A modest change in housing temperature alters whole body energy expenditure and adipocyte thermogenic capacity in mice

Abstract

Typical vivarium temperatures (20–26°C) induce facultative thermogenesis in mice, a process attributable in part to uncoupling protein-1 (UCP1). The impact of modest changes in housing temperature on whole body and adipose tissue energetics in mice remains unclear. Here, we determined the effects of transitioning mice from 24°C to 30°C on total energy expenditure and adipose tissue protein signatures. C57BL/6J mice were housed at 24°C for 2 wk and then either remained at 24°C (n = 16/group, 8M/8F) or were transitioned to 30°C (n = 16/group, 8M/8F) for 4 wk. Total energy expenditure and its components were determined by indirect calorimetry. Interscapular brown adipose tissue (iBAT) and inguinal white adipose tissue (iWAT) proteins were quantified by Western blot and quantitative proteomics. Transitioning from 24°C to 30°C reduced total energy expenditure in both male (−25%) and female (−16%) mice, which was attributable to lower basal energy expenditure in males (−36%) and females (−40%). Total iBAT UCP1 protein content was 50% lower at 30°C compared with 24°C, whereas iWAT UCP1 protein content was similar between conditions. iBAT UCP1 protein content remained 20-fold greater than iWAT at 30°C. In iBAT and iWAT, 183 and 41 proteins were differentially expressed between 24°C and 30°C, respectively. iWAT proteins (257) differentially expressed between sexes at 30°C were not differentially expressed at 24°C. Thus, 30°C housing lowers total energy expenditure of mice when compared with an ambient temperature (24°C) that falls within the National Research Council’s guidelines for housing laboratory mice. Lower iBAT UCP1 content accompanied chronic housing at 30°C. Furthermore, housing temperature influences sexual dimorphism in the iWAT proteome. These data have implications regarding the optimization of preclinical models of human disease.

NEW & NOTEWORTHY Housing mice at 30°C reduced the basal and total energy expenditure compared with 24°C, which was accompanied by a reduction in brown adipose tissue UCP1 content. Proteomic profiling demonstrated the brown adipose tissue and white adipose tissue proteomes were largely influenced by housing temperature and sex, respectively. Therefore, 30°C housing revealed sexual dimorphism in the white adipose tissue proteome that was largely absent in animals housed at 24°C.

INTRODUCTION

Mice are widely used to study human disease. Typically, laboratory mice are housed at temperatures that fall between 20 and 26°C, in line with recommendations from the National Research Council’s Guide for the Care and Use of Laboratory Animals (1). Despite this, it is acknowledged that these temperatures subject mice to cold stress, resulting in compensatory nonshivering thermogenesis and increased energy expenditure (2). Indeed, in mice housed at ∼21°C, the ratio of total to basal energy expenditure falls between 2.5 and 3.5 (3). Conversely, when laboratory mice are housed between 28 and 32°C—thought to reflect thermoneutrality (3–5)—minimal energy is expended to defend core body temperature. As a result, the total to basal energy expenditure ratio in mice housed near thermoneutrality lies between 1.6 and 1.9 (3, 4), which better reflects the total to basal energy expenditure ratio of humans (6, 7). Accordingly, housing temperature is an important experimental variable that may impact the translational potential of preclinical studies in rodents.

The main site of nonshivering thermogenesis in rodents is brown adipose tissue (BAT). During cold acclimation, the thermogenic capacity of BAT is markedly increased by chronic elevation of sympathetic tone (8). BAT recruitment involves mitochondrial biogenesis and increased uncoupling protein-1 (UCP1) expression (9, 10). UCP1 is essential for cold-induced nonshivering thermogenesis (11, 12), but unlike other membrane carrier proteins, such as the adenine nucleotide translocator (ANT), UCP1 does not contribute to basal mitochondrial proton leak (13, 14). Rather, upon cold exposure, increased fatty acid concentrations overcome purine nucleotide inhibition and activate UCP1, thereby uncoupling oxygen consumption from ATP synthesis. Although BAT is the main site of nonshivering thermogenesis, rodent white adipose tissue (WAT) becomes thermogenic after prolonged adrenergic stress (15, 16). However, the thermogenic capacity of WAT UCP1 in mice housed at 4°C for 4 wk is between one-third and one-tenth of that of BAT (16), suggesting that BAT is the predominant site of UCP1 thermogenesis, even when WAT undergoes a marked browning response to severe and prolonged cold stress.

Although it is clear that subthermoneutral housing augments nonshivering thermogenesis in mice, it is less clear to what extent modest changes in housing temperature impact whole body and adipose tissue energetics. To this end, we investigated the impact of transitioning mice from 24°C to 30°C on total energy expenditure and its components, as well as the chronic impact of this temperature transition on BAT and WAT protein signatures. We further aimed to determine the role of biological sex in metabolic response to altered housing temperature.

