Mature male Sprague-Dawley (SD) and Long-Evans (LE) rats were instrumented with telemetry transmitters for measurement of heart rate (HR) and housed in room calorimeters for assessment of food intake and oxygen consumption (V̇o2) at standard laboratory temperatures (23°C) to examine physiological responses to caloric restriction (CR; 60% of baseline ad libitum calories for 2 wk) and refeeding. Ad libitum controls had stable food intake (84–88 kcal/day) and gained weight at rates of 3–4 g/day. Groups from both strains assigned to CR exhibited similar patterns of weight loss and reductions in V̇o2 and HR. Upon refeeding, SD rats exhibited a mild, transient hyperphagic response (1 day) accompanied by sustained suppression of V̇o2 and HR that remained evident 8 days after refeeding. In contrast, LE rats exhibited sustained daily hyperphagia that persisted 8 days after refeeding and was accompanied by a complete restoration of HR and V̇o2. The lower HR and V̇o2 observed during refeeding in SD rats were not due to reduced locomotor activity. The results reveal a strain-dependent divergent response to recovery from CR. We conclude that during recovery from CR, homeostatic stimulation of appetite or suppression of energy expenditure may occur selectively to restore body weight.
- indirect calorimetry
- energy homeostasis
caloric restriction (CR) produces a multifaceted set of adaptations that reduce energy expenditure and increase appetite in an attempt to defend body fat and body weight (29, 41, 44). Frequently, sexually mature but rapidly growing male rats are used to examine mechanisms of energy homeostasis. These animals are in positive energy balance, consuming excess calories that are used to increase both lean and fat mass at rates of weight gain often exceeding 25 g/wk. Thus growing Sprague-Dawley (SD) rats treated with CR (rats fed 60% of control caloric intake) lose very little weight and reach a fairly stable weight within a week, whereas ad libitum control rats continue to rapidly grow. As a result, the weight differences are due more to growth of controls than weight loss in treated rats. When SD rats are refed, this strain of rats exhibits a sustained reduction in energy expenditure that appears to contribute to restoration of fat mass (3, 7).
The primary purpose of this study was to examine physiological responses to CR in another commonly used rat strain. As the results indicate, we observed that outbred Long-Evans (LE) rats and SD rats exhibit similar physiological responses during CR but exhibit unexpectedly divergent strategies for normalization of weight during refeeding. The findings indicate that suppressed thermogenesis is not always observed after CR.
All experimental procedures and protocols were approved by the Institutional Animal Care and Use Committee at Florida State University. Male LE [(Crl:LE)BR; Charles River, Raleigh, NC] and SD rats [Crl:CD(SD)IGSBR; Charles River Laboratories] were purchased at 5–6 wk of age and individually housed in polycarbonate cages containing wood chip bedding at standard conditions [ambient temperature (Ta) = 23 ± 0.1°C, 12:12-h light-dark cycle]. They were provided ad libitum access to pelleted rat chow [LabDiet rodent diet 5001; physiological fuel value 3.3 kcal/g, 4.5% fat, 23.4% protein; estimated food quotient 0.914 (2)] and deionized water. At ∼8 wk of age, animals in experiment 1 were anesthetized with pentobarbital sodium (50 mg/kg ip, Nembutal sodium solution; Henry Schein, Melville, NY) and instrumented with a catheter in the descending aorta coupled with a sensor and transmitter (TA11PA-C40; Data Sciences, St. Paul, MN) for telemetric monitoring of blood pressure and heart rate as described previously (32). Animals recovered for at least 10 days after surgery before being studied.
Indirect Calorimetry and Behavioral Monitoring
Measurements of metabolism and monitoring of animal behavior were performed using previously published approaches (32). Rats were transferred to standard shoe box cages fitted with custom-made polycarbonate lids providing a near air-tight seal for continuous determination of oxygen consumption (V̇o2, in ml/min) and carbon dioxide production (V̇co2, in ml/min) (35, 47). Respiratory quotient (RQ; V̇co2/V̇o2) was derived from these measurements.
