Norepinephrine (NE) microdialyzed in the preoptic area (POA) raises core temperature (Tc) via 1) α1-adrenoceptors (AR), quickly and independently of POA PGE2, and 2) α2-AR, after a delay and PGE2 dependently. Since systemic lipopolysaccharide (LPS) activates the central noradrenergic system, we investigated whether preoptic NE mediates LPS fever. We injected LPS (2 μg/kg iv) in guinea pigs prepared with intra-POA microdialysis probes and determined POA cerebrospinal (CSF) NE levels. We similarly microdialyzed prazosin (α1 blocker, 1 μg/μl), yohimbine (α2 blocker, 1 μg/μl), SC-560 [cyclooxygenase (COX)-1 blocker, 5 μg/μl], acetaminophen (presumptive COX-1v blocker, 5 μg/μl), or MK-0663 (COX-2 blocker, 0.5 μg/μl) in other animals before intravenous LPS and measured CSF PGE2. All of the agents were perfused at 2 μg/min for 6 h. Tc was monitored constantly. POA NE peaked within 30 min after LPS and then returned to baseline over the next 90 min. Tc increased within 12 min to a first peak at ∼60 min and to a second at ∼150 min and then declined over the following 2.5 h. POA PGE2 followed a concurrent course. Prazosin pretreatment eliminated the first Tc rise but not the second; PGE2 rose normally. Yohimbine pretreatment did not affect the first Tc rise, which continued unchanged for 6 h; the second rise, however, was absent, and PGE2 levels did not increase. SC-560 and acetaminophen did not alter the LPS-induced PGE2 and Tc rises; MK-0663 prevented both the late PGE2 and Tc rises. These results confirm that POA NE is pivotal in the development of LPS fever.
- prostaglandin E2
- cyclooxygenase inhibitors
- noradrenergic receptor antagonists
- pyrogen signaling
- body temperature
abundant evidence has implicated PGE2, elevated by a systemic pyrogenic challenge in the preoptic area (POA) of the anterior hypothalamus (the fever-producing locus), as an essential mediator of the febrile response (for reviews, see Refs. 5 and 36). It is generally agreed that the increased production of this PGE2 is catalyzed by inducible cyclooxygenase (COX)-2. How its level comes to rise in the POA is, however, a debated issue. Various mechanisms have been suggested, including that PGE2 is produced and released peripherally by circulating (e.g., monocytes) or fixed (e.g., hepatic macrophages) mononuclear phagocytes, or, alternatively, by cerebral microvascular endothelial cells; these are presumed to be activated by the circulating exogenous pyrogen [e.g., lipopolysaccharide (LPS)] or by pyrogenic cytokines [e.g., interleukin (IL)-1β] induced by it. PGE2 is then postulated to pass into the brain by diffusion, by transport through the blood-brain barrier (BBB), or through the organum vasculosum laminae terminalis (OVLT), a circumventricular organ in which the BBB is deficient (for reviews, see Refs. 2, 7, 9, 57). These mechanisms, thus, are predicated on the humoral transport of PGE2 or its activating factor, i.e., LPS or IL-1β, to the brain.
A neural mechanism to account for the elevation of preoptic PGE2 under febrile conditions has also been proposed (60, 63; for review, see Ref. 10). It is based on evidence suggesting that the peripheral pyrogenic signal could be transmitted centripetally via the vagus nerves to the medulla oblongata, thence via the ventral noradrenergic bundle to the POA. This pathway is quicker than circulatory transport, and it is well documented that the systemic administration of exogenous and endogenous pyrogens provokes the prompt activation of noradrenergic terminals in the POA and the consequent release of norepinephrine (NE; 39, 42; for reviews, see Refs. 20 and 43). It is also well established that NE induces the synthesis of PGE2 in brain tissue in vitro and in vivo (23, 31; for reviews, see Refs. 5 and 45).
