The FASEB Journal. 2020;34:5863–5876. wileyonlinelibrary.com/journal/fsb2
|
5863R E S E A R C H A R T I C L E
The generation of immune-induced fever and emotional
stress-induced hyperthermia in mice does not involve brown
adipose tissue thermogenesis
Anna Eskilsson
1|
Kiseko Shionoya
1|
Sven Enerbäck
2|
David Engblom
1|
Anders Blomqvist
1This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
© 2020 The Authors. The FASEB Journal published by Wiley Periodicals, Inc. on behalf of Federation of American Societies for Experimental Biology
Abbreviations: 5HT, 5-hydroxytryptamine; 8-OH-DPAT, 8-hydroxy-2-(di-n-propylamino)tetralin hydrobromide; BAT, brown adipose tissue; IL-1β,
interleukin-1β; i.p., intraperitoneally; i.v., intravenously; KO, knockout; LPS, lipopolysaccharide; PGC-1α, peroxisome proliferator-activated receptor-γ coactivator-1α; UCP, uncoupling protein; WT, wild type.
1Department of Biomedical and Clinical
Sciences, Linköping University, Linköping, Sweden
2Department of Medical Biochemistry and
Cell Biology, Institute of Biomedicine, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden
Correspondence
Anders Blomqvist, Division of
Neurobiology, Department of Biomedical and Clinical Sciences, Linköping University, Linköping, Sweden. Email: anders.blomqvist@liu.se
Funding information
Vetenskapsrådet (VR), Grant/Award Number: 2016-01301; Hjärnfonden (Swedish Brain Foundation), Grant/Award Number: FO2019-0033; Swedish Research Council, Grant/Award Number: 2016-01301; Brain Foundation, Grant/Award Number: FO2019-0033
Abstract
We examined the role of brown adipose tissue (BAT) for fever and emotional stress-induced hyperthermia. Wild-type and uncoupling protein-1 (UCP-1) knock-out mice were injected with lipopolysaccharide intraperitoneally or intravenously, or subjected to cage exchange, and body temperature monitored by telemetry. Both genotypes showed similar febrile responses to immune challenge and both displayed hyperthermia to emotional stress. Neither procedure resulted in the activation of BAT, such as the induction of UCP-1 or peroxisome proliferator-activated receptor-γ coactivator-1α (PGC-1α) mRNA, or reduced BAT weight and triglyceride content. In contrast, in mice injected with a β3 agonist, UCP-1 and PGC-1α were strongly
induced, and BAT weight and triglyceride content reduced. Both lipopolysaccharide and the β3 agonist, and emotional stress, induced UCP-3 mRNA in skeletal muscle.
A β3 antagonist did not attenuate lipopolysaccharide-induced fever, but augmented
body temperature decrease and inhibited BAT activation when mice were exposed to cold. An α1/α2b antagonist or a 5HT1A agonist, which inhibit vasoconstriction,
abolished lipopolysaccharide-induced fever, but had no effect on emotional stress-in-duced hyperthermia. These findings demonstrate that in mice, UCP-1-mediated BAT thermogenesis does not take part in inflammation-induced fever, which is dependent on peripheral vasoconstriction, nor in stress-induced hyperthermia. However, both phenomena may involve UCP-3-mediated muscle thermogenesis.
K E Y W O R D S
lipopolysaccharide, peroxisome proliferator-activated receptor-γ coactivator-1α, uncoupling protein-1, uncoupling protein-3, vasoconstriction
1
|
INTRODUCTION
Brown adipose tissue (BAT) has since long been known to be present in, for example, rodents as well as in human in-fants and to be critical for cold-induced non-shivering ther-mogenesis.1,2 The discovery of functional BAT also in adult
humans3-5 has propelled a novel interest in BAT because of
its metabolic importance, especially in relation to obesity and type II diabetes.2 BAT has also been implicated in
thermogen-esis unrelated to cold exposure. It has hence been suggested from observations in rats that both inflammation-induced fever and emotional stress-induced hyperthermia depend on non-shivering thermogenesis through the sympathetic activa-tion of BAT.6-9 However, the evidence for such roles for BAT
is largely indirect, such as the demonstration of increased py-rogen induced activity in peripheral sympathetic nerves as-sociated with but not solely innervating BAT10 and observed
temporal differences between pyrogen induced temperature increases in BAT and other body sites, such as the rectum and peritoneal cavity.6,7,9
BAT thermogenesis is dependent on uncoupling protein-1 (UCP-1), which works as a proton translocator within the inner mitochondrial membrane that, upon activation, short circuits the respiratory chain, thereby dissipating chemical energy as heat.11 Studies of mice with a genetic deletion of
UCP-112 have demonstrated the critical role of this protein
for adaptive adrenergic non-shivering thermogenesis.13,14
UCP-1 knockout mice are therefore an ideal tool to exam-ine the role of BAT thermogenesis also for immune-induced fever and emotional stress-induced hyperthermia. However, little such data exists. Okamatsu-Ogura et al examined the febrile response to an intraperitoneal injection of interleu-kin-1β (IL-1β) in UCP-1 knockout and wild-type mice and found no difference between genotypes.15 Unfortunately,
these authors examined only a brief time window (1 hour) during which handling stress-induced hyperthermia may ob-scure the immune-induced febrile response. Szentirmai and Kapás16 determined the effects of systemic inflammation on
sleep and body temperature in UCP-1 knockout and wild-type mice. They found that intraperitoneal injections of inflamma-tory substances abolished non-rapid-eye movement sleep in UCP-1 knockout animals but did not affect fever. However, injections were given immediately prior to dark onset and recordings were carried out during the dark, active period. Hence, the temperature data likely were compromised by the animals’ activity patterns. Finally, Riley and collaborators17
reported that UCP-1 knockout mice had significantly elevated thermoregulatory responses to immune challenge compared to wild-type mice; however, the sample size was very small.
Here, we examined the febrile response of UCP-1 knock-out mice, using both intraperitoneal and intravenous injec-tions of bacterial wall lipopolysaccharide, a well-established pyrogen that elicits an inflammatory cascade similar to that
seen during natural infections.18 We also examined the
tem-perature response of these mice to emotional stress. We find that UCP-1 is dispensable for both phenomena and that nei-ther immune stimulation nor emotional stress evokes any signs of BAT activation. Instead, our findings indicate that fever and stress-induced hyperthermia are dependent on other mechanisms such as peripheral vasoconstriction and/or mus-cle thermogenesis.
2
|
MATERIALS AND METHODS
2.1
|
Animals
Adult UCP-1 knockout mice, in which exons 2 and 3 are re-placed by a neomycin resistant gene, were progeny of mice created by Enerbäck and collaborators.12 The mice were held
on a C57BL/6 background and bred at an ambient tempera-ture of 21°C and on a 12 hours light/dark cycle (lights on at 7:00 AM) in a specific pathogen-free facility, with free access to food and water. In all experiments on these mice, littermates were used as wild-type controls and groups were balanced with respect to sex and age. For experiments using only wild-type animals, adult male C57BL/6 mice were pur-chased from Janvier Labs (Le Genest-Saint-Isle, France). All animal experiments were approved by the Linköping Animal Ethics Committee and followed international guidelines.
2.2
|
Substances
Lipopolysaccharide (LPS) from Escherichia coli (O111:B4) was purchased from Sigma-Aldrich (St. Louis, MO). CL 316243 (β3 agonist), SR 59230A hydrochloride (β3
antago-nist), 8-hydroxy-2-(di-n-propylamino)tetralin hydrobromide (8-OH-DPAT; 5HT1A agonist), and prazosin hydrochloride
(α1/α2B antagonist) were purchased from Tocris Bioscience
(Bristol, UK).
