The critical role of leukotriene B₄ in antigen-induced mechanical hyperalgesia in immunised rats

Abstract

1. We investigated the mediators responsible for mechanical hypersensitivity induced by antigen challenge in rats immunised with ovalbumin (OVA).

2. Challenge with OVA (12.5–100 μg, intraplantar) caused a dose- and time-dependent mechanical hypersensitivity, which peaked 3 h after, decreased thereafter and reached control levels 24 h later.

3. Levels of TNFα, IL-1β and cytokine-induced neutrophil chemoattractant 1 (CINC-1) were increased in paw skin after antigen challenge.

4. OVA-evoked hypersensitivity was partially inhibited (about 51%) by pretreatment with anti-TNFα , IL-1β and IL-8 sera or with IL-1 receptor antagonist (IL-1ra), but not anti-NGF serum. Pretreatment with thalidomide (45 mg kg⁻¹) or pentoxifylline (100 mg kg⁻¹) also partially inhibited the hypersensitivity at 1–3 h after challenge.

5. Pretreatment with indomethacin (5 mg kg⁻¹) or atenolol (1 mg kg⁻¹) reduced the OVA-induced hypersensitivity at 1 and 3 h, but not at 5 h after challenge, while the combination of B₁ and B₂ bradykinin receptor antagonists was ineffective over the same times.

6. Pretreatment with MK886 (5-lipoxygenase-activating protein inhibitor, 3 mg kg⁻¹), CP 105696 (LTB₄ receptor antagonist; 3 mg kg⁻¹) or dexamethasone (0.5 mg kg⁻¹) inhibited the hypersensitivity from 1 to 5 h. Furthermore, LTB₄ levels were increased in the paw skin of challenged rats.

7. In conclusion, our results suggest that the TNFα-, IL-1β- and CINC-1-driven release of prostaglandins, sympathetic amines and LTB₄ mediates the first 3 h of mechanical hypersensitivity induced by antigen challenge in rats. At 5 h after OVA administration, although TNFα has some role, LTB₄ is the critical nociceptive mediator.

Introduction

The sensitisation of nociceptors is the common denominator of inflammatory pain and represents a functional upregulation of nociceptors that leads to a state known as hyperalgesia/allodynia. Following this event, previously mild or ineffective stimuli cause ‘overt pain' in humans, or a characteristic behavioural response that is used as an end point in animal nociceptive tests (Handwerker, 1976; Perl, 1976). There are two groups of directly acting hyperalgesic mediators that satisfy the experimental and clinical criteria for agents that directly sensitise nociceptors: eicosanoids and sympathetic amines. The capacity of prostaglandins and sympathetic amines (noradrenaline and dopamine) to sensitise nociceptors has been shown in man and in animals using both behavioural and electrophysiological techniques (Hannington-Kiff, 1974; Lol & Nathan, 1978; Lol et al., 1980; Ferreira, 1983; Nakamura & Ferreira, 1987; Duarte et al., 1988).

There is a great deal of evidence in the literature that in inflammation induced by carrageenin or lipopolysaccharide (LPS) administration, the release of eicosanoids and sympathetic amines is secondary to the generation of bradykinin (BK), which stimulates TNFα production. This stimulates two pathways, each of which leads to the release of cytokines and the final nociceptive mediators that sensitise the nociceptor. The two pathways are: (i) carrageenin/LPS → BK → TNFα → IL-6 → IL-1β prostaglandins (which sensitise nociceptors) and (ii) carrageenin/LPS → BK → TNFα → CINC-1 (rat IL-8-related chemokine; Watanabe et al., 1991) → sympathetic amines that also sensitise the nociceptors. Besides the mediators described above, platelet-activating factor leukotriene B₄ (LTB₄) and endothelin also participate in the genesis of inflammatory hyperalgesia (Bonnet et al., 1981; Levine et al., 1984; Ferreira et al., 1989, respectively). The LTB₄-induced hyperalgesia appears to be dependent on polymorphonuclear leukocytes, but independent of prostaglandin synthesis (Levine et al., 1984; Bisgaard & Kristensen, 1985).

Regarding the inflammatory reaction to antigen challenge, it is well accepted that activation of the immune system leads to the release of proinflammatory cytokines such as TNFα, IL-1β, IL-6 (Nathan, 1987; Decker, 1990), IL-18 (Gracie et al., 1999) and chemokines (König et al., 2000), which in turn, leads to the release of a series of inflammatory mediators, including eicosanoids (for review, see Bayon et al., 1998). Using different experimental models, it was shown that, among these mediators, TNFα (Nordahl et al., 2000), IL-1β (Alstergren et al., 1998; Kopp, 1998), prostaglandins (Portanova et al., 1996; Omote et al., 2002) and also NGF, a neurotrophic factor (Ma & Woolf, 1997), are involved in the onset of the hyperalgesia that follows challenge with antigen. However, there is no study, in a single model of inflammation following antigen challenge, of the individual contribution of the above-mentioned mediators to the onset of mechanical hyperalgesia. We have, therefore, examined the roles of BK, TNFα , IL-1β , CINC-1, NGF, prostaglandins and sympathetic amines in hyperalgesia in immunised rats, following antigen challenge. Recently, our group demonstrated that neutrophil migration induced by antigen challenge in mice depends on TNFα released by CD4+T cells, which acts through an LTB₄-dependent mechanism (Canetti et al., 2001). However, since the possible involvement of LTB₄ in hyperalgesia induced by antigen challenge remained to be investigated, we also addressed this point in the present study.

