Fever and lethargy induced by subcutaneous pyrogen infusion in unrestrained rats

We have investigated the effects of continuous subcutaneous infusion of lipopolysaccharide (LPS), muramyldipeptide (MDP), or saline on abdominal temperature and voluntary activity in unrestrained rats. Both pyrogens were infused via osmotic pumps at a rate of approximately 2 microg.kg-1.min-1 for 7 d. LPS infusion evoked a 3-d and MDP a 1-d elevation in body temperature. Night-time activity was suppressed on days 1 and 2 during LPS infusion and on day 1 of MDP infusion. Body mass was significantly decreased on infusion day 4 in rats receiving either LPS or MDP; however, the rate of weight gain had been restored by day 8 (1 d after cessation of pyrogen infusion). We further tested the body temperature response of the same experimental animals to a single subcutaneous bolus injection (250 microg/kg) of the same pyrogen that had been infused for 7 d, 2 d after cessation of pyrogen infusion (day 9). The fever response in rats receiving a bolus injection of either LPS or MDP was significantly attenuated in rats that had previously been infused with the same pyrogen. These data suggest that tolerance developed to continuous infusion of both Gram-negative and Gram-positive pyrogens, and that mechanisms of tolerance development set in early during the 7-d infusion period of both pyrogens and persisted for at least 2 d after the cessation of pyrogen infusion. We propose that cytokine intermediates were involved or required in inducing these responses to continuous infusion of both LPS and MDP.     

Fever in rats during normal and dehydrated conditions

The effect of endogenous pyrogen (EP, from rabbit) and endotoxin (Salmonella typhosa) on rectal temperature (Tre) was investigated in normal and dehydrated rats of both sexes. Intraperitoneal injection of either EP or endotoxin did not affect body temperature. In addition, no changes in Tre were observed when endotoxin was injected intravenously in normally hydrated male rats, but significant falls in Tre occurred in normal female rats. However, intravenous injection of EP produced fever in both sexes, but females generally showed smaller responses. A second intravenous injection of endotoxin, given 3 days after the first injection, always produced fever in normally hydrated rats. The pattern of this febrile response was monophasic. In contrast to the response in normal rats, intravenous endotoxin produced significant fevers with a biphasic pattern in dehydrated rats of either sex, but the febrile responses of male rats were greater than those of female rats. On the other hand, there were no significant differences between febrile responses to intravenous EP exhibited by normal and dehydrated animals. These results show that rats of both sexes possess physiological mechanisms capable of producing a fever following intravenous injections of EP.     

Differences in endogenous pyrogen fevers induced by iv and icv routes in rabbits

We have compared the characteristics of fevers produced by endogenous pyrogen administered by the intravenous (iv) and by the intracerebroventricular (icv) routes in conscious rabbits. Fevers induced by the intracerebroventricular route have a longer latency to onset, a less steep rise in body temperature, and a longer time to peak elevation in body temperature than do fevers induced by the intravenous route. Furthermore, a dose of indomethacin (2 mg/kg) administered intravenously, which is effective in markedly attenuating fevers produced by the intravenous route, was completely without effect on fevers induced by the intracerebroventricular route. On the other hand, when indomethacin (500 micrograms) was infused intracerebroventricularly, it markedly reduced fevers induced by the subsequent injection of endogenous pyrogen into the contralateral cerebral ventricle, but such pretreatment had little effect on fevers elicited by intravenous injections of endogenous pyrogen. It is concluded that the sites of action of endogenous pyrogen in response to intravenous injections of pyrogen are different from those responding to intracerebroventricular injections of pyrogen and that this is manifest in several distinct differences in the characteristics of the two fevers. These results indicate that the intracerebroventricular model of fever production is not appropriate for the study of the normal pathogenesis of fever.     

