Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Interleukin-1 (IL-1) is a cytokine that is synthesized in and released from phagocytic macrophages when they are activated by bacterial lipopolysaccharide (LPS) (Dinarello, 1984; Kluger, 1991). It produces activation of the hypothalamo-pituitary adrenocortical axis, and the resultant increase in glucocorticoids leads to modulation of the multiple IL-1-induced physiological responses (Rivier, 1993; Schobitz et al., 1994; Rivier, 1995; Buckingham et al., 1996). For example, glucocorticoids inhibit IL-1-induced fever (Davidson et al., 1990; Watanabe et al., 1995) and play a crucial role in the development of IL-1-induced acute-phase responses (Gordon and Koj, 1985). A relatively small amount of IL-1, whether administered centrally or systematically, has been shown to evoke a rise in blood pressure in both rats (Morimoto et al., 1992; Takahashi et al., 1992; Bataillard and Sassard, 1994; Kannan et al., 1996) and humans (Haefeli et al., 1993). We recently found that glucocorticoids modulate the pressor response induced in rats by the systemic injection of IL-1 and we suggested that endogenous glucocorticoids may actually enhance the IL-1-induced pressor response (Watanabe et al., 1996). The mechanism underlying this modulation, however, still remains to be elucidated.

Interestingly, it has been shown that IL-1 stimulates the release of nitric oxide (NO) (Busse and Mulsch, 1990; Schini et al., 1991; Kosaka et al., 1992; Kanno et al., 1993). This substance is a potent endogenous vasodilator synthesized from L-arginine by NO synthase (NOS) (Moncada et al., 1991; Lancaster, 1992; Snyder and Bredt, 1992; Szabo, 1995). There is increasing evidence that three isoforms of NOS exist: endothelial constitutive NOS (endothelial cNOS), brain cNOS, and cytokine (such as IL-1)-inducible NOS (iNOS) (Moncada et al., 1991; Szabo, 1995). Because glucocorticoids are potent inhibitors of the induction of iNOS, and because NO dilates blood vessels and lowers blood pressure (Moncada et al., 1991; Lancaster, 1992; Snyder and Bredt, 1992; Szabo, 1995), we wondered whether the effect of glucocorticoids on IL-1-induced NO release might contribute to their enhancement of the IL-1-induced pressor response.

To test this hypothesis, we examined the effects of an exogenous glucocorticoid, dexamethasone (DEX), and of adrenalectomy on the NO production and pressor response that are both induced by IL-1 in rats. Blood pressure was measured in freely moving rats using a biotelemetry system. Because NO is unstable and is degraded into nitrite (NO2) and nitrate (NO3) ions, we measured plasma NO2 and NO3 (NOx) as our indices of endogenous NO production. For this purpose, we used capillary zone electrophoresis (Ueda et al., 1995), which is a newly developed and reliable assay that measures NO2 and NO3 directly from plasma samples. The present results suggest that, in freely moving rats, glucocorticoids enhance the IL-1-induced pressor response at least in part by inhibiting endogenous NO release.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals

The animals used in this study were male Wistar rats, weighing 270 to 350 g. They were housed in individual plastic cages (40 × 25 × 25 cm; length × width × depth) with wood-chip bedding in a room maintained at 26 ± 1°C, a temperature within the thermoneutral zone for rats. They experienced a photoperiod of 12 h light:12 h dark, lights coming on at 7:00 AM. The animals' living conditions and the experimental protocols satisfied criteria laid down by the ethics committee of Yamaguchi University.

