INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

FEVER is the most common manifestation of disease, and it is now recognized that it is an important component of the host defense response. There is good evidence that animals in which the febrile response is experimentally suppressed show increased morbidity and mortality to infection (15, 28). Given this fact, it is surprising that there are certain times when the ability to develop a fever is reduced or absent; this phenomenon has been termed "endogenous antipyresis" (33). For example, suppression of fever caused by peripheral pyrogens such as lipopolysaccharide, interleukin-1, and centrally administered prostaglandin E (PGE) in pregnant animals, including rats, has been reported by several authors (25, 26, 38). With regard to fevers induced by peripheral pyrogens, the phenomenon is most evident when pregnancy is close to term and is less profound after the delivery. Neither the impact of this reduced febrile responsiveness on the maternal-infant unit nor the mechanism by which fever responses are reduced is known. The purpose of this research is to determine a possible mechanism for the reduced febrile responses. The most likely possibilities include either reduced capability to mount a thermogenic response to pyrogens or an active suppression of thermogenic and heat conservation pathways by one of the central neurotransmitters responsible for bringing about defervescence. Several authors have suggested that the antipyretic peptide arginine vasopressin (AVP) may be responsible for the increased endogenous antipyretic activity (5, 36). In favor of the latter possibility, there is increased immunoreactivity for AVP during pregnancy in the parvocellular paraventricular nucleus and terminal areas in the lateral septum and the amygdala (27). Furthermore, basal release of AVP is greater in the septal area in nonfebrile pregnant and postparturient rats compared with virgin females (22). These findings provide correlative evidence suggesting that endogenous antipyresis is responsible for fever suppression in term pregnancy, inasmuch as the septal area is one of the sites that is sensitive to AVP-mediated antipyresis (30).

Nonetheless, other mechanisms could also account for fever suppression at term. One possibility is that the central nervous system response to pyrogens, or the neural output to generate a febrile response, is attenuated at term. Prostaglandin E₂ (PGE₂), a naturally occurring mediator of fever within the brain, provides a useful tool to ask if the neural response to pyrogens is altered at near term. Central PGE₂ is found to activate a number of central autonomic areas (18, 37), including loci where it increases the frequency of sympathetic nerve activity (39), elevates plasma catecholamines (8), induces pressor responses and tachycardia (12), and, most importantly for fever generation, increases the sympathetic nerve activity to the interscapular brown adipose tissue (iBAT) (2), an important site for nonshivering thermogenesis (11). This suggests that PGE₂-induced sympathetic activation of thermogenesis pathways could be less prominent at full-term pregnancy and may therefore contribute to the diminished febrile responses. Given that fever is thought to have survival value (15), it is important to determine how neurally evoked fever is suppressed at near term.

This work used central infusion of PGE₂ to induce fever in conscious and urethan-anesthetized rats to answer the following questions: 1) is PGE₂-induced fever specifically suppressed at near term compared with earlier pregnancy and postparturition? 2) Is sympathetically driven thermogenesis (iBAT temperature and sympathetic activity to the iBAT) reduced at near term? 3) Is there an alteration in AVP release in the ventral septal area (VSA) or in its action at near term? 4) Is fever reduction at near term associated with changes in cardiovascular responses (as an indicator of general sympathetic tone) to PGE₂?

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Time-bred Sprague-Dawley rats were purchased from the University of Calgary vivarium and were housed singly in temperature-controlled quarters under a normal 12:12-h light-dark cycle (lights on 0700). Pellet chow and water were accessible ad libitum. All experimental protocols were approved by the University of Calgary Animal Care Committee and were carried out in accordance with the Canadian Council of Animal Care guidelines.