We reported recently that α-adrenoceptor (AR) agonists microdialyzed in the POA of conscious guinea pigs evoke two differentially modulated core temperature (Tc) rises, viz., one occurring very promptly and involving the activation of α1-AR but with no associated PGE2 release and the other occurring significantly later and involving the activation of α2-AR, the participation of, specifically, COX-2, and the consequent production of PGE2 (23). In guinea pigs, pyrogenic doses of intravenously injected LPS rapidly and characteristically evoke two successive Tc rises associated with concurrent PGE2 rises (64). Based on these findings, we investigated whether LPS fever could be mediated by preoptic NE and its two phases regulated in the same way.
MATERIALS AND METHODS
Male, pathogen-free, Hartley guinea pigs (301–350 g body wt on arrival; Charles River Laboratories, Wilmington, MA) were used in these experiments. The animals were quarantined for 1 wk, three to a cage, before any experimental use. Tap water and food (Agway Prolab guinea pig diet) were available ad libitum except during the experiments. The ambient temperature (Ta) in the animal room was 23 ± 1°C, the housing Ta recommended by the Institute of Laboratory Animal Resources Commission on Life Sciences (32); light and darkness alternated, with light on from 0600 to 1800. After quarantine, to moderate the psychological stress associated with the experiments, the animals were trained for 1 wk daily for 4 h to the experimental procedures by handling and placement in individual, locally fabricated, semicircular wire mesh confiners designed to prevent their turning around and to minimize their forward and backward movements, but without causing excessive restraint stress; rodents readily adapt to such confinement and show no signs of discomfort or anxiety (53, 59). 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 (1).
All glassware, plasticware, instruments, and cannulas used in these studies were sterilized by autoclaving. Electrochemical-grade, high-purity water (Baxter Healthcare, Muskegan, MI) was used exclusively in the preparation of all the solutions. Before use, the stock solutions were passed through a sterile 0.22-μm Miller-GS filter unit (Millipore, Bedford, MA) as an added precaution against bacterial contamination. Absence of endotoxic contamination in all fluids not containing LPS by design were verified by the Limulus amebocyte lysate test (Pyrochrome; Associates of Cape Cod, Falmouth, MA).
LPS was Salmonella enteritidis LPS B (batch no. 651628; Difco Laboratories, Detroit, MI), the same LPS batch we have used in all our previous studies; it was suspended in pyrogen-free saline (PFS, 0.9% NaCl, USP; Abbott Laboratories, Chicago, IL). Heparin was purchased from Elkins-Sinn (Cherry Hill, NJ) and dissolved in PFS. PFS was also the control drug for these solutions.
The microdialysis perfusate (and vehicle for all the drugs administered ic) was artificial cerebrospinal fluid for guinea pigs (aCSF; final concentration in mM: 140.0 NaCl, 2.7 KCl, 1.0 MgCl2·2H2O, 1.2 CaCl2·2H2O, 2.0 Na2HPO4; osmolality: 290 mosmol/kgH2O; pH 7.4, adjusted with 85% H3PO4); it was prepared freshly each day and prewarmed to ∼38°C for administration. Prazosin hydrochloride (catalog no. P7791, an α1-AR antagonist; 1 μg/μl of aCSF) and yohimbine hydrochloride (catalog no. Y3125, an α2-AR antagonist; 1 μg/μl of aCSF) were prepared just before use and stored in amber glass vials at room temperature. The selective COX-1 inhibitor SC-560 (catalog no. S2064; 5 μg/μl of aCSF), the selective COX-2 inhibitor MK-0663 (0.5 μg/μl of aCSF), the COX-1 splice variant COX-1v presumptively selective antagonist acetaminophen, USP (catalog no. A5000; 0.5 μg/μl of aCSF), and their solvent DMSO (6% in aCSF) were similarly prepared freshly for each use. These drug concentrations replicate those in our previous study (23). MK-0663 was generously donated by Merck (Rahway, NJ); DMSO (catalog no. 081-4) was purchased from Burdick and Jackson (Muskegon, MI). All the other drugs were purchased from Sigma-Aldrich (St. Louis, MO).
All animals received the antibiotic chloramphenicol (20 mg/kg body wt sc) 1 h before surgery and one time a day for the following 2 days. Immediately after surgery, the animals also received a subcutaneous bolus, 10 ml PFS injection and pain medication (butorphanol, 0.05 mg/kg body wt). All procedures were performed under ketamine-xylazine (35–5 mg/kg body wt im) anesthesia and aseptic conditions. Experiments were performed 7 days after the last of the following two surgical procedures, when the animals had recovered; 7 days also separated the two surgeries. Retraining was performed during the latter 4 days of this recovery period.