Body temperature, behavior, and plasma cortisol changes induced by chronic infusion of Staphylococcus aureus in goats

Most experimentally induced fevers are acute, usually lasting approximately 6-12 h, and thus do not mimic chronic natural fevers, which can extend over several days or more. To produce a model of chronic natural fever, we infused eight goats (Capra hircus) intravenously with 2 ml of 2 x 10(11) cell walls of Staphylococcus aureus (S. aureus) for 6 days using osmotic infusion pumps (10 microl/h) while measuring changes in body temperature, behavior, and plasma cortisol concentration. Seven control animals were infused with sterile saline. Abdominal temperature-sensitive data loggers and osmotic infusion pumps were implanted under halothane anesthesia. To compare our new model with existing models of experimental fever, we also administered 2-ml bolus intravenous injections of 2 x 10(11) S. aureus cell walls, 0.1 microg/kg lipopolysaccharide (Escherichia coli, serotype 0111:B4), and sterile saline in random order to six other goats. Bolus injection of lipopolysaccharide and S. aureus induced typical acute phase responses, characterized by fevers lasting approximately 6 h, sickness behavior, and increased plasma cortisol concentration. Infusion of S. aureus evoked prolonged fevers, which lasted for approximately 3 days, starting on day 4 of infusion (ANOVA, P < 0.05), and did not disrupt the normal circadian rhythm of body temperature. However, pyrogen infusion did not cause plasma cortisol concentration to rise (ANOVA, P > 0.05) or the expression of sickness behavior. In conclusion, infusion of S. aureus produced a fever response resembling that of sustained natural fevers but did not elicit the cortisol and behavioral responses that often are described clinically and during short-term experimental fevers.     

Responses of baboons to traditionally pyrogenic agents

It is not clear whether baboons develop fever in response to endotoxin or other pyrogens. We injected various pyrogens intravenously in 12 unrestrained baboons (Papio ursinus) and measured their body temperature using intra-abdominal radiotelemeters. Serum iron concentration was also measured. The baboons developed fever after injection of killed Staphylococcus aureus (5 X 10(7) organisms/kg). No significant fever was measured after injection of lipopolysaccharide (Salmonella typhosa) (0.1, 8, 40, and 100 micrograms/kg), bovine serum albumin (4 mg/kg), killed Salmonella minnesota (5 X 10(7) organisms/kg), and killed Salmonella typhi (5 X 10(7) organisms/kg). A significant decrease in serum iron concentration was found only after injection of S. aureus and lipopolysaccharide, 100 micrograms/kg. The phagocytic synthesis of interleukin-1 following pyrogen stimulation in baboons and some other primates appears to differ from that in man and in nonprimates.     

A comparison of the febrile responses of the Brattleboro and Sprague-Dawley strains of rats to endotoxin and endogenous pyrogens

The febrile responses of homozygous (di/di) Brattleboro rats, to both intravenous endogenous pyrogen and to a lipopolysaccharide endotoxin, were compared with those of normal Sprague-Dawley rats. There were no detectable differences between the fever curves of the two strains in response to endogenous pyrogen. Brattleboro rats, which are deficient in the neuropeptide arginine vasopressin (AVP), displayed fevers that were both qualitatively and quantitatively indistinguishable from those of normal Sprague-Dawley rats that do not suffer from congenital diabetes insipidus. It is concluded that the absence of AVP-containing cells in Brattleboro rats is not an important factor in determining the nature of their febrile responses to endogenous pyrogen. More remarkable, however, were the divergent febrile responses of the two strains to intravenously injected endotoxin. Normal rats displayed hypothermic responses, whereas the Brattleboro rats became febrile. By 2 h after the injection of endotoxin, body temperatures in both strains had returned to normal. Three hours after the rats had been exposed to endotoxin, both strains were found to be totally refractory to endogenous pyrogen. However, when both strains of rats were tested to endogenous pyrogen 3 days later, their febrile responses were more than double the magnitude of their initial control responses. These alterations in the febrile responsiveness of rats occurring at different times after the injection of endotoxin appear to be related to the effects that endotoxin has on the cells of the reticuloendothelial system, over the same time course.     

Antipyrogenic properties of oxytocin

Effects of oxytocin on pyrogenal or endogenous pyrogen-induced fever were studied. Intramuscular injection of oxytocin (0.2 micrograms/kg every half an hour) did not significantly affect the pattern of pyrogenal-induced fever. Constant intravenous injection of oxytocin (0.4 and 4 micrograms/kg/h) 2-4.5 h after pyrogenal decreased the rectal temperature, on an average, by 24% and 31%, respectively. Endogenous pyrogen fever was not attenuated by intravenous oxytocin (4 micrograms/kg/h). The antipyrogenic effect of oxytocin is related to inhibition of endogenous pyrogen synthesis rather than to blockade of its action, which is indicated by a decreased second peak of the temperature curve, inhibition of endogenous pyrogen synthesis in vitro, and persistence of the hyperthermic effect of endogenous pyrogen.     