This study comprised four types of experiment, all on freely moving rats. Each experimental group was divided into two subgroups: blood pressure was measured in one subgroup and the plasma concentration of NOx in the other. We took these measurements from different animals because the rats were hand-held while blood samples were being taken for NOx measurement. This procedure is a kind of very mild stress and we wanted to record blood pressure changes without any stress. In addition, to use the same rat on two different occasions would also have been inappropriate, because chronic (Kornel et al., 1995) or repeated (Suzuki et al., 1995) administration of DEX, the glucocorticoid used in this study, increases the resting blood pressure level (hypertension). For this reason, we wanted to administer DEX only once in each animal in this study. In experiment 1, we investigated the effects of a single i.p. injection of IL-1 (10 µg/kg) on either blood pressure or the plasma level of NOx. This was done with or without pretreatment, by s.c. injection, with the glucocorticoid, DEX. In experiment 2, a single s.c. injection of the nonspecific NOS inhibitor, N-nitro-L-arginine methyl ester (L-NAME, 15 mg/kg), was given, and its effect on the IL-1-induced pressor or NOx response was examined. In experiment 3, the blood pressure or the plasma concentration of NOx was measured in adrenalectomized (ADX) and in sham-ADX rats before and after i.p. IL-1. In experiment 4, ADX rats were given a single s.c. injection of DEX 1 h before the i.p. injection of IL-1 to examine the effect of glucocorticoid treatment on the blood pressure or NOx response to IL-1 in ADX rats. All animals had ad libitum access to water (see below) and to standard laboratory rat chow. Sham-ADX and intact rats drank tap water, whereas ADX rats were given 0.9% salt water.

Surgery

Blood pressure was measured using a biotelemetry system (Data Science, Inc., St. Paul, MN) (Lange et al., 1991). Each rat was anesthetized with sodium pentobarbitone (50 mg/kg, i.p.) and a battery-operated transmitter (model TA11PA-C40) was implanted i.p. The transmitter includes a sensor and a radio-frequency transmitter. The transmitters have a fluid-filled intra-arterial catheter attached to the sensor in the body of the transmitter. The output of the transmitter was monitored by antennae mounted in a receiver board (model RA1310) placed under each animal's cage. The data were fed into a peripheral processor (matrix model BCM100) connected to a Sanyo MBC-17J AX computer (IBM-compatible). The blood pressure monitoring transmitter was calibrated at three levels of pressure, 750, 850, and 950 mm Hg. To obtain the calibration values, air pressure set at a given value (750, 850, or 950 mm Hg) is read by a sensor in the transmitter and converted to voltage. Subsequently, a voltage-frequency conversion is performed and this signal data is fed into the computer. This procedure is repeated for each level of pressure. The actual blood pressure reported in our manuscript is the pressure difference between the atmospheric pressure and the pressure read by the transmitter implanted in the rat's abdomen. During surgery, the tip of the blood pressure catheter was inserted directly into the abdominal aorta. In brief, abdominal section was performed, and the catheter tip (12 mm in length) was inserted into the abdominal aorta (from a point near the bifurcation) so that it was directed toward the heart. At least 10 days were allowed to elapse before the start of the experiment.

When required, bilateral adrenalectomy was carried out under general anesthesia (sodium pentobarbitone, 50 mg/kg, i.p.) at least 10 days after the implantation of the transmitter. A dorsal midline skin incision was made, the muscle layer below the rib cage on each flank was cut and the adrenal gland was removed with some surrounding tissue; this was repeated on the contralateral side. For the sham operation, similar incisions were made and the adrenal glands were located but not removed. At least 10 days were allowed to elapse before experimentation began.

Rats without transmitters (intact, ADX, or sham-ADX) were used to measure the plasma concentration of NOx. The animals were anesthetized with sodium pentobarbitone (50 mg/kg, i.p.), and a polyvinyl tube was inserted into the jugular vein so that its tip lay in the superior caval vein near the right atrium (Harms and Ojeda, 1974). The free end of the catheter was passed s.c. to the mid-scapular region, where it was exteriorized dorsally behind the neck. It was kept patent by flushing it every day with heparinized 0.9% saline (50 U/ml). This implantation was performed at least 1 week after the rat's adrenalectomy or sham operation and at least 3 days before any blood samples were taken. All rats were handled for 10 min each day for at least 3 days to accustom them to the experimenters.