Methods

Animals

Male Wistar rats (100–200 g) were housed in a temperature-controlled room, with access to water and food ad libitum, until use. All experiments were conducted in accordance with NIH guidelines on the welfare of experimental animals and with the approval of the Ethics Committee of the School of Medicine of Ribeirão Preto (University of São Paulo).

Procedures for active immunisation

OVA was dissolved in phosphate-buffered saline (PBS) to an appropriate concentration (2 mg ml⁻¹) and mixed with an equal volume of complete Freund's adjuvant (CFA) at a concentration of 1 mg ml⁻¹ of Mycobacterium tuberculosis in 85% paraffin oil and 15% mannide monoleate. CFA was used to augment the efficiency of the immunisation procedure (Freund, 1956) by prolonging the lifetime of injected autoantigen and by stimulating its effective delivery to the immune system; this results in altered leukocyte proliferation and differentiation (for a review, see Billiau & Matthys, 2001). Rats weighing approximately 100 g were injected subcutaneously at two different sites on their back to give a total dose of 200 μg of OVA dissolved in an emulsion containing an equal volume of PBS plus CFA. Control (sham immunised) rats were injected with this emulsion without OVA. After 21 days, the rats were challenged by the intraplantar (i.pl.) administration of OVA (at different doses dissolved in 100 μl of PBS) to one of the hindpaws.

Nociceptive test: mechanical nociceptive hypersensitivity

Mechanical nociceptive hypersensitivity was tested in rats as described previously (Ferreira et al., 1978). In this method, a constant pressure of 20 mmHg (measured using a sphygmomanometer) is applied (via a syringe piston moved by compressed air) to a 15-mm² area on the dorsal surface of the hindpaw, and discontinued when the rat presents a ‘freezing reaction'. This reaction typically is comprised of brief apnoea, concomitant with retraction of the head and forepaws and a reduction in the escape movements that animals normally make to free themselves from the position imposed by the experimental situation. Usually, the apnoea is associated with successive waves of muscular tremor. For each animal, the latency to the onset of the freezing reaction is measured before administration (zero time) and at different times after administration of the hyperalgesic agents. The intensity of mechanical hypersensitivity is quantified as the reduction in the reaction time, calculated by subtracting the value of the second measurement from the first (Ferreira et al., 1978). Reaction time was 30.2±0.5 s (mean±s.e.m.; n=15) before injection of the hyperalgesic agents or after intraplantar injection of saline. The shortened reaction time observed after inflammatory stimulus injection is prevented by steroidal and nonsteroidal anti-inflammatory drugs (NSAIDs) (Ferreira et al., 1988; 1997; Cunha et al., 1991; 1992; Lorenzetti et al., 2002). This method has been used to demonstrate the contribution of eicosanoids, sympathetic amines, cAMP and cytokines to the development of peripheral inflammatory hyperalgesia (Ferreira & Nakamura, 1979a; Ferreira et al., 1988; 1993; Francischi et al., 1988; Cunha et al., 1991; 1992; 1999), as well as the peripheral analgesic effect of opiates (Ferreira & Nakamura, 1979b) and NSAIDs (Ferreira et al., 1988; Cunha et al., 1991; 1992). These concepts and findings have been extensively confirmed with other methodologies such as formalin-induced flinching (Vinegar et al., 1976; Choi et al., 2001; Granados-Soto et al., 2001), chemically induced writhing (Duarte et al., 1988; Follenfant et al., 1988; Ribeiro et al., 2000b), the classical Randall-Sellito method (Aley et al., 1995) and others (Vinegar et al., 1976; Safieh-Garabedian et al., 2002). Furthermore, this method is able to discriminate between peripheral and central analgesic effects of drugs (Ferreira et al., 1978; Duarte & Ferreira, 1992).

The nociceptive hypersensitivity was measured at the indicated times after intraplantar injection of OVA (12.5–100 μg paw⁻¹, in 100 μl), keyhole limpet haemocyanin (KLH, 25 μg paw⁻¹, in 100 μl) or PBS (100 μl) into the hindpaw of rats. Different individuals were responsible for preparing the solutions to be injected, performing the injections and measuring the reaction times. Multiple paw treatments with PBS did not alter basal reaction time, which was similar to that observed in noninjected paws. Moreover, despite the fact that there is evidence in the literature that intraplantar administration of CFA in naive mice induces long-lasting mechanical hypersensitivity in both paws (Ferreira et al., 2001), in our experimental model the subcutaneous administration of CFA to the back of the animals did not alter basal reaction times.

Experimental protocols

Hyperalgesic effect of intraplantar injection of OVA in immunised rats

Mechanical hypersensitivity was measured 1, 3, 5, 7, 12 and 24 h after the administration of OVA (12.5, 25, 50 or 100 μg paw⁻¹) injected in a final volume of 100 μl into the hindpaw of immunised rats. Sham-immunised animals received i.pl. injection of OVA (100 μg paw⁻¹, in 100 μl). The OVA-challenge was performed 21 days after immunisation. As a control, OVA-immunised rats were also challenged with PBS (100 μl paw⁻¹) or KLH (25 μl paw⁻¹), an unrelated antigen.

