Difference in Housing Temperature-Induced Energy Expenditure Elicits Sex-Specific Diet-Induced Metabolic Adaptations in Mice

Abstract

Objective:

Test whether increased energy expenditure (EE), independent of physical activity, reduces acute diet-induced weight gain through tighter coupling of energy intake to energy demand and enhanced metabolic adaptations.

Methods:

We utilized indirect calorimetry and qMRI to assess energy metabolism and body composition during 7-day high-fat/high-sucrose (HFHS) feeding in male and female mice housed at divergent temperatures (20°C vs. 30°C).

Results:

As previously observed, 30°C housing resulted in lower total EE and energy intake compared to 20°C mice regardless of sex. Interestingly, housing temperature did not impact HFHS-induced weight gain in females, while 30°C male mice gained more weight than 20°C males. Energy intake coupling to EE during HFHS was greater in 20°C versus 30°C, with females greater at both temperatures. Fat mass gain was greater in 30°C mice compared to 20°C, while females gained less fat mass than males. Strikingly, female 20°C mice gained considerably more fat-free mass than 30°C. Reduced fat mass gain was associated with greater metabolic flexibility to HFHS, while fat-free mass gain was associated with diet-induced adaptive thermogenesis.

Conclusions:

These data reveal that EE and sex interact to impact energy homeostasis and metabolic adaptation to acute HFHS altering weight gain and body composition change.

GRAPHIC ABSTRACT

Legend: Male and female mice housed at 30°C had lower energy expenditure (EE) & energy intake (EI), while having greater energy balance (EB), during 7-day high-fat/high-sucrose (HFHS) feeding compared to male and female mice, respectively, housed at 20°C. However, female mice had lower EB compared to males at both housing temperature. Female mice housed at 30°C gained less weight than 30°C males but gained the same relative amount of fat mass during acute HFHS feeding. Interestingly, 20°C females gained the same amount of weight as 20°C males but gained primarily fat-free mass, while the males gained the same proportion of fat as 30°C males and females.

INTRODUCTION

Current recommendations to prevent weight gain and treat obesity include increasing physical activity or daily exercise with a goal of increasing total energy expenditure (EE) and improving energy balance (1,2). Improved energy balance at higher physical activity levels is proposed to be achieved through greater pairing of energy intake to energy demand (3, 4). Hypothetically, coupling of energy intake to EE improves as energy flux - the sum of EE and energy intake - increases with increasing physical activity (4, 5). This was first proposed by Jean Mayer following his studies in humans and rodents showing that food intake is more closely matched to energy demand at higher physical activity levels and assumedly greater EE & energy flux (6, 7). However, investigating energy flux and pairing of energy intake to energy demand in human subjects is complicated by the large inter-individual variability of total and physical activity EE (8, 9). Also, increased physical activity produces multiple systemic and tissue-specific adaptations independent of increases in EE (reviewed in (10, 11)), which is further complicated by sex-specific differences in physical activity level and physiological adaptation (12, 13). As such, new approaches are necessary to more specifically assess the independent impact of changes in EE on energy balance and protection from weight gain.

The assessment of EE in mice using indirect calorimetry is a common experimental tool. However, while indirect calorimetry experiments are impacted by several factors (14), ambient housing temperature is perhaps the most debated (15, 16, 17, 18). This controversy centers around the well described linear increases in cold-induced nonshivering thermogenesis as ambient temperature drops below the thermoneutral zone (19, 20). This increased thermogenesis to defend body temperature is accompanied by concomitant increases in total EE and energy intake (19, 21), and as such energy flux. While the debate regarding rodent housing temperature is critical to the appropriate design and execution of future mouse studies, the impact of ambient temperature on EE and energy flux can be utilized to investigate the role of these outcomes in the adaptation to metabolic challenges.

In this study, we tested whether differences in EE, independent of physical activity level, impacts energy homeostasis and metabolic adaptation of males and females during a short-term dietary challenge. We utilized qMRI and indirect calorimetry to assess body composition and energy metabolism outcomes in male and female mice during 1-week high-fat/high-sucrose (HFHS) feeding housed at either 20°C or 30°C. We observed the composition of HFHS-induced weight gain (fat vs. fat-free mass) was different between male and female mice with different EE. Additionally, in support of the energy flux hypothesis we observed higher baseline EE during 20°C housing results in greater energy flux, coupling of energy intake to EE, and less positive energy balance. Further, increased baseline EE produced greater adaptation of fat utilization and diet-induced non-shivering thermogenesis during short-term HFHS feeding, particularly in females.

