Address correspondence and reprint requests to Rosaria Meli, Department of Experimental Pharmacology, via D. Montesano, 49–80131 Naples, Italy. E-mail: email@example.com
Thrombin (THR) plays a key role in the brain under physiological and pathological conditions. Several of the biological activities of thrombin have been shown to be mainly driven through activation of protease-activated receptor-1 (PAR-1)-type thrombin receptor. Here we have studied the effect of THR and PAR-1-activating peptide (PAR1-AP), SFLLRN, on cytokine-induced expression of inducible nitric oxide (iNOS), a prominent marker of astroglial activation using the rat C6 glioma cells. In this cell line, THR (1–10 U/mL) and PAR1-AP (1–100 µm) induced a significant concentration-dependent increase both of IFN-γ- (250 U/mL) or TNF-α- (500 U/mL) induced NO release. The observed increase of NO production was related to an enhancement of iNOS expression as measured in cell lysates prepared from different treatments by using SDS–PAGE followed by western blot analysis. The effect of THR, but not that of PAR1-AP, was significantly inhibited by hirulogTM (60 µg/mL), a specific and stochiometric THR inhibitor or by cathepsin-G (40 mU/mL), an inhibitor of PAR-1. In conclusion our data suggest a role for THR through activation of PAR-1 in the induction of astroglial iNOS, and further support the hypothesis that THR may function as an important pathophysiological modulator of the inflammatory response.
Following trauma, injury, or diseases, breakdown of blood–brain barrier (BBB) occurs, causing activation of several different repairing mechanisms. In the complex pattern of events involved the protease thrombin (THR), generated after vascular injury as part of the coagulation cascade, plays an important role. Indeed, in addition to its central role in coagulation, THR has numerous biological functions, that are related to inflammation, tissue remodelling, and healing (for a review see Cirino et al. 2000).
Recent observations suggest an equally important role for THR in the brain under normal conditions and following injury (Nishino et al. 1993; Turgeon and Houenou 1997). Many of its actions are mediated by the activation of protease-activated receptor-1 (PAR-1)-type thrombin receptor, a member of the G-protein coupled receptor family. There are four known members of this receptor family of PARs, namely PAR-1, PAR-3 and PAR-4, which are activated by THR, while PAR-2 is a receptor for trypsinlike-enzymes (Coughlin 2000). The peculiar mechanism whereby THR activates its receptor is a proteolytic unmasking of a new –NH2 terminus, which acts as a new self-activating ligand. In particular activation of PAR-1, localised to neurones and glial cells, leads to a widespread effects such as extensive retraction of processes on neurones and astrocytes, inhibition of neurite outgrowth (Gurwitz and Cunningham 1988), stellation reverse in type-1 astrocytes (Cavanaugh et al. 1990; Grabham and Cunningham 1995), and cell proliferation (Grabham and Cunningham 1995). All these effects contribute to gliosis commonly seen at sites of injury in the brain, and limiting tissue regeneration. In addition, astroglia has been proposed to have a role in the cerebral injury as it can synthesise and secrete inflammatory cytokines (Benveniste 1992; Norris et al. 1994; Campbell et al. 1997), and express adhesion molecules (Aloisi et al. 1992), responsible for interaction with T-lymphocytes and extracellular matrix protein. The potential importance of THR effects in the brain is also strengthened by studies where THR inhibitors have been shown to produce beneficial effects. In particular glia-derived nexin, a specific THR inhibitor, secreted mainly by astrocytes, blocks astrocyte stellation reversal by THR (Cavanaugh et al. 1990) and argatroban or hirudin, reduces secondary brain damage following inflammation (Motohashi et al. 1997; Kubo et al. 2000).
Nitric oxide (NO) has been shown to be involved in many physiological and pathological processes in the brain. Small amount of NO synthesised in the brain during neural activity mediates physiological functions, such as neural morphogenesis, development and plasticity, while excess NO production contributes to neuronal injury during cerebral ischemia and may lead to neurodegeneration in various pathological conditions (Koprowski et al. 1993). NO is synthesised by three different types of NO-synthase (NOS) including the constitutive neuronal (nNOS) and endothelial (eNOS) isoforms and the inducible isoform (iNOS) (Knowles and Moncada 1994). The contribution of iNOS to delayed ischemic injury was demonstrated by using iNOS knockout mice (Iadecola et al. 1997) and a selective iNOS inhibitor (Parmentier et al. 1999).