Each cage was sampled once every 4 min for 23 h. Resting V̇o2 was determined as the lowest 30 values (2 h) for both the dark and light phases. Averages for both the dark and light phases were also determined. To further examine the influence of CR and refeeding on V̇o2, we generated cumulative frequency plots using all 4-min bins (28). Energy expenditure was estimated using the Weir equation: energy expenditure (kcal/min) = 3.91(V̇o2) + 1.1(V̇co2)/1,000 (43). Energy expenditure was calculated separately for the dark and light phases. Because recordings were made for 23 h, the equivalent of an additional hour of average dark-phase energy expenditure was added to estimate total daily energy expenditure. Daily energy balance was estimated as the difference between caloric intake and caloric expenditure.
The shoe box cage was positioned on a custom-designed force platform to obtain quantification of locomotor activity. Rats lived in these cages for the entire 3- to 4-wk experimental protocol. They consumed powdered rat chow (LabDiet rodent diet 5001, from a bowl feeder) and deionized water. Feeding behavior was monitored using a photo beam sensor positioned across the opening of the feeder. The pattern of photo beam break was analyzed using a custom-written program to determine the number of meals consumed in dark and light phases. A meal was defined as at least 30 s of cumulative beam breakage, initiated by at least 5 s of beam breakage within a 1-min period and terminated by a period of at least 15 min with <5 s of beam breakage. Drinking behavior was monitored using a lickometer (38).
After recovery of animals from surgery and acclimation to the metabolic cages, baseline data were collected for at least 3 days, during which all animals were provided ad libitum access to food. The length of the baseline period varied slightly so that the body weight between groups of rats could be matched before CR was imposed. Groups of LE (LE-AL, n = 8) and SD (SD-AL, n = 8) rats were maintained on ad libitum access to food for 14 days, whereas other groups (LE-CR and SD-CR, n = 8 for each group) were subjected to 14 days of CR to 60% of average baseline food intake. Food and water intake and body weight of each rat were determined during a daily maintenance period that occurred 1 h before the beginning of the dark phase. At this time, daily cardiovascular and metabolic data were compiled and transferred for off-line processing. In the refeeding period, food was available ad libitum to all groups for 4 days, during which refeeding patterns and recovery kinetics in cardiovascular and metabolic variables were measured. After observing an unexpected sustained hyperphagia in the first four LE rats studied after CR, we extended the recovery period for subsequent SD and LE rats to 8 days.
To determine whether physiological differences in recovery from CR were associated with differences in body composition or leptin levels, we assigned additional LE and SD rats to AL, CR (14 days at 60% of AL controls), and refeeding (refed ad libitum for 4 days) groups (6 groups; n = 6 for each group). Rats were housed individually in standard shoe box cages at Ta = 23°C and fed pelleted rat chow (LabDiet rodent diet 5001) once daily at the onset of the dark phase for this study, as in experiment 1. Late in the light phase (generally associated with postabsorptive state), animals were anesthetized (65 mg/kg ip pentobarbital sodium) and decapitated for collection of trunk blood and determination of organ and visceral fat depot weights. Subcutaneous fat was estimated by weighing a 2-cm2 section of the lower abdominal skin (previously shaved) with attached fat (15). Serum leptin levels were determined by radioimmunoassay (Leptin RIA kit; Linco Research, St. Louis, MO).
Data Analysis and Statistics
The final hour of the light phase (during which daily chamber maintenance procedures were performed) was excluded from analysis, resulting in 12-h averages for the dark phase and 11-h averages for the light phase. SPSS 11.0 software was used for statistical analyses of dependent variables. Baseline strain differences were analyzed using an independent samples t-test. Changes from baseline were calculated for each rat and reported as responses to CR and refeeding. Treatment and strain effects on responses to CR and refeeding were statistically assessed using two-way ANOVA and Tukey's post hoc test. Significance levels of P < 0.05 were accepted.