Jugular vein cannulation.
In preparation for intravenous injections, a siliconized cannula (ID 0.020 in., OD 0.037 in.; Baxter Healthcare, McGraw Park, IL), prefilled with heparinized (10 IU/ml) PFS, was inserted in the left jugular vein and gently guided in the superior vena cava of each guinea pig. The free end of the cannula was passed subcutaneously toward the head, exteriorized on the top of the head, knotted, and rolled into a coil. This coil was then placed inside a protective polypropylene shield (a centrifuge microtube with a screw-cap, with its cone cut off) that was fixed to the skull with dental acrylic cement and four self-tapping, miniature stainless steel screws. The neck wound was sutured and cleansed with 10% povidone-iodine solution and treated with nitrofurazone powder. To maintain the patency of the inserted cannulas, these were flushed with 0.5 ml of heparinized (3 IU/ml) PFS every day after surgery until 3 days before an experiment, when PFS alone was used because of the inhibitory effect of heparin on complement activation (72); complement component 5a is a critical mediator of the febrile response of guinea pigs and mice to LPS (8).
A sterile, 17-mm-long, 17-gauge, thin-walled stainless steel guide cannula with a tightly fitting indwelling stylet was implanted stereotaxically (Mechanical Developments, Trent H. Wells, Jr., South Gate, CA) in the left medial POA [anterior-posterior = 11.6 mm, lateral = 1.0 mm, ventral = −9.0 mm; according to the atlas of Luparello (44)] and fixed to the skull with four self-tapping, miniature, stainless steel screws and dental acrylic cement. Concentric microdialysis probes were constructed in our laboratory as previously described (23). Approximately 2 h before an experiment, the indwelling stylets of the guide cannulas were replaced by sterile microdialysis probes so that their dialysis membrane tips protruded exactly 1 mm beyond the guide cannulas. They were fixed to the skull with tissue adhesive and immediately perfused with sterile guinea pig aCSF via sterile 1-ml tuberculin syringes clamped to a syringe pump (model no. A-99; Razel Scientific Instruments, Stamford, CT) as the driver, for the duration of an experiment. To run six animals simultaneously, the pushers of two pumps were modified so that each could accommodate three syringes at a time. During the first 90 min after insertion of the probes [thermal stabilization period (see below)], the flow rate of the perfusion was adjusted to 4, 3, and 2 μl/min at 30-min intervals; it was maintained at the latter rate for a second 90-min period (the pretreatment control period) and also for the following experimental period. The effluents from the microdialysis probes were collected in vials chilled on ice over 30-min intervals continuously throughout the experiments from 90 min before to up to 360 min after a treatment (time 0 = initiation of treatment) and analyzed later for their NE or PGE2 content.
After an experiment, the guinea pigs were killed by isoflurane overdose, and the brains were quickly removed and stored in 10% phosphate-buffered Formalin for later histological verification of the placement of the dialysis probe tips. Localization of the center of the dialysis probe within 0.5 mm of the medial POA was regarded as the correct placement. Only the data from confirmed preoptic placements of the probes are included in this report.
To obviate possible effects of circadian variations, all of the experiments were begun at the same time of day (0830). A 90-min thermal stabilization period preceded all the experimental treatments. After their last surgery (7 days), the guinea pigs, fully conscious, were loosely restrained in the individual wire mesh confiners to which they had been trained at Ta 23 ± 1°C. The Tc values of the animals were monitored constantly from the beginning of the stabilization period and recorded at 2-min intervals for the duration of an experiment on a Macintosh Plus 1-Mb microcomputer through an analog-to-digital converter, using precalibrated copper-constantan thermocouples inserted 5 cm in the colon. The data were displayed digitally on a monitor, printed on an ImageWriter printer, and stored on a diskette for later analysis.
