Neurophysiological, Neuroimmunological, And Neuroendocrine Basis Of Pruritus

Truth Seeker

New Member
Pruritus (itch) can be defined as an unpleasant cutaneous sensation associated with the immediate desire to scratch. Recent findings have identified potential classes of endogenous "itch mediators" and establish a modern concept for the pathophysiology of pruritus. First, there in no universal peripheral itch mediator, but disease-specific sets of involved mediators. Second, numerous mediators of skin cells can activate and sensitize pruritic nerve endings, and even modulate their growth. Our knowledge of itch processing in the spinal cord and the involved centers in the central nervous system is rapidly growing. This review summarizes the current information about the significance of neuron—skin interactions, ion channels, neuropeptides, proteases, cannabinoids, opioids, kinins, cytokines, biogenic amines, neurotransmitters, and their receptors in the pathobiology of pruritus. A deeper understanding of these circuits is required for the development of novel antipruritic strategies.

Pruritus (itch) can be defined as an unpleasant cutaneous sensation associated with the immediate desire to scratch. Teleologically, pruritus may be interpreted as part of the body's defense mechanism by which we dispose of potential dangerous organisms or stimuli. Chronic or intense scratching leads to the development of skin lesions and the release of inflammatory mediators that potentially induce or aggravate pruritus resulting in reinforcement of scratching. This "itch—scratch" cycle, unfortunately, is frequently resistant to topical and systemic therapy (Figure 1, Table S1).

Pruritus is one of the most common symptoms in dermatology and general medicine. It can be initiated during inflammation, cancer, metabolic diseases, infection, psychiatric diseases, drug application, stress, and others. Current evidence clearly indicates the existence of an interactive network between the skin and the peripheral as well as the central nervous system to regulate and respond to pruritic stimuli (Figure 1). It became evident that specified sensory nerves and their receptors are crucially involved in the pathophysiology of pruritus. Thus, pruritus is not merely a submodality of pain, but an individual sensation of our sensory nerve system.

Recent observations clearly expand our knowledge of potential classes of endogenous "itch mediators" (reviewed in Stander and Steinhoff, 2002; Yosipovitch et al., 2003) and established a modern concept of the pathophysiology of pruritus (Table 1). In other words, different peripheral itch mediators and receptors may be involved with a different impact among various pruritic diseases (e.g. atopic dermatitis, urticaria, renal, and cholestatic pruritus). However, the importance of the central nervous "itch centers" under physiological and pathophysiological conditions is still at the beginning of being understood. This review summarizes the current knowledge about the significance of various mediators and receptors in the pathophysiology of pruritus. Owing to limited space, several important papers could not be cited. A prolonged reference version can be found in the supplement.

Neurophysiology of Itch
Primary afferent pruriceptive neurons
According to the intensity hypothesis of itch, low-level activation of unspecific nociceptors would induce pruritus, whereas higher discharge frequencies would provoke pain. However, application of low concentrations of algogens generally does not cause itch, just less intense pain. Further, intraneural electrical microstimulation of human afferent nerves induces either pain or, less commonly, pruritus. Increasing the stimulation frequency increases the intensity of pain or itch, but no switch from pruritus to pain is observed. Therefore, a dedicated neuronal system for encoding itch could be predicted. C-fibers, responding to histamine application in parallel to the itch ratings of subjects, have been discovered among the group on mechano-insensitive C-afferents confirming that there is a specific pathway for itch (Figure 1, Tables 1, 2, and S1). In contrast, the most common type of C-fibers, "polymodal" nociceptors are either insensitive to histamine or only weakly activated by it (Schmelz et al., 2003). Thus, this fiber subtype cannot account for the prolonged itch induced by the intradermal application of histamine.

The histamine-sensitive or "itch" fibers or pruriceptors are characterized by a particular low conduction velocity, large innervation territories, mechanical unresponsiveness, and high transcutaneous electrical thresholds (Schmelz et al., 2003). In line with the large innervation territories of these fibers, two-point discrimination for histamine-induced itch is poor (15 cm in the upper arm).

The relative prevalence of the different C-fiber types has been estimated from recordings in the superficial peroneal nerve (Schmidt et al., 1997). About 80% are polymodal nociceptors, which respond to mechanical, heat, and chemical stimuli. The remaining 20% do not respond to mechanical stimulation. These fibers are mainly "mechano-insensitive nociceptors", which are activated by chemical stimuli, and can be sensitized to mechanical stimulation in the presence of inflammation (Schmelz et al., 2000b). This latter characteristic led to the name "sleeping nociceptor". Among the mechano-insensitive afferent C-fibers, there is a subset of units, which have a strong and sustained response to histamine. They comprise about 20% of the mechano-heat-insensitive class of C-fibers.

The pruritogenic potency of inflammatory mediators is characterized by their ability to activate histamine-positive mechano-insensitive C-nociceptors. However, concomitant activation of mechano-sensitive and mechano-insensitive histamine-negative nociceptors will decrease the itch. Therefore, itch sensation is apparently based on both, activity in the pruriceptors and absence of activity in the pain-mediating nociceptors.

