New Insight into the Origins of Itch and Pain: How are Itch and Pain Signals Coded and Discriminated by Primary Sensory Neurons?

The unpleasant sensations itch and pain induce distinct behavioral patterns. Although both acute itch and pain serve protective functions, they are different sensory modalities. Itch stimuli lead to the scratching reflex or the desire to scratch, which may be helpful for removing potential irritants. In contrast, pain stimuli lead to withdrawal behaviors but not scratching [1]. Chronic itch and pain are significant clinical symptoms of many systemic diseases, and markedly reduce the quality of life of affected patients. Clinically, effective treatments for chronic itch and pain are still limited. The incomplete understanding of how itch and pain are processed by the somatosensory system is a major hurdle to developing new effective analgesics and itch relievers.

Although itch and pain are clearly distinct sensations, they are closely related and share some neuroanatomical pathways. The peripheral terminals of primary afferent sensory neurons, whose cell bodies are localized in the dorsal root ganglia (DRGs) and trigeminal ganglia, detect a number of sensory stimuli, such as touch, temperature, pain, and itch, and their central branches transmit these signals to the spinal cord. Although our understanding of how itch and pain are transduced at the molecular and cellular levels has improved dramatically for decades, how itch and pain are encoded and discriminated by the somatosensory system remains largely unclear [1].

The current theories of itch coding are unable to explain all the experimental phenomena. First, the intensity theory argues that primary sensory neurons are polymodal and can convey both pain and itch signals. Strong stimulation of a subpopulation of primary sensory neurons can induce pain sensations, while weak stimulation can induce itch sensations. Second, the labeled-line theory suggests that itch and pain are sensed or coded by non-overlapping subpopulations of sensory neurons. Third, the selectivity theory (also called population coding theory) suggests that itch-selective neurons also respond to painful stimuli and itchy stimuli exclusively activate itch-selective neurons to induce itch, while painful stimuli are able to activate both itch-selective neurons and other nociceptive populations, whose activation induces pain sensation with masked itch sensation. Fourth, the spatial contrast theory of itch argues that the activation of a subpopulation of nociceptors is able to produce itch sensation and contrasts this stimulus against the inactivated nociceptors of the surrounding nerve fibers. Nevertheless, it is still elusive whether and how primary afferent sensory neurons encode itch and pain signals differently.

Mas-related G-protein-coupled receptors (Mrgprs) constitute a family of G-protein-coupled receptors expressed by non-peptidergic primary sensory neurons. MrgprA3 is the primary receptor for the pruritogen chloroquine, and MrgprA3-expressing primary sensory neurons are sufficient and essential for mediating itch in mice [1]. In a new paper published in Neuron, Sharif et al. provide compelling evidence that metabotropic and ionotropic stimulation of the itch-selective MrgprA3⁺ neurons induce itch and pain sensations, respectively [1]. Sharif et al. first employed adeno-associated virus (AAV)-mediated delivery in hemizygotic MrgprA3^Cre-EGFP mice to selective express the Gq-coupled DREADD hM3Dq-mCherry in MrgprA3⁺ primary sensory neurons in mice. Injection of clozapine N-oxide (a DREADD agonist) into the nape of the neck of AAV-injected Cre⁺ mice induced robust scratching, but not in control mice, suggesting that metabotropic activation of MrgprA3⁺ C-fibers predominantly triggers the itch sensation in mice. Next, by crossing transgenic lines to produce MrgprA3^{Cre-EGFP+/−}: Rosa26 ^{ChR2-EYFP+/−} mice, Sharif et al. selective expressed the excitatory light-gated actuator channelrhodopsin-2 (ChR2) in MrgprA3⁺ primary sensory neurons. In these mice, blue laser illumination of the nape of the neck evoked pain behaviors. Furthermore, cheek injection of clozapine N-oxide induced more scratching than wiping in AAV-injected MrgprA3^{Cre-EGFP+/−} mice. Moreover, optogenetic activation of MrgprA3⁺ neurons in the cheek evoked pain-associated wiping in mice. Thus, these data indicate that metabotropic activation of MrgprA3⁺ C-fibers conveys itch signals, while ionotropic stimulations on MrgprA3⁺ C-fibers conveys pain signals. Finally, Sharif et al. found that peripheral transient receptor potential (TRP) channels (TRPV1 and TRPA1) are involved in itch, but not pain induced by ionotropic activation of MrgprA3⁺ C-fibers in mice. The MrgprA3⁺ primary sensory neurons have been proposed as a labeled line for itch. However, Sharif et al. demonstrated that MrgprA3⁺ neurons not only convey itch signals, but also transmit pain signals, depending on the stimulation modalities, thus challenging the labeled-line theory (Fig. 1A).