iBAT temperature and nerve activity measurements. At stages of pregnancy days 16-17 (P16-17) and 19-20 (P19-20 or near term, term = 21 days) and days 1-2 postpartum (L1-2), rats were anesthetized with urethan (1.4 g/kg ip). Body temperature was monitored by a rectal thermistor and was allowed to stabilize at around 36°C with a heating pad under the animal (23, 24). Although the anesthetized rat displays slightly reduced thermogenic responses compared with the conscious, unrestrained rat, the responses appear qualitatively identical. In particular, such animals are capable of developing fevers to all pyrogens tested to date. Furthermore, use of this preparation allows distinction between behavioral responses (which are often significant in females at parturition) and physiological responses to the pyrogen. A guide cannula (23 gauge thin wall) was implanted over the lateral ventricle for intracerebroventricular infusion. An incision was made on the back, the iBAT was exposed, and a needle thermistor (YSI 524) was inserted into the iBAT for temperature recording. Records of both rectal and iBAT temperature were made on a Gould chart recorder. One of the parallel nerves to the iBAT was isolated, cut distally as it enters the iBAT, and placed on a bipolar electrode. The preparation was bathed in prewarmed mineral oil. Activity of the isolated nerve was preamplified, the signal was then amplified and fed to a spike processor where spikes were counted over 5-s bins, and the output was passed to a Gould recorder. After all parameters had stabilized for a minimum of 30 min, PGE₂ (50 ng/5 µl sterile saline; Sigma) or a V₁ AVP receptor antagonist, [d(CH₂)¹₅-O-Me-Tyr²-Arg⁸]-vasopressin (1 nmol; Manning compound; Bachem, Torrence, CA) plus PGE₂ (50 ng) was infused into the lateral ventricle, and measurements were taken for at least another hour.

Cardiovascular measurements and microdialysis of the VSA. Pregnant and postpartum animals were anesthetized, a lateral ventricle was cannulated, and rectal temperature was monitored as described in iBAT temperature and nerve activity measurements. A CMA/12 microdialysis probe (length of microdialysis membrane, 2 mm; 100-kDa cutoff; in vitro AVP recovery rate, 6-7%; CMA/Microdialysis, Stockholm, Sweden) was directed to the VSA contralateral to the ventricular cannula at the following coordinates: anterior-posterior +0.4; lateral +0.6; and dorsal 7.2 mm from the dura (32). Artificial cerebrospinal fluid (CSF) containing (in mM) 126 Na⁺, 2.5 K⁺, 1.3 Ca²⁺, 1.0 Mg²⁺, and 135 Cl (pH 7.4) was passed through the probe at the rate of 2 µl/min. A femoral artery catheter filled with sterile, heparinized saline was connected to a pressure transducer and a Gould chart recorder for heart rate and blood pressure recording. Rectal temperature information was also registered on a channel of the chart recorder. Stable baseline values for temperature, heart rate, and blood pressure were recorded for a minimum of 60 min. During baseline recording, 2 microdialysis samples of 30 min each were collected (first sample discarded), after which 50 ng PGE₂/5 µl sterile, pyrogen-free saline was infused into the lateral ventricle, and four further 30-min microdialysis samples were collected. Collected samples were frozen at 20°C and lyophilized until assay. On completion of the experiment, rats were perfused via the heart, and brain sections of 60 µm were stained with neutral red for histological confirmation of the site of microdialysis. AVP levels were determined in lyophilized microdialysates without extraction before RIA (19, 21). Media containing known levels of AVP were used to calibrate the assay. All samples were processed in a single assay with an intra-assay variation of 7%. The sensitivity of this assay was 0.1 pg at 95% binding (50% inhibition at ~1.5 pg), and cross reactivity with oxytocin and structurally related peptides was <0.7%.

PGE₂ fever and AVP antagonism in conscious pregnant rats. Pregnant rats at day 14 were anesthetized with pentobarbital sodium (50 mg/kg ip). Under aseptic conditions, a precalibrated, battery-driven temperature transmitter (Mini-Mitter, Sun River, OR) was inserted into the abdominal cavity, and a guide cannula (23 gauge thin wall) was implanted above the upper floor of the lateral ventricle. Rats were transferred to an environmentally isolated and temperature-controlled (22°C) testing room at day 19 of pregnancy and were infused on the 20th day of pregnancy with either 25 ng PGE₂ or 1 nmol Manning compound plus 25 ng PGE₂ in 5 µl of sterile, pyrogen-free saline. Core temperatures were monitored with a telemetry system (DataQuest II; Data Sciences, St. Paul, MN) that automatically takes a reading every 10 min for 2 h after PGE₂ infusion.

Data analysis. All data are presented as means ± SE. Original temperature (except for conscious animals testing, which was at 10-min intervals), heart rate and blood pressure, and BAT nerve activity recordings were taken at 5-min intervals. Temperature, heart rate, and blood pressure data are presented as net deviation from the mean of baseline points; unless specified, all baseline data were statistically identical between groups. Mean arterial pressure (MAP) was calculated as diastolic and one third of the pulse pressure. iBAT nerve activity is presented as fold increase from the baseline because activity levels varied considerably from nerve to nerve. Data were analyzed by one-way ANOVA of values over time, with Student-Newman-Keuls pairwise comparisons (between groups) or Dunnett's test (vs. controls), or by two-way ANOVA with gestational groups and treatments as factors. Alternately, and only when specifically indicated in the results, areas under the curve were calculated in arbitrary units, and values were compared by ANOVA followed by Student-Newman-Keuls post hoc. Significance was accepted at the 0.05 level.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

PGE₂-induced rectal and iBAT temperature responses are reduced at near term. Central PGE₂ infusion caused an elevation in rectal and iBAT temperatures with iBAT response having an earlier onset and peak (20 min) and a bigger magnitude than the change in rectal temperature (Fig. 1A). Increase in rectal temperature after PGE₂ infusion, which was present in all groups (F = 7-171, P < 0.05; ANOVA), peaked at about 30 min and decreased gradually thereafter. Whereas baseline temperatures were similar at all times (P16-17, 36.1 ± 0.2; P19-20, 36.3 ± 0.3; L1-2, 36.0 ± 0.2°C; P > 0.05), the febrile response was greatly reduced near term (P19-20, P < 0.05; Student-Newman-Keuls post hoc comparisons) and recovered after delivery (Fig. 1B).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Rectal and interscapular brown adipose tissue (iBAT) temperature (T) responses to central prostaglandin E2 (PGE2; 50 ng icv) recorded at days 16-17 of pregnancy (P16-17), near term at days 19-20 of pregnancy (P19-20), and days 1-2 postpartum (L1-2). Time 0 and arrows indicate PGE2 infusion. A: representative chart recorder traces of rectal and iBAT temperature responses to PGE2 in 3 different animals. B: net changes () in rectal temperature in response to PGE2. C: net changes in iBAT temperature in response to PGE2. Data are means ± SE; n = 11-14 for each group.

Baseline temperatures for iBAT were similar at all times (P16-17, 36.2 ± 0.2; P19-20, 36.5 ± 0.3; L1-2, 36.3 ± 0.2°C; P > 0.05). PGE₂ also induced an iBAT temperature response at P16-17 and after delivery (F = 26-248, P < 0.05, ANOVA; Fig. 1C), whereas at near term iBAT temperature did not increase in response to PGE₂ (F = 1.8, P > 0.05, ANOVA). Thus at near term, iBAT temperature change was significantly less than that at P16-17 and post partum (P < 0.05, Student-Newman-Keuls post hoc comparison; Fig. 1C).

iBAT nerve activity is reduced at near term. We wished to determine if the reduced iBAT temperature response to PGE₂ resulted from reduced peripheral responsiveness or reduced neural activity in iBAT nerves. Because PGE₂ fever is mostly caused by nonshivering thermogenesis in the iBAT (2, 9), we asked if the specific sympathetic activity to the iBAT was changed at near term. PGE₂ induced a fast onset increase in iBAT nerve activity in all groups (F = 7.8-12.3, 1-way ANOVA; Fig. 2, A and B) whose time course correlated well with that of iBAT temperature. At near term, PGE₂-induced increase in iBAT nerve activity was also reduced when compared with P16-17 (F = 4.4, P < 0.05, 1-way ANOVA with Student-Newman-Keuls post hoc); the iBAT nerve activity in response to PGE₂ partially recovered after delivery (Fig. 2B).

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Sympathetic nerve activity to iBAT in response to PGE2 at P16-17, P19-20, and L1-2. A: representative chart recorder traces of iBAT nerve activity in response to PGE2 in 3 different rats at time indicated by arrows. B: fold increase in iBAT nerve activity in response to PGE2 at time 0. Data are means ± SE; n = 5-6 for each group.

PGE₂-induced cardiovascular responses are reduced at near term. The following experiment was carried out to test whether there is a decrease in other sympathetic responses to PGE₂ associated with pregnancy. PGE₂ induced a fast increase in heart rate and arterial blood pressure similar in time course to iBAT temperature responses (Fig. 3A). Central PGE₂ induced an immediate increase in heart rate that peaked at 10 min and lasted about 30 min in all groups (F = 9-18, P < 0.01, ANOVA; Fig. 3B), but the increase was smaller at near term when compared with the other two groups (P < 0.05, Student-Newman-Keuls post hoc). Nonetheless, baseline heart rates were indistinguishable at all ages (P16-17, 297 ± 15; P19-20, 311 ± 9; L1-2, 307 ± 16 beats/min; P > 0.05).

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Cardiovascular responses to central PGE2 at P16-17, P19-20, and L1-2. A: chart recorder traces showing heart rate (HR) and arterial pressure (AP) responses to PGE2 given at time indicated by arrow. B: net changes in HR in response to PGE2 at time 0. C: net changes in mean AP (MAP) in response to PGE2 at time 0. Data are means ± SE; n = 7-8 for each group.

Baseline MAP tended to decrease with the time of pregnancy and was significantly lower in postpartum rats than in pregnant groups (P16-17, 84.3 ± 2.4; P19-20, 80.5 ± 1.9; L1-2, 70.5 ± 3.5 mmHg; F = 57, P < 0.01, ANOVA). The pressor effects of PGE₂ were present in all groups and followed a similar time course to that of heart rate and also preceded the rise in rectal temperature (F = 6-80, P < 0.01, ANOVA; Fig. 3C). The overall increase in blood pressure was reduced at near term compared with that in the P16-17 group (F = 6, P < 0.05, ANOVA), and it did not recover after delivery (Fig. 3C).

V₁ AVP receptor blockade does not enhance temperature and iBAT responses to PGE₂. Enhanced AVP antipyresis has been proposed to underlie fever suppression at term; therefore, we determined if coinfusion of a V₁ AVP receptor antagonist, Manning compound, with PGE₂ would result in an elevated fever. We used a dose of the antagonist that was previously shown to be effective in elevating fever height and prolonging its duration as a result of an identical dose of PGE₂ (3). Rectal temperature (P16-17, 36.3 ± 0.1; P19-20, 36.4 ± 0.1; L1-2, 36.0 ± 0.2°C) and iBAT temperature (P16-17, 36.4 ± 0.1; P19-20, 36.6 ± 0.3; L1-2, 36.3 ± 0.3°C) were identical in all groups (P > 0.5). Rectal temperature changes followed a similar course and magnitude as had been seen with PGE₂ alone in that responses were greatly reduced at near term (Fig. 4, A and B; 2-way ANOVA followed by Student-Newman-Keuls post hoc, P < 0.05). Thus coinfusion of Manning compound, an AVP receptor antagonist, did not reverse the reduced PGE₂-induced temperature changes seen at near term. Similarly, coinfusion of Manning compound and PGE₂ produced comparable iBAT nerve activity responses to those seen with PGE₂ alone (Fig. 4C). Analysis of iBAT nerve activities indicated that the PGE₂-induced firing rate was reduced at near term despite the presence of Manning compound (2-way ANOVA followed by Student-Newman-Keuls post hoc).

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Rectal temperature (A), iBAT temperature (B), and iBAT nerve activity (C) responses to coinfusion of Manning compound (1 nmol) and PGE2 at P16-17, P19-20, and L1-2 (n = 5) at time 0.

An alternate way of analyzing time-dependent responses such as temperature is to calculate the integrated response (i.e., area under the curve). Analysis of the integrated changes in rectal and iBAT temperatures and iBAT nerve activities in response to PGE₂ or Manning compound plus PGE₂ in anesthetized rats revealed a significant change as a result of gestational age, but not as a result of blockade of V₁ receptors by Manning compound. These results are identical to those obtained after analysis of time-dependent data.

AVP V₁ receptor antagonism does not change PGE₂ fever in conscious rats. Although the AVP antagonist was without apparent effect at near term in anesthetized pregnant rats, we wished to verify in conscious, free-moving rats that enhanced AVP antipyresis was not responsible for the suppression of fever at near term. Thus we gave PGE₂ (25 ng) or PGE₂ plus Manning compound (1 nmol) to near-term pregnant rats instrumented with Mini-Mitter telemetry devices. Fevers in both groups of rats were similar to those seen in anesthetized females and resembled those we had previously observed in conscious, unanesthetized females in our lab (3). Of particular note is that coinfusion of PGE₂ with the Manning compound did not enhance the febrile responses (F = 1.3, P > 0.05; Fig. 5A), a finding identical to that seen in the anesthetized animals.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Effects of V1 antagonist on PGE2 fever in conscious near-term pregnant rats and arginine vasopressin (AVP) release in ventral septal area (VSA). A: net changes in core temperature responses to PGE2 (25 ng in 5 µl saline) alone or in combination with Manning compound (MC; 1 nmol) in conscious rats at day 20 of pregnancy (n = 5-6). Time 0 indicates time of infusion. Data are means ± SE. Baseline temperatures in each group were similar (36.48°C, PGE2; 36.53°C, PGE2/MC). B: drawings showing placement of microdialysis probes. Dots are tip of probe, lines are range of effective dialysing membrane. lv, Lateral ventricle; ac, anterior commissure; ox, optic nerve. C: AVP content in 30-min microdialysates from VSA before and after icv PGE2 infusion (arrow, n = 7-8) given immediately after collection of perfusate 1.

AVP release in the ventral septal area is not altered. An additional way to examine potential AVP involvement in the febrile response is to measure the actual levels of the peptide in the active antipyretic locus in the VSA (3, 20). In the present experiment, the sites of microdialysis probe placement were identified histologically and were within the range defined previously as the VSA (Fig. 5B). Baseline release of AVP in the VSA was no different in P16-17 (0.42 ± 0.06 pg), near-term (0.48 ± 0.1 pg), or postparturient rats (0.40 ± 0.7 pg; F = 0.2, P > 0.05). After PGE₂ infusion, no increase over corresponding control levels in any of the groups was observed (Dunnett's test, P > 0.05; Fig. 5C).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The results indicate that PGE₂ fever was reduced at P19-20 (near term) compared with several days earlier, and the febrile response recovered 1-2 days after delivery. These changes were accompanied by reduced sympathetically mediated cardiovascular responses and, of particular importance, reduced iBAT temperature and nerve activity to the iBAT. There was, however, no apparent increase in AVP release in the VSA during fever, nor did blockade of AVP receptors in the brain reverse the suppression of fever or sympathetic responses. Thus AVP did not appear to be the central neurotransmitter involved in suppression of fever at near term.

Fever suppression at full-term pregnancy has been suggested to result from increased central AVP activity because its synthesis and terminal release are increased in certain brain regions during pregnancy (27, 36). It has been shown previously that central PGE₂ stimulates AVP release in the VSA in male rats but not in virgin females, indicating that female rats do not appear to use AVP as a major endogenous antipyretic (3). We postulated that the central AVP antipyresis system in females may become functional only at certain times, i.e., during the fever suppression at term. In this work, no increased release of AVP was seen in the VSA that would account for a suppression of fever, suggesting that PGE₂ does not activate this pathway in females, even at term. As there are other vasopressinergic sites that are implicated in antipyresis, for example the medial amygdala where AVP immunoreactivity is also increased during pregnancy (7, 27), it remains to be determined if pregnant females make use of other sites for endogenous antipyresis.

A lack of increase in AVP release does not preclude the possibility that pregnant females respond to basal levels of AVP in a sensitized manner as the result of receptor upregulation or increased cross talk with oxytocin receptors (34). However, whereas there is ample evidence for upregulation of brain oxytocin receptors during pregnancy (14), there is no evidence for a similar change in AVP receptors. Nonetheless, it was necessary to determine if blocking the AVP receptors would reveal an action of endogenous AVP, perhaps acting on sensitized receptors. However, the antagonist was without effect on fever. We used a dose of antagonist that we and others have found to be effective in blocking endogenous AVP activity during experimental fevers and in antagonizing the action of exogenously administrated AVP, even in sensitized animals (3, 16, 35, 41). Thus it is likely that the antagonist at the dose administered was adequate to block any of the AVP receptors that might have participated in the antipyretic response near term. In conclusion, the fact that the AVP antagonist did not enhance fever like it does in male rats indicates that AVP-mediated neurotransmission is unlikely to be activated during fever in the near-term pregnant rat.

Although we could find no evidence for suppressive influence of AVP at term, an alternative possibility is that sympathetic activation in response to PGE₂ could be reduced. If so, it is possible that other aspects of the sympathetic activation by PGE₂ would be similarly reduced. As shown in this study, the reduced febrile response was accompanied by reduced pressor and tachycardiac effects of PGE₂ at near term. There are several possible mechanisms for this hyporesponsiveness to central PGE₂. One possibility is a change in number or affinity of central prostaglandin receptors during pregnancy so that PGE₂ does not activate the sympathetic system to the same extent. We are unaware of any studies that have addressed this question. A second possibility is that there is increased catabolism and/or clearance of PGE₂ from the CSF such that there were lower effective concentrations at the receptors. In the sheep, the clearance of prostaglandin caused by uptake into the choroid plexus in vitro is slightly higher in the pregnant than in the nonpregnant adult (17), but information as to whether such an uptake process changes throughout the gestational period or indeed what happens in the rat is not available.

We have also considered the possibility that animals with reduced fevers were unable to vasoconstrict sufficiently to conserve heat and develop a fever. In support of this possibility are our findings of reduced pressor responses to central PGE₂ and the fact that the peripheral vasculature in late pregnancy is less sensitive to vasoconstrictive substances (6). However, we feel that this is unlikely because the febrile response to PGE₂ recovered during the post partum period when the pressor responses were still suppressed. Furthermore, in the rat PGE₂ appears to raise body temperature largely by activation of thermogenesis, in particular in the iBAT (9), and peripheral vasoconstriction appears to contribute little to the increase in body temperature.

Another possibility for the reduced fever is that there is reduced peripheral sympathetic drive to effector organs. To directly examine this possibility in terms of fever responses we examined the temperature response of the iBAT and the activity of its sympathetic nerve. We found that the sympathetic nerve activity to the iBAT after PGE₂ was reduced at near term, strongly suggesting that fever reduction at this time arises from diminished sympathetically driven iBAT thermogenesis. However, even at near term, iBAT nerve activity was still slightly increased by PGE₂ infusion (although to a much lesser extent compared with other gestation groups), but iBAT temperature change was very small. This could be a result of the fact that there is not a linear relationship between the level of sympathetic activity and the metabolic response in the iBAT. In addition to reduced central drive, there could be also additional reduced iBAT responsiveness to sympathetic drive. There is indeed evidence that estradiol, which is high just before term (40), inhibits iBAT thermogenesis by reducing iBAT responsiveness (29). Therefore, fever reduction at near term can be accounted for by a combined effect of reduced sympathetic drive to and reduced responsiveness of the iBAT.

Our results indicate that there is a transient but quite profound decrease in the sympathetic response to central PGE₂ at near term. On the basis of a dissociation between CSF PGE₂ levels and febrile responses in newborn lambs, a similar conclusion was reached by Coceani and colleagues concerning the reduced fevers in these animals (4). The mechanism for this is still obscure, but it is unlikely to be a result of an active suppression of sympathetic activation by AVP, because blockade of V₁ receptors did not change the sympathetic nerve activity to the iBAT. An alternate mechanism for a suppression of sympathetic activation would be via an increase in GABA-mediated inhibition of sympathetic effector neurons in the ventral lateral medulla caused by progesterone metabolites (10). Given the ubiquity of GABAergic inhibition in the brain, this could account for generalized inhibition of diverse sympathetic responses such as thermogenesis and cardiovascular activation. An equally distinct possibility is that there are pregnancy-related changes in prostaglandin receptors in the brain which are involved in thermogenesis (1, 31) such that their number or affinity are reduced at near term. Finally, other antipyretic substances such as melanocortins (13) may become more active at near term.