METHODS

Animals

Male and female C57Bl/6J (No. 000664, Jackson Laboratories, Bar Harbor, ME) mice (6 to 8 wk old) were used in the current study. Upon arrival at our facility, animals were group housed at 24°C on a reverse light cycle (light 7:00 PM–7:00 AM) with ad libitum access to a standard chow diet [TD.95092 (18.8% protein, 17.2% kcal fat, 63.9% kcal carbohydrate, 3.8 kcal/g), Envigo Teklad Diets, Madison, WI] and drinking water. After around a 2-wk acclimation period, animals were randomized to one of two ambient temperature groups that both underwent a 6-wk study. One group of animals (n = 16, 8 males and 8 females) was housed at 24°C for a total of 6 wk (24°C group). The standard housing temperature in our research vivarium is 24°C. A second group of animals (n = 16, 8 males and 8 females) was housed at 24°C for a total of 2 wk, and were then transitioned to a housing temperature of 30°C for a further 4 wk (30°C group). At the conclusion of the 6-wk study, all animals in both groups were euthanized in a rising concentration of CO₂, and tissue samples were collected. This protocol was approved by the Institutional Animal Care and Use Committee at the University of Arkansas for Medical Sciences.

Body Composition Analysis

Body composition was measured by quantitative magnetic resonance imaging (qMRI) using the EchoMRI-500 (EchoMRI, Houston, TX). Fat-free mass (FFM) was calculated as the difference between body weight and fat mass (FM). Body composition was determined at the end of the 6-wk study period.

Metabolic and Behavioral Phenotyping

From day 12 to day 16 of the 6-wk study, all mice underwent metabolic and behavioral phenotyping. Mice were individually housed for five consecutive days in specialized cages that allowed V̇o₂ and V̇co₂ to be continuously measured to calculate metabolic rate (Sable Systems International, Las Vegas, NV). During this time, food and water intake, activity, and voluntary wheel running were also continuously recorded. Mice only had access to running wheels during the 5-day recording period, after which mice were housed in cages with no running wheel access. During metabolic and behavioral phenotyping, animals were housed in environmental cabinets to control ambient temperature. In the 24°C group, cabinet temperature was set to 24°C for the 5 days where data were being recorded. In the 30°C group, the temperature was increased in the cabinet from 24°C to 30°C by 1°C per hour on day 3 of metabolic data collection. Daily metabolic and behavioral data for the entire 5-day data collection period are presented in Supplemental Fig. S1 (all Supplemental Material is available at https://doi.org/10.6084/m9.figshare.20173955).

Data files from metabolic and behavioral phenotyping experiments were processed using macros to distill data into hourly averages/totals and provided circadian reports of metabolic and behavioral data for each 12-h light cycle (Sable Systems International, Las Vegas, NV). Rates of energy expenditure were calculated from V̇o₂ and V̇co₂ using the Weir equation:

$$\text{EE\hspace{0.17em}}\left(\frac{\text{kcal}}{\text{h}}\right)\hspace{0.17em}=60\times \left(0.003941\times {\hspace{0.17em}\dot{}\text{V}\text{O}}_2)+(0.001106\times {\dot{}\text{V}\text{CO}\hspace{0.17em}}_2\right).$$

Total daily energy expenditure was calculated as the sum of the average rate of energy expenditure (kcal/h) for both the light and dark cycle times 12, summed and averaged across the final 2 days of the experimental period for both the 24°C and 30°C groups. Basal energy expenditure was calculated from the average rate of energy expenditure during the 30-min period with the lowest energy expenditure as kcal/h, extrapolated to 24 h and averaged across the final 2-day measurement periods. Resting metabolic rate was calculated from the average energy expenditure during the lowest 30-min period of activity as kcal/h, extrapolated to 24 h and averaged across the final 2-day measurement periods. Total daily wheel energy expenditure was calculated as the summed energy expenditure of each wheel running bout per day and averaged across the final 2-day measurement period. Energy expenditure of individual wheel running bouts was calculated as the energy expenditure exceeding the basal energy expenditure associated with each running bout.

Energy expenditure data were not normalized to body mass given the limitations with this approach (17). ANCOVA analyses were performed to attempt to account for the confounding effects of body composition (i.e., body mass, lean body mass) on measures of mouse energy expenditure, but we found no significant confounding effects of body mass or lean body mass on total energy expenditure or basal energy expenditure (Supplemental Fig. S2, A–D). Therefore, energy expenditure data are presented as the daily average values per animal. Energy intake was calculated as the total food intake for each feeding period times the energy density of each diet and summed across each day. All meters were the measurement of cage ambulation that included all gross and fine movements. Wheel running and cage physical activity for one male/female was excluded due to technical issues.

Tissue Collection

Mice were euthanized in a rising concentration of CO₂. Thereafter, interscapular brown adipose tissue (iBAT) and inguinal white adipose tissue (iWAT) depots were excised and weighed. Tissue was then immediately frozen and stored at −80°C for further analyses.