The effect of α-MSH on fever caused by Staphylococcus aureus cell walls in rabbits

The role of endogenous pyrogens induced by gram-positive bacterial pyrogens is not known. Intravenous alpha-MSH (2.5 micrograms) significantly reduced only the first phase of the biphasic thermal response to IV S. aureus cell walls (5 x 10(7)). Intracerebroventricular alpha-MSH (200 ng) had no effect on the fever response. The fall in serum iron concentration was significantly attenuated by the IV alpha-MSH but was not affected by the ICV alpha-MSH. Intravenous alpha-MSH had no effect on fever or the serum iron response caused by muramyl dipeptide (MDP). We conclude that the first phase of the thermal response to S. aureus cell walls is mediated by an endogenous pyrogen (EP) and the second phase of the response by a mechanism not involving EP, but possibly a muramyl peptide.     

Effects of nitric oxide synthase inhibitors on the febrile response to muramyl dipeptide and lipopolysaccharide in rats

We have administered aminoguanidine, a relatively specific inhibitor of inducible nitric oxide synthase, and N-nitro-L-arginine methyl ester (L-NAME), an unspecific nitric oxide synthase inhibitor, to rats made febrile with the gram-positive pyrogen, muramyl dipeptide and gram-negative pyrogen, lipopolysaccharide. Sprague-Dawley rats, housed individually at approximately 25 degrees C with a 12:12 h light:dark cycle (lights on 0700 hours), were injected (at 0900 hours) intraperitoneally with 50 mg/kg aminoguanidine, 25 mg/kg or 50 mg/kg L-NAME, and intramuscularly with 500 microg/kg muramyl dipeptide or 100 microg/kg lipopolysaccharide. Pyrogen injections were spaced at least 14 days apart. Body temperature was measured throughout the study in unrestrained animals using radio-telemetry. Neither muramyl dipeptide nor lipopolysaccharide-induced fevers were affected by aminoguanidine. However, L-NAME administration inhibited muramyl dipeptide and lipopolysaccharide-induced fevers, but only for the 1st 2-4 h of the fevers (two-way ANOVA, P<0.05). After the initial inhibition, lipopolysaccharide fevers developed normally. Therefore, constitutively expressed nitric oxide synthase appears to be involved in the initial phases of fever genesis of gram-negative and gram-positive fevers in rats. On the other hand, inducible nitric oxide synthase appears not to play a role in these fevers.     

Immunoadjuvants enhance the febrile responses of rats to endogenous pyrogen

The febrile responses of male Sprague-Dawley rats to a semipurified endogenous pyrogen produced from human monocytes were characterized by establishing fever dose-response curves. The animals were then injected intravenously with a number of substances that possessed the common properties of stimulating the phagocytic activity of the cells of the reticuloendothelial system and of acting as immunoadjuvants. The substances used were zymosan, lipopolysaccharide endotoxin, and muramyl dipeptide. Three days after any of these immunoadjuvants were injected, the fever sensitivity of the rats was remeasured. In each case, the slope of the fever dose-response curve tripled, and in some instances the response threshold for fever response was reduced by factors of three to eight. Furthermore, the maximum increase in body temperature produced by the endogenous pyrogen was more than doubled after immunoadjuvant treatment. By contrast latex beads, which are also phagocytized by the cells of the reticuloendothelial system but do not subsequently increase their phagocytic index nor do they enhance immune responses, had no effect on the fever sensitivity of rats in response to endogenous pyrogen. In the light of these findings, it is suggested that the febrile responses of rats to endogenous pyrogen are mediated in some manner by cells that possess some of the properties of reticuloendothelial cells. The location of these putative cells must be close to the circulation, because the immunoadjuvants used in this study were, for the most part, large molecular weight molecules that could not cross the blood-brain barrier easily.     