MATERIALS and METHODS

Animals

Male and female C57BI/6J (#000664, Jackson Laboratories, Bar Harbor, ME, USA) mice (6-weeks old) were individually housed (with huts and cotton nestlet) at either 20°C or 30°C on a reverse light cycle (light 10P – 10A) with ad lib access to low-fat, control diet [LFD, D12110704 (10% kcal fat, 3.5% kcal sucrose, 3.85 kcal/gm), Research Diets, Inc., New Brunswick, NJ, USA] for three weeks. Based on previous publications (19, 21), the difference in ambient housing temperature was proposed to produce an ~70 – 80% difference in total EE. At 9-weeks of age, animal and food weight was monitored prior to and following the 7 days of both LFD and the subsequent high-fat, high-sucrose diet [HFHS, D12451, 45% kcal fat, 17% kcal sucrose, 4.73 kcal/gm] at the assigned ambient temperature. The animal protocols were approved by the Institutional Animal Care and Use Committee at the University of Kansas Medical Center and the Subcommittee for Animal Safety at the Kansas City Veterans’ Hospital.

Body Composition Analysis

Body composition was measured by qMRI using the EchoMRI-1100 (EchoMRI, Houston, Texas, USA). Fat-free mass was calculated as the difference between body weight and fat mass. Though not reported, lean mass was determined during the body composition analysis and showed the same patterns as calculated fat-free mass. Body composition was determined prior to, and after the 7-day HFHS feeding.

Indirect Calorimetry & Energy Metabolism

Starting at 9-weeks of age (n=12), energy metabolism was assessed at 20°C or 30°C ambient temperature for 7 days on LFD followed by 7 days of HFHS by measuring VO₂ and VCO₂ in a Promethion continuous metabolic monitoring system (Sable Systems International, Las Vegas, NV, USA). Animals were acclimated to the indirect calorimetry cages for 5 days prior to initiation of data collection. Rate of energy expenditure was calculated with a modified Weir equation [EE (kcal/hr) = 60*(0.003941*VO2+0.001106*VCO2)], and respiratory quotient as VCO₂/VO₂. Total EE was calculated as the daily average rate of energy expenditure for each day times 24 and summed across the 7 days of each diet. Resting EE was determined from the average rate of EE during the 30-minute period with the lowest daily EE as kcal/hr and extrapolated to 24hrs for each day and summed as with total EE. Non-resting EE was calculated as the difference between total EE and resting EE. Metabolic flexibility to HFHS was assessed as the diet-induced change in respiratory quotient as the difference in the 7-day HFHS data minus the 7-day LFD data. Diet-induced adaptive non-shivering thermogenesis was calculated as the difference in resting EE on HFHS and LFD. Energy intake was assessed as the total food intake for each feeding period times the energy density of each diet. Energy balance was determined as the difference between the total energy intake and total EE. Energy flux represents the sum of total EE and energy intake (22). To assess the coupling of energy intake to energy demand, we calculate the percent energy coupled as: 1 – (energy balance/energy flux). From this calculation, the higher the percentage the greater the coupling of energy intake to EE. Food intake, energy intake, and energy balance data during HFHS feeding from two 20°C female mice was not included in data analysis due to excessive food spillage. Thermic effect of food was determined from the consensus thermic effect of food for fat (2.5%), carbohydrate (7.5%), and protein (25%), and the manufacturer provided diet information for each diet (19). As such, the thermic effect of food for HFHS (D12451, Research Diets, 4.73 kcal/g, 45% kcals fat, 35% kcal carbohydrate, 20% kcals protein) is 8.75% or 0.4139 kcal/g. This method of determining thermic effect of food reduces the potential influence of neurobehavioral adaptations of the fed/fasted transition impacting changes in EE through calculation of thermic effect of food across the entire 7-day HFHS feeding. Activity energy expenditure was calculated as the difference between non-resting EE and thermic effect of food. All_Meters is an assessment of cage activity including gross and fine movements; and is determined using the summed distances calculated from the Pythagoras’ theorem that the mouse moved since the previous data point based on XY second by second position coordinates. Cost of movement was calculated as the Activity EE divided by total meters traveled over the 7-days of HFHS. All data from one 20°C female mouse was excluded after discovering malocclusion at necropsy.

