Stimulation of epidermal cells elicits nociceptive behaviors.

(A) Fraction of larvae that exhibited optogenetic-induced rolling (roll probability) using the indicated GAL4 lines to drive UAS-CsChrimson expression. All experimental genotypes, except for larvae expressing UAS-CsChrimson in C4da neurons, included elav-GAL80 to suppress neuronal GAL4 activity. Genotypes: GAL4, UAS-CsChrimson, elav-GAL80/+. (B) Roll probability of larvae following optogenetic stimulation using the indicated GAL4 lines in combination with elav-GAL80 (or tsh-GAL80 + cha-GAL80 in the case of A58- GAL4) to drive UAS-CsChrimson expression in epidermal cells. All epidermal drivers except for sr-GAL4, which is expressed in apodemes but no other epidermal cells, elicited rolling responses. Genotypes: GAL4, UAS-CsChrimson, elav-GAL80/+. (C) Roll probability of larvae following thermogenetic stimulation using the indicated GAL4 line to express the warmth (35°C) -activated UAS-TrpA1. The number of rolling larvae (out of 50) is indicated for each group. Genotypes: GAL4, UAS-TrpA1, GAL80 (as indicated)/+. Control, UAS-TrpA1/+. Sample sizes are indicated in each panel. Asterisk (*) indicates p<0.05 in this and subsequent figures. Raw data for all figures is provided in Source Data File 1 and details of statistical analyses, including tests performed, p-values, and q-values are provided in Supplementary File 1.

Stimulation of epidermal cells evokes multimodal behavioral responses. (A) Larval behaviors were scored for 10 s before, during, and after optogenetic stimulation. (B-E) Fraction of larvae exhibiting indicated behaviors over time in one second bins expressing CsChrimson in (B) epidermal cells, (C) C4da neurons, (D) C3da neurons, and (E) Cho neurons in the presence and absence of all-trans retinal (ATR). Red line indicates the presence of light stimulation. (F) The latency to the first roll of the larvae that rolled from Epi>Chrimson ATR+ and C4da>Chrimson ATR+ treatment groups (n = 14, 17, respectively). (G) The duration of indicated behaviors of the larvae that displayed those behaviors during optogenetic stimulation. (H) The fraction of larvae that exhibited indicated behaviors following removal of the light stimulus of all larvae from panels (B-E). Genotypes: GAL4, UAS-CsChrimson/+.

(A) Optogenetic activation of CsChrimson-expressing epidermal cells in the body wall triggers calcium transients in the axon terminal of GCaMP6s-expressing nociceptive SSNs. Images show responses from one representative animal. Plots depict mean GCaMP6s fluorescence intensity of the axon terminals of (B) C4da, (C) C3da, (D) Cho, and (E) C1da neurons following optogenetic activation (light stimulus, yellow box) of epidermal cells over time. Solid lines depict mean GCaMP6s fluorescence across replicates (n=15 larval fillet preparations), shading indicates SEM, red traces are GAL4+ ATR+, blue traces are GAL4+ ATR-, black trace is GAL4- ATR+. (F) The fraction of larvae exhibiting indicated behaviors during optogenetic epidermal stimulation in combination with SSN silencing via Tetanus Toxin (TnT) expression. We note that although baseline rolling probability is elevated in all genetic backgrounds containing the AOP-LexA-TnT insertion, silencing C4da and C3da neurons significantly attenuates responses to epidermal stimulation. (G) The duration of the behavioral responses during optogenetic epidermal stimulation with neuronal TnT expression. Genotypes: (A-B) R27H06-LexA (C4da neurons), AOP-GCaMP6s, UAS-CsChrimson/+; R38F11-GAL4/+ or R27H06-LexA (C4da neurons), AOP-GCaMP6s, UAS-CsChrimson/+ (GAL4-ATR- effector-only control); (C) AOP-GCaMP6s, UAS-CsChrimson/+; R38F11-GAL4/NompC- LexA (C3da neurons); (D) UAS-GCaMP6s, AOP-CsChrimson, R61D08-GAL4 (Cho neurons)/R38F11-LexA; (E) UAS-GCaMP6s, AOP-CsChrimson, R11F05-GAL4 (C1da neurons)/R38F11-LexA; (F-G) R38F11-GAL4, UAS-CsChrimson, AOP-LexA-TnT/+ (Epi>CsChrimson); R38F11-GAL4, UAS-CsChrimson, AOP-LexA-TnT/ppk-LexA (Epi>CsChrimson + C4da>TnT); R38F11-GAL4, UAS-CsChrimson, AOP-LexA- TnT/NompC-LexA (Epi>CsChrimson + C3da>TnT).

