Lipopolysaccharide (LPS) administration induces hypothalamic nitric oxide (NO); NO is antipyretic in the preoptic area (POA), but its mechanism of action is uncertain. LPS also stimulates the release of preoptic norepinephrine (NE), which mediates fever onset. Because NE upregulates NO synthases and NO induces cyclooxygenase (COX)-2-dependent PGE2, we investigated whether NO mediates the production of this central fever mediator. Conscious guinea pigs with intra-POA microdialysis probes received LPS intravenously (2 μg/kg) and, thereafter, an NO donor (SIN-1) or scavenger (carboxy-PTIO) intra-POA (20 μg/μl each, 2 μl/min, 6 h). Core temperature (Tc) was monitored constantly; dialysate NE and PGE2 were analyzed in 30-min collections. To verify the reported involvement of α2-adrenoceptors (AR) in PGE2 production, clonidine (α2-AR agonist, 2 μg/μl) was microdialyzed with and without SIN-1 or carboxy-PTIO. To assess the possible involvement of oxidative NE and/or NO products in the demonstrated initially COX-2-independent POA PGE2 increase, (+)-catechin (an antioxidant, 3 μg/μl) was microdialyzed, and POA PGE2, and Tc were determined. SIN-1 and carboxy-PTIO reduced and enhanced, respectively, the rises in NE, PGE2, and Tc produced by intravenous LPS. Similarly, they prevented and increased, respectively, the delayed elevations of PGE2 and Tc induced by intra-POA clonidine. (+)-Catechin prevented the LPS-induced elevation of PGE2, but not of Tc. We conclude that the antipyretic activity of NO derives from its inhibitory modulation of the LPS-induced release of POA NE. These data also implicate free radicals in POA PGE2 production and raise questions about its role as a central LPS fever mediator.
- nitric oxide donors
- nitric oxide scavengers
- prostaglandin E2
- body temperature
- free radicals
reports that the gaseous transmitter nitric oxide (NO) stimulates the biosynthesis of PGE2 by increasing the activities of both isoforms of cyclooxygenase (COX; for reviews, see Refs. 10 and 49) have prompted suggestions that NO could have a propyretic function in the central mediation of the febrile response to pyrogens (for reviews, see Refs. 19, 65, 70, 73, and 74). Indeed, the three isoforms of nitric oxide synthase [endothelial (e), neural (n), and inducible (i) NOS], the enzyme that converts l-arginine to citrulline and NO, occur in the hypothalamus (7, 22), and circulating bacterial endotoxic lipopolysaccharides (LPS) and pyrogenic cytokines stimulate the release of NO in the preoptic (POA, the fever-producing locus) and paraventricular regions of the hypothalamus (23, 37, 64, 67, 80, 85, 88). Hence NO could mediate fever production by further upregulating COX-2, the isozyme implicated in inflammation (69), and thereby enhance the febrile response (3, 5, 40–43). Other studies have indicated that, on the contrary, NO inhibits, in particular, LPS-induced COX-2 activity (9, 11, 12, 78) and therefore prevents the development of fever (46). Indeed, previous data regarding a potential role of central NO in fever have predominantly indicated antipyretic effects (20, 36, 50, 56, 60, 74, 76). The mechanism of its antipyretic action has so far, however, eluded explanation.
