a target of cooling-mediated inhibition [67]. Members of a two20 pore-domain K+ channel family have indeed been demonstrated to be temperatureresponsive. A background two-pore-domain K+ channel, TREK- 1, is expressed in DRG neurons and in preoptic anterior hypothalamus, and when transfected into heterologous cells, TREK- 1 can be activated by heat (T>27"C) and inactivated by cold (12°C) [68]. Since intracellular K' concentration is higher inside the DRG and preoptic anterior hypothalamic neurons than outside of the cells, activation of TREK-1 is predicted to hyperpolarize the cell while its inhibition should cause depolarization. Thus, heatactivation and cold-inactivation of TREK-1 might contribute to temperature responses [68]. Mice lacking TREK-1 are hypersensitive to heat (between 30°C and 45°C as assessed in neuronal responses and between 46°C and 50°C as assessed in behavioral assays) but exhibit largely normal responses to cold [69]. However, following injury (sciatic nerve ligation), TREK-1 knockout mice exhibit decreased responses to acetoneinduced cooling, suggesting that TREK- 1 might participate in responses to cold [69]. Because heat-activation of TREK-1 is lost when the channel is excised fiom the cell in a membrane patch, Maingret, et al., suggested that a cytosolic factor might be required for TREK- 1's temperature responsiveness [68]. In fact, it is also possible that TREK- 1 might act as a down-stream effector of a trans-acting temperature sensor(s). Additional two-pore-domain K+ channels have been demonstrated to be capable of exhibiting temperature-responsiveness [70]. Both TREK-2 and TRAAK are activated by warming (24"C<T<42"C) in heterologous cells and in cultured neurons [70]. Interestingly, temperature activation of TREK-2 and of TRAAK is modulated by membrane potential [70], an effect also observed for a number of thermoTRPs (see below). Another two-pore-domain K+ channel implicated in mediating temperaturedependent responses is the C. elegans TWK-18. When transfected into Xenopus oocytes, TWK- 18-mediated K+ currents increase dramatically as temperature is increased from 25OC to 35OC [71]. Consistent with its proposed role as a thermo-responsive channel, twk-18(cnll O), a semi-dominant gain-of-function allele, exhibits temperature-sensitive paralysis at temperatures above 30°C [72]. The 4.7 fold higher K+ current in the mutant TWK-18(cnl10) than in the wild-type channels led Kunkel, et al., [71] to propose that an additional increase in K+ current at elevated temperatures is the basis of temperaturedependent paralysis in the twk-l8(cnll0) worms. Null twk-18 alleles have been generated but have not been examined for thermotaxis, perhaps because twk-18 promoter drives GFP expression in the body wall muscles [711. Besides two-pore-domain K+ channels, the ATP-gated cation channel P2X3 has been implicated in regulating responses to warm temperatures in mice [73]. P2X3 knockout mice are impaired for electrophysiological responses elicited by non-noxious heat (32OC<T<42OC) in wild-type mice [73,74]. Interestingly, mice lacking P2X3 exhibited enhanced preference for 30°C-35°C temperature range over higher or lower temperatures when placed onto a thermal gradient [74]. Behavioral responses to noxious heat stimuli, however, were normal [73, 741. As is the case for thermoTRPs, P2X3 is expressed in a subset of DRG neurons [73]. Thus, P2X3 might also contribute to temperature perception. Members of the DegenerinIEpithelial Sodium Channel (DEGENaC) family have also been demonstrated to exhibit temperature-dependent responses. When expressed in + is substantially stimulated by cooling and appears to Xenopus oocytes, ENaC ~ a current be inhibited by warming (the current was maximal at 6OC and minimal above 40°C) [75]. Other members of DEGENaC family, BNC1, ASIC and DRASIC did not respond to temperature changes on their own, but cooling strongly potentiated the currents induced by their agonist, acid (protons) [75]. DEGIENaC family members are expressed in DRG neurons, TG neurons and in taste-receptor cells in the tongue, and the DEGIENaC inhibitor amiloride attenuates cold activation of cultured TG neurons, raising their threshold of activation (requiring stronger cold stimulus for activation) [59]. The twopore-domain K+ channels, DEGlENaC family members and P2X3 channel are thus additional temperature-responsive proteins that may help temperature-sensing cells to integrate thermal stimuli and transduce the temperature information to downstream targets. Likely the types of neurons activated, as well as the levels of their activation, contribute to our conscious perception of temperature. Drosophila thermoTRPs. Drosophila genome contains sixteen putative TRP channels, 24 members of the DEGIENaC family and eleven two-pore-domain K+ channels [76-791. There are no known P2X3 channel homologs. Thus far, only members of the TRP family have been implicated in mediating thermosensation and thermosensory behaviors in flies. Drosophila TRPs include four TRPAs, three TRPCs, two TRPVs, one TRPM, one TRPML, one TRPN, one TRPP and three TRPP-like proteins ([76] and Fig. 1). Of these, three have been implicated in mediating thermosensory behaviors. TRPA painless is required for larval responses to a high-temperature nociceptive stimulus (38OC<T<52OC) [20]. Alleles of a different TRPA, pyrexia (pyx), slightly shift distributions of adult flies on a thermal gradient, although R N A - ~ pyx3 U ~ ~flies are distributed statistically similarly to wild-type [80]. However, pyx mutants are strongly defective for resisting paralysis at high temperature (40°C) [go]. Finally, RNAi against dTRPAl (ortholog of the mammalian TRPAI) disrupts larval thermotactic behavior [2 11. Both dTRPAl and PYX 23 proteins can be activated by elevated temperatures when expressed in heterologous cells, at D-27'C for dTRPAl [81] and T>-3540°C for PYX [80]. PAINLESS has not been reported to be temperature-activatable in heterologous cells. However, electrophysiological recordings from the abdominal nerves of wild-type larvae (abdominal nerves contain axons of PAINLESS-expressing md neurons, among others) demonstrate increasing activity in response to elevations in temperature (42'C), and this temperature-dependent activity increase is not detected in painless mutant larvae [20]. Thus, while it's unclear whether PAINLESS is a temperature-activated ion channel, PAINLESS is required for mediating behavioral responses to temperature in flies. It is possible that PAINLESS must oligomerize with another TRP channel to exhibit activation by temperature. Alternatively, PAINLESS might not function in Drosophila thermosensation as a thermal sensor but might regulate temperature-dependent behavioral responses downstream of a temperature sensation step. Thermotmis in C. elegans. The nematode C. elegans exhibits remarkably precise temperature discrimination behavior. When placed on a thermal gradient, the worms thermotax to the temperature they were cultivated on in the presence of food and can travel within that temperature zone straying less than 0.05'C from it [82]. Acclimating the worms to a different temperature for 2-4 hours in the presence of food switches the thermal preference towards the new temperature [82, 831. Starving the worms at their cultivation temperature for 1-3 hours also changes their thermal preference as the worms now avoid their cultivation temperature [83]. Thus, C. elegans can remember temperature information and these memories can be intimately tied to the presence of food. 24 Genetic analysis of C. elegans thermotaxis identified a number of molecules essential for aspects of this behavior. These molecules include transcription factors TTX1 and CEH- 14, which are required for the development and/or function of thermosensory neurons AFD [84, 851, and transcription factors TTX-3 and LIN- 11, which are required for the development and/or function of interneurons AIY and AIZ respectively [86, 871. Worms lacking either TTX-1 or CEH-14 are athermotactic, while the loss of TTX-3 makes the worms cryophilic and the loss of LIN-11 makes the worms thermophilic. This supports a model in which AFD neurons are responsible for temperature sensation while AIY and AIZ interneurons function downstream of AFD by counterbalancing the response to perceived temperature input [12]. Other factors required for thermotaxis include a cyclic nucleotide-gated channel TAX-214 [88, 891 and its upstream regulators guanylyl cyclases GCY-8, GCY- 18 and GCY-23 [90], ~a~+/calmodulin-dependent protein kinase I (CaMKI) CMK- 1 [911, neuronal calcium sensor-1 NCS- 1 [92], calcineurin AITAX-6 [93], diacylglycerol (DAG) kinase DGK- 1 [94] and protein kinase C (PKC)/TTX4 [94]. Modulation of thermotaxis by food is further mediated by the neurotransmitters serotonin and octopamine [83], by the secretory protein HEN- 1 [95], by insulin/insulin-like growth factor-1 (IGF- 1) receptor DAF-2, by the transforming growth factor P (TGFP) DAF-7 pathways [96] as well as by other, as yet uncloned, mutants affecting food modulation of thermotaxis [83]. The C. elegans thermal sensors that mediate thermotaxis, however, have not yet been described. Mechanisms of temperature sensation: Regulation of thermoTRPs. Temperature perception can exhibit profound adaptations, as anyone who ever dove into a cold pool or got out of a hot shower can attest to. These adaptations can be 25 fairly rapid, or take years (such as, for instance, acclimating to a particular climate). Members of same species are often able to inhabit diverse thermal environments. C. elegans, for instance, can prefer any of a range of temperatures within its cognate functional range, depending on the worm's prior experiences [82,83,97]. In mammals, the sensation of temperature can often be sensitized or desensitized by factors such as prolonged exposure to a particular temperature or by inflammation. While some of the plasticity in sensation of, and responses to, temperature might be accounted for by changes in neuronal connectivity accompanying such temperature shifts, the regulation of neuronal signaling, including modulation of the activity of the temperature-regulated ion channels likely plays a major role. Such modulations may also underlie in vivo adjustments to temperature thresholds and activation of various thermoTRPs and probably contribute to the apparently overlapping roles for therrnoTRPs in temperature perception. The molecular mechanisms of temperature sensor regulation are potentially complex, with multiple pathways involved. Potential modulation of temperature sensation and perception might include sensitization or desensitization of temperature receptors by various signaling pathways and by endogenous and/or exogenous ligands. Furthermore, thermosensory properties of thermoTRPs are sometimes altered by interacting proteins, including other TRP channels through changes in subunit composition, and by membrane potential. Over a longer time scale, for instance during injury or inflammation, regulation of temperature perception might also include regulation of thermoTRP expression, perhaps at the level of transcription andlor translation a d o r by controlling thermoTRPs' delivery to the cell surface. Mechanisms for modulation of TRPVl activity. TRPVl is thought to play a central role in temperature sensation and pain perception [23] and this has made TRPVl and the regulation of its activity a target of intense investigation. TRPVl is activated by natural exogenous ligands such as the hot chili pepper "hot" ingredient capsaicin [24,26,27] and the warm temperature sensitizer camphor [50]. Endogenous ligands that can activate TRPVl currents (reviewed in [98, 991) include inflammation-upregulated molecules such as protons [24,25, 100, 1011, polyamines such as spermine [102, 1031, endovanilloids such as anandamide [99, 1041, certain lipoxygenase products [105-1071, and some N-acyl dopamines [108, 1091. Other endogenous lipids can also activate TRPV 1 [109]. A number of molecules released during inflammation, including protons [25, 100, 101, 110, 1111, ATP [112-1141, Bradykinin [28, 115-1171, prostaglandins [ll8, 1191, certain chemokines (CCL3) [ 1201, certain proteases (such as trypsin and tryptase) [ 121, 1221 and nerve growth factor (NGF) [115, 123, 1241 cause thermal hyperalgesia in part by increasing temperature sensitivity of TRPVl through their cognate receptors, thus sensitizing TRPV 1. Protons can gate TRPV 1 directly and lower the threshold of TRPV 1 activation evoked by heat and by other ligands [24,25]. Bradykinin sensitizes TRPVl currents in part by stimulating production of arachidonic acid, which is a substrate of lypoxygenases [125], and ATP potentiates TRPV 1 by stimulating the production of anandamide in a ca2+-dependentmanner [104]. On the other hand, pro-analgesic adenosine, released upon TRPVl activation, binds to and inhibits TRPVl [126]. TRPVl activity can be modulated by phosphatidylinositol4,5-bisphosphate (PIP2) and by the enzymes that regulate PIP2 metabolism. PIPz inhibits TRPVl activity and activation of enzymes that reduce PIP2 quantities in the membrane, such as 27 phospholipase C (PLC) and phosphatidylinositol3-kinase (PI3K), potentiates TRPV 1 currents elicited by capsaicin, heat or protons [115]. PUK, which converts PIP2 into PIP3 [127], mediates Insulin- [I281 and NGF-induced [123, 129, 1301 potentiation of TRPVl currents and NGF-induced thermal hyperalgesia [123], although a major role of PI3K in potentiation of TRPV 1 activity might be the stimulation of TRPV1 delivery to the cell surface [129]. Inhibition of PLC attenuates TRPVl currents [I151 and potentiation of TRPVl currents by inflammatory factors Bradykinin [1311, NGF [115], prostaglandins [118], proinflamrnatory chemokine CCL3 [I201 and proteases [1211 requires PLC activity. Despite the generation of IP3 and DAG following PLC cleavage of PIP2, and the ability of DAG to activate protein kinase C (PKC), a known enhancer of TRPVl currents (see below), Chuang, et al., [I151 found that potentiation of TRPVl currents required PLC but not PKC, consistent with the hypothesis that the loss of PIPz can facilitate TRPV 1 activity. In addition to modulation by lipids, TRPVl activity is regulated by phosporylation. For instance, prostaglandins E2 and I2 cause thermal hyperalgesia in part through sensitization of TRPV1 by PKC [I181 and protein kinase A (PKA) phosphorylation [118, 119, 132, 1331. PKC also potentiates heat-evoked currents in sensory neurons and heterologous cells [28, 134-1361 and mediates TRPVl currents potentiated by Bradykinin [134, 1371, ATP [113], anandamide [135, 1381, NGF [129, 1301, chemokine CCL3 [ 1201, insulin [128] and trypsidtryptase [121, 1221. Sensitization of TRPVl and potentiation of TRPV1 currents might also be mediated by ca2+calmodulin kinase I1 (CaMKII) [139, 1401 and by the cytoplasmic tyrosine kinase Src [ 1411. Furthermore, NGF-induced thermal hyperalgesia involves phosphorylation- dependent sensitization of TRPV1 currents and this phosphorylation is mediated by PI3K, Erk, PLC, PKC, Src and CaMKII [123,129, 1301. PKC, PI3K and Src are also required for the delivery of TRPV 1 to the plasma membrane [128, 129, 1421. Activation of TRPVl by heat or agonists, such as capsaicin, in sensory neurons or in heterologous cells, is followed by TRPVl desensitization [24,25, 143, 1441. At least for the capsaicin-induced activation of TRPV1, this desensitization is mediated by ca2+ [24, 139, 143, 1441 and involves PIP2 [115], ~a~+/calmodulin [145, 1461 and TRPVl dephosphorylation mediated by ~a~+/calmodulin-dependent phosphatase calcineurin [139, 143, 1471. ca2+-dependentdesensitization of TRPV 1 is antagonized by activation of PKA [133, 147, 1481 and by CaMKII [I 391. Activation of PKC has also been demonstrated to potentiate currents of desensitized TRPV 1 [149, 1501. Interestingly, activation of PKC doesn't potentiate capsaicin-evoked TRPVl currents in ca2+-free media, consistent with de-phosphorylation of TRPV 1 being an important mechanism of ca2+-dependentdesensitization [1351. Liu, et al., however, found that activation or inhibition of PKA or PKC did not affect the inability of capsaicin to elicit currents from the desensitized TRPV 1 [15 11. Instead, the authors found that PIP2 synthesis was critical for the recovery of TRPVl from desensitization. As assessed by the activity of PIP2sensitive K' channel Kir2.1, PIP2 levels were high when TRPVl was responsive to capsaicin, low when TRPVl was desensitized and high again once TRPVl recovered fiom desensitization. Removal of PIP2 from the membrane by activation of PLC or inhibition of phosphatidylinositol4-kinase (PI4K) strongly inhibited recovery of TRPVl fiom desensitization [1511. The authors explain the discrepancies between their report and other studies of TRPVl desensitization by suggesting that the use of a prolonged capsaicin application at a high concentration by Liu, et al., [15 11 ensured a more complete desensitization of TRPV 1 than accomplished in the studies using short repeated pulses of capsaicin. It is worth noting that inhibition of calcineurin attenuated desensitization of TRPVl even during a prolonged capsaicin application [I471 while activation of PKA was able to inhibit desensitization of TRPVl following short repeated pulses of capsaicin, but not following prolonged capsaicin stimulation [148]. In contrast to the capsaicin-evoked currents, desensitization of heat-evoked TRPV 1 currents seems to be largely ca2+-independed[25, 1351. As mentioned above, NGF and insulin can stimulate surface delivery of T V l [128, 129,1421. In addition to cell surface expression being a potential mechanism of TRPV 1 regulation, injury, inflammation and growth factors released during inflammation (such as NGF) can increase expression of TRPVl in part by increasing the number of neurons that express TRPV 1 [152-1571. Additional levels of TRPVl regulation may be provided by TRPVl interactions with other signaling molecules. Immunoprecipitation experiments demonstrated that in heterologous cells TRPV 1 can bind to PLC and receptor-tyrosine kinase TRKA [115]. This or similar signaling complex in sensory neurons may couple TRKA activation (for instance by NGF) to TRPV 1 stimulation by removing TRPV 1 inhibitor PIP2 [l 151. Other TRPV 1-interacting proteins include Fas-associated factor 1 (FAFI), which binds to TRPV 1 in DRG neurons and negatively regulates TRPV 1 activation by agonists such as heat, capsaicin and acid [1581. Eferin (EF-hands-containing Rab 11125-interacting protein) is a putative vesicle trafficking protein of unknown function. Eferin can bind to TRPVl in vitro but does not dramatically affect TRPVl activation by capsaicin [159]. TRPV 1 can also bind tubulin dimers and polymerized microtubules [I601 and TRPV 1 activation induces microtubule depolymerization, providing a potential link between TRPV 1 activation and cytoskeleton dynamics [1611. However, the effect of TRPV 1:tubdin interaction on TRPV 1 activity remains to be examined. Mechanisms for modulation of other thermoTRPs. Mechanisms that regulate activation of other thermoTRPs have not been reported in as voluminous manner as they have been for TRPVI. The information that has been gathered thus far, however, implicates some of the same signaling pathways and similar regulatory mechanisms in controlling the activity of other thermoTRPs as are thought to modulate the activity of TRPV 1. For TRPV2, a high-heat (W52OC) activated ion channel [35, 1621, the endogenous agonists are unknown. However, the activity of TRPV2 is strongly potentiated by a number of growth factors such as IGF- 1 and PDGF [I621 and by neuropeptide Head Activator [163]. TRPV2 protein levels can be increased by injury and inflammation [157, 1641. Growth factor binding to their cognate receptors also stimulates delivery of TRPV2 to the cell membrane from the cytoplasmic vesicles TRPV2 occupies prior to growth factor stimulation [162, 1631 and both the potentiation of TRPV2 activity, and its delivery to the cell surface following stimulation, requires PI3K [162,163] and CaMKII [163]. In mast cells TRPV2 currents can be enhanced by the elevation of cyclic AMP levels and by the four-transmembrane domain intracellular protein RGA (Recombinase Gene Activator Protein) [I 651. Furthermore, TRPV2 is phosphorylated by, and is found in a complex with, PKA and an A-kinase adapter protein (AKAP) ACBD3 [166]. TRPV3 is activated by warm temperatures and noxious heat (W33OC) [8,36-38, 1671, by warm sensation-inducing oregano, savory and thyme [49] and by the warm 31 temperature perception-sensitizing compound camphor [8,50]. Interestingly, rather than desensitizing, warming sensitizes TRPV3 responsiveness to further increases in temperature [8,36-38, 1671, and TRPV3 currents are rapidly attenuated by even relatively small decreases in temperature [37]. TRPV3 currents can also be potentiated by inflammation-induced arachidonic acid and other unsaturated fatty acids [168], PKC activation [168], G-protein coupled receptor (GPCR)-stimulated PLC activation [49] and by increases in membrane potential [167, 1681. Heat-evoked TRPV3 currents are also partially inhibited by ca2+via an unknown mechanism [36, 1671. TRPV4 is activated by innocuous temperatures and by heat (-27OC<T<-42OC) [7, 39,169,170], epoxygenase (cytochrome P450) products of arachidonic acid metabolism [171, 1721, by phorbol esters [171, 173-1761 (in part through PKC activation [174, 1751) and by hypo-osmolarity [39, 177- 1801. In mice, besides mediating thermosensation [54, 181, 1821, TRPV4 is required for osmotic regulation [I831 and also mediates development of multiple types of hyperalgesia, including thermal, osmotic and mechanical [I8 1, 182, 184, 1851. The heat-evoked currents of TRPV4 are potentiated by hypo-osmolarity [7,391 and inhibited by hypertonic solutions [391. Hypo-osmolarity activation of TRPV4 is potentiated by prostaglandin E2 [185]. TRPV4 currents are enhanced by Src-family kinase phosphorylation [186, 1871, and are regulated by ca2+and ~a~+/calmodulin binding to TRPV4 [176, 1881. The amount of TRPV4 on the cell membrane is increased by the interaction of TRPV4 with microtubule-associated protein 7 (MAP7) [I891 and in renal cells is decreased by serinelthreonine kinases WNK1 and WNK4 [190]. TRPM8 is activated by cooling (T>-2S°C), by the cooling compound menthol and by the synthetic super-cooling compound icilin [41,42,5 1, 1911. Menthol and icilin 32 