The NE content of the samples was evaluated using HPLC with EC detection. Microdialysate effluents were collected in 1 μl of 5% perchloric acid on ice and stored at −20°C until analysis. For analysis, samples (10 μl) were injected using a CMA 200 refrigerated automatic sampler (CMA Microdialysis, North Chelmsford, MA) on a 150 × 3-mm ODS C18
The PGE2 content of the microdialysate samples was analyzed using a commercial PGE2 enzyme immunoassay kit (High Sensitivity PGE2 EIA Kit no. 931-001; Assay Designs, Ann Arbor, MI). The PG synthetase inhibitor indomethacin (10 μg/ml) was added to all the blood samples immediately after collection. All samples were diluted before analysis in the assay buffer system provided by the manufacturer, according to the manufacturer's instructions. All samples were analyzed in duplicates. The detection limit of this assay was 8.26 pg/ml of sample. The data were expressed as changes (ΔPGE2) relative to their values over the 90 min before a treatment (P0).
Experiment 1: Effect of LPS injected intravenously on the level of NE in the POA.
To verify in our model that NE is released in the POA in virtually immediate response to the peripheral administration of LPS, collection of microdialysate effluents was begun 60 min before, immediately after the 90-min stabilization period, and continued, in this instance, for 300 min after (time 0) the intravenous injection of PFS (0.6 ml/kg) or LPS (2 μg/kg in 0.6 ml of PFS). The guinea pigs were conscious throughout this experiment. The samples were analyzed for their NE content as described above.
Experiment 2: Effects of α-AR antagonists microdialyzed in the POA on the febrile and preoptic PGE2 responses to LPS intravenously.
To determine whether NE presumptively released promptly in the POA after the intravenous administration of LPS mediates the ensuing febrile and PGE2 responses, prazosin or yohimbine was microdialyzed (each at 1 μg/μl of aCSF) immediately following the 90-min stabilization period in the POA of conscious guinea pigs for 1 h, beginning 30 min before the intravenous injection of PFS or LPS. This design was chosen to replicate the conditions of our previous study (23). Tc values and preoptic PGE2 levels were measured from 90 min before until 360 min after PFS or LPS injection, as described earlier.
Experiment 3: Effects of COX inhibitors microdialyzed in the POA on the febrile and preoptic PGE2 responses to LPS intravenously.
To identify the COX isozyme responsible for the rise in preoptic PGE2 during the two characteristic rising phases of the febrile response of guinea pigs to intravenous LPS, the selective COX-1 inhibitor SC-560 (5 μg/μl), the selective COX-1v inhibitor acetaminophen [0.5 μg/μl; COX-1v (putative COX-3) has been implicated as a putative central fever mediator (11, 12, 16)], or the selective COX-2 inhibitor MK-0663 (0.5 μg/μl) was microdialyzed immediately following the 90-min stabilization period in the POA of conscious guinea pigs and continued for the duration of the experiment; the solvent of these inhibitors was 6% DMSO in aCSF. PFS or LPS was injected intravenously at time 0. Tc values and preoptic PGE2 levels were determined as described previously.
The results are presented as means ± SE. The values of Tc are reported as changes (ΔTc, in oC) from basal values [initial Tc (Tci), the Tc at 2-min intervals averaged over the last 10 min of the preceding 90-min stabilization period] plotted at 6-min intervals. Latencies of fever onset were defined as the intervals (in minutes) between the time of LPS injection (0 min) and that of the first Tc rise >0.2°C (the SD of the mean Tci) that continued uninterruptedly beyond 0.5°C. The NE data are expressed as their absolute values (in pg/ml) and the PGE2 data as changes (ΔPGE2, in pg/ml) relative to the mean ± SD of their three values before a treatment. A repeated-measures ANOVA was used to compare all changes between groups; factor 1 was the between-groups factor (the treatment) and factor 2 the within-subjects factor (the different sampling periods), followed, if significant differences were found, by a point-by-point Tukey-Kramer multiple-comparison test. The analyses were performed using Instat 3 (Graph Pad Software; Instant Biostatistics, San Diego, CA). Each variable was considered to be independent. The 5% level of probability was accepted as statistically significant.