In studies not related to fever research, it has been shown that norepinephrine (NE) activates e-, n-, and iNOS in brain (1, 8, 21). Thus, in the case of the control of the release of luteinizing hormone-releasing hormone (LHRH) from the hypothalamus, the NE-induced increased activity of NOS has been specifically attributed to the activation of α1-adrenoceptors (AR) on NOergic neurons (8). The inverse, an ultrashort loop negative feedback by which NO at higher concentrations, as in inflammatory states, inhibits the release of NE, thereby attenuating the action of NO and decreasing the release of LHRH, has also been described (68). Furthermore, the modulation by NO of the release of NE from locus coeruleus neurons has also recently been reported (33). These findings therefore indicated that NO could, in turn, modulate monoaminergic neurotransmission in the central nervous system. Evidence for the putative involvement of NE and α1-AR in the initiation of fever was presented in our companion paper (15). Hence the existence of a mutually antagonistic NE-NO interaction in the POA in response to intravenous LPS is plausible. Of particular relevance in this regard are the findings that subdiaphragmatic vagotomy attenuates the LPS- and interleukin (IL)-1β-induced increases in hypothalamic microdialysate NE (16, 30, 83, 84) and NO (29). The febrile response to LPS (17, 66, 71, 82, 83) and the increase in LPS-induced preoptic PGE2 (66; for review, see Ref. 6) have been shown previously to be similarly inhibited by subdiaphragmatic vagotomy.
To determine the potential NE-NO interplay in fever induction, we investigated whether an NO donor or scavenger microdialyzed in the POA of conscious guinea pigs would enhance or depress, respectively, the secretion of NE in the POA in response to intravenous LPS (15) and thereby modulate one and/or both phases of the biphasic febrile response to intravenous LPS. The preoptic COX-2-catalyzed PGE2 that is generated consequent to the noradrenergic activation of local α2-AR (14) was also evaluated.
Finally, on the presumption that the increase in COX-2-independent preoptic PGE2 observed during the early phase of the febrile rise in our companion study (experiment 2 in Ref. 15) could be derived from the peroxidation of arachidonic acid by oxidative products of NE metabolism and/or by NO and its related reactive oxygen species (ROS), we administered the antioxidant (+)-catechin to prevent the accumulation of ROS and thereby evaluate their role, if any, in the increase in preoptic PGE2.
MATERIALS AND METHODS
The animals, drugs, surgical procedures, physiological measurements, biochemical analyses, and statistical methods employed in this study were completely analogous to those of our companion study (15), with which it was continuous. All animal protocols were approved by our institutional Animal Care and Use Committee and fully conform to the standards established by the United States Animal Welfare Act and by the American Physiological Society documents entitled Guiding Principles for Research Involving Animals and Human Beings. The additional drugs used in this study were purchased from Sigma-Aldrich (St. Louis, MO), viz., clonidine hydrochloride (catalog no. C7897), 3-morpholinosydnonimine hydrochloride (SIN-1; catalog no. M5793), 2-[4-carboxyphenyl]-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide, potassium salt (carboxy-PTIO, catalog no. C221), and (+)-catechin hydrate [catalog no. C1788, an antioxidant flavonoid (34)]; all were dissolved in artificial cerebrospinal fluid (aCSF), which was also the control perfusate. These solutions were prepared just before use and stored in amber glass vials at room temperature.
Experiment 1: Effects of an NO donor and scavenger microdialyzed in the POA on the level of preoptic NE released in response to intravenous LPS.
Collection of microdialysate effluents was begun immediately after the 90-min stabilization period (time 0; see Ref. 15) and continued for 300 min. The NO donor SIN-1 or the NO scavenger carboxy-PTIO (both 20 μg/μl of aCSF) was perfused in the POA throughout the same period. Pyrogen-free saline (PFS, 0.6 ml/kg) or LPS (from Salmonella enteritidis; 2 μg/kg in 0.6 ml of PFS/kg) was injected intravenous 60 min after the initiation of theses collections. The animals were conscious throughout the experiments. The samples were analyzed for their NE content as described in our companion paper (15).
Experiment 2: Effects of an NO donor and scavenger microdialyzed in the POA on the thermal and preoptic PGE2 responses to an α2-AR agonist comicrodialyzed in the POA.
Immediately following the 90-min stabilization period, clonidine (a selective α2-AR agonist; 2 μg/μl of aCSF), and clonidine + SIN-1 or carboxy-PTIO (both 20 μg/μl of aCSF) were microdialyzed in the POA of conscious guinea pigs for 6 h; aCSF alone was the control solution. This design was chosen to replicate that of our original study. Core temperatures (Tc) and preoptic PGE2 levels were determined as described in our companion paper (15).