Statistical Analysis

Data are presented as scatter plots with means and standard error. The two-standard deviation test was utilized to test for outliers in each group. SPSS version 25 (SPSS Inc., Armonk, New York) was utilized for all statistical analysis. Two-way ANOVA analysis was performed to assess interaction and main effects of sex and temperature on outcome variables. Where significant interactions or main effects were observed, post hoc analysis was performed using least significant difference to test for any specific pairwise differences. Main effects are discussed only when all pairwise treatment comparisons within that parameter were significant. Additionally, two-way ANCOVA with fat- and fat-free mass as co-variates was performed for total EE and energy intake to statistically adjust for body composition differences between sexes. Adjusted marginal means and partial eta-squared values as approximations of effect size were calculated. Statistical significance was set at p<0.05.

RESULTS

20°C vs 30°C housing temperature produces divergent energy metabolism

In this investigation we utilized the well described capacity different ambient housing temperature to produce differences in mouse EE (19, 20). As expected from previous publications (19, 20, 21), utilizing indirect calorimetry we observed male and female 30°C mice to have lower baseline total EE (~40%), energy intake (~35%), and resting EE (~55%) compared to 20°C on LFD (Table1). Non-resting EE was not observed to be different between temperatures, while females were lower at both temperatures. Average daily respiratory quotient was not different across all conditions (Table 1), suggesting similar macronutrient utilization. Importantly, while female mice showed greater cage activity (All_Meters) at both temperatures (Table1), no effect of housing temperature within sex was observed. Prior to HFHS feeding, no difference in body weight or body composition were observed between mice housed at 20°C vs. 30°C, while females weighed less than males at both temperatures.

Table 1.

LFD Energy Metabolism & HFHS Anthropometrics

	20		30
	Male	Female	Male	Female
Baseline Energy Metabolism (1-Week LFD)
Total EE (kcal)	81.77 ± 1.34	74.51 ± 2.20††	48.79 ± 0.97%	44.80 ± 0.78%
Energy Intake (kcal)	92.21 ± 2.91	78.77 ± 3.15††	58.33 ± 1.22%	52.98 ± 1.41%
Resting EE (kcal)	56.13 ± 1.77	53.33 ± 1.69	25.58 ± 0.59%	23.80 ± 0.51%
Non-Resting EE (kcal)	24.24 ± 0.83	21.53 ± 0.72†	22.92 ± 0.48	21.00 ± 0.55†
Respiratory Quotient	0.860 ± 0.010	0.850 ± 0.008	0.861 ± 0.007	0.841 ± 0.008
All_Meters (m)	1839.8 ± 117.2	2149.4 ± 185.5†	2007.3 ± 97.8	2308.6 ± 108.9†
Body Weight (1-Week HFHS)
Day 0	24.21 ± 0.34	18.67 ± 0.31 †	25.26 ± 0.35	19.46 ± 0.24†
Day 7	26.91 ± 0.52	21.13 ± 0.48†	28.98 ± 0.51%%	21.76 ± 0.40†
Body Composition (1-Week HFHS) Fat Mass (g)
Day 0	1.74 ± 0.17	1.53 ± 0.08	2.47 ± 0.14%	1.84 ± 0.09†
Day 7	3.37 ± 0.32	2.13 ± 0.18†	5.18 ± 0.30%	3.28 ± 0.28†,%
Fat-free Mass (g)
Day 0	22.72 ± 0.43	17.21 ± 0.33†	22.79 ± 0.26	17.63 ± 0.24†
Day 7	23.54 ± 0.34	19.01 ± 0.38†	23.80 ± 0.27	18.48 ± 0.20†

All values expressed as mean ± SEM (n=11-16).