The intravenous injection of LPS produced a tripling of the basal level of NE in the POA interstitial fluid of the conscious guinea pigs, measured 30 min after this treatment (Fig. 1). NE then gradually returned toward its initial value over the following 2 h. The level of preoptic NE was not affected by the intravenous injection of PFS.
Intravenous LPS produced its characteristic, prompt, biphasic fever; the onset latency was 11 ± 0.4 min (mean ± SD; Fig. 2A). The first febrile peak of ∼1.4°C occurred at ca. 60 min after LPS administration, and the second of ∼1.7°C ca. 90 min after the first. The return to Tci was gradual, but essentially completed by ca. 300 min after LPS injection. Preoptic PGE2 levels increased significantly in these conscious guinea pigs in close temporal correspondence with their Tc rise (Fig. 2B). Intravenous PFS had no effect on either of these variables.
The intra-POA microdialysis of prazosin significantly delayed the onset and slowed the rate of development of the first Tc rise; the onset latency was 19 ± 0.5 min (P < 0.05). It also greatly attenuated the height of the first and significantly reduced the maximum of the second febrile rises. Thus the first Tc rise increased only ∼0.5°C, stabilized for ca. 30 min, and then resumed its rise to a second peak of ∼1.1°C at ca. 150 min (Fig. 2C). Tc then gradually decreased. Prazosin microdialyzed in the POA, however, did not block the initial LPS-induced preoptic PGE2 rise, which continued to its peak at 180 min; but it depressed its level at 60 min and augmented it at 180 min post-LPS (Fig. 2D) by comparison with its contour in the LPS only-treated counterparts (Fig. 2B).
Intra-POA-microdialyzed yohimbine did not affect the slope and height of the first febrile rise, but it suppressed the second rise so that the fever continued at its initial, lower, first-phase level uninterruptedly through the end of the experiment (Fig. 2E). However, strikingly, it totally suppressed the normally associated first and second POA PGE2 rises (Fig. 2F).
Neither intra-POA prazosin nor yohimbine demonstrably affected the Tc and preoptic PGE2 levels of the PFS intravenously treated controls (Fig. 2, A–F).
Neither the microdialysis of SC-560 nor of acetaminophen altered the prototypical, temporally correlated effects of intravenous LPS on Tc and POA PGE2 levels (Fig. 3, A–F). MK-0663, on the other hand, blocked the second rise of Tc normally evoked by intravenous LPS; it had, however, no effect on the early response of this variable (Fig. 3G). However, it depressed the early LPS-induced rise of preoptic PGE2 and reversed its late rise (Fig. 3H). Neither PFS injected intravenously nor aCSF or DMSO, the vehicle of these COX inhibitors, microdialyzed intra-POA per se over the same duration, affected the Tc or the preoptic PGE2 levels of the conscious guinea pigs (data not shown).
The present results verify that the biphasic fever characteristically evoked in conscious guinea pigs by the intravenous injection of a moderate, subseptic dose of LPS is attended by coincident increases in preoptic PGE2 levels, as reported previously (64). They further show that this challenge rapidly provokes the release of NE in the POA of these animals and, hence, would support the notion that preoptic NE may play a pivotal role in febrigenesis. Thus α1-AR antagonism by intra-POA prazosin greatly slowed the development and reduced the height of both the early and late phases of the febrile response to intravenous LPS, but without reducing the normally associated preoptic PGE2 rises. α2-AR antagonism by yohimbine, on the other hand, affected neither the development nor the height of the first febrile rise, but it prevented both the development of the second Tc rise and the preoptic PGE2 rises associated with both intravenous LPS-induced febrile phases. Consequently, remarkably, fever was initiated and maintained under the latter conditions at its original height throughout the experiment in the total absence of a corresponding increase in the level of POA PGE2. These findings would infer, therefore, that intra-POA PGE2 is not essential for the manifestation of the second phase of intravenous LPS fever. The present results indicate, moreover, that, such as it exists, only the rise in preoptic PGE2 that is associated with the second, but not with the first, phase of LPS fever is COX-2 dependent. This is also a novel finding, at variance with the conventional view (for reviews, see Refs. 5 and 36). Neither COX-1 nor COX-1v appeared to influence Tc or the production of POA PGE2 during either phase.
