The behavioral characteristics of thermal nociception in adult Drosophila
Vertebrates exhibit escape/avoidance behavior when they are exposed to painful stimuli. We observed a similar behavior, the jump response, when adult Drosophila was exposed to noxious heat. Therefore, we established two methods to measure this behavior. Our results indicate that the nociceptive behavior in adult flies has three distinct characteristics, which resemble those in mammals.
Second, adult flies are able to distinguish thermal nociception from innoxious thermosensation, which resembles mammals’ ability to distinguish painful heat from warmth. Our studies demonstrated that the nociceptive behavior is significantly different from the behavior induced by innoxious thermosensation in locomotor pattern and response time (see Results). Moreover, the normal nociceptive behavior of biz flies (Fig. 3b) further confirms the difference between thermal nociception and innoxious thermosensation.
Third, adult Drosophila are able to respond to different kinds of noxious stimulus. We observed that when given a cotton cord wet with 1 m NaOH or HCl, each fly caught the cord (only touching it with its legs) for a few seconds, then released it suddenly and refused the cord for several minutes. This result suggests that adult flies also respond to noxious chemical stimuli in addition to noxious thermal stimuli. Additionally, strong electrical stimuli can elicit the jump response in adult flies (T. Tully, personal communication). The observations above suggest that the nociceptors in adult Drosophila can be activated by different kinds of noxious stimuli, as do the C-fibers in mammals (Lynn 1984).
Besides the three characteristics above, there may be another two shared by mammals and Drosophila adults. First, the threshold temperature for thermal nociception in mammals is approximately 45 °C (Vyklicky 1984). Our results suggest that the nociceptive threshold temperature in adult Drosophila is also approximately 45 °C. A previous study reported that biz flies distributed randomly across the temperature gradients from 23 to 36.5 °C, and even to 45 °C (Sayeed & Benzer 1996), but we found that biz flies behaved normally in the CLB assay (Fig. 2d), suggesting that the nociceptive threshold temperature in adult Drosophila is approximately 45 °C. This hypothesis is also supported by the result that the mean latencies decreased significantly with an increase of plate temperatures from 43 to 47 °C in the HP assay (Fig. 2b). However, we still lack a direct evidence to prove this hypothesis because of the technical difficulty in measuring the cuticle temperature of adult flies (see below).
Second, we tried to heat the different parts of thorax and abdomen with a focused laser beam. Adult flies were also able to jump away from the stimuli, suggesting that the nociceptors in adult flies are systemically located in their bodies, including the legs, the abdomens and the thoraxes. The systemic distribution of nociceptors may be a common characteristic shared by mammals and adult Drosophila (Lynn 1984).
The advantages and disadvantages of the two behavioral assays
Secondly, our results illustrated that the mean latency was stable in the CLB assay (Fig. 2c), as in the paw-withdraw test in rats (Hargreaves et al. 1988), despite repeated stimuli. This advantage is considered to be an important criterion for a well-developed nociceptive assay (Vyklicky 1984).
Thirdly, similar to the rotarod treadmill test, which served as the control for the HP test in mice (Chen et al. 2001), the jump test was established to serve as the control for the two assays. Although similar methods have been established (Elkins & Ganetzky 1990; Kaplan & Trout 1974; Snowball & Holmqvist 1994), our assay is simple and effective. The results observed in glc1 flies demonstrate its effectiveness. glc1 is a mutant of a glutamate-gated chloride channel subunit gene (DmGlula) (Kane et al. 2000). glc1 flies exhibited a behavioral defect in response to noxious heat in the two assays (see Supplemental material Fig. S1). However, their mean latency in the jump test was significantly longer than that of wild-type flies (see Supplemental material Fig. S1), suggesting that glc1 is defective in the output pathway of the jump response. Thus, the abnormal behavior of glc1 in the two assays does not prove that glc1 is indeed defective in thermal nociception.