SDS-PAGE and Immunoblotting

Mitochondrial and thermogenic proteins in iBAT and iWAT were quantified by Western blot. Approximately 30–80 mg of iBAT and ∼75–150 mg of iWAT were homogenized with a glass pestle and tube in ice-cold 1× precipitation assay buffer containing: 25 mM Tris-HCl pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, and 0.1% SDS, supplemented with 1× HALT protease inhibitor cocktail (Thermo Fisher Scientific). Tissue lysates were kept on ice for 1 h before being centrifuged for 15 min at 20,000 g (4°C). The supernatant was collected while carefully avoiding the lipid layer. Samples were then centrifuged for 15 min at 18,000 g (4°C) for an additional two times to minimize the amount of contaminating lipids (18), before the supernatant was stored at −80°C. Total sample protein was quantified by the Pierce BCA assay. Samples were subsequently resuspended in 4× Laemmli buffer (Bio-Rad Laboratories, Hertfordshire, UK) containing reducing agent (1× working concentration: 31.5 mM Tris-HCl [pH 6.8], 10% glycerol, 1% SDS, 0.005% bromophenol blue, and 355 mM 2-mercaptoethanol) and were heated between 37 and 95°C depending upon recommended sample preparation from primary antibody suppliers. Between 20 and 45 µg of total protein was loaded depending upon the protein and tissue of interest and electrophoresed on 12% Mini-TGX SDS polyacrylamide gels. Proteins were transferred to a PVDF membrane by one of two methods: 1) semidry transfer using the Trans-Blot Turbo Transfer System and 2) tank transfer at 0.15 A for 60 min in CAPS (pH 11) transfer buffer. Ponceau staining was performed to confirm equal loading between lanes. After removal of ponceau stain, membranes were blocked in Tris-buffered saline Tween-20 (TBS-T) or phosphate buffered saline (PBS) containing 2%–5% nonfat dry milk for 1 h at room temperature or overnight at 4°C. After blocking, membranes were incubated overnight with rabbit or mouse anti total antibodies: UCP1 (1:5,000; Abcam, ab10983), ANT1/2 (1:1,000; Abcam, ab110322), and OXPHOS (1:1,000; Abcam, ab110413) in 5% BSA or 1% milk. After overnight incubation, the membranes were washed three times in TBS-T or PBS-T for 5 min and incubated for 1 h in horseradish peroxidase (HRP)-conjugated anti-rabbit/mouse antibodies (Cell Signaling Technology, London, UK) at a dilution of 1:10,000–1:20,000. Proteins were visualized by enhanced chemiluminescence (Thermo Fisher Scientific Inc., Waltham, MA) and quantified by densitometry (Amersham Imager 600, GE Healthcare, Life Sciences, NJ).

Quantitative Proteomics

Total protein from tissue (n = 6/group/tissue) was reduced, alkylated, and purified by chloroform/methanol extraction before digestion with sequencing grade-modified porcine trypsin (Promega, Madison, WI). Tryptic peptides were then separated by reverse-phase XSelect CSH C18 2.5 μm resin (Waters, Milford, MA) on an in-line 150 × 0.075 mm column using an UltiMate 3000 RSLCnano system (Thermo). Peptides were eluted using a 60-min gradient from 98:2 to 65:35 buffer A:B ratio (Buffer A = 0.1% formic acid, 0.5% acetonitrile; Buffer B = 0.1% formic acid, 99.9% acetonitrile). Eluted peptides were ionized by electrospray (2.4 kV) followed by mass spectrometric analysis on an Orbitrap Exploris 480 mass spectrometer (Thermo). To assemble a chromatogram library, six gas-phase fractions were acquired on the Orbitrap Exploris with 4 m/z data-independent acquisition (DIA) spectra (4 m/z precursor isolation windows at 30,000 resolution, normalized AGC target 100%, maximum inject time 66 ms) using a staggered window pattern from narrow mass ranges using optimized window placements. Precursor spectra were acquired after each DIA duty cycle, spanning the m/z range of the gas-phase fraction (i.e., 496–602 m/z, 60,000 resolution, normalized AGC target 100%, maximum injection time 50 ms). For wide-window acquisitions, the Orbitrap Exploris was configured to acquire a precursor scan (385–1,015 m/z, 60,000 resolution, normalized AGC target 100%, maximum injection time 50 ms) followed by 50× 12 m/z DIA spectra (12 m/z precursor isolation windows at 15,000 resolution, normalized AGC target 100%, maximum injection time 33 ms) using a staggered window pattern with optimized window placements. Precursor spectra were acquired after each DIA duty cycle.

Proteomic Data Analysis

Following acquisition, data were searched using an empirically corrected library and a quantitative analysis was performed to obtain a comprehensive proteomic profile. Proteins were identified and quantified using EncyclopeDIA and visualized with Scaffold DIA using 1% false discovery thresholds at both the protein and peptide levels (19). The UniProtKB Mus musculus database was used for the database search. Protein-exclusive intensity values were assessed for quality and normalized using ProteiNorm (20). The data were normalized using Cyclic Loess and statistical analysis was performed using Linear Models for Microarray Data (limma) with empirical Bayes (eBayes) smoothing to the standard errors (21). Proteins with an FDR-adjusted P value <0.05 and a fold change >2 were considered to be significant. Gene ontology enrichment analysis was performed to determine significantly enriched biological processes using the online tool GOrilla (22). The background gene list used for analysis included only those proteins detected in the proteomic experiment, and a statistical threshold of 1 × 10⁻³ was used to determine significantly enriched processes (23).