There were generally few differences in baseline cardiovascular physiology, metabolism, and behavior between weight-matched male SD and LE rats (Table 1). At this time, SD rats were slightly but significantly older (SD: 79 ± 1 vs. LE: 72 ± 4 days; P < 0.01). At this age, both SD and LE rats were growing and in positive energy balance (Fig. 1B). Food and water intake (Table 1 and Fig. 1, C and D), mean arterial pressure (Table 1), and locomotor activity (Table 1 and Fig. 2, G and H) were similar. Resting V̇o2 and HR were significantly higher (∼15 beats/min; P < 0.05; Fig. 2, A and B) and the standard deviation of interbeat interval (SDIBI) was significantly lower in LE rats in both light and dark phases (P < 0.05; Table 1). In addition, LE rats engaged in greater amounts of light-phase ingestive behavior as measured by feeding and drinking activity (Table 1). This likely contributed to a significant elevation in total light-phase V̇o2 in LE rats.
Response to CR
AL groups gained weight at the rate of 3–4 g/day (Fig. 1A), exhibited stable caloric intake over the course of the study (Fig. 1C), and remained in positive energy balance (Fig. 1B). Although LE rats exhibited greater metabolic efficiency during the 2-wk period (Table 2), body weight at the conclusion of the 2-wk period of ad libitum feeding was not significantly different between SD and LE rats (P = 0.08; Table 2). Rats assigned to CR exhibited mild initial weight loss during the first week of CR (SD-CR: −21 ± 3 and LE-CR: −25 ± 3 g from baseline) but generally stable weight during the second week of CR (SD-CR: −25 ± 4 and LE-CR: −26 ± 4 g from baseline; Fig. 1A). This is consistent with an estimated energy balance near zero during the second week of the CR period for both strains (Fig. 1B). Restricted SD and LE rats exhibited nearly identical body weights at the end of the 2-wk period of CR (Table 2 and Fig. 1A).
CR produced small but significant reductions in mean arterial pressure (P < 0.05), systolic blood pressure (P < 0.01), and pulse pressure (P < 0.01), with no effect on diastolic pressure in both SD and LE rats (Table 2). As expected, CR also reduced HR (P < 0.01) and V̇o2 (P < 0.01) and increased HR variability (P < 0.05) in both strains (Table 2 and Fig. 2). The reduction in total V̇o2 was generally associated with lower resting V̇o2, although CR did not significantly decrease dark-phase resting V̇o2 in SD rats (Table 2). CR-induced reductions in HR and V̇o2 developed sooner and were more evident in the light phase (Fig. 2) than in the dark phase. This could be due, in part, to the shift in ingestive patterns of rats treated with CR implemented by providing one feeding at the onset of the dark phase. Clearly, these rats consumed their calories during the dark phase and thus exhibited reduced light-phase feeding and drinking behavior (Table 2). It also should be noted that during the light phase, particularly early in the response to CR, RQ was substantially reduced (Fig. 2F). By the end of the 2-wk period, light-phase RQ had returned to control levels in SD-CR rats but remained significantly lower in LE-CR rats. CR had no effect on home cage locomotor activity (Table 2 and Fig. 2, G and H).
Response to Refeeding
One day of ad libitum refeeding produced significantly greater caloric intake in both CR groups than in either AL group or baseline values (Figs. 1C and 3A). After the second day of refeeding, the LE-CR group remained hyperphagic, whereas in the SD-CR group, caloric intake was no different than in AL controls of both strains. Total caloric intake over baseline in the LE-CR group during the 4-day recovery period was significantly greater than in the SD-CR group (70.8 ± 7.2 vs. 42.4 ± 10.4 kcal, respectively; P < 0.005; Fig. 3A). The greater caloric intake in LE rats was associated with greater cumulative positive energy balance (LE: 181.5 ± 17.8 vs. SD: 143.7 ± 10.9 kcal) and weight gain (LE: 61.8 ± 3.6 vs. SD: 50.8 ± 4.7 g) during refeeding. Both strains exhibited a pronounced elevation of RQ over 1.0 that was clearly evident during the first light-phase period (12–24 h after ad libitum access to food; Fig. 2, E and F) of refeeding, indicative of net lipogenesis (9). There was no significant difference in RQ between strains during refeeding.