Additional primary afferents involved in producing itch
Histamine-sensitive C-fibers have been found among the mechano-insensitive afferent C-fibers. They are characterized by very high transcutaneous electrical thresholds and are involved in the generation of the axon reflex erythema (Schmelz et al., 2000a). Recently, focal electrical stimulation with low intensity but high frequency has been shown in humans to induce itch (Ikoma et al., 2005). Interestingly, the electrically induced itch was not accompanied by an axon reflex erythema, suggesting that the activated fibers do not belong to the histamine-sensitive pruriceptors described above.
Specific spinal pruriceptive neurons
Using the cat spinal cord, a specific class of dorsal horn neurons projecting to the thalamus has been demonstrated, which respond strongly to histamine administered to the skin by iontophoresis (Andrew and Craig, 2001). The time course of these responses was similar to that of itch in humans and matched the responses of the peripheral C-itch fibers. These units were also unresponsive to mechanical stimulation and differed from the histamine-insensitive nociceptive units in lamina I of the spinal cord. In addition, their axons had a lower conduction velocity and anatomically distinct projections to the thalamus. Thus, the combination of dedicated peripheral and central neurons with a unique response pattern to pruritogenic mediators and anatomically distinct projections to the thalamus provides the basis for a specific neuronal pathway for itch.
Central itch processing
The itch-selective units in lamina I of the spinal cord form a distinct pathway projecting to the posterior part of the ventromedial thalamic nucleus, which projects to the dorsal insular cortex, a region that has been shown to be involved in a variety of interoceptive modalities like thermoception, visceral sensations, thirst, and hunger. The supraspinal processing of itch and its corresponding scratch response have recently been investigated in man by functional positron emission tomography (Darsow et al., 2000). Induction of itch by intradermal histamine injections and histamine skin-prick-induced co-activation of the anterior cingulate cortex, supplementary motor area, and inferior parietal lobe predominantly in the left hemisphere (Mochizuki et al., 2003 and references therein). The significant co-activation of motor areas supports the familiar observation that itch is inherently linked to a desire to scratch. The multiple activated sites in the brain after itch induction argue against the existence of a single itch center and reflect the multidimensionality of itch. Thus, a broad overlap of activated brain areas is evident for pain and itch (Drzezga et al., 2001 and references therein). However, subtle differences in the activation pattern between itch and pain have been described. In contrast to pain, itch seems to be characterized by a lack of secondary somatosensory cortex activation on the parietal operculum and by left hemispheric dominance (Drzezga et al., 2001). In addition, the periaqueductal gray matter was activated only when painful and itching stimuli were simultaneously applied. This activation was combined with reduced activity in the anterior cingulate, dorsolateral prefrontal cortex, and parietal cortex, suggesting that the periaqueductal gray matter might be involved in the central inhibition of itch by pain (Mochizuki et al., 2003).

Skin—nerve interactions and itch
The skin is highly innervated by primary sensory nerves, postganglionic cholinergic parasympathetic, and postganglionic sympathetic nerves, resulting in a complex afferent/efferent neuronal network in the skin. It is further discussed that in the facial skin some parasympathetic nerves exist associated with flushing. These neurons utilize, among others, classical neurotransmitters (catecholamines, acetylcholine), certain neuropeptides (e.g. substance P (SP), calcitonin gene-related peptide (CGRP), opioids, cannabinoids (CB)), and certain neurotrophins (nerve growth factor (NGF), neurotrophin-4).

Certain subtypes of unmyelinated C-fibers also project into the epidermis (Kennedy et al., 1994). Thus, neuromediators may directly communicate with keratinocytes or Langerhans cells, and vice versa. For example, CBs stimulate the release of -endorphin from keratinocytes, thereby activating sensory neurons resulting in the modulation of pain (Ibrahim et al., 2005). From that it is evident that dysregulation of skin function (pH changes, trauma, disrupted barrier function, inflammation, infection, UV light) can directly or indirectly stimulate sensory nerve endings, thereby inducing pruritus (Figure 2). Therefore, cell—nerve interactions in the epidermis as well as the dermis influence skin-derived pruritus (Figure S1, Tables 1 and 2).