Epidermal stimulation augments nociceptive responses. (A) Mean GCaMP6s responses (F/F0) in C4da axons during optogenetic stimulation (yellow box) of C4da neurons alone (green) or of C4da neurons and epidermal cells (magenta), shading indicates SEM. (B) Simultaneous epidermal stimulation increased the peak calcium response (Fmax/F0), (C) total calcium influx (area under the curve), and (D) duration of C4da neuron calcium responses compared to stimulation of C4da neurons alone. Genotypes: ppk-LexA, AOP-GCAMP6s/+; R27H06-GAL4/UAS-CsChrimson (C4da) and ppk-LexA, AOP-GCAMP6s/+; R27H06-GAL4/R38F11-GAL4, UAS-CsChrimson (C4da+epi). (E-J) Characterization of the behavioral responses to low-intensity optogenetic stimulation of C4da neurons, epidermal cells, or simultaneous C4da neurons and epidermal cells. (E) Cumulative and (F) total roll probability during optogenetic stimulation (indicated by the red bar). n = 33 (C4da>CsChrimson), 30 (Epi>CsChrimson), and 31 (C4da + Epi>CsChrimson) larvae. (G, H) Number and frequency distribution of rolls, (I) latency to the first roll observed for larvae of the indicated genotypes, and (J) the duration of the indicated behaviors during light stimulus. Genotypes: UAS-CsChrimson/+; R27H06-GAL4/+ (C4da), UAS- CsChrimson/+; R38F11-GAL4/+ (Epidermis), UAS-CsChrimson/+; R27H06- GAL4/R38F11-GAL4 (C4da+Epidermis). (K) Roll probability of larvae to a 20 mN or 50 mN von Frey mechanical stimulus and epidermal optogenetic activation (a light stimulus, 1.16 μW/mm2 that was insufficient on its own to induce nocifensive rolling). Larvae were reared in the presence or absence of ATR, as indicated. Genotypes: UAS- CsChrimson/+; R38F11-GAL4/+. (L-N) Prior epidermal but not nociceptor stimulus potentiates mechanical nociceptive responses. (L) Roll probability of control larvae (UAS-TrpA1/+) or larvae expressing TrpA1 in the epidermis (Epi-GAL4: R38F11-GAL4) or C4da neurons (UAS-TrpA1/+; C4da-GAL4 #1: R27H06-GAL4, UAS-TrpA1/+; C4da- GAL4 #2: ppk-GAL4, UAS-TrpA1/+), or control larvae (no GAL4: UAS-TrpA1/+;) in response to 40 mN mechanical stimulus 10 s following 10 s of a thermal stimulus (25° or 32° C). To control for effects of genetic background, we confirmed that each of the experimental genotypes exhibited mechanically induced nociceptive sensitization (Fig. 4S1C). (M) Roll probability of control larvae (UAS-TrpA1/+) or larvae expressing TrpA1 in the epidermis (Epi>TrpA1: R38F11-GAL4, UAS-TrpA1/+) in response to a 40 mN mechanical stimulus delivered at the indicated time interval following a 32° C thermal stimulus. (N) Nociceptive enhancement (difference in the roll probablity to the first and second stimulus) is plotted against the recovery duration and results were fit to an exponential curve to derive the decay time constant. The red line indicates nociceptive enhancement of a mechanical stimulus by a prior epidermal thermogenetic stimulus; the black line indicates nociceptive enhancement by a prior mechanical stimulus.