2.3
|
Temperature recordings
Mice were implanted intraperitoneally (i.p.) with a tran-sponder that records core body temperature (E-Mitter; Starr Life Sciences, Oakmont, PA). Surgery was performed under isoflurane anesthesia or injection anesthesia with ketamine (70 mg/kg) and dexmedetomidin (0.4 mg/kg). Where applicable, the mice were during the same surgical session provided with an indwelling jugular catheter that was exteriorized at the back of the neck and connected to a swivel system (CMA Microdialysis, Solna, Sweden) on the top of the cage, permitting injections without handling the mice (for details, see ref. 19). Following the surgery, mice
were either kept at an ambient temperature of 29°C, pro-viding near-thermoneutral conditions20 or at the ordinary
animal facility temperature of 21°C. About 1 week after transponder implantation, animals were injected i.p. with LPS (120 µg/kg) or with a β3 agonist (CL 316243; 0.1 or
0.5 mg/kg). Where applicable, the LPS or saline injection was immediately preceded by an intraperitoneal injection of a β3 antagonist (SR 59230A; 10 µg/kg), a 5HT1A agonist
(8-OH-DPAT; 0.5 mg/kg), or an α1/α2B antagonist
(prazo-sin hydrochloride; 1-2 mg/kg). Saline was used as a vehi-cle and as a control. Animals supplied with an indwelling jugular catheter were injected with LPS (30 µg/kg) three days after surgery. To elicit emotional stress, animals were placed in another animal’s cage. As a control, animals were lifted and placed back in their own cage (for details, see ref. 21).
2.4
|
Temperature recordings during
cold stress
Mice that had been kept at 29°C for 1 week after transponder implantation were injected i.p. with a β3 antagonist (SR
59230A; 10 µg/kg) or saline and placed in an indirect calo-rimetric system (INCA; Somedic, Hörby, Sweden), in which the ambient temperature gradually decreased to 7°C, and the body temperature continuously recorded. After the ani-mals were killed, their interscapular BAT was removed for analysis.
2.5
|
Quantitative real-time PCR
Mice were housed at an ambient temperature of 29°C for 1 week. They were then injected i.p. with LPS (120 µg/ kg), a β3 agonist CL 316243 (0.5 mg/kg), or saline, or
subjected to cage exchange as described above, and killed by asphyxiation with CO2 4 hours (LPS) or 2 hours (β3
agonist) postinjection, or 2-3 hours after cage exchange, respectively. Interscapular BAT and skeletal muscle from the right hind limb were excised and placed in RNA later stabilization reagent (Qiagen, Hilden, Germany), and kept at −70°C until further processing. RNA was extracted with RNeasy Universal Plus kit (Qiagen) and reverse transcrip-tion was performed with High Capacity cDNA Reverse Transcription kit (Applied Biosystems; Foster City, CA). Quantitative real-time PCR was performed using Gene Expression Master Mix (Applied Biosystems) on a 96-well plate (7900HT Fast RT-PCR system; Applied Biosystems). TaqMan assays used were for UCP-1: Mm01244861_m1; UCP-2: Mm00627599_m1; UCP-3: Mm00494077_m1; PGC-1: Mm01208835_m1; and GAPDH: Mm99999915_ g1 (all from Applied Biosystems).
2.6
|
Triglyceride quantification
Mice were killed by asphyxiation with CO2 6 hours after the
i.p. injection of LPS or β3 agonist. Interscapular BAT was
excised and immediately frozen on dry ice and kept at −70°C until further processing. Triglyceride quantification was car-ried out using the Triglyceride Quantification kit (Ab65336; Abcam, Cambridge, UK). Samples were homogenized in lysis buffer consisting of 5% NP40 (Sigma-Aldrich) in dH2O
and triglycerides were extracted according to the manufac-turer’s description.
2.7
|
BAT weight
Before RNA extraction or triglyceride quantification, both lobes of the excised BAT were weighed on a precision labo-ratory scale.
2.8
|
Statistics
Statistical analyses were performed in GraphPad PRISM version 6 (GraphPad Software, San Diego, CA). Data are presented as mean ± SEM. Group differences were ana-lyzed with t test or when three or more groups were involved with one-way or two-way ANOVA, followed by Tukey’s or Sidak’s post hoc tests, respectively, except in the cases of un-equal variance when non-parametric tests were used (Mann-Whitney and Kruskal-Wallis followed by the Dunn’s post hoc test for multiple comparisons).
3
|
RESULTS
3.1
|
UCP-1 is not necessary for
LPS-induced fever
We first examined the febrile response to immune chal-lenge with LPS in wild-type and UCP-1 knockout mice. Experiments were carried out at a near-thermoneutral tem-perature (29°C), that is, during conditions at which the ratio between mean energy expenditure and basal metabolic rate of mice (1.7-1.8) is similar to that found in humans.22 LPS
is a well-established exogenous pyrogen23 that has been
ex-tensively used for research on the mechanisms of fever.24,25
When injected i.p. (120 µg/kg) in wild-type mice, the follow-ing response is seen (Figure 1A): After an initial momentarily appearing hyperthermia that is elicited by the handling stress (and hence seen also in animals injected with vehicle) there is a rapid temperature fall, followed after 1-1½ hours of a rising body temperature that peaks at around 5-6 hours. As demonstrated in Figure 1A, this response was not affected
by UCP-1 deletion, implying that the activation of BAT was not necessary for the febrile response in this experimental setting. However, because the initial handling stress-induced hyperthermia may obscure early occurring febrile responses to LPS,24 we injected in the second set of experiments LPS
intravenously (i.v.; 30 µg/kg) through an indwelling venous catheter that had its free end outside the animal’s cage, thus permitting injections without handling the animal.19,24 As
described previously,24 this procedure results in a triphasic
fever (Figure 1B) with the first peak occurring at around 20 minutes postinjection (note in Figure 1B that this initial
peak is absent in mice injected with vehicle), a second peak at 1.5-2 hours, and a third phase of fever starting at around 3 hours. Similar to what was seen following the i.p. injection of LPS, i.v. administration resulted in the same febrile response in UCP-1 knockout mice as in wild-type mice (Figure 1B).
3.2
|
BAT thermogenesis is not activated
during LPS-induced fever
Next, we examined if LPS induced the expression of UCP-1 mRNA in interscapular BAT. We also examined the ex-pression of the peroxisome proliferator-activated receptor-γ coactivator-1α (PGC-1α), a transcription coactivator that in-teracts with a broad range of transcription factors involved in, for example, adaptive thermogenesis, mitochondrial biogen-esis, and glucose/fatty acid metabolism.26 PGC-1α induces
UCP-1 and has been shown to be necessary for cold toler-ance.27 As shown in Figure 2A,B, there was no effect of LPS
on the UCP-1 or PGC-1α mRNA expression. Neither was there any significant change in the weight of the BAT or its triglyceride content (Figure 2C,D). In contrast, the adminis-tration of a selective β3 receptor agonist (CL 316243; 0.5 mg/
kg), a procedure known to increase thermogenesis by BAT in both rodents and humans,3,28 resulted in prominent UCP-1
and PGC-1α expression in wild-type mice (UCP-1: Kruskal-Wallis H = 12.83, P = .0001; P = .0294 for WT NaCl vs WT agonist; PGC-1α: Kruskal-Wallis H = 12.13, P = .0003;
P = .0126 for WT NaCl vs WT agonist; Figure 2A,B) as well
as a reduction of BAT weight (Kruskal-Wallis H = 11.49,
P = .0009; P = .0029 for WT NaCl vs WT agonist) and a
tendency to reduce the triglyceride content (Kruskal-Wallis
H = 5.129, P =.0728; P =.0945 for WT NaCl vs WT agonist;
Figure 2C,D). It also elicited a temperature elevation in wild-type but not in UCP-1 knockout mice (One-way ANOVA
F3,27 = 3.012, P =.0474 for 60-180 minutes; WT agonist vs KO agonist: P <.05 for various time points; Figure 2E). As expected, the β3 agonist had no effect on BAT weight or
tri-glyceride content in UCP-1 knockout mice (data not shown).