Statistical Analyses

Data are presented as means ± standard error. Two-way ANOVAs were performed to detect main effects (i.e., temperature or sex) and potential significant interactions between temperature and sex. Multiple comparisons between experimental conditions were adjusted for multiple tests, using Dunnett’s or Sidak’s where appropriate. Statistical analyses were carried out using GraphPad Prism version 9 (GraphPad Software, LLC, San Diego, CA). ANCOVA analyses were performed in R Studio (v.1.4.1717, RStudio, PBC, Boston, MA). Statistical significance was set at P < 0.05.

RESULTS

Thermoneutral Housing Lowers Energy Expenditure

First, we determined how housing temperature and sex impacted mouse body mass and composition. Females weighed 22% and 26% less than males at 24°C and 30°C, respectively (Fig. 1A). Absolute FFM was 24% and 25% lower in females versus males at 24°C and 30°C, respectively (Fig. 1B). Absolute fat mass was similar between temperature conditions and sexes (Fig. 1C).

Figure 1.

Energy expenditure is attenuated in mice housed at 30°C. A: body mass of male and female mice (n = 8M/8F per group) housed at 24°C for 2 wk then either maintained at 24°C or transitioned to 30°C for a further 4 wk. Absolute fat-free mass (FFM; B) and absolute body fat content (C). Indirect calorimetry was used to determine average daily total energy expenditure (D) and basal energy expenditure (E) (n = 8M/8F mice per group). F: basal energy expenditure as a % of total energy expenditure. G: Energy intake calculated as the sum of daily food intake (g) times the energy density (kcal/g) of the provided diet (n = 8M/8F mice per group). All meters excluding wheel running (H) and wheel running meters (I). Representative experimental energy expenditure traces from mice housed at 24°C (J) or 24°C then transitioned to 30°C (K). Blue and red bars in A–I indicate 24°C and 30°C conditions, respectively. Green and pink lines/points in J–K reflect males and females, respectively. χ main effect of sex. τ main effect of temperature. χ × τ sex and temperature interaction. *P < 0.05, **P < 0.01, and ***P < 0.001.

Next, we determined the impact of housing temperature on total energy expenditure and its components. Total energy expenditure was 25% and 16% lower in males and females respectively at 30°C (Fig. 1D). Total energy expenditure of females was 13% lower than males at 24°C, but similar between sexes at 30°C. The basal energy expenditure of males and females was 36% and 40% lower at 30°C versus 24°C, respectively (Fig. 1E). Basal energy expenditure comprised 50% and 43% of total energy expenditure in males at 24°C and 30°C, respectively (Fig. 1F). Whereas in females, basal energy expenditure comprised 56% and 40% of total energy expenditure at 24°C and 30°C, respectively. Wheel energy expenditure was 82% greater in females versus males at 30°C, though wheel energy expenditure was similar between sexes at 24°C (Supplemental Fig. S1I). A 62% increase in wheel energy expenditure was observed in females at 30°C versus 24°C.

Average daily energy intake decreased by 46% in males at 30°C (Fig. 1G), and energy intake of females at 24°C was 34% lower than males at 24°C. Weight-adjusted energy intake was not different from nonadjusted energy intake (data not shown). All meters covered by animals were similar between conditions (Fig. 1H). Wheel running activity decreased by 50% in males at 30°C versus 24°C (Fig. 1I). At 30°C, females ran 100% further than males. Representative energy expenditure traces from mice housed at 24°C and 30°C are shown in Fig. 1, J and K.

Housing Temperature Alters Brown Adipose Tissue Mass and Thermogenic Capacity

The absolute iBAT dissected depot mass was 73% and 77% greater at 30°C versus 24°C in males and females, respectively (Fig. 2A). Absolute iBAT mass of females was 12% lower than males at 30°C. Relative iBAT depot mass was greater (∼74%) at 30°C compared with 24°C in both males and females (Fig. 2B). Compared with 24°C, iBAT protein density (µg/mg tissue) was 37% and 32% lower in males and females at 30°C, respectively (Fig. 2C). However, total iBAT depot protein density was not different between conditions (Fig. 2D). iBAT UCP1 protein expression was 57% and 55% lower in males and females at 30°C, respectively, when compared with animals housed at 24°C (Fig. 2E). iBAT ANT1/2 protein expression was 78% and 46% greater in males and females at 30°C compared with 24°C (Fig. 2F).

Figure 2.

