INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

IN THE LAST THREE YEARS OB protein (OB, also known as leptin) has become a leading candidate for the long-sought lipostatic signal in the control of feeding and energy balance (5, 19). OB is expressed in adipose tissue of many species, including humans and rodents, is secreted into the blood, and binds to receptors in several sites (e.g., 7, 8, 11, 25). Plasma OB concentration is positively related to the degree of adiposity and is lowered by fasting or weight loss (24). Three models of genetic obesity in rodents that are accompanied by massive hyperphagia have been linked to defects in the OB system: OB is absent in ob/ob mice, which have a mutation in the OB gene, and is overexpressed in db/db mice and fa/fa Zucker rats, which lack functional OB receptors (6, 44). OB reduces food intake and increases energy expenditure after peripheral or central administration in genetically obese and normal rodents (4, 16, 18, 29, 33). Sensitivity to these actions of OB varies considerably among rodent strains (ob/ob mouse > normal > fa/fa Zucker rat > db/db mouse). OB inhibits feeding much more potently when it is administered centrally than when given peripherally (35, 40). Therefore, the effect of OB on food intake probably results from activation of receptors in the central nervous system.

Whether OB inhibits feeding by decreasing meal size, meal frequency, or both is not yet clear. Because these parameters of spontaneous feeding are differentially affected by various test procedures (e.g., 15, 17, 20, 41), the answer to this question will have important consequences for future work on the mechanisms of OB's effects on feeding. There are only two reports of the effects of acute OB administration on meal patterns, and they produced conflicting results. In the accompanying paper, Flynn et al. (12) reported that intracerebroventricular injection of OB decreased meal size and not meal frequency, whereas Blevins et al. (3) reported that intravenous or paraventricular hypothalamic injection of OB decreased meal frequency and not meal size. Each of these was an acute test of a single bolus administration of OB. To better model the physiological situation of chronically increased release of OB from an enlarged adipose tissue, we tested the effects of eight consecutive daily subcutaneous injections of a novel modified OB (Mod-OB) with a prolonged biological half-life (>48 h) on food intake, meal patterns, and body weight. We describe the results in male rats here and those in female rats in an accompanying report (9).

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals and housing conditions. Twenty-four male Sprague-Dawley rats (ZUR:SD, Institut für Labortierkunde, University of Zurich) with a mean body weight of 292 g at study onset were used. The rats were individually housed in Plexiglas cages (43 × 27 × 18 cm) with grated stainless steel floors. The cages were in a quiet, temperature-controlled (22 ± 2°C) colony room that was kept on an artificial 12:12-h dark-light cycle with the lights off at 1330. Ground normal rat chow (no. 890 25 G3, NAFAG, Gossau, Switzerland), with 3.0% fat and a metabolizable energy of 12.2 kJ/g, and tap water were available ad libitum. A 16-cm-long, gray, hard plastic pipe (diameter 7.5 cm) was placed in every cage in an effort to achieve more naturalistic meal patterns (27). Before the experiment, the rats were adapted to the diet and the maintenance conditions for 2 wk.

Test procedure. Every day between 0930 and 1030, the feeding cups were weighed and refilled and the rats were weighed and, beginning 10 days before the experiment, they received intrascapular subcutaneous injections of 0.5 ml/kg 0.9% NaCl. Two groups of rats (n = 12 each) were then matched for body weight and food intake during 3 baseline days (experimental days 1-3). Test injections were administered on each of the next 8 days (experimental days 4-11) using a between-subjects design. Mod-OB (Hoffmann-La Roche, Nutley, NJ) was produced by the covalent linkage of authentic recombinant human OB protein and polyethylene glycol polymers. The result of polyethylene glycolation of several other proteins has been a prolonged half-life (13, 26, 28). In pharmacokinetic experiments, Mod-OB had a half-life in the serum of rats of >48 h. Mod-OB binds to and activates the long form of the OB protein receptor (OB-Rb or OB-R_L) in transfected cell lines and exhibits the same spectrum of biological activities observed with recombinant human OB protein in rats. Mod-OB was formulated into a stock solution in a 10 mM sodium acetate buffer, pH 4.5, containing 80 mM sodium chloride. The vehicle was identical except for the absence of Mod-OB. The stock solution was diluted daily with sterile 0.9% NaCl to yield a concentration of 8 mg Mod-OB for injection. The 12 Mod-OB rats received 0.5 ml/kg of this solution subcutaneously for a daily dose of 4 mg Mod-OB/kg. We selected that dose on the basis of preliminary results indicating that 2 mg Mod-OB/kg was about the threshold dose for an inhibition of feeding under similar conditions. The dose of 4 mg/kg Mod-OB is similar to the doses used in the original studies demonstrating feeding effects in ob/ob mice (16, 29). The 12 control rats received 0.5 ml/kg of similarly diluted Mod-OB vehicle. Daily procedures were all completed by 1030, i.e., 3 h before dark onset, after which the rats were not further disturbed.