Role of the epidermis in itch sensation
Itch sensations can be induced owing to damage to the skin barrier function in the dry skin (xerosis) and atopic eczema as well as papulosquamous diseases such as psoriasis and lichen planus. In this context, itching most probably is not related to mast cells (MCs). The epidermis itself is innervated by sensory nerve endings anatomically associated with keratinocytes and Langerhans cells. Recent studies have demonstrated that — upon stimulation — keratinocytes are capable of releasing pruritic as well as antipruritic mediators such as endovanilloids, endorphins, neuropeptides, proteases, and cytokines (Figure 2, Table 1). With respect to the role of dry skin in pruritus, Nojima et al. (2004) have shown in a murine model that dry skin induces enhanced c-fos expression, which reflects activation of spinal cord neurons. SP as well as CRF decrease the electric potential in keratinocytes as well as did skin barrier disruption. Thus, ion channels including the voltage-gated, ATP-gated (Inoue et al., 2005), and transient receptor potential (TRP)V-gated may be directly involved in the transmission of pruritus via keratinocyte activation in the dry skin. Keratinocytes express a variety of receptors (neuropeptide receptors, neurotrophin receptors (for NGF, neurotrophin-4), CB receptors, proteinase-activated receptor-2 (PAR2), and the TRPV1 ion channel, all of which have been demonstrated to be involved in transmitting itch sensations. Interestingly, different trigger factors are capable of stimulating the release of pruritogenic or antipruritogenic factors from keratinocytes. For example, prostaglandin E2 induces the release of neurotrophin-4 from keratinocytes (Kanda et al., 2005). Both, the histamine H1 and H2 receptors are expressed by keratinocytes and may be involved in epidermal barrier dysfunction (Ashida et al., 2001). In addition, nociceptin may stimulate the release of leukotriene B4 from keratinocytes via opioid receptor-like 1 receptor (Andoh et al., 2004). Moreover, Miyamoto et al. (2002) have shown that antagonists of opioid receptors suppressed scratching behavior in a mouse model of dry skin, most probably by a central effect. In summary, dry skin induces itch by stimulating the release of various pruritogenic mediators from keratinocytes. Less, however, is known about the potential release of endogenous antipruritogenic factors from keratinocytes.
The MC—nerve connection: functions beyond itch induction?
The depicted MC—nerve interactions may not only promote but also terminate pruritus. This yet unproven hypothesis is supported by several independent lines of evidence (Figure S1). First, MCs express and release large amounts of proteases (tryptase, chymase, carboxypeptidase A, and cathepsins), which have been shown to degrade/inactivate pruritogenic peptides. For example, MC enzymes have been implicated to downregulate axon reflex-mediated neurogenic inflammation in skin and other tissues by cleaving SP, CGRP, and vasoactive intestinal polypeptide (VIP). Also, MC chymases efficiently degrade and terminate the activity of endothelin-1 (ET-1) (Maurer, 2002). ET-1 (a neuropeptide that is also produced by, among others, endothelial cells and MCs) has been implicated in host defense against bacterial infections (Maurer, 2002) and cardiovascular diseases, and is known to induce a burning and painful sensation as well as itch upon injection into human skin (Katugampola et al., 2000). Interestingly, degradation of ET-1 by MC chymase requires prior activation, which is, at least in part, provided for by ET-1 itself by binding to endothelin-A receptors on MCs. In other words, when ET-1 induces damage (e.g. pruritus), it also starts an "automatic self-destruction program", that is, endothelin-A-dependent release of ET-1-degrading chymase from MCs. This novel MC function may be relevant for many skin conditions, in which neuropeptides are implicated such as atopic dermatitis or psoriasis.
Why is the degradation of neuropeptides by MC proteases a likely mechanism of MC-mediated control of pruritus? MCs undergo degranulation and protease release in response to several neuromediators including SP and CGRP (Bienenstock et al., 1987). However, processing of peptide mediators by MC proteases does not require prior activation of MCs as some proteases are constitutively expressed on cell surfaces, thus allowing for continuous regulation of neuropeptide levels in the skin even in the absence of MC-degranulating signals (Hagermark, 1974). In addition, MCs may also contribute to the termination of neurogenic inflammation and pruritus by regulating tissue expression of neutral endopeptidases and angiotensin-converting enzyme, two zinc metalloproteinases that effectively control neuropeptide skin levels. Levels of angiotensin-converting enzyme and neutral endopeptidase in human skin MCs were found to be increased in response to various proinflammatory mediators that are produced by MCs. Interestingly, those proinflammatory mediators, particularly tumor necrosis factor alpha (TNF-), are also known to downregulate the expression of receptors for SP and corticotropin-releasing hormone (CRH). Thus, MC-derived mediators could also limit pruritus at sites of inflammation by increasing the clearance of neuropeptides via neutral endopeptidases and/or angiotensin-converting enzyme and by reducing the numbers of their binding sites.

Finally, several independent studies using MC-deficient animals have shown that the frequency and/or duration of scratching in response to itch-inducing agents are not reduced in the absence of MCs (Hayashi et al., 2001). On the contrary, pruritus appears to be similar to or even stronger in genetically MC-deficient KitW/KitW-v mice than in normal mice. For example, KitW/KitW-v mice exhibit a 20% increase in scratching behavior induced by histamine or SP as compared to normal Kit+/+ mice (Hossen et al., 2003 and references therein), and both the time and the incidence of scratching responses induced by intracutaneous injections of compound 48/80 exceeded those of Kit+/+ mice (Inagaki et al., 2002).

Taken together, these findings suggest that MCs can be involved both in the activation as well as termination of pruritus. The use of novel mouse models including techniques to surgically denervate defined skin areas as well as mice deficient for neuropeptides, neuropeptide-degrading proteases, and/or MCs will help to clarify these important questions.