Housing temperature alters the thermogenic capacity of brown adipose tissue. A: interscapular brown adipose tissue (iBAT) mass. B: relative iBAT mass. C: iBAT protein content per mg tissue. D: iBAT protein content per depot (mg/depot). E: iBAT uncoupling protein-1 (UCP1) protein levels (n = 4 mice/group). F: iBAT ANT1/2 protein levels (n = 4/group). G: iBAT total OXPHOS protein levels. H: iBAT OXPHOS complex I-V protein levels. I: representative Western blot images. Blue and red bars reflect 24°C and 30°C, respectively. Animal n = 4–8/group. χ main effect of sex. τ main effect of temperature. *P < 0.05 and ***P < 0.001.

To further explore the impact of housing temperature on iBAT thermogenic capacity, we quantified the levels of key mitochondrial proteins. Although a significant main effect of temperature was observed for total iBAT OXPHOS protein contents (Fig. 2G), pairwise comparisons were not statistically significant. iBAT complex I protein expression was 22% and 34% lower at 30°C versus 24°C in males and females, respectively (Fig. 2H). Representative Western blot images are presented in Fig. 2I.

Housing Temperature Does Not Alter the Thermogenic Capacity of White Adipose Tissue

Absolute iWAT mass was similar between conditions (Fig. 3A). Although there were no differences in relative iWAT mass between sexes, relative iWAT mass was 15% greater in females at 30°C versus 24°C (Fig. 3B). iWAT protein content per mg tissue was 38% lower in males at 30°C versus 24°C. Female iWAT protein content (per mg tissue) was 13% greater than males at 30°C (Fig. 3C). Total iWAT depot protein content was 38% lower in males at 30°C versus 24°C, but 35% greater in females at 30°C versus 24°C (Fig. 3D). Total protein content of females iWAT depot at 30°C was 125% greater than males at 30°C (Fig. 3D).

Figure 3.

Housing temperature alters the thermogenic capacity of white adipose tissue. A: inguinal white adipose tissue (iWAT) mass. B: relative iWAT mass. C: iWAT protein content per mg tissue. D: iWAT protein content per depot (mg/depot). E: iWAT uncoupling protein-1 (UCP1) protein levels (n = 4 mice/group). F: iWAT ANT1/2 protein levels (n = 4/group). G: iWAT total OXPHOS protein levels. H: iWAT OXPHOS complex I-V protein levels. I: representative Western blot images. Blue and red bars reflect 24°C and 30°C, respectively. Animal n = 4–8/group. χ main effect of sex. τ main effect of temperature. *P < 0.05, **P < 0.05 and ***P < 0.001.

iWAT UCP1 protein content was not affected by temperature or sex as measured by proteomics and Western blot (Fig. 3E). There were no differences in ANT1/2 protein content between conditions (Fig. 3F). iWAT total OXPHOS contents and OXPHOS complex protein contents were lower at 30°C versus 24°C (Fig. 3, G–I).

Brown Adipose Tissue Is the Predominant Site of Nonshivering Thermogenesis in Mice

To better understand the potential physiological relevance of temperature-induced changes in brown and white adipose tissue thermogenic capacity, we compared the protein contents of both depots, and how housing temperature impacted these parameters. Per mg tissue, the protein contents of iBAT were 5.5-fold and 3.7-fold greater than iWAT at 24°C and 30°C, respectively (Fig. 4A). Total protein per depot was similar between conditions (Fig. 4B). UCP1 content per µg protein was 20- and 10-fold greater in iBAT versus iWAT at 24°C and 30°C, respectively (Fig. 4C). Total UCP1 protein per depot was 39- and 20-fold greater in iBAT versus iWAT at 24°C and 30°C, respectively (Fig. 4D).

Figure 4.

Brown adipose tissue is the predominant site of nonshivering thermogenesis. A: tissue protein content (µg/mg tissue). B: total tissue protein (mg/depot). C: uncoupling protein-1 (UCP1) protein content per µg protein. D: UCP1 protein per depot. Animal n = 8–16/group. Blue bars reflect 24°C (blue bars) and red bars 30°C. τ main effect of temperature. θ main effect of tissue. ***P < 0.001.

Quantitative Proteomics Reveal a Depot-Specific Impact of Housing Temperature and Sex on Adipose Tissue Proteomes

To further characterize how housing temperature and sex influence iBAT and iWAT protein expression, we performed label-free quantitative proteomics. In total, 183 proteins were differentially expressed in iBAT from mice housed at 30°C compared with 24°C, where 92 and 91 proteins were significantly up- and downregulated versus 24°C, respectively (Fig. 5A). Proteins differentially expressed in iBAT between temperatures were largely related to lipid metabolic biological processes (Supplemental Table S1). iBAT proteins of lower abundance at 30°C compared with 24°C included UCP1, UCP3, fatty acid synthase (Fasn), ATP citrate synthase (Acly), acetyl–CoA carboxylase-1, glycerol kinase, long-chain-fatty-acid–CoA ligase 5 and medium-chain acyl–CoA ligase, mitochondrial. The top 30 differentially expressed proteins in iBAT between conditions are reported in Supplemental Table S2.

Figure 5.