Upon observing this strain-dependent difference in response to refeeding, we extended the recovery period in one group of SD and LE rats to 8 days. The LE rats studied for 8 days remained hyperphagic, consuming 152 ± 25 calories over baseline, whereas the SD rats consumed a total of only 63 ± 32 calories during the 8-day recovery period (n = 4 for each group; P < 0.01). Hyperphagia in LE rats was accompanied by increased meal size in both dark (SD-CR: 9.7 ± 0.8 vs. LE-CR: 13.0 ± 1.3 min; P < 0.01) and light phases (SD-CR: 9.6 ± 1.6 vs. LE-CR: 14.4 ± 2.2 min; P < 0.01).
In addition to greater post-CR hyperphagia, the LE-CR group exhibited a more rapid recovery of V̇o2 during the refeeding period than did the SD-CR group (Figs. 2, C and D, and 3B). In the light phase, LE-CR change in (Δ)V̇o2 was no different than LE-AL V̇o2 by refeeding day 3 (0.31 ± 0.26 vs. 0.69 ± 0.26 ml/min, respectively; Fig. 3B), whereas SD-CR V̇o2 remained significantly lower than SD-AL V̇o2 at that time (−0.54 ± 0.24 v −0.13 ± 0.23 ml/min, respectively; Fig. 3B). SD-CR V̇o2 failed to recover by the end of the 4-day refeeding period (Fig. 3B). HR recovery was similar to recovery of V̇o2 in that LE-CR HR returned to control levels by refeeding day 3, whereas SD-CR HR remained low (Fig. 3C). By the end of the recovery period, SD-CR ΔHR was significantly greater than in all other groups (SD-CR: −36.8 ± 3.7 vs. SD-AL: −19.3 ± 3.5, LE-CR: −15.6 ± 3.5, and LE-AL: −11.7 ± 3.5 beats/min; P < 0.005). By day 8 of the recovery period in animals studied for an additional 4 days, strain differences in V̇o2 (dark phase: SD-CR: 9.71 ± 0.31 vs. LE-CR: 12.58 ± 0.65 ml/min; light phase: SD-CR: 7.24 ± 0.21 vs. LE-CR: 9.20 ± 0.40 ml/min; P < 0.005) and HR (dark phase: SD-CR: 374 ± 6 vs. LE-CR: 395 ± 16 beats/min; light phase: SD-CR: 326 ± 4 vs. LE-CR: 350 ± 13 beats/min; P < 0.01) were still evident.
The more rapid recovery in HR and V̇o2 during refeeding in LE rats was not due to greater locomotor activity. There was no effect of refeeding on locomotor activity in LE rats. Interestingly, SD rats exhibited a slight transient increase in both dark-phase and light-phase locomotor activity (Fig. 2, G and H). Surprisingly, the delayed recovery in SD rats was not explained by sustained suppression of resting V̇o2. This is illustrated by examination of the cumulative frequency plots of V̇o2 (Fig. 4). CR clearly produced a leftward shift in the plots of both SD and LE rats. One day of refeeding was associated with a return toward baseline patterns, which was completed by 4 days of refeeding in LE rats. However, the SD curve is still shifted to the left on day 4 of refeeding, although the shift is not evident for the lowest 20% of values.
Body Composition and Serum Leptin
As expected, CR was associated with significant reductions in weights of metabolically active organs (liver, kidney, and heart), visceral fat depots, and serum leptin levels (Table 3). We observed no strain differences in the effects of CR or refeeding on organ weights, fat depots, or serum leptin levels (Table 3).