Restraint increases afebrile body temperature but attenuates fever in Pekin ducks (Anas platyrhynchos)

In mammals, procedures such as handling, restraint, or exposure to open spaces induces an increase in body temperature (T(b)). The increase in temperature shares some characteristics with pyrogen-induced fever and so is often called "stress fever." Birds also respond to acute handling with a stress fever, which may confound thermoregulatory studies that involve animal restraint. We have measured the T(b) responses of Pekin ducks on days when they were restrained and compared them to days when the birds remained unrestrained. Restraint induced a 0.5 degrees C increase in T(b) that was sustained for the entire 8 h of restraint. To determine whether the restraint-induced increase in T(b) is mediated by prostaglandins (PGs) we compared the T(b) responses during restraint after intraperitoneal injection with saline to the responses during restraint after injection with diclofenac sodium (15 mg/kg). There was no difference in response, suggesting that restraint affects T(b) by a PG-independent mechanism. We also compared the T(b) response to intramuscular injection of lipopolysaccharide (LPS; 100 microg/kg), a bacterial pyrogen, when the ducks were restrained or unrestrained. Despite T(b) being higher at the time of LPS injection when the ducks were restrained, the maximum temperature reached after LPS injection was higher, and the period that T(b) remained elevated was longer when the ducks were unrestrained. We conclude that restraint should be considered as a potential confounder in thermoregulatory studies in birds and presumably other species too.     

Blunted febrile response to intravenous endotoxin in starved rats

The effects of fasting on the febrile responses to intravenous injection of bacterial lipopolysaccharide (LPS; endotoxin) of Escherichia coli were investigated in rats. Ad libitum-fed rats (C) produced a biphasic fever with an increase in the temperature difference between brown adipose tissue and colon and shivering activity (SA). Measurement by a direct calorimeter showed no particular changes in heat loss. Rats starved for 4 days (F4) responded to intravenous LPS with a monophasic fever accompanied by an increase in SA only. However the maximal rise in colonic temperature (Tco) did not differ from C rats. Subsequent 2-day fasting reduced SA and the maximal fever height. Endogenous pyrogen (EP) injected intravenously produced a prompt rise in Tco followed by prolonged hyperthermia in C rats. In the F4 rats, there was no such sustained rise in Tco as a result of intravenous EP. The response in Tco to intravenous prostaglandin E2 (PGE2) was the same in fed and starved rats. The administration of LPS, EP, and PGE2 into the lateral ventricle evoked a similar extent of hyperthermia in C and F4 rats. Because the second phase of fever has been shown to occur after pyrogens are translated into a febrile stimulus within the blood-brain barrier, it is assumed that the functional changes of the blood-brain barrier such as in the permeability of pyrogens or in the sensitivity of pyrogen receptors resulted in the absence of the second phase of fever in starved rats.     

Circadian Rhythm of Blood Ethanol Clearance Rates in Rats: Response to Reversal of the L/D Regimen and to Continuous Darkness and Continuous Illumination

Phase relationships of the circadian rhythms of blood ethanol clearance (metabolic) rates and body temperature were studied in rats successively exposed to 4 illumination regimens: LD (light from 0800-2000 hr), DL (light from 2000-0800 hr), constant darkness (DD) and, lastly, constant light (LL). After a 4-wk standardization to each regimen, body temperatures were taken at 9 × 4-hr intervals to establish baseline circadian profiles. One week later, groups (N = 8) received 1.5 g/kg ethanol (i.p.) at 6 equally spaced timepoints during a 24-hr span, when temperatures were again measured. Ethanol clearance rates were estimated from decreasing blood ethanol levels sampled every 20 min from 60-200 min after dosing, and the resultant elimination curves were subjected to cosinor analysis. These studies show for the first time that the high amplitude circadian rhythm in ethanol metabolism persists under constant conditions of illumination (DD and LL), demonstrating that it may well be a truly internal circadian rhythm and not a response to exogenous cues of the light/dark cycle. During both LD and DL, maximal and minimal ethanol clearance rates fell near the end of the dark and light phases, respectively, and followed circadian peak and trough control temperatures by approximately 6 hr. A fixed internal phase relationship between the core body temperature and the circadian rhythm in ethanol metabolism is demonstrated, thus establishing the rhythm in body temperature as a suitable and convenient internal marker rhythm for studies of the metabolism of low-to-moderate ethanol doses. These studies demonstrate that the phase relationships of blood ethanol clearance rate and body temperature can be manipulated by the illumination regimen selected, an observation of both basic and practical importance.     