The impact of housing temperature and sex on the interscapular brown adipose tissue (iBAT) and inguinal white adipose tissue (iWAT) proteome. Volcano plots of temperature (A and B) and sex (C–F) effects on the iBAT and iWAT proteomes. iBAT and iWAT tissue were harvested from male and female mice housed at 24°C or 30°C (n = 6 mice/group). Individual proteins (circles) in light blue, black, and red represent proteins significantly downregulated, not differentially expressed, and those significantly upregulated at 30°C, respectively (A and B). Points (circles) in dark blue, black, and pink represent proteins significantly upregulated in males, proteins not differentially expressed, and proteins significantly downregulated in males versus females, respectively (C–F). Data represent n = 6–12 mice/group. Proteins are labeled by gene name.

A total of 41 total proteins were differentially expressed in iWAT from mice housed at 30°C. Of those proteins, 17 and 24 proteins were up- and downregulated at 30°C compared with 24°C, respectively (Fig. 5B). Proteins of lower abundance in iWAT at 30°C included V-type proton ATPase subunit C 1, acetyl–CoA carboxylase-1, acetyl–CoA carboxylase 2, acetyl-coenzyme A synthetase, cytoplasmic and acyl carrier protein, mitochondrial. No biological processes were significantly enriched by gene ontology enrichment analysis after adjusting for the false discovery rate (Supplemental Table S3). The top 30 differentially expressed proteins in iWAT between conditions are reported in Supplemental Table S4.

Few proteins were differentially impacted by sex in iBAT, regardless of housing temperature. At 24°C, α-1-antitrypsin 1–5 (Serpina1e) and eukaryotic translation initiation factor 2 subunit 3, Y-linked (Eif2s3y) were significantly less abundant in females versus males (Fig. 5C). At 30°C, 11 proteins were differentially expressed between sexes, which included four down- and seven upregulated proteins in females versus males (Fig. 5D). Interestingly, two proteins with key roles in de novo lipogenesis—Fasn and Acly—exhibited greater expression in females at 30°C, but not at 24°C.

The iWAT proteome exhibited marked sexual dimorphism, particularly in mice housed at 30°C. A total of 141 proteins were differentially expressed between males and females housed at 24°C (Fig. 5E). Of those proteins, 64 and 77 were significantly up- and downregulated in females compared with males, respectively. According to gene ontology analysis, these proteins were related to cell adhesion, one-carbon compound transport, and developmental processes (Supplemental Table S5). The top 30 differentially expressed proteins between sexes at 24°C are shown in Supplemental Table S6. In iWAT from mice housed at 30°C, 331 proteins were differentially expressed between sexes (Fig. 5F). Specifically, 150 and 181 proteins were significantly up- and downregulated in females compared with males, respectively. Gene ontology enrichment analysis revealed that no biological processes were significantly enriched in iWAT at 30°C (Supplemental Table S7). The top 30 differentially expressed iWAT proteins between sexes at 30°C are shown in Supplemental Table S8. Notably, a total of 257 proteins in iWAT were differentially expressed between males and females at 30°C, which were not differentially expressed at 24°C.

DISCUSSION

We investigated how a modest change in housing temperature affects whole body energetics and adipocyte protein signatures in mice. Housing mice at 30°C decreases basal energy expenditure and thus total energy expenditure when compared with housing at 24°C. Lower total energy expenditure at 30°C was associated with less iBAT, but not iWAT, UCP1 protein abundance. Total iBAT UCP1 protein content was 20-fold greater than iWAT at 30°C, suggesting that iBAT UCP1 is the principal contributor to facultative thermogenesis in mice housed at 30°C. Furthermore, proteomic profiling of iWAT revealed sexual dimorphism, which was more prominent when mice were housed at 30°C rather than at 24°C.

30°C Housing Lowers Total Energy Expenditure Largely by Lowering Basal Energy Expenditure

The first key finding of this study was that basal energy expenditure and total energy expenditure were markedly lowered by transitioning mice from a standard housing temperature (24°C) to one thought to better reflect thermoneutrality (30°C). Even this rather modest temperature change (i.e., 6°C) had a striking impact on total energy expenditure, largely by reducing basal energy expenditure, suggesting that facultative thermogenesis is a significant contributor to total energy expenditure in mice housed at 24°C. Our current data align with previous work demonstrating that housing temperatures approaching thermoneutrality attenuate mouse energy expenditure (2–5). Together, these findings illustrate that even modest changes in housing temperature can markedly influence whole body energetics in mice. Moreover, our data lend further support to the notion housing temperature is a critical variable to consider when optimizing rodent studies to best align mouse energy metabolism to that of free-living humans (3–5, 24).

Reduced energetic demands at thermoneutrality are accompanied by lowered daily food intake (5, 25–27). Here, we observed that male, but not female mice decreased daily energy intake at 30°C versus 24°C. This finding contrasts data from a previous study where female mice lowered their daily energy intake to 30°C versus 20°C when fed a high-fat diet (26). However, in the current study mice had access to running wheels, where female mice housed at 30°C ran twice the distance that male mice ran, which clearly influenced energy expenditure. Another observation of our study was that daily wheel running volume was less in male mice housed at 30°C compared with those housed at 24°C, which supports the outcomes of some (28, 29), but not all previous studies (25). Broadly speaking, these data raise interesting questions regarding the use of running wheels in rodent energetics studies, where our findings indicate both sex and housing temperature influence voluntary wheel running in mice. However, it is interesting to note that mice ran between ∼5 and 10 km each day, which clearly exerts significant control of total energy expenditure and the relationship between basal energy expenditure and total energy expenditure.