The key new finding from these studies is the discovery of divergent strategies to restore body fat and body weight in SD and LE rats. In mature but growing SD male rats allowed ad libitum refeeding after 2 weeks of 40% CR, we observed a very mild and transient hyperphagia during refeeding, accompanied by a sustained suppression of both V̇o2 and HR for several days. This pattern of suppressed thermogenesis after refeeding in male SD rats has been reported previously, although in those studies SD rats were generally given controlled refeeding to prevent hyperphagia (5, 8). Our findings indicate that suppressed thermogenesis is observed in SD rats allowed unrestricted access to food during refeeding. In sharp contrast to SD rats, we observed that LE rats exhibit sustained hyperphagia accompanied by a rapid recovery in HR and V̇o2 upon refeeding. The more rapid increase in HR and V̇o2 during refeeding in LE rats was not due to greater locomotor activity. As discussed below, it is frequently suggested that the homeostatic controls for appetite and energy expenditure are concurrently regulated. The observation of distinct strain differences in the kinetics of recovery of appetite and energy expenditure suggests they may be independently regulated during recovery from CR.
Strain Differences and Similarities
There are several other reports in the literature comparing LE and SD rat strains. Compared with SD rats, LE rats exhibit 1) greater anorexia and body weight reduction in response to immobilization stress (13), 2) greater reductions in HR and body temperature following neurotoxicant exposure (16), 3) reduced drinking and greater sodium excretion in response to a number of dipsogenic agents (14), 4) diminished blood pressure and sympathetic responses to several weeks of cold exposure (36), and 5) resistance to DOCA hypertension (19). This list is neither comprehensive nor particularly revealing concerning fundamental underlying differences in physiology between these outbred rat strains. Not all comparisons of these strains reveal differences. Consistent with our findings in male rats, LE and SD female rats exhibit similar patterns of weight loss and levels of disruption of the estrous cycle during food deprivation (45). Furthermore, although LE rats exhibited diminished sympathetic and blood pressure responses to cold, they did exhibit cold-induced hyperphagia very similarly to SD rats (36). In addition to these between-strain comparisons, it should be noted that there also are differences in the body weight and body composition of SD rats, depending on their source (24).
Using continuous telemetry, calorimetry, and behavioral monitoring, we have observed some additional strain differences between weight-matched male SD and LE rats. During baseline recordings at 23°C, LE rats exhibited higher resting HR and reduced HR variability in both light and dark phases, suggesting the possibility of reduced tonic vagal suppression of HR. Furthermore, we observed that LE rats exhibit more ingestive behavior in the light phase. This was evident in terms of both greater time spent at the food hopper (photo beam break time) and greater drinking behavior at baseline. In addition, we observed that light-phase metabolic rate was slightly elevated in this strain.
Responses to CR
We imposed a level of CR (rats consumed 60% of ad libitum calories) that is routinely used in studies that examine the life span extension effect of reduced caloric intake (22, 46). During the 18–22 days of study, ad libitum controls exhibited very steady caloric intake but gained weight. Thus these growing male rats are in positive energy balance and appear to regulate caloric intake to maintain positive energy balance. When CR was imposed on these growing rats, the magnitude of total weight loss was minimal (Fig. 1A); instead, the primary effect was cessation of growth. CR animals have reduced visceral fat mass and reduced mass of metabolically active organs, including the heart, kidney, and liver.
CR produced expected reductions in HR, V̇o2, and RQ in both strains. In general, these responses were not different between strains, although one difference was that LE rats exhibited a slower normalization of light-phase RQ during the 2 wk of CR (see Fig. 2F). Interestingly, SD rats sensitive to diet-induced obesity exhibit greater CR-induced body fat loss and sympathoinhibition (assessed by 24-h urinary norepinephrine excretion) compared with SD rats resistant to obesity (26). However, we observed no strain difference in the pattern of CR-induced decrease in fat pad or organ weights, suggesting no major difference in the body composition response to CR between SD and LE rats. Nonetheless, both strains exhibited the well-established bradycardia during CR that is likely due to both decreases in sympathetic activity and increased vagal tone.
In this study, CR rats were provided with their daily aliquot of food just before the beginning of the dark phase. This strategy allows the animals to consume their food at the normal time and avoids the production of feeding-entrained circadian rhythm in activity and blood cortisol levels at another time in the light-dark cycle (10). Frequently, restricted feeding is used as a method to increase wheel-running activity in a paradigm known as activity-based anorexia (17, 42). It is interesting to note that in our studies, there was no evidence of increased home cage locomotor activity, in either the dark or the light phase, during 2 wk of CR in rats. Severe CR has been shown to reduce home cage activity in SD rats (11). The key point is that although CR can clearly influence locomotor activity in some conditions, the physiological responses we observed during CR in SD and LE rats apparently were not influenced by changes in locomotor activity.
