The Generation of Nitric Oxide and Carbon Monoxide ... - CiteSeerX

0 downloads 0 Views 198KB Size Report
The pic- ture is even more complex if we consider that the two gases may or may .... the ductus arteriosus in the lamb: evidence against the prime role of guanylyl.
0013-7227/98/$03.00/0 Endocrinology Copyright © 1998 by The Endocrine Society

Vol. 139, No. 3 Printed in U.S.A.

The Generation of Nitric Oxide and Carbon Monoxide Produces Opposite Effects on the Release of Immunoreactive Interleukin-1b from the Rat Hypothalamus in Vitro: Evidence for the Involvement of Different Signaling Pathways* CESARE MANCUSO†, GIUSEPPE TRINGALI†, ASHLEY GROSSMAN, PAOLO PREZIOSI, AND PIERLUIGI NAVARRA Institute of Pharmacology (C.M., G.T., P.P., P.N.), Catholic University Medical School, Largo Francesco Vito 1, 00168 Rome, Italy; and Department of Endocrinology (A.G.), St. Bartholomew’s Hospital, London EC1A 7BE, United Kingdom ABSTRACT Both the cytokine, interleukin-1 (IL-1), and the gaseous neurotransmitters, nitric oxide (NO) and carbon monoxide (CO), have been implicated in the control of neuroendocrine functions, such as the release of CRH and luteotropic hormone-releasing hormone from the hypothalamus. Though increased levels of IL-1 in this brain region are unambiguously associated with enhanced CRH and reduced luteotropic hormone-releasing hormone release, the net effects of the two gases are still unclear, but in vivo and in vitro evidence suggests that the generation of NO and CO within the hypothalamus might counteract the stimulatory effects of IL-1 and bacterial lipopolysaccharide on the neuroendocrine stress axis. In this study, we have investigated the effects of NO and CO on the release of immunoreactive (ir)-IL-1b from the rat hypothalamus in vitro. It was observed that the NO donor, sodium nitroprusside (SNP), stimulates ir-IL-1b release under basal conditions, whereas the increase in CO levels obtained with hemin, the CO

I

T IS NOW generally agreed that central nervous system (CNS)-borne interleukin-1b (IL-1b) is able to influence neuroendocrine function, including the release of CRH (1) and luteotropic hormone-releasing hormone (2) from the hypothalamus. IL-1b gene expression and synthesis are induced in astrocytes and microglia by the systemic or intracerebroventricular administration of bacterial lipopolysaccharide (LPS) (3), an experimental procedure currently used to mimic pathological conditions such as endotoxemia and septic shock. In addition, IL-1b production is constitutive in a neuronal population whose cell bodies have been predominantly localized to the paraventricular nucleus of the rat hypothalamus, with sparse intra- and extrahypothalamic projections (4). Although these interleukinergic neurons Received June 12, 1997. Address all correspondence and requests for reprints to: Prof. Pierluigi Navarra, Institute of Pharmacology, Catholic University Medical School, Largo Francesco Vito 1, 00168 Rome, Italy. * This work was carried out within the Programma nazionale di ricerca su Farmaci II fase, contract awarded by Ministero dell’Universita` e Ricerca Scientifica e Tecnologica to Consorzio Siena Ricerche. Part of this work was also supported by Consiglio Nazionale delle Ricerche Grant 96.05324.ST74. † Contributed equally to this work.

precursor through the heme oxygenase pathway, has no effect on basal ir-IL-1b release but inhibits release stimulated by high K1 concentrations. The opposite effects of the two gases on cytokine release seemed to be caused by the activation of different signaling pathways, because: 1) SNP, but not CO-saturated solutions, is able to increase cyclic GMP levels in hypothalamic tissue; 2) CO-saturated solutions increase PGE2 production and release from the hypothalamic explants, whereas SNP has no effect; 3) SNP-stimulated ir-IL-1b release is counteracted by a selective inhibitor of soluble guanylyl cyclase, LY 83583, but not by a cyclooxygenase inhibitor, indomethacin; and 4) conversely, indomethacin, but not LY 83583, reverses the inhibitory effect of hemin on K1-stimulated ir-IL-1b release. It is concluded that NO and CO signal in the rat hypothalamus via the activation of soluble guanylyl cyclase and cyclooxygenase, respectively. (Endocrinology 139: 1031–1037, 1998)

seem to be involved in the activation of the hypothalamicpituitary-adrenal (HPA) axis after restraint stress (5), it is unclear whether neuronal IL-1 participates in the activation of stress response secondary to immune-inflammatory challenges. More recently, a new class of neuromodulatory agents has been shown to play a role in the control of neuroendocrine function: the gaseous compounds nitric oxide (NO) and carbon monoxide (CO) (6). The mode of action of these gases is exquisitely paracrine, because they act only at short distance from their sites of generation. All of the known isoforms of NO synthase and heme oxygenase (HO), i.e. the enzymes leading to NO and CO biosynthesis, respectively, have been detected within the rat hypothalamus (7–10), providing the basis for the paracrine control of neuropeptide release. An interplay seems to exist between IL-1 and the gases NO and CO in the control of the HPA axis. Thus, IL-1- and LPS-induced increases in circulating adrenocorticotropic hormone and corticosterone levels in the rat were significantly potentiated by the inhibition of NO synthase activity by l-nitroarginine methyl ester, indicating that the generation of NO induced by IL-1 or LPS serves as a counterregu-