3.3
|
No compensatory upregulation of
UCP-2 and UCP-3 in UCP-1 KO mice
To check that no compensatory expression of other un-coupling proteins could explain the intact febrile response to LPS in the UCP-1 knockout mice, we examined the ex-pression of UCP-2 and UCP-3 in BAT as well as in skel-etal (striated) muscle (rich in UCP-3). There was a tendency toward somewhat elevated levels of UCP-2 mRNA in BAT (Figure 3A), but not in skeletal muscle (Figure 3B) of UCP-1 knockout mice. An elevation of UCP-2 in UCP-1 knockout mice has been reported previously and has been suggested
FIGURE 1 The febrile response to immune challenge is not affected by UCP-1 deletion. A, Intraperitoneal injection of 120 µg/ kg LPS (at time point 0), at an ambient temperature of 29°C, resulted in similar fever in both UCP-1 knockout mice (KO) and wild-type (WT) mice. Note that both genotypes displayed initial stress-induced hyperthermia as the consequence of the handling during the i.p. injection, irrespective of whether LPS or saline was injected. B, Both genotypes developed a similar fever after i.v. injection of 30 µg/ kg LPS. Note that the i.v. injection, carried out without handling the animals, resulted in a triphasic fever,24 with the first phase appearing
around 20 minutes after injection. This phase is obscured by the handling stress-induced hyperthermia when animals are injected i.p. (cf. A). Solid lines represent mean and dotted lines show SEM
$
to reflect the increased amount of white fat in the BAT, in turn, caused by the lower usage of fat, when UCP-1 is de-leted.12 However, in neither BAT nor skeletal muscle, UCP-2
mRNA levels were regulated by LPS (Figure 3A,B). In con-trast, UCP-3 mRNA levels were induced by LPS; however, only in skeletal muscle (Two-way ANOVA F1,19 = 14.04, P =.0014; P =.0442 for WT NaCl vs WT LPS, and P =.0220
for KO NaCl vs KO LPS; Figure 3D) and not in BAT (Figure
FIGURE 2 Immune challenge with LPS does not activate brown adipose tissue (BAT), whereas prominent activation is seen after the administration of a β3 agonist. A, UCP-1 mRNA,
which was strongly induced by a β3 agonist (CL 316243, 0.5 mg/
kg i.p.), was unaffected by LPS (120 µg/kg i.p.). B, PGC-1 mRNA was not induced by LPS, but strongly induced by the β3 agonist.
C,D, LPS administration had no effect on BAT weight (C) or triglyceride (TGs) content (D), whereas the β3 agonist resulted in
significantly reduced BAT weight and a tendency toward reduced TGs. E, Injection of the β3 agonist to wild-type mice resulted in
hyperthermia. For the analyses in A-B, animals injected with NaCl and LPS were killed at 4 hours after injection, which is just prior to the peak of LPS fever (see Figure 1A), whereas animals injected with a β3 agonist were killed at 2 hours, corresponding to the time
point when mice treated with the agonist showed significantly increased body temperature (see E). For analyses in C-D, data are from mice killed 6 h after injection, a time point chosen to permit triglycerides to be consumed. Experiments were performed at an ambient temperature of 29ºC. Data are shown as mean ± SEM. * in E indicates P < .05 between WT mice given agonist and KO mice given agonist $ & ( % '
FIGURE 3 No difference in UCP-2 and UCP-3 expression in BAT or skeletal muscle between WT and UCP-1 KO mice, but an induction of UCP-3 in muscle by LPS and a β3 agonist. A,B, No
significant differences in UCP-2 expression in BAT (A) or skeletal muscles (B) between genotypes or treatments (LPS, 120 µg/kg i.p., or saline). C, UCP-3 expression in BAT did not differ between genotypes and was not affected by immune challenge with LPS. D, UCP-3 was induced in skeletal muscle by LPS, but there was no difference between genotypes. E, UCP-3 was induced in skeletal muscle by a β3 agonist (CL 316243, 0.1 mg/kg i.p.). F, No statistically significant
induction of UCP-2 in skeletal muscle by the β3 agonist. Data were
obtained 4 hours after the injection of LPS and 2 hours after the injection of the β3 agonist and are shown as mean ± SEM. Experiments
were performed at an ambient temperature of 29ºC $
&
( )
%
3C), and there was no difference between genotypes (Figure 3C,D). UCP-3 mRNA was also induced by the β3 agonist CL
316243 (0.1 mg/kg) (Kruskal-Wallis H = 6.838, P =.0251;
P =.0382 for WT agonist vs control; Figure 3E). There was
no statistically significant effect on UCP-2 mRNA levels by the β3 agonist (Figure 3F).
3.4
|
β
3receptor blockage attenuates
cold-induced thermogenesis but does not affect
LPS-induced fever
We next examined if the administration of a β3 antagonist that
has been shown to be able to block the adrenergic activation of UCP-129 and to inhibit BAT thermogenesis,30 could block
or attenuate LPS-induced fever. We found that the administra-tion of 10 µg/kg i.p. of SR59230A, a β3 antagonist that,
how-ever, also possesses α1 antagonist properties,31 had no effect
on the febrile response to i.p. injected LPS, although it attenu-ated the handling stress hyperthermia (Two-way ANOVA
F1,19 = 12.46, P = .0022 at peak hyperthermia; P = .0137 for antagonist + LPS vs saline + LPS; P = .1192 for antagonist + saline vs saline + saline; Figure 4A). We also administered repeated injections of the β3 antagonist (at 0 minutes, 120
min-utes, and 240 minutes after LPS injection), but this procedure did neither affect the fever response to LPS. Nor did adminis-tration of the β3 antagonist 3 hours after LPS injection, when
fever was present, affect body temperature (data not shown). To ascertain that the SR59230A in the present experimen-tal set up was able to affect thermogenesis, we transferred wild-type mice that either were treated with SR59230A (with the same dose as given in the fever experiment) or with vehicle to an indirect calorimetric system with an ambi-ent temperature that gradually fell to 7°C (Figure 4B2). We found that mice that had been given vehicle much better de-fended their body temperature than mice that had been given SR59230A, with the latter displaying a rapidly decreasing body temperature (Mann-Whitney P = .0196 at 45 minutes; Figure 4B1). The animals given SR59230A also showed a reduced induction of UCP-1 mRNA as well as PGC-1 mRNA after SR59230A injection (Unpaired t test P = .054 and P = .0011, respectively; Figure 4B3,B4), indicating a blockage of BAT activation.
3.5
|
Inhibition of peripheral
vasoconstriction abolishes LPS-induced fever
Because the data obtained imply that uncoupling protein-me-diated thermogenesis in BAT is not critical for inflammation-induced fever, we examined alternative mechanisms that could explain the fever seen after LPS injection. To this end, we ad-ministered to wild-type mice prazosin, an α1 and α2B antagonist
known to inhibit peripheral vasoconstriction, and to attenuate fever in rabbits.32-34 Mice given 1 or 2 mg/kg prazosin i.p.
together with saline showed a temperature response similar to mice that had been given vehicle and saline, demonstrat-ing that prazosin by itself had no effect on body temperature (Figure 5A). However, when given together with LPS, 2 mg/kg prazosin blocked the febrile response (Two-way ANOVA for 120-240 minutes: F3,20 = 3.133, P = .0484; prazosin + LPS vs
NaCl + LPS: P = .0418; Figure 5A), whereas 1 mg/kg resulted in an attenuated febrile response (data not shown).