Fetal bovine serum (FBS), tissue culture media, and supplements were purchased from Hy-Clone. Human α-thrombin, recombinant mouse TNF-α, cathepsin-G and 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) were from Sigma (St Louis, MO, USA). Recombinant mouse IFN-γ was obtained from Calbiochem (San Diego, CA, USA). PAR1-AP (SFLLRN) and scrambled peptide (FSLLRN) were prepared by solid phase synthesis with a continous-flow instrument with on-line UV monitoring (Milligen/Biosearch 9050). The stepwise synthesis was carried out by Fmoc chemistry. Peptides were purified by RP-HPLC and their identity was confirmed by mass spectroscopy in the Department of Medicinal Chemistry of the Federico II University of Naples. HirulogTM was obtained from Biogen (Cambridge, MA, USA). iNOS was detected on western blot with a mouse monoclonal antibody IgG1 from Transduction Laboratories (Lexington, KY, USA; N39120, clone 54); the peroxidase conjugated secondary antibody was purchased from Jackson (West Grove, PA, USA). The rat C6 glioma cell line was a generous gift of Dr M. Taglialatela, Department of Neurosciences, University of Naples ‘Federico II’, Italy.
C6 glioma cell line was cultured in 75-cm2 flasks in Dulbecco's modified essential medium (DMEM) with 10% FBS, 2 mm l-glutamine, 100 U/mL penicillin and 100 µg/mL streptomycin at 37°C under 5%CO2 humidified air. Cells were passaged at confluence using a solution of 0.025% trypsin and 0.01% EDTA and used between passage 49–57.
Cell viability studies were performed by 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay in a 96-well plate. C6 cells (5 × 104/well) were cultured for three days. After 4 h of starvation in serum free DMEM/F12, the cells were treated with THR or PAR-1AP at increasing concentrations (1–10 U/mL and 1–100 µm, respectively). After 24 h of incubation at 37°C, 25 µL of MTT (5 mg/mL) were added to each well, 3 h later the cells were lysated with 100 µL of lysis buffer (20% SDS and 50% DMF, pH 4.7). After an additional 18 h incubation at 37°C, the optical densities (OD620) for the serial dilutions of drugs were compared with the OD of the control wells to assess the cell viability (Mosmann 1983). Cell proliferation was confirmed by [3H]thymidine incorporation according to Luo and Miller (1996).
Cell treatments and nitrite assay
Cells (5×104/well) were cultured for 3 days, to confluence, in 24 well-plates in complete medium. After 4 h of starvation in serum free DMEM/F12, cells were stimulated with THR (1–10 U/mL), PAR1-AP (1–100 µm), or scrambled peptide (100 µm) for 24 h. Immediately thereafter the supernatants were separated and kept frozen at − 20°C until were assayed. In another set of experiments, following a 30-min incubation with IFN-γ (250 U/mL) or TNF-α (500 U/mL) C6 cells were treated with THR (1–10 U/mL) or PAR-1AP (1–100 µm). The effect of hirulogTM (60 µg/mL) (Maraganore et al. 1990), a specific THR inhibitor, was assessed by adding it at the same time of either THR or PAR-1AP.
In another set of experiments, cathepsin-G (40 mU/mL), a PAR-1 inhibitor (Renesto et al. 1997), was added 30 min before IFN-γ in presence or in absence of THR or PAR-1AP.
Nitrite accumulation, an indicator of NO synthesis, was measured in the culture medium by Griess reaction. In brief, 100 µL of cell culture medium were mixed with 100 µL of Griess reagent [equal volumes of 1% (w/v) sulfanilamide in 5% (v/v) phosphoric acid and 0.1% (w/v) naphtylethylenediamine-HCl] and incubated at room temperature (22°C)for 10 min. Immediately, thereafter absorbance at 550 nm was measured in a microplate reader Titertek. Fresh culture medium was used as the blank for all the experiments. The amount of nitrite in the samples was calculated from a sodium nitrite standard curve freshly prepared in culture medium and expressed as micromolar unit.
Western blot analysis
C6 glioma cells were cultured in complete medium in P-60 well plates (3×105/dish). After 3 days, about 90% confluence, cells were starved for 4 h in DMEM/F12 serum free medium and subsequently treated with IFN-γ (250 U/mL) or TNF-α (500 U/mL) and 30 min thereafter stimulated with THR or PAR1-AP at increasing concentrations, in presence or absence of hirulogTM (60 µg/mL). The same scheme of cell treatment was performed also in presence of cathepsin-G (40 mU/mL), that was added 30 min prior to IFN-γ stimulation.