Drugs

Human recombinant IL-1, supplied by Otsuka Pharmaceutical (Tokushima, Japan), was produced from recombinant strains of Escherichia coli. The activity of the IL-1 was found to be 2 × 10⁴ units/µg by a thymocyte coproliferation assay. The IL-1 preparation was shown to be free of significant endotoxin contamination by the Limulus amoebocyte assay (<0.05 pg/µg protein). For injection, the recombinant IL-1 was dissolved in sterile saline. These solutions were divided between several vials and stored at 40°C until use. We used each vial within the 2 days after thawing, and thus avoided repeated freezing and thawing. DEX (Sigma Chemical Co., St. Louis, MO) or L-NAME (Sigma) was dissolved in sterile saline.

Experimental Protocols

Experiment 1. Cardiovascular measurement. On the day of the experiment, each rat was gently picked up, and its transmitter switched on using a magnet. The blood pressure was then allowed to stabilize for a period of 90 min before any injections. IL-1 was administered i.p. in a volume of 1 ml/kg over a period of 15 s into hand-held rats. To minimize the confusing effects of the rats' circadian rhythm, IL-1 was always given between 10:00 AM and 11:00 AM. Either DEX or saline was given by s.c. injection 1 h before the IL-1.

NOx measurement. Another group of rats was given IL-1 to induce a NOx response. Blood samples were withdrawn from the cannula in the jugular vein to allow the plasma concentration of NOx to be measured in intact rats before and after the injection of IL-1. Either DEX or saline was given s.c. to each animal 1 h before the IL-1. Blood samples were taken three times: 60 min before and 30 and 180 min after the injection of IL-1. On each occasion, about 0.4 ml of blood was withdrawn, collected into a test tube containing EDTA and centrifuged at 2000 rpm (454 g) for 15 min at 4°C. The plasma was then transferred into a fresh test tube and stored at 40°C until the measurement of NOx. The blood cells were resuspended in sterile saline, stored at 4°C, and returned to each animal at the end of each day of experiments. Plasma NOx concentration was measured by capillary zone electrophoresis. The method has been described in detail elsewhere (Ueda et al., 1995). In brief, electrophoresis was carried out in a Waters AccuSep polyimide-coated, 60 cm  75 µm inside diameter fused-silica capillary at a potential of 20 kV, with on-column UV detection at 214 nm. The running buffer was composed of 750 mM sodium chloride containing 5% NICE-Pac Osmosis Flow Modifier Anion-BT (Waters Corp., Milford, MA) (Ueda et al., 1995).

Experiment 2. The effect of s.c. injection of L-NAME (15 mg/kg) on either the pressor or NOx response induced by IL-1 was examined in this experiment. L-NAME was administered 150 min after the IL-1 injection. The procedures used were essentially the same as those described for experiment 1, except that plasma samples were taken only twice: 60 min before and 180 min after IL-1 injection.

Experiments 3 and 4. ADX or sham-ADX rats were given IL-1, and their blood pressure or NOx was measured (experiment 3). Some ADX rats received glucocorticoid treatment 1 h before the IL-1 injection (experiment 4). The procedures used were similar to those described for experiment 1, except that plasma samples were taken 60 min before and 180 min after the IL-1 injection.

Confirmation of Adrenalectomy

At the end of each experiment on ADX rats, the animal was sacrificed with an overdose of ether, and blood was taken by cardiac puncture. The resulting plasma samples were assayed for the presence of corticosterone, using a commercial corticosterone radioimmunoassay kit (Diagnostic Products Corp., Los Angeles, CA). In addition, each animal was visually checked for regeneration of the adrenal glands. Rats with very low levels of corticosterone (<20 ng/ml) in their plasma samples even under ether stress and with no apparent regeneration of the adrenal glands qualified as ADX rats.

Statistical Analysis

All results are expressed as mean ± S.E.M. Data were analyzed for statistical significance by a repeated measures ANOVA (Macintosh, StatView 4.0) to assess the overall effect.