Circadian variation in circulating pyrogen: possible role in resistance to infection

In the rat, body temperature (bt) is highest, and plasma iron (Fe) and zinc (Zn) concentrations are lowest at night while the rat is most active; the inverse is true during the day. Based on data implicating endogenous pyrogen (EP) as a mediator of the rise in body temperature and fall in plasma trace metal levels during infection we hypothesized that the circadian rise in body temperature and fall in plasma Fe and Zn levels may be attributed to a cyclic release of EP. To test this hypothesis: (1) Rats were injected ip with an antipyretic dose of sodium salicylate (300 mg/kg). The result was a reduction (P less than 0.05) in bt at night. (2) Rats were injected during the day with 1 ml each of plasma collected from rats during the night. As a control, rat plasma collected during the day was injected at this same time point. A rise (P less than 0.05) in bt was observed only in animals who had received plasma collected at night. These results support the hypothesis that a pyrogen, perhaps EP, is present in the plasma of rats at night. The release of EP during periods of greatest activity may have an adaptive role since rats are more likely to come into contact with pathogens during these times. If EP were released during periods of activity, the likelihood of severe infection occurring would be diminished. To test this hypothesis, two groups of rats were injected with Salmonella typhimurium, one group at midnight (A) and one group at noon (B). The mortality rate was 25% in group A and 60% in group B (P less than 0.025). These data support the hypothesis that the immune/host defense of rats to S. typhimurium is more effective at night, possibly due to an increased level of circulating pyrogen.     

Response of rabbits to pyrogen in a helox environment

1. 1.|The effectiveness of a gas environment consisting of 80% helium and 20% oxygen (Helox) in reducing a rabbit's fever due to an i.v. injection of endotoxin was found to be dependent on the amount of pyrogen injected.

2. 2.|When a relatively large dose of pyrogen was injected, the helox environment used in these experiments reduced the mean maximum temperature reached during the fever from 41.5 to 40.5°C, but the helox did not significantly alter the change in temperature from baseline levels prior to the injection (a 1.4°C increase in air and a 1.1°C increase in helox).

3. 3.|When a relatively low dose of pyrogen was injected, the helox environment increased the change in temperature from baseline at peak fever, but did not produce a significant change in the actual temperatures attained during the fever.

Central and peripheral injections of ACTH (1–24) reduce fever in adrenalectomized rabbits

Central administration of ACTH (1-24) reduces fever in normal rabbits in doses that have no effect on afebrile body temperature. Previous experimental and clinical reports indicate that peripheral administration of both ACTH and corticosteroids reduces fever, and since central injection of corticosteroids can also lower fever it might be that the antipyretic effect of intracerebroventricular (ICV) ACTH (1-24) is due to adrenal stimulation. To learn whether this endogenous central peptide can produce antipyresis independently, ACTH (1-24) was injected ICV in bilaterally adrenalectomized (ADX) rabbits made febrile by IV injections of leukocytic pyrogen (LP). ACTH (250 ng) given ICV reduced fever in these animals and had a slight hypothermic effect when given to the same rabbits when they were afebrile. Doses of 25-75 ng reduced fever without influencing normal body temperature. Intravenous injections of ACTH (2.5 micrograms) also lowered fever caused by IV LP in ADX rabbits. The present findings raise the possibility that release of endogenous central ACTH, and perhaps entry into the brain of circulating ACTH, the release of which is known to increase in fever, limits the magnitude of the febrile response by influencing central temperature controls.     

α-MSH injected into the septal region reduces fever in rabbits

In previous research the concentration of α-MSH within the septal region of rabbits increased with fever. This finding raises the possibility that the septal concentration of this peptide, which reduces fever when given both peripherally and intracerebroventricularly, is important to limitation of fever. To test this idea, rabbits with cannulas in the septal region were made febrile by IV injections of leukocytic pyrogen (LP). Injection of α-MSH (1 μg bilaterally) into the septal region did reduce fever, consistent with the idea that the increase in septal α-MSH concentration which occurs naturally in fever limits the febrile response. We also noted late rises in body temperature when experimental and control septal injections were given close together in time. These increases in temperature were similar to those known to occur after injections into the primary temperature control in the PO/AH region. This commonality further strengthens the possibility that septal neurons are important to central modulation of body temperature.     