Reduced wheel running volume of male mice at 30°C might be anticipated to explain a portion of the reported decrease in total energy expenditure, however, previous work has demonstrated that wheel running is unlikely to account for differences in total energy expenditure between groups of animals because of compensatory reductions in energy expenditure from other metabolic processes (30). Interestingly, we observed that patterns of wheel energy expenditure did not exactly align with wheel running volume. This finding could be explained by differences in wheel running economy since energetic costs of wheel running are greater in mice housed at 28°C versus 24°C (30). Moreover, the high-resolution respiratory gas exchange data generated in the current study allowed for energy expenditure to be determined in the basal state and during wheel running. Warmer housing temperature exerted its most pronounced effect on basal energy expenditure, suggesting that despite altered running behavior, the primary determinant of lower total energy expenditure in mice housed at 30°C reduced basal energy expenditure, presumably the result of attenuated facultative thermogenesis.

iBAT Thermogenic Capacity Is Blunted at 30°C

Mitochondria serve specialized functions within BAT and WAT (31). One such function for BAT mitochondria is UCP1-dependent heat production (32), a form of recruitable mitochondrial proton leak. Indeed, UCP1-dependent thermogenesis in BAT is thought to be the primary site of facultative heat production in rodents housed below thermoneutrality. We observed lower complex I protein expression in iBAT at 30°C compared with 24°C, which supports studies demonstrating mitochondrial density is increased during cold acclimation (33, 34). This finding is also consistent with the idea that complex I-linked substrates (i.e., pyruvate and malate) are the preferred respiratory substrates of brown adipose tissue mitochondria (14, 35, 36). Cold acclimation also increases mitochondrial and thermogenic protein contents in a subpopulation of white adipocytes known as brite/beige cells (16, 37, 38). Here, the contents of complexes I-IV in iWAT were modestly affected by a 6°C reduction in housing temperature, which supports some (28), but not all previous investigations (25). Overall, our data posit that 30°C housing lowers mitochondrial protein content in iBAT and iWAT.

Upon activation, BAT displays remarkable thermogenic capacity owing to its abundance of UCP1 protein (39, 40). As expected (33, 40, 41), we observed lesser BAT UCP1 protein abundance at 30°C compared with 24°C. Lower UCP1 protein content would be anticipated to substantially lower the thermogenic capacity of iBAT (10, 39). Indeed, isolated brown fat mitochondria from mice acclimated to 30°C display decreased UCP1-dependent respiration compared with those acclimated to 4°C (14, 16). Previous studies have reported greater iBAT UCP1 protein content in female rodents housed at room temperature compared with male counterparts (42–45), but we observed no sexual dimorphism for UCP1 protein content in this study, suggesting that in our experimental paradigm iBAT thermogenic capacity is comparable between sexes. Besides UCP1, the adenine nucleotide translocase—which facilitates ADP/ATP exchange—contributes to basal proton leak of BAT mitochondria (13, 46). We observed greater BAT ANT1/2 protein content at 30°C versus 24°C, and while BAT is not typically associated with ATP production, our observation may suggest that iBAT oxygen consumption is perhaps better coupled to ATP synthesis at 30°C compared with 24°C. Alternatively, this finding may suggest greater basal (ANT-dependent) proton leak in BAT of mice housed at 30°C, where UCP1 abundance is markedly reduced. Together, these data underscore the influence of housing temperature on iBAT UCP1 abundance.

It has long been appreciated that rodent WAT depots can express UCP1 following prolonged adrenergic stress (47). Although iWAT develops functional UCP1 after prolonged acclimation to severe cold (16), iWAT UCP1 abundance was not measurably different in animals housed at 30°C versus 24°C in the current study. This finding supports previous studies of mice housed at 21–22°C (25, 28, 41), and likely reflects the moderate shift in housing temperature studied here (i.e., 30°C vs. 24°C), which contrasts temperatures close to freezing that are often used to maximally recruit iWAT UCP1. Aside from UCP1, we report that iWAT ANT1/2 protein expression is lowered at 30°C. Increased capacity for ANT-dependent thermogenesis at temperatures below thermoneutrality may facilitate a heightened demand for ATP, where lipid turnover may be increased.