The rapid elevation in the concentration of NE in the preoptic dialysate observed in this study following the intravenous injection of LPS and its temporal association with the early phase of fever are in agreement with previous findings in rats and mice treated with intraperitoneal LPS (18, 39, 42, 43, 73, 74). Indeed, it is well established that the peripheral administration of exogenous and endogenous pyrogens stimulates the increased metabolism of NE, especially in the hypothalamus (for reviews, see Refs. 20 and 43). It is also generally accepted that, notwithstanding certain species and other experimental variations regarding the up or down direction of its thermal effect, the central noradrenergic system is involved in thermoregulation (for review, see Ref. 75). In guinea pigs, electrical stimulation of the ascending noradrenergic system in the brain stem (67, 68) or NE microinjected in the POA (55) evokes Tc rises. Recently, we (23) reported that NE microdialyzed in the POA of conscious guinea pigs mediates, in fact, not one but two distinct Tc rises, each associated with the activation of a different α-AR class. Thus one rise is rapid in onset and α1-AR-mediated, and the other is delayed, appearing significantly later than the former, and is α2-AR mediated; the specific involvement of both AR classes was verified by the blockade of their thermal effect by their respective, selective antagonists. We also found in the same study that the early Tc rise provoked by intra-POA microdialyzed NE occurred without any demonstrable change in the level of PGE2 in the POA, whereas the later rise was associated with a concurrent elevation of this level. Again, the noninvolvement of PGE2 in the α1-AR-induced Tc rise and, in contrast, the involvement of COX-2-dependent PGE2 in that caused by α2-AR activation were confirmed by their respective responses to selective COX-1 and COX-2 inhibitors (23). Also, in a parallel study (6), we found that the thermal responses of conscious wild-type Cox-1 and Cox-2 gene-deleted mice to the intracerebroventricular microinjection of α1- and α2-AR agonists corresponded exactly with those observed in these guinea pigs. In the present study, the biphasic febrile response of the conscious guinea pigs to intravenous LPS was associated with NE released in the POA and evidently developed in the same order as the two successive Tc rises induced by intra-POA microdialyzed NE per se, that is, the present first rise of Tc was induced by α1-AR stimulation and was not associated with PGE2 formation, whereas the second elevation of Tc was mediated by α2-AR stimulation and associated with the delayed production of COX-2-dependent PGE2. We suggest, therefore, that noradrenergic terminals in the POA, through their NE-induced, presumptively contemporaneous activation of both α-ARs, mediated the observed febrile response to LPS.
The present finding that prazosin attenuated the initial febrile rise, i.e., that α1-AR stimulation mediated it without the intermediation of PGE2, suggests that the α1-ARs activated by NE are located on postsynaptic warm-sensitive or thermoinsensitive neurons in the POA and that NE directly reduces or augments, respectively, the activities of these neurons. According to the classical model of Hammel (29), both responses promote heat conservation. In support, the direct inhibition and excitation by cirazoline, an α1-AR agonist, of the firing rates of, respectively, warm-sensitive and thermoinsensitive neurons have recently been reported in rat POA slice preparations (33). Because these neurons, moreover, are thought to inhibit synaptically connected cold-sensitive neurons, these are concomitantly facilitated, stimulating heat production, i.e., in combination, these effector mechanisms raise Tc. The specific α1-AR subtype involved in this hyperthermic effect remains to be identified.