1031

NO AND CO AFFECT IL-1b RELEASE FROM HYPOTHALAMUS

1032

latory mechanism impeding the exaggerated activation of the HPA axis (11, 12). To date, there is no in vivo evidence of a similar interaction between the cytokines and CO, although hemin, the CO precursor via the HO pathway, has been shown to inhibit the stimulatory effect of IL-1b on CRH release from rat hypothalamic explants (13). This finding matched similar observations with IL-1b and NO-donors on the releases of CRH and vasopressin in vitro (14, 15), suggesting that the hypothalamus is the major locus for interactions among IL-1b, NO, and CO. To explore such interactions in more detail, we have used a previously characterized in vitro model to investigate the effects of NO and CO on the release of immunoreactive (ir)-IL-1b from rat hypothalamic explants. Such release was stimulated by high K1 concentrations, and K1-induced release was dependent on Ca21, suggesting that the fraction of IL-1 released under the depolarizing stimulus belongs to a preformed pool, which is probably neuronal in origin (16). While previous studies have suggested that the biological actions of CO in the CNS mimic those of NO (17), we unexpectedly found that the generation of NO and CO in hypothalamic explants was associated with opposite effects on ir-IL-1b release. This also suggested that the two gases may exert the above actions through the activation of different signaling mechanisms. Further studies provided pharmacological and biochemical evidence that NO and CO signal in the rat hypothalamus via the activation of soluble guanylyl cyclase (sGC) and cyclooxygenase (COX), respectively. Materials and Methods Animals Male and female Wistar rats, weighing 200 –300 g, were used. They were kept five to a cage and maintained at a temperature of 23 6 1.5 C, with a relative humidity of 65 6 2%. The animals were exposed to 12 h of light (0600 –1800 h) followed by 12 h of dark, and they had free access to food and water. The use of animals for this experimental work has been approved by the Italian Ministry of Health (licensed authorization to P. Navarra).

Hypothalamic dissection On the day of the experiment, the animals were decapitated between 0900 h and 1000 h to avoid circadian variation. The brains were removed, and the hypothalami were dissected within the following limits: anterior border of the optic chiasm, anterior border of the mamillary bodies, and the lateral hypothalamic sulci. The depth of dissection was 4 mm. The hypothalami were then bisected longitudinally through the midsagittal plane, and two hypothalamic halves from the same animal were incubated in the same vial. The total dissection time was less than 2 min from decapitation.

Hypothalamic incubations Two hypothalamic halves from the same animal were incubated in 7-ml polyethylene vials containing 500 ml incubation medium for ir-IL-1b and PGE2 assays and 1 ml for cyclic GMP (cGMP) assay. In experiments for the determination of ir-IL-1b release, the incubation medium was Earle’s balanced salt solution (EBSS; Sera-Lab Ltd, Crawley Down, UK) supplemented with 0.1% BSA (Sigma Chemical Co., St Louis, MO), 50 mg/ml ascorbic acid (Sigma), and 80 IU/ml aprotinin (Trasylol, Bayer, Germany), pH 7.4, in an atmosphere of 95% O2/5% CO2. In the first 60 min of incubation after hypothalamic explant, the medium was replaced every 20 min; during this time, irIL-1b release rate tended to stabilize (data not shown). Therefore, in all the subsequent experiments, the first hour was taken as preincubation time. Thereafter,

Endo • 1998 Vol 139 • No 3

the explants were subjected to a 20-min control incubation in plain medium to assess basal ir-IL-1b release. This was followed by a second 20-min incubation in medium containing test substances or, for the control group, in medium alone. In experiments with KCl, in the second 20-min period, EBSS was replaced by a medium consisting of 56 mm KCl and 67 mm NaCl, with the same concentrations of the other ions as found in EBSS. Such medium was able per se to significantly increase ir-IL-1b release, with respect to incubations with plain EBSS (16). Medium samples were stored at 230 C until assayed for ir-IL-1b. Experiments for the determination of cGMP and PGE2 were conducted according to the same protocol as above, with the following modifications: in experiments with CO, saturated solutions of the gas were prepared by bubbling pure CO (Caracciolossigeno, Rome, Italy) for 1 h before the experiments (i.e. during the preincubation time) and then incubating the tissue in CO-saturated medium in an atmosphere of O2/CO2/CO. The use of EBSS in these experiments was prevented by the observation that gassing of EBSS for 1 h with pure CO increases pH of the solution to about pH 8.8, which inhibits sGC activity (18). Therefore, an incubation medium consisting of Tris (0.1 m) 1 NaCl (95 mm), pH 7.4, was employed in experiments for the determination of cGMP, and the same medium, with the addition of 0.05% BSA, pH 7.4, was used when PGE2 release was assessed. Experiments with sodium nitroprusside (SNP) for the evaluation of PGE2 and cGMP were conducted in EBSS supplemented with 0.05% BSA and in Tris (0.1 m) 1 NaCl (95 mm), respectively. Incubation media and tissues were stored at 230 C until assayed for PGE2 and cGMP, respectively.