Neuropeptides, Neurotrophins And Receptors
Tachykinins: "the MC connection"
One of the best — albeit probably not most important — neuromediators studied thus far is SP (Table 1). Upon stimulation by exogenous or endogenous trigger factors, the above introduced slow-conducting histamine-sensitive C-fibers are not only capable of transporting the itch signal to the central nervous system (orthodromic) but also release neuropeptides such as SP and CGRP (antidromic). Because of their anatomical association to cutaneous nerves, MCs and their released products appear to play an important role in the pathophysiology of itch and inflammation (Steinhoff et al., 2003a, 2003b). Independent of neurokinin-1 receptor, intradermal application of SP releases histamine from MCs, which acts as a pruritogen (Thomsen et al., 2002). Moreover, SP enhances intradermal concentrations of nitric oxide, which may enhance SP-induced pruritus (Andoh and Kuraishi, 2003). Specific neurokinin-1 receptor effects of SP on MCs include TNF upregulation, which in turn can sensitize pruriceptive afferent endings.
Therefore, it is apparent that various neuropeptides may exert indirect effects on sensory nerves by triggering the release of pruritic agents from various target cells such as MCs (histamine, tryptase, chymase, TNF-), endothelial cells (kinins, endothelin), keratinocytes (prostanoids, NGF), and immune cells (cytokines). This mechanism is probably the main cause for itch in atopic dermatitis, psoriasis, and prurigo nodularis (Ständer et al., 2003).

In the past, it was believed that MC activation was all-or-nothing, with IgE cross-linking inducing the functional consequences of allergic reactions and anaphylaxis. However, the activity of MCs in health and disease is clearly much more complex. Using the patch-clamp technique, Janiszewski et al. (1994) reported that MCs do not respond electrophysiologically to very low (picomolar) concentrations of SP, but activation and delayed degranulation occurred after a second exposure. Therefore, MCs can be primed when exposed to physiologically relevant low concentrations of SP, and lower their thresholds to subsequent activation. Thus, if MCs are indeed primed by SP exposure, it would enable a subthreshold stimulus to initiate MC activation (Figure S1). Furthermore, secretion can occur without the evidence of degranulation, and even molecules stored within the same granules can be released and secreted in a discriminatory pattern. Of course, this mechanism may also be true for other mediators not studied thus far.

Another group of neuropeptides, the opioids, have been used as analgetic agents for more than 2000 years. So far, more than 20 endogenous analogs have been described, which are subdivided into three classes (endorphins, enkephalins, and dynorphins) and exert their effects by triggering opioid receptors (, , , and orphan receptors) (Table 1).
Peripheral effects With respect to itch, intradermally injected opioids activate MCs by a non-receptor-mediated mechanism. In contrast to morphine, the highly potent -opioid agonist, fentanyl does not provoke any MC degranulation, even at concentrations having -agonistic effects exceeding those of morphine. Thus, one can conclude that morphine-induced MC degranulation is not mediated by -opioid receptors. As high local concentrations of opioids are required to degranulate MCs, itch induced by systemic administration in therapeutic opioids doses is probably not owing to peripheral MC degranulation, but to central mechanisms. There is also no evidence for direct neuronal excitation by peripherally applied opioids as potent opioid agonists, even at high concentrations, do not provoke itch (Blunk et al., 2004 and references therin). Thus, the impact of the observed increased expression of -opioid receptors in atopic dermatitis (Bigliardi-Qi et al., 2005) is unclear. However, according to the inhibitory effects of neuronal -opioid receptors, it would be expected that their increased expression act antipruritic. Interestingly, peripherally applied CBs suppress histamine-induced pruritus, with inhibitory CB receptors CB1 and CB2 being found on skin nerves (Stander et al., 2005). However, the inhibitory effects of peripheral CBs have been suggested to be — at least in part — mediated by a secondary release of endogenous opioids in the skin (Ibrahim et al., 2005). Moreover, endogenous CBs, such as anandamide, have also been shown to activate TRPV1 receptors (see below), which adds to the complex role of CBs in the modulation of pruritus. A very similar complex interaction has been shown for nociceptin: it activates the inhibitory opioid receptor-like 1 receptor, whereas the direct neuronal effect on nociception appears to be inhibitory. Additionally, the secondary release of leukotriene B4 by nociceptin in the skin provokes itch behavior (Andoh et al., 2004).
Central effects It is common experience that itch sensation can be reduced by painful sensations caused by scratching or by various painful (thermal, mechanical, chemical) stimuli. Cutaneous field stimulation has also been successfully used to inhibit histamine-induced itch for several hours in a relatively large (20 cm in diameter) area around a stimulated site suggesting a central mode of action. Not only is itch inhibited by enhanced input of pain stimuli, but vice versa inhibition of pain processing may reduce its inhibitory effect, and thus enhance itch. This phenomenon is particularly relevant to spinally administered -opioid receptor agonists, which induce segmental analgesia often combined with segmental pruritus (Onigbogi et al., 2000). This mechanism might well account for the antipruritic effect of -opioid antagonists (nalmefene, naloxone, and naltrexone) observed in experimentally evoked histamine-induced itch as well as pruritus in different dermatoses (Metze et al., 1999; Heyer et al., 2002). This phenomenon is particularly relevant to spinally administered -opioid receptor agonists, which induce segmental analgesia combined with a naloxone-sensitive segmental pruritus in about half of the patients.
Interestingly, while -opioid receptor antagonists significantly diminish itch, in animal experiments, -opioid antagonists enhanced itch (Kamei and Nagase, 2001). In line with these results, the -opioid agonist nalbuphine as well as the new developed -opioid receptor agonist, TRK-820, have been shown to reduce pruritus. As - as well as -opioid receptors are expressed by cutaneous nerve fibers, a crucial role of opioids in the modulation of itching can be expected.