This analysis was followed by an unpaired Student's t test with Bonferroni's correction to enable the two groups of rats to be compared in terms of the values at each time point. Even if the repeated measures ANOVA showed no treatment effect, we considered data that showed significant interaction (ANOVA) and a significant difference (Student's t test) to be statistically significant (see Results, Fig. 1, B and C). Furthermore, a one-way ANOVA, followed by Scheffe's test (post hoc test), was carried out to compare the values in each group at 60 min with those at each subsequent time point.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   A and B, changes in mean arterial blood pressure after i.p. injection at time 0 of IL-1 (10 µg/kg). DEX (0.5 mg/kg) (n = 10) or saline (n = 10) was administered s.c. 1 h before injection of IL-1. Small inset (A) depicts entire time course of blood pressure changes; values shown in "B" were taken from inset. C, changes in plasma concentration of NOx after i.p. injection at time 0 of IL-1 (10 µg/kg). DEX (0.5 mg/kg) (n = 5) or saline (n = 6) was administered s.c. 1 h before the injection of IL-1. From ANOVA: for treatment effect, p > .05 in B and C; for time effect, p < .001 in B and C; for interaction, p < .001 in B and C. ***p < .001 versus corresponding value at 60 min.

Statistical analysis was performed on the data illustrated in parts B and C of Figs. 1 to 4. Details of the results of the various forms of analysis are given in each figure legend. Differences were considered significant at p < .05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of s.c. Treatment with DEX on IL-1-Induced Pressor and NOx Responses in Intact Rats (Experiment 1). Injection of IL-1 (10 µg/kg, i.p.) induced biphasic increases in mean arterial blood pressure in intact rats (Fig. 1A). Pretreatment with DEX (0.5 mg/kg, s.c.) reduced the early rise in blood pressure, but enhanced the second phase (Fig. 1A and B). The injection of IL-1 evoked a rise in the plasma concentration of NOx at 180 min after the injection, although it had no effect at 30 min (Fig. 1C). This IL-1-induced increase in NOx was suppressed by DEX (0.5 mg/kg, s.c.).

Effect of s.c. Treatment with L-NAME on IL-1-Induced Pressor and NOx Responses in Intact Rats (Experiment 2). Figure 2 shows the effect of L-NAME (15 mg/kg, s.c.) on the IL-1 (10 µg/kg, i.p.)-induced pressor and NOx responses in intact rats. As shown in Fig. 2, L-NAME, given at 150 min, enhanced the IL-1-induced rise in blood pressure at 180 min (Fig. 2, A and B) and reduced the increase in plasma NOx at the same time point (Fig. 2C). It can be seen from Figs. 1 and 2 that both DEX and L-NAME completely suppressed the IL-1-induced increase in plasma NOx at 180 min, and that these agents enhanced the pressor response at that time point. This confirmed that endogenous NO, the release of which increased after IL-1 injection, exerts a moderating action on the pressor response to IL-1. In additional experiments, L-NAME (15 mg/kg, s.c.) or saline was given s.c. immediately before IL-1 to examine the effect of L-NAME on the IL-1-induced early increase in blood pressure seen in this study. The early phase of the IL-1-induced pressor response was significantly enhanced by L-NAME. The relevant data at each time point for saline+IL-1 (n = 5) and L-NAME+IL-1 (n = 5), respectively, were: 100 ± 3 mm Hg and 99 ± 3 mm Hg at time of injection; 119 ± 4 mm Hg and 137 ± 2 mm Hg at time 30 min after the injection.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   A and B, changes in mean arterial blood pressure after i.p. injection at time 0 of IL-1 (10 µg/kg). L-NAME (15 mg/kg) (n = 6) or saline (n = 6) was administered s.c. 150 min after the injection of IL-1. Small inset (A) depicts entire time course of blood pressure changes; values shown in B were taken from inset. C, changes in plasma concentration of NOx after i.p. injection at time 0 of IL-1 (10 µg/kg). L-NAME (15 mg/kg) (n = 6) or saline (n = 6) was administered s.c. 150 min after injection of IL-1. From ANOVA: for treatment effect, p < .01 in B, p < .05 in C; for time effect, p < .001 in B, p < .01 in C; for interaction, p < .01 in B, p < .05 in C. *p < .05, **p < .01, ***p < .001 versus corresponding value at 60 min.

