The results of this study show that ET-1 is a very potent pruritogen in mice, causing dose-dependent scratching, the behavioural correlate of itch, with a mean apparent ED50 of 1 pmol (2.5 ng). ET-1 is therefore about six million times more potent as a pruritogen than histamine, which has an apparent ED50 of 5.8 μmol in this strain of mice (Bell et al., 2004). The log dose–response plot for ET-1-induced scratching was bell-shaped and cumulative dosing with the peptide led to desensitization of the scratch reflex.
Two previous studies found that ET-1 causes pruritus when administered to human skin. Both Ferreira et al. (1989) and Katugampola et al. (2000) described the development of a burning pruritus in human subjects following intradermal injection of ET-1. The present experiments in mice used an automated repetitive movement detector (Brash et al., 2005) to measure accurately the scratching induced by injecting ET-1, and we thought this was the first report of the pruritogenic action of ET-1 in BALB/c mice. However, after the work was submitted for publication, a paper was published which demonstrated that ET-1 causes pruritus in mice via similarly potent actions involving ETA receptors in murine skin (Trentin et al., 2006).
Scratching in animals as a behavioural correlate of itch
It is impossible to determine what sensation any animal experiences, so behavioural correlates have to be used. Animals may scratch in response to painful stimuli as well as to those that are associated with different human sensations, such as itch and tickle (McMahon and Koltzenburg, 1992). However, it has been shown that whereas the pruritogenic agents compound 48/80 and substance P cause scratching when injected subcutaneously in mice, the algogenic (pain producing) agents capsaicin and formalin do not, but instead usually cause vocalization, biting and withdrawal (Kuraishi et al., 1995). We found that histamine or 5-HT caused dose-related scratching in mice, whereas injection of algogens evoked slight scratching which did not differ in intensity from that induced by PBS (Bell et al., 2004). Injection of histamine into human skin is invariably associated with the sensation of itch and reflex scratching (Keele and Armstrong, 1964), and even very high doses of ET-1 (up to 50 μg i.d.) only caused itch (Ferreira et al., 1989). We therefore consider it probable, by analogy with the results from man, that mice in the present study were scratching in response to ‘itch’ rather than ‘pain’ evoked by low doses of ET-1. It is the reflex scratching which causes damage to the skin in response to pruritogens both in man and mice, and that behaviour can be studied objectively in animals.
Involvement of ETA and ETB receptors in ET-1-induced scratching
The selective ETA antagonist, BQ123 (Ihara et al., 1992), significantly reduced scratching evoked by a mid-range dose of ET-1 (Figure 3), which strongly suggests the involvement of ETA receptors in ET-1-induced itch. The selective ETB agonist, IRL-1620 (Takai et al., 1992), also induced scratching, which could be interpreted as evidence for involvement of ETB receptors in the scratch reflex response to ET-1. However, the ED50 for the scratch response to IRL-1620 was 875 times greater than that of ET-1 (Figure 4a), and only scratching evoked by the highest dose of IRL-1620 was significantly greater than the vehicle-induced response. It therefore seemed probable that the high dose of IRL-1620 evoked scratching via weak agonist action at ETA receptors (Takai et al., 1992). This was confirmed by showing that the ETA antagonist, BQ123, significantly reduced scratching in response to IRL-1620 (Figure 4b).
The possibility existed that BQ123 might exert its antipruritic action through sedation, as occurs with sedating histamine H1 receptor antagonists (Nicholson, 1983; Shuster, 1988). However, this was rejected because there is no evidence that ETA antagonists such as BQ123 cause sedation, and we did not see any change in spontaneous exploration or grooming behaviour following BQ123 administration.
Taken together, this functional integrative behavioural evidence from intact animals suggested that ET-1 acts via ETA receptors to evoke reflex scratching. Further studies were undertaken using the ETB antagonist, BQ788, in the expectation that these would provide additional evidence that IRL 1620 evokes scratching via an ETA receptor-dependent mechanism, that is, they would serve as a negative control. In fact, we found that scratching evoked by IRL1620 was antagonized by BQ788, whereas that caused by ET-1 was unaffected. This evidence therefore may mean that ETB receptors are involved in scratching evoked by the ETB agonist, IRL1620, but not in scratching caused by ET-1, which is an endogenous non-selective ETA and ETB receptor agonist.
Trentin et al. (2006) used male Swiss mice and measured bouts of scratching evoked by ET-1 (1–20 pmol). Scratching was antagonized by co-injecting the ETA antagonist BQ123 into the skin, or by i.p. administration of a different ETA receptor antagonist, atrasentan. Our results with ET-1 and BQ123 in female BALB/c mice confirm some of the findings of Trentin et al. (2006) and provide additional evidence that ET-1 is a potent pruritogen in mice, apparently acting via ETA receptors in the skin. These authors also reported that IRL1620 (10 pmol) was not pruritic, and that accords with our findings using this dose (Figure 4a). However, we observed that higher doses of the ETB agonist are pruritic, whereas they found that, when co-injected, it was antipruritic, inhibiting scratching evoked by ET-1 or histamine, an effect we did not investigate.