Determination of TNFα, IL-1β and CINC-1 levels in paw skin of OVA-injected rats

OVA (25 μg paw⁻¹, in 100 μl) or PBS (100 μg paw⁻¹) was injected into the hindpaw of immunised rats and 0.5, 1, 3, 5 or 24 h later, the animals were killed. The skin of the whole of the plantar area of paws was obtained and homogenised in 500 μl of the appropriate buffer containing protease inhibitors, and TNFα, IL-1β and CINC-1 levels were determined as described previously (Safieh-Garabedian et al., 1995; Rees et al., 1999) by enzyme-linked immunosorbent assay (ELISA). Briefly, microtitre plates (Nunc-Maxisorb) were coated overnight at 4°C with a rabbit anti-rat CINC-1 polyclonal antibody, sheep anti-rat IL-1β or sheep anti-rat TNFα. After blocking the plates, rat CINC-1, IL-1β or TNFα standards at various dilutions in a medium and 50 μl of samples were added in triplicate and maintained at room temperature for 2 h. Sheep anti-CINC-1, IL-1β or TNFα biotinylated polyclonal antibodies were added at a 1 : 500 dilution, followed by incubation at room temperature for 1 h. Finally, 100 μl of avidin-HRP (1 : 5000 dilution) was added to each well and, after 30 min, the plates were washed and the colour reagent OPD (40 μg, 50 μl well⁻¹) was added. After 15 min, the reaction was terminated with H₂SO₄ (1 M, 50 μl well⁻¹) and the optical density measured at 490 nm. The results were obtained by comparing the optical density with standard curves. In addition, the results were adjusted to 500 μl, the volume used to extract the cytokine from the paw skin, and were expressed as nanograms of respective cytokine per paw. As a control, the levels of these cytokines were determined in naive rats and sham-immunised animals injected with PBS (100 μg paw⁻¹) or with OVA (25 μg paw⁻¹, in a final volume of 100 μl). The cytokine levels in skin samples from control rats were determined 3 h later.

Effects of anticytokines sera or IL-1ra treatments on OVA-induced mechanical hypersensitivity

Rats were treated with preimmune serum (control group), sheep anti-rat TNFα, IL-1β, NGF or sheep anti-human IL-8 (each at doses of 50 μg paw⁻¹), and 15 min later, the same paw was injected with OVA (25 μg paw⁻¹, in 100 μl). IL-1ra (300 μg paw⁻¹, in 100 μl) was injected 30 min before OVA challenge (at the same dose). The sham-immunised group received an OVA injection (also at the dose of 25 μg paw⁻¹, in 100 μl). Mechanical hypersensitivity was measured 1, 3 and 5 h after OVA injection. The effectiveness of these doses of the antisera or of IL-1ra against their respective cytokines has been demonstrated previously (Ferreira et al., 1988; Cunha et al., 1991; 1992; 2000; Lorenzetti et al., 2002) and was also confirmed in this study.

Effects of dexamethasone, thalidomide, pentoxifylline, HOE 140 plus des-Arg⁹-Leu⁸-BK (DALBK), indomethacin, atenolol, MK 886 and CP 105696 on OVA-induced mechanical hypersensitivity

Immunised rats were treated with PBS (100 μl paw⁻¹; control), TRIS buffer (1 ml, i.p.;. control), 1% DMSO in PBS (1 ml, i.p.; control), 0.1% methyl cellulose in H₂O (1 ml, p.o.; control), HOE 140 dissolved in PBS (1 mg kg⁻¹, i.p.), DALBK dissolved in PBS (500 μg paw⁻¹), indomethacin dissolved in 0.1 M TRIS buffer, pH 8.0 (5 mg kg⁻¹, i.p.) or atenolol dissolved in PBS (1 mg kg⁻¹, i.p.), and 30 min later, the animals received an i.pl. injection of OVA (25 μg paw⁻¹, in 100 μl). Pretreatments with thalidomide dissolved in 1% DMSO in PBS (45 mg kg⁻¹, i.p.), pentoxifylline dissolved in PBS (100 mg kg⁻¹, i.p.), MK 886 dissolved in 0.1% methyl cellulose in H₂O (1 mg kg⁻¹, orally) or dexamethasone dissolved in PBS (0.5 mg kg⁻¹, s.c.) were performed 1 h before OVA challenge. CP 105696 dissolved in 1% DMSO in PBS (3 mg kg⁻¹, i.p.) was administered 45 min before OVA injection. Reinforcement doses of these drugs (at the same concentrations given above) or of solvents were injected 3 h after the OVA challenge. The sham-immunised group received an OVA injection (also at the dose of 25 μg paw⁻¹, in 100 μl). Mechanical hypersensitivity was evaluated 1, 3 and 5 h after the OVA challenge. The effectiveness of these doses of the drugs against their respective mediators has been demonstrated previously (Nakamura & Ferreira, 1987; Ferreira et al., 1988; 1997; Poole et al., 1999a; Ribeiro et al., 2000a; Canetti et al., 2001; Viana et al., 2002).

Determination of LTB₄ level in paw skin of OVA-injected rats

OVA (25 μg paw⁻¹, in 100 μl) or PBS (100 μl) was injected into the hindpaws of immunised and sham-immunised rats and 3 h later the animals were killed. Total skin samples of the plantar area of paws from these groups and from naive rats (control values) were obtained and homogenised in 500 μl of the appropriate buffer containing protease inhibitors (as described previously by Safieh-Garabedian et al., 1995). LTB₄ levels were determined by enzyme immunoassay using a commercial kit (Biotrak^TM, Amersham Pharmacia Biotech, England, U.K.) according to the manufacturer's instructions. The results were obtained by comparing the optical density with standard curves. In addition, the results were adjusted to 500 μl, the volume used to extract LTB₄ from the paw skin, and were expressed as pg paw⁻¹.