Blood Pressure and NOx Responses Induced in ADX or Sham-ADX Rats by IL-1 (10 µg/kg, i.p.) (Experiment 3). The resting mean arterial blood pressure (measured at 15-min intervals over the hour before IL-1 injection) was significantly lower in the ADX rats (88 ± 1 mm Hg) than in the sham-ADX rats (105 ± 2 mm Hg). Hence, the changes in blood pressure induced by IL-1 are expressed as absolute deviations from the resting level, measured at time 0 (Fig. 3, A and B). The IL-1-induced pressor response was smaller in the ADX rats than in the sham-operated rats, the second phase being the more strongly affected (Fig. 3, A and B). In contrast, the ADX rats showed an evoked increase in NOx at 180 min that was significantly greater than that seen in sham-ADX rats (Fig. 3C).

View larger version (14K): [in this window] [in a new window]   Fig. 3.   A and B, absolute changes in mean arterial blood pressure from its level at time 0 in ADX (n = 10) and sham-ADX (n = 10) rats. Each rat received an i.p. injection at time 0 of IL-1 (10 µg/kg). Small inset (A) depicts entire time course of blood pressure changes; values shown in B were taken from inset. C, changes in plasma concentration of NOx in ADX (n = 8) and sham-ADX (n = 7) rats after i.p. injection at time 0 of IL-1 (10 µg/kg). From ANOVA: for treatment effect, p < .001 in B, p < .05 in C; for time effect, p < .001 in B and C; for interaction, p < .001 in B, p < .05 in C. *p < .05, ***p < .001 versus corresponding value at 60 min.

Effect of s.c. Treatment with DEX on IL-1-Induced Pressor and NOx Responses in ADX Rats (Experiment 4). Pretreatment with DEX (0.5 mg/kg, s.c.) 1 h before the IL-1 injection resulted in a marked enhancement of the IL-1-induced changes in mean arterial blood pressure in ADX rats (Fig. 4, A and B). In contrast, the evoked increase in NOx was attenuated by DEX (Fig. 4C).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   A and B, changes in mean arterial blood pressure in ADX rats after i.p. injection at time 0 of IL-1 (10 µg/kg). DEX (0.5 mg/kg) (n = 9) or saline (n = 8) was administered s.c. 1 h before injection of IL-1. Small inset (A) depicts entire time course of blood pressure changes; values shown in B were taken from inset. C, changes in plasma concentration of NOx in ADX rats after i.p. injection at time 0 of IL-1 (10 µg/kg). DEX (0.5 mg/kg) (n = 5) or saline (n = 6) was administered s.c. 1 h before injection of IL-1. From ANOVA: for treatment effect, p < .05 in B and C; for time effect, p < .001 in B and C; for interaction, p < .001 in B and C. **p < .01, ***p < .001 versus corresponding value at 60 min.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present results showed that i.p. administration of IL-1 in freely moving rats induced a biphasic increase in mean arterial blood pressure, and that a rise in the plasma concentration of NOx occurred during the second phase of the pressor response (at 3 h after the injection of IL-1). Furthermore, systemic pretreatment with DEX enhanced the second phase of the pressor response, thus confirming our previous finding (Watanabe et al., 1996), and inhibited the evoked increase in the plasma level of NOx. These results suggest that NO is involved as a vasodilator in the regulation of the late phase of the pressor response, and that the enhancement by DEX of the IL-1-induced rise in blood pressure was, at least in part, due to its inhibition of NO release. This idea is further supported by the finding that one of the most frequently used nonselective NOS inhibitors, L-NAME, both enhanced the IL-1-induced pressor response and attenuated the NOx response seen in this study. This result indicates that, whichever type of NOS (cNOS or iNOS) is responsible for the NO synthesis that follows IL-1 injection, the endogenous NO, once produced, does indeed exert an inhibitory effect on the IL-1-induced pressor response.