Perfusion of the septum of the rabbit with vasopressin antiserum enhances endotoxin fever

The septal region of the brains of conscious, adult, male New Zealand White rabbits were perfused by means of a push-pull system before and after an intravenous administration of bacterial pyrogen extracted from Salmonella abortus equi. Perfusion of the septal area with sucrose solution (260 mM) had no significant effect on the resulting fever (1.13 +/- 0.09 degrees C) when compared to a control fever without the push-pull perfusion (1.06 +/- 0.12 degrees C). Arginine vasopressin (AVP) added to the perfusing solution (20 micrograms/ml) caused a significant attenuation of the fever (0.81 +/- 0.20 degrees C). An antiserum specific to AVP when added to the perfusing solution resulted in a fever which was significantly greater (2.38 +/- 0.13 degrees C) than the control. Radioimmunoassay of perfusates collected from the control perfusions before and during fever showed that, as the body temperature rose in response to the pyrogen, the level of AVP in the perfusate collected from the septal area decreased. These results provide further evidence that AVP may act in the septal area of the brain to modulate the febrile response.     

Fever in young lambs: hypoxemia alters the febrile response to a small dose of bacterial pyrogen.

Experiments were done on eight young lambs to investigate the effects of hypoxemia on the body temperature, metabolic and cardiovascular responses to intravenous administration of a small dose of bacterial pyrogen (0.3 micrograms lipopolysaccharide extracted from Salmonella Abortus Equi; SAE). Each lamb was anaesthetized with halothane and prepared for sleep staging and measurements of cardiac output, arterial and mixed-venous haemoglobin oxygen saturations, body-core and ear-skin temperatures. Three experiments were done on each lamb, the first being done no sooner than three days after surgery. The first experiment consisted of establishing the thermal neutral environment during normoxemia (ie, environmental temperature at which total body oxygen consumption was minimal while body temperature was maintained) for each lamb. The second and third experiments were done at the lamb's thermoneutral environment as determined on day 1. One experiment was done during normoxemia (ie, control condition, SaO2 approximately 90%) and one experiment was done during hypoxemia (ie, experimental condition, SaO2 approximately 50%). Measurements were made during a control period and during one-minute experimental periods at 10 minute intervals for 120 minutes following administration of 0.3 micrograms of bacterial pyrogen in sterile saline. Administration of SAE produced a short-lived fever of about 0.8 degrees C in the normoxemic lambs, whereas no change in body-core temperature was observed in the hypoxemic lambs. During normoxemia, the increase in body-core temperature was preceded by peripheral vasoconstriction, the onset of shivering, and a surge in total body oxygen consumption. The increase in total body oxygen consumption was met primarily by an increase in total body oxygen extraction during the development of fever. Cardiac index, heart rate, and systemic oxygen transport increased during the peak body-core temperature response. Systemic arterial blood pressure did not change significantly during the febrile response; however, pulmonic arterial blood pressure increased. During hypoxemia, peripheral vasoconstriction and shivering occurred following administration of SAE, but there was no change in total body oxygen consumption or body-core temperature. Thus, our data provide evidence that hypoxemia alters the febrile response of young lambs to bacterial pyrogen. The precise mechanism remains to be determined.     

Effects of malaria on O2 consumption and brown adipose tissue activity in mice

Increased energy expenditure often occurs during illness or after injection of endotoxin and can contribute to the generation of fever. In laboratory rats and mice the thermogenic response has been attributed to the sympathetic activation of brown adipose tissue (BAT), although mice often fail to show pyrexia. In this study the effects of malaria on O2 consumption and BAT were studied in mice inoculated with Plasmodium berghei. Parasitemia was maximal (greater than 50% of erythrocytes showing positive Leishman staining) 72 h after inoculation. Up to this time body weight and food intake were similar to values for control mice, although colonic temperatures were slightly depressed in infected mice. Thereafter, infected mice showed marked hypophagia, loss of body weight, and severe hypothermia; colonic temperature was less than 31 degrees C at 96 h when the experiment was terminated. Resting O2 consumption (VO2) measured at 24 degrees C was slightly elevated in infected mice 12 h after inoculation and reached a peak value (31% above controls) at 48 h. VO2 returned to the same value as controls at 96 h. In vitro thermogenic activity of BAT (assessed from binding of guanosine diphosphate to isolated mitochondria) was not significantly altered in infected mice. These data demonstrate a marked thermogenic response to malarial infection, but this is not accompanied by fever in mice and is dissociated from stimulation of BAT activity.