Introduction

Itch is an unpleasant sensation that produces the desire to scratch [1]. Moreover, scratching relieves the itch, at least temporarily. This behavior—scratching in response to an aversive stimulus—is highly conserved across species. All mammals and birds show itch behavior. Fish will rub themselves against a rough surface to remove an irritant. Even fruit flies display site-directed grooming behavior in response to an irritating substance [2]. The conservation of this behavioral response across the animal kingdom underscores the idea that itch has an important protective function: itch triggers a behavior that removes harmful agents from the body’s surface in the short term, and because itch feels unpleasant, an organism learns to avoid itch-inducing situations in the long term.

However, there are many pathological conditions for which itch is no longer protective. In such cases, itch becomes a chronic condition that significantly decreases the quality of life. Numerous diseases are associated with severe itch, including atopic dermatitis, liver disease, postherpetic itch, small fiber neuropathies, and even some cancers [3]. Pathological itch from these and other diseases can be just as debilitating as pain, and there is a great unmet need for new treatment options [4].

In many ways, itch is like pain—both are aversive sensations that evolved to protect us, and both are initially detected by similar (though likely distinct) subsets of primary afferents [5]. Indeed, the transformation from an aversive somatosensory stimulus in the periphery to the sensory percept of either pain or itch in the brain appears to involve parallel neural pathways comprised of peripheral neurons that detect pruritogens, the superficial dorsal horn of the spinal cord, the parabrachial nucleus, the dorsal raphe, the thalamus, and several regions of the cortex. Thus, functional magnetic resonance imaging performed on the brain of a person experiencing either pain or itch might be almost indistinguishable [6]. Indeed, we do not yet understand how itch and pain are differentially encoded in the nervous system. Nevertheless, we know that they must be because itch and pain feel qualitatively different and elicit distinct behavioral responses.

Which Sensory Neurons Signal Itch?

There is great interest in determining which sensory neurons mediate itch and how these afferents are different than those that are involved in signaling pain. Identifying such fibers has important therapeutic implications because this knowledge could potentially allow the development of selective therapies for pruritus. Several years ago, microneurography experiments in humans led to the long-awaited discovery of itch-specific sensory neurons—these fibers responded to histamine, and their activity corresponded to the sensation of itch [7]. However, there was no molecular marker that allowed us to identify these cells. (Histamine responsiveness on its own is insufficient to identify an itch fiber because histamine activates numerous sensory neurons, likely including those that are involved in signaling pain). Thus, although there was good evidence for the existence of itch-selective sensory afferents, their identity remained elusive.

In the last few years, several groups have reported the discovery of markers that define itch-mediating sensory afferents, including gastrin releasing peptide (GRP), Mas-related G-protein receptor A3 (MrgprA3), and Naturietic Polypeptide b (Nppb) [8–11]. But these reports are somewhat contradictory, and so the identity of itch fibers remains controversial. Both GRP and Nppb are attractive candidates as potential markers of itch afferents because both of these peptides cause itch when injected intrathecally [9,10]. Thus, it stands to reason that primary afferents that release these peptides might be involved in itch. However, the idea that GRP is a marker of itch neurons has come under scrutiny of late. The analysis of GRP expression in dorsal root ganglia (DRG) using antibodies originally suggested that GRP is expressed in subsets of primary sensory afferents [9]. Yet, it has been difficult to corroborate these findings with methods that detect GRP mRNA, and the inability to detect GRP message in DRG sensory neurons has now called the specificity of the GRP antibody into question [12], though this subject remains controversial [13]. Rather, GRP mRNA is found in abundance in spinal interneurons, and so it seems likely that these spinal interneurons (rather than DRG neurons) are the main source of endogenous GRP in the spinal cord. Nppb, on the other hand, remains a potential marker for itch neurons. Nppb mRNA is clearly expressed in subsets of primary afferents, and Nppb is required for itch sensation [10]. What remains to be determined is the degree to which Nppb is a bona fide marker of itch neurons—in other words, one that marks all itch afferents and only itch afferents.

The strongest evidence for a marker that defines a specific population of sensory afferents that are tuned for itch comes from the analysis of a small population of DRG neurons that coexpress the MrgprA3 and MrgprC11. These related GPCRs were originally discovered in a screen for genes that are selectively expressed in peptidergic afferents [14]. Subsequently it was found that these two proteins are receptors for pruritogens. Specifically, MrgprA3 is a receptor for chloroquine, whereas MrgprC11 is activated by Bam8-22 and SLIGRL [15,16]. Because MrgprA3 and MrgprC11 are activated by pruritogens, it was logical to infer that the neurons that express these receptors are itch afferents. However, direct evidence that MrgprA3/C11-expressing neurons mediate itch was only possible on the genetic targeting of this population. The selective expression of cre recombinase in these cells made it possible to visualize this population clearly and enabled loss- and gain-of-function (LOF and GOF) studies to rigorously test the role of these afferents in itch. Importantly, the ablation of the MrgprA3 population with diphtheria toxin significantly reduced (although did not completely eliminate) itch, and selective activation of the MrgprA3 population resulted in scratching behavior [11]. The human counterpart to these receptors is thought to be MrgprX1, which also responds to a broad array of pruritogens when ectopically expressed, but whether this gene is a marker for itch-mediating neurons in humans is unknown [15,17].

The genetic labeling of the MrgprA3-expressing population of sensory neurons in a mouse was an important advance because it allowed us to see itch fibers for the first time [11]. Intriguingly, this population of neurons coexpresses CGRP and IB4, two mostly nonoverlapping markers that are thought to define distinct subsets of C fibers, the so-called peptidergic and nonpeptidergic classes, respectively. Furthermore, the central terminals of the MrgprA3 population ramify within a very narrow band of lamina II within the superficial spinal cord (newly termed II middle) that is between the peptidergic and nonpeptidergic layers. Hence the identification of MrgprA3-expressing itch afferents has challenged the conventional classification scheme for sensory neurons by revealing the existence of yet a third subtype that mediates itch and shows intermediate properties with respect to neurochemical expression as well as the laminar distribution of its central terminals. Importantly, the peripheral targeting of these neurons is entirely consistent with a role for these cells in itch: these neurons exclusively target structures in which we feel itch, such as the skin, but not any other regions of the body such as muscle or internal organs. Furthermore, within the skin, these neurons show an extremely superficial pattern of innervation, with terminations in the stratum granulosum. The loss- and gain-of function studies along with the specialized distribution of MrgprA3-expressing cells provide very compelling evidence that MrgprA3-expressing neurons mediate itch. Nonetheless, it should be noted that these fibers likely represent just one of several subpopulations of itch afferents [13,18].

Activating TRP Channels to Block Itch

Although we are still sorting out identity of all of the subtypes of primary sensory afferents that mediate itch, there is good evidence that these fibers belong to a larger population of sensory neurons that express TRPV1. This idea implies that, although TRPV1-targeted therapies may not be specific, they should nevertheless be effective at blocking itch. In mice, chemical ablation of TRPV1-expressing sensory neurons caused an almost complete loss of itch behavior [19]. Similarly in humans, repeated topical application of capsaicin is commonly used as a treatment for itch [20]. For instance, in a double-blind trial, low-concentration capsaicin treatment (0.025%, four times daily) caused a significant decrease in pruritic psoriasis [21]. This treatment seems paradoxical, and in some ways it is, for the initial response to capsaicin is intense burning, stinging, and itch. However, when TRPV1-expressing sensory nerve fibers are exposed to repeated applications of TRPV1 agonists, TRPV1-expressing sensory neurons show retraction of peripheral processes and impaired signaling (Figure 16.1), an effect that has been termed defunctionalization [22]. More recently, the use of a high-concentration capsaicin treatment (8% capsaicin patch) has been developed [23]. The advantage to this approach is that a single application can cause the defunctionalization of TRPV1-expressing sensory neurons. In anecdotal reports, an 8% capsaicin patch was successfully used to treat notalgia paraesthetica [24]. In addition, the fact that capsaicin patches significantly reduce postherpetic neuralgia strongly suggests that such treatment will also be effective for postherpetic itch [25].

However, the treatment of intractable itch with capsaicin, though it can be effective, is a long way from a perfect solution. The treatment itself (not surprisingly) can be quite painful—indeed, concentrated capsaicin is applied in the presence of a local anesthetic. Furthermore, on the retraction of TRPV1-expressing fibers, there is a multimodal loss in sensitivity, including pain, itch, and warm/hot temperatures, and so the treatment is not particularly specific for itch. For these reasons, capsaicin therapy for pruritus only makes sense if the itch is localized to a specific region of the skin. Finally, the effects of capsaicin treatment are temporary, lasting up to 12weeks, whereas the peripheral terminals from TRPV1-expressing sensory neurons regenerate and reinnervate the skin. Despite these shortcomings, low-dose capsaicin treatment is commonly used for the treatment of itch [20], and the newer high-concentration formulas have the potential to be broadly effective for numerous types of localized, neuropathic itch.

Activating TRPV1 is not the only strategy to combat itch—the other is via TRPM8. Many over-the-counter itch remedies contain menthol as an active ingredient. Now research from our own lab may have uncovered the underlying mechanism for how menthol inhibits itch [26]. Menthol is known to activate TRPM8, which is expressed on sensory afferents that convey cool [27]. Our work suggests that menthol acts as a counterstimulus that inhibits itch via the activation of a specific population of inhibitory neurons in the dorsal spinal cord, which we have termed B5-I neurons. In particular, we show that B5-I neurons, which function to inhibit itch, get direct input from TRPM8 afferents. In addition, we find that, whereas menthol inhibits itch in wild-type mice, this counterstimulus has no effect in mice that are lacking B5-I neurons. These data suggest that B5-I neurons are the cellular basis for the inhibition of itch by menthol. In this regard, menthol and scratching may be two types of counterstimulation that work via analogous spinal circuits. But activation of TRPM8 as a strategy to inhibit itch has a notable advantage over scratching because TRPM8 activation—unlike scratching—does not cause tissue damage that further exacerbates itch. These findings suggest that TRPM8 agonists are a promising therapy for pruritus and that they reduce itch by harnessing endogenous anti-itch circuits.

Inhibiting TRP Channels to Block Itch

Numerous chemical agents cause itch, and recently we have begun to get a clearer picture of how they do so. Activated dermal mast cells release histamine and serotonin, which bind to receptors on sensory neurons and cause itch. Foreign materials from plants and insects often have protease activity that can cause itch [28]. For instance, spicules from the tropical plant cowhage contain a protease that activates PAR2, and proteolytic cleavage products including BAM8-22 and SLIGRL cause itch by activating MrgprC11 [15,16,28]. Often itch is part of an immune response that involves the release of numerous cytokines, such as interleukin 31 and TSLP, which act on cytokine receptors [29,30]. What these itch receptors have in common is that they are all metabotropic rather than ionotropic receptors—in other words, they are not sufficient to trigger an action potential by themselves. Instead, these metabotropic receptors need to be coupled through signaling pathways to a channel that allows current influx to support the generation of an action potential.

Notably, in every case that has been examined in detail, it has been found that the channels that are activated downstream of itch receptors are TRP channels. In particular, H1 receptor for histamine is coupled to TRPV1 via phospholipase A2 [19,31]. Analogously, MrgprA3, MrgprC11, and TSPLR appear to be coupled to TRPA1, whereas IL-31R may be able to couple to either TRP channel [29,30,32]. These studies emphasize the important idea that TRPV1 and TRPA1 are more than just pain sensors. Rather, it is now emerging that these channels are integrators of diverse noxious stimuli, including those that induce sensations of itch (Figure 16.2).