In fact, glucocorticoids have been shown to inhibit the induction of iNOS, but to have no apparent effect on the constitutive form of the enzyme (Radomski et al., 1990; Moncada et al., 1991). Indeed, although L-NAME induced a marked pressor effect, no change in resting blood pressure was observed after the acute injection of DEX in this study. This is in contrast to the effect of chronic administration (Kornel et al., 1995) or repeated injection (Suzuki et al., 1995) of DEX, which is to induce hypertension. Furthermore, there are reports suggesting that iNOS activity is stimulated within 3 h (the time at which we saw a rise in plasma NOx) after a challenge with proinflammatory agents such as bacterial LPS or IL-1 (Rosa et al., 1990; Kanno et al., 1993; Szabo et al., 1993a,b; Szabo, 1995). LPS has been shown to induce iNOS through the mediation of LPS-induced cytokines such as IL-1 (Szabo, 1995). Taken together, all this makes it seem likely that DEX inhibited the IL-1-induced stimulation of iNOS, and that this resulted in an attenuation of the ensuing NO release and an enhancement of the pressor response evoked by IL-1 in freely moving rats.

By comparison with sham-ADX rats, our ADX rats showed a suppression of the IL-1-induced pressor response and an enhancement of the plasma NOx increase. Furthermore, DEX pretreatment restored to normal both the pressor and the NOx responses that had undergone alterations after adrenalectomy. These results suggest that the absence of endogenous glucocorticoids in the ADX rats led to a disinhibition of iNOS activity, and consequently to an enhanced NO release. This would then be responsible, at least in part, for the suppression of the pressor response seen in the ADX rats. In other words, endogenous glucocorticoids might normally act to attenuate NO formation and so enhance the IL-1-induced pressor response in freely moving rats. Szabo et al. (1993a,b) and Wu et al. (1995) found that the effect of LPS administration in anesthetized intact and ADX rats (viz. hypotension) was attenuated by DEX treatment, which also inhibited the evoked increase in iNOS activity in the lung. These findings are in accord with ours inasmuch as they suggest that endogenous glucocorticoids exert a tonic suppressive effect on iNOS induction. As a consequence, glucocorticoids might tend to prevent cardiovascular failure being induced by LPS in anesthetized rats (Szabo et al., 1993b). The above findings are supported by the fact that an inhibitory action of glucocorticoid on the NO release evoked by IL-1 or LPS has also been demonstrated in vitro (Radomski et al., 1990; Rosa et al., 1990; Kanno et al., 1993; Jun et al., 1994).

There is a possibility that an IL-1-induced rise in blood pressure could have the effect of increasing the shear stress on endothelial cells (Ranjan et al., 1995), and so lead to the activation of cNOS. A part of the regulatory action of NO on the IL-1-induced blood pressure rise could then simply derive from a cNOS-induced NO release. If this were so, the elevation in NO would represent a nonspecific, counterregulatory response to the blood pressure rise. In this study, however, the plasma level of NOx was unchanged 30 min after the injection of IL-1, but it was raised 180 min after, while the (raised) levels of blood pressure at these time points were almost the same. The activation of cNOS is rapid, whereas the induction of iNOS is slow. Stimulation of iNOS activity has been reported to occur within 180 min after a challenge with LPS or IL-1 (Rosa et al., 1990; Kanno et al., 1993; Szabo et al., 1993a,b; Szabo, 1995). All the above findings suggest that iNOS plays an important role in the NOx response to IL-1, although we cannot exclude the possibility of the participation of cNOS.

