The involvement of a cyclooxygenase 1 gene-derived protein in the antinociceptive action of paracetamol in mice

Abstract

Paracetamol is a widely used analgesic and antipyretic with weak anti-inflammatory properties. Experimental evidence suggests that inhibition of prostaglandin biosynthesis contributes to its pharmacological actions. Three cyclooxygenase (COX) isoenzymes are involved in prostaglandin biosynthesis, COX-1, COX-2 and a recently discovered splice-variant of COX-1, COX-3. Our aim was to identify the relative roles for these enzymes in the antinociceptive action of paracetamol in mice. We compared the antinociceptive action of paracetamol with the non-selective non-steroid anti-inflammatory drug, diclofenac and studied paracetamol antinociception in COX-1 and COX-2 knockout mice.

Paracetamol (100–400 mg/kg) inhibited both acetic acid- and iloprost-induced writhing responses. In contrast, diclofenac (10–100 mg/kg) inhibited only acetic acid-induced writhing. Only diclofenac reduced peripheral prostaglandin biosynthesis whereas both drugs reduced central prostaglandin production. Prostaglandin E₂ (PGE₂) concentrations were reduced in different brain regions by administration of paracetamol. COX-1, COX-2 and COX-3 enzyme proteins were expressed in the same brain regions. The effects of paracetamol on writhing responses and on brain PGE₂ levels were reduced in COX-1, but not COX-2, knockout mice. The selective COX-3 inhibitors, aminopyrine and antipyrine also reduced writhing responses and brain PGE₂ biosynthesis. These results suggest that the antinociceptive action of paracetamol may be mediated by inhibition of COX-3.

Introduction

Three isoforms of the prostaglandin synthesising enzyme, cyclooxygenase (COX), have been characterised, COX-1, COX-2 and COX-3 (DeWitt and Smith, 1988, Xie et al., 1991, Chandrasekharan et al., 2002). COX-1, generally expressed constitutively, is responsible for the synthesis of prostaglandins involved in the regulation of physiological functions (Crofford, 1997). Thus, COX-1 is regarded as a ‘house-keeping’ enzyme, unlike COX-2, which is induced by various pro-inflammatory stimuli (Jones et al., 1993) and in pathological conditions such as inflammation, pain and fever (Vane et al., 1994, Yamamoto and Nozaki-Taguchi, 1996, Cao et al., 1996). COX-3 is a splice-variant of COX-1 that retains the intron-1 gene sequence at the mRNA level which encodes a 30 aa sequence inserted into the N-terminal hydrophobic signal peptide of the enzyme protein. In the dog, COX-3 protein was expressed predominantly in the central nervous system (CNS) and in the heart (Chandrasekharan et al., 2002). Paracetamol selectively inhibited this COX-3 protein in preference to COX-1 and COX-2. In the canine, COX-3 mRNA intron-1 is within frame, but in humans and rodents there is a frameshift mutation in intron-1 of the COX-3 transcript. The mechanism for conversion of COX-3 mRNA to active enzyme protein in mice is therefore not clear as yet. However, in rat tissues, an approximately 64 kDa protein was detected using antibody against the intron-1 moiety of a COX-1 variant protein (Snipes et al., 2005). Also, an intron-retained COX-1 protein was identified in human tissues (Qin et al., 2005). Although this human COX-1 variant (named COX-1b₂) synthesised prostaglandin F_2α from arachidonic acid, its activity was not selectively inhibited by paracetamol (Qin et al., 2005).

Prostaglandin E₂ (PGE₂), is a potent hyperalgesic mediator at the site of injury (Ferreira et al., 1973), in the spinal cord (Yaksh and Malmberg, 1993, Malmberg and Yaksh, 1994) and in the brain (Hori et al., 2000, Abe et al., 2001). An increased release of PGE₂ into the splenic vein of an anaesthetised dog was detected after a nociceptive injection of bradykinin into the spleen (Ferreira et al., 1973). Subsequently, PGE₂ release from spinal cord slices in vitro was reported after stimulation with capsaicin or K ⁺(60 mM) (Malmberg and Yaksh, 1994, Dirig et al., 1997). PGE₂ was also collected by microdialysis with a probe implanted in the spinal cord of rats in vivo. The concentrations of PGE₂ in the dialysate increased after an injection of formalin into the hind paw of the rat (Malmberg and Yaksh, 1995, Muth-Selbach et al., 1999, Tegeder et al., 2001) or after an intraspinal injection of substance P (Yaksh et al., 2001). A hyperalgesic, systemic injection of bacterial lipopolysaccharide (Abe et al., 2001) raised the levels of PGE₂ in the preoptic area of the rat hypothalamus and PGE₂ concentrations in mouse brains fell after systemic administration of cyclooxygenase inhibitors (Ferrari et al., 1990).

Paracetamol is a widely used analgesic and antipyretic with weak anti-inflammatory effects (Clissold, 1986). Although its mode of pharmacological action has not been fully elucidated, paracetamol is generally believed to act centrally rather than by a peripheral action (Clissold, 1986). The initial report of Flower and Vane (1972), describing an inhibitory effect of paracetamol on COX activity in the CNS is now supported by a number of other studies (Malmberg and Yaksh, 1994, Muth-Selbach et al., 1999, Ferrari et al., 1990). For example, in mice, antinociceptive doses of paracetamol attenuated the ex vivo synthesis of brain PGE₂ in a dose-related manner (Ferrari et al., 1990) and in human volunteers, paracetamol raised the threshold to painful electrical stimulation of the sural nerve (Piletta et al., 1991). In addition, in the formalin test in the rat, antinociceptive doses of paracetamol reduced flinching behaviour and release of PGE₂ from the spinal cord (Muth-Selbach et al., 1999). In spite of its potent inhibition of the biosynthesis of prostaglandins in the CNS in vivo (Muth-Selbach et al., 1999, Ferrari et al., 1990, Flower and Vane, 1972, Grèen et al., 1989), paracetamol was shown to be a weak inhibitor of both COX-1 and COX-2 in vitro (Warner et al., 1999). It has been suggested that the recently discovered COX-1 variant, namely COX-3, may be the target for the antinociceptive action of paracetamol, which inhibits the activity of COX-3 at therapeutic concentrations in vitro (Chandrasekharan et al., 2002).