Materials

The following drugs were used: OVA, indomethacin, atenolol, dexamethasone, pentoxifylline, HOE 140, des-Arg⁹-Leu⁸-BK (DALBK), CFA (H37Ra, ATCC 25177), KLH (keyhole limpet haemocyanin) (Sigma, St Louis, MO, U.S.A.), thalidomide (RBI, MA, U.S.A.), MK 886 (5-lipoxygenase-activating protein inhibitor; Calbiochem, CA, U.S.A.), CP 105696 (LTB₄ receptor antagonist, a gift from Pfizer/Groton Laboratories, Groton, CT, U.S.A.) and IL-1ra (National Institute for Biological Standards and Control, U.K.). The following antisera were used: sheep anti-rat IL-1β, sheep anti-rat TNFα , sheep anti-rat NGF and sheep anti-human IL-8 (National Institute for Biological Standards and Control, U.K.).

Statistical analysis

Results are presented as means and standard errors of the mean for groups of five animals (for in vivo experiments) or three animals (for in vitro experiments) and they are representative of two or three different experiments. The amount of variation between the different experiments was not statistically different (ANOVA) and it was not higher than 10%. The differences between the experimental groups were also compared by ANOVA and, in the case of significance, individual comparisons were subsequently made with Bonferroni's t-test. The level of significance was set at P<0.05.

Results

Hyperalgesic effect of OVA in immunised rats

Intraplantar injection of OVA into the hindpaw of immunised rats evoked mechanical hypersensitivity in a dose- (12.5–100 μg paw⁻¹, in 100 μl) and time- (1–24 h) dependent manner when compared to sham-immunised rats injected with OVA (100 μg paw⁻¹). The mechanical hypersensitivity of immunised animals injected with OVA at doses of 25, 50 or 100 μg paw⁻¹ was already significant 1 h after the OVA injection (P<0.05), and this peaked 3 h later, declining thereafter and reaching the control level 24 h later. In the group injected with 12.5 μg paw⁻¹ of OVA, the hypersensitivity response was not significant after 1 h (P>0.05), but increased to significant values at 3 and 5 h (P<0.05) after OVA injection, declining thereafter (Figure 1). Injection of OVA (100 μg paw⁻¹, in 100 μl) into the hindpaw of sham-immunised (control) rats or, alternatively, of an unrelated antigen, KLH (25 μg paw⁻¹) in OVA-immunised animals did not alter the mechanical nociceptive threshold significantly when compared to the PBS-injected immunised animals (Figure 1).

Figure 1.

Hyperalgesic effect of ovalbumin (OVA) in immunised rats. The mechanical hypersensitivity was evaluated at 1, 3, 5, 7, 12 and 24 h after intraplantar injection of OVA in immunised (I.) rats at doses of 12.5, 25, 50 or 100 μg paw⁻¹, in 100 μl and in sham-immunised rats (S.I.) at a dose of 100 μg paw⁻¹. Control immunised rats were injected with PBS (100 μg paw⁻¹) or an unrelated antigen, KLH (25 μg paw⁻¹, in 100 μl). See method for immunisation protocol. Results are expressed as means±s.e.m. in groups of five rats. *P<0.05 when compared to S.I. rats treated with OVA (ANOVA followed by Bonferroni's t-test).

OVA-induced TNFα, IL-1β and CINC-1 production in paw skin of immunised rats

Intraplantar injection of OVA (25 μg paw⁻¹, in 100 μl) induced a significant increase in the production of TNFα, IL-1β and CINC-1 in the paw skin of immunised rats when compared to OVA-treated sham-immunised rats. The patterns of cytokine production were broadly similar, although there were some differences in the time courses (Figure 2, panels a–c). Concentrations of TNFα and CINC-1 peaked 1 h after OVA injection (P<0.05), remaining significantly elevated 3 h later (P<0.05) and decreasing to control levels at 5 h (Figure 2, panels a and c). The concentration of IL-1β peaked 3 h after OVA challenge (P<0.05), decreasing thereafter and reaching the control level 24 h later (Figure 2, panel b). The injection of PBS (100 μl paw⁻¹) in immunised rats did not induce a significant increase in TNFα, IL-1β and CINC-1 levels in rat paw skin when compared to PBS-injected sham-immunised rats.

Figure 2.

Concentrations of TNFα (panel a), IL-1β (panel b) and CINC-1 (panel c) in rat hindpaw injected with PBS or OVA. PBS (100 μl) or OVA (25 μg paw⁻¹, in 100 μl) was injected in the sham-immunised (S.I.) and immunised groups. Immunised rats were killed at 0.5, 1, 3, 5 or 24 h after the injection and paw skin samples were obtained for measurement of cytokines by ELISA. The cytokine levels shown for naive rats (N, open bar), PBS or OVA-treated S.I. rats (grey bars) represent values obtained 3 h after the treatments. Results are expressed as means±s.e.m. of three samples for each group. *P<0.05 when compared to OVA-injected S.I. rats (ANOVA followed by Bonferroni's t-test).