Kinins, kallikreins
Pruritic activity of kinins has also been investigated decades ago (Table 1). It was shown that both the tryptic enzymes (such as kallikreins) and the resulting peptide fragments (like the mainly pain-inducing bradykinin) can induce itch by activating histamine-sensitive C-fibers (Schmelz et al., 2003). Furthermore, there is some evidence that bradykinin type-2 receptor antagonists can reduce deoxycholic acid-induced itch behavior in rodents (Hayashi and Majima, 1999). Moreover, in this model, inhibition of tissue kallikrein suppressed the itch indicating a major role for epidermal kallikrein during itch responses. Recent studies on itch mediators mainly focus on the role of proteases in pruritus (see also in the following chapters): increased kallikrein activity and reduced kininogen levels were found in pruritic popular eruptions. It is noteworthy that intracutaneous injections of kallikrein did not provoke marked itch in rats. However, massive itch behavior is observed in mice overexpressing epidermal kallikrein-7 (Ny and Egelrud, 2004 and references therein).
Other neuropeptides and neurotrophins
Several mediators from the central and peripheral nervous system, which are involved in neuroimmune and neuroendocrine interactions (Steinhoff et al., 2003a, 2003b and references therein). Some of these mediators have been implicated in the pathophysiology of pruritus (Table 1). For example, CGRP modulates inflammation and pruritus (Brain and Grant, 2004), and prolongs itch latency following SP injection suggesting an inhibitory effect of CGRP on SP-induced itching. However, increased amounts of CGRP were observed in nerve fibers of pruritic diseases such as atopic dermatitis, nummular eczema, and prurigo nodularis.
Intradermal application of vasoactive intestinal polypeptide, neurotensin, and secretin also led to a histamine-dependent itch response, associated with pruritus, whealing, and axon-reflex erythema. In normal human skin, vasoactive intestinal polypeptide showed a comparable potency with respect to pruritus as compared to SP. Moreover, pituitary adenylate cyclase-activating polypeptide, somatostatin, and CRH were demonstrated to stimulate histamine release and MC degranulation from human and rat skin (reviewed in Steinhoff et al., 2003a, 2003b and references therein). CRH is well known as an important neuroendocrine mediator involved in stress response, thereby modulating inflammation, immunity, and pruritus. For example, it was shown that CRH triggers release of several mediator involved in itch response (Lytinas et al., 2003), possibly via CRH-R1 (Cao et al., 2005) (Table 1).

Neurotrophins such as NGF and neurotrophin-4 have also been implicated in itch pathophysiology. NGF is released by keratinocytes, MCs, and fibroblasts (Groneberg et al., 2005). Activation of its high-affinity receptor (trk A) on sensory nerves leads to nerve sprouting and sensitization. Serum levels of NGF are increased in atopic dermatitis patients, which induces release of the pruritogenic mediator tryptase (Groneberg et al., 2005). Also, MCs and keratinocytes of these patients generate high levels of NGF, which can be stimulated by histamine (Kanda and Watanabe, 2003). Similarly, neurotrophin-4 levels were also found to be enhanced in atopic dermatitis (Grewe et al., 2000), and brain-derived neurotrophic factor (BDNF) induces chemotaxis of eosinophils of these patients (Raap et al., 2005). Together, these results are clearly in favor of an important role of neurotrophins in the pathophysiology of itching, although direct evidence is still lacking.