Epidermal cells are intrinsically mechanosensitive. (A) Schematic of preparation to measure radial stretch evoked calcium responses of dissociated epidermal cells. (B) Representative calcium responses of a dissociated epidermal cell to 0.5% and 1% radial stretch (successive stimuli), 2.5% radial stretch, and 5% radial stretch. (C) Dose response curve displaying the fraction of epidermal cells activated by increasing magnitudes of stretch. Red trace displays the mean ± SEM across six independent dissociated cell preparations, obtained from a minimum of 6 larvae. Gray traces display fraction responding in each dissociated cell preparation replicate. (D) Subsets of epidermal cells display varying stretch thresholds, n = 6 dissociated cell preparations, for a total of 654 epidermal cells. (E) Representative mechanically induced epidermal calcium responses in the larval body wall. Images show GCaMP6s fluorescence intensity 100 ms prior to (i) and 20 s following (ii) a 25 μm membrane displacement (poke). (F) Distribution of the peak calcium response (Fmax/F0) to a 25 μm membrane displacement (poke) of 24 cells from 24 independent larval fillets. Cells were classified as responders (>10% increase in normalized GCaMP6s fluorescence). (G) Mean calcium responses (F/F0) of poke responders and non-responders (n = 12 cells each). Solid lines depict mean normalized GCaMP6s fluorescence and shading indicates SEM. Genotype: R38F11-GAL4, UAS-GCaMP6s.

CRAC channels are required for epidermal mechanosensory responses and epidermal nociceptive potentiation. (A) RNAi screen for epidermal ion channels required for mechanically induced nociceptive potentiation. Bars depict nociceptive potentiation index (difference in the larval roll probability to the first and second mechanical stimuli divided by roll probability to the first mechanical stimulus). Candidate channels were chosen for further analysis if they had a z-score greater than 2 (absolute value). (B) The CRAC channels Orai and Stim are required in epidermal cells for mechanically evoked nociceptive potentiation. Roll probability of larvae of the indicated genotypes (Control RNAi, R38F11-GAL4, UAS-RFP-RNAi/+; Stim RNAi, R38F11-GAL4, UAS-Stim-RNAi/+; Orai RNAi, R38F11-GAL4, UAS-Orai-RNAi/+) to a 40 mN mechanical stimulus followed by a second 40mN mechanical stimulus 10 s later. (C) Drosophila epidermal cells display classical store-operated calcium entry (SOCE). Treatment with the drug thapsigargin (TG) in the absence of extracellular calcium promoted depletion of intracellular calcium stores and calcium influx, following extracellular calcium re- addition. (D) Like TG, 1% stretch in the absence of extracellular calcium induced depletion of intracellular calcium stores and calcium influx, following extracellular calcium re-entry. (E) 69% of stretch responsive cells displayed greater calcium influx during intracellular calcium stores release than during the calcium re-entry phase. (F-G) The Orai blocker, lanthanum chloride (500 nM) or the depletion of intracellular stores by thapsigargin (1 µM) reduces the fraction of stretch-sensitive epidermal cells. (H-I) The fraction of stretch-sensitive epidermal cells is significantly decreased in cells isolated from larvae expressing Stim RNAi, or Orai RNAi, as compared to control RNAi. (J) Stretch stimuli evoke dose-dependent calcium signals in the human keratinocyte HaCaT cell line. (K) Representative stretch evoked SOCE calcium response in HaCaT cells. Stretch induces calcium release from stores in the absence of extracellular calcium and a greater calcium influx in the presence of extracellular calcium. (L) Epidermal hyperpolarization enhances mechanical nocifensive responses. Roll probability of larvae expressing GtACR in epidermal cells (R38F11-GAL4, UAS-GtACR/+) or control larvae (R38F11-GAL4/+) to a single 70 mN mechanical stimulus. (M) Epidermal Stim overexpression enhances mechanical nocifensive responses. Roll probability of Stim- overexpressing larvae (R38F11-GAL4, UAS-Stim/+) and control larvae (R38F11- GAL4/+) to two successive 40 mN mechanical stimuli delivered 10 s apart. (N) Epidermal potentiation of mechanical nociceptive responses requires exocytosis. Roll probability of control larvae (UAS-shits/+) or larvae expressing temperature-sensitive dominant-negative shi in epidermal cells (R38F11-GAL4, UAS-shits/+) in response to two successive mechanical stimuli that followed 10 min of conditioning at the permissive (25° C) or non-permissive (30° C) temperature. (O) Model of epidermal-neuronal signaling. Mechanically evoked Stim/Orai calcium signaling in epidermal cells drives calcium influx and vesicle release that drives nociceptor activation and mechanical sensitization via activation of C4da nociceptors.