^%p<0.05 main effect of 20°C vs. 30°C.

^†p<0.05 main effect of male vs. female,

^††p<0.05 male vs. female within temperature by diet group.

Increased EE produces unique short-term HFHS-induced weight gain phenotype in female mice

Interestingly, baseline EE did not impact weight gain in female mice; whereas, HFHS-fed male mice housed at 30°C gained more weight compared to 20°C males and 30°C females (Figure 1A). Next, we utilized qMRI to assess whether baseline differences in EE influenced HFHS-induced changes in body composition. Values for fat- and fat-free mass at the initiation and end of HFHS feeding are presented in Table 1. The lower EE associated with 30°C housing resulted in greater increases in fat mass on HFHS compared to 20°C, while females gained less fat mass compared to males at either temperature. (Figure 1B). Strikingly, fat-free mass increased ~2.6-fold more in 20°C female mice on HFHS compared to 20°C male or 30°C female (Figure 1C). To further highlight the interaction of sex and temperature in short-term weight gain, Figure 1D displays the combined fat and fat-free mass components of the one week change in body weight. This representation highlights the increased proportion of HFHS-induced weight gain comprised of fat-free mass in 20°C female mice (~80%), compared to all other groups (~30%). Together, these data demonstrate an interaction of sex and temperature to impact diet-induced weight gain and changes in body composition, particularly, in females.

Figure 1: Higher energy metabolism during 20°C housing changes composition of weight gained in female mice during one week HFHS feeding.

Body weight and body composition analysis utilizing qMRI was performed before and after 7-days of HFHS feeding. A) One week change in body weight (n=10-16). The difference in the initial and final values during the one week HFHS are displayed as (B) change in fat mass (FM) and (C) fat-free mass (FFM). D) One week change in body weight presented as change in FM and FFM. Values are means ± SEM. n=10-16. % p<0.05 main effect of 20°C vs. 30°C, † p<0.05 main effect of sex.

Short-term HFHS weight gain is not associated with energy balance or energy intake coupling in female mice

From the indirect calorimetry data, we assessed whether the observed sex-specific HFHS-induced weight gain phenotypes were associated with energy balance and coupling of energy intake to EE. Interestingly, energy balance during short-term HFHS feeding did not mirror the observed weight gain but was associated with change in fat mass (Figure 2A). Further, energy balance during HFHS was more positive in 30°C in male and female mice, while female mice had less positive energy balance compared to males regardless of temperature. The primary components of energy balance – total EE & energy intake – are presented in Figure 2B & 2C, respectively. As with LFD baseline, total EE was lower during HFHS in 30°C mice compared to 20°C for both sexes, and 30°C females were lower than males. Notably, differences in total EE due to temperature were not associated with any differences in activity level or activity EE during HFHS (Supplemental Figure 1). Energy intake was lower in 30°C mice regardless of sex, and females at both temperature had lower intake during the 1-week HFHS.

Figure 2: Female mice have reduced energy balance and improved coupling of energy intake to energy expenditure regardless of housing temperature.

Indirect calorimetry was utilized to assess energy metabolism in male and female C57BI/6J mice during 7-days of HFHS (n=8-12). A) Energy balance was calculated as difference in energy intake and total EE for each mouse during each diet exposure. B) Total EE was determined as the sum of the 24hr EE during 7-days HFHS. C) Energy intake during each dietary exposure was determined as the sum of food intake (g) times the energy density (kcal/g) for each diet. D) Energy flux was calculated as the sum of one week total EE and energy intake. E) Percent energy coupled represents the coupling of energy intake to EE and was calculated as 1 minus the energy balance divided by energy flux. Values are means ± SEM. % p<0.05 main effect of 20°C vs. 30°C, † p<0.05 main effect of sex, †† p<0.05 male vs. female within temperature, %% p<0.05 20°C vs. 30°C within sex.

Increased EE has been proposed as a mechanism to increase energy flux – sum of total EE and energy intake – which improves energy balance through greater pairing of energy intake to EE (3, 4, 5). Energy flux was lower in HFHS fed 30°C mice compared to 20°C for both sexes (Figure 2D). Due largely to the observed lower energy intake in female mice, energy flux was lower at both temperatures in females compared to males. In Figure 2E, the efficiency of coupling of energy intake to EE was calculated as 1 minus the energy balance divided by energy flux (percent energy coupled). Male and female mice housed at 30°C have reduced energy coupling during 1-week HFHS compared to 20°C. Further, while 20°C female mice tended to have greater energy coupling compared to males, only the 30°C females were observed to have significantly greater energy coupling compared to 30°C males. Together, these data support the hypothesis that increased EE can improve energy balance through increased energy flux and coupling of energy intake to EE. Further, these data highlight that energy balance is a powerful predictor of acute diet-induced fat mass gain in both sexes but does not predict weight gain in female mice.

Female mice have greater total EE following ANCOVA

The observed sex-specific differences in HFHS-induced weight gain and changes in body composition suggested that the differences in body size may be impacting the energy metabolism outcomes. Co-variate analysis was performed to assess the effect of differences in the components of body weight (fat- and fat-free mass) on the interpretation of total EE and energy intake. Adjusted estimated marginal means and estimate of co-variate effect size, partial eta-squared values, are shown in Table 2. Following ANCOVA adjustment for differences in fat- and fat-free mass, females had higher total EE at both temperatures compared to males, while total EE in 20°C mice remained higher than 30°C as observed in absolute total EE data (Figure 2B). Importantly, fat mass was not a significant co-variate of total EE; while fat-free mass was significant and showed a moderate effect size (partial eta-squared – 0.38) on total EE. During ANCOVA of energy intake neither fat- or fat-free mass was found to be a significant co-variate, therefore no data is presented. These findings suggest that the smaller female mice have greater inherent EE per unit fat-free mass which is associated with reduced energy balance and improved energy coupling.

Table 2.

HFHS Energy Expenditure Component Analysis & ANCOVA

	20°C		30°C
	Male	Female	Male	Female
1-Week HFHS Energy Expenditure Component Analysis
Resting EE (kcal)	67.41 ± 1.71	66.44 ± 2.08	31.69 ± 0.83%	27.97 ± 0.86%
Non-Resting EE (kcal)	19.49 ± 0.72	18.69 ± 0.67	22.00 ± 0.67%	20.74 ± 0.50%
Weight Adjusted Energy Expenditure
TEE (kcal/day, co-variate(s) Fat- & Fat-Free Mass)	76.66 ± 2.12	93.83 ± 2.21†	45.21 ± 2.54%	58.86 ± 2.45%,†
Co-variate Effect Size	Co-variate(s)	p-value	Partial Eta2
	Fat-Free Mass	p<0.001	0.388
	Fat Mass	p=0.261	0.032

All resting- and non-resting EE values expressed as mean ± SEM (n=10 - 12). All ANCOVA values expressed as estimated marginal mean ± SEM (n=10 - 12).

^%p<0.05 main effect of 20°C vs. 30°C,

^†p<0.05 main effect of male vs. female. Estimated effect size of the co-variates for ANCOVA analysis are presented as partial eta squared.

Reduced EE of 30 °C housing is associated with impaired metabolic flexibility

Obese human subjects do not increase fat utilization to the same extent as lean following a high-fat diet (23), and this lack of metabolic flexibility to acute high-fat diet is predictive of future weight gain (24). To assess whether differences in baseline EE impacted substrate utilization and the capacity to adapt substrate utilization during short-term HFHS, we determined daily respiratory quotient. As described above, respiratory quotient was not different on LFD (Table 1). During HFHS, daily average respiratory quotient was higher in 30°C males compared to both 20°C males and 30°C females (Figure 3A), suggesting lower fat utilization. No difference was observed between females. We calculated metabolic flexibility - the capacity to adapt substrate utilization based on changes in substrate availability - as the difference in average daily respiratory quotient during HFHS and LFD (Figure 3B) (25, 26). 30°C mice are less metabolically flexible during short-term HFHS as revealed by the smaller HFHS diet-induced reduction in daily average respiratory quotient compared to 20°C mice. However, 30°C female mice had greater HFHS diet-induced changes in respiratory quotient compared to males. Figure 3C shows the average change in daily respiratory quotient during the transition from LFD to HFHS. The figure highlights the rapid respiratory quotient decrease in all groups and slower, more transient response of the 30°C mice. These data show greater baseline EE results in increased metabolic flexibility during short-term HFHS, which is associated with improved energy balance, coupling of energy intake to EE, and reduced fat mass gain.

Figure 3: Decreased energy metabolism in mice housed at 30°C produces reduced metabolic flexibility to short-term high-fat/high-sucrose diets.

A) Indirect calorimetry was utilized to determine average daily respiratory quotient (RQ). Metabolic flexibility to HFHS feeding was assessed as the B) the difference in average daily RQ during HFHS feeding and average daily RQ during LFD feeding. C) RQ change from LFD RQ during each day of the HFHS exposure. Values are means ± SEM. n=10-16. + p<0.05 main effect of LFD vs. HFHS, % p<0.05 main effect of 20°C vs. 30°C, †† p<0.05 male vs. female within temperature by diet group.

Sex and baseline EE interact in the non-shivering thermogenic response to short-term HFHS feeding

Diet-induced non-shivering thermogenesis is the centrally-mediated, adaptive capacity to increase EE in response to increases in energy intake, and is proposed as a compensatory mechanism for limiting increased energy balance during transitions to energy dense diets (27, 28). Diet-induced non-shivering thermogenesis is calculated as the difference in HFHS and LFD resting EE (diet-induced resting EE). Interestingly, the lower baseline EE of 30°C mice was associated with lower diet-induced non-shivering thermogenesis in both male and female mice compared to 20°C Figure 4A). Additionally, 20°C females had greater diet-induced non-shivering thermogenesis compared to males. Figure 4B depicts the daily increase in resting EE due to HFHS feeding and demonstrates the rapid and sustained responses observed across the 7-day intervention. In Figure 4C, the diet-induced non-shivering thermogenesis is presented as the percent of total EE to highlight the relative contribution of this adaptive response on overall EE during HFHS. While, no difference is observed between male mice, female mice show a divergent response with 30°C mice being higher and 20°C being lower than the males. Importantly, these non-shivering thermogenesis findings are not associated with the thermic effect of food (Supplemental Figure 1), which necessarily closely matched food intake (Supplemental Figure 2) based on the calculation method. Diet-induced changes in resting EE could result in alterations in the other major component of total EE, non-resting EE. Diet-induced changes in non-resting EE were lower in 20°C mice compared to 30°C, with males having the greatest decrease. Interestingly, very little change in non-resting EE was observed mice 30°C (Figure 4D). These data demonstrate that differences in baseline EE produces divergent adaptive thermogenic response to HFHS, particularly in females, which is associated with the observed HFHS-induced gains in fat-free mass.

Figure 4: Baseline energy expenditure and sex interact impacting high-fat/high-sucrose-induced non-shivering adaptative thermogenesis.

HFHS-induced nonshivering thermogenesis was calculated as the diet-induced change in A) resting EE as the difference of one week HFHS values and the one week LFD values. B) Daily HFHS-induced change in resting EE above 1-week average LFD resting EE. C) Percentage of total EE comprised of diet-induced resting EE during HFHS. D) HFHS-induced change in non-resting EE. Values are means ± SEM. n=10-12. % p<0.05 main effect of 20°C vs. 30°C, †† p<0.05 male vs. female within temperature, %% p<0.05 20°C vs. 30°C within sex.

DISCUSSION

Optimal maintenance of energy balance and body weight is putatively achieved at higher levels of EE via greater energy flux and improved coupling of energy intake to energy demand (3, 4, 5). However, the direct impact of EE on energy balance and weight gain is complicated by potentially confounding factors produced by the common experimental tools (e.g. – physical activity, chemical uncouplers, etc.). Also, there are limited studies examining the interaction of EE and metabolic adaptations on weight gain regulation between sexes. Herein, we utilized differing ambient housing temperatures (20 vs. 30°C) to modulate EE in male and female mice as a novel experimental tool to assess the independent role of EE on diet-induced weight gain and adaptation of energy metabolism.