Antinociceptive effect of cytokine antisera on OVA-induced nociceptive hypersensitivity

Next, we used anticytokine sera to investigate whether the cytokines released in the paw after OVA administration mediated the OVA-evoked nociceptive hypersensitivity in immunised rats (Figure 3). The control group was treated with preimmune serum. Pretreatment with anti-rat TNFα serum (50 μl paw⁻¹, 15 min before OVA injection) significantly reduced the nociceptive hypersensitivity induced by OVA (25 μg paw⁻¹), when evaluated 1, 3 or 5 h after challenge (P<0.05). Pretreatment with IL-1ra (300 μg paw⁻¹, 30 min before OVA injection), sheep anti-rat IL-1β or sheep anti-human IL-8 (50 μl paw⁻¹, 15 min before the OVA injection) significantly inhibited the nociceptive hypersensitivity induced by OVA (25 μg paw⁻¹, in 100 μl) when evaluated 1 or 3 (P<0.05), but not 5 h (P>0.05) after OVA challenge. Pretreatment with sheep anti-rat NGF (50 μl paw⁻¹, 15 min before OVA injection) did not alter the nociceptive effect of OVA (25 μg paw⁻¹) in immunised rats (P>0.05). As observed previously, injection of OVA (25 μg paw⁻¹) in the sham-immunised group did not cause hypersensitivity (Figure 3). The doses of the antisera or of IL-1ra inhibit the nociceptive hypersensitivity induced by the respective cytokines by more than 90% (data not shown).

Figure 3.

Effect of anticytokine sera or interleukin 1 receptor antagonist (IL-1ra) on OVA-evoked mechanical hypersensitivity. Preimmune serum (control group; α-Control), sheep anti-rat TNFα (α-TNFα, 50 μl), IL-1β (α-IL-1β, 50 μl) serum, sheep anti-human IL-8 serum (α-IL-8, 50 μl) or sheep anti-rat NGF serum (α-NGF, 50 μl) were administered 15 min before intraplantar injection of OVA (25 μg paw⁻¹). IL-1ra (300 μg paw⁻¹, in 100 μl) was administered 30 min before OVA challenge. The sham-immunised group (S.I.) received an intraplantar injection of OVA (25 μg paw⁻¹). The mechanical hypersensitivity was assessed 1, 3 and 5 h after OVA challenge. Results are expressed as means±s.e.m. in groups of five rats. *P<0.05 when compared to preimmune serum-treated rats (ANOVA followed by Bonferroni's t-test).

Effect of HOE 140 plus DALBK, dexamethasone, thalidomide, pentoxifylline, indomethacin and atenolol on OVA-induced nociceptive hypersensitivity

Immunised rats were treated twice (30 min before and 3 h after intraplantar OVA injection) with the combination of B₁ and B₂ bradykinin receptor antagonists (HOE 140 plus DALBK), cytokine synthesis inhibitors (dexamethasone, thalidomide and pentoxifylline), a cyclooxygenase inhibitor (indomethacin) and a beta adrenoceptor antagonist (atenolol). We found that treatment with HOE 140 (1 mg kg⁻¹, i.p.) plus DALBK (500 ng paw⁻¹) did not inhibit the hyperalgesic effect of OVA in immunised rats when this was evaluated 1, 3 and 5 h after OVA challenge (P>0.05, Figure 4). HOE 140 (at the same doses) or DALBK (at the same doses) administered individually also did not inhibit the hyperalgesic effect of OVA (data not shown). Pretreatment with thalidomide (45 mg kg⁻¹, i.p.) or pentoxifylline (100 mg kg⁻¹, i.p.) significantly reduced the OVA-induced nociceptive hypersensitivity when evaluated 1 and 3 (P<0.05, Figure 4), but not 5 h after OVA challenge (P>0.05, Figure 4). Dexamethasone (0.5 mg kg⁻¹, s.c.) eliminated the nociceptive hypersensitivity induced by OVA (25 μl paw⁻¹) when this was evaluated 1, 3 or 5 h after OVA challenge (P<0.05, Figure 4). Pretreatment with indomethacin (5 mg kg⁻¹, i.p.), atenolol (1 mg kg⁻¹, i.p.) or atenolol plus indomethacin (at the same doses given above) significantly inhibited the nociceptive hypersensitivity induced by OVA when evaluated 1 and 3 (P<0.05) but not 5 h after OVA challenge (P>0.05, Figure 5). The intraplantar pretreatment with dexamethasone (5 μg paw⁻¹), indomethacin (100 μg paw⁻¹) or atenolol (25 μg paw⁻¹) produced results similar to those obtained with the systemic treatments described above (data not shown). Furthermore, the doses of HOE 140, DALBK and atenolol inhibited the nociceptive hypersensitivity induced by BK and dopamine, respectively, by more than 90% (data not shown). The doses of indomethacin, dexamethasone, thalidomide and pentoxifylline inhibited the nociceptive hypersensitivity induced by carrageenin by 50, 95, 90 and 88%, respectively (data not shown).

Figure 4.

Effect of thalidomide, dexamethasone, pentoxifylline or HOE 140 plus DALBK on OVA-evoked mechanical hypersensitivity. PBS (equivalent volume), dexamethasone (0.5 mg kg⁻¹), pentoxifylline (100 mg kg⁻¹, i.p.) or thalidomide (45 mg kg⁻¹, i.p.) were administered 1 h before and 3 h after (indicated by arrow) OVA challenge and HOE 140 (1 mg kg⁻¹, i.p.) plus des-Arg⁹-Leu⁸-bradykinin (DALBK, 500 μg paw⁻¹) were administered 30 min before and 3 h after (indicated by arrow) OVA treatment. The sham-immunised group (S.I.) received an intraplantar injection of OVA (25 μg paw⁻¹). The mechanical hypersensitivity was assessed at the time indicated after OVA challenge. Results are expressed as means±s.e.m. in groups of five rats. *P<0.05 when compared to PBS-treated rats (ANOVA followed by Bonferroni's t-test).

Figure 5.