After 24 h of incubation cells were washed twice with ice-cold phosphate-buffered saline (PBS) and resuspended in Tris-HCl 20 mm pH 7.5, 10 mm NaF, 150 mm NaCl, 1% Nonidet P40, 1 mm phenylmethylsulfonylfluoride, 1 mm Na3VO4, leupeptin (10 µg/mL) and trypsin inhibitor (10 µg/mL). After 1 h, cell lysates were obtained by centrifugation at 100 000 g for 15 min at 4°C. Protein concentrations were estimated by the Bio-Rad protein assay using bovine serum albumin as standard.
Equal amounts of protein (70 µg) of the cell lysates were dissolved in Laemmli's sample buffer, boiled for 5 min, and subjected to sodium docecyl sulfate-polyacrylamide gel electrophoresis (8% polyacrylamide). Western blotting was performed by transferring proteins from a slab gel to a sheet of polyvinylidene difluoride membrane at 240 mA for 40 min at room temperature. The filter was then blocked with 1X PBS and 5% non-fat dried milk for 40 min at room temperature and incubated with the first antibody diluted 1 : 10 000 in 1X PBS, 5% non-fat dried milk and 0.1% Tween 20 overnight at 4°C. The day after, the filter was incubated with a secondary antibody (anti-mouse IgG-horseradish peroxidase conjugate 1 : 10 000 dilution in the same buffer) for 1 h at room temperature. Subsequently, blots were extensively washed with PBS, developed using ECL-detection reagents (Amersham) according to the manufacturer's instructions, and exposed to Kodak X-Omat film. The protein bands of iNOS on X-ray film were scanned and densitometrically analysed by Model GS-700 Imaging Densitometer.
Data are reported as mean ± SEM values of three independent determinations. All experiments were done at least three times, each time with three or more independent observations. Statistical analysis was performed by two-way Anova test and multiple comparisons were made by Bonferroni's test. A value of p < 0.05 was considered significant.
Effect of THR and PAR-1AP on cell proliferation
The proliferative effect of THR (1-5-10 U/mL) or PAR1-AP (1-10-100 µm) was evaluated. When comparing optical density (OD) values of drug-treated and control cells, the proliferative effect was significant only at the highest concentration of THR used (0.882 ± 0.028 versus 0.781 ± 0.024, p < 0.05), while PAR-1AP was ineffective at all concentrations tested. The proliferative effect of THR was confirmed by [3H]thymidine incorporation assay (874.33 ± 108.73 versus 525.00 ± 29.30 counts per min, p < 0.05). In both assay used IFN-γ by itself did not show a proliferative effect (data not shown). Combination of IFN-γ (250 U/mL) with THR (10 U/mL) did not further increase THR activity.
Effect of THR and PAR-1AP on NO2− production
THR (1-5-10 U/mL) and PAR-1 AP (1-10-100 µm) induced a weak non-significant increase in NO2− accumulation in C6 glioma cell supernatant (control was 0.70 ± 0.34, THR was 0.88 ± 0.35, 1.03 ± 0.34 and 1.14 ± 0.29 µm at 1, 5 and 10 U/mL, respectively, PAR-1AP was 0.89 ± 0.16, 0.94 ± 0.17 and 0.97 ± 0.20 µm at 1, 10 and 100 µm, respectively).
Effect of THR and PAR-1AP on IFN-γ-induced NO2− production
Incubation of C6 cells with THR (1–10 U/mL), after IFN-γ (250 U/mL) stimulation, induced an increase of nitrite accumulation in a concentration-dependent manner (Fig. 1a). The synergic effect of THR was significant at 5 and 10 U/mL (*p < 0.05 and **p < 0.01, versus IFN-γ alone). To confirm that THR action was mediated via the activation of the proteolytical cleavage of THR receptor, the specific THR protease inhibitor hirulogTM was used. HirulogTM (60 µg/mL) was able to significantly block THR effect, at the highest concentration used, on IFN-γ-induced NO2− production (#p < 0.05, versus THR 10 U/mL plus IFN-γ).
PAR-1AP showed a similar but weaker effect on IFN-γ-induced NO2− production by these cells (Fig. 1b). This effect appears to be specific, as the scrambled peptide (100 µm) was inefficacious. The effect of PAR-1AP was significant only at the highest concentration tested (100 µm, *p < 0.05 and **p < 0.01, versus IFN-γ alone) and it was not reversed by hirulogTM.
Effect of THR and PAR-1AP on TNF-α-induced NO2− production
As shown in Fig. 2 (a and b) both THR (1-5-10 U/mL) and PAR-1AP (1-10-100 µm) had the same profile of activity on cells stimulated by TNF-α. Their effect was concentration-dependent and significant at the higher concentrations both of THR at 5 and 10 U/mL (*p < 0.05 and **p < 0.01 versus TNF-α alone) and PAR-1AP at 100 µm (*p < 0.05 and **p < 0.01, versus TNF-α alone).