To further examine the role of peripheral vasoconstriction for LPS-induced fever, we gave wild-type mice 8-hydroxy-2-(di-n-propylamino)tetralin hydrobromide (8-OH-DPAT), a 5-HT1A agonist which previously has been shown to reduce
ear pinna vasoconstriction and attenuate fever in LPS-treated rabbits.35 The administration of 8-OH-DPAT i.p. at a dose of
0.5 mg/kg given together with a saline injection resulted in a temperature response that was similar to that of mice only given saline, with the exception for an attenuated handling stress-induced hyperthermia (Figure 5B). However, when 8-OH-DPAT was administered together with LPS, a strong hypothermic response was seen, as well as a subsequent at-tenuation of the LPS-induced fever (Figure 5B).
3.6
|
LPS-induced fever does not require or
involve UCP-1-mediated thermogenesis at
sub-thermoneutral ambient temperature
The above data hence suggest that peripheral vasoconstric-tion and not UCP-1-mediated thermogenesis is critical for LPS-induced fever. However, the experiments were carried out during near-thermoneutral conditions (29°C) and may hence not reflect the mechanisms involved at lower ambi-ent temperature.7 We, therefore, also examined the febrile
response to i.p. injected LPS in wild-type and UCP-1 knock-out mice at an ambient temperature of 21°C, but there was neither any difference in the febrile response between geno-types at this temperature (Figure 6A). Furthermore, there was no induction of UCP-1 or PGC-1α mRNA in LPS-treated wild-type mice as compared to the levels observed in saline-treated mice (Figure 6B,C) and there was no reduction in the weight of BAT (Figure 6D).
3.7
|
Emotional stress-induced
hyperthermia does not require or activate
UCP-1-mediated thermogenesis but is also
independent of peripheral vasoconstriction
In addition to the observations obtained in the experiments in which animals were briefly restrained to permit intra-peritoneal injections (and which showed no difference in
the hyperthermic response between wild-type and UCP-1 knockout mice neither at an ambient temperature of 29°C nor at normal room temperature of 21°C; Figures 1 and 6), we further examined the role of UCP-1-mediated thermogen-esis for emotional stress-induced hyperthermia using a cage exchange paradigm. Wild-type and UCP-1 knockout mice showed a similar rapid temperature increase when placed in another animal’s cage, as well as a similar less pronounced response when lifted up and placed back in their own home
cage irrespective of whether this procedure was performed at 29°C or 21°C (Figure 7A,B). The only difference noted was that the body temperature of the mice normalized faster when the ambient temperature was 21°C than when it was 29°C. Examination in wild-type mice of UCP-1 and PGC-1α mRNA expression at 3 hours, a time point at which body temperature was still elevated in the cage exchange group but normal in the control group, showed no difference be-tween the two groups when experiments were carried out
FIGURE 4 β3 receptor blockage does not affect immune-induced fever but attenuates cold-induced thermogenesis. A, LPS-induced fever
was similar in mice given saline or a β3 antagonist (SR 59230A; 10 µg/kg i.p. injected immediately prior to the LPS injection [120 µg/kg i.p.]). Note
that the β3 antagonist attenuated the initial handling stress-induced hyperthermia. Experiments were performed at an ambient temperature of 29ºC. *
indicates P < .05 for β3 antagonist + LPS versus NaCl + LPS at peak of hyperthermia. B, The administration of the β3 antagonist to mice subjected
to cold (B2) caused a rapid temperature fall (B1) and attenuated expression of UCP-1 mRNA (B3) and PGC-1α mRNA (B4), as measured 45 minutes after injection and cold exposure. Basal levels of UCP-1 and PGC-1α mRNA 45 minutes after saline injection at near thermoneutral conditions (29°C) are given as comparison. Data are shown as mean ± SEM
$
at 29°C (Figure 7C,D), and nor did the analysis of BAT weight (Figure 7E). In contrast, injection of a β3 agonist to
naïve wild-type mice showed strong induction, as meas-ured at 2 hours postinjection, of both transcripts (UCP-1: Kruskal-Wallis H = 7.714, P = .013; P = .0365 for control vs β3 agonist; PGC-1α: Kruskal-Wallis H = 8.189, P = .005; P = .0129 for control vs β3 agonist) and reduced BAT weight
(Kruskal-Wallis H = 8.214, P = .0062; P = .0176 for control vs β3 agonist; Figure 7C-E).
When cage exchanged was performed at an ambient tem-perature of 21°C, there was neither any indication of BAT activation (Figure 7F-H), with BAT weight, in fact, being
somewhat higher and UCP-1 and PGC-1α mRNA levels somewhat lower than in unstimulated control mice, although the differences did not reach statistical significance. A β3
ago-nist given to naïve mice held at 21°C yielded a small, but sta-tistically significant increase in UCP-1 mRNA levels (Figure 7F; One-way ANOVA F2,19 = 13.73; P = .0002, P = .0061
for control vs β3 agonist), as well as strongly induced PGC-1α
mRNA (Figure 7G; Kruskal-Wallis H = 15.99, P < .0001;
P = .0340 for control vs β3 agonist). BAT weight was reduced
but the difference was not statistically significant (Figure 7H; Kruskal-Wallis H = 11.97, P = .0006; P = .1059 for control vs β3 agonist). However, BAT weight was significantly lower
in control mice held at 21°C than in control mice held at 29°C (73 ± 5 vs 96 ± 7 g; Unpaired t test P = .026) and compari-son of ΔCt values for UCP-1 mRNA (vs GAPDH mRNA) at
21°C and 29°C showed that the basal levels (in unstimulated mice) of UCP-1 mRNA in BAT were 20 times higher when mice were held at the lower ambient temperature, being at the same level as mice kept at 29°C and stimulated with the β3 agonist, hence indicating that the mice held at 21°C
expe-rienced cold stress.
The administration of an α1/α2B antagonist (prazosin) did
not affect the hyperthermic response to cage exchange (data not shown), being consistent with its absence of effect on the handling stress-induced hyperthermia elicited during the in-traperitoneal injections (see Figure 5A).
3.8
|
Emotional stress-induced
hyperthermia is associated with increased
expression of UCP-3 in skeletal muscle
Finally, we examined if cage exchange stress influenced UCP expression in skeletal muscles. Analysis of UCP-3 mRNA in skeletal muscle of wild-type mice 2 hours after cage ex-change showed increased expression when compared to naïve mice (1.7 times; t test: P = .0082, n = 8 and 6). There was also a similar difference at 3 hours after cage exchange, but it was not statistically significant. The levels of UCP-2 mRNA in skeletal muscle did not differ between animals subjected to cage exchange and naïve controls, whereas UCP-1 lev-els were increased (by 2.5 times). However, the amount of UCP-1 mRNA in muscle was very small, being in the order of 10 000 times lower than the amount of UCP-3 mRNA, and hence likely of little functional significance.