The findings that the second febrile rise and the associated increase of preoptic PGE2 were both inhibited by yohimbine and the selective COX-2 inhibitor MK-0663 indicate that these subsequent, characteristic features of the febrile response to intravenous LPS were specifically mediated by α2-AR-activated COX-2-dependent PGE2, in agreement with the demonstrated actions of intra-POA microdialyzed clonidine, an α2-AR agonist, and NE per se (23). It is significant in this regard that, under the present experimental conditions, the second peak of fever was ∼0.5°C higher than the first and that this was also the extent of the fever attenuated by yohimbine (Figs. 2, A and E). The brain cell type expressing COX-2 in response to NE in these guinea pigs remains to be determined. Because in rat brain COX-2 mRNA becomes detectable in astrocytes, microglia, perivascular cells, and cerebromicrovascular endothelial cells, but only irregularly in neurons, ca. 1 h after the intravenous administration of LPS and pyrogenic cytokines (13, 15, 22, 70), hence after the onset of fever (latency ca. 10 min), we conjecture that the increased PGE2 in the POA of the present guinea pigs was generated by astrocytic processes contacting noradrenergic synaptic regions. The delay imposed by its synthetic process probably accounts for the interval between α2-AR stimulation and the appearance of PGE2. The identity of the α2-AR subtype involved in this effect also remains to be elucidated. The subsequent effect of the thus released PGE2 on the activities of POA warm-sensitive and thermoinsensitive neurons is presumptively similar to that indicated earlier for α1-AR-mediated responses. The PGE2-sensitive receptor involved in these neuronal effects is probably the EP3 subtype; it has been linked to the development of fever and is present in the POA (21, 69; for review, see Ref. 52). The receptor implicated in the inhibition of presynaptic NE release by PGE2 has also been previously identified as the EP3 subtype (62).
The notion that NE and PGE2 may interact in the POA in fever production is not new, albeit that the mechanism of their cooperation in this function is in dispute. Thus it has been suggested that the hyperthermic action of NE may be exerted through PGE2 because the Tc rise induced in cats by intracerebroventricularly microinjected NE is inhibited by pretreatment with aspirin (50). A similar involvement of NE was described originally for the PGE2-mediated release of luteinizing hormone-releasing hormone (51). However, the converse has also been suggested, that PGE2 may be thermogenic through NE because the destruction by 6-hydroxydopamine of the noradrenergic nerve terminals in the POA of rabbits attenuated the rise in Tc produced by intravenous LPS (38, 41). On the other hand, no mutual interaction between NE and PGE2 was found in monkeys (49). The present data would indicate that, in guinea pigs, PGE2 is the thermogenic agent during the second phase of intravenous LPS fever, produced consequent to the activation of α2-ARs by NE released in the POA. The in vitro release of PGE2 by NE-stimulated brain tissue and the rapidly increased production of PGE2 by the microdialysis of NE in the POA of conscious guinea pigs have been demonstrated previously (31, 65). It is, in fact, well documented both in the peripheral and central nervous systems that the stimulation of noradrenergic neurons induces the postsynaptic release of PGE2, which then limits the further presynaptic release of NE, thereby modulating the activity of noradrenergic neurons (30, 45).
As already noted, the increase in preoptic NE that attended the febrile response of the present guinea pigs to intravenous LPS and, previously, also that of other species to intraperitoneal LPS and IL-1β was coincident with its early rather than its late phase. This would suggest that the fever-triggering message evoked in the periphery that signals its release in the POA was very quickly conveyed centripetally. In view of the rapidity of this transmission, we deduce that it was conveyed neurally. Indeed, ample data have implicated the vagus, especially its hepatic branch afferents, as the carrier of peripheral pyrogenic signals to the brain stem (60, 63; for reviews, see Refs. 19 and 61) and identified PGE2 released by Kupffer cells stimulated by LPS-activated complement component 5a as this fever signal (40, 54; for reviews, see Refs. 7, 8, 36, 57). Other data have corroborated that signals can proceed from the brain stem to the POA by way of noradrenergic projections originating in the A1 and A2 regions of the medulla oblongata (25) and arriving in the POA via the ventral noradrenergic bundle (27, 29, 71). Furthermore, subdiaphragmatic vagotomy blocks intraperitoneal IL-1β-induced hypothalamic NE activation in rats (26, 34, 73) and mice (74).