sensitize TRPM8 to cold, allowing the channel to be activated at higher temperature thresholds [41,51, 191, 1921. The application of menthol onto human skin results in cold hyperalgesia [193]. TRPM8 can be negatively regulated by PKC phosphorylation [194, 1951, which can facilitate ca2+-dependentdesensitization of the channel [195], and TRPM8 can be positively regulated by androgen [196]. TRPM8 rnRNA is also upregulated in prostate cancer cells [196]. Furthermore, TRPMS activity is intimately tied to the levels of PIP2 [19 1, 1921. Unlike PIP2's role in inhibition of TRPV 1, PIP2 activates TRPM8, with high concentrations of PIP2 activating the channel at -32-37OC, around the regular body temperature. PIP2 also potentiates TRPM8 activation by cold, menthol and icilin [191, 1921. ca2+-dependentdesensitization of TRPM8 activity [42, 5 1, 1911 is mediated by the depletion of PIP2 levels: Activation of TRPMS allows ca2+ influx into the cell, which causes ca2+-dependentPLC to translocate to the membrane, via its pleckstrin homology (PH) domain, where it reduces local PIP2 concentration, causing loss of TRPM8 activity [191]. Another study, however, reported that inhibition of PLC attenuated cold- and menthol-evoked TRPM8 currents [47], which appears at odds with the above findings suggesting positive regulation of TRPM8 activity by PIP2. The different consequences of PLC inhibition may be due to the difference in cellular models used in these studies (Xenopus oocytes, COS cells and HEK.293 cells by Rohacs, et al., [I911 and by Liu and Qin [I921 versus CHO cells used by Bandell, et al., [47]). It is also possible that in different cellular contexts PLC signaling could modulate TRPM8 activity in either positive or inhibitory manner. TRPM4 and TRPM5 are also activated by temperature, and the magnitude of this activation is modulated by membrane potential [40, 1971. At resting potentials, TRPM4 and TRPM5 are activated above -1 5OC, and the magnitude of activation by temperature increases with membrane potential [40, 1971. TRPM4 is a ca2+-actviatedcation channel [198-2001, impermeable to divalent cations [199,201], and intracellular ca2+is absolutely required for TRPM4 activity, including for the potentiation of TRPM4 by modulators [202,203], temperature [40] and membrane potential [202,203]. Activation of TRPM4 by intracellular ca2+is followed by the channel's desensitization [198,2022061, accompanied by the progressive loss of ca2+sensitivity by TRPM4 [198,202]. At least in part, the regulation of TRPM4 activation by ca2+is mediated by calmodulin, as TRPM4 activity correlates with the amount of functional calmodulin present. In the presence of ca2+,calmodulin can bind to TRPM4 and interfering with this binding abolishes most, though not all, ca2+-dependentactivation of TRPM4 [202]. Similar to its regulation of TRPM8, PIP2 antagonizes TRPM4 desensitization. PIP2 appears to increase ca2+sensitivity of TRPM4 [198,204], perhaps by slowing the rate of ca2+dissociation fiom TRPM4, because removal of ca2+fiom the medium results in a much slower rate of TRPM4 desensitization when PIP2 levels are elevated [198]. Increases in PIP2 levels shifts activation of TRPM4 by membrane potential to more negative potentials, increasing the "open state" probability (Po) of TRPM4. ATP can also antagonize TRPM4 desensitization, perhaps by promoting PIP2replenishment during desensitization andlor by increasing of association of ca2+with TRPM4 [198,202]. However, ATP can also inhibit TRPM4 activity under certain conditions [203,207,208]. In addition to ATP, PIP2 and calmodulin, PKC activity enhances ca2+sensitivity of TRPM4 [197,202] while the polyamine spermine can block TRPM4 [208]. TRPM5 is closely related to TRPM4 (-45% identity for human proteins [206]) and is also a ca2+-actviated,monovalent cation-permeable (but divalent cation impermeable) channel that requires intracellular ca2+for its activity [206,209-2111. As is the case for TRPM4, membrane potential can modulate TRPMS activity, including its activation by temperature, and this regulation also requires ca2+[40,206,209,210]. Activation of TRPMS by intracellular ca2+is followed by the channel's desensitization [209,2 101, and PIP2 can re-sensitize TRPMS [19 1,2091. TRPMS can be activated downstream of GPCR stimulation [206,209,210,212], as, for example, is thought to be the case for the mechanism of taste sensation [212,2131. GPCR stimulation is thought to activate TRPMS currents through the action of PLC, which cleaves PIP2 into DAG and IP3,with the later releasing ca2+from the intracellular stores [209]. However, while some groups reported TRPMS activation upon ca2+release from the intracellular stores [206,210,2 11,2141, other groups have failed to observe this [209,2 121. The explanation for this apparent discrepancy may be a finding that fast increase in ca2+ levels is much more effective at evoking TRPMS currents then slow ca2+increase [2101. Similar to TRPM4, TRPMS currents are blocked by spermine [208]. Unlike TRPM4, however, TRPMS does not appear to be inhibited by ATP [208], but is inhibited by protons, while TRPM4 is not [2151. TRPAl channel can be activated by cooling below -17OC [43,44,47,57], although not all groups have been able to reproduce this finding [45,57,60]. Pungent compounds found in mustard oil (isothiocyanate), garlic (allicin), cinnamon oil (cinnamaldehyde) and tear gas (acrolein) can activate TRPAl currents and produce pain and thermal and mechanical hyperalgesia [44-481. Carvacrol, an ingredient of warm sensation-evoking oregano, which is also perceived as pungent, can activate and rapidly desensitize TRPA 1 [49]. The cooling compound icilin can also activate TRPA 1 currents [43] while the warmth-sensitizing compound camphor inhibits TRPAl [SO]. TRPAl can also be activated by Bradykinin-responsive signaling pathway [47,48] through activation 35 of PLC and perhaps DAG and arachidonic acid synthesis [47], and TRPAl knockout mice do not develop thermal or mechanical hyperalgesia in response to Bradykinin [48]. TRPA1 expression is increased in response to injury or inflammation, such as that induced by NGF [64], and reduction of TRPAl levels alleviates cold hyperalgesia induced by spinal nerve injury or NGF injection [64,65]. TRPAl levels have also been reported to be deregulated in many human cancers [2 16-2 181. TRPA 1 is ubiquitinated following activation with icilin and its ubiquitination is regulated in part by a tumor suppressor de-ubiquitinating enzyme (ubiquitin hydrolase) CYLD [2171. CYLD physically interacts with and is able to de-ubiquitinate TRPAl, increasing its levels in cultured cells. Thus, TRPAl activity may be regulated by ubiquitination. ca2+might also regulate TRPA1, as release of ca2+from the internal ca2+stores has also been reported to activate TRPAl currents [45], although another group could not replicate this finding [47]. For the three putative Drosophila thermoTRPs (see above), no agonists other than temperature, antagonists or interacting proteins are known. However, dTRPAl currents in heterologous cells exhibit strong outward rectification, indicating that dTRPAl currents can be potentiated by the increases in membrane potential (see below) [811. pyrexia locus is differentially transcribed to produce two PYREXIA proteins [80]. These proteins feature the same transmembrane domains and C-terminal sequences but PYX-PB lacks much of the N-terminus found in PYX-PA. In heterologous cells, heating above 35OC activates both PYX-PA and PYX-PB currents, although the PYX-PA currents exhibit larger amplitude. Interestingly, when PYX-PA and PYX-PB are co-transfected together into cells, the current profile following activation by heat is different than that of cells in which either channel is transfected alone. Furthermore, the ion permeabilities of 36 the PYX-PA and PYX-PB co-transfected cells following heat activation were different than those of cells in which either channel was transfected alone [go]. Thus, PYREXIA channel activity appears to be regulated by differential subunit composition [go], just as the activity of some mammalian thennoTRPs (see below). Regulation of thermoTRP activity by membrane potential. Recent work indicates that membrane potential can regulate TRP activity (reviewed in [197]). Currents of a number of thermoTRPs (and some TRPs not demonstrated to be activated by temperature) exhibit outward rectification as a function of membrane potential. These thermoTRPs include TRPV 1 [197,2 191, TRPV3 [37, 1671, TRPM4 [197, 198,201-2051, TRPM5 [40,206,208-2101, TRPM8 [197,219], TRPAl [43,45] and dTRPAl [811. Positive membrane potentials act to sensitize thermoTRPs to agonists, including temperature, allowing warming- or heat-activated thermoTRPs to open at lower temperatures than at lower potentials and allowing TRPM8 to be activated at higher temperatures than at lower potentials [197,2 191. In fact, at extremely high positive potentials (+100mV) TRPM8 can be activated at 32OC, while at -80mV TRPM8 current is only detectable after cooling below 25OC [219]. Similarly, TRPVl currents can be elicited by a 17OC stimulus at potentials above +1OOmV, while at -100mV TRPVl currents are detectable only above 35OC and are significant above 40°C [219]. Biophysical analysis of thermoTRP membrane potential-gating mechanisms suggest that positive potentials increase Po (the probability that a channel is open at given conditions) while temperature and other agonists affect gating of the channels by potentially different, and sometimes separable mechanisms [197,219-22 11. One of the 37 best examples of the differences in mechanisms of thermoTRP activation employed by membrane potential and agonist gating is exhibited by TRPM4. As mentioned above, TRPM4 requires ca2+for its activity - without intracellular ca2+even the most positive potentials can't elicit TRPM4 currents - and modulation of TRPM4 currents by membrane potentials does not affect ca2+dependence of the channel, suggesting that ca2+sensitivity of TRPM4 current is independent of the voltage sensitivity of TRPM4 [205]. Further evidence for the independence of membrane potential and agonist gating of TRPM4 can be gleamed fiom the effects of decavanadate, a synthetic agonist of TRPM4 thought to interact with the ATP-binding sites of TRPM4. Decavanadate elicits TRPM4 currents at -1 00mV but not at +100mV holding potentials [203]. In a protocol where a pre-pulse to +100mV is followed by voltage clamping to negative potentials, decavanadate does not enhance the amplitude of TRPM4 current at any potential [203]. Also, decavanadate potentiates the inhibition of TRPM4 by ATP and this inhibition of TRPM4 currents by ATP in the presence of decavanadate is voltage-independent. Thus, gating of TRPM4 by ca2+(and its modulation by effector molecules) can be separated fiom the effects of membrane potential on gating [203]. While the cell is unlikely to reach membrane potentials of +100mV or -1 OOmV, modulators of thermoTRP activation, such as temperature, lipids or protein kinases may shift voltage-dependence enough, i.e. increase the Po enough, so that even the more modest voltage-dependent Po increases observed at the physiological potentials (compared to those at very positive potentials) may be sufficient to produce a cellular response [197,2 191. How does the modulation of current by membrane potential interact with the temperature activation of the channel? In one model, rise in membrane potential increases the Po of the channel while the temperature affects opening or closing rates of the channel [197,2 19,2201. For TRPV 1, if the opening rate is temperature-sensitive while the closing rate is not, an increase in temperature will activate the channel and will require less positive potential to activate it [220]. In other words, membrane potential and temperature are acting fairly independently of each other to modulate Po of the channel. This "two-state" (opened-closed) model can approximate behavior of thermoTRPs relatively well [197,219,220], although the kinetics of TRPM8 activation appear to be more complex [197,2 191 and might be more accurately described by an "allosteric" model, where temperature and membrane potential induce interacting changes in the channel's structure [22 11. Regulation of thermoTRPs by subunit composition. One additional mechanism for altering the properties of the thermal sensors is the use of alternative subunit compositions by thermoTRPs. TRP channels function as tetramers and different subunits have been shown to associate with each other [33]. For example, TRPV 1 can hetero-oligomerize with TRPV2 [222-2241 and with TRPV3 [381. When TRPVl is co-transfected with TRPV3 into heterologous cells, TRPVl currents elicited by protons or capsaicin are enhanced, suggesting that hetero-oligomerization can alter properties of thermoTRPs [38]. TRP family members can also homo-oligomerize [199,222,223]. Further increasing the diversity of TRP channels, different isoforms of individual thermoTRP have been identified. Mouse TRPVI RNA is spliced into at least two different isoforms, a (which appears to encode a full-length "canonical" channel) and f3, with mTRPV 1f3 lacking 1Oaa within the N-terminal cytoplasmic domain [225]. mTRPVl f3 does not appear to form a functional channel, although it can be trdficked to the cell membrane and can interfere with the activation of mTRPVla by agonists, acting 39 as a dominant negative [225]. mTRPVl is expressed in sensory neurons and in tongue, tissues where TRPVl is known to function [225]. Human TRPV1 also exhibits alternative splicing, with hTRPVlb identical to the canonical TRPVl except for a 60aa deletion in the N-terminal domain of hTRPV 1b [226]. hTRPV 1b can be activated essentially normally by heat (threshold of -47OC), but unlike canonical TRPV 1, hTRPVlb can't be activated by capsaicin or protons. Analysis of TG neurons identified a subpopulation of neurons that respond to heating (-47OC) but not to capsaicin, and RTPCR analysis revealed that hTRPVl b isoform is present in these neurons [226]. An isoform of rat TRPV 1, TRPVlvARencodes only a 253a.a cytoplasmic portion of the TRPVl N-terminal region, lacking the rest of the protein, including the transmembrane domains [227]. TRPVlvARpotentiates currents of the canonical TRPV 1 in HEK253 cells but inhibits TRPVl currents in COS-7 cells, suggesting context-specific regulation of TRPV 1 activity by TRPV 1vAR[227]. TRPV 1vARdoes not seem to regulate expression or cell surface delivery of TRPV1, and while the mechanism of its regulation of TRPVl is unclear, TRPVl vARprotein can bind to TRPVl and is expressed in sensory neurons [227]. A different isoform of rat TRPV 1 lacks almost the entire N-terminal domain. This isoform, VR.S1sv,is not activated by TRPVl agonists, and while its interactions with the canonical TRPVl protein have not been reported, VR.5'sv is expressed in DRG and brain neurons [228]. For human TRPV4, five splicing isoforms have been reported: TRPVB4-B, C and E lack portions of the N-terminal ankyrin repeats present in the canonical TRPV4-A, while TRPV4-D lacks 34 amino acids more N-terminal of the ankyrin repeats. TRPV4-D responds to tested stimuli as canonical TRPV4-A while B, C and E isoforms are unresponsive [229]. B, C and E isoforms are improperly glycosylated and appear to be retained in the ER [229]. Interestingly, stimuli-responsive A and D 40 isoforms form both homo-oligomers and hetero-oligomers when transfected into heterologous cells [229]. These findings suggest that regulation of expression andlor modulation of various TRPs and TRP isoforms allows for further plasticity in neuronal mechanisms for detection of temperature. Conclusions and future directions. During the last decade, several candidate ion channels for mediating the initial step of temperature sensation have been identified. These thermoTRPs respond to a broad spectra of temperatures, and various examples of regulating thermoTRPs' temperature responsiveness have been reported and are detailed above. In addition, temperature-activatable non-TRP proteins have been identified, increasing the potential diversity of the thermal sensors and the mechanisms for the cells and for the organism to sense temperature. With this apparent complexity of thermosensory mechanisms and interactions between them, the thermal sensors might be expected to exhibit quite a bit of redundancy and compensation. Indeed, mice lacking individual thermoTRPs are still capable of sensing temperatures, even within the ranges the knocked-out channels were predicted to sense. While this redundancy is likely very beneficial to the animal, given the importance of temperature perception, the redundancy among thermosensory mechanisms also makes it difficult to assess the relative contributions of the individual genes to thermal sensation. In that regard, invertebrate model systems, with their powerfid genetics and likely smaller redundancy may provide an avenue for identifying and studying molecules required for temperature perception and the mechanisms of modulation of these thermal sensors. While much has been learned about the mechanisms of temperature sensation, the story is far fiom complete. Even though under controlled conditions thermoTRPs can exhibit a remarkable expansion of the temperature ranges that can activate them, how large of a contribution to temperature sensation do such expansions make in the animal? How much do different non-TRP thermosensitive molecules contribute to the perception of the temperature stimulus? For noxious heat perception, for instance, what are the mechanisms that mediate responses to high heat in neurons lacking TRPVl and TRPV2? How do the neurons sense noxious cold in the absence of TRPMS and TRPAl? Do other thermoTRPs pick up the temperature slack or do other, perhaps as yet unidentified, molecules shoulder most of the burden? Stay tuned. The answers to these questions, as well as to other key questions such as how does temperature actually gate an ion channel on a structural level and how the modulation of the temperature gating is accomplished by interacting factors are coming. Using Drosophila thermotaxis to study molecular mechanisms of temperature sensation. The perception of temperature is a complex process requiring multiple proteins. The sensation of temperature stimuli is thought to be mediated by temperature-responsive proteins, the thermal sensors. In mammals, multiple temperature-responsive proteins have been identified. These putative thermal sensors include members of the Transient Receptor Potential (TRP) family, members of the DegeneridEpithelial Sodium Channel (DEGIENaC) family, certain two-pore potassium channels and the ATP receptor P2X3. TRP channels in particular are excellent candidates for serving as thermal sensors for mediating temperature sensation because different temperature-responsive TRPs (thermoTRPs) can be activated by different temperature spectra [ 5 ] . Furthermore, mouse knockout studies of some thermoTRPs support the involvement of thermoTRPs in temperature sensation (see above). It is interesting to note, however, that both temperature responsiveness and mouse knockouts of discrete thermoTRPs support the requirement of different thermoTRPs in the perception of discrete and relatively narrow temperatures. The putative thermal sensors are all expressed in tissues and cells known to be temperature-responsive, such as the subsets of sensory neurons in the skin and tongue. Similar to the thermals sensors, subsets of temperature-responsive cells can be activated by different temperatures ( [ 5 ] ; see above). Temperature-responsiveness of temperature-activatable proteins and of temperature-sensitive cells can be further modulated by mechanisms such as the endogenous and exogenous agonists, phosphorylatiodde-phosphorylation, different subunit composition of the thermal sensors and by membrane voltage potential. These 43 modifications of the thermal sensors' temperature responsiveness likely expand the temperatures capable of activating a given thermal sensor. The modifications of temperature-responsiveness of thermal sensors, as well as overlaps and redundancies among the thermal sensors likely contribute to animals' temperature perception. However, such redundancies and potential overlaps between the activities of thermal sensors also make it difficult to assess the relative contributions of individual proteins, and the interactions between these proteins and the modulators of their activity to thermal sensation and temperature perception. The mechanisms of temperature perception are still poorly understood. It is likely that both as yet unidentified genes and unappreciated interactions between known genes are required for temperature sensation. To better define molecular mechanisms of temperature sensation we initiated the study of thermotaxis (directed movements in response to changes in temperature) in a higher genetically tractable organism, Drosophila melanogaster. Drosophila adults and larvae exhibit a variety of robust thermosensory behaviors [15, 19-2 1, 801 including thermotaxis. The molecular mechanisms that underlie thermotactic behavior were most thoroughly investigated in the nematode C. elegans (reviewed in [12]; see above). However, while the neurons and some of the molecules required for therrnotaxis in C. elegans have been identified, the molecular mechanisms by which the worm senses temperature and translates that sensation into a behavioral response are still not well understood. When we began our studies of thermotaxis, nothing was known regarding the molecular mechanisms of thermotaxis in Drosophila, virtually nothing was known about the cells that mediate thermotaxis, and overall, very little was known about the neurons and molecules that mediate thermosensation and thermosensory behaviors in Drosophila. As a first step towards elucidating the molecular mechanisms of thermotaxis, we wished to identify the thermal sensors that mediate thermotactic behavior. We used RNA interference (RNAi)-based reverse genetic approach to knock down levels of various Drosophila relatives of mammalian thermoTRPs (since most evidence at the time suggested that thennoTRPs were excellent candidates to be thermal sensors to mediate thermotaxis). Through this approach, we identified dTRPA1, a Drosophila ortholog of mammalian thermoTRP TRPA 1, as a critical regulator of thermotaxis ([2 11; see below). Subsequent analysis of dhpA1 mutants that we have generated confirmed that dTRPAl is required for thennotaxis. Interestingly, dtrpAl mutants exhibited defects for thermotaxis at both elevated and cold temperatures, suggesting that unlike most known thermoTRPs, dTRPAl is required for behavioral responses to broad ranges of temperatures. We also identified dTRPAl -expressing cells and implicated a subset of these cells in mediating thermotaxis at elevated temperatures. How dTRPAl mediates thermotaxis is currently unclear. dTRPAl might mediate thermotaxis by serving as a temperature sensor andlor by functioning in thermotaxis downstream of the temperature-sensation step. Work from Tim Jegla's and Ardem Patapoutian's labs suggests that dTRPAl protein can be activated by warming [8 11, suggesting that dTRPAl might function as a temperature sensor at least during thermotaxis at elevated temperatures. The temperature responsiveness of dTRPAl in animals, however, has not yet been assessed. Our findings provide important initial characterization of themotaxis in Drosophila and implicate novel neurons and molecules in mediating thermotactic behavior. References. Gilbert, S.F., Developmental Biology. 