4
|
DISCUSSION
The findings of the present study, performed in mice, show that UCP-1-dependent thermogenesis is not critical for pyrogen-induced fever or emotional stress-induced hyper-thermia. Mice with deletion of UCP-1 displayed a normal
FIGURE 5 Blockage of peripheral vasoconstriction abolishes immune-induced fever. A, Injection of the α1 and α2B antagonist
prazosin (2 mg/kg i.p.) immediately before the injection of LPS (120 µg/kg i.p.) inhibited the fever response for about 4 hours but did not affect body temperature when injected together with saline. B, Injection of the 5-HT1A agonist 8-OH DPAT (0.5 mg/kg i.p.)
immediately before the injection of LPS (120 µg/kg i.p.) resulted in a hypothermic response and subsequently attenuated fever throughout most of the observation period. Mice given 8-OH DPAT together with saline showed similar body temperature to mice injected with saline only, except for attenuated handling stress-induced hyperthermia. Experiments were performed at an ambient temperature of 29ºC
$
febrile response to the peripheral administration of lipopoly-saccharide, both at near thermoneutral conditions and at an ambient temperature of 21°C. They also showed a normal emotional stress-induced hyperthermia, both when handled for performing intraperitoneal injections, and in a cage ex-change paradigm. Furthermore, there was neither any differ-ence when LPS was given i.v. instead of i.p. Hdiffer-ence, also the first phase of fever, only discerned when animals are immune challenged during conditions not involving any handling stress,24 was found to be UCP-1 independent.
While it could be argued that UCP-1 knockout mice may have developed compensatory mechanisms for fever and emo-tional stress-induced hyperthermia, there was neither any evi-dence in wild-type mice that LPS injection or cage exchange elicited any activation of BAT. Thus, there were no UCP-1 or PGC-1α mRNA upregulation, and BAT weight and triglycer-ide content were unaffected (see also ref.15), which was in stark
contrast to what was seen when a β3 agonist, known to activate
BAT, was administered. Furthermore, the administration of a β3 antagonist had no effect on LPS-induced fever but made
the mice much more sensitive to a cold environment, being in line with the role of BAT for cold-induced thermogenesis.36
However, an α1/ α2a antagonist that blocks peripheral
vaso-constriction abolished LPS-induced fever, suggesting that heat conservation, but not UCP-1 activation, is a critical mechanism behind pyrogen-induced fever, at least in mice. Furthermore, Okamatsu-Ogura et al15 found no activity independent increase
in oxygen consumption after immune challenge, whereas the administration of a β3 agonist significantly increased oxygen
consumption. Taken together, these findings strongly suggest that immune-induced fever does not involve BAT activation.
The above conclusion is clearly at odds with a large, and widely disseminated literature in the field, claiming that fever, and also emotional stress-induced hyperthermia, is
FIGURE 6 The febrile response to immune challenge at normal room temperature (21°C) is not affected by UCP-1 deletion and is not associated with induced UCP-1 or PGC-1α mRNA expression or BAT weight reduction. A, UCP-1 KO mice and WT mice displayed similar fever after injection of LPS (120 µg/kg i.p.). B-D, Injection of LPS (120 µg/kg i.p.) did not affect the mRNA levels of UCP-1 (B) or PGC-1 (C) or the weight of the BAT (D), as determined 4 hours postinjection. Data are shown as mean ± SEM
$
&
FIGURE 7 Emotional stress-induced hyperthermia is not dependent on UCP-1. A,B, UCP-1 KO and WT mice subjected to cage exchange at an ambient temperature of 29ºC (A) or 21ºC (B) displayed prominent hyperthermia, whereas control mice that were taken up and placed back in their home-cage showed a weaker temperature response, but neither response differed between genotypes. C-E, UCP-1 (C) and PGC-1α mRNA levels (D) at 3 hours post-handling in wild-type mice kept at 29ºC did not differ between treatments, and nor did BAT weight (E). In contrast, injection of a β3 agonist (CL 316243; 0.5 mg/kg i.p.) resulted in strong expression of both transcript as well as reduced BAT weight compared to
what was seen in the control mice. F-H, UCP-1 (F) and PGC-1α mRNA levels (G) at 3 hours post-handling in wild-type mice kept at 21ºC did not display any statistically significant differences between treatments, and nor did BAT weight (H). Injection of a β3 agonist (CL 316243; 0.5 mg/
kg i.p.) resulted in only a small increase of UCP-1 mRNA (F) and a prominent upregulation of PGC-1α mRNA (G), but no statistically significant difference in BAT weight (H)
$
& (
%
'
mediated by BAT thermogenesis.8-10,37-39 However, it should
be recognized that this idea is based on studies in rats and not mice. While species differences hence provide a possible explanation for the difference between the present findings and those reported from studies in rats, it should nevertheless be pointed out that the evidence for an involvement of BAT in fever and emotional stress-induced hyperthermia in rats largely is indirect, and that in rats, no true functional studies of these phenomena, i.e. studies in which BAT thermogenesis has been blocked, do exist.40 Furthermore, there is meagre
evidence in rats for the activation of BAT during inflam-mation-induced fever, such as UCP-1 induction, or the con-sumption of BAT or its triglycerides,41 signs that are robustly
elicited by adrenergic β3 stimulation or cold exposure.28,36,42
It should also be noted that critical evidence for BAT activa-tion during febrile condiactiva-tions relies on anesthetized rat prepa-rations, that is, a model in which normal thermoregulation is compromised.43 In such a model, prostaglandin E
2 (PGE2)
was injected in the preoptic area of the hypothalamus and increased sympathetic innervation of BAT was recorded from nerve bundles from the ventral surface of the BAT pads.10
Hence, this condition cannot be considered as physiological fever, which in addition to central PGE2 synthesis also
in-volves the peripheral and central release of various cytokines. Moreover, since the nerve trunks that during their passage through the BAT pads give off fibers to the BAT are inter-costal nerves,44 increased activity in these nerve trunks does
not necessarily reflect sympathetic BAT activation but may instead indicate, for example, increased vasoconstriction in the innervated territory of these nerves.
Other evidence for an involvement of BAT in fever, as well as in emotional stress-induced hyperthermia, is the ob-servation that an increase in BAT temperature precedes the increase in whole body temperature.6,9 While this
observa-tion is intriguing, its interpretaobserva-tion as evidence for BAT ther-mogenesis has been questioned: It has been suggested that increased temperature in the interscapular area could arise from other sources than BAT such as the activation of neck and shoulder muscles and increased blood flow to these areas during emotional stress45 (see also ref. 46 for the account of
various technical problems associated with BAT temperature measurements). Shunting of blood from, for example, vis-cera to axial muscles during immune as well as emotional stress47-49 could thus explain the temporal difference between
the observed temperature rise over the neck and shoulder re-gion and that of core body temperature measured by abdom-inally implanted probes.
In the present study, we found vasoconstriction to be a crit-ical mechanism for immune-induced fever. Indeed, fever has been shown to be associated with vasoconstriction in both rats and mice, with each phase of fever, as monitored during ther-moneutrality, closely matching reduced skin heat loss.20,24
However, it has been suggested that there are different
thresholds for the activation of vasoconstriction and BAT thermogenesis for the generation of fever. This idea is in line with the presence of different thresholds for vasoconstriction and shivering thermogenesis during cold defense, with the former being much more sensitive to a reduced body tempera-ture than the latter.43 Thus, Szekely and Szekely7 reported,
again in rats, that in a warm environment, vasoconstriction alone was sufficient to produce fever, whereas at lower tem-peratures, both vasoconstriction and BAT thermogenesis were in play. Such a differential activation of the different modes of non-shivering thermogenesis is in line with the anatomi-cal segregation of the preoptic pathways activating presym-pathetic thermoregulatory neurons in the rostral medullary raphe nucleus: Whereas the projection from prostaglandin E2 sensitive preoptic neurons to the rostral medullary raphe
nucleus that elicits vasoconstriction is a direct one, the pro-jections that elicit activation of BAT and shivering thermo-genesis include a relay in the dorsomedial hypothalamus.50,51
In the present study, we examined the febrile response (and BAT activation) of wild-type and UCP-1 knockout mice both at near-thermoneutrality (29°C) and at lower ambient temperature (21°C). At neither temperature, the febrile re-sponse was dependent on UCP-1 and similar to what was seen at near-thermoneutrality, there were no signs of activation of BAT in wild-type mice at the lower ambient temperature. Thus, within the examined temperature interval, BAT ther-mogenesis did not appear to contribute to the febrile response, but it cannot be ruled out that at ambient temperatures lower than 21°C, immune challenge may elicit BAT thermogenesis. However, at low ambient temperature, it is difficult to elicit sustained fever in rodents; instead, rats and mice show hy-pothermia, especially after high LPS doses.20,52-55
Immune-induced hypothermia, which is also seen in humans with sepsis,56 has been shown to be biologically adaptive, being
the preferred response during unfavorable metabolic condi-tions.57 It is elicited by vasodilation and augmented by cold
seeking behavior that is dependent on a neuronal circuit in the dorsomedial hypothalamic nucleus.58 It hence appears that
vasoregulation is the main tool by which the rodents respond to inflammatory challenge, by reducing heat loss in a warm environment and promoting heat loss in a cold environment.