Analytical methods IL-1b was measured by RIA as previously described (16). The detection limit of the assay was 8 pg/tube, and the estimated EC50 was 170 pg/tube. The intra- and interassay variability was 2% and 10%, respectively. PGE2 was measured by RIA as previously described (19). The detection limit of the assay was 2 pg/tube, with an EC50 of 28 pg/tube. cGMP was assayed in hypothalamic homogenates. Hypothalamic tissues were homogenated in 1 ml Tris (0.05 m) 1 4 mm EDTA at 4 C using a Labsonic 2000 sonicator (B. Braun, Melsungen, Germany). Fifty ml of the homogenate were taken for the subsequent protein assay, and the remainder was heated for 3 min in a boiling water bath to coagulate proteins. After 20-min centrifugation at 4,000 rpm at 4 C, the supernatants were assayed for cGMP using a commercial RIA kit (Amersham, Little Chalfont, UK). Briefly, 100 ml of unknown or standard (the latter in the range 0.125– 8 pmol/tube) were incubated for 90 min at 4 C with 50 ml of tracer (about 9,500 cpm/tube) and 50 ml of appropriately diluted antiserum. Separation of bound from free cGMP was achieved by precipitation with 1 ml of 60% cold ammonium sulfate. Tubes were then centrifuged at 4,000 rpm for 6 min, and the supernatants were discarded. Pellets were resuspended in 1 ml of cold double-distilled water, and radioactivity was measured in 10 ml of scintillation fluid. The detection limit of the assay was 0.125 pmol/tube, and the EC50 was 1.64 pmol/tube. Proteins from hypothalamic homogenates were assayed using the bicinchoninic acid method (BCA Protein Assay Reagent, Pierce, Rockford, IL).

Drugs and chemicals Hemin HCl, biliverdin free base, indomethacin (INDO) free base, SNP, and 3-isobutyl-1-methyl-xanthine (IBMX) free base were obtained from Sigma. Tin-mesoporphyrin-9 (SnMP9) was obtained from Porphyrin Products Inc. (Logan, UT). LY 83583 (6-anilino-5, 8-quinolinequinone), a selective inhibitor of soluble guanylyl cyclase (20), was purchased from Calbiochem (La Jolla, CA). NG-nitro-l-arginine methyl ester (L-NAME) was obtained from Bachem Feinchemikalien AG (CH 4416, Bubendorf, Switzerland). Hemin and SnMP9 were dissolved in 100 mm NaOH; SNP in normal saline; biliverdin in absolute methanol; and INDO, IBMX, and LY 83583 in absolute ethanol. All substances were further diluted in incubation media to obtain working solutions. The latter had a final pH of 7.4. All drugs tested were found to produce no shift in the standard curves of the assays for ir-IL-1b, cGMP, and PGE2.

NO AND CO AFFECT IL-1b RELEASE FROM HYPOTHALAMUS Statistical analysis Results are given as means 6 1 sem, unless otherwise stated. In the experiments conducted to estimate cGMP content in the hypothalamic tissue, data are expressed as fmol/mg of protein. In experiments for assay of ir-IL-1b and PGE2 released into the incubation medium, data are expressed as pg/hypothalamus and pg/ml, respectively. Furthermore, ratios were calculated by dividing the amount of ir-IL-1b or PGE2 released in the second 20-min incubation period by those released in the previous period, the latter providing a paired control for the second 20-min period in each tissue block. Expression of data as ratios allows for compensation in irIL-1b and PGE2 variations among different tissue explants. To clarify the ratio calculation procedure, in Table 1, both the individual values of IL-1 released in two consecutive 20-min periods and the ratio calculated for each hypothalamus are reported. The data were then analyzed by ANOVA and subsequent Newman-Keul’s test for multiple comparisons among group means. Differences were considered statistically significant if P , 0.05.

1033

showed that NO and CO produced by the tissue under basal conditions did not influence cytokine release [data expressed as ir-IL-1b ratio, m 6 sem of (n) animals per group: controls 1.032 6 0.029 (4); 100 mm L-NAME 1.002 6 0.032 (4); 10 mm SnMP9 1.137 6 0.071 (4)].