Experiment 3: Effects of an NO donor and scavenger microdialyzed in the POA on the febrile and preoptic PGE2 responses to intravenous LPS.
Immediately following the 90-min stabilization period, aCSF, SIN-1, or carboxy-PTIO (both 20 μg/μl of aCSF) was microdialyzed in the POA of conscious guinea pigs for 6 h beginning with the intravenous injection of PFS (0.6 ml/kg) or LPS (2 μg/kg in 0.6 ml of PFS/kg). Tc values and preoptic PGE2 levels were determined as described previously (15).
Experiment 4: Effect of an antioxidant microdialyzed in the POA on the febrile and preoptic PGE2 responses to intravenous LPS.
Immediately following the stabilization period, the antioxidant (+)-catechin hydrate (3 μg/μl of aCSF) was microdialyzed in the POA of conscious guinea pigs for the duration of the experiment. PFS or LPS was injected intravenously at time 0. Tc values and preoptic PGE2 levels were determined as before.
The intra-POA microdialysis of the NO donor SIN-1 (in aCSF) significantly depressed the amount (from ca. 800 to ca. 500 pg/ml) and shortened the presence (from ca. 90 to ca. 30 min) of NE released in the POA in response to the intravenous administration of LPS compared with the effect on these variables of the microdialysis of aCSF alone. By contrast, the NO scavenger carboxy-PTIO (in aCSF) significantly augmented the level (to ca. 1,250 pg/ml) and prolonged the presence (to ca. 180 min) of NE released in the POA in response to intravenous LPS. These results are depicted in Fig. 1, A and B.
The intra-POA comicrodialysis of the α2-AR agonist clonidine and SIN-1 delayed the onset and lengthened the duration of the early fall of Tc induced by clonidine alone, preventing its late rise; however, it inhibited both the early fall and late rise of PGE2 [previously shown to be COX-2 dependent (14)] induced by clonidine alone. The comicrodialysis of clonidine and carboxy-PTIO, on the other hand, attenuated the early fall of both Tc and PGE2 but significantly increased their late rises. Intra-POA-microdialyzed clonidine induced the same effects observed by us earlier (14), i.e., an initial decrease followed by a rise of both Tc and preoptic PGE2 amounts (Fig. 2, A and B). SIN-1 delayed the onset but not the magnitude of the initial Tc fall but prevented its subsequent rise; however, it abrogated both the fall and rise of PGE2 (Fig. 2, C and D). Carboxy-PTIO attenuated the initial falls in both Tc and PGE2 levels and delayed, but did not prevent their subsequent, full rises (Fig. 2, E and F); indeed, Tc was significantly elevated compared with its control (Fig. 2A vs. 2E). Neither the microdialysis of aCSF, SIN-1, nor of carboxy-PTIO alone affected the Tc or preoptic PGE2 levels of these animals (data not shown).
The intra-POA microdialysis of SIN-1 suppressed both the early and late rises of Tc and PGE2 induced by intravenous LPS, whereas that of carboxy-PTIO significantly increased them. Thus SIN-1-microdialyzed intra-POA in conscious guinea pigs did not affect the onset, but greatly reduced the magnitude, of the first of the prototypic two Tc rises evoked by intravenous LPS; and it abolished the second rise so that the fever continued at its first level until it abated after ∼3 h (compare Fig. 3, A and C). This treatment also abrogated the PGE2 increases associated with both Tc rises (compare Fig. 3, B and D). Carboxy-PTIO, on the other hand, greatly enhanced the late, but not the early, LPS-induced Tc and PGE2 rises so that the magnitudes of both variables were significantly increased during the course of this phase (Fig. 3, E and F). Neither the microdialysis of aCSF, SIN-1, nor of carboxy-PTIO alone affected the basal Tc or preoptic PGE2 levels of these animals (data not shown).
