Cooling Relief of Acute and Chronic Itch Requires TRPM8 Channels and Neurons

Cooling inhibits pruritogen-induced behaviors

Next, we asked whether cooling inhibits pruritogen-evoked behaviors by examining licks/bites induced by either Cqx (Figure 2a), compound 48/80 (Figure 2b), or 100 μg histamine (Figure 2c) in mice placed on a surface set to a range of temperatures. At the preferred temperature of 30°C (Lee et al., 2005; Moqrich et al., 2005), robust licking/biting was observed with each pruritogen, responses that were not statistically different at 24°C (Figure 2a-​-2c2c;P > 0.05). However, at 20°C and below, licking/biting evoked by Cqx was essentially abolished (Figure 2a, P < 0.001 compared to 30°C). Similarly, compound 48/80 responses were significantly reduced at 20°C (Figure 2b, P < 0.01) with a further reduction as the surface was lowered into the noxious cold range (10°C, P < 0.001). Flinches and lifts were not observed at any temperature tested with Cqx (Figure 2d) or compound 48/80 (Figure 2e), demonstrating that licking/biting is the predominant behavior to these stimuli.

Open in a separate window

Figure 2.

Cooling inhibits acute itch.

The duration of licking and biting behaviors at different plate temperatures (30, 24, 20, 17, and 10°C) in mice injected with chloroquine (Cqx) (a, n = 8–10), 48/80 (b, n = 8), or histamine (c, n = 9). Neither Cqx (d) or 48/80 (e) induced flinches or lifts, whereas 100 μg histamine (f) induced robust pain-like behaviors that were inhibited at plate temperatures of 17°C and 10°C. (g–i) Duration of all behaviors showed cooling inhibition to all three substances tested. ***P < 0.001, **P < 0.01 and all data compared to 30°C via one-way ANOVA, Neuman-Keuls post-test.

Histamine (100 μg) also evoked robust licking/biting (Figure 2c) at control temperatures, behaviors that, unlike Cqx and compound 48/80, were not reduced at 20°C (P > 0.05). However, at 17°C and below, these behaviors were reduced significantly (Figure 2c, P < 0.001). Of note, significant flinching/lifting was observed at this dose of histamine, yet these responses were also reduced at 17°C and 10°C (Figure 2f; P < 0.01). Thus, even as this dose of histamine is likely producing pain, cooling is inhibitory, results consistent with our prior analysis of cooling analgesia (Knowlton et al., 2013).

Lastly, analysis of all behaviors (flinches, lifts, licks, bites) showed that cooling can block these responses (Figure 2g-​-2i),2i), data that largely mirror the findings when just licking and biting was quantified (Figure 2a-​-2c).2c). Taken together, these data show that cooling, in addition to being analgesic, also alleviates pruritic behaviors, suggesting that regardless of whether these behaviors are itch or pain, the use of hind paw is a suitable method to determine the underlying mechanisms of antipruritic cooling.

Cooling delays pruritogen-induced itch

Pruritogen-evoked behaviors resolve within 30 minutes post-injection in the cheek and nape assays (LaMotte et al., 2011), and we found this to be true in the hind paw of mice injected with Cqx or the serotonin receptor agonist αMe5-HT (30 μg) at 24°C (Figure 3a, ​,3c).3c). Human studies show that cooling-induced itch relief requires constant cooling and that itch sensations will return upon rewarming of the skin (Fruhstorfer et al., 1986). This is also the case in mice, as Cqx-behaviors inhibited at 20°C (Figure 3b; P < 0.001 compared to mice at 24°C after 15 minutes) return when the surface was warmed to 24°C (Figure 3a). In vehicle-injected mice exposed to the 20°C to 24°C temperature shift, a slight increase in attention to the hind paw was observed upon warming, but clearly not to the extent observed if Cqx-injected animals. Moreover, total behaviors were similar in mice at 24°C for the entirety of the recording period compared to those placed at 20°C initially, then raised to 24°C after 15 minutes (Figure 3b, P > 0.05).

Open in a separate window

Figure 3.

Antipruritic cooling requires constant stimulation.

(a) Chloroquine (Cqx; 200 μg) induced licking/biting behaviors resolved within 25 minutes post injection when the test plated was held at 24°C (n = 7). A plate temperature of 20°C showed inhibition with robust behaviors produced when the plate was warmed to 24°C 15 minutes after Cqx injections, responses not observed in vehicle-injected mice. (b) Summary of the data in (a) shows inhibition in the first 15-minute period post-injection when the plate is at 20°C (***P < 0.001 compared to 24°C, one-way ANOVA, Neuman-Keuls post-test), with behaviors uncovered in the second 15–45 minute period when the plate was warmed to 24°C (*P < 0.05 compared to when the plate was held constant at 24°C, one-way ANOVA, Neuman-Keuls post-test). Summary of the total behaviors showed no difference between the two stimulation protocols (P>0.05, one-way ANOVA, Neuman-Keuls post-test). (c) In a test of the duration in which cooling inhibits itch, α-methyl 5-HT (aMe5-HT; 30 μg) behaviors were blocked by cooling to 17°C, but were observed when the plate was warmed to 24°C after 30 minutes. (d) Summary of the αMe5-HT responses in mice on a 17°C plate for 30 minutes followed by warming to 24°C compared to mice placed at 24°C only. *P < 0.05, ^nsP> 0.05, n = 6–8, one-way ANOVA, Neuman-Keuls post-test.