Results The effects of increased NO and CO generation on ir-IL-1b release from hypothalamic explants

The NO-donor, SNP, was used to increase NO levels in the hypothalamic tissue. SNP stimulated, in a concentrationdependent manner, both basal (Fig. 1A) and KCl-stimulated ir-IL-1b release (Fig. 1B) from the hypothalamus. Statistically significant increases were obtained with 0.1 and 1 mm SNP. Hemin, the physiological substrate of HO, was used to generate CO in the hypothalamic explants. In the dose-range 1–10 mm, hemin did not modify basal ir-IL-1b release, but inhibited release stimulated by 56 mm KCl (Fig. 2, A and B). The inhibitory effect of hemin was attributed to CO, because it was blocked by the selective HO inhibitor, SnMP9 (21) (Fig. 2C); and the stable HO end-product, biliverdin, did not antagonize K1-induced ir-IL-1b release (not shown). Hemin also tended to reduce SNP-stimulated ir-IL-1b release, although inhibition did not reach statistical significance (Table 1). Experiments with L-NAME and SnMP9 given alone

FIG. 1. Sodium nitroprusside stimulates ir-IL-1b release from the hypothalamus, both under basal conditions (A) and in the presence of 56 mM KCl (B). Data are expressed as IL-1b ratio, means 6 1 SEM of eight animals per group. *, P , 0.05; **, P , 0.01 (both vs. controls); °°, P , 0.01 vs. KCl alone.

TABLE 1. The effect of hemin on SNP-stimulated ir-IL-1b release Treatment

ir-IL-1b, pg/hypothalamus

Ratios

1st 20 min

2nd 20 min

Controls

527.92 523.28 539.41 539.41

501.49 607.11 511.44 480.97

0.950 1.160 0.948 0.892

1 mM SNP

699.24 674.73 635.80 485.23

949.56 895.75 848.67 682.95

1.358 1.327 1.335 1.407

10 mM Hemin

574.86 527.47 577.96 521.07

605.48 552.07 516.80 549.78

1.053 1.047 0.894 1.055

SNP 1 hemin as above

593.36 573.07 504.85 536.40

703.64 693.40 698.46 716.96

1.185 1.209 1.383 1.337

Mean 6

SEM

(n)

0.988 6 0.059 (4)

1.357 6 0.018 (4)a

1.012 6 0.039 (4)

1.259 6 0.036 (4)a

Data are individual values of picograms of ir-IL-1b released in two consecutive 20-min periods, and their respective ratios. For the latter, the means 6 SEM are also reported. a P , 0.01 vs. controls.

1034

NO AND CO AFFECT IL-1b RELEASE FROM HYPOTHALAMUS

Endo • 1998 Vol 139 • No 3

was able to completely reverse the inhibitory effect of hemin on stimulated ir-IL-1b release (Table 3). INDO given alone had no significant effect on K1-stimulated cytokine release [data expressed as ir-IL-1b ratio, m 6 sem of (n) animals per group: controls 0.981 6 0.041 (9); 56 mm KCl 1.164 6 0.036 (9), P , 0.01 vs. controls; 56 mm KCl 1 1 mg/ml INDO 1.125 6 0.021 (9), P , 0.01 vs. controls, NS difference vs. KCl alone]. The effects of CO and NO on soluble guanylyl cyclase activity in the hypothalamus

Guanylyl cyclase activity was assessed in terms of cGMP production. As expected, the addition of 1 mm IBMX to the incubation medium significantly increased cGMP content in the hypothalamic tissue. Incubation of the explants in medium saturated with CO did not further increase cGMP levels (Fig. 3A). Hemin was also used to increase CO levels in the tissue; however, the former is known to exert a direct inhibitory effect on sGC (22). In fact, we found that 10 mm hemin significantly decreased cGMP levels within the hypothalamic tissue [data expressed as fmol of cGMP/mg of protein, m 6 sem of (n) animals per group: controls 74.24 6 10 (6); 10 mm hemin 46.03 6 11.8 (6), P , 0.05 vs. controls]. Conversely, 1 mm SNP produced a significant increase in cGMP levels in the presence of IBMX (Fig. 3B). The effects of CO and NO on COX activity in the hypothalamus

In these experiments, COX activity was monitored by measuring the concentration of a COX end-product, PGE2, released into the incubation medium. This was taken as an index of total PG produced by the hypothalamus (23). Incubation of the tissue with a CO-saturated solution produced a significant increase in PGE2 (Table 4): this finding matches our previous observations that hemin increases PGE2 production and release by the rat hypothalamus in vitro (23). Once again, the effect of SNP diverged from those of CO or hemin, because 1 mm of the drug did not elicit any change in PGE2 released from the hypothalamic explants (Table 4). Discussion

FIG. 2. One (A) and 10 mM hemin (B) inhibit K1-stimulated ir-IL-1b release. Sn-mesoporphyrin-9 counteracts the inhibitory effect of hemin (C). Data are expressed as IL-1b ratio, means 6 1 SEM of eight animals per group. **, P , 0.01 vs. controls; °, P , 0.05; °°, P , 0.01 (both vs. KCl alone); #, P , 0.05 vs. KCl 1 Hemin.

The effects of specific antagonists of sGC and COX on NOand CO-modulated IL-1b release

The selective inhibitor of sGC, LY83583, at 1 mm, significantly reduced the increase in ir-IL-1b release elicited by 1 mm SNP but failed to counteract the inhibitory effect of 10 mm hemin on K1-stimulated cytokine release (Table 2). LY83583 given alone had no effect on basal ir-IL-1b release but reduced release stimulated by KCl (Table 2). Conversely, 1 mg/ml of the COX inhibitor, INDO, did not significantly antagonize the stimulatory effect of SNP but

In this paper, we report that the generation of NO and CO in rat hypothalamic explants is associated with opposing effects on ir-IL-1b release from the tissue, with NO producing an increase in cytokine release, whereas the increase in CO levels after treatment with hemin inhibited K1-stimulated ir-IL-1b release. The opposite actions of the two gases seemed to be caused by the activation of different signaling pathways: NO exerted its effect via sGC, whereas the inhibitory effect of CO was attributed to an action on COX. The increased cGMP production induced by NO in the CNS is known to elicit both stimulatory and inhibitory responses on the release of neurotransmitters or neurohormones. Stimulation is often associated with an action on neurons signaling via excitatory amino acids (24). Among inhibitory effects attributed to the activation of the NOcGMP pathway are the release of CRH and vasopressin (14, 25). Many effects of cGMP are thought to be mediated by the stimulation of cGMP-dependent protein kinase (24), but ad-

NO AND CO AFFECT IL-1b RELEASE FROM HYPOTHALAMUS

1035

TABLE 2. The effects of the sGC inhibitor, LY 83583, on the release of ir-IL-1b under basal conditions, after stimulation by SNP and KCl, as well as on hemin inhibition of K1-stimulated cytokine release pg ir-IL-1b/hypothalamus

Treatment

Ratios

1st 20 min

2nd 20 min

Controls 1 mM SNP 1 mM LY 83583 SNP 1 LY 83583 as above

654.76 6 14.97 (8) 602.42 6 13.75 (8) 656.24 6 15.11 (8) 649.56 6 7.70 (8)

651.56 6 16.47 (8) 779.71 6 10.35 (8) 635.27 6 18.20 (8) 738.46 6 15.30 (8)

0.997 6 0.027 1.300 6 0.041a 0.970 6 0.047 1.130 6 0.026c/b

Controls 56 mM KCl 56 mM KCl 1 10 mM Hemin 56 mM KCl 1 1 mM LY 83583 KCl 1 Hemin 1 LY 83583 as above

570.59 6 15.00 (9) 557.03 6 8.27 (8) 594.55 6 9.18 (9) 572.96 6 9.89 (9) 589.85 6 9.09 (9)

524.00 6 13.75 (9) 657.43 6 20.61 (8) 582.41 6 6.94 (9) 615.93 6 16.55 (9) 613.52 6 18.08 (9)

0.920 6 0.017 1.189 6 0.032a 0.980 6 0.016d 1.076 6 0.029d/a 1.039 6 0.023d

Data are expressed both as pg of ir-IL-1b/hypothalamus and as ratios, means 6 a P , 0.01 vs. controls. b P , 0.05 vs. controls. c P , 0.01 vs. SNP alone. d P , 0.01 vs. 56 mM KCl alone.

SEM

of (n) animals per group.

TABLE 3. The effects of the COX inhibitor, INDO, on the release of ir-IL-1b under basal conditions, after stimulation by SNP, as well as on hemin inhibition of K1-stimulated cytokine release pg ir-IL-1b/hypothalamus

Treatment

Ratios

1st 20 min (n)

2nd 20 min (n)

Controls 1 mM SNP 1 mg/ml INDO SNP 1 INDO as above

553.44 6 17.97 (8) 481.60 6 19.20 (8) 529.45 6 18.60 (6) 490.74 6 31.74 (8)

530.97 6 12.01 (8) 626.71 6 17.35 (8) 503.93 6 21.97 (6) 585.15 6 37.65 (8)

0.965 6 0.029 1.318 6 0.062a 0.953 6 0.034 1.210 6 0.073a

Controls 56 mM KCl 56 mM KCl 1 10 mM hemin KCl 1 hemin as above 1 1 mg/ml INDO

540.31 6 17.75 (8) 533.31 6 22.92 (8) 584.41 6 14.71 (8) 497.97 6 15.40 (8)

499.22 6 22.90 (8) 629.49 6 12.28 (8) 605.21 6 22.68 (8) 612.32 6 15.44 (8)

0.931 6 0.049 1.199 6 0.043a 1.037 6 0.037b 1.240 6 0.053c/a

Data are expressed both as pg of ir-IL-1b/hypothalamus and as ratios, means 6 a P , 0.01 vs. controls. b P , 0.05 vs. 56 mM KCl alone. c P , 0.01 vs. hemin alone.

SEM

of (n) animals per group.

TABLE 4. The effects of SNP (upper panel) and CO-saturated solutions (lower panel) on PGE2 production and release by hypothalamic explants Treatment

pg PGE2/ml

Ratios

1st 20 min (n)

2nd 20 min (n)

Controls 1 mM SNP

90.16 6 12.50 (15) 138.78 6 26 (16)

120.35 6 23 (15) 160.50 6 31.90 (16)

1.335 6 0.088 1.156 6 0.072

Controls CO

168.50 6 23.11 (6) 185.87 6 30.85 (5)

131.30 6 19.18 (6) 192.27 6 22.23 (5)

0.780 6 0.052 1.033 6 0.100a

Data are expressed both as pg of PGE2/ml of incubation medium and as ratios, means 6 a P , 0.05 vs. controls.

ditional effects of the cyclic nucleotide are thought to be caused by activation or modulation of specific cGMP-dependent ion channels, leading to enhanced influx of various cations, including Ca21 (26). However, it is unlikely that changes seen in brain slices are simply caused by a direct action of cGMP on nerve terminals, because complex interaction between neurons and glia may take place (24). Thus, although the activation of sGC by NO may well account for the observed effect of the NO-donor, SNP, on ir-IL-1b release in vitro, our experimental approach does not allow us to conclude that neuronal IL-1b is specifically involved. sGC has previously been regarded as the primary signal-

SEM

of (n) animals per group.

ing mechanism for endogenous CO (27). This does not seem to be the case for the effects of the gas on ir-IL-1b release from the hypothalamus. Instead, the present evidence suggests that CO acts on cytokine release via the stimulation of COX. We have previously shown that the precursor of CO through the HO pathway, hemin, induces concentration-dependent increases in PGE2 release from hypothalamic explants in the same experimental model as that used in this study (23). This effect was specifically attributed to the generation of CO after HO-dependent hemin catabolism, because it was blocked by HO inhibitors such as SnMP9 and Zn-protoporphyrin-9, as well as by ferrous hemoglobin, which inactivates CO by

1036

NO AND CO AFFECT IL-1b RELEASE FROM HYPOTHALAMUS

FIG. 3. The effects of CO-saturated solutions (A) and SNP (B) on the production of cGMP in hypothalamic tissue. Data are expressed as fmol cGMP/mg of protein, means 6 1 SEM of six to eight animals per group. *, P , 0.05 vs. controls.

direct binding. Moreover, the stable HO end-product, biliverdin, did not enhance PGE2 release from the explants (23). The direct evidence obtained here with CO-saturated solutions matches previous findings with hemin, thereby confirming the stimulatory role of CO on PG production in this experimental paradigm. Though PGE2 was taken in these studies as a marker of COX activity, it is not known whether PGE2 itself or other end-products of COX are responsible for the inhibition of K1-stimulated ir-IL-1b release, although we have previously observed that PGE2 inhibits unstimulated IL-1 release from hypothalamic explants (our unpublished data). We have been unable to show any significant effect of the NO-donor, SNP, on PGE2 production and release from the hypothalamus. This finding is in conflict with evidence showing that NO stimulates COX in various in vitro models, including short-term incubations of rat medio-basal hypothalami (28, 29). More recently, however, it has been demonstrated that NO can also inhibit COX activity (in particular, the inducible isoform COX2) in microglial cells (30). In our experimental paradigm, we cannot discriminate between the COX isoforms implicated in PGE2 production; but COX2 might play a role, insofar as it is the isoform constitutively expressed in neurons (31). Moreover, the possibility that NO activates COX in the same manner as it does sGC (i.e. binding to the heme moiety of the enzyme) has been questioned (32). Though the issue of the actions exerted by NO on COX remains under dispute, in any case, the COX pathway does not seem to be involved in SNP stimulation of IL-1b release, because its inhibition by INDO had no significant effect on SNP-stimulated cytokine release.

Endo • 1998 Vol 139 • No 3

As far as the lack of CO effect on cGMP production is concerned, a fundamental factor may be that the gas is only a weak activator of sGC, producing only 1- or 2-fold increases in enzymatic activity (33, 34). Therefore, effects of CO mediated by its activation of sCG are likely to be detected only in those brain areas, such as the hippocampus, endowed with high levels of sGC (17). On the contrary, two major studies have shown very low sGC levels in the rat hypothalamus (35, 36), which is consistent with our failure to show any relationship between the effects of CO and sGC activity in hypothalamic explants. In contrast, NO was reported to be more active than CO in stimulating sGC, because the former, under directly comparable conditions, increases cGMP production by a factor of 10 or more (33, 34). This may account for the fact that NO is able to induce significant increases in cGMP levels in the hypothalamus, as shown by us and others (37), in spite of low levels of hypothalamic sGC. In conclusion, it is apparent that the pattern of signaling mechanisms by the gases, NO and CO, is growing increasingly complex. Though it was previously thought that activation of sGC was the dominant, if not the sole, signal transduction mechanism of NO, subsequent studies have revealed that the latter may also exert biological activities via alternative mechanisms such as ADP ribosylation, activation of COX, protein nitrosylation, and the formation of peroxynitrite ions (38). The same profile seems now to emerge with CO, as alternative signaling pathways, such as the enzymes cytochrome P450 (39) or COX, are being proposed. The picture is even more complex if we consider that the two gases may or may not (as shown in the present study) share a common signaling pathway in different brain regions, depending on the amount of target proteins at the sites of gas generation and the relative efficacy of the gases in binding and activating such proteins. Thus, neuroregulation by gaseous transmitters assumes an increasingly important role in neuroendocrine function as their activities become more specifically defined. References 1. Busbridge NJ, Grossman AB 1991 Stress and the single cytokine: interleukin modulation of the pituitary-adrenal axis. Mol Cell Endocrinol 82:C209 –C214 2. Rivest S, Rivier C 1995 The role of corticotropin-releasing factor and interleukin-1 in the regulation of neurons controlling reproductive functions. Endocr Rev 16:177–199 3. Tilders FJH, De Rijk RH, Van Dam A-M, Vincent VAM, Schotanus K, Persons JHA 1994 Activation of the hypothalamo-pituitary-adrenal axis by bacterial endotoxins: routes and intermediate signals. Psychoneuroendocrinology 19:209 –232 4. Lechan RM, Toni R, Clark BD, Cannon JG, Shaw AR, Dinarello CA, Reichlin S 1990 Immunoreactive interleukin-1b localization in the rat forebrain. Brain Res 514:135–140 5. Shintani F, Nakaki T, Kanba S, Sato K, Yagi G, Shiozawa M, Aiso S, Kato R, Asai M 1995 Involvement of interleukin-1 in immobilization stress-induced increase in plasma adrenocorticotropic hormone and in release of hypothalamic monoamines in the rat. J Neurosci 15:1961–1970 6. Costa A, Poma A, Navarra P, Forsling ML, Grossman A 1996 Gaseous transmitters as new agents in neuroendocrine regulation. J Endocrinol 149:199 –207 7. Bredt DS, Hwang PM, Snyder SH 1990 Localization of nitric oxide synthase indicating a neural role for nitric oxide. Nature 347:768 –770 8. Grossman A, Rossmanith WG, Kabigting EB, Cadd G, Clifton D, Steiner RA 1994 The distribution of hypothalamic nitric oxide synthase mRNA in relation to gonadotropin-releasing hormone neurons. J Endocrinol 140:R5–R8 9. Sun Y, Rotenberg MO, Maines MD 1990 Developmental expression of heme oxygenase isozymes in rat brain. Two HO-2 mRNAs are detected. J Biol Chem 265:8212– 8217 10. Ewing JF, Haber SN, Maines MD 1992 Normal and heat-induced pattern of

NO AND CO AFFECT IL-1b RELEASE FROM HYPOTHALAMUS

11. 12. 13.

14. 15. 16. 17. 18. 19.

20. 21. 22. 23.

24.

expression of heme oxygenase-1 (HSP32) in rat brain: hyperthermia causes rapid induction of mRNA and protein. J Neurochem 58:1140 –1149 Rivier C, Shen GH 1994 In the rat, nitric oxide modulates the response of the hypothalamic-pituitary-adrenal axis to interleukin-1b, vasopressin and oxytocin. J Neurosci 14:1985–1993 Rivier C 1995 Blockade of nitric oxide formation augments adrenocorticotropin released by blood-borne interleukin-1b: role of vasopressin, prostaglandins, and a1-adrenergic receptors. Endocrinology 136:3597–3603 Pozzoli G, Mancuso C, Mirtella A, Preziosi P, Grossman A, Navarra P 1994 Carbon monoxide as a novel neuroendocrine modulator: inhibition of stimulated corticotropin-releasing hormone release from acute rat hypothalamic explants. Endocrinology 135:2314 –2317 Costa A, Trainer P, Besser M, Grossman A 1993 Nitric oxide modulates the release of corticotropin-releasing hormone from the rat hypothalamus in vitro. Brain Res 605:187–192 Yasin S, Costa A, Trainer P, Windle R, Forsling ML, Grossman A 1993 Nitric oxide modulates the release of vasopressin from rat hypothalamic explants. Endocrinology 133:1466 –1469 Tringali G, Mancuso C, Mirtella A, Pozzoli G, Parente L, Preziosi P, Navarra P 1996 Evidence for the neuronal origin of immunoreactive interleukin-1b released by hypothalamic explants. Neurosci Lett 219:143–146 Zhuo M, Small SA, Kandel ER, Hawkins RD 1993 Nitric oxide and carbon monoxide produce activity-dependent long-term synaptic enhancement in hippocampus. Science 260:1946 –1950 Hardman JG, Sutherland EW 1969 Guanylyl cyclase, an enzyme catalyzing the formation of guanosine 39,59-monophosphate from guanosine triphosphate. J Biol Chem 244:6363– 6370 Navarra P, Pozzoli G, Brunetti L, Ragazzoni E, Besser M, Grossman A 1992 Interleukin-1b and interleukin-6 specifically increase the release of prostaglandin E2 from rat hypothalamic explants in vitro. Neuroendocrinology 56:61– 68 Schmidt MJ, Sawyer BD, Truex LL, Marshall WS, Fleisch JH 1985 LY 83583: an agent that lowers intracellular levels of cyclic guanosine 39,59-monophosphate. J Pharmacol Exp Ther 232:764 –769 Meffert MK, Haley JE, Schuman EM, Schulman H, Madison DV 1994 Inhibition of hippocampal heme oxygenase, nitric oxide synthase, and long-term potentiation by metalloporphyrins. Neuron 13:1225–1233 Ignarro LJ, Ballot B, Wood KS 1984 Regulation of soluble guanylate cyclase activity by porphyrins and metalloporphyrins. J Biol Chem 259:6201– 6207 Mancuso C, Pistritto G, Tringali G, Grossman AB, Preziosi P, Navarra P 1997 Evidence that carbon monoxide stimulates prostaglandin endoperoxide synthase activity in rat hypothalamic explants and in primary cultures of rat hypothalamic astrocytes. Mol Brain Res 45:294 –300 Wang X, Robinson PJ 1997 Cyclic GMP-dependent protein kinase and cellular signaling in the nervous system. J Neurochem 68:443– 456