In this study, L-NAME given at 150 min after IL-1 enhanced the late phase of the IL-1-induced pressor response and attenuated the IL-1-induced NOx response. On the other hand, Yamamoto et al. (1994) found that an IL-1-induced decrease in blood pressure was blocked by the administration of L-NAME in conscious rats. These two findings suggest that, regardless of the effect of IL-1 on blood pressure, the blood pressure after IL-1 administration is certainly under the influence of IL-1-induced NO. In the present study, there was no increase in plasma NOx at 30 min after IL-1, although L-NAME given immediately before IL-1 augmented the IL-1-induced early rise in blood pressure. We believe that at 30 min after IL-1, iNOS is not yet induced and that it is the cNOS-induced (not IL-1-induced) basal release of NO that affects the IL-1-induced early pressor response.

The mechanisms responsible for the induction of the IL-1-induced pressor response were not explored here. However, the involvement of centrally acting prostaglandins (PGs) (Morimoto et al., 1992; Takahashi et al., 1992; Bataillard and Sassard, 1994; Kannan et al., 1996) and corticotropin-releasing factor (Nakamori et al., 1993) in such pressor responses has been suggested. In our previous report, we showed an increase in the plasma concentration of norepinephrine after injection of IL-1 (Watanabe et al., 1996). Therefore, the pressor response to IL-1 may be, at least in part, neurosympathetically mediated. However, we cannot exclude the possibility that the IL-1-induced pressor response is due in part to a direct increase in peripheral smooth muscle tone. It is also possible that an endothelium-derived vasoconstrictor, endothelin, contributes to the pressor response, because IL-1 has been shown to stimulate endothelin production from endothelial cells (Katabami et al., 1992). Because blood pressure is affected by the endothelium-dependent regulation of vascular smooth muscle tone by NO and endothelin, it would be interesting to examine the above-mentioned possibility in future research.

There is a discrepancy between our results and previous reports that needs to be considered. It has repeatedly been reported that administration of a high dose of IL-1 or LPS induces a decrease in blood pressure in anesthetized rats (Kilbourn et al., 1992; Szabo et al. 1993a,b; Wu et al. 1995). However, it seems likely that a high dose of either of these agents would greatly stimulate the induction of iNOS, leading to an overproduction of NO and thence to shock (Szabo, 1995). Under pathological conditions, such as septic shock, such NO overproduction might become a major cause of death. In contrast, our freely moving conscious rats showed an increase in blood pressure after injection of a relatively small dose of IL-1. Furthermore, other investigators have found an increased blood pressure after the systemic administration of a small dose of IL-1 (100 ng/rat) in anesthetized rats (Takahashi et al., 1992). Interestingly, a pressor response to a small dose of IL-1 has been reported in humans as well (Haefeli et al., 1993). For this reason, we believe that the actual doses of IL-1 administered are important factors determining the reported effect of IL-1 on blood pressure. The physiological significance of the IL-1-induced pressor response is, at the present time, unknown. Future studies should investigate this point.

The present results represent the first evidence that the enhancement by glucocorticoid of the IL-1-induced pressor responses seen in freely moving rats results, at least in part, from its inhibitory effect on endogenous NO release. However, it should be noted that in the present results there was no change in plasma NOx at 90 min after the administration of the IL-1, while the increase in blood pressure at the same time point was augmented by pretreatment with DEX (see Fig. 1A). It is suggested that iNOS was not yet induced at 90 min after the IL-1 and, therefore, that other mechanisms might also be involved in the enhancement of the pressor response that is seen with glucocorticoid. One such candidate might be a change in the IL-1-induced PGs in the blood. Indeed, glucocorticoids interfere with the induction of PG synthesis through inhibition of the activity of phospholipase A₂ and cyclooxygenases, which are the key enzymes of PG biosynthesis (Goppelt-Struebe, 1997), and some of the circulating PGs, especially PGE, are well known to lower blood pressure. This possibility should be examined in future research. Changes in plasma NOx were measured in this study: this provides an index of the entire body's endogenous NO production. In the future, we need to examine the in vivo effect of glucocorticoids on localized IL-1-induced NO release in tissues such as blood vessels. This would help us determine in which tissue the NO that is involved in blood pressure regulation is induced.