The intra-POA microdialysis of (+)-catechin did not change the quick onset and height of the first LPS-induced febrile rise, but it attenuated the second rise (Fig. 4A vs. Fig. 3A), reminiscent of the effect of intra-POA yohimbine (Fig. 2E and Ref. 15). Remarkably, it also totally prevented the rise in preoptic PGE2 (Fig. 4B) characteristically associated with both febrile rises (Fig. 3B).
These results indicate that the administration of an NO donor (SIN-1) suppresses the increase in preoptic PGE2 levels induced by intravenous LPS. It also eliminates both the early PGE2 fall and its later rise caused by intra-POA clonidine. By contrast, the removal of NO (via carboxy-PTIO) enhances both the LPS- and clonidine-associated PGE2 rises, but not the clonidine-associated fall. Furthermore, the addition of NO depresses the LPS-induced intra-POA secretion of NE, whereas its elimination augments it. The role of NE in the induction of PGE2 and, hence, its function in fever production were the subject of our companion study (15). Taken together with those findings, the present results therefore indicate that intra-POA-microdialyzed SIN-1, an NO donor, and carboxy-PTIO, an NO scavenger, inhibits and promotes, respectively, the entire course of the febrile response to LPS by reducing and augmenting, respectively, the preoptic release of NE and, in this manner, regulating the activation of both the postsynaptic α1 (PGE2-independent)- and α2 (PGE2-dependent)-AR induced by NE, and, in the case of the latter, also the consequent production of PGE2. Hence, in view of its restraining effect on the release of NE and, thereby, its modulation of the actions of NE on both α-ARs, NO, presumably released in the POA coincidentally with or very shortly after NE, would appear to act as a negative regulatory factor, i.e., to be a central endogenous antipyretic mediator of LPS fever. To our best knowledge, the present in vivo modulating effects of an NO donor and scavenger on preoptic NE secretion and their interaction in the mechanism of fever production are novel observations. No study has previously employed the in vivo intra-POA microdialysis technique in conscious animals to validate the antipyretic activity of NO. Finally, these data show for the first time that an antioxidant, (+)-catechin, microdialyzed in the POA prevents the rise in preoptic PGE2 induced by intravenous LPS but does not affect the associated Tc rise.
Analyzing the present results piece by piece, first, the involvement of central NO in the control of body temperature, both in health and in disease, is well documented (for review, see Refs. 19, 65, 70, 73, 74). Previous data regarding its potential role in fever have, however, been conflicting, some indicating no effect (31, 57) and others pyretic (3, 5, 40–43, 53) or antipyretic (20, 46, 50, 56, 60, 76) effects. This confusion of results is probably because of methodological differences among the studies, e.g., routes and timing of the injections, doses and types of the various agents administered, animal species, etc. Indeed, the time of appearance, cell source and thermal effect of the NO generated and, hence, also, the NOS isoforms activated vary with the nature of the pyrogenic stimulus (36). Nevertheless, the preponderance of the data suggest that NO in the brain attenuates fever. The present results, therefore, conform to the majority findings and, moreover, suggest a probable mechanism for its antipyretic effect.
Second, the association between NO and PGE2 production found in this study is also not a new discovery. Indeed, the activation of COX enzymes by NO in most cells is well established (for review, see Refs. 10 and 49). Regarding the brain, in support, the cytokine-induced, COX-2-mediated release of PGE2 by astrocytes, which may be a source of preoptic PGE2 in the febrile response (15), has been shown to be modulated in vitro by iNOS (48), and the release of PGE2 from brain tissue stimulated by IL-1β was curtailed by an iNOS synthesis inhibitor (57). These observations would seem to support the case for a propyretic synergy between NO and PGE2, as suggested by various workers (40–43, 50, 56). Other studies, however, have indicated that, to the contrary, inflamed brain tissue from iNOS knockout mice contain less PGE2 than their controls (52). It has also been shown that NO inhibits both the expression and activity of COX-2 (9, 11, 77). These results would imply that the absence of iNOS-catalyzed NO upregulates COX-2 expression and PGE2 production and, hence, enhances fever. From this, the concept developed that, under physiological conditions, the c- or eNOS-mediated release of NO activates COX-1, resulting in the basal release of PGE2, whereas in inflammatory states the additional expression of iNOS contributes larger amounts of NO and induces the consequent activation of COX-2. The present data would be consistent with the latter conclusion.