5 ed. 1997, Sunderland: Sinauer Associates, Inc . Margesin, R., F. Schinner, and (Eds.), Cold-Adapted Organisms: Ecology, Physiology, Enzymology and Molecular Biology. 1 ed. 1999: Springer. 4 16. Purves, D., et al., Chapter 10: Pain, in Neuroscience. 200 1, Sinauer Associates, Inc. p. 209-222. Nomoto S, S.M., Iriki M, Riedel W, Role of afferentpathways of heat and cold in body temperature regulation. Int J Biometeorol., 2004.49: p. 67-85. Patapoutian, A., et al., ThermoTRP channels and beyond: mechanisms of temperature sensation. Nat Rev Neurosci, 2003.4(7): p. 529-39. Axelrod, F.B. and M.J. Hilz, Inherited autonomic neuropathies. Semin Neurol, 2003.23(4): p. 381-90. Chung, M.K., et al., TRPV3 and TRPV4 mediate warmth-evoked currents in primary mouse keratinocytes. J Biol Chem, 2004.279(20): p. 2 1569-75. Moqrich, A., et al., Impaired thermosensation in mice lacking TRPV3, a heat and camphor sensor in the skin. Science, 2005.307(5714): p. 1468-72. Boulant, J.A., Role of the preoptic-anterior hypothalamus in thermoregulation andfever. Clin Infect Dis, 2000.31 Suppl5: p. S 157-61. Boulant, J.A., Hypothalamic control of thermoregulation: Neurophysiological basis, in Handbook of the Hypothalamus: Behavioral Studies of the Hypothalamus, P.J. Morgane and J. Panksepp, Editors. 1980, Marcel Dekker: New York. p. 1-82. Barker, J.L. and D.O. Carpenter, Thermosensitivity of neurons in the sensorimotor cortex of the cat. Science, 1970.169(945): p. 597-8. Mori, I., Genetics of chemotaxis and thermotaxis in the nematode Caenorhabditis elegans. Annu Rev Genet, 1999.33: p. 399-422. Kimura, K.D., et al., The C. elegans thermosensory neuron AFD responds to warming. Curr Biol, 2004. 14(14): p. 1291-5. Mori, I. and Y. Ohshima, Neural regulation of thermotaxis in Caenorhabditis elegans. Nature, 1995.376(6538): p. 344-8. Sayeed, 0 . and S. Benzer, Behavioral genetics of thermosensation and hygrosensation in Drosophila. Proc Natl Acad Sci U S A, 1996.93(12): p. 607984. Wustmann, G., et al., A new paradigm for operant conditioning of Drosophila melanogaster. J Comp Physiol [A], 1996. 179(3): p. 429-36. Glaser, F.T. and R. Stanewsky, Temperature synchronization of the Drosophila circadian clock. Curr Biol, 2005. 15(15): p. 1352-63. Zars, T., Two thermosensors in Drosophila have dverent behavioral functions. J Comp Physiol [A], 2001. 187(3): p. 235-42. Liu, L., et al., IdentzBcation andfunction of thermosensory neurons in Drosophila larvae. Nat Neurosci, 2003.6(3): p. 267-73. Tracey, W.D., Jr., et al., painless, a Drosophila gene essentialfor nociception. Cell, 2003. 113(2): p. 261-73. Rosenzweig, M., et al., The Drosophila ortholog of vertebrate TRPAI regulates thermotaxis. Genes Dev, 2005. 19(4): p. 4 19-24. Murphy Jr., B.F. and J.E. Heath, Temperature sensitivity in the prothoracic ganglion of the cockroach, Periplaneta americana, and its relationship to thermoregulation. J. Exp. Biol., 1983.105: p. 305-3 15. Tominaga, M. and M.J. Caterina, Thermosensation and pain. J Neurobiol, 2004. 61(1): p. 3-12. Caterina, M.J., et al., The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature, 1997.389(6653): p. 816-24. Tominaga, M., et al., The cloned capsaicin receptor integrates multiple painproducing stimuli. Neuron, 1998.2 l(3): p. 53 1-43. Caterina, M.J., et al., Impaired nociception andpain sensation in mice lacking the capsaicin receptor. Science, 2000.288(5464): p. 306-13. Davis, J.B., et al., Vanilloid receptor-1 is essentialfor inflammatory thermal hyperalgesia. Nature, 2000.405(6783): p. 183-7. Cesare, P. and P. McNaughton, A novel heat-activated current in nociceptive neurons and its sensitization by bradykinin. Proc Natl Acad Sci U S A, 1996. 93(26): p. 15435-9. Nagy, I. and H. Rang, Noxious heat activates all capsaicin-sensitive and also a su b-population of capsaicin-insensitive dorsal root ganglion neurons. Neuroscience, 1999. 88(4): p. 995-7. Reichling, D.B. and J.D. Levine, Heat transduction in rat sensory neurons by calcium-dependent activation of a cation channel. Proc Natl Acad Sci U S A, 1997.94(13): p. 7006- 11. Montell, C., L. Birnbaumer, and V. Flockerzi, The TRP channels, a remarkably functional family. Cell, 2002. 108(5): p. 595-8. Montell, C., The TRP superfamily of cation channels. Sci STKE, 2005. 2005(272): p. re3. Schaefer, M., Homo- and heteromeric assembly of TRP channel subunits. Pflugers Arch, 2005.451(1): p. 35-42. Owsianik, G., et al., Permeation and selectivity of trp channels. Annu Rev Physiol, 2006. 68: p. 685-717. Caterina, M.J., et al., A capsaicin-receptor homologue with a high thresholdfor noxious heat. Nature, 1999. 398(6726): p. 436-4 1. Peier, A.M., et al., A heat-sensitive TRP channel expressed in keratinocytes. Science, 2002.296(5575): p. 2046-9. Xu, H., et al., TRPV3 is a calcium-permeable temperature-sensitive cation channel. Nature, 2002. 418(6894): p. 181-6. Smith, G.D., et al., TRPV3 is a temperature-sensitive vanilloid receptor-like protein. Nature, 2002. 418(6894): p. 186-90. Guler, A.D., et al., Heat-evoked activation of the ion channel, TRPV4. J Neurosci, 2002.22(15): p. 6408-14. Talavera, K., et al., Heat activation of T M M 5 underlies thermal sensitivity of sweet taste. Nature, 2005. 438(7070): p. 1022-5. Peier, A.M., et al., A TRP channel that senses cold stimuli and menthol. Cell, 2002. 108(5): p. 705-15. McKemy, D.D., W.M. Neuhausser, and D. Julius, Identzfication of a cold receptor reveals a general role for TRP channels in thermosensation. Nature, 2002. 416(6876): p. 52-8. Story, G.M., et al., ANKTMl, a TRP-like channel expressed in nociceptive neurons, is activated by cold temperatures. Cell, 2003.112(6): p. 819-29. Macpherson, L.J., et al., The pungency of garlic: activation of TRPAl and TRPVl in response to allicin. Curr Biol, 2005. 15(10): p. 929-34. Jordt, S.E., et al., Mustard oils and cannabinoids excite sensory nerve fibres through the TRP channel ANKTMl. Nature, 2004.427(6971): p. 260-5. Bautista, D.M., et al., Pungent productsflorn garlic activate the sensory ion channel TRPAl. Proc Natl Acad Sci U S A, 2005.102(34): p. 12248-52. Bandell, M., et al., Noxious cold ion channel TRPAl is activated by pungent compounds and bradyknin. Neuron, 2004.41(6): p. 849-57. Bautista, D.M., et al., TRPAl mediates the inflammatory actions of environmental irritants andproalgesic agents. Cell, 2006.124(6): p. 1269-82. Xu, H., et al., Oregano, thyme and clove-derivedflavors and skin sensitizers activate specific TRP channels. Nat Neurosci, 2006.9(5): p. 628-635. Xu, H., N.T. Blair, and D.E. Clapham, Camphor activates and strongly desensitizes the transient receptor potential vanilloid subtype 1 channel in a vanillid-independent mechanism. J Neurosci, 2005.25(39): p. 8924-37. Chuang, H.H., W.M. Neuhausser, and D. Julius, The super-cooling agent icilin reveals a mechanism of coincidence detection by a temperature-sensitive TRP channel. Neuron, 2004.43(6): p. 859-69. Cruz, A. and B.G. Green, Thermal stimulation of taste. Nature, 2000.403(6772): p. 889-92. Temussi, P., The history of sweet taste: not exactZy a piece of cake. J Mol Recognit, 2006. Lee, H., et al., Altered thermal selection behavior in mice lacking transient receptor potential vanilloid 4. J Neurosci, 2005.25(5): p. 1304-10. Kwan, K.Y., et al., TRPAl contributes to cold, mechanical, and chemical nociception but Is not essentialfor hair-cell transduction. Neuron, 2006.50: p. 277-289. Babes, A., D. Zorzon, and G. Reid, Twopopulations of cold-sensitive neurons in rat dorsal root ganglia and their modulation by nerve growthfactor. Eur J Neurosci, 2004.20(9): p. 2276-82. Reid, G., ThermoTRP channels and cold sensing: what are they really up to? Pflugers Arch, 2005.451(1): p. 250-63. McKemy, D.D., How cold is it? TRPM8 and TRPAl in the molecular logic of cold sensation. Mol Pain, 2005. l(1): p. 16. Thut, P.D., D. Wrigley, and M.S. Gold, Cold transduction in rat trigeminal ganglia neurons in vitro. Neuroscience, 2003.119(4): p. 1071-83. Nagata, K ., et al., Nociceptor and hair cell transducerproperties of TRPAI, a channelfor pain and hearing. J Neurosci, 2005.25(16): p. 4052-6 1. Dhaka, A*,V. Viswanath, and A. Patapoutian, TRP Ion Channels and Temperature Sensation. Annu Rev Neurosci, 2006. Story, G.M. and R.W.t. Gereau, Numbing the senses: role of TRPAl in mechanical and cold sensation. Neuron, 2006.50(2): p. 177-80. Park, C.K., et al., Functional expression of thermo-transient receptor potential channels in dental primary afferent neurons: Implication for toothpain. J Biol Chem, 2006. Obata, K., et al., TRPAl induced in sensory neurons contributes to cold hyperalgesia a f t r inflammation and nerve injury. J Clin Invest, 2005.115(9): p. 2393-401. Katsura, H., et al., Antisense knock down of TRPAI, but not TRPM8, alleviates cold hyperalgesia after spinal nerve ligation in rats. Exp Neurol, 2006. Woodbury, C.J., et al., Nociceptors lacking TRPVl and TRPV2 have normal heat responses. J Neurosci, 2004.24(28): p. 64 10-5. Reid, G. and M. Flonta, Cold transduction by inhibition of a background potassium conductance in rat primary sensory neurones. Neurosci Lett, 2001. 297(3): p. 171-4. Maingret, F., et al., TREK-1 is a heat-activated background K(+) channel. Embo J,2000.19(11): p. 2483-9 1. Alloui, A., et al., TREK-1, a K(+) channel involved in polymodal pain perception. Embo J , 2006. Kang, D., C. Choe, and D. Kim, Thermosensitivityof the two-pore domain K+ channels TREK-2 and TRAAK J Physiol, 2005.564(Pt 1): p. 103-16. Kunkel, M.T., et al., Mutants of a temperature-sensitivetwo-P domainpotassium channel. J Neurosci, 2000.20(20): p. 75 17-24. Hosono, R., S. Kuno, and M. Midsukami, Temperature-sensitive mutations causing reversible paralysis in Caenorhabditis elegans. J Exp 2001, 1985.235(3): p. 409-21. Souslova, V., et al., Warm-coding deficits and aberrant inflammatorypain in mice lacking P2X3 receptors. Nature, 2000.407(6807):p. 1015-7. Shimizu, I., et al., Enhanced thermal avoidance in mice lacking the ATP receptor P2X3. Pain, 2005.116(1-2):p. 96-108. Askwith, C.C., et al., DEG/ENaC ion channels involved in sensory transduction are modulated by cold temperature. Proc Natl Acad Sci U S A, 2001.98(11): p. 6459-63. Montell, C., Drosophila TRP channels. Pflugers Arch, 2005.451(1): p. 19-28. Kellenberger, S. and L. Schild, Epithelial sodium channevdegenerinfamily of ion channels: a variety offunctions for a shared structure. Physiol Rev, 2002.82(3): p. 735-67. Buckingharn, S.D., et al., Structure andfunction of two-pore-domain K+ channels: contributionsfrom genetic model organisms. Trends Pharmacol Sci, 2005.26(7): p. 361-7. Littleton, J.T. and B. Ganetzky, Ion channels and synaptic organization: analysis of the Drosophila genome. Neuron, 2000.26(1): p. 35-43. Lee, Y., et al., Pyrexia is a new thermal transient receptor potential channel endowing tolerance to high temperatures in Drosophila melanogaster. Nat Genet, 2005.37(3): p. 305-10. Viswanath, V., et al., Opposite thermosensor infiuitjly and mouse. Nature, 2003. 423(6942): p. 822-3. Hedgecock, E.M. and R.L. Russell, Normal and mutant thermotaxis in the nematode Caenorhabditis elegans. Proc Natl Acad Sci U S A, 1975.72(10):p. 4061-5. Mohri, A., et al., Genetic control of temperature preference in the nematode Caenorhabditiselegans. Genetics, 2005.169(3): p. 1437-50. Satterlee, J.S., et al., Specification of thermosensory neuronfate in C. elegans requires ttx-1, a homolog of otd/Otx. Neuron, 200 1.31(6): p. 943-56. Cassata, G., et al., The LIM homeobox gene ceh-14 confers thermosensory finction to the AFD neurons in Caenorhabditis elegans. Neuron, 2000.25(3): p. 587-97. Hobert, O., et al., Regulation of interneuronfunction in the C. elegans thermoregulatory pathway by the tbc-3 LIM homeobox gene. Neuron, 1997.19(2): p. 345-57. Hobert, O., et al., Control of neural development andfinction in a thermoregulatory network by the LIM homeobox gene lin-11. J Neurosci, 1998. 18(6): p. 2084-96. Coburn, C.M., et al., A cyclic nucleotide-gated channel inhibits sensory axon outgrowth in larval and adult Caenorhabditis elegans: a distinct pathway for maintenance of sensory axon structure. Development, 1998.125(2): p. 249-58. Komatsu, H., et al., Mutations in a cyclic nucleotide-gated channel lead to abnormal thermosensationand chemosensation in C. elegans. Neuron, 1996. 17(4): p. 707- 18. Inada, H., et al., Identification of guanylyl cyclases thatfinction in thermosensory neurons of Caenorhabditis elegans. Genetics, 2006. Satterlee, J.S., W.S. Ryu, and P. Sengupta, The CMK-1 CaMKland the TAX-4 Cyclic nucleotide-gated channel regulate thermosensory neuron gene expression andfinction in C. elegans. Curr Biol, 2004. 14(1): p. 62-8. Gomez, M., et al., Ca2+ signaling via the neuronal calcium sensor-1 regulates associative learning and memory in C. elegans. Neuron, 2001.30(1): p. 24 1-8. Kuhara, A., et al., Negative regulation and gain control of sensory neurons by the C. elegans calcineurin TAX-6. Neuron, 2002.33(5): p. 75 1-63. Okochi, Y., et al., Diverse regulation of sensory signaling by C. elegans nPKCepsilodeta TTX-4. Embo J, 2005.24(12): p. 2 127-37. Ishihara, T., et al., HEN-1, a secretory protein with an LDL receptor mot$ regulates sensory integration and learning in Caenorhabditis elegans. Cell, 2002. 109(5): p. 639-49. Murakami, Hayet al., Aging-dependent and -independent modulation of associative learning behavior by insulirdinsulin-like growthfactor-1 signal in Caenorhabditis elegans. J Neurosci, 2005.25(47): p. 10894-904. Ito, H., H. Inada, and I. Mori, Quantitative analysis of thermotaxis in the nematode Caenorhabditiselegans. J Neurosci Methods, 2006. De Petrocellis, L. and V. Di Marzo, Lipids as regulators of the activity of transient receptor potential type Vl (TRPVl) channels. Life Sci, 2005. 77(14): p. 165 1-66. Van Der Stelt, M. and V. Di Marzo, Endovanilloids. Putative endogenous ligands of transient receptor potential vanilloid 1 channels. Eur J Biochem, 2004. 271(10): p. 1827-34. Bevan, S. and P. Geppetti, Protons: small stimulants of capsaicin-sensitive sensory nerves. Trends Neurosci, 1994.17(12):p. 509- 12. Jones, N.G., et al., Acid-inducedpain and its modulation in humans. J Neurosci, 2004.24(48): p. 10974-9. Ahern, G.P., X. Wang, and R.L. Miyares, Polyamines Are Potent Ligands for the Capsaicin Receptor TRPVI. J Biol Chem, 2006.281(13): p. 899 1-5. Zhang, M., H. Wang, and K.J. Tracey, Regulation of macrophage activation and inflammation by spermine: a new chapter in an old story. Crit Care Med, 2000. 28(4 Suppl): p. N60-6. Van Der Stelt, M., et al., Anandamide acts as an intracellular messenger amplzjjing Ca2+ influx via TRPVl channels. Embo J, 2005.24(17): p. 3026-37. Shin, J., et al., BradyRnin-12-lipoxygenase-VRI signaling pathway for inflammatory hyperalgesia. Proc Natl Acad Sci U S A, 2002.99(15): p. 10150-5. Hwang, S.W., et al., Direct activation of capsaicin receptors by products of lipoxygenases: endogenous capsaicin-like substances. Proc Natl Acad Sci U S A, 2000.97(11): p. 6155-60. Levine, J.D., et al., Hyperalgesic properties of 15-lipoxygenaseproducts of arachidonic acid Proc Natl Acad Sci U S A, 1986.83(14):p. 533 1-4. Chu, C.J., et al., N-oleoyldopamine, a novel endogenous capsaicin-like lipid that produces hyperalgesia. J Biol Chem, 2003.278(16): p. 13633-9. Movahed, P., et al., Endogenous unsaturated C18 N-acylethanolamines are vanilloid receptor (TRPVI) agonists. J Biol Chem, 2005.280(46): p. 38496-504. Khasar, S.G., et al., A novel nociceptor signalingpathway revealed in protein kinase C epsilon mutant mice. Neuron, 1999.24(1):p. 253-60. Leffler, A., B. Monter, and M. Koltzenburg, The role of the capsaicin receptor TRPVl and acid-sensing ion channels (ASICS) in proton sensitivity of subpopulations ofprimary nociceptive neurons in rats and mice. Neuroscience, 2006.139(2): p. 699-709. Moriyama, T., et al., Possible involvement of P2Y2 metabotropic receptors in ATP-induced transient receptor potential vanilloid receptor 1-mediated thermal hypersensitivity. J Neurosci, 2003.23(14): p. 6058-62. Tominaga, M., M. Wada, and M. Masu, Potentiation of capsaicin receptor activity by metabotropic ATP receptors as a possible mechanismfor ATP-evoked pain and hyperalgesia. Proc Natl Acad Sci U S A, 2001.98(12): p. 695 1-6. Hamilton, S.G., A. Wade, and S.B. McMahon, The effects of inflammation and inflammatory mediators on nociceptive behaviour induced by ATP analogues in the rat. Br J Pharrnacol, 1999.126(1):p. 326-32. Chuang, H.H., et al., Bradykinin and nerve growthfactor release the capsaicin receptor @om PtdIm(4,5)P2-mediated inhibition. Nature, 200 1 . 41l(6840): p. 957-62. Manning, D.C., et al., Pain and hyperalgesia afleer intradermal injection of bradyknin in humans. Clin Pharmacol Ther, 1991.50(6): p. 721-9. Sugiura, T., et al., Bradykinin lowers the threshold temperaturefor heat activation of vanilloid receptor I. J Neurophysiol, 2002.88(1): p. 544-8. Moriyama, T., et al., Sensitization of TRPVl by EPI and IP reveals peripheral nociceptive mechanism ofprostaglandins. Mol Pain, 2005. l(1): p. 3. Rathee, P.K., et al., PWAKAPNR-I module: A common link of Gs-mediated signaling to thermal hyperalgesia. J Neurosci, 2002.22(11): p. 4740-5. Zhang, N., et al., A proinflammatory chemokine, CCL;1,sensitizes the heat- and capsaicin-gated ion channel TRPVI. Proc Natl Acad Sci U S A, 2005.102(12): p. 4536-41. Amadesi, S., et al., Protease-activated receptor 2 sensitizes the capsaicin receptor transient receptor potential vanilloid receptor 1 to induce hyperalgesia. J Neurosci, 2004.24(18): p. 4300- 12. Dai, Y., et al., Proteinase-activated receptor 2-mediatedpotentiation of transient receptor potential vanilloid subfamily 1 activity reveals a mechanismfor proteinme-induced inflammatorypain. J Neurosci, 2004.24(18): p. 4293-9. Zhuang, Z.Y., et al., Phosphatidylinositol3-kinaseactivates ERK in primary sensory neurons and mediates inJammatory heat hyperalgesia through TRPVl sensitization. J Neurosci, 2004.24(38): p. 8300-9. Julius, D. and A.I. Basbaum, Molecular mechanisms of nociception. Nature, 2001. 413(6852): p. 203-10. Thayer, S.A., T.M. Pemey, and R.J. Miller, Regulation of calcium homeostasis in sensory neurons by bradyknin. J Neurosci, 1988.8(11): p. 4089-97. Puntambekar, P., et al., Direct interaction of adenosine with the TRPVl channel protein. J Neurosci, 2004.24(14): p. 3663-71. Cantley, L.C., The phosphoinositide 3-kinase pathway. Science, 2002.296(5573): p. 1655-7. Van Buren, J.J., et al., Sensitization and translocation of TRPVl by insulin and IGF-I. Mol Pain, 2005. l(1): p. 17. Zhang, X., J. Huang, and P.A. McNaughton, NGF rapidly increases membrane expression of TRPVl heat-gated ion channels. Embo J, 2005.24(24): p. 42 1 1 -23. Bonnington, J.K. and P.A. McNaughton, SignaNingpathways involved in the sensitisation of mouse nociceptive neurones by nerve growth factor. J Physiol, 2003.551(Pt 2): p. 433-46. Tang, H.B., et al., Sensitization of vanilloid receptor 1 induced by bradykinin via the activation of second messenger signaling cascades in rat primary afferent neurons. Eur J Pharmacol, 2004.498(1-3): p. 37-43. Lopshire, J.C. and G.D. Nicol, The CAMPtransduction cascade mediates the prostaglandin E2 enhancement of the capsaicin-elicited current in rat sensory neurons: whole-cell and single-channel studies. J Neurosci, 1998.18(16):p. 608 1-92. Bhave, G., et al., CAMP-dependentprotein kinase regulates desensitization of the capsaicin receptor (VRl) by direct phosphorylation. Neuron, 2002.35(4): p. 72 131. Cesare, P., et al., Specifc involvement of PKC-epsilon in sensitization of the neuronal response to painful heat. Neuron, 1999.23(3):p. 6 1 7-24. Vellani, V., et al., Protein kinase C activationpotentiates gating of the vanilloid receptor VRl by capsaicin, protons, heat and anandamide. J Physiol, 2001. 534(Pt 3): p. 813-25. Numazaki, M., et al., Direct phosphorylation of capsaicin receptor VRl by protein kinase Cepsilon and identifcation of two target serine residues. J Biol Chem, 2002.277(16): p. 13375-8. Premkumar, L.S. and G.P. Ahem, Induction of vanilloid receptor channel activity by protein kinase C. Nature, 2000.408(6815):p. 985-90. Lee, S.Y., et al., Sensitization of vanilloid receptor involves an increase in the phosphorylated form of the channel. Arch Pharm Res, 2005.28(4): p. 405-12. Jung, J., et al., Phosphorylation of vanilloid receptor I by Ca2+/calmodulindependent kinase 11regulates its vanilloid binding. J Biol Chem, 2004.279(8): p. 7048-54. Price, T.J., et al., Pharmacological interactions between calciudcalmodulindependent kinase 11alpha and TRPVI receptors in rat trigeminal sensory neurons. Neurosci Lett, 2005.389(2): p. 94-8. Jin, X., et al., Modulation of TRPVI by nonreceptor tyrosine kinase, c-Src kinase. Am J Physiol Cell Physiol, 2004.287(2): p. C558-63. Morenilla-Palao, C., et al., Regulated exocytosis contributes to protein kinase C potentiation ofvanilloid receptor activity. J Biol Chem, 2004.279(24): p. 2566572. Docherty, R.J., et al., Inhibition of calcineurin inhibits the desensitization of capsaicin-evoked currents in cultured dorsal root ganglion neuronesfrom adult rats. Pflugers Arch, 1996.431(6): p. 828-37. Koplas, P.A., R.L. Rosenberg, and G.S. Oxford, m e role of calcium in the desensitization of capsaicin responses in rat dorsal root ganglion neurons. J Neurosci, 1997.17(10):p. 3525-37. Rosenbaum, T., et al., Ca2+/calmodulin modulates TRPVI activation by capsaicin. J Gen Physiol, 2004.123(1):p. 53-62. Numazaki, M., et al., Structural determinant of TRPVI desensitization interacts with calmodulin. Proc Natl Acad Sci U S A, 2003. lOO(13):p. 8002-6. Mohapatra, D .P. and C. Nau, Regulation of Ca2+-dependent desensitization in the vanilloid receptor TRPVI by calcineurin and CAMP-dependentprotein kinase. J Biol Chem, 2005.280(14):p. 13424-32. Mohapatra, D.P. and C. N a yDesensitization of capsaicin-activated currents in the vanilloid receptor TRPVI is decreased by the cyclic AMP-dependent protein kinase pathway. J Biol Chem, 2003.278(50):p. 50080-90. Mandadi, S., et al., Increased sensitivity of desensitized TRPVI by PU4 occurs through PKCepsilon-mediatedphosphorylationat S800. Pain, 2006. Mandadi, S., et al., Activation ofprotein kinase C reverses capsaicin-induced calcium-dependent desensitization of TRPVI ion channels. Cell Calcium, 2004. 