Nevertheless, vasoconstriction is unlikely to result in and maintain fever on its own, especially in a sub-thermoneu-tral ambient temperature, in which additional thermogen-esis would be needed to increase the body temperature. A possible candidate is fatty acid-mediated proton transport by UCP-3, expressed in skeletal muscle.59 Consistent with
pre-vious reports,60-65 we found here that LPS, as well as a β 3
agonist, induced UCP-3 mRNA in skeletal muscle (but not in BAT). Furthermore, it has also been reported that a genetic deletion of UCP-3 abolishes LPS-induced fever and that its selective reexpression in skeletal muscle partly rescues this response.17 Taken together, the available data hence indicate
that immune-induced fever is dependent both on heat pres-ervation by vasoconstriction and on UCP-3-mediated heat generation in skeletal muscles, and probably also shivering thermogenesis. However, UCP-1-mediated thermogenesis in BAT is not involved, at least not in mice.
There was neither any evidence in the present study that emotional stress-induced hyperthermia was dependent on BAT thermogenesis. The immobilization stress associated with intraperitoneal injections resulted in similar hyperther-mia in UCP-1 knockout mice as in wild-type mice, both at near-thermoneutral temperature (29°C) and at normal room temperature (21°C). Furthermore, there were no signs of BAT activation following cage exchange stress, with mRNA levels of UCP-1 and PGC-1α being similar to those seen in control mice and much lower than in mice treated with a β3 agonist
when animals were kept at near-thermoneutral temperature. When the cage exchange experiment was performed at an ambient temperature of 21°C, there were neither any signs of BAT activation. However, data showed that mice exposed to this ambient temperature experienced cold stress, as evi-denced by the much higher basal level of UCP-1 mRNA and lower BAT weight. The fact the BAT already was strongly activated may explain the small effect of the β3 agonist on
the UCP-1 mRNA expression. Indeed, basal UCP-1 mRNA levels in BAT of mice exposed to 21°C were as high as those seen in mice kept at 29°C and stimulated with the β3 agonist,
suggesting a close to maximal BAT activation at the lower ambient temperature. It is noteworthy that BAT weight was higher and UCP-1 and PGC-1α mRNA levels were lower in the cage exchange mice than in the control mice. While these differences were not statistically significant, they were con-sistent across analyses. Although speculative, it is tempting to suggest that when body temperature increases during emo-tional stress, and does so by BAT-independent thermogen-esis, the cold-induced BAT thermogenesis is turned down.
While the present data hence imply that BAT is not neces-sary nor activated during emotional stress, it remains unclear how the rapidly increased temperature is generated. Consistent with previous reports on emotional stress-induced hyperther-mia in rats and mice,66,67 the administration of a β
3 antagonist
attenuated the handling stress-induced hyperthermia seen in association with intraperitoneal injection (Figure 4A), sug-gesting the presence of a β3 adrenergic mechanisms, although
it should be noted that the β3 antagonist used (SR59230A) also
possesses α1 antagonist properties.31 However, when prazosin,
which is an α1 and α2B antagonist and which robustly
inhib-ited LPS-induced fever, was given there was no effect on the handling stress-induced hyperthermia. Moreover, the admin-istration of the centrally acting 5HT1A agonist 8-OH-DPAT35
resulted in attenuated handling stress-induced hyperthermia. It is conceivable that different sets of central presympathetic vasoconstrictors are activated by emotional stress and immune stress, respectively, and that they involve distinct peripheral
mechanisms. This idea is supported by a recent study by Machado et al, showing that the activation of glutamatergic neurons in the dorsal hypothalamic area drove an increase in body temperature without causing vasoconstriction, hence suggesting a neural substrate for the hot feeling and flushing phenomenon experienced during certain emotional states.68
Along the same line, Marks et al, also reported that the hyper-thermic response of fear was not due to skin vasoconstriction but due to sympathetically, adrenoceptor-mediated thermo-genesis, that, however, did not involve BAT.45
Since the β3 agonist induced UCP-3 mRNA, adrenergic
UCP-3-dependent thermogenesis in muscle could play a crit-ical role in emotional stress-induced hyperthermia. This idea, which is supported by our finding that cage exchange stress induced UCP-3 mRNA in skeletal muscle, fits well into the observations reported above of activation of neck and shoul-der muscles and increased blood flow to these areas during emotional stress.45 Emotional stress, such as restraint stress
and cage exchange stress, evokes a fight-flight response during which increased muscle metabolism would seem bio-logically adaptive.
It has long been known that human embryos and infants display BAT,69 but it was only recently that BAT was
un-equivocally demonstrated also in adult humans.4,5,70 It seems
to be present in most, if not all individuals, and has been shown to take part in cold-induced thermogenesis.71 To the
best of our knowledge, no studies have been performed as to the role of BAT for the febrile response in humans. Fever is a hallmark of infection and inflammation. The elevation of the body temperature following an immune challenge is a stereotypic response seen in all vertebrates, including poiki-lotherms, which have been shown to prefer a warmer environ-ment when infected.72 Because fever is so well preserved, one
would expect that in homeotherms, such as mammals, similar thermogenic mechanisms should be at play across species. The present data indicate that BAT thermogenesis is not in-volved in fever (and stress-induced hyperthermia) in mice [of note, a similar conclusion was recently reached for thy-roid-induced thermogenesis73] and calls for further
examina-tion of this quesexamina-tion in other species, such as rats and humans.
ACKNOWLEDGMENTS
Supported by grants from the Swedish Research Council (#2016-01301) and Swedish Brain Foundation (#FO2019-0033). We thank Drs. Barbara Cannon and Jan Nedergaard for providing UCP-1 KO mice.
CONFLICT OF INTEREST
The authors declare no conflict of interest.
AUTHOR CONTRIBUTIONS
A. Eskilsson and A. Blomqvist designed the research; A. Eskilsson and K. Shionoya performed the research; S.
Enerbäck contributed new reagents or analytic tools; A. Eskilsson, D. Engblom, and A. Blomqvist analyzed the data; A. Eskilsson and A. Blomqvist wrote the paper.
REFERENCES
1. Cannon B, Nedergaard J. Brown adipose tissue: function and phys-iological significance. Physiol Rev. 2004;84:277-359.
2. Betz MJ, Enerbäck S. Human brown adipose tissue: what we have learned so far. Diabetes. 2015;64:2352-2360.
3. Cypess AM, Weiner LS, Roberts-Toler C, et al. Activation of human brown adipose tissue by a beta3-adrenergic receptor ago-nist. Cell Metab. 2015;21:33-38.