Third, the reciprocal interaction between NE and NO in the hypothalamus found in this study is also not a new observation, but it has not been considered previously in the context of fever production. Thus, in experiments related to the regulation of the secretion of various neuropeptides by the paraventricular nucleus (PVN), it was shown that NE quickly upregulates nNOS and, more slowly, iNOS by activating α1-AR on neurons and astrocytes, respectively, inducing the release of NO (1, 8, 21, 58). The released NO, in high concentrations, in turn inhibits both the basal and stimulated release of NE, thus providing a negative feedback that limits the further stimulation of α1-AR (33, 68). It is possible that the relatively transitory rise of NE and its differentiation from the sustained rise of Tc, as observed in the present and previous studies (for reviews, see Ref. 15), was the result of this mechanism, i.e., the inhibition of noradrenergic terminals by the large amount of NO presumably generated in the POA in response to intravenous LPS. The released NO, in high concentrations, also inactivates, in particular, iNOS, thereby reducing the activation of COX-2 (8, 9, 77). It is therefore possible that the inhibition of iNOS by the large amount of NO presumably generated in response to LPS and the consequent NE stimulation of α1-AR played a part in both the rise and fall of the early phase of fever in the present experiments. It could also account in part for the attenuation and amplification of the febrile and PGE2 rises observed when SIN-1 and carboxy-PTIO, respectively, were microdialyzed in the POA of the present guinea pigs.
Fourth, the present finding that NE, NO, and PGE2 are functionally interrelated in the hypothalamus is also not a new observation, but it too has not been considered previously in the context of fever production. Thus, as already reviewed, the sequence NE activation of α1-AR → upregulation of n- and iNOS → production of NO → activation of COX-1 and -2 → induction of PGE2 has been demonstrated in various cell types under both inflamed and noninflamed conditions (1, 8, 13, 21, 32, 44, 59, 77, 79). Given that the hypothalamic noradrenergic system is quickly activated by intravenous LPS (15), its activity is modulated by NO, and PGE2 is produced, the occurrence of a similar cascade in the POA could, a priori, also plausibly account for the initiation of the febrile response. Thus NE quickly released in the POA in response to intravenous LPS could initially activate nNOS, causing the local release of NO, which then, in turn, could activate constitutive COX-2 via α1-AR, resulting in the first rises of PGE2 and Tc. According to this scheme, the second PGE2 and febrile rises would develop independently of iNOS-derived NO, since its further production would presumably be inhibited by the large amount already present; the second phase, hence, should be elicited by the stimulation of α2-AR by NE and the consequent, direct activation of inducible COX-2 (15). The data depicted in Fig. 3, C–F, are consonant with this interpretation. There is, moreover, extensive evidence from numerous sources (for review, see Ref. 6) that the trigger that initiates endotoxic fever is PGE2 released in virtually instantaneous response to LPS by complement 5a-activated Kupffer cells, its signal being transmitted from the liver to the nucleus of the solitary tract by the vagus and thence via the ventral noradrenergic bundle to the POA, ending in the quick release of NE. Although the operation of this pathway has been questioned (63), recent reports that subdiaphragmatic vagotomy inhibits the systemic LPS-induced release of both NE and NO in the POA (29, 30, 83, 84) are consistent with both the present data and the existence of an afferent vagal pyrogenic signaling pathway.