35(5):p. 47 1-8. Liu, B., C. Zhang, and F. Qin, Functional recoveryfrom desensitization of vanilloid receptor TRPVI requires resynthesis of phosphatidylinositol4,5bisphosphate. J Neurosci, 2005.25(19): p. 4835-43. Ji, R.R., et al.,p38 M P K activation by NGF in primary sensory neurons afrer inflammation increases TRPV. levels and maintains heat hyperalgesia. Neuron, 2002.36(1): p. 57-68. Amaya, F., et al., NGF and GDNF d$ierentially regulate TRP VI expression that contributes to development of inflammatory thermal hyperalgesia. Eur J Neurosci, 2004.20(9): p. 2303-10. Price, T.J., et al., Treatment of trigeminal ganglion neurons in vitro with NGF, GDNF or BDNF: efects on neuronal survival, neurochemicalproperties and TRPVI-mediated neuropeptide secretion. BMC Neurosci, 2005.6(1): p. 4. Matthews, P.J., et al., Increased capsaicin receptor TRPVl nerve fibres in the infamed human oesophagus. Eur J Gastroenterol Hepatol, 2004.16(9): p. 897902. Biggs, J.E., et al., Changes in vanilloid receptor 1 (TRPVl) expressionfollowing lingual nerve injury. Eur J Pain, 2006. Shimosato, G., et al., Peripheral inflammation induces up-regulation of TRPV2 expression in rat DRG. Pain, 2005.119(1-3):p. 225-32. Kim, S., et al., TRPVl recapitulates native capsaicin receptor in sensory neurons in association with Fas-associatedfactor I. J Neurosci, 2006.26(9): p. 2403- 12. Lee, S.Y., Identification of a protein that interacts with the vanilloid receptor. Biochem Biophys Res Commun, 2005.331(4): p. 1445-51. Goswami, C., et al., IdentiJication and characterization of a Ca2+ -sensitive interaction of the vanilloid receptor TRPVl with tubulin. J Neurochem, 2004. 91(5):p. 1092-103. Goswami, C., et al., Rapid disassembly of dynamic microtubules upon activation of the capsaicin receptor TRPVl. J Neurochem, 2006.96(1): p. 254-66. Kanzaki, M., et al., Translocation of a calcium-permeable cation channel induced by insulin-like growth factor-I. Nat Cell Biol, 1999. l(3):p. 165-70. Boels, K., et al., The neuropeptide head activator induces activation and translocation of the growth-factor-regulated Ca(2+)-permeable channel GRC. J Cell Sci, 2001.114(Pt 20):p. 3599-606. Gaudet, A.D., et al., Regulation of TRPV2 by axotomy in sympathetic, but not sensory neurons. Brain Res, 2004.1017(1-2): p. 1 55-62. Stokes, A.J., et al., Formation of a physiological complex between TRPV2 and RGA protein promotes cell surface expression of TRP V2. J Cell Biochem, 2005. 94(4):p. 669-83. Stokes, A.J., et al., A TRPV2-PKA signaling modulefor transduction ofphysical stimuli in mast cells. J Exp Med, 2004.200(2): p. 137-47. Chung, M.K., A.D. Guler, and M.J. Caterina, Biphasic currents evoked by chemical or thermal activation of the heat-gated ion channel, TRPV3. J Biol Chem, 2005.280(16): p. 15928-41. Hu, H.Z., et al., Potentiation of TRPV3 channelfunction by unsaturatedfatty acids. J Cell Physiol, 2006. Watanabe, H., et al., Heat-evoked activation of TRPV4 channels in a HEK293 cell expression system and in native mouse aorta endothelial cells. J Biol Chem, 2002. 277(49):p. 47044-5 1. Chung, M.K., H. Lee, and M.J. Caterina, Warm temperatures activate TRPV4 in mouse 308 keratinocytes. J Biol Chem, 2003.278(34):p. 32037-46. Watanabe, H., et al., Anandamide and arachidonic acid use epoxyeicosatrienoic acids to activate TRPV4 channels. Nature, 2003.424(6947): p. 434-8. Vriens, J., et al., Modulation of the Ca2permeable cation channel TRPV4 by cytochrome P450 epoxygenases in vascular endothelium. Circ Res, 2005.97(9): p. 908-15. Watanabe, H., et al., Activation of TRPV4 channels FVRL-2/mTRP12) byphorbol derivatives. J Biol Chem, 2002.277(16): p. 13569-77. Xu, F., E. Satoh, and T. Iijima, Protein kinase C-mediated Ca2+ entry in HEK 293 cells transiently expressing human TRPV4. Br J Pharmacol, 2003. 140(2):p. 413-21. Gao, X., L. Wu, and R.G. O'Neil, Temperature-modulated diversity of TRPV4 channel gating: activation by physical stresses and phorbol ester derivatives throughprotein kinase C-dependent and -independent pathways. J Biol Chem, 2003.278(29): p. 27 129-37. Watanabe, H., et al., Modulation of TRPV4 gating by intra- and extracellular Ca2+. Cell Calcium, 2003.33(5-6): p. 489-95. Liedtke, W., et al., Vanilloid receptor-related osmotically activated channel (VROAC), a candidate vertebrate osmoreceptor. Cell, 2000.103(3): p. 525-35. Strotmann, R., et al., OTRPC4, a nonselective cation channel that confers sensitivity to extracellular osmolarity. Nat Cell Biol, 2000.2(10): p. 695-702. Wissenbach, U., et al., Trp12, a novel Trp relatedprotein fLom kidney. FEBS Lett, 2000.485(2-3): p. 127-34. Delany, N.S., et al., Identijication and characterization of a novel human vanilloid recepror-likeprotein, VRL-2. Physiol Genomics, 200 1.4(3): p. 165-74. Todaka, H., et al., Warm temperature-sensitive transient receptor potential vanilloid 4 (TRPV4)plays an essential role in thermal hyperalgesia. J Biol Chem, 2004.279(34): p. 35133-8. Suzuki, M., et al., Impairedpressure sensation in mice lacking TRPV4. J Biol Chem, 2003.278(25): p. 22664-8. Liedtke, W. and J.M. Friedman, Abnormal osmotic regulation in trpv4-/- mice. Proc Natl Acad Sci U S A, 2003. lOO(23): p. 13698-703. Alessandri-Haber, N., et al., A transient receptor potential vanifloid 4-dependent mechanism of hyperalgesia is engaged by concerted action of inflammatory mediators. J Neurosci, 2006.26(14): p. 3864-74. Alessandri-Haber, N., et al., Hypotonicity induces TRPV4-mediated nociception in rat. Neuron, 2003.39(3): p. 497-5 11. Xu, H., et al., Regulation of a transient receptor potential (TRP) channel by tyrosinephosphorylation. SRC family kinase-dependent tyrosine phosphorylation of TRPV4 on TYR-253 mediates its response to hypotonic stress. J Biol Chem, 2003.278(13): p. 11520-7. Alessandri-Haber, N., et al., Transient receptor potential vanilloid 4 is essential in chemotherapy-induced neuropathic pain in the rat. J Neurosci, 2004.24(18): p. 4444-52. Strotmann, ReyG. Schultz, and T.D. Plant, Ca2+-dependentpotentiationof the nonselective cation channel TRPV4 is mediated by a C-terminal calmodulin binding site. J Biol Chem, 2003.278(29): p. 2654 1-9. Suzuki, M., A. Hirao, and A. Mizuno, Microtubule-associated [correctedlprotein 7 increases the membrane expression of transient receptor potential vanilloid 4 (TRPV4). J Biol Chem, 2003.278(5 1): p. 5 1448-53. Fu, Y., et al., WNK kinases influence TRPV4 channeljiknction and localization. Am J Physiol Renal Physiol, 2006. Rohacs, T., et al., PI(4,5)P2 regulates the activation and desensitization of TRPM8 channels through the TRP domain. Nat Neurosci, 2005. 8(5): p. 626-34. Liu, B. and F. Qin, Functional control of cold- and menthol-sensitive TRPM8 ion channels by phosphatidylinositol 4,5-bisphosphate. J Neurosci, 2005.25(7) :p. 1674-81. Narner, B., et al., TRPAl and TRPM8 activation in humans: effects of cinnamaldehyde and menthol. Neuroreport, 2005.16(9): p. 955-9. Premkumar, L. S., et al., Downregulation of transient receptor potential melastatin 8 by protein Rinase C-mediated dephosphorylation. J Neurosci, 2005. 25(49): p. 1 1322-9. Abe, J., et al., Ca2+-dependent PKC activation mediates menthol-induced desensitization of transient receptor potential M8. Neurosci Lett, 2006.397(1-2): p. 140-4. Zhang, L. and G.J. Barritt, Evidence that TRPM8 is an androgen-dependent Ca2+ channel requiredfor the survival ofprostate cancer cells. Cancer Res, 2004. 64(22): p. 8365-73. Nilius, B., et al., Gating of TRP channels: a voltage connection? J Physiol, 2005. 567(Pt 1): p. 35-44. Nilius, B., et al., The Ca2+-activated cation channel TRPM4 is regulated by phosphatidylinositol4,5-biphosphate.Embo J, 2006.25(3): p. 467-78. Launay, P., et al., TRPM4 is a Ca2+-activated nonselective cation channel mediating cell membrane depolarization. Cell, 2002.109(3): p. 397-407. Xu, X.Z ., et at., Regulation of melastatin, a TRP-relatedprotein, through interaction with a cytopZasmic isoform. Proc Natl Acad Sci U S A, 200 1.98(19): p. 10692-7. Nilius, B., et al., The selectivityfiller of the cation channel TRPM4. J Biol Chem, 2005.280(24): p. 22899-906. Nilius, B., et al., Regulation of the Ca2+ sensitivity of the nonselective cation channel TWM4. J Biol Chem, 2005.280(8): p. 6423-33. Nilius, B., et al., Decavanadate modulates gating of TRPM4 cation channels. J Physiol, 2004.560(Pt 3): p. 753-65. Zhang, Z., et al., Phosphatidylinositol4,5-bisphosphaterescues TRPM4 channels from desensitization. J Biol Chem, 2005.280(47): p. 391 85-92. Nilius, B., et al., Voltage dependence of the Ca2+-activated cation channel TRPM4. J Biol Chem, 2003.278(33): p. 308 13-20. Hofinann, T., et al., TRPM5 is a voltage-modulated and Ca(2+)-activated monovalent selective cation channel. Curr Biol, 2003.13(13): p. 1 153-8. Nilius, B., et al., Intracellular nucleotides andpolyamines inhibit the Ca2+activated cation channel TRPM4b. Pflugers Arch, 2004.448(1): p. 70-5. Ullrich, N.D., et al., Comparison offinctional properties of the Ca2+-activated cation channels TWM4 and TRPMSfrom mice. Cell Calcium, 2005.37(3): p. 267-78. Liu, D. and E.R. Liman, Intracellular Ca2+ and the phospholipid PIP2 regulate the taste transduction ion channel TRPM5. Proc Natl Acad Sci U S A, 2003. lOO(25): p. 15 160-5. Prawitt, D., et al., TRPM5 is a transient Ca2+-activated cation channel responding to rapid changes in [Ca2+]i. Proc Natl Acad Sci U S A, 2003. lOO(25): p. 15 166-71. Perez, C.A., et al., A transient receptor potential channel expressed in taste receptor cells. Nat Neurosci, 2002.5(11): p. 1 169-76. Zhang, Y., et al., Coding of sweet, bitter, and umami tastes: dgerent receptor cells sharing similar signalingpathways. Cell, 2003. 112(3): p. 293-301. Damak, S., et al., Trpm5 null mice respond to bitter, sweet, and umami compounds. Chem Senses, 2006.31(3): p. 253-64. Perez, C.A., et al., Making sense with TRP channels: store-operated calcium entry and the ion channel Trpm5 in taste receptor cells. Cell Calcium, 2003.33(56): p. 541-9. Liu, D., Z. Zhang, and E.R. Liman, Extracellular acid block and acid-enhanced inactivation of the Ca2+-activated cation channel TRPMS involve residues in the S3-S4 and SJ-St5 extracellular domains. J Biol Chem, 2005.280(21): p. 20691 -9. Jaquemar, D., T. Schenker, and B. Trueb, An ankyrin-like protein with transmembrane domains is specifcally lost afer oncogenic transformation of humanfibroblasts. J Biol Chem, 1999.274(11):p. 7325-33. Stokes, A., et al., TRPAl is a substrate for de-ubiquitination by the tumor suppressor CYLD. Cell Signal, 2006. Schenker, T. and B. Trueb, Down-regulated proteins of mesenchymal tumor cells. Exp Cell Res, 1998.239(1):p. 161-8. Voets, T., et al., The principle of temperature-dependentgating in cold- and heatsensitive TRP channels. Nature, 2004.430(7001):p. 748-54. Liman, E.R., Thermal gating of TRP ion channels: food for thought? Sci STKEY 2006.2006(326):p. pe12. Brauchi, S., P. Orio, and R. Latorre, Clues to understanding cold sensation: thermodynamics and electrophysiologicalanalysis of the cold receptor TRPM8. Proc Natl Acad Sci U S A, 2004. lOl(43): p. 15494-9. Liapi, A. and J.N. Wood, Extensive co-localization and heteromultimer formation of the vanilloid receptor-like protein TRPV2 and the capsaicin receptor TRPVl in the adult rat cerebral cortex. Eur J Neurosci, 2005.22(4): p. 825-34. Hellwig, N., et al., Homo- and heteromeric assembly of TRPV channel subunits. J Cell Sci, 2005.118(Pt 5): p. 9 17-28. Rutter, A.R., et al., Heteromerization and colocalization of TrpVl and TrpV2 in mammalian cell lines and rat dorsal root ganglia. Neuroreport, 2005.16(16): p. 1735-9. Wang, C., et al., An alternative splicingproduct of the murine trpvl gene dominant negatively modulates the activity of TRPVl channels. J Biol Chem, 2004.279(36): p. 37423-30. Lu, G., et al., TRPVlb, a@nctional human vanilloid receptor splice variant. Mol Pharmacol, 2005.67(4): p. 1 1 19-27. Tian, W., et al., Regulation of TRPVl by a novel renal& expressed rat TRPVl splice variant. Am J Physiol Renal Physiol, 2006.290(1): p. F117-26. Schumacher, M.A., et al., Molecular cloning of an N-terminal splice variant of the capsaicin receptor. Loss of N-terminal domain suggestsfunctional divergence among capsaicin receptor subtypes. J Biol Chem, 2000.275(4): p. 2756-62. Amiges, Meyet al., Human TRPV4 channel splice variants revealed a key role of ankyrin domains in multimerization and traflcking. J Biol Chem, 2006.281(3): p. 1 580-6. Chapter 2. The Drosophila ortholog of vertebrate TRPAl regulates thermotaxis. Rosenzweig, M., Brennan, K.M., Tayler, T.D., Phelps, P.O., Patapoutian, A., and P.A. Garrity. As published in Genes & Development (2005) Feb 15; 19(4):419-24. Epub 2005 Jan 28. Karen Breman generated the dTRPAl antibody, optimized the dTRPA 1 antibody conditions and provided technical assistance with the dsRNA injections. Tim Tayler helped to optimize the dTRPAl antibody conditions, acquired the images in Figure 4, AF, and provided assistance with the dsRNA injections during the early phases of the project. Paul Phelps provided technical assistance with the dsRNA injections and generated the antibody stainings and the Western blot for the Supplementary Figure 1. Paul Garrity performed the high-temperature nociception experiments in Figure 3B and C and Figure 4H and directed the Supplementary movies. ABSTRACT. Thennotaxis is important for animal survival, but the molecular identities of temperature sensors controlling this behavior have not been determined. We demonstrate dTRPA1, a heat-activated Transient Receptor Potential (TRP) family ion channel, is essential for thermotaxis in Drosophila. dTrpAI knockdown eliminates avoidance of elevated temperatures along a thermal gradient. We observe dTRPAl expression in cells without previously ascribed roles in thermosensation and implicate dTRPAl -expressing neurons in mediating thermotaxis. Our data suggest that thennotaxis relies upon neurons and molecules distinct from those required for high-temperature nociception. We propose dTRPAl may control thermotaxis by sensing environmental temperature. INTRODUCTION Animals exhibit strong behavioral responses to temperature, and many are able to thermotax, undergoing directed migration guided by differences in temperature. Central to thennotaxis are the abilities to sense environmental temperature and to execute the appropriate behavioral response. Animal thennotaxis has been studied most extensively in the nematode Caenorhabditis elegans, where ablation studies have defined the neuronal circuitry involved in thermotactic behavior and molecular genetic studies have identified several molecules required for the development and function of the thermosensory system [I]. However, the molecular identity of the thermal sensors themselves has remained unknown. Larvae and adults of the f f i t fly Drosophila melanogaster also exhibit strong thermotactic behaviors ([2-41). No regulators of Drosophila thennotaxis have been identified at the molecular level, and little is known of the neural circuitry that controls thermotaxis, aside fiom a small group of terminal organ neurons involved in larval cold avoidance [4]. Several classes of molecules have been implicated in potentially mediating temperature sensation and could be involved in thermotaxis. The two-pore-domain K+channel TREK-1 [5] and members of the DEGIEnaC-family of Na+ channels [6] are regulated by temperature in cultured cells, while mice lacking the ATP-gated cation channel P2X3 are defective for electrophysiological responses to moderate warmth (32OC-45"C) [7]. In addition, several members of the Transient Receptor Potential (TRP) family of ion channels have been shown to act as temperature-responsiveion channels in heterologous cells [8,9], and mice lacking one of these proteins, the heat-activated TRPV I, have been shown to be defective in a withdrawal response to noxious high temperature as well as thermal hyperalgesia upon inflammation [lo, 111. While the mouse, C. elegans, and Drosophila melanogaster genomes all encode two-pore-domain K+ channels, DEGlEnaC proteins, and TRP proteins, it has not been established whether any of these molecules play important roles in thermotaxis. The temperature-responsiveTRPs (TRPV1-V4, TRPM8, and TRPA1) have been dubbed thermoTRPs and include members of three distinct families of TRP channels: TRPV, TRPM, and TRPA [8,9]. The Drosophila genome encodes two TRPV family members, one TRPM, and four TRPAs. Of these proteins, f'unctions have been described for the TRPVs Inactive and Nanchung, which act together in hearing [12,13], and the TRPA Painless, which mediates larval nociceptive responses to high-temperature mechanical stimulation [14]. One Drosophila TRP protein has been shown to function as a temperature-responsiveion channel in heterologous cells (i.e., is a thermoTRP), dTRPA 1 (formerly dANKTM1) [I 51. dTRPA1 is the Drosophila ortholog of the single mammalian TRPA protein TRPA 1, and dTRPA1 opens in response to warming [151. However, the in vivo function of dTRPAl (and of its mammalian ortholog) in thermosensory behavior has not been explored. Here we develop a novel RNAi-based strategy for studying thermotactic behavior and use this approach to demonstrate that the warmth-activated ion channel dTRPAl is essential for thermotaxis. We proceed to identify a novel group of dTRPAl-expressing neurons in the CNS that appear important for thermotactic behavior, and find that the proteins and neurons essential for thermotaxis differ from those previously implicated in high-temperature nociceptive behavior. This work identifies a candidate environmental temperature sensor for themotaxis and provides a cellular and molecular starting point for the dissection of thermoTRP signaling and thermotaxis in Drosophila. In addition, the RNAi-based strategy described here should be applicable to the study of other behaviors in Drosophila. RESULTS AND DISCUSSION Drosophila larvae exhibit robust thermotactic behavior We examined Drosophila thermotactic behavior by using a thermal preference assay, placing larvae on a gradient of temperatures warmer than their optimal growth temperature (24OC) [16] and allowing the larvae to migrate from the release zone of 3 1OC-35OC into a region of even higher temperature or a region of lower temperature (Fig. 1A). Wild-type late-first/early-second instar larvae rapidly migrated down the thermal gradient into the cooler zone (Fig. 1C; Supplementary Movie S1). Some larvae explored the warmer zone but rapidly reoriented and headed down the gradient. Larval thermotactic behavior in this thermal preference assay was quantified with an avoidance index (AI) (Fig. 1B; [4]). Wild-type larvae achieved A1 scores >0.9 within 2 min (Fig. 1B), demonstrating strong heat avoidance. dTrpA1 is requiredfor thermotaxis We devised a simple RNA interference (RNAi) strategy to survey whether any Drosophila TRPA, TRPV, or TRPM family members might contribute to larval thermotactic behavior. In this approach, embryos were injected with double-stranded RNAs (dsRNAs) corresponding to the genes of interest, and the resulting larvae were analyzed for their ability to thermotax. The injected animals contained a neuronally expressed green fluorescent protein (GFP) transgene, and dsRNA targeting GFP expression was included in all injections, serving as an internal control for a successful injection (Fig. 2A). Injection of a mixture of dsRNAs corresponding to the four TRPA family members strongly altered thermotaxis (A1 scores of c0.3) (Fig. 2B). In contrast, injection of dsRNAs corresponding to the TRPVs Inactive and Nanchung or the TRPM CG30078 had no detectable effect on thermotaxis (Fig. 2B). These latter experiments served as controls demonstrating that the dsRNA injection procedure itself did not affect thermotaxis. Taken together, our data suggest that one or more Drosophila TRPA family members is required for thermotaxis. Strikingly, RNAi of dTrpAl alone strongly disrupted avoidance of elevated temperature. Unlike wild-type larvae, similar numbers of dTrpA1(RNAi) larvae migrated into both warmer and cooler zones, yielding A1 scores near zero (Fig. 2D), and many dTrpA1(RNAi) larvae traveled deep into the warmer area (Fig. 2C; Supplementary Movie Figure 1. Thermal preference assay. (A) Larvae are placed in the middle of the therrnal gradient on an agarose-covered plate, within the release zone. Agar surface temperatures are as indicated. (B) Larval behavior on gradient quantified by an avoidance index (AI) [4] as indicated. Graph depicts behavior of uninjected wild-type (elav-Gal4; UAS- mCD8:GFP) late-first instarlearly-second instar larvae (n = 9 assays). (C) Still images from thermal preference assay of uninjected wild-type larvae. Red dotted lines demarcate the release zone, with heated zone at lefr and unheated zone at right. In all figures, data are mean k SEM. More than 39 larvaelassay. FIGURE 1: A =2 mln S2). Injection of dsRNA against a second, nonoverlapping region of dTrpAl also generated A1 scores near zero (Fig. 2D). dTrpA1 knockdown disrupted thennotactic behavior throughout larval life, as third instar dTrpA1(RNAi) larvae also exhibited thermotaxis defects (Fig. 2E). Injection of dsRNAs against the three other TrpAs had no effect on thermotaxis (Fig. 2D). We further examined the TRPA Painless, which mediates responses to high-temperature mechanical stimulation [14], a thermosensory behavior potentially distinct from thermotaxis. Neither late-first instarlearly-second instar nor third instar painless mutant larvae (pain' andpain3) were defective for thennotaxis in our assay (Fig. 2D,E). Our data demonstrate that dTrpAl is absolutely required for proper behavior in the