As described above, PAR2 is expressed in pancreatic acinar cells (Kawabata et al., 2002b) and ductal epithelium (Nguyen et al., 1999), and its activation stimulates pancreatic juice secretion (Kawabata et al., 2000d). Although PAR2 might not play critical roles in pancreatic exocrine secretion under physiological conditions, increasing evidence suggests the emerging roles played by PAR2 during pancreatitis (Olejar et al., 2001; Namkung et al., 2004; Maeda et al., 2005; Sharma et al., 2005a; Matej et al., 2006; Kawabata et al., 2006b; Singh et al., 2007). PAR2 expression in the pancreas appears to increase during taurocholate-induced acute pancreatic lesion development in rats, although the physiological relevance of PAR2 upregulation remains to be determined in this model (Olejar et al., 2001). Systemic administration of PAR2 agonists suppresses caerulein-induced acute pancreatitis in rats and mice (Namkung et al., 2004; Sharma et al., 2005a; Kawabata et al., 2006b). PAR2-deficient mice exhibit more severe inflammatory signs than wild-type animals in a relatively severe pancreatitis model induced by 12 hourly injections of caerulein at 50 μg kg−1 (Sharma et al., 2005a), suggesting a protective role for activation of PAR2 by endogenous proteinase such as trypsin. However, the difference in the severity of inflammatory symptoms between PAR2-deficient and wild-type animals is not clear in a mild pancreatitis model induced by 6 hourly injection of caerulein at the same dose (Kawabata et al., 2006b). The protective mechanisms for PAR2 in pancreatitis appear to, at least in part, involve inhibition of translocation of phosphorylated ERK to the nucleus in pancreatic cells (Sharma et al., 2005a), although phosphorylation of ERK in whole cells is unaffected or rather facilitated by PAR2 activation (Namkung et al., 2004; Sharma et al., 2005a). The most recent evidence indicates that trypsin released during the early stages of pancreatitis activates PAR2 on the acinar cells and stimulates secretion of digestive enzymes including trypsinogen from these cells, leading to decreased intrapancreatic enzyme levels and limitation of the severity of pancreatitis (Singh et al., 2007) (Figure 4). In contrast, a study using anti-PAR2-antibodies implies a pro-inflammatory role for PAR2 in caerulein-induced pancreatitis in rats (Maeda et al., 2005), which is inconsistent with evidence from studies employing PAR2-activating peptides and PAR2-knockout mice. Although the discrepancy has yet to be explained, it is likely that trypsin-induced activation of PAR2 present in intrapancreatic sensory neurons (Steinhoff et al., 2000; Hoogerwerf et al., 2001) might promote inflammation, since pancreatitis appears to involve neurogenic inflammation (Nathan et al., 2001, 2002; Hutter et al., 2005). There are plenty of clinical and fundamental studies showing that inhibitors of pancreatic proteinases that are capable of activating PAR2 improve acute pancreatitis (Iwaki et al., 1986; Otsuki et al., 1990; Harada et al., 1991; Takeda et al., 1996; Chen et al., 2000; Maeda et al., 2005; Ishikura et al., 2007).
Figure 4. A scheme for roles of PAR2 expressed on sensory neurons and acinar cells during pancreatitis. Trypsin, released from acinar cells in response to cytotoxic stimulation such as caerulein, causes pancreatic cell damage by proteolytic digestion, and would also activate neuronal PAR2, leading to pancreatic pain. In the early stages of pancreatitis, trypsin could activate PAR2 on the acinar cells and decrease intrapancreatic trypsin levels by stimulating exocrine secretion of trypsinogen into the duodenum, limiting the extent of pancreatitis and related pain. Unknown non-PAR2 receptors on sensory neurons should mediate pancreatic pain in response to trypsin and/or unknown nociceptive mediators, derived from acinar cells, particularly in PAR2-knockout mice.
Download figure to PowerPoint
Clinically, acute pancreatitis is accompanied with a sharp and severe pain from the upper abdominal area to the back, and treatment of the pancreatitis-related pain is very important. Apart from pancreatitis itself, PAR2 expressed in sensory neurons is involved in pancreatic pain (Hoogerwerf et al., 2001, 2004; Kawabata et al., 2006b; Ishikura et al., 2007). Administration of PAR2-activating peptides and trypsin into the pancreatic duct causes activation of nociceptive neurons, as measured by expression of Fos protein, in the superficial layers of the thoracic spinal cord in anesthetized rats, and induces a behavioural pain response in awake rats (Hoogerwerf et al., 2001, 2004; Ishikura et al., 2007). The ductal trypsin-evoked spinal Fos expression can be blocked by pretreatment with camostat mesilate, a proteinase inhibitor (Ishikura et al., 2007). The mice with mild pancreatitis caused by 6 hourly repeated systemic administration of caerulein exhibit referred hyperalgesia in the skin of the upper abdomen. This referred hyperalgesia during the mild pancreatitis can be abolished by not only repeated but also single administration of the proteinase inhibitor, camostat mesilate (Ishikura et al., 2007), and nafamostat mesilate (Kawabata et al., a manuscript in preparation). This suggests a possibility that endogenous proteinases including trypsin might directly stimulate PAR2 present in intrapancreatic sensory neurons during pancreatitis, resulting in pancreatic pain/referred hyperalgesia (Figure 4). Nonetheless, the referred hyperalgesia during the pancreatitis in PAR2-knockout mice is more severe than that in wild-type animals, while the inflammatory symptoms in this mild pancreatitis model are not significantly different between the PAR2-knockout and wild-type animals (Kawabata et al., 2006b). Further, repeated co-administration of PAR2-activating peptides with caerulein suppressed the referred hyperalgesia in wild-type animals, but not PAR2-knockout mice. Thus, the role of PAR2 in pancreatitis-related pain is very complex. One possibility is, as mentioned above, that trypsin released during the early stages of pancreatitis might stimulate PAR2 on the acinar cells and decrease intrapancreatic levels of nociceptive mediators including trypsin through enhancement of exocrine secretion of acinar cell contents such as trypsinogen into the duodenum (Figure 4). In PAR2-knockout mice, however, receptors other than PAR2 expressed in intrapancreatic sensory neurons should mediate the actions of trypsin and/or the other unknown nociceptive messengers present in the acinar cells (Figure 4). It is likely that PAR4 might mediate the nociceptive actions of trypsin and kallikrein released from the acinar cells, since PAR4 can be activated directly by those acinar cell enzymes (Kawabata, 2002; Ossovskaya and Bunnett, 2004; Oikonomopoulou et al., 2006a, 2006b). Bradykinin B2 receptors could also mediate the nociception through the kallikrein–bradykinin pathway, known to be activated during pancreatitis (Griesbacher and Lembeck, 1992; Griesbacher et al., 2002), since even a single dose of HOE-140, a B2 receptor antagonist, partially inhibited the established referred hyperalgesia during pancreatitis in mice (Kawabata et al., unpublished data). These hypotheses have yet to be evaluated by more in-depth studies. Together, proteinase inhibitors and PAR2 antagonists, if available, might be clinically useful for the treatment of pain accompanying established acute pancreatitis, although the use of PAR2 antagonists might not be recommended in the early stages of acute pancreatitis. Interestingly, there is clinical evidence that proteinase inhibitors such as nafamostat mesilate and gabexate mesilate are highly effective against established acute pancreatitis-related pain (Harada et al., 1991; Takeda et al., 1996; Chen et al., 2000).