Early work of Mayer and Edholm independently demonstrated in rodents and humans that energy intake is highly regulated at higher levels of EE, measured as oxygen consumption or increased physical activity (6, 7, 29, 30). This work has been extended to describe two zones. The regulated zone occurs when energy intake and greater EE are highly coupled, and the unregulated zone occurs at lower EE where the relationship between EE and energy intake become uncoupled (unregulated zone) (31). Recently, increased coupling of energy intake to EE and associated improved energy balance was observed during stepwise increases in physical activity and EE (32). Further, higher EE was associated with better pairing of energy intake to energy demand and increased fat oxidation during either energy balance feeding or overfeeding (33). Here we show that utilizing different housing temperatures produces divergent EE in mice independent of physical activity. Leveraging this effect shows that greater EE resulted in increased energy flux, less positive energy balance, and greater coupling of energy intake to EE during HFHS feeding in both males and females. However, female mice have less positive energy balance and greater energy intake coupling compared to males at either EE level. The previously cited studies did not investigate whether the increases in energy intake coupling and energy balance associated with higher EE were different in males versus females (32, 33). Future work is necessary to specifically investigate the observed sex differences in coupling of energy intake to EE and the responses in energy metabolism to increased EE. Together, these findings confirm and support that higher EE aids in improving energy balance regulation and body weight homeostasis while also revealing critical differences between sexes.

Previous studies have consistently revealed sexual dimorphism for weight gain and anatomical location of fat mass gains during hypercaloric conditions. In this study, female mice at higher EE gained less fat mass despite having the same change in body weight as lower EE females. Interestingly, the higher EE female mice (20°C) were the only group to gain more fat-free mass than fat mass during HFHS feeding. Thus, the fat mass gain in the 20°C females represented only 20% of the body weight gain, compared to ~70% of the weight gain for the other three groups. Typically, between 60 – 80% of weight gained during most periods of positive energy balance is fat mass (34). Recent human data showed that individuals with reduced EE and reduced coupling of energy intake to EE had increased fat mass gain over time (35). While others have observed an increase in fat-free mass with a decrease in fat mass in overweight women following 3 months of physical training, which was not observed in male subjects (36). These data highlight a unique metabolic response to short-term HFHS feeding in females with higher EE, and the need for more studies on how differences in EE producing sex-specific responses in body composition and weight gain during increased energy intake.

The capacity to adapt energy metabolism through changes in substrate utilization and non-shivering thermogenesis likely drives susceptibility for obesity and metabolic disease and are thus a focus of treatment (25, 27, 28, 37, 38). The capacity to adapt substrate utilization from fat (low respiratory quotient) to carbohydrate (high respiratory quotient) during the fed-to-fasted transition was coined metabolic flexibility (39), and was demonstrated to be reduced in obese compared to lean subjects. The definition of metabolic flexibility has expanded to encompass adaptation of substrate utilization in response to high-fat diet feeding (26), and human data demonstrates that a lower ratio of fat to carbohydrate oxidation or the inability to increase fat utilization during high-fat overfeeding predicts future weight gain (24, 40). Previously, we have observed increased weight gain in a rodent model with reduced metabolic flexibility and reduced fat oxidation during short-term HFHS (41). In this study, while 7-days of HFHS feeding resulted in reduced average daily respiratory quotient in all mice, increased baseline EE was associated with more extensive lowering of average daily respiratory quotient following the HFHS transition. Additionally, females demonstrated greater metabolic flexibility to short-term HFHS feeding. The causative role of metabolic flexibility in the development and progression of obesity and metabolic disease is still unclear, however, these data support an association of increased EE with improved metabolic flexibility and reduced gains in adiposity during HFHS feeding.