Responses to Refeeding After CR
Reduced basal metabolic rate and selective lipogenesis are recognized features of the response to restoration of ad libitum feeding after CR imposed on young growing rats (3, 5, 6, 12, 23, 30). This response has been observed in SD rats studied at both standard laboratory temperatures and at thermoneutrality (8). The sustained reduction in metabolic rate after CR is not unique to SD rats, because it also has been observed in both lean (12) and obese (31) Wistar rats. In golden hamsters, reduced thermogenesis is also the primary mechanism to restore energy homeostasis after fasting (27). Evaluation of cumulative frequency plots indicates that the suppression of 24-h V̇o2 during refeeding in SD rats is not due to reduced resting V̇o2 but to a greater frequency of values in the midrange of V̇o2. Interestingly, this pattern of reduced total energy expenditure, with no difference in resting energy expenditure, has been observed in humans after 10% weight loss (37). Although the specific mechanisms underlying sustained suppression in V̇o2 in SD are not yet well understood, Dulloo and Jacquet (6, 7) have hypothesized that a component of the increased energy efficiency after refeeding is regulated by and is in proportion to the depletion of fat stores.
We observed that the suppression of metabolism in SD rats was accompanied by a sustained reduction in HR. We have observed a similar pattern of long-term suppression of HR and V̇o2 in SD rats after fasting (33). However, it is clear that this suppression of metabolism is not always observed during refeeding. Instead, LE rats exhibited sustained hyperphagia after refeeding, accompanied by a rapid normalization of HR and V̇o2. It is possible that greater thermic effects of feeding associated with hyperphagia contributed to more rapid normalization of the V̇o2 in LE rats. Sustained hyperphagia after CR was previously reported in LE rats (4), male Wistar rats (21), and young female SD rats (20). In addition, we also recently studied the effects of 2 wk of CR in FBNF1 male rats and observed rapid recovery of HR and V̇o2, with sustained hyperphagia (unpublished data). It is very intriguing to us that HR seems to closely follow V̇o2 during refeeding in our studies to date. The observation suggests the possibility that endocrine or neural mechanisms associated with suppressed metabolism during refeeding also have a direct influence on HR control.
Currently, the mechanisms associated with the strain differences in physiological and behavioral responses to refeeding after CR are unknown. Given the multiple afferent signaling mechanisms and central nervous system pathways engaged by negative energy homeostasis, there are many possibilities. It is interesting to note that differential regulation of appetite and energy expenditure also has been observed after central administration of recombinant adeno-associated virus containing leptin cDNA (39, 40). In these studies, the treatment produced reduced food intake for periods of time significantly shorter than for the increase in energy expenditure. The sustained hyperphagia in LE rats is somewhat reminiscent of the response observed after central administration of agouti-related protein (AgRP) (18). Although hypothalamic AgRP is increased during acute food deprivation, 2 wk of CR did not increase hypothalamic AgRP in young SD rats (1). CR generally stimulates melanin-concentrating hormone (MCH) in the lateral hypothalamus (34). MCH appears to have a much more potent effect on V̇o2 than on stimulation of feeding (25). Thus one testable hypothesis is that CR promotes long-lasting activation of AgRP neurons in LE rats but long-term activation of MCH neurons in SD rats. Clearly, there are numerous other possibilities, but it appears that additional studies designed to dissect the differential mechanisms of weight gain after CR in these rats strains could reveal pathways that are relatively specific for appetite and energy expenditure regulation.
This work was supported by National Heart, Lung, and Blood Institute Grant HL-56732.
We gratefully acknowledge the assistance of Chris Deluise and Erin Mariano. We thank Ross Henderson for assistance with instrumentation and computerized data collection.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society