Effect of atenolol and indomethacin treatment on the OVA-evoked mechanical hypersensitivity. PBS (equivalent volume), indomethacin (5 mg kg⁻¹, i.p.), atenolol (1 mg kg⁻¹, i.p.) or indomethacin plus atenolol (at the same doses) were administered 30 min before and 3 h after (indicated by arrow) OVA challenge. The sham-immunised group (S.I.) received an intraplantar injection of OVA (25 μg paw⁻¹). The mechanical hypersensitivity was assessed at the time indicated after OVA challenge. Results are expressed as means ±s.e.m. in groups of five rats. *P<0.05 when compared to PBS-treated rats (ANOVA followed by Bonferroni's t-test).

Inhibition of OVA-induced nociceptive hypersensitivity by 5-lipoxygenase activating protein (FLAP) inhibitor (MK 886) or LTB₄ receptor antagonist (CP105696)

Immunised rats were pretreated with CP 105696 (3 mg kg⁻¹) or MK 886 (3 mg kg⁻¹), 45 min or 1 h before OVA challenge, respectively. Reinforcement doses of CP 105696 (at the same dose) or MK 886 (at the same dose) were given 3 h after intraplantar injection of OVA (25 μg paw⁻¹). The treatment of immunised rats with MK 886 or CP 105696 significantly reduced the OVA-induced hypersensitivity when this was evaluated 1, 3 or 5 h after OVA challenge (P<0.05, Figure 6, panel a). The dose of CP 105696 inhibited the nociceptive hypersensitivity induced by LTB₄ by 95% (data not shown in figure).

Figure 6.

Participation of LTB₄ in OVA-evoked mechanical hypersensitivity. Panel (a) Effect of LTB₄ synthesis inhibitor (MK 886) and LTB₄ receptor antagonist (CP 105696) on OVA-induced mechanical hypersensitivity. PBS (equivalent volume, i.p.) or CP 105696 (3 mg kg⁻¹, i.p.) was administered 45 min before and 3 h after (indicated by arrow) OVA injection (25 μg paw⁻¹). MK 886 (1 mg kg⁻¹, orally) was administered 1 h before and 3 h after (indicated by arrow) OVA injection. The sham-immunised group (S.I.) received an i.pl. injection of OVA (25 μg paw⁻¹). The mechanical hypersensitivity was assessed at the time indicated after OVA treatment. Results are expressed as means±s.e.m. in groups of five rats. *P<0.05 when compared to PBS-treated rats (ANOVA followed by Bonferroni's t test). Panel (b) Levels of LTB₄ in paw skin of immunised, sham-immunised and naive rats. The sham-immunised group (S.I.) and immunised group (I) received an intraplantar injection of PBS (100 μl paw⁻¹) or OVA (25 μg paw⁻¹) and, 3 h later, rats were killed and paw skin samples were obtained for measurement of LTB₄ levels by EIA. The levels of LTB₄ in paw skin samples from naive rats were determined for control values. Results are expressed as means±s.e.m. of three samples from each group. *P<0.05 when compared to OVA injected sham-immunised group (S.I.), PBS-treated immunised rats or naive animals (ANOVA followed by Bonferroni's t-test).

Increase in LTB₄ levels in paw skin of immunised rats stimulated by OVA

OVA (25 μg paw⁻¹, in 100 μl) and PBS (100 μl paw⁻¹) were injected into the hindpaws of immunised or sham-immunised (control) rats. The animals were killed 3 h later and skin samples were obtained for the determination of LTB₄ levels. The injection of OVA into the hindpaws of immunised rats increased LTB₄ levels three-fold when compared to the sham-immunised rats stimulated with OVA (at the same dose; P<0.05), immunised rats injected with PBS or naive rats (Figure 6, panel b).

Discussion

In this study, we investigated the individual contribution of cytokines, BK, prostaglandins, leukotrienes, sympathetic amines and NGF, a neurotrophic factor, to the mechanical hypersensitivity exhibited following antigen challenge. The intraplantar injection of antigen (OVA) induced a dose- and time-dependent mechanical hypersensitivity, which peaked 3 h after challenge, declined thereafter and reached control levels 24 h later. This response was an antigen-specific immune reaction, because the same injection of OVA did not induce hypersensitivity in sham-immunised or naive animals. Furthermore, injection of KLH, a protein to which the rats had not been immunised, did not induce hypersensitivity in rats immunised to OVA.

The first 3 h of OVA-induced mechanical hypersensitivity was related to the endogenous release of TNFα, IL-1β and CINC-1. The OVA-induced hypersensitivity was significantly inhibited at 1 and 3 h by specific anti-rat TNFα, IL-1β and anti-human IL-8 (which cross-reacts with rat CINC-1) antisera (Lorenzetti et al., 2002) or IL-1ra treatments. Although NGF causes thermal and mechanical hyperalgesia (Lewin et al., 1993; 1994), it is not involved in the OVA-evoked mechanical hypersensitivity, because intraplantar pretreatment with antiserum that neutralises this cytokine had no effect on OVA-induced hypersensitivity. Consistent with the involvement of endogenous TNFα, IL-1β and CINC-1 in OVA-evoked hypersensitivity, levels of these mediators were found to be elevated in rat paw skin subsequent to the intraplantar injection of the same dose of OVA that evoked mechanical hypersensitivity. Moreover, the pattern of cytokine expression in the paw was similar to the time course of OVA-evoked mechanical hypersensitivity. The involvement of these cytokines in mechanical hyperalgesia has also been shown in carrageenin- or LPS-induced inflammation (Ferreira et al., 1988; Cunha et al., 1991; 1992; Lorenzetti et al., 2002) and in the nociceptive acetic acid- or zymosan-induced writhing response in mice (Ribeiro et al., 2000b). Furthermore, there is evidence that TNFα mediates the development of temporomandibular joint pain in patients with chronic connective tissue disease (Nordahl et al., 2000) and that IL-1β, IL-8 and IL-18 are involved in the pathogenesis of rheumatoid arthritis (Eastgate et al., 1988; Endo et al., 1991; Gracie et al., 1999, respectively).