Next, we determined whether increasing the duration of cooling was more effective in inhibiting pruritogen-induced behaviors and whether such delay in behaviors occurs with other pruritogens, in this case with aMe5-HT. As with Cqx, animals robustly attend to their injected hind paw when the plate was at 24°C, behaviors which resolved by 20 minutes post-injection (Figure 3c). When μMe5-HT–injected animals were placed on a 17°C plate, these behaviors were inhibited for the 30-minute test duration, but were observed immediately after the plate was warmed to 24°C (Figure 3c). Furthermore, the summed responses of these mice were not significantly different from αMe5-HT–injected mice placed at 24°C immediately after injection (Figure 3d; P > 0.05). Thus, consistent with human studies, antipruritic cooling requires constant stimulation, and pruritogen-evoked behaviors return when the skin is warmed to ambient temperatures (Fruhstorfer et al., 1986).

Cooling antipruritus is TRPM8-channel and TRPM8-neuron – dependent

Next, we examined the molecular and cellular basis for the effect of cooling on itch. TRPM8 is the principal detector of cold temperatures in the mammalian peripheral nervous system (McKemy et al., 2002;Peier et al., 2002). With the knowledge that cooling can alleviate pruritogen-induced behaviors in wild-type mice, we next asked whether this effect of cooling was dependent on functioning TRPM8 ion channels. Wild-type and TRPM8 nulls (Trpm8^−/−) (Bautista et al., 2007; Knowlton et al., 2013) were injected with Cqx and then subjected to a range of cold temperatures (Figure 4). As mentioned, wild-type mice showed robust responses at 24°C, behaviors that were significantly reduced at 20°C and 17°C (Figure 4a; P < 0.001). Similarly, Cqx evoked robust behaviors in Trpm8^−/− mice at 24°C, responses that were not different than wild-types (P > 0.05 wild-type vs. Trpm8^−/− at 24°C). However, unlike wild-type animals, Cqx-evoked behaviors were not attenuated at 20°C in Trpm8^−/− mice (Figure 4b; P > 0.05), demonstrating that TRPM8 channels are required at this temperature. However, these behaviors were reduced in Trpm8^−/− mice when the plate was cooled to near the noxious cold threshold (17°C, P < 0.001 compared to Trpm8^−/− at 24°C and P > 0.05 compared to wild-type at 17°C). These data suggest that TRPM8 is required for inhibition of Cqx-induced behaviors with mild cooling, but that the channel is not needed at near-noxious temperatures.

Open in a separate window

Figure 4.

Antipruritic cooling requires TRPM8 channels and neurons.

(a) Chloroquine (Cqx)-induced licking/biting behaviors were inhibited in wild-type mice (n = 8–10) at 20°C and 17°C (***p < 0.001 compared to 24°C, one-way ANOVA, Neuman-Keuls post-test). (b) In mice lacking functional TRPM8 channels (Trpm8^−/−, n = 8–9), 20°C was ineffective in blocking these behaviors (^nsP > 0.05, one-way ANOVA, Neuman-Keuls post-test), while a colder temperature of 17°C did inhibit (**P < 0.01 compare to both 24° and 20°C, one-way ANOVA, Neuman-Keuls post-test). (c) In comparison, mice in which TRPM8-expressing afferent neurons are ablated, cooling was unable to inhibit Cqx-induce behaviors, even at 10°C (^nsP > 0.05 compared to 24°C for all temperatures tested, one-way ANOVA, Neuman-Keuls post-test, n = 8).

These results raise several hypotheses, including that cold is inhibiting nerve conduction in pruriceptive afferents (unlikely at 20°C), or that other cold transduction mechanisms are involved in antipruritic cooling at noxious cold temperatures. We and others have shown that TRPM8 neurons represent a labeled line for cold and mediate all aspects of cold perception, even in the absence of TRPM8 channels (Knowlton et al., 2013; Pogorzala et al., 2013). Therefore, to test whether block by noxious cold requires TRPM8 neurons, we repeated the assays described here using our Trpm8^DTR mouse line in which TRPM8⁺ neurons are ablated in the adult mouse via genetically targeted expression of the cell-death–promoting human diphtheria toxin receptor (Knowlton et al., 2013). As with Trpm8^−/− mice, we observed robust behaviors at 24°C and 20°C in TRPM8-neuron–ablated animals, but in contrast Cqx still evoked behaviors at 17°C (Figure 4c, P > 0.05 compared to 24°C). As 17°C is a suprathreshold noxious cold temperature, we tested Cqx responses at subthreshold temperature of 10°C, also observing robust behaviors that were not significantly different than those at 24°C (P > 0.05 compared to 24°C). While there was a downward trend as temperatures decreased, there was no significant difference between all four temperatures tested (P > 0.05, repeated-measures of ANOVA), showing that TRPM8 neurons are critical for antipruritic cooling.

