Cathepsin S is a cysteine protease linked to inflammatory processes including atherosclerosis and asthma. The possibility that this or other cysteine proteases might evoke itch or be part of a classical ligand-receptor signaling cascade has not been considered previously. We show here that human cathepsin S evokes itch and activates human protease-activated receptors 2 and 4.
The sensation of itch is mediated by two distinct non-overlapping populations of cutaneous nerve fibers that evoke comparable degrees of itch (Namer et al., 2008; Johanek et al., 2008). One set of fibers, the mechano-insensitive population, is more responsive to histamine than to cowhage. The other set is mechanosensitive and is more responsive to cowhage than to histamine (Namer et al., 2008; Johanek et al., 2008). Histamine is a classical mediator of itch and is associated with a wheal and flare. Since most clinical itches do not have a wheal or flare and do not respond to antihistamines, histamine is not thought to contribute to most itches (Ikoma et al., 2006). Cowhage refers to a tropical legume or, in this case, the loose hairs that cover the pods of Mucuna pruriens, and evoke itch. The active component of cowhage is mucunain, a cysteine protease that serves as a ligand for protease-activated receptors (PARs) 2 and 4 (Reddy et al., 2008). The identification of an endogenous mediator with properties similar to cowhage could lead to insights into non-histamine mediated itch. We focused on human cathepsin S because it shares active site sequence homology with mucunain and is selectively up-regulated in human keratinocytes upon stimulation with interferon-gamma, consistent with a possible pruritic role in inflammatory skin disease (Schwarz et al., 2002).
There are fifteen human cathepsins, including eleven cysteine, two aspartic and two serine proteases. Cathepsins were traditionally considered lysosomal proteases. It is now recognized that the broad expression and range of pH dependence of some cathepsins reveal that they have many functional roles including tissue remodeling, metastasis and inflammation. Examples of cysteine cathepsin activities include cleavage of collagen by cathepsin L to generate endostatin (Felbor et al. 2000), an endogenous inhibitor of angiogenesis, and cleavage of the invariant chain in antigen presenting cells by cathepsin S (Nakagawa et al., 1999) as part of the inflammatory cascade.
There are four PARs and they are members of the G-protein coupled receptor family. Their identified endogenous activators are all serine proteases. These proteases trigger the activation of PARs by unmasking ‘tethered ligand’ sequences near the N-termini of the receptors. Certain kallikrein-related peptidases and mast cell tryptase, which are serine proteases, can activate PAR2 in vitro (Oikonomopoulou et al., 2006, Stefansson et al., 2008) but they have not been demonstrated to be endogenous pruritic agents. Serine proteases and PAR2 have also been linked to barrier function (Hachem et al., 2006). The presence of PAR2 on free nerve endings in the skin, keratinocytes and dorsal root ganglia link this receptor to itch and pain (Steinhoff et al., 2000, Steinhoff et al., 2003, and Shpacovitch et al., 2008). Data on PAR4 are more limited, revealing that it is expressed in rat dorsal root ganglia and its activation, at least in this species, appears to be antinociceptive (Asfaha et al., 2007). A PAR4 hexapeptide agonist has recently been reported to cause scratching in mice (Tsujii et al., 2008).
Cathepsin S and mucunain are endogenous and exogenous cysteine proteases, evoke similar itch and nociceptive sensations, and serve to activate PAR-2 and 4 (Figure 1, Figure 2 and Reddy et al., 2008). The relative contribution of PAR2 versus PAR4 activation in cathepsin S-induced sensations is not addressed by these data. Cathepsin S has been implicated in neuropathic pain, manifest by increased gene expression in rat dorsal root ganglia following peripheral nerve injury. Such pain was lessened by a cathepsin S inhibitor (Barclay et al., 2007). These observations suggest that cathepsin S may have an excitatory effect on neurons. As keratinocytes have been demonstrated to express cathepsin S (Schwarz et al., 2002), a role for this protease in keratinocyte-neuronal communication may be expected. Cathepsin S may contribute to the pruritus of inflammatory skin conditions including atopic dermatitis and psoriasis and could have a role in barrier function. Cathepsin S and other endogenous or exogenous cysteine proteases may activate PARs as part of additional inflammatory processes. For example, Der p1, a mite cysteine protease associated with asthma activates PAR2 (Asokananthan et al., 2002). Taken together, we suggest that exogenous and endogenous cysteine proteases interact with PARs as part of signaling cascades in homeostasis and disease.