The conclusion that endogenous TNFα, IL-1β and CINC-1 are involved in the first 3 h of OVA-evoked hypersensitivity in immunised animals is reinforced by our finding that thalidomide and pentoxifylline significantly reduced the hypersensitivity response at 1 and 3 h, but not at 5 h after OVA-challenge. These drugs, among other effects, inhibit the synthesis of TNFα and IL-1β (Strieter et al., 1988; Zabel et al., 1989; Sampaio et al., 1991; Bienvenu et al., 1992; Schandené et al., 1992; Weinberg et al., 1992; Moreira et al., 1993). Furthermore, dexamethasone inhibited OVA-induced hypersensitivity at 1, 3 and also at 5 h after the challenge. This difference in the pattern of inhibition among thalidomide, pentoxifylline and dexamethasone can be explained by the fact that thalidomide and pentoxifylline did not inhibit the synthesis of LTB₄, which is inhibited by dexamethasone (Hirata et al., 1980), and, as discussed below, LTB₄ mediates the later phase of OVA-evoked hypersensitivity.

During the last decade, investigations of the mechanism by which the mechanical hyperalgesia following LPS or carrageenin injection is induced by cytokines have shown that they constitute a link between these stimuli and the release of the final hyperalgesic mediators, that is, the prostaglandins and sympathetic amines (Ferreira et al., 1988; Cunha et al., 1991; 1992; Lorenzetti et al., 2002). TNFα , IL-1β and CINC-1 act sequentially: TNFα is released first after the injury and stimulates the production of IL-1β , which, in turn, stimulates the production of cyclooxygenase metabolites (Ferreira et al., 1988; Cunha et al., 1992) and of CINC-1, which stimulates the release of sympathetic amines (Lorenzetti et al., 2002). In the present study, indomethacin, a nonsteroidal anti-inflammatory drug that inhibits prostaglandin synthesis, and atenolol, a β₁-adrenoceptor antagonist, significantly inhibited the OVA-induced hypersensitivity in the first 3 h but were ineffective at 5 h after the OVA-challenge. These profiles of inhibition were similar to those observed with antisera against the above-mentioned cytokines, or with thalidomide and pentoxifylline, suggesting that, as with carrageenin- and LPS-induced inflammation, it is the cytokine-driven release of prostaglandins and sympathetic amines that mediates the early phase of OVA-induced hypersensitivity in immunised animals. The fact that atenolol, a selective β₁ adrenoceptor antagonist, inhibits the OVA-induced hypersensitivity suggests that the sympathetic amines released after the OVA challenge are activating the β₁-adrenoceptor on the sensitive neurone. In previous studies, it was demonstrated that in carrageenin- or LPS-induced inflammation, sympathetic amines are also released and activate the β₁ (Nakamura & Ferreira, 1987) or β₂ (Levine et al., 1988) adrenoceptor. Our results do not exclude the possibility that the β₂ adrenoceptor is also activated after OVA challenge.

It seems that TNFα has a role in hypersensitivity 5 h after OVA administration, a conclusion supported by the significant inhibition of the hypersensitivity obtained at this time point with a specific anti-TNFα antiserum. However, this TNFα effect is not mediated by the sequential release of IL-1β/eicosanoids and of CINC-1/sympathetic amines, because neither antisera against these cytokines nor indomethacin and atenolol affected the OVA-evoked hypersensitivity at this time point. Furthermore, thalidomide and pentoxifylline were also ineffective on OVA-induced hypersensitivity at this time point. The conclusion that TNFα participates in OVA-induced hypersensitivity at 5 h seems to contradict the observation that soluble TNFα was not detected in the paw skin at this time point after OVA challenge. A possible explanation for these apparently contradictory results is that, at this time point, the hyperalgesic effect of the membrane-associated form of TNFα predominates, which is antagonised by antiserum, but not detected by ELISA. In fact, we recently showed that LPS-stimulated macrophages release soluble TNFα and express the membrane-associated form of this cytokine, which is not detected by ELISA (Crossara-Alberto et al., 1997). However, if this hypothesis is correct, thalidomide did not inhibit the expression of this form of TNFα, since the pretreatment of the rats with this drug did not inhibit the OVA-induced hypersensitivity at this time point. Further experiments are needed to clarify this point.

It was proposed that BK acting synergistically on B₁ and B₂ receptors stimulates the release of TNFα involved in carrageenin- or LPS-induced mechanical hyperalgesia (Ferreira et al., 1993; Poole et al., 1999a). To investigate whether BK generation also preceded the release of cytokine involved in the genesis of OVA-induced hypersensitivity, we used a combination of B₁ and B₂ receptor antagonists. It was found that pretreatment of the immunised rats with DALBK plus HOE 140, selective B₁ and B₂ receptors antagonists, respectively, did not significantly inhibit OVA-induced mechanical hypersensitivity. Despite this observation, it has been reported in the literature that there is a substantial increase in the expression of the B₂ receptor in the lumbar dorsal root ganglion neurones in antigen-induced arthritis in the rat knee and that this increase is relevant for the generation of acute and chronic inflammatory pain (Segond von Banchet et al., 2000). The use of different experimental models could explain these apparently contradictory results.