1037

25. Akamatsu N, Inenaga K, Yamashita H 1993 Inhibitory effects of natriuretic peptides on vasopressin neurons mediated through cGMP and cGMP-dependent protein kinase in vitro. J Neuroendocrinol 5:517–522 26. Finn JT, Grunwald ME, Yau KW 1996 Cyclic nucleotide gated-ion channels: an extended family with diverse functions. Annu Rev Physiol 58:395– 426 27. Verma A, Hirsch DJ, Glatt CE, Ronnett GV, Snyder SH 1993 Carbon monoxide: a putative neural messenger. Science 259:381–384 28. Salvemini D, Misko TP, Masferrer JL, Seibert K, Currie MG, Needleman P 1993 Nitric oxide activates cyclooxygenase enzymes. Proc Natl Acad Sci USA 90:7240 –7244 29. Rettori V, Gimeno M, Lyson K, McCann SM 1992 Nitric oxide mediates norepinephrine-induced prostaglandin E2 release from the hypothalamus. Proc Natl Acad Sci USA 89:11543–11546 30. Minghetti L, Polazzi E, Nicolini A, Cre´monin C, Levi G 1996 Interferon-g and nitric oxide down-regulate lipopolysaccharide-induced prostanoid production in cultured rat microglial cells by inhibiting cyclooxygenase-2 expression. J Neurochem 66:1963–1970 31. Yamagata K, Andreasson KI, Kaufmann WE, Barnes CA, Worley PF 1993 Expression of mitogen-inducible cyclooxygenase in brain neurons: regulation by synaptic activity and glucocorticoids. Neuron 11:371–386 32. Tsai A 1994 How does NO activate hemeproteins? FEBS Lett 341:141–145 33. Kharitonov VG, Sharma VS, Pilz RB, Magde D, Koesling D 1995 Basis of guanylate cyclase activation by carbon monoxide. Proc Natl Acad Sci USA 92:2568 –2571 34. Burstyn NJ, Yu AE, Dierks EA, Hawkins BK, Dawson JH 1995 Studies of the heme coordination and ligand binding properties of soluble guanylyl cyclase (sGC): characterization of Fe(II)sGC and Fe(II)sGC(CO) by electronic absorption and magnetic circular dichroism spectroscopies and failure of CO to activate the enzyme. Biochemistry 34:5896 –5903 35. Matsuoka I, Giuili G, Poyard M, Stengel D, Parma J, Guellaen G, Hanoune J 1992 Localization of adenylyl and guanylyl cyclase in rat brain by in situ hybridization: comparison with calmodulin mRNA distribution. J Neurosci 12:3350 –3360 36. Burgunder JM, Cheung PT 1994 Expression of soluble guanylyl cyclase gene in adult rat brain. Eur J Neurosci 6:211–217 37. Bhat G, Mahesh VB, Aguan K, Brann DW 1996 Evidence that brain nitric oxide synthase is the major nitric oxide synthase isoform in the hypothalamus of the adult female rat and that nitric oxide potently regulates hypothalamic cGMP levels. Neuroendocrinology 64:93–102 38. Dawson VL, Dawson TM 1996 Nitric oxide actions in neurochemistry. Neurochem Int 29:97–110 39. Coceani F, Kelsey L, Seidlitz E 1996 Carbon monoxide-induced relaxation of the ductus arteriosus in the lamb: evidence against the prime role of guanylyl cyclase. Br J Pharmacol 118:1689 –1696