Our results showed that systemic administration of the ETB antagonist, BQ788, did not affect the scratch response to the non-selective ETA/B receptor agonist ET-1 (Figure 5a), in contrast to the finding of Trentin et al. (2006) that co-injecting BQ788 enhanced scratching evoked by ET-1, suggesting that ETB receptors inhibit scratching. Our results are thus discrepant, but Trentin et al. (2006) did comment on their unexpected finding that systemic administration of the ETB antagonist A-1292621 diminished scratching induced by ET-1.
Attempting to characterize receptors on the basis of behavioural studies involving high doses of agonists and antagonists is complicated by the issue of the selectivity of drugs for receptors in particular species. Thus, we found that only high doses of the ETB agonist IRL1620 cause scratching, which was antagonized by the ETA receptor antagonist BQ123 and also by the ETB receptor antagonist BQ788. This suggests that ETB receptors, as well as ETA receptors are involved in evoking scratching, but that could be a misinterpretation if high doses of agonist and/or antagonist are acting non-selectively in mice to affect ETA receptors. We did not co-inject drugs locally, and the situation is complicated by the possibility that systemic and local effects of a drug will differ in vivo. Also, as suggested by the delay to the start of scratching, ET-1 may be acting indirectly on pruriceptors via release of pruritogenic mediators (e.g. histamine from mast cells; Yamamura et al., 1994b). It seems probable that ETA and ETB receptors interact via receptor ‘crosstalk’ (Mickley et al., 1997), which makes it difficult to establish what contribution is made by the individual receptor subtypes in response to the endogenous ligand, ET-1. Further studies using in vitro techniques to establish the location and the type(s) of ET receptor involved in evoking the scratching reflex are clearly needed.
Mechanisms of ET-1-induced itch: possible indirect action via histamine release
The relatively long delay to onset of scratching following ET-1 injection into the skin is suggestive of an indirect action of the peptide. Several lines of evidence support the involvement of histamine in ET-1-induced pruritus. When injected into human skin ET-1 evokes pallor at the injection site, with flare in the surrounding tissue of similar character to that evoked by histamine (Brain et al., 1992). This flare response is markedly reduced by H1 receptor antagonists (Crossman et al., 1991; Brain et al., 1992). ET-1-induced release of substance P from primary afferent neurons in human skin has been proposed (Bunker et al., 1992), and substance P induces itch via histamine-dependent and independent mechanisms in both humans (Hagermark et al., 1978; Hagermark, 1992) and mice (Andoh et al., 1998).
Research on murine foetal skin-derived mast cells shows the presence of ETA receptors on their surface and ET-1 degranulates these cells in a dose- and ETA receptor-dependent manner (Matsushima et al., 2004). Histamine release caused by ET-1 has also been demonstrated in both peritoneal and bone marrow-derived mast cells harvested from BALB/c mice (Yamamura et al., 1994a, 1994b). It is thus possible that ET-1 degranulates mast cells in BALB/c mice and the delayed onset of itch is caused by liberated histamine. However, against this possibility, Brain et al. (1992) described the inability of ET-1 to activate human skin mast cells in vitro, and Katugampola et al. (2000) found only a minimal rise in histamine in the skin microvasculature after infusing ET-1, suggesting that histamine release is not important for the development of ET-1-induced pruritus in humans. A reciprocal interaction between ET-1 and mast cells has been reported (Hultner and Ehrenreich, 2005) and it will be of considerable interest to perform more detailed studies in normal and genetically modified mice, for example. mast cell deficient (Wershil and Galli, 1994), to establish whether or not ET-1 acts via ETA and possibly ETB receptors to trigger scratching indirectly, through release of histamine or other mediator(s) from neurones, glia, immune cells and blood vessels in the skin (Pomonis et al., 2001).
Direct activation of pruriceptive C-fibres
Although the overall onset of scratching following injection of ET-1 or IRL1620 was significantly slower in comparison with the pruritogens histamine and 5-HT, several mice began scratching quite rapidly after injection, and were close to the median delay for histamine or 5-HT (Figure 2). The high variance in time to onset of response may reflect the time taken for injected substances such as ET-1 and histamine to reach a threshold concentration for activating pruriceptive C-fibres in a particular receptive field, which may result from slight differences in the positioning of the needle in the skin. The fact that scratching started fairly soon after injection in some animals could be interpreted as direct action of ET-1 on ETA receptors located on the small-diameter pruriceptive C-fibres described by Schmelz et al. (1997). ET-1 binding sites have been identified within the dorsal root ganglia (DRG) of the rat, rabbit and monkey, with ETA receptors localized to small diameter DRG neuronal cell bodies (Kar et al., 1991; Pomonis et al., 2001), which may be involved in nociception or pruriception. Moreover, colocalization of ETA receptors and the C-fibre marker, CGRP, has been demonstrated on these cell bodies. This evidence points to the presence of ETA receptors on neurones in DRG which might be pruriceptive, but this is speculative. Detailed neuropharmacological investigation is required to identify mechanically insensitive pruriceptive C-fibres in murine skin which respond to low doses of ET-1 in an ETA and possibly ETB receptor-dependent manner.