HirulogTM suppressed the synergistic effect of THR on TNF-α-induced NO release (#p < 0.05, versus TNF-α plus THR 10 U/mL) but did not modify PAR-1AP effect.
As before, the scrambled peptide did not modify TNF-α-induced NO2− production by C6 cells, excluding an aspecific effect of PAR-1AP.
THR and PAR-1AP potentiation of IFN-γ-induced iNOS expression
To determine whether the potentiation caused by THR and PAR-1AP on IFN-γ-induced nitrite release was due to enhanced iNOS expression, western blot analysis was carried out on whole lysates using a monoclonal antibody against the mouse macrophage iNOS. A representative blot of iNOS induction in lysates prepared from cells stimulated with THR and PAR-1AP after a pre-tratment with IFN-γ is shown in Fig. 3. In unstimulated control cells there was not detectable protein band of ∼130 kDa. As expected hirulogTM reversed the potentiation of iNOS expression induced by THR, while it did not modify PAR-1AP effect. The densitometric analysis was performed on three different experiments. C6 cell iNOS expression was increased by 39.75 ± 6.92%, 248.75 ± 36.99% and 411 ± 74.78% at 1, 5 and 10 U/mL of THR, respectively. Moreover IFN-γ-induced iNOS was also increased by PAR-1AP treatment (30.33 ± 9.62%, 123.60 ± 17.61% and 164.3 ± 30.10% increase at 1, 10 and 100 µm, respectively).
THR and PAR-1AP potentiate TNF-α-induced iNOS expression
To evaluate the potentiation of THR and PAR-1AP of TNF-α-induced iNOS expression, C6 cell lysates were subjected to immunoblot analysis. Both THR and PAR-1AP enhanced TNF-α-induced expression of the enzyme (Fig. 4). HirulogTM inhibited THR effect on TNF-α-induced iNOS expression, but did not modify PAR-1AP effect. Densitometric analysis performed on three different experiments showed that THR enhanced iNOS expression by 112.65 ± 37.27%, 227.30 ± 83.37% and 357.25 ± 144.67% at 1, 5 and 10 U/mL, respectively, while PAR-1AP induced an increase of enzyme band of 60.95 ± 43.04%, 118.00 ± 41.57% and 265.20 ± 59.37% at 1, 10 and 100 µm, respectively.
Cathepsin-G inhibits THR but not PAR-1AP effect on IFN-γ-induced NO release
To further confirm PAR-1 involvement in the synergistic effect of THR on IFN-γ-induced NO release, we used cathepsin-G. When added to cells, it reversed significantly (p < 0.05) THR potentiation on IFN-γ-induced NO release, as well as iNOS expression. As expected, cathepsin-G did not modify PAR-1AP non-proteolytic activity (Fig. 5).
The inflammatory nature of ischemic damage has led to investigate on the involvement of several mediators and enzymes that are characteristic of this phenomenon. It goes without any doubt that activation of the inflammatory pathways leads invariably to the activation of the coagulation cascade. Here by using the rat C6 glioma cell line, where it has been demonstrated the existence of PAR-1 receptor (Ubl et al. 1998), we have evaluated the effect of THR and PAR-1AP in basal and inflammatory conditions. C6 glioma cell line exhibits many properties of astrocytes, as expression of astrocyte-specific markers, including glial fibrillary acidic protein (GFAP) and S-100 (Pfeiffer et al. 1970; Bissel et al. 1974). In addition, compared with primary culture of astrocytes, which can contain small amounts of contaminating microglia, C6 provide a pure source of astroglia-derived cells. Astrocytes have been suggested to play a role in cerebral injury by synthesising and secreting cytokines and inflammatory mediators, such as NO (del Zoppo et al. 2000).
Astrocyte-derived NO could contribute to neuronal cell death in a variety of pathological conditions. The cytotoxic activity of NO is due not only to a direct proinflammatory effect on brain tissue, but also to an increase of cerebral vessel permeability with an ischemia- and inflammation-induced disruption of the BBB that allows the recruitment of other proinflammatory cells, such as neutrophils and lymphocytes (Boje and Lakhman 2000; Mayhan 2000).