Non-shivering adaptive thermogenesis is the centrally-mediated increase in heat production in response to either cold or diet stimuli (28). The recent discoveries of thermogenic adaptations in both adipose depots and skeletal muscle, many laboratories have explored non-shivering thermogenesis as a target for obesity treatment (reviewed in (27, 28, 37)). Rodent work has highlighted the potential importance of diet-induced non-shivering thermogenesis through findings of increased susceptibility or protection for diet-induced weight gain following knockout (42, 43, 44, 45) or overexpression (46) of genes involved in various thermogenic pathways. In this study we observed that HFHS feeding produced rapid and sustained increases in non-shivering thermogenesis (diet-induced changes in resting EE). Similar findings in human subjects have been observed as change in sleeping EE from 1 day (47, 48), up to 56 days (48), of overfeeding. Importantly, we observed that mice with higher baseline EE have a greater diet-induced non-shivering thermogenic response. Which was further increased in 20°C female mice, where diet-induced non-shivering thermogenesis represented approximately twice the proportion of total EE (~14%) as 30°C females (~7%). Interestingly, diet-induced non-shivering thermogenesis was closely associated with the observed changes in fat-free mass in female mice during short-term HFHS. Suggesting the increased non-shivering thermogenesis of the 20°C female mice is due to the greater gain in fat-free mass that was only found in these mice. However, due to the experimental design, it is not possible to determine the daily change in fat-free mass across the HFHS feeding. Together, these findings are the first to show that EE and sex interact to alter short-term thermogenic response to an energy dense diet.

Limitations

Despite the wide breadth of energy metabolism data collected during these experiments, several potentially confounding factors and limitations should be considered. First, while the C57Bl/6J mouse strain is extensively utilized in obesity studies, the use of other inbred and outbred mouse strains for future studies, particularly related to assessment of sex differences, are necessary. Second, the increased EE of mice at sub-thermoneutral ambient temperatures is primarily mediated by centrally regulated non-shivering thermogenic pathways in adipose and skeletal muscle. These pathways differ from those potentially activated through increased physical activity or exercise and thus findings may not be directly comparable. Further, from the data herein we cannot determine the magnitude of activation of the different non-shivering thermogenic tissues, which could potentially differ by baseline EE or sex. Also, it is not clear how the use of different ambient housing temperatures impacts other metabolic homeostatic systems (e.g. – glucose homeostasis), which could influence the outcomes of the current study. Third, the assessment of diet-induced weight gain in 9 – 11 week old mice could be confounded by the previously observed dependence of weight gain on age of diet initiation and sex (49). Fourth, previous mouse work has demonstrated that male mice defend different body core temperatures at different ambient temperatures (19, 50). The lack of these thermal biology data prevents a comprehensive dissection of sexual differences in energy metabolism, and the impact on metabolic responses to short-term HFHS feeding. Also, no determination of protein utilization for energy is possible under the experimental conditions. Finally, the lack of fecal energy excretion data prevents the calculation of net energy intake during both the LFD and HFHS feeding, and potentially confounds the calculation of EB.

Conclusions

Because the prevention of weight gain is putatively easier than weight loss (3), it is critical that mechanisms underlying the protection or susceptibility to weight gain during short-term hypercaloric conditions be elucidated. This study used ambient temperature to determine if higher or lower EE in male and female mice would change metabolic adaptations and weight gain during a transition to acute HFHS feeding. Herein, these data support that higher levels of EE enhance coupling of energy intake to energy demand, producing lower positive energy balance with reduced weight- and adiposity gain during short-term HFHS. Additionally, these data demonstrate that EE level plays an important role in determining weight gain and body composition changes in female mice exposed to acute HFHS feeding. Further, it demonstrates the utility of different mouse housing temperature as a tool to study the independent role of EE in metabolic phenotypes. Finally, these findings have further significance when one considers that the vast majority of obesity research conducted in mice occurs at sub-thermoneutral housing, near, the 20°C temperature.

Study Importance.

What is already known about this subject?

• Increased energy expenditure associated with physical activity hypothetically prevents or reduces short-term diet-induced weight gain by increasing coupling of energy intake to expenditure.

What are the new findings in your manuscript?

• Supports energy flux hypothesis, where increased energy expenditure results in greater energy flux which improves energy balance via enhanced coupling of energy intake to energy demand.

• Increased energy expenditure in female mice results in predominantly fat-free mass gain during short-term HFHS feeding.

• Increased energy expenditure is associated with greater metabolic flexibility to HFHS and diet-induced non-shivering thermogenesis, particularly in female mice.

How might your results change the direction of research or the focus of clinical practice?

• Highlights the importance of energy expenditure levels in the prevention of weight gain and adiposity.

• First evidence that energy expenditure level may play a role in the composition of weight gained by females during acute HFHS feeding.