Proteases and their receptors
Another intriguing pruritogenic candidates are the proteases. These molecules comprise about 5% of the human genome making them the largest protein family within the human being. Classically, proteases are still regarded as destructive enzymes, which break down proteins or merely activate or inactivate peptides by processing molecules. However, little is known about the role of proteases as signaling molecules during neuronal transmission. Already in the 1950s, proteases have been suggested to be ultimately involved in itch responses, burning, pain, and inflammation. Arthur and Shelley, and later Rajka demonstrated that exogenous as well as endogenous serine proteases are capable of inducing pruritus (Bodo et al., 2005 and references therein) (Figure 2, Table 1). Trypsin and MC chymase provoke itching and visible changes (edema, flare) when injected intracutaneously, at least in part via MC activation. In contrast, papain-induced pruritus seems to be histamine-independent.
A milestone in our understanding of protease-mediated signaling was the cloning of the first proteinase-activated receptor, PAR1. Although PAR1, PAR2, and PAR4 have been described in neurons, a role of PARs in the pathophysiology of itching has only been shown for PAR2 (Vergnolle et al., 2003). Importantly, functional PAR2 is present on primary spinal afferents, and releases neuropeptides upon stimulation by tryptase (Steinhoff et al., 2000, 2003a, 2003b, 2005). From this observation, one may speculate that proteinases activate PAR2 on sensory neurons, thereby triggering pruritus in skin diseases such as atopic dermatitis. Additionally, keratinocyte-derived PAR2, which is upregulated in the epidermis of atopic dermatitis patients (Buddenkotte et al., 2005) may mediate pruritus induced by endogenous (trypsins, kallikreins) or exogenous proteases (bacteria, house-dust mite). Indeed, in atopic dermatitis patients, the endogenous PAR2 agonist tryptase was increased up to four-fold and PAR2 expression was markedly enhanced on primary afferent nerve fibers of lesional skin. In contrast, no significant differences in histamine concentrations were observed between the diseased and healthy samples, suggesting that tryptase might be more important than histamine for the transmission of itch responses in atopic dermatitis. Moreover, intracutaneous injection of specific PAR2 agonists provoked sustained and prolonged itch in such patients. This observation may also explain why non-sedative antihistamines are poorly effective in atopic dermatitis. Thus, PAR2 activation on cutaneous sensory nerves and keratinocytes may be a novel pathway for the transmission of itch responses during atopic dermatitis and probably other skin diseases. Hence, PAR2 antagonists or protease inhibitors may be promising therapeutic targets for the treatment of pruritus (Figure 2).

The TRP channel family in itch
Recent findings suggest that itch mediators exert their effects also by activating ion channels of the TRP channel family. TRP channels comprise six groups of molecules: the canonical (TRPC), the vanilloid (TRPV), the melastatin (TRPM), the polycystin (TRPP), and mucolipin (TRPML) subfamilies, and the anhyrin (TRPA). In general, these molecules act as calcium-permeable sensory transduction channels to detect, for example, taste, warmth, heat, cold, and osmotic/mechanical stress both at cellular and multicellular (i.e., whole organism) levels (reviewed in Zhang et al., 2003 and references therein). With respect to the development and, most intriguingly, therapy of itch, recently certain temperature-sensitive members of the TRPV subfamily and the TRPM8 gained substantial interest (Figures 1 and 2; Tables 2 and S1).
TRPV1: a putative central role in the pathogenesis and therapy of itch TRPV1 was originally described on C-type nociceptive sensory neurons (Caterina et al., 1997) as a molecular target for capsaicin, the pungent ingredient of hot chili peppers. The activation of the receptor first excites these neurons by initiating ionic fluxes and concomitant action potential firing and neuropeptide release. At higher doses and longer times, capsaicin induces the desensitization the sensory afferents (Caterina and Julius, 2001 and references therein).
In addition to capsaicin, TRPV1 can also be activated/sensitized by numerous endogenous substances collectively referred to as "endovanilloids" (Table 2). The receptor was first shown to be stimulated by low threshold (>43°C), heat and acidosis. Later, several other molecules were also described to either directly and/or indirectly act on the TRPV1. These agents are, for example, eicosanoids, bradykinin, prostaglandins, and various neurotrophins (such as NGF, neurotrophin-3 and -4) (Lazar et al., 2004 and references therein). It was also shown that histamine-induced excitation of sensory neurons and PAR2 activation (Amadesi et al., 2004) does involve the activation/sensitization of TRPV1. Taken together, these findings strongly implicate that TRPV1 is indeed a central integrator molecule in the pain and itch pathway.

Prolonged or repeated vanilloid application results in a depletion of neuropeptides such as SP in C-type neurons, hence suspending the interplay between skin sensory neurons and MCs. Indeed, topical capsaicin effectively prevented histamine-induced itch under experimental conditions. In addition, capsaicin is widely used in the therapy of pruritus in numerous skin diseases such as prurigo nodularis, notalgia paresthetica, pruritus ani, hemodialysis-related pruritus, uremic pruritus, etc. (Biro et al., 1997; reviewed in Yosipovitch et al., 2003).

Furthermore, recent findings provide a new "hot" twist to the field further highlighting the pathophysiological and therapeutic importance of TRPV1 signaling in itch. Namely, functional TRPV1 channels were described on numerous non-neuronal cells types including, of greatest importance, human skin epidermal keratinocytes, dermal MCs, dendritic cells, and various keratinocyte populations of the hair follicle (Bodo et al., 2005 and references therein). In addition, activation of TRPV1 — beside markedly affecting proliferation, differentiation, and apoptosis — resulted in the upregulation of IL-1b and tumor growth factor-, whereas IGF and HGF were downregulated in the non-neuronal cells (Bodo et al., 2005 and references therein).