The role of endogenous LTB₄ in the genesis of OVA-induced-mechanical hypersensitivity was evaluated using MK 886, an LTB₄ synthesis inhibitor (FLAP inhibitor; Miller et al., 1990) and CP 105696, an LTB₄ receptor antagonist (Koch et al., 1994). It was observed that this eicosanoid, together with prostaglandins and sympathetic amines, mediates the first 3 h of OVA-induced hypersensitivity. Moreover, LTB₄ also mediates the OVA-induced hypersensitivity at 5 h after the challenge. This is supported by the observation that OVA-evoked hypersensitivity was significantly inhibited by MK 886 and CP 105696 at 1, 3 and also 5 h after challenge. Confirming the involvement of LTB₄ in the process, a significant increase in LTB₄ levels was observed in the paw of sensitised rats after intraplantar injection of OVA. Mechanical and thermal hyperalgesic effects of LTB₄ in rats and humans have been described previously and appear to be dependent on polymorphonuclear leukocytes (Levine et al., 1984; Bisgaard & Kristensen, 1985). The possible participation of neutrophils in the onset of the OVA-evoked hypersensitivity is under investigation. Although this finding is the first demonstration that LTB₄ mediates the hyperalgesia observed in antigen-induced hyperalgesia, a number of studies have reported significant levels of LTB₄ in human inflammatory diseases in which hyperalgesia occurs, such as rheumatoid arthritis (Davidson et al., 1983; Ahmadzadeh et al., 1991; Gursel et al., 1997). Taken together, these results suggest that LTB₄ antagonists could be useful for the treatment of pain observed in immune inflammatory diseases. Whether the release of LTB₄, which participates in the OVA-induced hypersensitivity, is mediated by cytokines was not investigated in the present study. However, there is evidence in the literature that TNFα, which was increased during the first 3 h after the OVA challenge, is capable of inducing LTB₄ production (Meyer et al., 1988; Camussi et al., 1989; Canetti et al., 2001).

Recently, Feitosa et al. (2002), investigating the mediators involved in OVA-induced oedema in sensitised rats, demonstrated that dexamethasone and serotonin antagonists, but not indomethacin, MK 886, thalidomide or pentoxifylline, inhibited oedema formation significantly. Similarly, we also demonstrated, in our experimental model, that OVA-induced paw oedema is not inhibited by the pretreatment of the animals with indomethacin, atenolol, thalidomide, pentoxifylline or MK 886 (results not shown), treatments that inhibited the hyperalgesia (Figures 4, 5 and 6). Dexamethasone treatment reduced both OVA-induced oedema and mechanical hyperalgesia (Figure 4). These results clearly showed that there are different mechanisms involved in the genesis of oedema and hyperalgesia. Therefore, it may be that the oedema formation induced by OVA challenge in immunised rats depends on the release of serotonin and kinins that promote an increase in venous permeability. On the other hand, sensitisation of primary sensory neurones by eicosanoids or sympathetic amines depends on the release of cytokines by resident cells (Cunha et al., 2000). In line with this idea, we have shown that the administration of TNFα, IL-1β and IL-8, at doses that caused hyperalgesia, did not cause oedema (Poole et al., 1999b). Moreover, thalidomide, a drug that inhibits the release of TNFα, also reduces mechanical hyperalgesia, but not oedema formation induced by intraplantar administration of carrageenin (Ribeiro et al., 2000a).

It is well known that cAMP and protein kinases A and C are involved in molecular events in the primary sensory neurone associated with acute and persistent hyperalgesia triggered by inflammatory stimuli, as well as by final hyperalgesic mediators such as prostanoids and sympathetic amines (Taiwo & Levine, 1991; Aley & Levine, 1999; Cunha et al., 1999; Aley et al., 2000). However, the involvement of these intracellular mediators in the mechanical hypersensitivity induced by antigen challenge is not yet fully understood, and this remains an aim of our future studies.

In conclusion, our results suggest that the cytokine (TNFα, IL-1β and CINC-1)-driven release of prostaglandins, sympathetic amines and LTB₄, but not of bradykinin, mediates the first 3 h of mechanical hypersensitivity induced by OVA in immunised rats. At 5 h after OVA administration, although TNFα has a minor role, LTB₄ is the key nociceptive mediator. Therefore, inhibition of the synthesis of this eicosanoid could be beneficial for the control of hyperalgesia in immune inflammatory diseases.

Abbreviations

BK

bradykinin

CFA

complete Freund's adjuvant

CINC-1

cytokine-induced neutrophil chemoattractant 1

DALBK

des-Arg⁹-Leu⁸-BK

EIA

enzyme immunoassay

ELISA

enzyme-linked immunosorbent assay

FLAP

5-lipoxygenase activating protein

HOE 140

D-Arg-Arg-Pro-Hyp-Gly-3-[2-thienyl]-Ala-Ser-D-1,2,3,4-tetrahydro-3-isoquinolinecarbonyl-L-2α,3β,7αβ]-octahydro-1H indole-2-carbonyl-Arg

i.p.

intraperitoneal

i.pl.

intraplantar

IL

interleukin

IL-1β

interleukin 1 beta

IL-1ra

interleukin 1 receptor antagonist

IL-8

interleukin 8

KLH

keyhole limpet haemocyanin

LPS

lipopolysaccharide

LTB₄

leukotriene B₄

NGF

nerve growth factor

OPD

o-phenylenediamine

OVA

ovalbumin

PBS

phosphate-buffered saline

s.c.

subcutaneous

TNFα

tumour necrosis factor alpha