Chemical cooling requires TRPM8 channels

TRPM8 channels are also activated by the cold mimetic menthol, which is also a TRPM8-dependent analgesic (Proudfoot et al., 2006). Menthol and other TRPM8 agonists have been reported to alleviate severe pruritus in several studies (Andersen et al., 2017; Frolich et al., 2009; Stander et al., 2017), with recent evidence that an over-the-counter topical menthol solution inhibits Cqx-induced itch in mice (Kardon et al., 2014). However, the TRPM8-dependence of this inhibition has not been established. Thus, to confirm that its effects are mediated by TRPM8, we used the cheek assay in wild-type and Trpm8^−/− mice (Shimada and LaMotte, 2008). First, to establish that the topical menthol product (8% solution, Stopain; Troy Healthcare LLC, Hazelton, PA) used before to block itch (Kardon et al., 2014) stimulates sensory afferents in the cheek in a TRPM8-dependent manner, we applied it to the cheek and measured wipes, an indicator of irritation and not itch. As expected, robust wiping was induced by this solution in wild-type mice (P < 0.001 vs. control, Figure 5a), responses that were significantly reduced in Trpm8^−/− animals (P < 0.001). However, there was a slight increase in Stopain responses in Trpm8^−/− mice (P < 0.05 compared to vehicle), consistent with non-TRPM8–dependent effects that have been attributed to menthol when used at high concentrations in vivo (Liu et al., 2013). Thus, while TRPM8 channels are the predominant mediator of responses to menthol, additional factors lead to behaviors induced by high-dose menthol in vivo.

Open in a separate window

Figure 5.

Menthol inhibits Cqx-induced itch in a TRPM8-dependent manner.

(a) Topical menthol (8%) produced robust cheek wiping behaviors in wild-type mice (***P < 0.001 compared to vehicle, one-way ANOVA, Neuman-Keuls post-test, n = 8) that were significantly reduced in Trpm8^−/− animals (***P < 0.001 compared to vehicle, one-way ANOVA, Neuman-Keuls post-test, n = 6–8). Menthol did evoke significant wiping compared to vehicle (*P < 0.05, one-way ANOVA, Neuman-Keuls post-test, n = 6). (b) Topical menthol (8%) significantly inhibited Cqx-induced cheek scratches in wild-type (*P < 0.05, one-way ANOVA, Neuman-Keuls post-test, n = 8), an effect that was not observed in Trpm8^−/− animals (^nsP > 0.05, one-way ANOVA, Neuman-Keuls post-test, n = 6–8). Cqx, chloroquine; PBS, phosphate buffered saline.

Next, to determine whether the effects of topical menthol on itch reported previously were TRPM8-dependent, we applied Stopain 10 minutes before an intradermal injection of Cqx (50 μg) into the cheek and measured the number of hind-paw scratches directed at the injected cheek in wild-type and Trpm8^−/− mice (Figure 5b). As predicted, Cqx induced a significant number of scratches, responses that were significantly reduced with prior application of the menthol solution (P < 0.05);results similar to those reported previously (Kardon et al., 2014). Cqx also induced scratching in Trpm8^−/− mice, but these responses were not significantly affected by treatment with the menthol solution (Figure 5b, P > 0.05). Thus, similar to cooling, cold mimetics can alleviate itch and do so in a manner that is largely dependent on TRPM8 channels.

Animals

All experiments were performed with adult C57BL6/J mice between the ages of 8 and 14 weeks. Wild-type, TRPM8-knockout mice, and TRPM8-DTR mice were on the same C57Bl/6 genetic background, and all animals were provided standard mouse chow and water ad libitum. All procedures and tests were approved by the University of Southern California Institutional Animal Care and Use Committee and conducted in accordance with the recommendations of the International Association for the Study of Pain and the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Ablation of TRPM8 neurons was performed and validated as described previously (Knowlton et al., 2013).

Behavior tests

Dry skin model of chronic itch

Animals were habituated to the cold plate 1 h/day for 2 days before recording at both the beginning and end of the experiment. Behavior was recorded on day 0 and day 15 for 30 minutes at each of the following temperatures: 30°C, 20°C, and 30°C again. From day 1 to day 15, mice were treated with 1:1 acetone-ether for 30 seconds, followed by water for 15 seconds on one hind paw. Mice in the control group were treated only with water. Treatments were administered twice daily 8 hours apart.

The authors wish to thank members of the McKemy Laboratory for their assistance in the completion of this work. We thank Diana Bautista, Qiufu Ma, and Wendy Knowlton for their guidance and advice in establishing the itch behavioral paradigms in this study. The present study was funded by grant NS087542 from the National Institutes of Health.