These novel results invite an attractive hypothesis with further potential therapeutic implications (Figures 1 and 3; Tables 2 and S1). Namely, topically applied capsaicin may not exclusively target sensory neurons but also TRPV1-expressing MCs and keratinocytes, thereby modulating the proposed neuronal—non-neuronal network. We do not know the complete pattern of mediator changes following TRPV1 stimulation on non-neuronal cells. However, TRPV1 expression was dramatically increased in epidermal keratinocytes of prurigo nodularis patients and normalized after successfully treating the characteristic nodular pruritic lesions with topical capsaicin (Stander et al., 2004). This example strongly suggests a major role of TRPV1 expression in non-neuronal cells in pruritus patients.

TRPV2, TRPV3, and TRPV4: further intriguing targets to be investigated Originally, these channels were also described as cellular temperature sensor molecules as all are activated by increasing temperatures (>53°C for TRPV2; >31°C for TRPV3; and >25°C for TRPV4) (Peier et al., 2002b and references therein) (Table 2). Importantly, TRPV3 exerts a very similar neuronal expression pattern to that of TRPV1. In addition, it was also postulated that TRPV3 subunits might form heteromultimeric structures by interacting with TRPV1 monomers and therefore may act as signal co-transducers and/or regulators of the TRPV1-mediated pain and itch sensation. Indeed, mice lacking the TRPV3 have strong deficits in response to both innocuous (mostly TRPV3 range) and noxious heat (rather TRPV1 range) (Moqrich et al., 2005).
Most intriguingly, functional TRPV3 and TRPV4 channels (similarly to TRPV1) are also expressed at high levels on epidermal keratinocytes (Peier et al., 2002b and references therein; Moqrich et al., 2005). In addition, TRPV4 was shown to be activated by such lipid peroxidation products as eicosanoids, which may also function as TRPV1-activating pruritogenic substances (Watanabe et al., 2003). Thus, sensitization and activation of TRP channels could underly pruritic activity of prostanoids and adding to their complex role in pruritus induction.

Finally, with respect to the central involvement of MCs in the initiation of itch, it is worth noting that TRPV2 and TRPV4 (along with TRPV1) are also expressed by MCs (Stokes et al., 2004). These authors also revealed that physical and thermal activation of TRPV2 on MCs resulted in a proinflammatory degranulation event, which depended on the activity of the protein kinase A-related signaling, one of the chief mechanisms in initiating sensitization of TRPV1 (a key event in initiating itch and pain, see above) as well.

Although more studies are necessary to further define the itch-related "non-thermosensor" role of these channels (e.g. in keratinocyte-specific cytokine and mediator synthesis and release), the close resemblance to TRPV1 with respect to cell-specific expression, activation, sensitization, and to the initiated cellular mechanisms highlights their putative roles in itch pathophysiology.

Icilin and the "cool" TRPM8 channel Another intriguing member of the TRP family is TRPM8, which is selectively expressed in C-type sensory neurons (Figures 1 and 3; Tables 2 and S1). TRPM8 is thought to be a thermosensor for coolness and cold (8—28°C) and also activated by chemicals — menthol, menthol analogs, and icilin — which produce sensations of cold (Peier et al., 2002a and references therein).
Icilin was first synthesized with the intent of finding a morphine-like analgesic, but when tested in rodents it produced unusual behavioral events. Icilin injected into the systemic circulation of fur-coated animals produces rotational movements similar to those manifested by a dog when wet ("wet dog shakes"). Icilin initiates punctate sensations of coolness, harmonizing with the idea of "cold spots", when the particles come into contact with the various mucosal surfaces (Wei and Seid, 1983). Icilin was compared to menthol and, for in vivo effects such as "wet shake behavior", it was 400 more active than menthol. Unlike menthol, it did not smell or irritate the eyes.

Comparisons continued after the cloning of TRPM8 with calcium entry into cells as an index of activity. The EC50 of icilin in TRPM8-transfected cells was reported to be 200 to 800 less than menthol, but potencies of the two cannot be directly matched because the maximum efficacy of icilin for calcium entry was greater than that of menthol. Icilin also activates TRPA1 (ANKM1), another TRP channel, but at a lower potency than TRPM8.

In an animal model of scratching in hairless rats, provoked by a magnesium-deficient diet, a 2% icilin ointment reduced the degree of excoriations by 55—60% (E Wei and Meingassner T, unpublished data). In this model, it was found that 2—3% icilin ointment, elicited robust cooling sensations for 2—4 h without any irritancy. The onset of maximal cooling occurred within 10—15 minutes after application.