From these and other well-substantiated findings that the rises in Tc and preoptic PGE2 levels induced by peripherally administered pyrogens are prevented by the intra-POA administration of low doses of COX-2 inhibitors, it would seem highly likely that the PGE2 collected in the POA in the present experiments was generated inside rather than outside the BBB. Indeed, the capacity of brain tissue to generate PGE2 is well established (for reviews, see Refs. 5 and 66). Moreover, PGE2 infused in the internal carotid artery of conscious guinea pigs could not be detected in the POA and caused a fall rather than a rise of Tc (64). It follows from these results, therefore, that, since its formation was consequent to its induction by locally released NE, the signal that evoked it was, in the first instance, the very same as that which initially evoked the release of NE in the POA, that is, it was a nervous signal transmitted from the liver via the vagus. This contradicts an alternative view, that the fever-producing PGE2 that acts in the POA originates in the periphery and is transported to it by the bloodstream; it, being lipophilic, then either diffuses across the BBB (47, 58) or enters the POA through the leaky OVLT (56). However, direct evidence that blood-borne PGE2 passes in the brain and, especially, that PGE2 derived from the blood in this way raises Tc is controversial (48, 64). Indeed, because PGE2 is generally considered to be a paracrine rather than an endocrine mediator and, as an organic anion at physiological pH, enters cells poorly by simple diffusion, it would seem improbable a priori that the very rapid Tc rise in response to intravenous LPS should depend on its blood-borne transport from peripheral sources to the POA and ready passage across the BBB. In support, the in vivo expression in rat hypothalami of prostaglandin transporter, a principal carrier of PGs from the blood in cerebral endothelial cells, is not affected by intraperitoneal LPS challenge (37). Moreover, PGE2-inactivating enzymes are scarce in the hypothalamus and not upregulated in response to peripheral LPS (35; for review, see Ref. 36) so that, to mitigate its central effects, PGE2 is usually transported from the brain into the blood (4, 17). An alternative pathway of signal transduction from the periphery to the brain, that the fever-mediating PGE2 is produced by the cerebral endothelial cells themselves (for reviews, see Refs. 46 and 57), although based on well-substantiated evidence that circulating LPS and pyrogenic cytokines induced by it upregulate the expression of COX-2 in cerebral endothelial cells (14, 15), is attended by the difficulty that the hypothalamic expression of multidrug resistance-associated protein 4, which actively transports PGs out of cells, also is not affected by LPS (37).
In view of the delay imposed by its synthetic process, the prompt elevation of PGE2 in the POA during the first phase of fever can therefore not be attributed to its COX-2-mediated formation. The possibility that COX-1 or COX-1v mediated this PGE2 production is negated by the present findings that neither SC-560 nor acetaminophen microdialyzed in the POA prevented the development of fever. The noninvolvement of COX-1 and COX-1v in fever production has also been shown by other workers (for review, see Ref. 7). The involvement of COX-2-dependent PGE2 is similarly excluded since the intra-POA microdialysis of MK-0663 did not inhibit the first febrile rise and since, moreover, it is in any case, as demonstrated herein, released in the POA in correlation with the second phase of fever. It is possible, therefore, that the early increase in preoptic PGE2 was mediated by a COX-independent mechanism, possibly by the nonenzymatic isoprostane pathway of free radical-catalyzed peroxidation of arachidonic acid to 8-iso-PGE2 (for review, see Ref. 3). The free radicals in this case could be generated by the auto-oxidation of NE and/or be nitric oxide; the latter is also released locally in the POA after LPS administration (24). The finding that this PGE2, however, was not integral to the initiation of the febrile response nor, indeed, to its continuation is novel and intriguing as regards LPS-induced fever, although other fevers caused by certain cytokines have been reported previously to occur independently of central PGE2 (see Ref. 24).
In conclusion, these results are consistent with the view that NE released in the POA in response to the vagally conveyed pyrogenic message of PGE2 generated by complement component 5a-stimulated Kupffer cells mediates the characteristic biphasic Tc rises evoked in conscious guinea pigs by the present moderate dose (2 μg/kg body wt) of intravenous LPS. The following two, successive mechanisms of action are hypothesized: 1) rapid induction of the first Tc rise by activation of α1-ARs without the mediation of PGE2 and 2) subsequent development of the second Tc rise by simultaneous stimulation of α2-ARs, consequently activating (after the delay imposed by the de novo synthesis of the relevant enzymes) the production and release of COX-2/mPGES-1-dependent PGE2 in the POA. It should be emphasized that the present conclusions concern specifically the initiating, not the sustaining, sequence of the febrile response of guinea pigs to intravenous LPS. Thus they do not contradict the participation in the late phase of this response of any of the various humoral pyrogen signaling pathways that have been proposed for this and other species.
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, Department 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