thermal preference assay, providing strong genetic evidence for dTrpA1 involvement in thermotaxis. Furthermore, these data also suggest that the TRPA proteins Painless and dTRPAl have distinct roles in temperature-regulated behavior. dTrpA1(RNA i) larvae are selectively defective for thermotactic behavior We next tested how specific the behavioral defect of dTrpA1(RNAi) animals was for thennotaxis. First, we determined that the motility indices (MIS) of wild-type and dTrpA1(RNAi) larvae in the thermal gradient were indistinguishable. (The MI represents the fraction of larvae no longer in the release zone at a given time point [see Materials and Methods].) Late-first instarlearly-second instar wild-type larvae had MIS of 0.56 * 0.08 and 0.64 0.08 at 2 and 5 min after release (n = 9 assays, GEM), while * dTrpAI(RNAi) larvae had MIS of 0.60 0.06 and 0.69 0.04 at these time points (n = 11 s (RNAi) animals were specifically defective in thermotaxis rather assays). ~ h u dTrpA1 than in their ability to migrate at elevated temperature. Second, we examined lanal Figure 2. dTrpA1 is required for therrnosensory behavior. (A) RNAi strategy (methods). GFP-negative (asterisk), but not GFP-positive (circles), RNAi animals were retained for analysis. (B) Late-first/early-second instar behavior. Uninjected (n = 9 assays); TrpVs:dsRNA, injected with Gal4, Inactive, and Nanchung dsRNAs (Gal4 + Inactive + Nanchung) (n = 4); TrpM:dsRNA:Gal4 + CG30078 (n = 4); TrpAs:dsRNA:Gal4 + dTrpAl + Painless + CG17142 + CG3 1284 (n = 6). (C) Preference assay (orientation as in Fig. 1C). (D) As in B except dTrpA1(RNAOregion I : Gal4 + dTrpAl (region 1) (n = 7). dTrpA1(RNAOregion 2:Gal4 + dTrpAl(region 2) (n = 4); TrpAs-dTrpAl :dsRNA:Gal4 + Painless + CG 17142 + CG3 1284 (n = 4); painless1 (n = 5); painless3 (n = 4); atonal(RNAO (n = 4). (I?) Third instar behavior. Uninjected (n = 9); dTrpA1(RNAi)region 1 (n = 13);painless1 (n = 7); painles? (n = 4); md-Gal4: UAS-TeTxLC (n = 12). (**) p < 0.000 1 (versus uninjected). Slight md-Gal4: UAS-TeTxLC effect was not statistically significant ( p = 0.09,2 min; p = 0.27,5 min) and may reflect their noted uncoordination [14]. Nineteen to 75 larvaelassay. M min -0s A 2nrln a min chemotaxis in response to an olfactory repellent (n-octyl acetate) [17], and found dTrpA1(RNAi) larvae responded similarly to wild type (Fig. 3A), indicating that dTrpAI(RNAi) larvae were not defective for all avoidance behavior. Finally, we examined responsiveness of dTrpA1(RNAi)larvae to higher temperatures (55°C). Because larvae rapidly died when exposed to gradients of higher temperature (data not shown), larvae were tested by using a previously described temperature-dependent nociceptive assay [14]. Crawling third instar larvae rapidly and dramatically curl when touched with a hot (55OC) probe (Fig. 3B) but not when touched with a 25OC or 36OC probe. dTrpA1 (RNAi) larvae responded indistinguishably from wild-type larvae in this assay (Fig. 3B,C). As a negative control for the assay, we examined larvae expressing tetanus toxin light chain (TeTxLC), an inhibitor of synaptic vesicle release [18], under md-Gal4 [19] control (md-Gal4 is expressed in multiple-dendritic neurons and 100 CNS neurons) [14]. As previously reported [14], md-Gal4:UAS-TeTxLC larvae were defective in response to contact with the hot probe (Fig. 3C). The ability of dTrpAI(RNAi)larvae to chemotax and respond to other thermal stimuli suggests a specific requirement for dTrpA1 in therrnotaxis. dTRPAl expression and the role of dTrpA1-Gal4-expressing cells in thermosensation Antisera raised against dTRPAl detected strong dTRPAl protein expression in a small number of central brain neurons (Fig. 4A,B) and in neuroendocrine cells of the corpus cardiacum (Fig. 4A). dTrpA1(RNAi) larvae had no detectable dTRPAl expression in these regions, demonstrating that our antisera was specific for dTRPAl and that the dTrpA1 dsRNA effectively reduced dTRPAl protein expression (Fig. 4C). Specific Figure 3. dTrpA1(RNAi) animals respond to other stimuli. (A) n-Octyl acetate avoidance of late-first/early-second instar elm-Gal#; UAS-mCD8:GFPlarvae, uninjected, or dTrpA1 (RNAi) (region 1 ) (n = 6 assays for each). Odor avoidance = [(number of larvae in no odorant zone) - (number of larvae in odorant zone)]l[(number of larvae in no odorant zone) + (number of larvae in odorant zone)]. (B) Response to contact with 55OC probe of third instar uninjected and dTrpA1(RNAi) larvae. (C) Time between probe contact and initiation of curling. Uninjected (n = 4 1 larvae); dTrpAI(RNAi)region 1 (n = 24); md- Gal4 (n = 23); UAS-TeTxLC (n = 30); md-Gal#: UAS-TeTxLC (n = 23). Most larvae not responding within 3 sec never exhibited detectable curling. FIGURE 3: N-octyl acetate avoidance 55'C probe uninjected dTrpAf(RNA uninjected dTpA f(RNAi) t=O sec c t=0.4 sec Response to contact with 55'C probe <I sec u 1- 2 sec 2-3sec ~ s e c dTRPAl expression was also detected in two pairs of cells adjacent to the mouthhooks and in the developing gut (data not shown). Interestingly, we did not detect dTRPAl expression in either multiple-dendritic neurons (implicated in temperaturedependent nociceptive responses) [14] or chordotonal neurons, both of which show temperature-dependent calcium changes [4]. We explored the role of multiple-dendritic neurons using md-Gal4: UAS-TeTxLC larvae. Consistent with no significant role of multiple-dendritic neurons in our thermotaxis assay, significant thermotaxis defects were not detected in md-Gal4: UAS-TeTxLC larvae (Fig. 2E). In addition, significant thermotaxis defects were not detected in atonal(RNAi) larvae, which lack chordotonal neurons (Fig. 2D; [20]; see Materials and Methods). These data suggest that there may be differences in the neuronal circuitry required for thermotaxis in response to elevated temperature and withdrawal fiom a high-temperature nociceptive stimulus. The role of dTRPAl -expressing cells in thermotaxis was further examined by expressing TeTxLC or the cell death-promoting gene Hid under control of putative dTrpA I promoter sequences. dTrpA I -Gal4 drove GFP expression in most dTRPA 1expressing central brain neurons (Fig. 4D-F), but not in the other dTRPAl-expressing cell populations (data not shown). dTrpA1-Gal4: UAS-Hid animals had reduced numbers of dTRPAl -expressing brain neurons (e.g., only one of the three dTPRAl -expressing cells in Fig. 4D-F was present), while the corpus caridiaca appeared unaffected (see Materials and Methods). Consistent with the participation of dTrpAI-Gal4-expressing cells in thermosensory behavior, dTrpA1-Gal4: UAS-Hid as well as dTrpAI-Gal4: UASTeTxLC third instar larvae were partially, but significantly (p < 0.005), compromised in thermotactic behavior (Fig. 4G). Figure 4. dTRPAl expression. (A) dTRPAl expression in anterior (arrowhead) and posterior (bracket) groups of neurons within the brain and in neuroendocrine cells of corpus cardiacum (asterisk). (B) Posterior group of dTRPAl-expressing neurons. (C) dTrpA1(RNAi) animals lack dTRPA1 staining. Asterisk denotes corpus cardiacum. (D-F) dTrpA-Gal4: UAS-mCD8:GFP (green) expression. dTrpA1-Gal4 drives GFP expression in dTRPA1-expressing (purple) neurons and 100-1 50 additional cells within the brain. (IT)GFP. (0Anti-dTRPAl. (G) Thermal preference behavior. Wild type (n = 9 assays); UAS-TeTxLC (n = 8); dTrpAl-Gal4(n = 9); dTrpAl-Gal4: UAS-InactiveTeTxLC (n = 6); UAS-Hid (n = 6); dTrpA1-Gal4: UAS-TeTxLC (n = 10); dTrpA1-Gal4: UAS-Hid (n = 10). (*)p < 0.005. (H) Response to 55OC probe contact. dTrpA1-Gal4: UAS-TeTxLC (n = 19); dTrpA1-Gal4: UAS-Hid (n = 20). " Thermal Prhrence Behavior (third instar) 1 I. ~ V. P . c d l l-GaW: UA n~ctiveTeTxLC UT? 8 0.4 T< = 'r UASJIM 1-08M: UA TeTxlC dr 14M: UAGid 0.2 0 Time (min) Response to W'C probe Interestingly, both dTrpA1-Gal4: UAS-Te TxLC and dTrpA1-Gal4: UAS-Hid larvae exhibited normal withdrawal from a high-temperature nociceptive stimulus (Fig. 4H), indicating that these animals were not defective for all thermosensory responses. Similar effects were also obtained using a second, independent dTrpAl -Gal4 insertion (data not shown). These data further support the notion that there may be differences in the neuronal circuitry required for thermotaxis and high-temperature nociception. dTRPA1 is an essential regulator of thermotaxis The ability to move toward favorable environmental temperatures and away from unfavorable environmental temperatures is critical for animal survival. While thermotactic behavior has been observed in many animals, the underlying thermal sensors have not been identified. Here we have used the heat-avoidance response of Drosophila larvae to identify dTrpA1 as a key mediator of thermotaxis. The failure of animals lacking dTRPAl expression to avoid elevated temperatures mirrors the ability of dTRPAl to function as a heat-activated ion channel in vitro. Taken together, these data make dTRPAl an attractive candidate for an environmental temperature sensor controlling thermotaxis. It will be of interest to explore how the temperature-sensing properties of dTRPAl observed in cultured cells relate to the physiological and behavioral responses of the intact animal. As closely related dTRPAl orthologs are present in the Drosophila pseudoobscura, Bombyx morii, and Anopheles gambiae genomes, as well as the C. elegans genome, TRPAl s could regulate thermotaxis in other insects as well as in C. elegans. Thermotaxis and high-temperature nociception involve distinct mechanisms Another Drosophila TRPA protein, Painless, also modulates thermosensory behavior [14]. However, dTRPAl and Painless appear to have distinct thermosensory hctions. While painless mutant larvae are strongly defective in responding to noxious high-temperature mechanical stimulation [14], painless mutants are normal in our thermotaxis assays. Further emphasizing the apparent differences between these different thermosensory behaviors, inhibition of the md-Gal4-expressing neurons disrupts responses to noxious high-temperature mechanical stimulation but has no significant effect on thermotaxis, while inhibition of dTrpAl-Gal4 cells has the opposite effect. Thus the migration of larvae away fkom moderately elevated temperatures (thermotaxis) and the high-temperature nociceptive response appear to rely upon distinct TRPA-family members and potentially distinct neural circuits. Determining the neuronal circuitry required for dTrpAl-dependent thermotaxis is an important goal for the future. Our dTrpAl-Gal4-mediated inhibition and ablation studies are consistent with a role for dTRPAl -expressing CNS neurons in dTrpAI dependent thermotaxis. CNS thermal sensors have been proposed within the head and thoracic ganglia of adult bees and cockroaches [21-231 and are found in the vertebrate preoptic-anterior hypothalamus [24]. As small insects (Drosophila larvae are <5 mg) have limited heat capacity [22,25], internal sensors of elevated temperature could be quite effective. Most importantly, identification of dTRPAl -expressing cells provides an entry point into defining the neuronal circuitry controlling thermotaxis in flies. It will be of interest to determine how dTRPAl-expressing cells participate in the sensation of environmental temperature, either as primary thermosensory cells or as higher-order neurons involved in processing thermosensory input. dTRPAl could have functions in addition to therrnotaxis. The strong expression of dTRPAl in the corpus cardiacum is intriguing as this neuroendocrine gland is involved in temperature-dependent developmental phenomena in some insects, including seasonal polyphenism in butterfly wings, where corpus cardiacum ablation prevents establishment of the summer wing pattern [26]. Furthermore, vertebrate TRPAl has been proposed as a candidate mechanotransduction channel for hearing [27]. Whether dTRPAl responds to mechanical stimulation is unknown, but future analysis of dTRPAl function in flies could potentially yield insights into mechanosensation as well as thermosensation. MATERIALS AND METHODS Molecular genetics and RNAi Fly strains were obtained from Bloomington, except for painless1 and painless3 from the EP collection (Exelixis) [28], md-Gal4 [19] from Y. Jan (University of California, San Francisco, San Francisco, CA), UAS-TeTxLC (UAS-TNT-C)[18] from Troy Littleton (Massachusetts Institute of Technology, Cambridge, MA), and UASHid(#14) fiom Hermann Steller (Rockefeller University, New York, NY). RNAi was performed as described 1291, except the concentration of each dsRNA in injection mixtures was fkom 1.3 m g l d for mixtures of five dsRNAs up to 5 mg/mL for mixtures of two dsRNAs. Injections were into elav-Gal4;UAS-mCD8:GFPanimals, and Gal4 dsRNA was included in each injection mix. Thirteen to 24 h after injection, animals with unaltered (Fig. 2A, open circle) or partially diminished GFP expression (Fig. 2A, closed circle) were eliminated (they comprised 2%-3% and QO% of total injected animals, respectively). Animals lacking GFP expression were grown on molasses plates at 25OC. On day 3 (injected day O), late-first/early-second instar larvae (approximately the size of 48-h-old uninjected elav-Gal#;UAS-mCD8:GFPlarvae) were used in analyses. Following the assay, larvae were allowed to recover on molasses plates overnight at 25OC; larvae significantly larger than 48-h elav-Gal4;UAS-mCD8:GFP larvae were removed and the rest assayed once more. Larvae were then allowed to develop to third instar for further analysis. dTrpAl dsRNAs corresponded to bases 1241-1787 (region 1) and 2872-3605 (region 2) of dTrpA1 cDNA (AY302598) [15]. Details of all dsRNAs are provided in the Supplemental Material. Chordotonal neuron loss in atonal(RNAi) larvae was confirmed by examining elav-Gal4;UAS-mCD8:GFP in animals injected with only Atonal dsRNA. dTrpAl-Gal4 was created by cloning 33-1301 bp upstream of the dTRPA1 translational start into pGATB [301. dTrpA1-Gal#: UAS-mCD8:GFP expression in both lines used did not overlap with dTRPAl protein expression in the larval anterior, gut, or corpus cardiaca. Both dTrpA1-Gal4 lines drove expression in small but nonoverlapping populations of peripheral neurons. Antibody production and histology Rat polyclonal dTRPAl antiserum was raised (Covance) against the 178 Cterminal amino acids of dTRPAl fused to 6XHis. dTRPAl staining of neurons in larval brain and corpus cardiacum was initially detected at first instar and persisted until third instar. Expression of a dTrpAl cDNA in the retina resulted in robust ectopic staining and permitted detection of dTRPAl protein by Western blot (Supplementary Fig. 1). Images were obtained using a Nikon PCM2000 confocal microscope and Zeiss Axioplan2 with Nikon DXM1200 camera. Behavioral assays Thermal gradient assay was similar in design to that of [4], but with different temperatures. Thermal gradient assay was performed on 2% agarose (25 mL) in a 150 x 15 mm Petri dish (VWR 25384-139) placed so that the release zone was at the interface of two metal slide warmers (one heated to 47OC, one unheated) 4 mm apart. Agarose surface was allowed to equilibrate for 50 min prior to each assay (with dish lid in place). Surface temperatures were determined using a Fluke 5211 thermometer with dual K-type flat surface probes. Larvae were washed in distilled water and lined up in the release zone with a paint brush. Plate was immediately covered with a lid (t = 0). Larvae were raised on molasses plates or in bottles and vials with food. Larvae expressing TeTxLC or Hid were examined at third instar to permit maximum TeTxLC or Hid accumulation. Hid ablation of dTRPAl cells was quantified using the three large dTRPAl -expressing CNS neurons bracketed in Figure 4. Three cells were observed in all 10 dTrpA1-Gal#:UASGFP hemispheres and all eight UAS-Hid hemispheres examined, while one cell was observed in all 20 dTrpA1-Gal#:UAS-Hidhemispheres examined. The MI represented the fraction of larvae no longer in the release zone at a given time point. MI = (number of larvae in cooler zone + number of larvae in warmer zone)/(total number of larvae). All statistical comparisons used two-tailed unpaired t-test versus uninjected. Animals expressing active toxins exhibited modest reductions in MI. MI of third instar dTrpAl-Gal#;UAS-TeTxLC larvae were 0.57 % 0.04 and 0.69 A 0.03 at 2 and 5 min (n = 10 assays) and those of dTrpAI -Gal#:UAS-Hid larvae were 0.70 A 0.05 and 0.78 * 0.04 (n = 1O), while those of wild-type larvae were 0.86 0.03 and 0.89 0.02 (n = 10). However, the value of the MI did not correlate with thermotaxis defects, suggesting the reduced MI is not the source of the thermotaxis defect. dTrpAl -Gal#:UAS-Hid animals had a higher MI than did dTrpAl-Gal#:UAS-TeTxLC animals, but were more defective for thermotaxis. Furthermore, the dTrpA1-Gal#:UAS-Hid MI was not significantly * different from that of md-Gal4:UAS-TeTxLC (0.75 0.05 and 0.85 0.03; n = 12), but thermotaxis in dTrpAl -Gal#:UAS-Hid was substantially reduced compared with that in md-Gal#:UAS-TeTxLC ( p < 0.002). In addition, third instar dTrpAl-Gal4 and dTrpAIGal4:UAS-Hid larvae performed indistinguishably for chemotaxis, with n-octyl acetate AIs of 0.67 * 0.07 and 0.66 * 0.09, respectively (n = 6 assays for each, 60-min time point), suggesting a specific defect in thermotaxis. Probe assay was performed using a Sears Craftsman Soldering Iron (1 13.540410) controlled by a Statco Variac (3PlV221B). Probe temperature was monitored with Physitemp BAT- 12 thermometer equipped with K-type flat surface probe and varied between 55°C and 57OC during assays. The Supplementary Movies were made with a Sony TRV38 and edited with Apple iMovie. Odor avoidance assays were performed as described [17], except 2.5 pL of undiluted n-octyl acetate (Sigma) was presented in microcentrifbge tube cap opposite an empty cap, and late-first/early-second instar larvae were assayed for uninjected and dTrpA1(RNA0. Acknowledgments We thank Y. Jan and T. Littleton for fly strains; T. Jegla, G. Story, and V. Viswanath for helpful discussions and communicating results prior to publication; L. Guarente, H.R. Horvitz, R. Huey, R. Hynes, T. Orr-Weaver, C. Quinn, W. Rhyu, R. Stevenson, D. Tracey, B. Wold, and Garrity lab members for helpful discussions; S. Gilbert, F. Nijhout, and R. Stevenson for reprints; and A. Chess, L. Huang, R. Huey, R. Khodosh, J. Whited, and B. Wold for comments on the manuscript. This work was supported by a grant to P.A.G. fiom the Raymond and Beverly Sackler Foundation and a Whitehead Family Career Development Professorship. M.R. was supported by an NSF Graduate Research Fellowship and an NIH Predoctoral Training Grant, and T.D.T. was supported by an NIH Predoctoral Training Grant. FOOTNOTES Supplemental material is available at http://www.genesdev.org. Article published online ahead of print. Article and publication date are at REFERENCES I. Mori, Annu Rev Genet 33,399-422 (1999). 0. Sayeed, S. Benzer, Proc Nail Acad Sci U S A 93,6079-84 (Jun 11,1996). T. Zars, J Comp Physiol [A] 187,235-42 (Apr, 2001). L. Liu, 0. Yermolaieva, W. A. Johnson, F. M. Abboud, M. J. Welsh, Nut Neurosci 6,267-73 (Mar, 2003). F. Maingret et al., Embo J 19,2483-91 (Jun 1,2000). C. C. Askwith, C. J. Benson, M. J. Welsh, P. M. Snyder, Proc Natl Acad Sci U S A 98,6459-63 (May 22,2001). V. Souslova et al., Nature 407, 1015-7 (Oct 26,2000). S. E. Jordt, D. D. McKemy, D. Julius, Curr Opin Neurobiol13,487-92 (Aug, 2003). A. Patapoutian, A. M. Peier, G. M. Story, V. Viswanath, Nat Rev Neurosci 4, 529-39 (Jul, 2003). M. J. Caterina et al., Science 288,306-13 (Apr 14,2000). J. B. Davis et al., Nature 405, 183-7 (May 11,2000). J. Kim et al., Nature 424,8 1-4 (Jul3,2003). Z. Gong et al., J Neurosci 24,9059-66 (Oct 13,2004). W. D. Tracey, Jr., R. 1. Wilson, G. Laurent, S. Benzer, CeN 113,261-73 (Apr 18, 2003). V. Viswanath et al., Nature 423,822-3 (Jun 19,2003). W. H. a. B. Siddiqui, C.A., Ann. Entomol. Soc. Am. 65,993-1 001 (1972). G. Heimbeck, V. Bugnon, N. Gendre, C. Haberlin, R. F. Stocker, JNeurosci 19, 6599-609 ( A u ~1,1999). S. T. Sweeney, K. Broadie, J. Keane, H. Niemann, C. J. OtKane,Neuron 14,34151 (Feb, 1995). F. B. Gao, J. E. Brenman, L. Y. Jan, Y. N. Jan, Genes Dev 13,2549-61 (Oct 1, 1999). M. L. Huang, C. H. Hsu, C. T. Chien, Neuron 25,57-67 (Jan, 2000). B. Heinrich, . I Exp. Biol. 85,6 1-72 (1980). B. Heinrich, m e hot-blooded insects: Strategies and mechanisms of thermoregulation. (Harvard University Press, Boston, MA., 1993). B. F. a. H. Murphy Jr., J.E., J: Exp. Biol. 105,305-315 (1983). J. A. Boulant, Clin Infect Dis 31 Suppl5, S157-61 (Oct, 2000). R. D. Stevenson, Am. Nat. 125, 102- 117 (1985). S. a. E. Fukuda, K., Proc. Japan. Acad 42,1082-1087 (1966). D. P. Corey et al., Nature 432,723-30 (Dec 9,2004). P. Rorth et al., Development 125, 1049-57 (Mar, 1998). H. C. Sears, C. J. Kennedy, P. A. Garrity, Development 130,3557-65 (Aug, 2003). A. H. Brand, N. Perrimon, Development 118,401-15 (Jun, 1993). N. H. Patel, in Methods in Cell Biology E. A. Fyrberg, Ed. (Academic Press, San Diego, 1994), vol. 44, pp. 445-487. SUPPLEMENTARY INFORMATION: Supplementary Movie 1: This movie shows the behavior of wild-type late firstlearly second instar larvae on the thermal gradient. This time lapse begins with the lining up of wild-type larvae in the release zone, followed by the assay. The portion of the movie containing the assay is speeded up 8X and covers from 30 seconds to 6 minutes of the assay. (QuickTime, 4 MB). Supplementary Movie 2: This movie shows the behavior of dTrpA1 (MA0 late fust/early second instar larvae on the thermal gradient. The video is speeded up 8X and covers from 30 seconds to 6 minutes of the assay. (QuickTime, 2.5 MB). Supplementary Methods Histology Anti-dTrpAl was used at 1500-1:1000, mouse Mab 2 2 ~ 1 0(DSHB) at 150. Cy-3conjugated goat-anti-rat secondary antibody (Jackson) was used at 1500 and FITCconjugated goat-anti-mouse secondary antibody (Jackson) at 1500. Sonicated first instar larvae and dissected third instar nervous systems were prepared as described [3 11. RNAi Sequences of primers used for generation of templates for dsRNA synthesis were as follows. T7 polymerase binding site sequence is underlined. Inactive (CG4536) Supplementary Figure 1: Additional controls for specificity of anti-dTRPAl antisera. a, Anti-dTRPAl staining in wild type animal. dTRPAlexpression is observed in the brain (bracket), but not in the retina (open mowhead). b, Ectopic dTRPAl expression in the retina is detected by anti-dTRPAl antisera (arrowhead). These animals express a dTrpAl cDNA under the control of the retina-specific GMR promoter. Endogenous dTRPAl expression in the brain (bracket) is barely visible as this image is underexposed to permit clear visualization of retinal staining. c) Western blots containing protein from wild type and GMR-Gal4; UAS-dTRPA1 larvae probed with anti-dTRPAl. Anti-dTRPAl detects overexpressed dTRPA1 in GMR-Gal4; UAS-dTRPA1,recognizing a protein species whose mobility is consistent with the predicted dTRPAl molecular weight of 135 kD. Blots were reprobed with anti-ELAV as a loading control. 