Renewed selection for heat loss in mice: Direct responses and correlated responses in feed intake, body weight, litter size, and conception rate1

Genetic and genomic analysis of oxygen consumption in mice

We estimated genetic parameters for oxygen consumption (OC), OC per metabolic body weight (OCMBW) and body weight at three through 8 weeks of age in divergently selected mice populations, with an animal model considering maternal genetic, common litter environmental and cytoplasmic inheritance effects. Cytoplasmic inheritance was considered based on maternal lineage information. With respect to OC, estimated direct heritability was moderate (0.32) and the estimated proportion of the variance of cytoplasmic inheritance effects to the phenotypic variance was very low (0.01), implying that causal genes for OC could be located on autosomes. To assess this hypothesis, we attempted to identify possible candidate causal genes through selective signature detection with the results of pooled whole-genome resequencing using pooled DNA samples from high and low OC mice. We made a list of possible candidate causal genes for OC, including those relating to electron transport chain and ATP-binding proteins (Ndufa12, Sdhc, Atp10b, etc.), Prr16 encoding Largen protein, Cry1 encoding a key component of the circadian core oscillator and so on. The results, although careful interpretation must be required, could contribute to elucidate the genetic mechanism of OC, an indicator for maintenance energy requirement, and therefore feed efficiency.

Selection for lower residual feed intake in mice is accompanied by increased body fatness and lower activity but not lower metabolic rate

Context Mice bred to be genetically different in feed efficiency were used in this experiment designed to help improve our knowledge of the biological basis of variation in feed efficiency between individual animals. Aims This experiment used mice to explore the metabolic basis of genetic variation in feed efficiency in the growing animal. Methods Mice bred to differ in residual feed intake (RFI) recorded over a postweaning test were used. After 11 generations of divergent selection, mice in groups were tested for RFI from 6 to 8, 8 to 10, and 10 to 12 weeks of age, and measured for traits describing the ability to digest feed, body composition, protein turnover, basal and resting metabolic rate, and level of activity. Key results Compared with the low-RFI (high efficiency) line mice, high-RFI mice consumed 28% more feed per day over their RFI-test, were no heavier, were leaner (16% less total fat per unit of bodyweight), did not differ in the fractional synthesis rate of protein in skeletal muscle or in liver, and had similar basal metabolic rates at 33°C. On an energy basis, the selection lines did not differ in energy retained in body tissue gain, which represented only 1.8% of metabolisable energy intake. The remaining 98.2% was lost as heat. Of the processes measured contributing to the higher feed intake by the high-RFI mice, 47% of the extra feed consumed was lost in faeces and urine, activity was 84% higher and accounted for 24%, the cost of protein gain was 6% higher and accounted for 2%, and the energy cost of digesting and absorbing the extra feed consumed and basal heat production could have accounted for 11 and 15% each. Conclusions Selection for low RFI (high efficiency) in mice was accompanied by an increase in body fat, an improvement in the process of digestion, a lower rate of protein turnover and a much lower level of activity. Selection did not result in major change in basal metabolic rate. Implications This experiment with mice provided new information on the biological basis of genetic differences in feed efficiency. The experiment investigated the relative importance of major energy-consuming metabolic processes and was able to quantify the responses in protein turnover and level of activity, being responses in energy-consuming processes that have proven difficult to quantitatively demonstrate in large farm animals.

Exposure to hot temperatures during lactation stunted offspring growth and decreased the future reproductive performance of female offspring

Among the important aspects of climate change, exposure to high temperatures (heat waves) is rapidly emerging as an important issue, in particular for female mammals during lactation. High temperatures adversely impact ability to dissipate heat, which has negative effects on reproductive output. The cumulative effects on growth of F1 offspring after weaning and future reproductive performance of offspring remain uncertain. In this study, the F1 mice that weaned from mothers lactating at 21°C and 32.5°C were housed at 21°C from day 19 till 56 of age; during which food intake and body mass were measured. The F1 adult females that had been weaned at the two temperatures were bred and then both exposed to 32.5°C during lactation. Energy intake, milk output and litter size and mass were determined. The F1 adults weaned at 32.5°C consumed less food and had lower body mass than their counterparts weaned at 21°C. Several visceral organs or reproductive tissues were significantly lower in mass in F1 weaned at 32.5°C than at 21°C. The exposure to 32.5°C significantly decreased energy intake, milk output and litter mass in F1 adult females during lactation. The F1 adult females weaned at 32.5°C produced less milk and raised lighter pups than those previously weaned at 21°C. The data suggest that transient exposure to hot temperature during lactation has long-lasting impacts on the offspring, including stunted growth and decreases in future reproductive performance when adult. This indicates that the offspring of females previously experiencing hot temperatures have a significant fitness disadvantage.