In our experimental conditions both THR and PAR-1AP at the concentrations used did not cause a significant increase in nitrite release in C6 cell supernatants. Conversely, they both potentiated IFN-γ-induced nitrite production and this effect was paralleled by an increased iNOS expression. In pathological situations, such as viral or allergic encephalitis, multiple sclerosis and brain diseases associated with inflammation, there are infiltrates of activated T-lymphocytes, which could be the main source of different inflammatory cytokines such as IFN-γ, TNF-α and IL-6. Moreover, it is well known that stimulation of cytokine receptors located on the astroglial cell membrane is followed by important biochemical changes including the induction of iNOS and consequent formation of NO (Feinstein et al. 1994; Sakai et al. 1995). Thus, THR released following an injury could induce, in co-operation with IFN-γ, iNOS expression and in turn NO release contributing to sustain the ongoing inflammatory response.
THR effect was inhibited by the specific THR inhibitor hirulogTM, while PAR-1 AP effect was unchanged. This latter result outlines the importance of THR proteolytic activity to exert the effect observed. It is well known that hirulogTM, a 20 amino acid peptide, binds the fibrinogen site of THR and inhibits THR proteolytic effect by a catalytic site inhibition domain (Maraganore et al. 1990). HirulogTM did not inhibit the response mediated by PAR-1AP, as would be expected based on the direct activation of the peptide agonist compared with the proteolitically response of THR.
The observation that PAR-1AP reproduces THR effect suggests that THR receptor PAR-1 is involved. This evidence is strengthened by the evidence that cathepsin-G, an inhibitor of PAR-1, abolishes THR but not PAR-1AP effect. Cathepsin-G is known to cleave PAR-1 downstream of the thrombin cleavage site, generating a ‘disarmed’ receptor, unresponsive to subsequent proteolytic activation of TRH, but still responsive to non-proteolytic activity of PAR-1AP (Vergnolle 2000).
In order to investigate whether the observed effects were solely linked to IFN-γ or even to another pro-inflammatory cytokine, we have also evaluated the capability of THR and PAR-1AP to potentiate TNF-α induction of iNOS. Indeed, TNF-α plays a major role in inflammation and it is also a potent inducer of iNOS. In our experiments, TNF-α-induced nitrite production was strongly increased in a concentration-dependent fashion by THR and PAR-1AP. Also with this different cytokine the increased nitrite production well correlated with enhanced and concentration-dependent up-regulation of iNOS protein expression. Our observation that cytokines synergised with THR and PAR-1AP to release NO from C6 cells suggests that activation of both cytokine and THR receptors may be a key interaction associated with inflammatory responses. Several possibilities can be considered to account for this synergy. One possibility is that cytokines increase the expression of functional THR receptors necessary for stimulation of NO synthesis. This mechanism is unlikely, as in our experimental conditions IFN-γ did not affect THR-induced cell proliferation. The exact mechanism leading to the this synergistic up-regulation of iNOS is presently unclear. However, it is likely that differential pathways of cellular activation are involved; in fact, it is well known that iNOS induction by cytokines occurs through pathways involving JAK/STAT proteins, while PARs, that are G protein coupled receptors, can activate alternative signalling transduction mechanisms (Kitamura et al. 1996).
In the system studied PAR-1AP is less efficacious than THR. This difference in potency could be due to different causes such as (1) an inefficient presentation of the soluble peptide to the binding domain of PAR-1, when compared with the tethered peptide, (2) the rapid inactivation of peptide by proteolysis and (3) the involvement of other receptors activated by THR but not by PAR-1AP.
The involvement of other PARs and non-PARs related activation can not be excluded. Till now there is no evidence for PAR-3 and PAR-4 expression on rat C6 cells, while Kaufman et al. (2000) reported a role for PAR-1 and PAR-4 in THR-induced calcium signalling in an astrocytoma cell line of human origin.
As injury of the CNS causes an astrogliosis, characterised by cell swelling and proliferation, similar to the effects of the serine protease THR on astrocytes (Pindon et al. 1998), we also evaluated the proliferative effect of THR and PAR-1AP on C6 cells. This effect was significant at the highest concentration of THR and uneffected by IFN-γ, while PAR-1AP was ineffective. We hypothesise that THR released concomitant to inflammatory cytokines such as TNF-α and IFN-γ present at the site of injury might amplify and/or sustain CNS inflammation.
In conclusion, the results of the present study show that THR has the potential to increase the production of proinflammatory mediators, such as NO, by activating glial cells in the CNS, supporting a role for THR as an important pathophysiological modulator of the inflammatory response.
This study was supported in part by a grant from the Ministero dell'Università e della Ricerca Scientifica e Tecnologica, Italy. GMR is at present a recipient of a grant of European Union.