Sex Influences Adipose Tissue Proteomic Signatures in a Temperature-Dependent Manner

Untargeted “omics” approaches are increasingly leveraged to survey physiological adaptations to environmental stimuli. Here, we report that the iBAT proteome was particularly sensitive to housing temperature, where over 180 proteins were differentially expressed between animals housed at 24°C and 30°C. Proteins involved in lipid metabolism and bioenergetics such as UCP1, fatty acid synthase, ATP-citrate lyase, long-chain-fatty-acid–CoA ligase 5, and glycerol kinase were all downregulated at 30°C. These findings underscore the impact of housing temperature on mouse iBAT physiology, which extends well beyond an upregulation of the thermogenic protein UCP1. Interestingly, sex did not considerably affect the murine iBAT proteome, regardless of housing temperature. One previous study has documented how sex influences the rat BAT proteome, where fatty acid synthase expression was elevated in females versus males (48). Likewise, we observed that fatty acid synthase, as well as ATP-citrate lyase, were upregulated in iBAT of females versus males at 30°C, suggesting that females’ capacity for de novo lipogenesis in iBAT is greater than males (49, 50). Yet, these effects were not apparent at 24°C, suggesting that cold stress and a subsequent increase in fuel use masks the sexual dimorphism seen in BAT at 30°C.

It is widely acknowledged that there are sex differences in lipid metabolism. For example, females have greater rates of triglyceride synthesis in subcutaneous WAT when compared with any WAT depot in males (51). In this study, we report substantial sexual dimorphism in the murine iWAT proteome. Proteins involved in de novo lipogenesis were more abundant in females versus males, regardless of housing temperature. These included key targets of the carbohydrate response element-binding protein (ChREBP), such as fatty acid synthase, ATP-citrate lyase, and acetyl-coenzyme A synthetase. Our findings support previous research reporting greater gene expression of ChREBP targets in wild-type females versus males (49), where female adipose tissue was more dependent upon de novo lipogenesis for lipid storage. One challenge for future research is to elucidate the potential mechanisms underlying sexual dimorphism in de novo lipogenesis, which could relate to sex hormone regulation (52). Notably, the number of iWAT proteins differentially expressed between sexes was twofold greater in animals housed at 30°C, suggesting that sexual dimorphism in the iWAT proteome is further revealed when animals are housed under conditions thought to reflect thermoneutrality. As recent work has suggested (25), future studies concerning iWAT function must carefully choose animal housing conditions to help promote the translation of findings to human populations. Our current data support this assertion, where the presence of iWAT sexual dimorphism was significantly influenced by housing temperature.

Limitations

Some limitations must be considered when interpreting the current data. With regards to whole body energetics, mouse energy expenditure was assessed during the temperature transition period in week 1. Energy expenditure may have diverged between conditions after several weeks of acclimation. However, we found that total energy expenditure of mice after 1 and 4 wk of 30°C housing to be comparable (data not shown). Furthermore, all mice had access to running wheels during the 5-day assessment of energy expenditure in the current study. This allowed us to assess voluntary wheel running and associated wheel running energy expenditure in males and females and their potential modulation by housing temperature. However, there is evidence to suggest that mice may require ∼2–3 wk to completely adjust to running wheels (53, 54), and therefore the voluntary wheel running recorded in the present study may not reflect habitual physical activity of mice fully acclimated to running wheels. Furthermore, studies comparing the impact of housing temperature on total energy expenditure and its components often do so in mice that do not have access to running wheels. Accordingly, this must be taken into consideration when comparing the current data to some of the published literature. A final limitation was that mouse food absorption efficiency was not assessed.

Conclusions

In summary, a modest 6°C change in housing temperature markedly alters the basal and therefore total energy expenditure of mice. Compared with housing temperatures in line with the National Research Council’s Guide for the Care and Use of Laboratory Animals, housing mice at 30°C markedly lowers total energy expenditure and iBAT UCP1 content. As well as dampening iBAT thermogenic capacity, housing mice at 30°C further reveals sexual dimorphism in the iBAT, and in particular iWAT proteome, emphasizing the importance of considering housing temperature when studying adipose tissue physiology. Collectively, these data have important implications regarding the optimization of preclinical rodent models of human metabolism and disease.

DATA AVAILABILITY

All datasets generated from the current study are included in this article and the supplemental files. Data files are available from the the corresponding author upon request.

SUPPLEMENTAL DATA

Supplemental Figs. S1 and S2 and Supplemental Tables S1–S8:https://doi.org/10.6084/m9.figshare.20173955.

GRANTS

This study was supported by the National Institute of General Medical Sciences (NIGMS) through a Center of Biomedical Research Excellence pilot Project 5P20GM109096 and R35GM142744. Support was also provided by the Arkansas Biosciences Institute and the United States Department of Agriculture-Agricultural Research Service (USDA-ARS) Grant USDA ARS 6026-51000-012-06S. We acknowledge the IDeA National Resource for Quantitative Proteomics for funding and conducting the proteomic work described in this study under Grant No. R24GM137786.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHORS CONTRIBUTIONS

C.P. conceived and designed research; D.G.S., L.T., J.D.S., and C.P. performed experiments; D.G.S. and C.P. analyzed data; D.G.S. and C.P. interpreted results of experiments; D.G.S. and C.P. prepared figures; D.G.S. and C.P. drafted manuscript; D.G.S. and C.P. edited and revised manuscript; D.G.S., L.T., J.D.S., and C.P. approved final version of manuscript.