Long-Term Trait Consistency in Mice Selected for Swim-Induced High Aerobic Capacity

The majority of studies show that metabolic rates are usually repeatable at the individual level, although their repeatabilities tend to decline with time and to be strongly affected by physiological changes. Changes in individual repeatabilities may therefore affect putative differences between experimental groups or populations. This problem is particularly relevant to artificial selection experiments that apply the selection protocol at early life stages, running the risk of a poor correlation of the trait with itself throughout the life cycle of individuals. Moreover, significant physiological changes (e.g., induced by reproduction) may affect traits under selection and therefore their postreproductive differentiation between selected lines. Here, using a unique animal model-mice from four lines selected for [Formula: see text] during swimming in 25°C water and four random-bred control (reference) lines-we analyzed the long-term consistency of aerobic capacity as well as postswim hypothermia in primiparous and nonreproducing females at 12, 25, and 29 wk of age. Our results show that significant between-line type divergence in [Formula: see text] and hypothermia persists over time and is only weakly affected by past reproduction. Furthermore, both traits are also repeatable within lines at the individual level. More generally, our results suggest that past reproduction events are unlikely to significantly affect between-population and between-individual differences in [Formula: see text] and related traits.

Selection for high and low oxygen consumption-induced differences in maintenance energy requirements of mice

Maintenance energy requirements (MER) of mice selected for high (H) or low (L) oxygen consumption (OC) were compared. Forty-four mice from H and L OC lines were weaned at 3 weeks and divided into four experimental groups: group A were sacrificed at 4 weeks; group B were fed ad libitum, and groups C and D were fed 2.8 and 2.4 g/day, respectively, from 4 to 8 weeks of age. Groups B-D were sacrificed at 8 weeks. Chemical components were estimated for all groups. MER was estimated using a model that partitioned metabolizable energy intake into that used for maintenance, and protein and fat deposition. The feed conversion ratio for the B group was significantly higher in the H than in the L line. Feed intake for metabolic energy content per metabolic body size was significantly also higher in the H line, whereas accumulated energy content per metabolic body size was significantly higher in the L line. MER of the H line was greater than that of the L line (P < 0.10). These results suggest that selection for H or L OC produced differences in chemical components, feed efficiency, and MER between the H and L lines.

Effect of reproduction on the consistency of the between-line type divergence in laboratory mice selected on Basal metabolic rate

Artificial selection experiments are an effective tool for testing evolutionary hypotheses, because they allow one to separate genetic and environmental variances of the phenotype. However, it is unclear whether trait divergence typically selected early in life persists over an animal's life and altered physiological states, such as reproduction. Here we analyzed the long-term consistency of the between-line type divergence in basal metabolic rate (BMR) selected at 12 wk of age in laboratory mice. We measured BMR in nonreproducing and reproducing females at the age of 22 wk and then at 27 wk of age. Our results show that within both the reproducing group and the control group, the between-line type separation in BMR is consistently retained over time and reproductive status. Metabolically active internal organs (heart, liver, kidneys, and small intestine) also consistently differed in size between the two line types with no significant long-term effect of reproduction. The observed consistency of the between-line type divergence in BMR suggests the existence of the persistent effect of the selection on metabolic traits applied early in life. Moreover, BMR variation achieved by means of artificial selection is considerably higher than that found in natural/unmanipulated populations. The latter may therefore be characterized by insufficient variance to statistically resolve correlations involving BMR.

Selection for high and low oxygen consumption altered hepatic mitochondrial energy efficiency in mice

Selection for high (H) and low (L) oxygen consumption (OC) as an indirect estimation of maintenance energy requirement was determined. Feed intake and body weight were measured and feed conversion ratio (FCR) of 4-8-week-old mice was calculated. Respiratory activity of hepatic mitochondria was measured at 12 weeks. Total feed intake (H: 103.74 g, L: 97.92 g, P < 0.01), daily feed intake (H: 3.70 g/day, L: 3.50 g/day, P < 0.01) and FCR (H: 18.79, L: 15.50, P < 0.01) were significantly different between lines. The line by sex interaction was significant for FCR. No line differences were observed in males; and the FCR of the H line was greater than in the L line in females. H line mice had the highest hepatic mitochondrial respiratory activity in state 2 (P < 0.01), the highest uncoupled respiratory rate of mitochondria in the presence of an uncoupling agent (P < 0.001), and the mitochondrial proton leak. The adenosine diphosphate/ O ratio was highest in the L line (P < 0.05). This suggests that the selection for high and low OC induced differences in basal mitochondrial respiration and basal metabolism, resulting in difference in FCR between H and L lines.

Life cycle biological efficiency of mice divergently selected for heat loss

Divergent selection in mice for heat loss was conducted in 3 independent replicates creating a high maintenance, high heat loss (MH) and low maintenance, low heat loss (ML) line and unselected control (MC). Improvement in feed efficiency was observed in ML mice due to a reduced maintenance energy requirement but there was also a slight decline in reproductive performance, survivability, and lean content, particularly when compared to MC animals. The objective of this study was to model a life cycle scenario similar to a livestock production system and calculate total inputs and outputs to estimate overall biological efficiency of these lines and determine if reduced feed intake resulted in improved life cycle efficiency. Feed intake, reproductive performance, growth, and body composition were recorded on 21 mating pairs from each line × replicate combination, cohabitated at 7 wk of age and maintained for up to 1 yr unless culled. Proportion of animals at each parity was calculated from survival rates estimated from previous research when enforcing a maximum of 4, 8, or 12 allowed parities. This parity distribution was then combined with values from previous studies to calculate inputs and outputs of mating pairs and offspring produced in a single cycle at equilibrium. Offspring output was defined as kilograms of lean output of offspring at 49 d. Offspring input was defined as megacalories of energy intake for growing offspring from 21 to 49 d. Parent output was defined as kilograms of lean output of culled parents. Parent input was defined as megacalories of energy intake for mating pairs from weaning of one parity to weaning of the next. Offspring output was greatest in MC mice due to superior BW and numbers weaned, while output was lowest in ML mice due to smaller litter sizes and lean content. Parent output did not differ substantially between lines but was greatest in MH mice due to poorer survival rates resulting in more culled animals. Input was greatest in MH and lowest for ML mice for both offspring and parent pairs, consistent with previous results in these lines. Life cycle efficiency was similar in MC and ML mice, while MH mice were least efficient. Ultimately, superior output in MC mice slightly outweighed the lower inputs in ML animals resulting from decreased maintenance energy requirements. Therefore, selection to reduce maintenance energy requirements may be more useful in terminal crosses or in a selection index to reduce possible negative effects on output, especially reproductive performance.

Lifetime reproductive performance and survival analysis of mice divergently selected for heat loss

Divergent selection for heat loss was implemented in mice creating maintenance high (MH) and low maintenance (ML) lines and an unselected control (MC) in 3 independent replicates. Mice from the ML line have improved feed efficiency, due to decreased maintenance energy requirement, but there is potential for a correlated decline in reproductive performance and survivability. Number fully formed (NFF), number born alive (NBA), number weaned (NW), litter weaning weight (LWW), pup weaning weight (PWW), fraction alive at birth (FAB), fraction alive at weaning, and birth interval were recorded at every parity on 21 mating pairs from each line × replicate combination cohabitated at 7 wk of age and maintained for up to 1 yr. Traits were summed over parities to evaluate lifetime production. Pairs were culled due to death or illness, no first parity by 42 d cohabitation, 2 consecutive litters with none born alive, 3 consecutive litters with none weaned, 42 d between parities, or average size of most recent 2 litters less than half the average of first 3 litters. Survival probabilities were produced and evaluated for each line and used to calculate mean number of parities using a Markov-chain algorithm assuming a maximum of 4, 6, 8, 10, or 12 parities or 1 yr. Line was insignificant for all litter traits while NFF, NW, and FAB decreased with parity (P < 0.05) and PWW tended to increase (P < 0.07). The MC mice had higher lifetime NW, LWW, and PWW (P < 0.04). Birth interval showed that MH mice had increasingly larger intervals while remaining the same in ML mice (P < 0.01). In the survival analysis, MC mice had the greatest survival rates overall, but ML mice had the greatest rates in the period up to 5 parities while MH mice had the greatest rates in later parities. This resulted in greater mean number of parities for ML mice up to maximum of 8 parities and higher means for MH mice when the maximum number of allowed parities was 10 or higher. Reproductive performance was not substantially affected by changing maintenance energy requirements. The ML animals appear to survive well in early parities and produce more parities when a low number of maximum parities is enforced, but this benefit declines in later parities and MH animals survive better and increase mean number of parities when turnover rates are low. Therefore, selection for low maintenance animals may be beneficial for systems desiring a short generation interval but less so for systems desiring longevity.

Body composition and feed intake of reproducing and growing mice divergently selected for heat loss

Changes in maintenance energy requirements and in feed efficiency have been achieved by divergent selection for heat loss in mice in 3 replicates, creating high heat loss, high maintenance (MH) and low heat loss, low maintenance (ML) lines and an unselected control (MC). However, feed intake has mainly been measured in mature animals and not during growth or reproduction. Additionally, there is evidence that reducing maintenance energy will increase fat content, an undesirable result. To evaluate if selection has altered body composition and lifecycle feed intake, mating pairs were continuously mated and maintained for up to 1 yr unless culled. Offspring pairs were sampled from each line at each parity and maintained from 21 to 49 d of age. Feed intake was recorded for mating pairs throughout the year and on offspring pairs. Body weight was recorded on all animals at culling as well as percent fat, total fat, and total lean, measured by dual X-ray densitometry. Average daily gain was also recorded for offspring. Energy partitioning was achieved using 2 approaches: Approach I regressed energy intake of the pair on sum of daily metabolic weight and total gain to obtain maintenance (bm) and growth (bg) coefficients for each line, replicate, feeding period, and sex (offspring pairs only); Approach II calculated bm for each pair assuming constant energy values for lean and fat gain. Energy coefficients and body composition traits were evaluated for effect of selection (MH vs. ML) and asymmetry of selection ([MH + ML]/2 vs. MC). Both MC mating and offspring pairs tended to have greater BW than the average of the selection lines (P < 0.08). Males of offspring pairs weighed more than females (P < 0.01), while females of mating pairs weighed more than males (P < 0.01). Line was insignificant (P > 0.15) for body composition traits. Using Approach I, MH mice had a greater bm than ML mice for mating pairs (P = 0.03) but not offspring pairs (P = 0.50). For Approach II, MH had a greater bm than ML mice for both mating (P = 0.01) and offspring pairs (P = 0.01). The effect of selection for heat loss on body composition was smaller than previously reported and unlikely to outweigh the benefit of reduced feed intake, which was shown to be maintained throughout an entire lifecycle that included reproducing animals. Additionally, the reduction in energy intake seems primarily due to reduced maintenance energy costs, validating the success of the selection procedure.

Locomotor activity and body temperature in selected mouse lines differing greatly in feed intake

Locomotor activity, body temperature, feed intake, and BW were measured on 382 mature male mice sampled from lines previously selected (25 generations) for either high (MH) or low (ML) heat loss and an unselected control (MC). Animals were from all 3 independent replicates of the 3 lines and across 4 generations (68 through 71). Locomotor activity and body temperatures were obtained using implanted transmitters with data collection over 4 d following a 3-d postsurgery recovery period. Data were collected every minute and then averaged into 30-min periods, thus providing 192 data points for each mouse. Least-squares means for feed intake adjusted for BW (Feed/BW, feed·BW(-1)·d(-1), g/g) were 0.1586, 0.1234, and 0.1125 (±0.0022) for MH, MC, and ML, respectively, with line being a highly significant source of variation (P < 0.0003). Line effects for locomotor activity counts, transformed to the 0.25 power for analysis, were significantly different, with MH mice being 2.1 times more active than ML mice (P < 0.003); MC mice were intermediate. Differences in body temperature were significant for both line (P < 0.03) and day effects (P < 0.001), with a 0.32°C difference between the MH and ML lines. Fourier series analysis used the combined significant periodicities of 24, 18, 12, 9, 6, and 3 h to describe circadian cycles for activity and body temperature. All 3 lines expressed daily peaks in body temperature and locomotor activity ∼3 h into darkness and ∼2 h after lights were turned on. There was a stronger relationship between locomotor activity and Feed/BW (P < 0.0001) than between body temperature and Feed/BW (P < 0.01); differences between lines in locomotor activity and body temperature explained 17% and 3%, respectively, of differences between lines in Feed/BW. Thus, line differences in locomotor activity contribute to line differences in maintenance, but approximately 80% of the differences between the MH and ML selection lines in Feed/BW remains independent of differences in locomotor activity.

Limits to sustained energy intake. XIII. Recent progress and future perspectives

Several theories have been proposed to explain limits on the maximum rate at which animals can ingest and expend energy. These limits are likely to be intrinsic to the animal, and potentially include the capacity of the alimentary tract to assimilate energy – the ‘central limitation’ hypothesis. Experimental evidence from lactating mice exposed to different ambient temperatures allows us to reject this and similar ideas. Two alternative ideas have been proposed. The ‘peripheral limitation’ hypothesis suggests that the maximal sustained energy intake reflects the summed demands of individual tissues, which have their own intrinsic limitations on capacity. In contrast, the ‘heat dissipation limit’ (HDL) theory suggests that animals are constrained by the maximal capacity to dissipate body heat. Abundant evidence in domesticated livestock supports the HDL theory, but data from smaller mammals are less conclusive. Here, we develop a novel framework showing how the HDL and peripheral limitations are likely to be important in all animals, but to different extents. The HDL theory makes a number of predictions – in particular that there is no fixed limit on sustained energy expenditure as a multiple of basal metabolic rate, but rather that the maximum sustained scope is positively correlated with the capacity to dissipate heat.