4. van Marken Lichtenbelt WD, Vanhommerig JW, Smulders NM, et al. Cold-activated brown adipose tissue in healthy men. New Engl
J Med. 2009;360:1500-1508.
5. Virtanen KA, Lidell ME, Orava J, et al. Functional brown adipose tissue in healthy adults. New Engl J Med. 2009;360:1518-1525. 6. Szekely M, Szelenyi Z, Sumegi I. Brown adipose tissue as a source
of heat during pyrogen-induced fever. Acta Physiol Acad Sci Hung. 1973;43:85-88.
7. Szekely M, Szelenyi Z. Endotoxin fever in the rat. Acta Physiol
Acad Sci Hung. 1979;53:265-277.
8. Nakamura K. Central circuitries for body temperature reg-ulation and fever. Am J Physiol Regul Integr Comp Physiol. 2011;301:R1207-R1228.
9. Kataoka N, Hioki H, Kaneko T, Nakamura K. Psychological stress activates a dorsomedial hypothalamus-medullary raphe circuit driving brown adipose tissue thermogenesis and hyperthermia.
Cell Metab. 2014;20:346-358.
10. Nakamura Y, Nakamura K, Matsumura K, Kobayashi S, Kaneko T, Morrison SF. Direct pyrogenic input from prostaglandin EP3 receptor-expressing preoptic neurons to the dorsomedial hypothal-amus. Eur J Neurosci. 2005;22:3137-3146.
11. Nicholls DG, Locke RM. Thermogenic mechanisms in brown fat.
Physiol Rev. 1984;64:1-64.
12. Enerback S, Jacobsson A, Simpson EM, et al. Mice lacking mi-tochondrial uncoupling protein are cold-sensitive but not obese.
Nature. 1997;387:90-94.
13. Golozoubova V, Hohtola E, Matthias A, Jacobsson A, Cannon B, Nedergaard J. Only UCP1 can mediate adaptive nonshivering ther-mogenesis in the cold. FASEB J. 2001;15:2048-2050.
14. Golozoubova V, Cannon B, Nedergaard J. UCP1 is essential for adaptive adrenergic nonshivering thermogenesis. Am J Physiol
Endocrinol Metab. 2006;291:E350-E357.
15. Okamatsu-Ogura Y, Kitao N, Kimura K, Saito M. Brown fat UCP1 is not involved in the febrile and thermogenic responses to IL-1β in mice. Am J Physiol Endocrinol Metab. 2007;292:E1135-E1139. 16. Szentirmai É, Kapás L. Brown adipose tissue plays a central role
in systemic inflammation-induced sleep responses. PLoS ONE. 2018;13:e0197409-e0197409.
17. Riley CL, Dao C, Kenaston MA, et al. The complementary and divergent roles of uncoupling proteins 1 and 3 in thermoregulation.
J Physiol. 2016;594:7455-7464.
18. Hesse DG, Tracey KJ, Fong Y, et al. Cytokine appearance in human endotoxemia and primate bacteremia. Surg Gynecol Obstet. 1988;166:147-153.
19. Engström L, Ruud J, Eskilsson A, et al. Lipopolysaccharide-induced fever depends on prostaglandin E2 production specifically in brain endothelial cells. Endocrinology. 2012;153:4849-4861.
20. Rudaya AY, Steiner AA, Robbins JR, Dragic AS, Romanovsky AA. Thermoregulatory responses to lipopolysaccharide in the mouse: dependence on the dose and ambient temperature. Am J
Physiol Regul Integr Comp Physiol. 2005;289:R1244-R1252.
21. Saha S, Engstrom L, Mackerlova L, Jakobsson PJ, Blomqvist A. Impaired febrile responses to immune challenge in mice deficient in microsomal prostaglandin E synthase-1. Am J Physiol Regul
Integr Comp Physiol. 2005;288:R1100-R1107.
22. Fischer AW, Cannon B, Nedergaard J. The answer to the question “What is the best housing temperature to translate mouse experi-ments to humans?” is: thermoneutrality. Mol Metab. 2019;26:1-3. 23. Kozak W, Conn CA, Kluger MJ. Lipopolysaccharide induces
fever and depresses locomotor activity in unrestrained mice. Am J
Physiol. 1994;266:R125-R135.
24. Romanovsky AA, Kulchitsky VA, Simons CT, Sugimoto N. Methodology of fever research: why are polyphasic fevers often thought to be biphasic? Am J Physiol. 1998;275:R332-R338. 25. Eskilsson A, Mirrasekhian E, Dufour S, Schwaninger M, Engblom
D, Blomqvist A. Immune-induced fever is mediated by IL-6 re-ceptors on brain endothelial cells coupled to STAT3-dependent induction of brain endothelial prostaglandin synthesis. J Neurosci. 2014;34:15957-15961.
26. Liang H, Ward WF. PGC-1alpha: a key regulator of energy metab-olism. Adv Physiol Educ. 2006;30:145-151.
27. Lin J, Wu PH, Tarr PT, et al. Defects in adaptive energy metabo-lism with CNS-linked hyperactivity in PGC-1alpha null mice. Cell. 2004;119:121-135.
28. Lowell BB, Flier JS. Brown adipose tissue, beta 3-adrenergic re-ceptors, and obesity. Annu Rev Med. 1997;48:307-316.
29. Nisoli E, Tonello C, Landi M, Carruba MO. Functional studies of the first selective beta 3-adrenergic receptor antagonist SR 59230A in rat brown adipocytes. Mol Pharmacol. 1996;49:7-14.
30. Ootsuka Y, Kulasekara K, de Menezes RC, Blessing WW. SR59230A, a beta-3 adrenoceptor antagonist, inhibits ultradian brown adipose tissue thermogenesis and interrupts associated ep-isodic brain and body heating. Am J Physiol Regul Integr Comp
Physiol. 2011;301:R987-R994.
31. Bexis S, Docherty JR. Role of α1- and β3-adrenoceptors in the modulation by SR59230A of the effects of MDMA on body tem-perature in the mouse. Br J Pharmacol. 2009;158:259-266. 32. Langer SZ, Hicks PE. Alpha-adrenoreceptor subtypes in blood
vessels: physiology and pharmacology. J Cardiovasc Pharmacol. 1984;6(suppl 4):S547-S558.
33. Bencsics A, Elenkov IJ, Vizi ES. α2−, α2A−, α2B/2C-Adrenocep-tor subtype antagonists prevent lipopolysaccharide-induced fever response in rabbits. Brain Res. 1995;705:302-306.
34. Szreder Z. Antipyresis as a result of alpha-1 adrenoceptor block-ade. Gen Pharmacol. 1993;24:455-460.
35. Blessing WW. 5-hydroxytryptamine 1A receptor activation reduces cutaneous vasoconstriction and fever associated with the acute in-flammatory response in rabbits. Neuroscience. 2004;123:1-4. 36. Nedergaard J, Cannon B. Chapter 9 – Brown adipose tissue as a
heat-producing thermoeffector. In: AA Romanovsky, ed. Handbook
of Clinical Neurology. Vol. 156. Elsevier, Amsterdam; 2018:137-152.
37. Morrison SF. Central neural control of thermoregulation and brown adipose tissue. Auton Neurosci. 2016;196:14-24.
38. Morrison SF. Chapter 17 – Efferent neural pathways for the con-trol of brown adipose tissue thermogenesis and shivering. In:AA Romanovsky, ed. Handbook of Clinical Neurology. Vol. 156. Elsevier, Amsterdam; 2018:281-303.