The TRPM8 agonist menthol and analogs were also examined (Behrendt et al., 2004). The most active analogs were comparable in activity to (-)menthol and at least 16-fold less active than icilin, with lower maximal responses. Carboxamides exert activity similar to icilin on TRPM8 in vitro. When applied to the skin, carboxamides produce cold sensations lasting from 30 minutes to 1 hour. Thus, carbocamides may serve as a model for endogenous TRPM8 agonists. Thus, with the identification of TRPM8 (and TRPA1) (Table 2) and the novel chemical ligands for these ion channels, understanding the linkage of psychic and somatic (e.g. dry skin, temperature changes) adjuncts of skin discomfort to molecular, cellular, and sensory inputs now seems more exact. Ultimately, the proposed mechanisms of chemical action will have to be resolved by further tests in human subjects.

Cytokines, IFNs
Recent findings clearly indicate a direct role of cytokines and chemokines on the regulation of primary afferent neurons via receptor activation (Steinhoff et al., 2003a, 2003b and references therein; Dillon et al., 2004) (Figures 1 and 2 and S1).
IL-2 Clinical observations suggest a role of IL-2 as an inducer of pruritus. High doses of recombinant IL-2, applied to cancer patients, for example, are capable of provoking flush, vasodilatation, and pruritus. Whether this is a direct, receptor-mediated process or an indirect, for example, via MCs or endothelial cells, is still unknown. Accordingly, treatment of Alzheimer disease (AD) patients with systemic or topical immunosuppressants such as tacrolimus, pimecrolimus, or cyclosporin A, for example, which inhibit the production of various cytokines including IL-2, experience attenuation of pruritus.
Another mechanism of action with respect to cytokine-induced itch may be synergistic amplification or receptor transactivation. For example, bradykinin appears to augment the effect of IL-2-induced pruritus on sensory nerves. However, the latency preceding the itch response after injection in AD patients suggests an indirect pruritogenic effect of IL-2 via other mediators.

IL-8 Recent observations suggested a role of IL-8 as a mediator of itch in AD patients. However, intracutaneous application of IL-8 was not sufficient to induce pruritus or erythema in human skin (Stander and Steinhoff, 2002 and references therein).
IL-31 IL-31 is a novel cytokine preferentially produced by T-helper type 2 cells, which induces both severe prurititis and dermatitis in mice. Dillon et al. (2004) recently demonstrated that transgenic overexpression of the novel cytokine IL-31 in T-lymphocytes induces severe pruritus and dermatitis in mice. IL-31 signals via a heterodimeric receptor composed of IL-31 receptor A and the oncostatin M receptor. Whether IL-31 exerts its effects via direct activation of the IL-31R on sensory nerves or indirectly, for example, via keratinocytes is unknown. The finding that keratinocytes express the IL-31 R suggests that IL-31 may induce pruritus through the induction of a yet unknown keratinocyte-derived mediator, which subsequently activates unmyelinated C fibers in the skin. From these observations, it is intriguing to speculate that IL-31 is upregulated in pruritic but not in non-pruritic forms of chronic skin inflammation. Thus, IL-31 may be a new link between the immune and nerval system by regulating inflammation as well as itch. This may also suggest that IL-31 and its signaling pathway represent a novel target for antipruritic therapy (Sonkoly et al., 2006).
INF- In a double-blind study, a beneficial effect of IFN- on itch responses has been clearly demonstrated also in AD patients. Pruritus was reduced by 50% even 1—2 years after long-term treatment with recombinant human IFN-. However, the mode of action and the receptor density of IFN receptors on sensory nerves in the skin is unknown.
Summary and perspectives
Itch research not only has proposed a variety of potential itch mediators but also identified non-neuronal cells contributing to the pathophysiology of pruritus. Being confronted with a variety of potential pruritic mediators generated by close interaction of neuronal and non-neuronal cells three major directions for future research arise:
First — which is the crucial pruritic mediator in a particular skin disease and how can it be modulated/suppressed? Second — what is the basis for an inflammatory process to release itch mediators and sensitize itch receptors rather than to produce pain and sensitization to pain? Third, which mechanisms can be used to modulate neuronal itch processing in the periphery, the spinal cord, and the central nervous system? The insight in anatomical and physiological structures as provided by central imaging will help to identify the involved central areas. Consequently, both, animal studies and clinical trials will further dissect the interaction and communication between the central and the peripheral nervous system under physiological and pathophysiological conditions. Here also, novel molecular imaging approaches may help to better understand the complexity within the central nervous system when processing itch information.

With respect to novel therapies, the potential redundancy of many pruritic signals complicates specific treatment (Figure 3, Table S1). Thus, it will be a major task to identify convergent peripheral and central targets in the itch pathway to increase the chances of clinical success.

Therefore, future research will have to clarify underlying mechanisms that induce, modulate, and transmit the itch signal on a molecular level to finally identify the essential molecules as targets for antipruritic therapy. Moreover, understanding perception and regulation of itch signals within the central nervous system will lead to novel strategies for the treatment of pruritus.

Source, Graphs and Figures: Journal of Investigative Dermatology - Neurophysiological, Neuroimmunological, and Neuroendocrine Basis of Pruritus
Top Bottom