Fig. 1.

Itch and pain sensations are encoded and discriminated by a single subpopulation of primary sensory neurons, possibly dependent on their distinct stimulation modes. A Slower metabotropic activation of MrgprA3 C-fibers evokes itch, while faster ionotropic activation of MrgprA3 C-fibers evokes pain. B Different ligands acting on TRPA1 channels evokes either itch or pain, depending on the stimulation pattern and stimulation intensity. C The peptide SLIGRL binds to PAR2 coupling to TRPV1 to induce pain, while SLIGRL binds to MrgprC11 in primary sensory neurons to elicit itch. D The neuropeptide SP binds to NK-1R coupling to TRPV1 to induce pain, while SP binds to Mrgprs in primary sensory neurons to elicit itch. CQ, chloroquine; CNO, clozapine-N-oxide; hM3Dq, mutated human Gq-coupled M3 muscarinic receptor; ChR2, channelrhodopsin-2; TRP, transient receptor potential; MGO, methylglyoxal; AITC, allyl isothiocyanate; miR-711, microRNA-711; IMQ, imiquimod; PAR-2, protease-activated receptor 2; NK-1R, neurokinin-1 receptor; SP, substance P; SLIGRL, Ser-Leu-Ile-Gly-Arg-Leu peptide.

Besides MrgprA3⁺ C-fibers, we asked whether and how other subpopulations of primary sensory neurons coded itch versus pain sensations. TRP cation channel, subfamily A, member 1 (TRPA1) induces Ca²⁺ influx after activation, and is strongly expressed in a subpopulation of nociceptors. Accumulating evidence shows that different activation of TRPA1 in the DRG neurons may convey either pain or itch signals (Fig. 1B). Our previous study showed that TRPA1 mediates H₂O₂-induced itch in mice [3]. We also found that methylglyoxal, a highly reactive dicarbonyl compound that is up-regulated in diabetic patients, directly activates TRPA1 in DRG neurons and contributes to diabetic itch in mice [4]. Extracellular miRNA-711 directly activates TRPA1 through its core sequence GGGACCC binding to the extracellular S5-S6 domains in TRPA1 to evoke acute and chronic cancer-related itch in mice [5]. Imiquimod (IMQ) has been found to directly activate TRPA1 to induce itch in zebrafish and mice [6]. Low concentrations of IMQ evoke itch, while higher concentrations elicit pain in zebrafish [6]. On the contrary, low concentrations of allyl isothiocyanate (AITC; a TRPA1 agonist) cause itch behaviors, while higher concentrations produce pain behaviors in mice and zebrafish [6]. Moreover, McNamara et al. showed that formalin excites nociceptors by directly activating TRPA1 to induce pain sensation in mice [7]. Thus, we proposed that different ligands of the same TRPA1 channel can differently induce itch and pain through different intensities and/or different modes of stimulation of TRPA1⁺ neurons that may lead to distinct neuronal firing patterns (Fig. 1B). Nevertheless, we cannot exclude the possibility that a subset of TRPA1⁺ afferents are activated by lower intensity of stimulation of TRPA1, while the remaining TRPA1⁺ afferents are additionally recruited by a higher intensity of stimulation.

An identical mediator can convey either itch and pain through the activation of distinct receptors and recruit downstream intracellular signaling pathways in primary sensory neurons. The synthetic peptide SLIGRL binds to two different G-protein-coupled receptors, protease-activated receptor 2 (PAR2) and MrgprC11 in primary sensory neurons, each of which may couple to distinct intracellular signal transduction pathways, leading to pain and itch, respectively. Moreover, activation of PAR2 elicits thermal and mechanical hyperalgesia by sensitizing TRP channels, such as TRPV1 (Fig. 1C). The neuropeptide substance P (SP) is expressed in a subset of unmyelinated nociceptive primary sensory neurons in the DRGs. Generally, SP regulates pain transmission by acting on neurokinin-1 receptor (NK-1R), which is expressed in postsynaptic dorsal horn neurons [8]. Interestingly, SP acts on NK-1R in primary sensory neurons to induce thermal hyperalgesia via protein kinase Cε-mediated sensitization of TRPV1 [8]. Intriguingly, SP can also directly activate Mrgprs on DRG neurons to induce itch, and SP-induced itch is independent of NK-1R [9] (Fig. 1D). Together, these data indicate that the identical ligand can bind to distinct receptors for mediating either itch and pain, possibly through different receptors and intracellular signal transduction pathways in the same subpopulation of primary sensory neurons.

To date, our understanding of how itch and pain signals are encoded and discriminated in the spinal cord is still limited. Sharif et al. have demonstrated that MrgprA3-expressing neurons that mediate itch are inhibited by intracisternal injection of the gastrin-releasing peptide receptor (GRPR) antagonist RC-3095, while MrgprA3-expressing neurons that mediate pain are inhibited by intracisternal injection of the μ-opioid receptor agonist DAMGO in mice. These data indicated that the metabotropic activation of MrgprA3⁺ afferents engages GRP-GRPR signaling, while the ionotropic activation of MrgprA3⁺ afferents engages an opioid-sensitive pathway in the spinal cord in mice. Primary sensory afferents release neurotransmitters and neuropeptides to activate postsynaptic receptors to transmit itch and/or pain signals in the spinal dorsal horn. MrgprA3-expressing primary sensory neurons form monosynaptic connections with excitatory interneurons that contain GRP [2]. Recently, it was found that itch signals are transmitted only when GRP⁺ neurons fire action potentials in bursts [10], which enables GRP release, and then the activation of GRPR-expressing neurons that carry itch signals from the spinal cord to the brain [10]. It is still unclear whether MrgprA3 afferents differentially encode itch versus pain through different firing patterns: bursting versus single action potentials. Recently, Dong et al. proposed a new leaky-gate model in the spinal cord to explain how itch and pain are encoded, revealing that GRP⁺ neurons in the spinal cord receive direct monosynaptic inputs from both primary nociceptive and pruriceptive neurons, indicating that GRP⁺ neurons can convey both itch and pain signals. For painful stimuli, the endogenous opioid system is recruited to close the gate, leading to decreased pain generated by parallel pathways [2].

In summary, accumulating evidence supports the hypothesis that primary nociceptive sensory neurons are intrinsically multimodal. Different stimulation modes, different stimulation intensities, and different intracellular signals in the same subpopulation of primary sensory neurons may evoke distinct somatosensory modalities (e.g. itch and pain), possibly via different neuronal firing patterns of nociceptors. Moreover, the same ligands may bind to different receptors to induce itch and pain through different intracellular signal transduction pathways in a subpopulation of primary sensory neurons. These findings provide new insights that the same subpopulation of primary sensory neurons may encode and discriminate itch and pain by distinct neuronal firing patterns and/or distinct intracellular secondary messengers, resulting in distinct outputs in their central terminals projecting to the dorsal horn of the spinal cord. This warrants further investigation to decode the spatiotemporal properties of the afferent firing patterns following different modes of activation of a single subpopulation of primary sensory neurons.