39. Morrison SF, Nakamura K. Central mechanisms for thermoregula-tion. Annu Rev Physiol. 2019;81:285-308.
40. Blomqvist A, Engblom D. Neural mechanisms of inflammation-in-duced fever. Neuroscientist. 2018;24:381-399.
41. Jennings G, Elia M. Effect of E. coli endotoxin on temperature, oxygen consumption and brown adipose tissue thermogenesis in rats and mice. Biosci Rep. 1987;7:517-523.
42. Nagase I, Yoshida T, Kumamoto K, et al. Expression of un-coupling protein in skeletal muscle and white fat of obese mice treated with thermogenic beta 3-adrenergic agonist. J Clin Invest. 1996;97:2898-2904.
43. Sessler DI. Perioperative thermoregulation and heat balance.
Lancet. 2016;387:2655-2664.
44. Foster DO, Depocas F, Zuker M. Heterogeneity of the sympathetic innervation of rat interscapular brown adipose tissue via intercostal nerves. Can J Physiol Pharmacol. 1982;60:747-754.
45. Marks A, Vianna DML, Carrive P. Nonshivering thermogenesis without interscapular brown adipose tissue involvement during conditioned fear in the rat. Am J Physiol Regul Integr Comp
Physiol. 2009;296:R1239-R1247.
46. Meyer CW, Ootsuka Y, Romanovsky AA. Body temperature measurements for metabolic phenotyping in mice. Front Physiol. 2017;8:520.
47. Joyner MJ, Casey DP. Regulation of increased blood flow (hyper-emia) to muscles during exercise: a hierarchy of competing physi-ological needs. Physiol Rev. 2015;95:549-601.
48. Archie JP Jr. Anatomic arterial-venous shunting in endotoxic and septic shock in dogs. Ann Surg. 1977;186:171-176.
49. Navaratnam RL, Morris SE, Traber DL, et al. Endotoxin (LPS) in-creases mesenteric vascular resistance (MVR) and bacterial translo-cation (BT). J Trauma. 1990;30:1104-1113; discussion 1113-1105. 50. Rathner JA, Madden CJ, Morrison SF. Central pathway for spon-taneous and prostaglandin E2-evoked cuspon-taneous vasoconstriction.
Am J Physiol Regul Integr Comp Physiol. 2008;295:R343-R354.
51. Nakamura K, Morrison SF. Central efferent pathways for cold-de-fensive and febrile shivering. J Physiol. 2011;589:3641-3658. 52. Oka T, Oka K, Kobayashi T, et al. Characteristics of
thermoregu-latory and febrile responses in mice deficient in prostaglandin EP1 and EP3 receptors. J Physiol. 2003;551:945-954.
53. Romanovsky AA, Shido O, Sakurada S, Sugimoto N, Nagasaka T. Endotoxin shock: thermoregulatory mechanisms. Am J Physiol
Regul Integr Comp Physiol. 1996;270:R693-R703.
54. Steiner AA, Chakravarty S, Robbins JR, et al. Thermoregulatory re-sponses of rats to conventional preparations of lipopolysaccharide are caused by lipopolysaccharide per se- not by lipoprotein contaminants.
Am J Physiol Regul Integr Comp Physiol. 2005;289:R348-R352.
55. Steiner AA, Dogan MD, Ivanov AI, et al. A new function of the leptin receptor: mediation of the recovery from lipopolysaccha-ride-induced hypothermia. FASEB J. 2004;18:1949-1951. 56. Clemmer TP, Fisher CJJ, Bone RC, Slotman GJ, Metz CA, Thomas
FO. Hypothermia in the sepsis syndrome and clinical outcome.
Crit Care Med. 1992;20:1395-1401.
57. Liu E, Lewis K, Al-Saffar H, et al. Naturally occurring hypo-thermia is more advantageous than fever in severe forms of lipopolysaccharide- and Escherichia coli-induced systemic in-flammation. Am J Physiol Regul Integr Comp Physiol. 2012;302: R1372-R1383.
58. Wanner SP, Almeida MC, Shimansky YP, et al. Cold-induced thermogenesis and inflammation-associated cold-seeking behav-ior are represented by different dorsomedial hypothalamic sites: a
three-dimensional functional topography study in conscious rats.
J Neurosci. 2017;37:6956-6971.
59. Pohl EE, Rupprecht A, Macher G, Hilse KE. Important trends in UCP3 investigation. Front Physiol. 2019;10:470.
60. Yu XX, Barger JL, Boyer BB, Brand MD, Pan G, Adams SH. Impact of endotoxin on UCP homolog mRNA abundance, thermo-regulation, and mitochondrial proton leak kinetics. Am J Physiol
Endocrinol Metab. 2000;279:E433-E446.
61. Sun X, Wray C, Tian X, Hasselgren P-O, Lu J. Expression of un-coupling protein 3 is upregulated in skeletal muscle during sepsis.
Am J Physiol Endocrinol Metab. 2003;285:E512-E520.
62. Harada N, Kusuyama A, Morishima M, Okada K, Takahashi A, Nakaya Y. Bezafibrate improves bacterial lipopolysaccharide-induced dyslipidemia and anorexia in rats. Metabolism. 2007;56:517-522. 63. Frisard MI, Wu Y, McMillan RP, et al. Low levels of
lipopolysac-charide modulate mitochondrial oxygen consumption in skeletal muscle. Metabolism. 2015;64:416-427.
64. Nakamura Y, Nagase I, Asano A, et al. β 3-adrenergic agonist up-regulates uncoupling proteins 2 and 3 in skeletal muscle of the mouse. J Vet Med Sci. 2001;63:309-314.
65. Boss O, Bachman E, Vidal-Puig A, Zhang C-Y, Peroni O, Lowell BB. Role of the β3-adrenergic receptor and/or a putative β4-adren-ergic receptor on the expression of uncoupling proteins and per-oxisome proliferator-activated receptor-γ coactivator-1. Biochem
Biophys Res Commun. 1999;261:870-876.
66. Lkhagvasuren B, Nakamura Y, Oka T, Sudo N, Nakamura K. Social defeat stress induces hyperthermia through activation of thermoregulatory sympathetic premotor neurons in the medullary raphe region. Eur J Neurosci. 2011;34:1442-1452.
67. Miyamoto T, Funakami Y, Kawashita E, et al. Repeated cold stress enhances the acute restraint stress-induced hyperthermia in mice.
Biol Pharm Bull. 2017;40:11-16.
68. Machado NLS, Abbott SBG, Resch JM, et al. A glutamatergic hypothalamomedullary circuit mediates thermogenesis, but not heat conservation, during stress-induced hyperthermia. Curr Biol. 2018;28:2291-2301.
69. Enerbäck S. Human brown adipose tissue. Cell Metab. 2010;11: 248-252.
70. Cypess AM, Lehman S, Williams G, et al. Identification and im-portance of brown adipose tissue in adult humans. New Engl J
Med. 2009;360:1509-1517.
71. Ouellet V, Labbé SM, Blondin DP, et al. Brown adipose tissue ox-idative metabolism contributes to energy expenditure during acute cold exposure in humans. J Clin Invest. 2012;122:545-552. 72. Boltaña S, Rey S, Roher N, et al. Behavioural fever is a
syner-gic signal amplifying the innate immune response. Proc R Soc B. 2013;280:20131381.
73. Dittner C, Lindsund E, Cannon B, Nedergaard J. At thermoneutral-ity, acute thyroxine-induced thermogenesis and pyrexia are inde-pendent of UCP1. Mol Metab. 2019;25:20-34.
How to cite this article: Eskilsson A, Shionoya K,
Enerbäck S, Engblom D, Blomqvist A. The generation of immune-induced fever and emotional stress-induced hyperthermia in mice does not involve brown adipose tissue thermogenesis. The FASEB Journal. 2020;34:
