Cytokine signaling mediates UV-induced nociceptive sensitization in Drosophila larvae

Abstract

Background: Heightened nociceptive (pain) sensitivity is an adaptive response to tissue damage and serves to protect the site of injury. Multiple mediators of nociceptive sensitization have been identified in vertebrates, but the complexity of the vertebrate nervous system and tissue-repair responses has hindered identification of the precise roles of these factors.

Results: Here we establish a new model of nociceptive sensitization in Drosophila larvae, in which UV-induced tissue damage alters an aversive withdrawal behavior. We find that UV-treated larvae develop both thermal hyperalgesia, manifested as an exaggerated response to noxious thermal stimuli, and thermal allodynia, a responsiveness to subthreshold thermal stimuli that are not normally perceived as noxious. Allodynia is dependent upon a tumor necrosis factor (TNF) homolog, Eiger, released from apoptotic epidermal cells, and the TNF receptor, Wengen, expressed on nociceptive sensory neurons.

Conclusions: These results demonstrate that cytokine-mediated nociceptive sensitization is conserved across animal phyla and set the stage for a sophisticated genetic dissection of the cellular and molecular alterations responsible for development of nociceptive sensitization in sensory neurons.

Figure 1. Larval responses to thermal stimuli and UV radiation

(A) Diagram of thermal probe illustrating the thermal control unit, feedback design (arrows) and brass tip. Box shows area of interest in B. (B) Schematic illustrating the major components of the heat probe. (C) Aversive response of third instar w¹¹¹⁸ larvae to stimulation at various temperatures. Behaviors were classified as “no response” (white, > 20 s), “slow withdrawal” (gray, between 5 and 20 s) or “fast withdrawal” (black, < 5 s). n = 50 for each temperature tested. (D) Average withdrawal latency at different temperatures. 44 °C = 9.6 s +/− 1.6 s and 48 °C = 1.5 s +/− 0.6 s. n = 50 for each temperature. Error bars indicate standard error of the mean. The median withdrawal time was significantly different between groups (p < 0.001, by Log-rank test). (E) Survival rate of third instar larvae to adulthood after treatment with increasing doses of ultraviolet radiation. n = 75 for each group.

Figure 2. Epidermal damage induced by UV radiation

(A-P) Wholemount staining showing the cellular effects of mock treatment (A-D) or UV treatment on larval tissues at 4 hours (E-H) 24 hours (I-L) or 48 hours (M-P) post-exposure. (A, E, I, M) Epidermal cell membranes of w¹¹¹⁸;ppk1.9-Gal4, UAS-eGFP larvae labeled with anti-Fasciclin III. (B, F, J, N) Class IV dendritic arborization sensory neurons labeled in the same ppk1.9-Gal4, UAS-eGFP larvae as in A, E, and I. (C, G, K, O) X-Gal staining of msn-lacZ larvae. (D, H, L, P) Apoptotic cells of w¹¹¹⁸ larvae labeled with an activated caspase-3 antibody. Scale bar in P, 100 μm for all panels.

Figure 3. Thermal allodynia and hyperalgesia after tissue damage

(A-B) Comparison of response thresholds between UV-treated and mock treated w¹¹¹⁸ larvae. (A), Response to the highest normally innocuous temperature (38 °C, see Figure 1B) at specified times after UV-induced tissue damage. White = no response, gray = response between 5 and 20 s, black = response in less than 5 s. Nociceptive threshold distributions are significantly different (p < 0.001, by Fisher’s exact test), at 8, 16, and 24 hours post-UV as compared to mock-treated larvae (asterisks). Mock-treated larvae were assessed 24 hours after mock irradiation, to compare with the greatest response seen in UV-treated larvae. (B) Responses to decreasing temperatures below the normal nociceptive threshold 24 hours after tissue damage. Gray = mock treated, hatched = UV treated. (C-D) Comparison of withdrawal latencies between UV-treated larvae and mock treated w¹¹¹⁸ larvae. (C) Response to a normally noxious temperature (45 °C) at specified times after UV-induced tissue damage. White = no response, gray = response between 5 and 20 s, black = response in less than 5 s. Nociceptive threshold classification is significantly different at 8 (p < 0.001) and 16 (p = 0.039) hours post-UV as compared to mock-treated larvae (asterisks). Mock-treated larvae were assessed 8 hours after mock irradiation, to compare with the greatest response seen in UV-treated larvae. (D) Average withdrawal latency to 45 °C stimulation at various times after tissue damage. Error bars indicate standard error of the mean. Mean withdrawal was significantly different (p < 0.001, by Student’s t-test) at 8 hours, as compared to mock-treated larvae (asterisks). n = 50 for all conditions.

Figure 4. Apoptosis in epidermal cells is required for thermal allodynia

(A-I) Wholemount staining of w¹¹¹⁸; A58-Gal4/+ (A-C), w¹¹¹⁸;UAS-dronc^IR/+ (D-F) and w¹¹¹⁸;UAS-dronc^IR/+;A58-Gal4/+ larvae (G-I). (A,D,G) Epidermal cell membranes (green) labeled with anti-fasciclin III. (B,E,H) Apoptotic cells (red) labeled with an activated caspase-3 antibody. (C,F,I) Merged panels. Bar in I, 100 μm for A-I. (J) Response to the highest normally innocuous temperature (38 °C) 24 hours post-UV. Larvae bearing both the A58-Gal4 and UAS-Dronc^IR inserts are significantly less sensitive (p < 0.001, by Fisher’s exact test) than the parental control strains when tested for thermal allodynia. (K) Response to a normally noxious temperature (45 °C, see figure 1C) 8 hours post-UV. Larvae bearing both the A58-Gal4 and UAS-Dronc^IR inserts displayed no statistically significant differences from the parental control lines or w¹¹¹⁸ control line (Figure 3D) when tested for thermal hyperalgesia (p = 0.659, by one-way ANOVA). (L-M) Response to the highest normally innocuous temperature (38 °C) 24 hours post-UV (L) or a normally noxious temperature (45 °C) 8 hours post-UV (M). Larvae lacking hemocytes (Pxn-Gal4>UAS-hid) displayed no differences from the parental control lines when tested for thermal allodynia or hyperalgesia. n = 30 for each group in J through M. Error bars indicate standard error of the mean.

Figure 5. Thermal allodynia is induced by TNF signaling

(A) Response to the highest normally innocuous temperature (38 °C) of eiger mutants or larvae lacking either Eiger, or Wengen in epidermal cells (A58-Gal4) or in nociceptive sensory neurons (ppk1.9-Gal4), 24 hours after UV treatment. (B) Response to a normally innocuous temperature (38 °C) of control larvae and of larvae with ectopic expression of Eiger in class IV sensory neurons in the absence of UV treatment. White = no response within 20 s, gray = response between 5 and 20 s, black = response in less than 5 s. n = 30 for each condition. Asterisk, p < 0.001 compared to wild type by Fisher’s exact test. (CE) Whole-mount staining of egr¹/egr³ larvae. (C) Epidermal cell membranes (green) labeled with anti-fasciclin III. (D) Apoptotic cells (red) labeled with an activated caspase-3 antibody. (E) Merged panel. (F) Response to innocuous (38 °C) and noxious (48 °C) temperatures in the absence of tissue damage. Both eiger mutants and control larvae do not respond within 20 seconds to normally subthreshold temperatures (white bars). However, all larvae within all groups rapidly withdraw from noxious stimuli (black bars), suggesting that normal nociception is not impaired in eiger mutants. Epidermal knockdown of Eiger and nociceptive sensory neuron knockdown of Wengen also did not affect normal nociception. Error bars indicate standard error of the mean. n = 25 for each condition. Bar in E, 100 μm for C-E.

Figure 6. Model

Top-down illustration depicting the relationship between UV-induced epidermal damage and changes in underlying sensory neurons. Ultraviolet light is absorbed by epidermal cells (1), which eventually undergo apoptosis/activate Dronc (2). As a result, Eiger is produced (3) and released from epidermal cells (4) and ultimately activates Wengen on the nociceptive sensory neuron membrane (5). This activation leads to intracellular signaling that sensitizes the nociceptor (6) to normally subthreshold stimuli. VNC = Ventral Nerve Cord.