5'- TAATACGACTCACTATAGGGAGACAGGTGATAAACCGTGCTGAG -3' 5'- TAATACGACTCACTATAGGGAGATCCCGTCTGGTTGACAATGC -3' Nanchung (CG5842) 5'- TAATACGACTCACTATAGGGAGACTTCGACAACCCATCGTCAC-3' 5'- TAATACGACTCACTATAGGGAGATTCTGCCATCCGACATCGAC-3' dTrpAl (CG5751) region 1 5'- TAATACGACTCACTATAGGGAGACGTGCATTAGTGTCAAGGACG -3' 5'-TAATACGACTCACTATAGGGAGATCCACCTGATCGAGCAGATG-3' dTrpAl (CG5751) region 2 5'-TAATACGACTCACTATAGGGAGATCGTGGTTCCGATTGCTACTG-3' 5'-TAATACGACTCACTATAGGGAGAAACTTTCGCCTCTGCCGTTC-3' Painless (CG15860) 5'- TAATACGACTCACTATAGGGAGAGATGAAAATGCCCCAGACGTG-3' 5'- TAATACGACTCACTATAGGGAGAATCAGCTTCGTTGATGTCTACGC-3' CG31284 5'- TAATACGACTCACTATAGGGAGAGAGGGTCAGAAGGAGATACTG-3' 5'-TAATACGACTCACTATAGGGAGACCAATCATCATCACCAGCACTG-3' CG17142 5'-TAATACGACTCACTATAGGGAGAACATATCCAGTGGCGGATCAG-3' 5'- TAATACGACTCACTATAGGGAGAGCACGTTCTCCATCTTAACGC -3' Atonal 5'-T7-CAGCAGCAGCAACTACTTGTCG-3' 5'-T7-TTGGAGGGTCTCGTGTTTGGAC-3 ' All PCR were performed on fly genomic DNA (CI55-Gal4; UAS-mCD8:GFPor WCS) except for dTrpM (GH22805)and dTrpAl region2 (AY302598) Chapter 3. TRPAl mediates both heat- and cold-avoidance behaviors in Drosophila. Generation of d t r p ~ pd, t r p l~Aeqd t r p ~Ae8, l and of dtrpA1Cre-Out(nO~) 1ines by allelic reduction of corresponding d t r p ~Ins l lines, as well as some PCR analysis of these lines, was accomplished by Caitlin Rex, a technical assistant in the Gamty lab. ABSTRACT. Temperature-gated members of the Transient Receptor Potential (TRP) ion channel family, dubbed thermoTRPs, help mediate temperature sensation in animals. Different thermoTRPs can be activated by specific ranges of temperatures, and loss of individual thermoTRPs leads to defective behavioral responses over narrow ranges of temperatures. No examples of individual thermoTRPs mediating animals' responses to diverse temperature ranges, such as responses to both hot and cold temperatures, are known. Here we report generation of strong loss of function mutations in the Drosophila TRPAI. It has been previously demonstrated that dTRPAl can be activated by elevated temperatures (-27OC) in heterologous cells and that dTRPAl is required for larval thermotaxis in response to elevated temperatures. The new dtrpA1 mutants were defective not only for the avoidance of elevated temperatures but also for the avoidance of cold temperatures, suggesting that dTRPAl can mediate responses to a wide range of temperatures. One interpretation consistent with our findings is that dTRPAl is a bimodal ion channel, capable of being activated by, and mediating animals' responses to, both hot and cold temperatures. INTRODUCTION. Proper responses to temperature are critical for animal survival. To sense temperature, animals appear to employ temperature-responsive ion channels [I]. The sensed temperature information is then processed and if necessary a behavioral adjustment is initiated. Temperature-responsive ion channels include members of the transient receptor potential (TRP) family [2]. These "thermoTRPs" are expressed in tissues known to be temperature responsive, such as sensory neurons innervating the skin and the tongue, and can be activated by specific ranges of temperatures in cultured neurons or in heterologous cells [I]. Distinct thermoTRPs are activated by noxious heat (TRPV 1 and TRPV2) [3,4], by moderately warm/innocuous temperatures (TRPV3, TRPV4, TRPM4 and TPM5) [5-91, by moderate cooling (TRPM8) [lo, 111 and possibly by noxious cold (TRPAI) [12, 131. Similarly, mouse knockouts lacking individual thermoTRPs exhibit defects in responses to relatively narrow ranges of temperatures [9, 14-171. These studies have led to the development of current models in which animals are proposed to sense temperature using a series of thermoTRPs to detect the broad spectrum of temperatures that animals sense, with each thermoTRP mediating responses to a limited portion of the temperature spectrum [I]. In particular, current models suggest that thermoTRPs are activated by, and mediate behavioral responses to, only fairly narrow ranges of temperatures. For an individual thermoTRP such a range could include hot or cold temperatures, but not both. Fruit fly Drosophila melanogaster larvae and adults can sense and avoid hot and cold temperatures, and Drosophila adults exhibit strong preferences for -24OC [18-201. Similar to mammals, temperature sensation in flies is mediated by multiple molecules and multiple neuronal mechanisms. Thus far, three putative Drosophila thermoTRPs have been identified. PYREXIA (PYX) is a thermoTRP activated by temperatures above -35OC-40°C and is required to prevent heat-induced paralysis at high temperatures (40°C) [21]. PYX has thus been proposed to protect flies from high-temperature stress. PAINLESS is required for larval response to high-temperature (-38OC<T<-52OC) mechanical stimulation [22]. While the ability of temperature to gate PAINLESS has not been reported, the abdominal nerves ofpainless mutants fail to exhibit proper increases in activity following temperature elevation to 42OC, suggesting that PAINLESS does regulate temperature-dependent neuronal activity [22]. Finally, dTRPAl has been demonstrated to be activated by warming (-27OC) in heterologous cells [23]. We have previously shown that RNA interference (RNAi)-mediated knockdown of dTRPAl expression disrupts larval thermotaxis (defined as directed migration in response to changes in temperature) at elevated temperatures (-3 1OC--3 5°C) [19]. As in mammals, it appears that Drosophila TRP channels exhibit specific temperature ranges and specific thermosensory behaviors for which they are required. For example, PAINLESS was dispensable in our elevated temperature thermotaxis assay, while dTRPAl appeared dispensable for larval responses to high temperature mechanical stimulation [19]. While several molecules have now been implicated in mediating thermosensory behaviors in Drosophila, molecules that regulate cold sensation and cold-dependent behaviors in Drosophila have not been identified. To facilitate further analysis of the role dTRPA1 plays in mediating Drosophila thermosensory behaviors we generated loss of function dtrpAlmutants. Consistent with our previous RNAi-based analysis [19], dtrpAl mutants were defective for avoidance of elevated temperatures. Surprisingly, dtrpA1 mutants were also defective for the avoidance of cold temperatures. dTRPAl is the first protein implicated in mediating behavioral responses to cold temperatures in insects. Furthermore, dTRPAl is the first thermoTRP involved in responses to both elevated and cold temperatures. Therefore, in contrast to previous models of thermoTRP function, which propose that an individual thermoTRP mediates responses to either heating or cooling, but not to both, dTRPA1 appears to control animals' responses to both increases and decreases in temperature. RESULTS. Generation of dtrpA1 alleles. We have previously shown that RNA interference (RNAi)-mediated knockdown of dTRPAl expression causes thermotaxis defects in the Drosophila larvae [19]. The dtrpA1(RNAi) larvae were defective for avoiding elevated temperatures. To facilitate further study of the role dTRPAl plays in thermosensory behaviors, we wanted to generate mutants in dtrpA1. To generate a lesion in dtrpAl, we used homologous recombination-mediated ends-in insertional strategy [24] to target the dtrpAl locus (Fig. 1A; See Materials and Methods). From this targeting screen we recovered 33 putative targeting events (at least 27 independent). PCR analysis (Fig. 1B and data not shown) revealed that 29 events (at least 24 independent) were insertions in the dTRPAl locus ( d t r p ~Insl alleles). The targeting construct was engineered to contain a frameshift mutation early in the coding sequence (predicted Dl 83->A, 27aa, STOP) and truncations at the 5' end (removing the first exon, including the putative endogenous start site, first intron and a portion of exon 2) and the 3' end (removing a portion of the predicted ionpore region, the sixth transmembrane domain and the C-terminal cytoplasmic tail). The left side of d t r p ~ l tandem duplication is then predicted to encode a non-functional protein lacking the final transmembrane domain and the C-terminal cytoplasmic tail. The right side of the duplication lacks the putative endogenous start site and any regulatory Figure 1. Generation of dtrpAl mutant alleles. (A) Strategy for targeting the dtrpAl locus using homologous recombination-mediated insertional targeting [24] and subsequent reduction of the dtrpA1Ins duplication to a single copy (Cre-out generation; [25]). Locations of the engineered fiameshift mutation, 5' and 3' end truncations, and of the lesions in the d t r p ~ l dtrp~l and dtrpAlinSlines are indicated. (B) Examples lines. Primers 2 and 3 anneal to the of the PCR analysis of the putative dtrpAlznS sequences of the white mini-gene, while primers 1 and 4 anneal to the sequences outside of the duplicated region. (C) Examples of the PCR analysis of the putative dtrpAlfSlines. The frameshift mutation was engineered by filling in a ClaI restriction enzyme site, creating an NruI restriction site in the process. PCR products of the chromosomes that bear the engineered frameshift mutation can be cut with NruI but not with ClaI, while the PCR products of the chromosomes lacking the engineered fiameshift mutation can be cut with ClaI but not with NruI. Figure 1. A Donor I FLP ---A dTRPA1 genomic locus m MCM? l a ~exon t -- Right side 3+4 Left side 1+2 -- yL - Nrul elements within the first intron and should not make functional protein because of the engineered frameshift mutation. Two of the 27 independent insertion lines contained not only the engineered frameshift mutation but also additional lesions within the dtrpA1 tandem duplication. d t r p ~iNde6 l contained a 205bp deletion within the right side of the tandem duplication, removing almost the entire exon 6 and 16bp of the intronic sequence 3' of exon 6 (the deletion resulted in the first two nucleotides of exon 6 being directly followed by the intronic sequence). It is unclear whether during splicing the two remaining base pairs of exon 6 would be included into the spliced mRNA or whether this lesion would result in skipping of the remaining two base pairs during splicing. If the remainder of exon 6 were skipped, this lesion would not be expected to shift the reading frame of the resulting contained a 307bp deletion on the left side of the duplication. That protein. dtrpAlimAe8 lesion consisted of a substitution of 320bp 5' of the I-SceI insertion site within the targeting construct with 13 unrelated nucleotides. This resulted in an internal deletion within exon 8, where the 5' 28bp of exon 8 were followed by 13 unrelated nucleotides and then by the remainder of exon 8. This lesion is predicted to shift the reading frame of the protein (predicted I4 16->N, 15aa, STOP). We used a Cre-out strategy [25] to reduce the tandem duplication in dlrp~l d t r p ~imAe6, l and dtrpA1 lAe8, and lines to a single copy, generating dhp~lf:dtrp~l 1Cre-outtno-fs)lines. Presence or absence of the fs, Ae6 and Ae8 mutations in the DNA of the Cre-out lines was determined by PCR, and in the case of the engineered fiameshift mutation, also by restriction digest (Fig. 1C and data not shown). The presence of the engineered frameshift mutation in the context of the otherwise wild- type dtrpAl gene in the dtrpAlfilines is predicted to shift the reading frame of the dTRPAl protein and the putative longest open reading fiame (ORF) in dtrpAlfilines is predicted to be out of fixme with the original dTRPAl-coding ORF. The d h p 1~cre-~'t(no-fs) lines generated during the reduction of tandem duplications in dtrpAlimlines to a single copy (the Cre-out strategy) did not contain either the engineered frameshift mutation, or the Ae6 or the Ae8 mutations. However, some of the lines originated from the DNA in the regenerated dtrpAl locus in the dtrp~lCreeo't(no-fs) targeting construct DNA. Thus, while the dtrpA1 locus in the dtrp~lCre4Ut(no~) lines is predicted to be wild-type, if the targeting construct DNA contained mutations in dtrpA.2 other than the engineered frameshift, these mutations could become part of the l lines during the Cre-out process. regenerated dtrpAl locus in the d t r p ~Cre-ou*(no-fs~ dTRPAl protein expression is absent in dhpAl mutants. In wild-type third instar larvae dTRPA1 protein is detected in a small group of neurons in the central brain and in the neurosecretory cells of Corpus Cardiaca (CC) [19]. We previously found that inhibition or ablation of the dTRPAl -expressing neurons in the central brain significantly attenuated larva1 thermotactic behavior [191. dTRPA 1 protein was undetectable in the brains or in the CCs of dtrpAlim3rdinstar larvae (Fig. 2), suggesting that dtrpAlimallele disrupts normal dTRPAl protein expression. dTRPAl is requiredfor larval avoidance of elevated temperatures. We next tested d r r p ~ l " mutant larvae for thermotactic behavior in response to elevated temperatures. When placed onto a thermal gradient within a -3 1°C--350C Figure 2. dTRPAl protein expression is absent in d t r p ~ l i nmutants. s (A)In third larval instar eye-brain complexes dTRPAl antibody detects expression in the Corpus Cardiaca neurosecretory cells (*) and in a subset of central brain neurons (brackets). (B) In the d t r p1~insU3#2-4 mutant animals, the dTRPAl irnmunoreactivity is absent. Figure 2. Figure 3. dtrpAl mutants are defective for themotaxis at elevated temperatures. (A) Thennotaxis assays at elevated temperatures on independent d t r p ~ l i lines m assayed as either homozygous mutant larvae or over a Df(dtrpA1). Behavior of all genotypes exhibit statistically significant differences from wild-type (P<0.0001). Wild type (n=l 1 assays); dWA insU3#2-2 homozygous (n=8); dtrpA1insU3#2-2lDf(dtrpA1) (n=5); d t r p ~insU3#2-4 l homozygous (n=9); dtrpAl insU3#2-4lDf(dtrpA1) (n=8); dtrgA1insUI-9 homozygous (n=5); d t r p ~insUz-9/~f(dtrp~l) l (n=5); d t r p ~insU3#22iz l homozygous (n=7); d t r p ~insU3#2l 4/~f(dtrp~ (n=4). I ) (B) dtrp~lCre-Out mutant animals that contain the engineered fiarneshift, the dtrpAlfiH8and dtrpAlfiG7animals, are significantly defective for thermotaxis at elevated temperatures #(P<0.05); * (P<0.0 1); * * (Pi0.00 1); * * * (P<0.000 1). Wild type (n=6); dtrpAI insU3#2-4 homozygous (n=7); dtrp~lfSHB~f(dtrp~l) (n=4); dtrPAlw homozygous (n=4); dtrPAlfsG7 homozygous (n=4) d t r p I~cre-0Ut(n0-fs)F6 homozygous (n=4) ;dtrpAl Cre-out(no-jkJF6lDfldtrpA1) (n=5). Figure 3. wild type d t r p1~ins#2-2 dtrpAIinm-2/Df(dtrpAI ) dtrpAIinm4 dtrpAIin*%f(dtrpA I ) dtrpAljn=#'" dtpAIins#7-g/Df(dtrpA I) 1indf2-11 d t w Iinmm7 I/Df(dtrpAI ) Time (min) wild type dtrpAIinm4 dtrpA1fSH8/Df(dtrpA I) dtrpAIfSH8 d t p AIbG7 d t r pI~Cm-out(no-fs)F6 dtrpAI C"+OM(no-fs)F6/Df(dtrpA I) Time (min) release zone, dtrpAli* larvae were significantly defective for avoidance of the highertemperature environment (Fig. 3A). Severe thennotaxis defects were observed in both the dtrpAlimhomozygous mutant larvae and in the larvae containing the dtrpAli* lesion in trans over a deficiency uncovering the dtrpAl locus (Df(dtrpA1)) Different d t r p ~'* l lines exhibited varying degrees of thermotactic behavioral defects (Fig. 3A) but the phenotypes of dtrp~l im/~f(dtrp~l) larvae varied less among the different dtrPAlimlines, suggesting that recessive modifiers were partially suppressing the thermotactic defects of )ldtrpAICreeout(no-fs)F6larvae some d t r p I~im lines. Df(dtrpAl)lTM3 and ~f(dtrpA1 exhibited normal thennotaxis in response to elevated temperatures (Fig. 3B and data not shown), suggesting that Df(dtrpA1) is recessive for heat avoidance thennotactic behavior. Furthermore, some of the d t r p ~ l mutant lines exhibited virtually identical avoidance indexes (AI) at elevated temperatures when assayed as homozygotes or as heterozygotes over Df(dhp1). While formally this data can be considered an indication of the d t r p ~iml mutation being a null allele, the effects of genetic background on thermotactic behavior make it difficult to draw that conclusion. The behavior of dtrpAI mutants on l the thermal gradient was also somewhat variable over time, even for the d ~ Amutants in trans with Df(dtrpAl), suggesting that some uncontrolled environmental factors can influence thennotactic behavior of dtrpA1 mutant animals. However, since a number of independent insertions into the dtrpA1 locus caused defects in larval thennotaxis at elevated temperatures, and since these insertions were defective for thennotaxis in trans over an independent dtrpA1 allele (Df(dtrpAI)), these results suggest that it is the mutation of the d ~ p Agene l in dtrp~lim mutant animals that was responsible for the observed thermotaxis defects. For subsequent analysis we focused on a single d t r p ~ l line, dtrpA1ins U3#2-4 since d t v p1~insU3#2-4 larvae exhibited similar thermotactic defects as homozygotes or as heterozygotes over Df(dtrpA1). We used this line to generate allelic substitutions of dtrpA1 (reduction of the d t r p ~ l to a single copy [25];see above and see Materials and l ) lines. The behavior of Methods), producing dtrPAlCre-Out (dtrpAlfSand d t r p ~cre-Out(nO$) dtrpAlfSmutants on the elevated temperature thermal gradient was similar to that of the dtrpAlinSanimals (Fig. 3B). However, animals from a Cre-out line without the engineered frameshift mutation in dTRPAl (dtrpA1Cre-out(n0-f)F6) exhibited wild-type avoidance of elevated temperatures, either as homozygotes or as heterozygotes over Df(dtrpA1) (Fig. 3B). Our preliminary data suggest that the Cre-out lines dtrpAlAe6and dtrpAlAe8are also defective for avoidance of elevated temperatures (data not shown). Taken together, the above results suggest that dTRPAl is required for larval thermotaxis at elevated temperatures. dTRPA1 is requiredfor larval avoidance of cold temperatures. When placed on a thermal gradient generated by cooling a portion of an agarosecovered plate by a cold generator set to 10°C (within -17OC -20°C release zone; Fig. 4A), wild-type larvae robustly avoided the colder half of the plate (cooled zone), rapidly migrating into the warmer (not cooled) zone (Fig. 4B). dtrpAl insU3#2-4 mutant larvae, however, migrated out of the release zone into both the cooled and warmer halves of the plate (Fig. 4B), gradually accumulating within the cooled zone. The observed accumulation of dtrpAl insU3#2-4 larvae on the cooled side with time might reflect the inability of the larvae to leave the cooled area once they migrate into sufficiently cold Figure 4. dtrpAl mutants are defective for thermotaxis at cold temperatures. (A) Diagram of the behavioral apparatus used to test thermotaxis at cold temperatures. The cold generator was set to the indicated temperature (Set T) while the slide warmer remained at room temperature. Points c and d on the diagram correspond to the temperature borders of the release zone. Temperatures for various points on the thermal gradients used in this study are provided in the table above the diagram. For the behavioral assays, larvae were placed within the release zone at t=O and their distribution on the plate was monitored with time. (B) Thermotaxis assays in response to the gradient generated by setting cold generator to 10°C. dtrpAl insU3#2-4 homozygous animals exhibited statistically significant behavioral differences from wild-type (P<0.000 1). Wild type (n=15); dtrpAl insU3#2-4 (n=15). (C) Thermotaxis assays in which the gradient was generated by a 17OC cold plate. dtrpAl insU3#2-4 homozygous animals exhibited no l statistically significant differences from wild-type. Wild type (n=3); d t r p ~insu3#2-4 (n=3). (D) Thermotaxis assays in which the gradient was generated by a 15OC cold plate. l (n=6 ) . Statistical differences are as described in Fig. 3. Wild type (n=4); d t r p ~insu3#2-4 (23) Thermotaxis assays in which the gradient was generated by a 13S°C cold plate. Statistical significance are as described in Fig. 3. Wild type (n=7); dtrpAl insU3#2-4 (n=9). (0Thermotaxis assays in which the gradient was generated by a 12OC cold plate. Statistical differences are as in Fig. 3. Wild type (n=3); dtrpAl insU3#2-4 (n=3). Figure 4. wild type Slide Warmer Cold Generator 2 6 10 20 Time (rnin) 16 16 20 Time (min) 26 30 40 2 6 10 2 6 1 16 20 Ttme (mln) 0 1 6 2 Time (min) 26 0 2 40 30 5 5 0 4 0 environment. Indeed, larval motility was noticeably slower within the cooled zone than insU3#2-4 within the wanner zone (data not shown). However, the dtrpA1 mutant larvae were able to move within the cooled zone and the larvae gradually migrated farther and faaher down the thermal gradient (towards even colder temperatures) during the course of the assay (data not shown). After the last time point (40 minutes) dtrpA1insU3#2-4 larvae typically were not found at temperatures lower than 10.4OC. Neither wild-type nor d t r p1~insU3#2-4 mutant animals exhibited significant preferences for either half of the plate in the absence of the cold stimulus (release zone -22.5OC--22.5OC; data not shown). These findings suggest that dTRPAl is required for thermotaxis at not only elevated temperatures, but also for thermotaxis at cold temperatures. dwA 1imU3#2-4 mutation exhibited strong defects in cold avoidance either when homozygous or when heterozygous over Df(dtrpA1) (Fig. 5A). Interestingly, Df(dtrpA1) and d t r p1~insU3#2-4 alleles exhibited partially dominant cold avoidance phenotypes (Fig. 5A). The cold avoidance defects exhibited by the Df(dtrpAl)/TM3,actin-GFPand by the d h p 1~imU3#2-4ITM3,actin-GFP animals appeared milder than the defects exhibited by the dwA 1insU3#2-4 homozygous mutant animals or by the dtrpA1insU3#2-4lDf(dtrpA1) larvae and only became pronounced 15-20 minutes into the assay. Neither Df(dtrpA1) nor dtrpA1insU3#2-4 alleles exhibited dominant thermotaxis phenotypes in response to elevated temperatures (Fig. 3B and data not shown). While