Fifth, there is, however, an apparent inconsistency with the validity of the NE-NO-PGE2 cascade just asserted in that it contradicts our earlier demonstration that, although the first phase of LPS fever is indeed mediated by α1-AR stimulation, this stimulation, contrary to its effect in the PVN cited above, does not produce PGE2 in the POA albeit that preoptic PGE2 levels rise promptly following intravenous LPS (15, 39, 55). It is therefore unlikely that catechin simply inhibited the expression of COX-2 and, hence, the production of COX-2-dependent PGE2. Such a blockade has been shown in colonic mucosa, platelets, chondrocytes, and various cancer cells (2, 24, 28, 54, 72); however, relevant to the present data, the expression of COX-1 was not affected (28). More likely, because the intra-POA microdialysis of selective COX-1, -2, and -3 inhibitors did not prevent the initial increase in PGE2 (Fig. 3 and Ref. 15), it may be conjectured that this rise was induced by a COX-independent pathway. Based on the present, novel finding that the intra-POA microdialysis of the antioxidant (+)-catechin suppresses both the LPS-induced early and late rises of preoptic PGE2 and the second, but not the first, phase of fever, we suggest that it may be produced by the nonenzymatic isoprostane pathway of free radical-catalyzed peroxidation of arachidonic acid (4, 18, 51). Although no direct measurements of biomarkers of oxidative stress were made in the present study, this interpretation is consonant with the involvement of free radicals in the observed, initial PGE2 elevation because ROS have been demonstrated previously to facilitate the production of LPS-induced PGE2 in microglia (81). In the present case, these could be NO, peroxynitrite, and/or have been generated by the autooxidation of NE. That fever is prevented when the production of ROS is blocked has been reported previously (25, 60–62). The effect was ascribed then not to the inhibition by antioxidants of isoprostane production but to their reduction of the thiol groups attached to N-methyl-d-aspartate receptors, thereby depressing glutamate-mediated neuronal excitability and, hence, limiting fever (61, 62); the activation by peripheral pyrogens of glutaminergic pathways both in the nucleus of the solitary tract and the organum vasculosum laminae terminalis has been demonstrated (26, 27, 45). Further indirect support for the involvement of ROS in the initial PGE2 elevation comes from the present finding that the intra-POA microdialysis of catechin throughout the febrile course had the same effect as pretreatment with the α2-AR antagonist yohimbine in our companion study, i.e., suppression of both the LPS-induced early and late rises of preoptic PGE2 and of the second peak of fever (Fig. 2 and Ref. 15); the latency of fever onset and its first peak were not affected, with the fever continuing at its early phase high level until it abated normally. The entire course of fever was therefore presumably driven by PGE2-independent α1-AR activation alone. Indeed, in view of this, the intriguing possibility exists that, in contrast to PGE2 generated in the liver (39, 55, 75), PGE2 in the POA, however it may occur there, may not be essential to the production of fever. The independence of febrigenesis from central PGE2 has been reported previously in connection with certain pyrogenic cytokines, e.g., macrophage inflammatory protein-1β (38, 47), IL-8 (87), and preformed pyrogenic factor (86); the fever, in those cases, was attributed to the direct action of the cytokines in the POA.
In conclusion, NO, presumptively released contemporaneously with NE in the POA in quick response to peripheral LPS-initiated vagally transmitted signals of PGE2, is a central negative-regulatory, i.e., antipyretic, mediator, acting by modulating the intra-POA secretion of the thus released NE. In addition, ROS generated in the POA by the autooxidation of NE and/or by NO and its derivatives may induce isoprostanes that are quickly converted to PGE2. However, it does not appear that this preoptic PGE2 contributes to the initial Tc rise, since it develops and continues independently of its presence or absence.
This study was supported, in part, by National Institute of Neurological Disorders and Stroke Grants NS-34857 and NS-38594 (to C. M. Blatteis).
We thank Daniel Morse and Gregg Short for outstanding graphic arts support.
Present addresses: C. Feleder, Dept. of Basic and Pharmaceutical Sciences, Albany College of Pharmacy, Union University, 106 New Scotland Ave., Albany, NY 12208; and V. Perlik, Zentiva a.s., Machova 18, 120 00 Prague 2, Czech Republic.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society