Despite the three advantages above, our assays also have two major disadvantages. First, it is difficult to find an available method to measure the temperature of local cuticle directly because of the small body size of adult flies. In contrast, this parameter is easily measured in rats (Hargreaves et al. 1988). Secondly, the surgically manipulated flies and biz flies responded slower than wild-type flies in the HP assay (Fig. 2d,e), suggesting that innoxious thermosensation sensitizes the jump response, although thermal nociception elicits it in this assay. Similarly, one of the behavior patterns of rats in the HP test was a response to a mixture of warmth and noxious heat (Espejo & Mir 1993). Finding a way to avoid the sensitization will be a goal of our future studies.
The role of the painless gene in thermal nociception
We examined painless1 flies in the two assays and the jump test, and further tested pain-rescue;painless1 flies (Fig. 3a,b,c). The results indicated that painless is required for thermal nociception in adults, as in larvae (Tracey et al. 2003), suggesting that this gene is expressed in the neurons of the nociceptive sensory pathway. Thus, we further investigated the expression pattern of painless in adult flies.
The pain-Gal4-driven expression of UAS-EGFP showed green fluorescence in the cells where painless was expressed (Tracey et al. 2003). By this method, we found the expression of this gene at the root of the femurs (Fig. 4a, black arrows) and at the tarsal segments four and five (Fig. 4b, black arrows). This suggested that painless is expressed in the nociceptors at the ends of legs, which were in contact with the copper plate. Therefore, we hypothesized that the behavioral defect of painless1 in the HP assay may have been caused by the abnormal function of those nociceptors.
In the TGs, most of the painless-expressing fibers, projected from the peripheral tissues into the prothoracic, mesothoracic and metathoracic neuromeres (Fig. 4c, red arrows), suggested that painless is expressed in the neural pathway that sends nociceptive information from the peripheral nervous system to the TG. Additionally, we did not observe obvious fiber projecting from the TGs to the brains, suggesting that painless is not expressed in the neural pathway that sends nociceptive information from the TG to the brain. The two suggestions above, plus the complex expression pattern of painless in the TGs (Fig. 4c), imply that the TG might be a primary nerve center that processes nociceptive information coming from the peripheral nervous system.
In the brains, green fluorescence invariably was found in three groups of small cells and in the MBs (Fig. 4d), implying that MBs play an important role in thermal nociception. However, the HU-treated flies and mbm1 females, whose MBs were very incomplete, exhibited a normal nociceptive behavior (Fig. 5a,b). This indicates that intact MBs are not required for thermal nociception. The result is not surprising, because mbm1 females and HU-treated flies exhibited normal learning scores and avoidance scores in a visual learning paradigm, in which heat punishment was used (S. Tang, personal communication; de Belle & Heisenberg 1994; Tang & Guo 2001). Considering that MBs may be unimportant for thermal nociception although painless is expressed in MBs, we hypothesized that, first, the expression of painless in the TG and the peripheral nervous system, rather than its expression in MBs, might be necessary for thermal nociception; secondly, the expression of painless in MBs might participate in other functions and consequently painless might be a pleiotropic gene.
The central complex is another brain structure in adult Drosophila. Several mutants (ccdKS135, ccbKS145, cbdKS96, ebo678, eboKS263, agnX1 and nobKS49), defective in the neural structure of central complex (Strauss & Heisenberg 1993), exhibited defective behavior in a visual learning paradigm, in which heat punishment was used (Li Liu, personal communication). The evidences implied that the mutants might also be defective in sensing noxious heat. Thus, we tried the mutants and found that none of them were obviously defective in the nociceptive behavior (data not shown), except nobKS49 (Fig. 6a,b). nobKS49 has a structural defect at the protocerebral bridge, the most dorso-posterior neuropil of the central complex (Strauss et al. 1992). Important for walking and flight behavior, this structure may be a part of the highly regulatory center for locomotion in adult Drosophila (Strauss et al. 1992). In our study, the abnormal nociceptive behavior of nobKS49 suggests that the protocerebral bridge also participates in the processing of nociceptive information (Fig. 5e), although what role this structure plays in thermal nociception needs further study.
In summary, the CLB assay, the HP assay and the jump test compose a behavioral testing system to assess thermal nociception in Drosophila. We demonstrated the behavioral comparability between mammals and adult flies, suggesting that adult Drosophila is a useful model for the study of nociception. The various genetic tools available in Drosophila, together with this behavioral testing system, will facilitate the identification of novel genes and neural pathways involved in thermal nociception.