at this point we can't rule out that TM3,actin-GFP balancer itself causes partially dominant cold avoidance phenotypes, the identical cold avoidance defects exhibited by d t r p ~imu3#224 l homozygous larvae and by the d t r p1~insU3#2-4lDf(dtrpA1) animals suggest that it is the mutation of dTRPAl that is the underlying cause of the observed thermotaxis defects. Additionally, if future Figure 5. Thermotaxis of dtrpA1 mutants at cold temperatures (lO°C cold generator). (A) Wild-type and dtrpA1insU3#2-4 homozygous animals re-graphed from Fig. 4B. d l r p1~i~U3#2-4lDf(dtrpA1) (n=8); dtrpA1u 3 ' 2 - 4 / ~ ~ 3 , a c t i(TM3,act-GFP) n-~~~ (n=3); Df(dtrpAl)/TM3,actin-GFP(n=9). NS = not statistically signifcant. (B) Wild-type and d t r p ~im"#2"'lof(dtrp~l) l animals re-graphed from Fig. 5A. dtrp~lXH8/~f(dtrp~l) (n=5); d t r p ~Cre-out(no-fs)F6 lDf(dtrpA1) (n=8); d h g ~ l ~ lTM3 ~ ~ ,actin-GFP - ~ ~ ~(n=3); ( ~ ~ ~ ) ~ ~ d t r p ~Cre-out(no$s)E5 lDf(dtrpA1) (n=5). Figure 5. wild type 15 20 Time (min) experiments determine that TM3,actin-GFP balancer does not cause a partially dominant cold avoidance phenotype, these data may suggest that proper levels of dTRPAl are essential for larval cold avoidance behavior. Interestingly, the dtrpAl insU3#2-4 mutant larvae were not equally defective for thermotaxis on all cold gradients tested. When the cold generator was set to 17OC (-2 1OC--22OC release zone), dtrpAl insU3#2-4 larvae avoided the cooled zone as well as wild-type larvae (Fig. 4C). As the temperature of the cooled zone was decreased, the thermotactic defect of dtrpAl insU3#2-4 larvae became more pronounced. At 15OC (-20°C- -2 1OC release zone), dtrpAl insU3#2-4 larvae initially avoided the cooled zone in an essentially wild-type manner (Fig. 4D). However, as the assay progressed, more dwA 1imU3#2-4 larvae began to migrate into the cooled zone, and by 40 minutes into the assay almost as many larvae were located in the cooled zone as in the wanner zone. In contrast, wild-type larvae possessed mechanisms to avoid the cooled zone both during the initial stages of the assay and as the assay progressed. As seen in Fig. 4 (E and F), progressive decrease in the temperature of the cooled zone resulted in more pronounced defects at earlier time points. These findings are consistent with two interpretations, which are not mutually exclusive. In the first possibility dTRPAl would be required for larval avoidance of stronger cold but not of more moderate cold temperatures. The animals would therefore possess multiple mechanisms for sensing and avoiding cool temperatures, but as the temperatures decrease, they increasingly rely on dTRPAl signaling in order to sense andlor properly respond to the aversive cold stimulus. Alternatively, d t r p ~ ~ i m U 3 # 2 -may 4 not be a null allele. Thus, there may be sufficient functional dTRPAl protein in d t r p1~insU3#2-4 larvae to signal avoidance of more moderate cold temperatures (17OC cold plate) and signal initial avoidance at colder temperatures (15OC, 13S°C and 12OC cold plates). At even colder temperatures, though, the residual level of functional dTRPAl in d t r p1~insU3#2-4 animals may be below the threshold necessary to mediate the appropriate thermotactic behavior and dtrpA1insU3#2-4 animals are therefore unable to avoid the cold stimulus either initially or later. Potential identijication of mutations that disrupt thermotaxis at cold temperatures but not at elevated temperatures. To further test whether lesions within the dtrpA1 locus disrupt cold avoidance lines for thermotaxis at cold temperatures. We used behavior, we examined dtrp~lCre~Out 10°C temperature for these assays in order to quantitatively compare the behavior of the larvae at early time points (the initial responses to temperature challenge). To avoid possible effects of recessive genetic modifiers on the behavior of the dtrp~lCreeoUt animals, we instead assessed the behavior of d t r p ~CreeOutl~f(dtrp~l) l animals. p ~ l ) exhibited identical defects in As shown in Fig. 5B, d t r p ~ l P H 8 / ~ f ( d t rlarvae cold avoidance behavior as those of the dtrpA1ins U3#2-4lDf(dtrpA1) animals, consistent with the predicted disruptive effect of the engineered fiameshift on the dTRPAl function. However, we were surprised to discover that dtrp~lCre-OM lines that did not contain the engineered frameshift mutation, dtrp~lCree0ut(n0fS)F6 and dhpAl ere-OU~(~O-~S)E~ ,also exhibited significantly defective cold avoidance behavior when assayed over Df(dtrpA1) (Fig. 5B). The cold avoidance defects of dtrp~lCree0ut~n0~F6 lDf(dtrpA1) and dtrp~lCre-Out~nO~ P J E 5 1 ~ f ( d t ranimals p ~ ~ ) were not as severe as those exhibited by the d t r p ~insU3#2l 4 / ~ f ( d t rI)por ~ by the d t r p ~ l f s H 8 / ~ f ( d tanimals r p ~ l ) (that retained the frameshift mutation), but were more severe than the phenotypes of Df(dtrpAI)/TM3,actin=GFP animals. Furthermore, larvae homozygous for the dtrpAl Cre-out(no-f)F6 chromosome also exhibited cold avoidance defects (data not shown), arguing that dtrpAI Cre-out(no-fs)F6 chromosome contains a mutation that affects larval thennotaxis at cold temperatures. dtrpAlere-OU~(~O-~S)F~ 1ine was derived from the dtrpA1insU3#2-4 insertion. We also generated d t r p ~Cre-Out l lines fiom the d t r p ~" l targeted insertions obtained independently from the d h p ~insU3#2-4 insertion in our targeting screen. We have performed preliminary line (assayed as a homozygous mutant), one dtrp~lAe8 characterizations of one dtrp~lAe6 line (assayed in trans over Df(dtrpA1)) and one dtrp~lCre-Out(nO~) line derived from the d t r p ~jNde6 l insertion (assayed as a homozygous mutant). All three of these lines exhibited defects in cold-avoidance behavior (data not shown), arguing that thermotactic defects in response to cold temperatures exhibited by the dtrpAI Cre-out(n0-f) animals were not specific for the dtrpAl insU3#2-4 insertion line. There are several potential explanations for the behavioral defects observed in the larvae, the animals that do not contain the responses to cold by the dtrp~lCre~OUt(nO-fs~ intended engineered fiameshift mutation in the dtrpA1 locus. One is that the thermotactic behavior in response to cold is very sensitive to perturbations of the genome and virtually any mutation will cause a cold avoidance phenotype. The thermotaxis phenotype of dtrpAI larvae in response to cold would thus be fairly nonspecific. While we have not yet ruled out this explanation, we have examined other fly lines that exhibited either no defects for cold avoidance over the whole time course of the assay or exhibited defects in cold avoidance only at later time points (1O°C cold plate; data not shown). In contrast, the animals with lesions in the dhpA1 locus were significantly defective for thermotaxis to cold temperatures from the beginning of the assay on. Another possible explanation lines contain additional unexpected mutations. for our results is that the dtrp~lCre-oUt(nO-fs) These mutations may have been introduced during the I-SceI-mediated integration of the donor DNA into dTRPAl locus. We know that this integration process can create unintended lesions as we have recovered two relatively large deletions resulting from targeted integration events into the dtrpA1 locus (dtrpal and dppal Alternatively, perhaps a more likely explanation for the observed results, is that the mutation(s) affecting cold avoidance behavior of the dtrp~lCre-OUt(nO~ animals may have originated within the targeting construct DNA. Thus far we have only assayed three d h p ~Cre-out(n0-f~)lines generated from two independent d t r p ~iml stocks. Especially in the case of mutations existing within the original targeting construct DNA, it is certainly l lines we have thus far assayed contain these possible that all three of the d t r p ~cre-Out(nO-fs) mutation(s). We are currently sequencing the targeting construct that we have employed for targeted insertional mutagenesis of the dhpAl locus. We will then analyze existing d k P lcre+ut(no-fs) ~ lines for the presence of any mutations discovered within the targeting construct and determine whether any of the mutations discovered correlate with the cold avoidance defect exhibited by dtrpAl Cre-out(no-fs) animals. l animals did not exhibit thermotaxis It is interesting to note that d t r p ~Cre-out(n0-fs)F6 defects in response to elevated temperature (Fig. 3B). Thus, if the unknown mutation(s) were responsible for the cold avoidance defects of the dtrp~lCre~0Ut(n0~F6 animals, these mutation(s) might be selectively disrupting the role dTRPAl plays in cold avoidance but not in heat avoidance. Thus, these mutation(s) might provide valuable insights into the mechanisms by which dTRPAl mediates temperature sensation. dTRPAl is requiredfor select larval behaviors. It is apparent from our cold avoidance experiments that d v A l mutant larvae retain some capacities for responding to select temperatures, indicating that global execution of sensory behaviors is not disrupted in the dtrpAl mutant larvae. To determine whether the behavioral defects exhibited by d w A l mutant animals might be specific for themotaxis we examined the responses of dhpA1insU3#2-4 mutant larvae to other sensory stimuli. dtrpA1insU3#2-4 larvae avoided the repellent odor n-octyl acetate [26] at least as well as wild-type (Fig. 5A). Similarly, there were no significant defects in the responses of dtrpAl insU3#2-4 mutant animals to the attractant odor propionic acid [26] (Fig. 5B). Furthermore, phototaxis of late lSt/early2ndinstar [27] d@Al insU3#2-4 larvae was indistinguishable from that of wild-type larvae (Fig. 5C). These findings suggest that while dTRPAl is required for a subset of temperature-driven behaviors, dTRPAl appears dispensable for other sensory modalities such as chemosensation and vision. DISCUSSION. Drosophila melanogaster exhibits strong temperature preferences and robust avoidance of unfavorable thermal environments [18-20,281. We have previously implicated dTRPAl in mediating larval responses to elevated temperature using RNAi [19]. Here, we continued our examination of the molecular mechanisms that mediate themotaxis in Drosophila. Figure 6. dtrpA1 mutant animals respond normally to temperature-independent sensory stimuli. (A) Wild-type (n=7) and dirpA1insU3#2-4 homozygous animals (n=7) robustly avoid a repellent odor n-octyl acetate. #p<.05 (B) Wild-type (n=6) and dtrpAl insU3#2-4 mutant animals (n=6) are equally attracted to an attractive odor propionic acid. (C) Wildtype (n= 10) and dtrpAl insU3#2-4 mutant animals (n=lO) exhibit indistinguishable phototactic behavior . Figure 6. Propionic acid noctyl acetate -1 J 20 30 Time (min) wild type dtrpAlina4 5 2 Time (min) To facilitate further studies of thennotaxis in Drosophila, we generated loss of h c t i o n alleles of dhpA1. As expected, we found that dhpA1 mutants were defective for thermotaxis at elevated temperatures. Interestingly, however, the thermotaxis defects exhibited by d h p A l mutant animals at elevated temperatures were less severe than the thermotaxis defects of the larvae from embryos injected with double stranded RNA (dsRNA) targeting dtrpA1, the dtrpA1(RNAi) larvae. The dtrpA1(RNAi) larvae exhibited no preference between warmer and cooler zones when placed onto a thermal gradient featuring a -3 1OC--3 5OC release zone (avoidance index (AI)=O) [19]. Those data were collected following multiple dsRNA injections throughout the course of a year and the behavior of the dtrpA1(RNAi) larvae was always consistent. The finding that dtrpA1 mutant larvae exhibited a slight preference for the cooler zone of the thermal gradient (AI=-0.2-0.4; A1 of 0.2 is equivalent to 60% of the larvae migrating away fiom the warmer area and 40% of the larvae migrating into the warmer zone) suggests that dtrpAl mutant larvae still possess some means to sense and appropriately respond to temperature. This difference between the thennotactic behavior of the dtrpA1(RNAi) larvae and of the dtrpA1 mutant larvae might be due to the dsRNA employed in our initial RNAi experiments causing off-target effects and altering expression of other genes in addition to dtrpAl. Alternatively, the dtrpAl alleles we have generated might not be null alleles. While the mutation we have engineered into the dtrpA1 locus should produce a truncated and non-functional dtrpAl protein, it is possible that due to alternate splicing, alternate transcriptional start site or due to the read-through of the frameshift mutation, some functional dTRPAl protein can be produced in the dtrpAl mutants. Finally, other mechanisms may compensate for the loss of dTRPAl in dtrpA1 mutants. Our finding that different dtrpA1 mutant lines we have generated can exhibit quantitatively different responses to elevated temperature (Fig. 3A) suggests that genetic modifiers can modulate the behavioral responses of dtrpA1 mutant animals. In the future, it might be possible to identify some of the genes that modulate thermotactic behavior of the dtrpA1 mutant animals in response to elevated temperatures. These genes might include other TRPs, potassium channels, and signaling molecules. One possibility is that in the dtrpA1 mutant background PYREXIA or PAINLESS can compensate for the loss of dTRPA1. While neither painless nor pyrexia mutants exhibited thermotaxis defects in our elevated temperature assay ([I 91 and data not shown), the temperature responsiveness of these channels might permit some thermotaxis in the absence of dTRPA1. We are currently generating pain1ess;dhpAI and pyrexia,dtrpAl double mutant animals to examine this possibility. Unexpectedly, we've discovered that dhpAI mutant animals were also defective for thermotaxis at cold temperatures. Multiple dtrpA1 mutants failed to avoid the cooled environment in our cold thermotaxis assay. While we need to further examine the nature l animals, of the therrnotaxis defects at cold temperatures exhibited by the d t r p ~CreeOut(nO~) our findngs that multiple dtrpA1 mutant lines exhibited strong cold avoidance defects suggest that dTRPAl is required for thermotaxis at cold temperatures as well as for themotaxis at elevated temperatures in Drosophila larvae. It is interesting that dtrpAl mutant animals were not defective for avoidance of all cold temperatures. Rather, the dtrpAl mutant animals exhibited more prominent and faster developing thermotactic defects to colder temperatures than to temperatures that were not as cold (Fig. 3). These findings suggest several possible models for how dTRPAl may be regulating thermotaxis to heat and to cold temperatures. One possibility is that dTRPAl is a temperature sensor that is activated by broad ranges of temperatures in vivo. Another possibility is that dTRPAl is a bimodal ion channel that can be activated by either hot or cold temperatures, but not by intermediate temperatures. A third possibility is that dTRPAl does not function as a temperature sensor at either elevated or cold temperatures, but instead plays another essential role in the function of the thermosensory circuit(s). With the available data we currently can't distinguish between these models. However, it has been reported that dTRPAl protein can be activated by elevated temperatures when expressed in heterologous cells [23], with a threshold of -27OC. This suggests that dTRPAl mediates thermotaxis to elevated temperatures by acting as a temperature sensor and that more moderate temperatures do not activate TRPAl. However, there is currently no direct evidence that dTRPAl is capable of being activated by temperature in Drosophila neurons. In addition to reporting that dTRPAl can be activated by elevated temperatures in heterologous cells, Ardem Patapoutian's and Tim Jegla's laboratories [23] also reported that dTRPAl was not activated by cold temperatures. We are currently re-examining the ability of dTRPAl to be activated by cold temperatures in collaboration with Tim Jegla's laboratory. Furthermore, because our dtrpAl mutants may be hypomorphic alleles, and since the threshold levels of dTRPA1 required for mediating avoidance of cold and heat and moderate temperatures may well vary, it is difficult to conclude from our behavioral data that dTRPAl only mediates avoidance of hot and cold temperatures. Instead, another mechanism could compensate for the loss of TRPAl at moderate temperatures. Future experiments may shed light on the mechanisms by which dTRPAl mediates thennotaxis. However, our results suggest that dTRPAl is the first thermoTRP that mediates animals' behavioral responses to both hot and cold temperatures in vivo. Previous models for temperature sensation in animals proposed that animals sense temperature using a series of thermoTRPs to detect the broad spectra of temperatures, with each thermoTRP mediating responses to a limited portion of the temperature spectrum [I]. Mouse knockouts lacking a particular thermoTRP not only exhibit behavioral defects in responses to only narrow ranges of temperatures but also exhibit only partial thennosensory behavioral defects at those temperature ranges [9, 14-171. The lack of strong thermosensory behavioral defects might be attributed to other thermoTRPs and non-TRP temperature responsive molecules partially compensating for the loss of a single thermoTRP. For instance, background K' channels TREK-1, TREK-2 and TRAAK, ATP receptor P2X3 and members of the DegenerinIEpithelial Sodium Channel (DEGIENaC) family are all non-TRP molecules that can either be activated by temperature or can regulate temperature-dependent responses in animals [29-331. Furthermore, the partial nature of the thermosensory behavioral defects exhibited by the animals lacking individual thermoTRPs might be due in part to the remaining thermoTRPs exhibiting sufficient plasticity in their temperature responsiveness to partially compensate for the loss of a particular individual thermoTRP. Indeed, numerous molecules such as membrane lipids, calcium, phosphorylation/de-phosphorylation,and membrane potential can dramatically alter temperature responsiveness of thermoTRPs in heterologous cells and in cultured neurons [34-411. For instance, injury or inflammation can decrease animals' thermal threshold, a phenomenon known as thermal hyperalgesia. Membrane potential in particular can shift the temperature activation threshold so dramatically that a heat-activated channel such as TRPVl can be activated at cool temperatures [40,41]. It is thought that membrane potential affects temperature thresholds of thermoTRPs by regulating the probability of channel opening [40-421. But while the temperature threshold of a thennoTRP can be shifted, other properties of the channel appear to be unchangeable: Regardless of the membrane potential or interactions with the channel modulators, higher temperatures are still more effective than lower temperatures at activating TRPV1, eliciting larger currents. Similarly, while the temperature threshold of cooling capable of activating TRPM8 can be shifted so that some activity is detected at higher temperatures, even at conditions that shift temperature threshold of TRPM8, TRPM8 is still more effectively activated at colder temperatures. In other words, regardless of the membrane potential, TRPVl is still activated by heating while TRPM8 is still activated by cooling. Our fndings that dtrpAl mutants exhibit thermotaxis defects at both elevated and cold temperatures suggest that dTRPAl controls animals' behavior at diverse temperature ranges and raises the possibility that dTRPAl protein could potentially act as a bi-modal thennoTRP, activated by both elevated temperatures and by cold. In addition to demonstrating that dtrpA1 mutants exhibited defects in thennotactic behavior at both elevated and cold temperatures, we also found that dtrpA1Cre-out(no-fs)F6 lines exhibited significant defects in thermotactic behavior at cold but not at elevated temperatures. These animals were not predicted to contain the engineered frameshift mutation or the he6 or the Ae8 mutations. While the molecular basis of the thermotaxis l animals is not yet clear, one exciting possibility is defects exhibited by d r r p ~CreeOUt(nO-fsS) that d t r p1'''-out(n0-f~) ~ mutation(s) selectively disrupts thermotactic response to cold temperatures but not to heat. How thermoTRPs, or any ion channels, are gated by temperature is unknown. The identification of the unintended mutation(s) in the d t r p ~Cre-out(no-fs)lines might reveal regions of dTRPAl required for cold sensation. Alternatively, these unintended mutation(s) in dtrpA1Cre-Out(nO-fs)lines may disrupt interactions of dTRPA1 with other modulators/effectors. Identification of these mutation(s) should significantly contribute to our understanding of the role dTRPAl plays in thermotaxis. What are the cells that mediate thermotaxis to cold temperatures? Liu, et al., [20] reported that the terminal organ located in the nose of the larvae exhibits changes in calcium transients in response to cooling and that expressing Tetanus Toxin Light Chain (TNT; an inhibitor of synaptic vesicle release) in the terminal organ results in defects in 3rdinstar larvae cold avoidance. dTRPAl protein expression was not detected in the terminal organ [19], however, suggesting that dTRPAl might be mediating thermotaxis to cold by acting in cells other than the terminal organ. Liu, et al., [20] used the same model cold generator we've employed, set at S°C which resulted in a thermal gradient of -1 1°C to -1 8OC measured at approximately same positions as points a and fin Fig. 4A. Thus, the temperatures they have used are significantly lower than the temperatures we have used in our thermal gradients. To determine the relationship between dTRPAl and the terminal organ neurons we will inhibit the function of the terminal organ neurons with TNT and will determine at which temperatures the terminal organs is required for thermotaxis on our gradients and at even lower temperatures. If the terminal organ is required only for very low temperatures, we will also examine whether dtrpAl mutants are defective for thennotaxis at those temperatures. We will also use the Gal4 drivers that express in the dTRPAl -expressing cells in the central brain (that are required for thennotaxis at elevated temperatures [19]) to express inhibitors of neuronal function, such as TNT, in those neurons. This might allow us to determine which cells are required for cold avoidance behavior, the relationship between dTRPAl and the terminal organ and whether same dTRPAl -expressing cells mediate both themotaxis at elevated and cold temperatures. MATERIALS AND METHODS. F& strains Flies were maintained in a humidified 25OC incubator and fed standard fly food. Fly strains used for generation of dtrp~lins and dtrpAlCre-Out alleles were obtained fiom the Bloomington Stock Center. Deficiencies uncovering the dtrpA1 locus (Df(dtrpA1): Df(3L)ED4415 and Df(3L)ED4416) were obtained from the Szeged stock center. p$ flies [21] were kindly provided by Craig Montell. pain2 flies [22] were kindly provided by Dan Tracey. Molecular genetics Mutants were balanced over balancers with larval markers (TM6B, SM6:TM6bY CyO-actinGFP or TM3-actinGFP). For d t r p ~ l " alleles generation, protocol described in [24,43] was followed with minor exceptions. Briefly, the targeting construct was generated by cloning BglI-KpnI dTRPAl genomic DNA fiagment fiom BAC RP98-10P9 (Open Biosystems) into pBluescriptI1. I-SceI site was inserted into BamHI site in exon 8 of dTRPA1, ClaI site in exon 4 was filled in to generate NruI site (the engineered fiameshift mutation), and the construct was cloned into NotIIKpnI digested pTV2. The targeting construct had 3.1kb homology fiom the 5' end to the I-SceI insertion and 2.2kb of homology from the I-SceI site towards the 3' end and the Not1 site in pTV2 faced the 5' end of the truncated dTRPAl gene. The targeting construct was introduced into Drosophila germline to generate donor flies. Two different donor insertions on the second chromosome were used for targeting. 2-4 males homozygous for the donor insertion were crossed to 3 w; h s ~ l ~ , h s - ~ - ~ c e virgins. ~ , ~ c 0-3 o / day ~ ~ old ,~~~ embryos/larvae were heat shocked once at 38OC for 65 5 minutes. 3 Non-CyO virgins were crossed to 1-4 w; eyFlp males and 394 crosses were set up for each donor (788 total, 2364 females). 33 putative targeting events, judged by presence of colored eyed flies (at least 27 independent), were recovered and analyzed. 17 events were recovered fiom one donor and 16 from the other. PCR analysis revealed that three of the events were donors, one a non-homologous insertion onto the second chromosome and 29 (24 independent) appeared to be insertions into the dTRPAl locus. Primer sequences are available upon request. Cre-out lines ( d t r p ~ l were generated as described [25] and analyzed by PCR and restriction digests to identify the lines containing the engineered frameshift, Ae6 and Ae8 mutations. Histology Rat polyclonal dTRPAl antiserum was used at 1:1000 as described [19]. Images were obtained using a Nikon PCM2000 confocal microscope and Zeiss Axioplan2 with Nikon DXM1200 camera. Behavioral assays Larvae for the behavioral assays were raised on molasses plates. Heat avoidance themotaxis assays were performed as described [19]. Generally 140 late lSt/early2nd instar larvae were harvested for each assay. Cold avoidance assays were similar to those of [20], but with different temperatures and 1" instar larvae used instead of 3rdinstar larvae assayed in [20]. 240 1St instar larvae were harvested for each cold assay, washed in distilled water, dried, and placed in the release zone with a paint brush. Cold avoidance assays were performed on 2% agarose (25 mL) in flat lids from 150 x 15 rnm Petri dish plates (VWR 25384-139) placed so that the release zone (-4mm) was in the middle of the intedace of a metal slide warmer (at room temperature) and a cold plate (AHP-301CP, TECA corporation, USA) -14 mm apart. Cold generator surface was covered with thin black plastic cut fiom 34.3 pm thick RNW 4050 trash bags (Office Depot) to ease the scoring of animals on the cooled surface. Agarose surface was allowed to equilibrate for at least one hour prior to each assay (covered by a glass slab to avoid temperature dissipation). For each time point, the number of larvae that migrated out of the release zone to the cooled side and to the not-cooled (RT) side was counted. The avoidance index (AI) = (number of larvae on the RT side - number of larvae on cooled side)/(number of larvae on the RT side + number of larvae on cooled side). Surface temperatures were determined using a Fluke 5211 thermometer with dual K-type flat surface probes (VWR). Larvae were washed in distilled water and lined up in the release zone with a paint brush as described [19]. Plate was immediately covered with a glass slab (t = 0) and a cardboard box was then placed over the setup so that the assay was performed in the dark. Odor avoidance assays were performed and scored as described [19,26]. Briefly 2.5 pL of undiluted n-octyl acetate (Sigma) or 1.5pL of undiluted Propionic acid (J.T. Baker) was presented in microcentrifuge tube cap opposite an empty cap in a standard 90 mm plastic Petri dish with 10 ml2% agarose. 60 late lst/early2ndinstar larvae were harvested for each assay. For larval phototaxis assays we used a 90 mm plastic Petri dish with four lighvdark quadrants, as described [27], except clear 2% agarose was used and the quadrants were generated using light-impermeable black electrical tape to cover two quarters on the bottom of the dish. Light was applied from underneath the behavioral plate by TW-26 transilluminator (VWR). The behavioral plate was raised -1 2.5 cm above the light source to avoid heating of the plate. Cardboard box was placed over the setup at the start of each assay to minimize the influence of the external light. Late lSt/early2ndinstar larvae were assayed. For each time point, the number of larvae that migrated out of the release zone to the stimulus zone (odor or light) and to the no stimulus zone (control: no odor or dark) side was counted. The A1 for odor or light was calculated. A1 = (number of larvae on the control side - number of larvae on stimulus side)/(number of larvae on the control side + number of larvae on stimulus side). All statistical comparisons used two-tailed unpaired t-test versus wild-type. ACKNOWLEDGEMENTS We thank members of the Garrity lab for helpful discussions, Craig Monte11 for pyrexia mutants and Dan Tracey for painless mutants. M.R. was supported by the Centocor fellowship, an NSF predoctoral grant and an NIH training grant. This research was also supported by a grant fiom the NIH to Paul Garrity. REFERENCES: Patapoutian, A., et al., ThermoTRP channels and beyond: mechanisms of temperature sensation. Nat Rev Neurosci, 2003.4(7): p. 529-39. Montell, C., The TRP superfamily of cation channels. Sci STKEY2005. 2005(272):p. re3. Caterina, M.J., et al., The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature, 1997.389(6653):p. 8 16-24. Caterina, M.J., et al., A capsaicin-receptor homologue with a high thresholdfor noxious heat. Nature, 1999.398(6726):p. 436-4 1. Peier, A.M., et al., A heat-sensitive TRP channel expressed in keratinocytes. Science, 2002.296(5575): p. 2046-9. Xu, H., et al., TRPV3 is a calcium-permeable temperature-sensitivecation channel. Nature, 2002.418(6894): p. 181-6. Smith, G.D., et al., TRPV3 is a temperature-sensitive vanifloid receptor-like protein. Nature, 2002.418(6894): p. 186-90. Guler, A.D., et al., Heat-evoked activation of the ion channel, TRPV4. J Neurosci, 2002.22(15):p. 6408-14. Talavera, K., et al., Heat activation of TRPM.5 underlies thermal sensitivity of sweet taste. Nature, 2005.438(7070): p. 1022-5. Peier, A.M., et al., A TRP channel that senses cold stimuli and menthol. Cell, 2002. 108(5):p. 705-15. McKemy, D.D., W.M. Neuhausser, and D. Julius, Identification of a cold receptor reveals a general role for TRP channels in thermosensation. Nature, 2002. 416(6876):p. 52-8. Story, G.M., et al., ANKTMI, a TRP-like channel expressed in nociceptive neurons, is activated by cold temperatures. Cell, 2003.1 12(6):p. 8 1 9-29. Bandell, M., et al., Noxious cold ion channel TRPAl is activated bypungent compounds and bradykinin. Neuron, 2004.41(6): p. 849-57. Caterina, M.J., et al., Impaired nociception andpain sensation in mice lacking the capsaicin receptor. Science, 2000.288(5464): p. 306- 13. Moqrich, A., et al., Impaired thermosensation in mice lacking TRPV3, a heat and camphor sensor in the skin. Science, 2005.307(5714): p. 1468-72. Lee, Heyet al., Altered thermal selection behavior in mice lacking transient receptor potential vaniNoid 4. J Neurosci, 2005.25(5): p. 1304-10. Kwan, K.Y., et al., TRPAl contributes to cold, mechanical, and chemical nociception but Is not essentialfor hair-cell transduction. Neuron, 2006.50: p. 277-289. Sayeed, 0. and S. Benzer, Behavioral genetics of thermosensation and hygrosensation in Drosophila. Proc Natl Acad Sci U S A, 1996.93(12):p. 607984. Rosenzweig, M., et al., The Drosophila ortholog of vertebrate TRPAl regulates thermotaxis. Genes Dev, 2005.19(4): p. 419-24. Liu, L., et al., Identification andfunction of thermosensory neurons in Drosophila larvae. Nat Neurosci, 2003.6(3):p. 267-73. Lee, Y., et al., Pyrexia is a new thermal transient receptor potential channel endowing tolerance to high temperatures in Drosophila melanogaster. Nat Genet, 2005.37(3): p. 305-10. Tracey, W.D., Jr., et al., painless, a Drosophila gene essentialfor nociception. Cell, 2003.113(2): p. 261-73. Viswanath, V., et al., Opposite thermosensor inpuitfly and mouse. Nature, 2003. 423(6942): p. 822-3. Rong, Y.S. and K.G. Golic, Gene targeting by homologous recombination in Drosophila. Science, 2000.288(5473):p. 20 13-8. Rong, Y.S., et al., Targeted mutagenesis by homologous recombination in D. melanogaster. Genes Dev, 2002. 16(12): p. 1568-81. Heimbeck, G., et al., Smell and taste perception in Drosophila melanogaster larva: toxin expression studies in chemosensory neurons. J Neurosci, 1 999. 19(15): p. 6599-609. Lilly, M. and J. Carlson, smellblind: a gene required for Drosophila olfaction. Genetics, 1990.124(2):p. 293-302. Zars, T., Two thermosensors in Drosophila have dzfferent behavioralfunctions. J Comp Physiol [A], 200 1. 187(3): p. 235-42. Maingret, Feyet al., TREK-1 is a heat-activated background K(+) channel. Embo J,2000.19(11): p. 2483-91. Kang, D., C. Choe, and D. Kim, Thermosensitivityof the two-pore domain K+ channels TREK-2 and T W K . J Physiol, 2005.564(Pt 1): p. 103-16. Shimizu, I., et al., Enhanced thermal avoidance in mice lacking the ATP receptor P2X3. Pain, 2005.116(1-2): p. 96-108. Souslova, V., et al., Warm-coding deJcits and aberrant inflammatorypain in mice lacking P2X3 receptors. Nature, 2000.407(6807):p. 1015-7. Askwith, C.C., et al., DEGIENaC ion channels involved in sensory transduction are modulated by cold temperature. Roc Natl Acad Sci U S A, 2001.98(11): p. 6459-63. Van Der Stelt, M. and V. Di Marzo, Endovanilloids. Putative endogenous ligands of transient receptor potential vanifloid 1 channels. Eur J Biochem, 2004. 271(10): p. 1827-34. De Petrocellis, L. and V. Di Marzo, Lipids as regulators of the activity of transient receptor potential type Vl (TRPVI) channels. Life Sci, 2005. 77(14): p. 165 1-66. Nilius, B., et al., The Ca2+-activated cation channel TRPM4 is regulated by phosphatidylinositol4,5-biphosphate.Embo J, 2006.25(3): p. 467-78. Nilius, Be, et al., Regulation of the Ca2+ sensitivity of the nonselective cation channel TRPM4.T. Biol Chem, 2005.280(8): p. 6423-33. Rohacs, T., et al., PI(4,5)P2 regulates the activation and desensitization of TRPM8 channels through the TRP domain. Nat Neurosci, 2005. 8(5): p. 626-34. Bhave, G., et al., CAMP-dependent protein kinase regulates desensitization of the capsaicin receptor (VRl) by direct phosphorylation. Neuron, 2002.35(4): p. 72 13 1. Nilius, B., et al., Gating of TRP channels: a voltage connection? J Physiol, 2005. 567(Pt 1): p. 35-44. 41 . 42. 43. Voets, T., et al., The principle of temperature-dependent gating in cold- and heatsensitive TRP channels. Nature, 2004.430(7001): p. 748-54. Liman, E.R., Thermal gating of TRP ion channels:food for thought? Sci STKE, 2006.2006(326):p. pe12. Sears, H.C., C.J. Kennedy, and P.A. Gmity,Macrophage-mediated corpse engdjhent is requiredfor normal Drosophila C .morphogenesis. Development, 2003.130(15):p. 3557-65. CONCLUSIONS AND FUTURE DIRECTIONS. We have identified dTRPAl as an important mediator of thennotaxis in Drosophila. Both the larvae in which dTRPAl expression was knocked down with RNA interference (RNAi) and the dtrpAl loss of function mutants were strongly defective for avoiding elevated temperatures (-3 1°C--3S°C). Furthermore, dtrpA1 mutants were also defective for avoiding cold temperature environments. These findings raise an exciting possibility that dTRPAl may be mediating thennosensory responses to both elevated and cold temperatures. For proper execution of a thermotactic behavior an animal must sense the temperature stimulus and then properly translate the temperature sensation into a behavioral response. dTRPAl has been demonstrated to be a warm temperatureactivated ion channel in heterologous cells [I]. This suggests that dTRPAl might control thennotactic behavior at elevated temperatures by functioning as a sensor of elevated temperatures. Viswanath, et al., [I] did not observe cold activation of dTRPA1, but that certainly doesn't rule out a possibility that dTRPAl can be activated by cold temperatures. Thus, cold-responsiveness of dTRPAl will be re-examined in cultured cells. It is possible that dTRPAl controls thermotactic responses to both elevated temperatures and to cold by directly sensing temperature. Alternatively, dTRPAl might mediate cold-avoidance behavior by acting at a step downstream of the initial temperature sensation step. An important direction for the fbture is to investigate whether dTRPAl functions as a temperature sensor for heat and/or cold in vivo. This may be accomplished by performing electrophysiological recordings from the neurons that require dTRPAl for thermotactic behavior in response to elevated temperatures andlor to cold. If dTRPA mediates thermotaxis by acting as a thermal sensor, warming or cooling larval brain might induce detectable changes in those neurons in the dTRPAl -dependent manner. While larval neurons may be too small to be a good target for direct electrophysiological recordings, it may be possible to perform these recordings in adults. However, while these experiments might be extremely informative, prior to these experiments we must identify the cells that mediate thermotaxis at cold temperatures and determine whether dTRPAl functions in those cells to mediate thermotaxis at cold temperatures. It will be also interesting to examine adult behavior of dtrpAl mutants. Adults can tolerate exposure to higher temperatures than the larvae, and thus in addition to potentially gaining a model system for physiological recordings, we might be able to more thoroughly examine the contributions different TRP channels make for thermosensation in the same assay. In our larval thermotaxis assay, we can't assess thermotaxis at temperatures much higher than we are currently using because the larvae rapidly die [2]. Therefore, to test whether a molecule mediates temperature sensation andlor thermosensory behavior at higher temperatures than we can achieve in our assay, we have to resort to different behavioral assays, such as a high-temperature (38OC-52OC) nociceptive stimulus [3]. For instance, we used a high-temperature nociception paradigm together with our lower temperature thermotaxis assay to test whether dTRPAl was required for responses to higher temperatures than we could achieve in our assay [2]. Adult choice test [4] and adult thermal gradient assays [4,5] should allow us to compare mutants' behavior at different temperatures while using the same behavioral paradigm. In the larvae, dTRPAl is expressed in restricted populations of cells. We have previously implicated one of these cell populations, dTRPAl -expressing cells in the central brain, in mediating larval responses for avoidance of elevated temperatures. The contribution of other dTRPAl -expressing cell populations is unknown and will need to be investigated in the fbture. Other important fbture directions include elucidating the dTRPAl thermosensory circuit. To what cells do dTRPAl -expressing cells project to mediate thennotactic behavior? In the case for thennotaxis at elevated temperatures, the dTRPAl -expressing cells in the central brain, that are required for thermotaxis at high temperatures, appear to project to the Kenyon cells of the mushroom body (Fig. 1). The mushroom body is thought to be the fly's learning center, but it appears to be dispensable for learning assays other than for the olfactory associated learning [6]. We ablated the mushroom bodies by expressing cell-death inducing genes in the mushroom body neurons. And while the mushroom body neurons in those larvae were undetectable, the afflicted larvae were still able to therrnotax away fiom the heated area (although they were significantly developmentally delayed and exhibited considerable delay in moving out of the release zone) (data not shown). It is possible that instead of projecting to the mushroom body neurons, dTRPAl -expressing cells instead make contacts with other neurons in the vicinity of the Kenyon cells. Alternatively, it is possible that other groups of cells transduce the signal from the dTRPAl -expressing cells in parallel with the mushroom body neurons. It is interesting to note that temperature can be used as a reinforcer in fly learning assays [7]. It will thus be very interesting to determine whether dTRPAl plays a role in regulating temperature-mediated learning. Figure 1. dTRPAl -expressing neurons project to the mushroom body. (A-C) dTRPAl antibody specifically stains Corpus Cardiaca neurons (*) and a few central brain neurons (brackets and arrowhead). (DOE) dTRPAl-expressing cells in the central brain appear to project towards the Kenyon cells of the mushroom bodies (expressing OK107-Gal4 driving UAS-GFP). Images courtesy of Tim Tayler. Figure 1. Activation of thermoTRPs by temperature can be modulated by other intracellular factors such as phospholipids, phosphorylationlde-phosphorylation,ca2+and membrane potential. What are the factors that control dTRPAl activity? Viswanath, et al., [I] reported that dTRPAl warming-evoked currents are outwardly rectifying, indicating that dTRPAl may be modulated by membrane potential. Other regulators of dTRPAl, however, are currently unknown. Since TRP channels are known to oligomerize [8], we attempted to determine whether any other Drosophila TRPs might function together with dTRPAl to mediate thermotaxis. We used RNAi to inhibit the other Drosophila TRP subfamilies. We injected double stranded RNA (dsRNA) targeting the members of the TRPA, TRPV, TRPM, TRPP, TRPN (NompC) and TRPML subfamilies. No dsRNAs except for those targeting dTRPAl yielded thermotaxis phenotypes ([2] and data not shown). RNAi-mediated targeting the members of the TRPC subfamily (TRP, TRP-like and TRP-y), however, has not yet been done. In the fbture, the lab will conduct genetic screens for thermotaxis mutants. These screens should yield regulators of dTRPAl activity, other molecules required for temperature sensation in flies, and genes necessary for the development and k c t i o n of the thermosensory circuit. Such screens could be done for regulators of both thermotactic behavior at elevated temperatures and for regulators of thermotaxis at cold temperatures. Other potentially useful screens that might facilitate identification of other regulators of themotaxis can take advantage of the partial thermotaxis defects exhibited by &pAl mutant animals at moderately cold temperatures. Thus, it might be possible to conduct enhancer/suppressor screens searching for modifiers of the partial dtrpAl mutant phenotype and such screens may identify different molecules than would be identified by screens in a wild-type background. In conclusion, this thesis represents the first molecular analysis of thermotaxis behavior in Drosophila. We have implicated a putative temperature sensor, dTRPA1, in mediating larval behavioral responses to both elevated and cold temperatures. We have also identified dTRPAl-expressing cells and implicated a subset of these cells in mediating thermotactic behavior at elevated temperatures. Finally, our analysis of dtrpA1 and painless mutant animals, as well as of the cells in which they are required for mediating the behavioral paradigms they were identified in, suggest that similar to mammals, Drosophila utilizes different neurons and different molecules for mediating different thennosensory responses. REFERENCES. Viswanath, V., et al., Opposite thermosensor i n h i t f l y and mouse. Nature, 2003. 423(6942): p. 822-3. Rosenzweig, M., et al., The Drosophila ortholog of vertebrate TRPAI regulates thermotaxis. Genes Dev, 2005.19(4): p. 4 19-24. Tracey,W.D., Jr., et al., painless, a Drosophila gene essentialfor nociception. Cell, 2003. 113(2): p. 261-73. Sayeed, 0. and S. Benzer, Behavioral genetics of thermosensation and hygrosensation in Drosophila. Proc Natl Acad Sci U S A, 1996.93(12): p. 607984. Lee, Y., et al., Pyrexia is a new thermal transient receptor potential channel endowing tolerance to high temperatures in Drosophila melanogaster. Nat Genet, 2005.37(3): p. 305-10. Wolf, R., et al., Drosophila mushroom bodies are dispensablefor visual, tactile, and motor learning. Learn Mem, 1998.5(1-2): p. 166-78. Wustmann, G., et al., A new paradigm for operant conditioning of Drosophila melanogaster. J Comp Physiol [A], 1996. 179(3): p. 429-36. Schaefer, M., Homo- and heteromeric assembly of TRP channel subunits. Pflugers Arch, 2005.451(1): p. 35-42. ACKNOWLEDGEMENTS. I would like to thank my advisor, Paul Garrity, for everything he has done to help me to become a better scientist, for all his help, advise and guidance. I would also like to thank my thesis committee members, Bob Horvitz, Terry Orr-Weaver and Chip Quinn for truly helpful discussions and much encouragement. Furthermore, I'd like to thank my family, friends and colleagues for their unwavering support. I further thank Karen Brennan, Paul Phelps and Caitlin Rex for technical assistance with this project, Betsey Walsh and Hilda Harris-Ransom for being there with answers to pretty much any administrative questions, Jessica Whited for making the lab life much more entertaining and for all the goods she baked for the lab over the years, Sunny Wong, Mandy Tarn and Anna and Sujith Vijayen for being there for me over the years. Special acknowledgements go to the NSF, NIH and the Centocor foundation for supporting this research. And thank you for reading this to the end.