Data acquisition and analysis. The rats fed from spill-resistant feeding cups fixed on scales (Mettler, PM 3000). The weight of the cups (±0.06 g) was recorded each 30 s by an Olivetti personal computer (M240), using custom-made software (VZM, Software-Entwicklung Albert Krügel, Munich, Germany). Meals were defined as food removals 0.2 g, with the time between any two removals 15 min. VZM used this meal definition to record the time of meal onset, the size (g) and duration (min) of the meal, and the intermeal interval (min). Meals defined and recorded in this way accounted for 96% of total daily food intake. Data are reported as means ± SE. Data for the vehicle and Mod-OB groups were statistically compared using Student's t-test with a minimum level of significance set at P = 0.05. Analyses were done on daily food intake and body weight and on meal parameters cumulated for each day, each nocturnal and diurnal period, and each 3-h quartile of the nocturnal period.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Single daily subcutaneous injections of 4 mg/kg Mod-OB for 8 consecutive days reduced daily food intake compared with vehicle injections (Fig. 1A). Significant reductions in daily food intake occurred during 9 consecutive days, beginning on injection day 3 (experimental day 6) through postinjection day 3 (experimental day 14). During this period, Mod-OB-injected rats ate between 4.5 and 7.1 g/day (13-20%) less than the control group. From the beginning of Mod-OB injection to experimental day 14, the cumulative reduction in food intake relative to the control mean was 51 ± 6 g/rat. Body weight was significantly less in the Mod-OB group for 5 days, from injection day 8 through postinjection day 4 (experimental days 11-15; Fig. 1B). The maximum body weight difference between the two groups was 24 g on postinjection day 3 (experimental day 14). Body weight of Mod-OB-injected rats slowly approached control body weight thereafter. No compensatory hyperphagia was evident in Mod-OB-injected rats after the injection period (Fig. 1A).

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Effects of single daily subcutaneous injections of 4 mg/kg recombinant modified OB protein (Mod-OB) for 8 consecutive days (Injections) on daily food intake (A) and body weight (B) in male rats. Data are means ± SE. *§# Significant differences between Mod-OB and vehicle group: * P  0.05; § P < 0.01; # P < 0.001.

Decreased nocturnal food intake accounted for most of the hypophagic effect of Mod-OB. Nocturnal food intake, like total food intake, was reduced on experimental days 6-14 (Fig. 2). The reduction of nocturnal food intake by Mod-OB was entirely due to a 21-34% decrease of spontaneous meal size (Fig. 3). Nocturnal meal size was reduced by Mod-OB for 8 consecutive days, from injection day 3 to postinjection day 2 (experimental days 6-13). Nocturnal meal duration tended to be decreased by Mod-OB, but this reached statistical significance only on experimental day 7 (Fig. 4). Mean eating rate during nocturnal meals (meal size/meal duration) was not significantly changed on any day. Nocturnal meal frequency tended to be increased by Mod-OB, but this never reached statistical significance (Fig. 5). The satiety ratio (meal size/meal duration) of nocturnal meals was not significantly changed on any day. Analysis of meal parameters by quartiles of the nocturnal period failed to reveal any ultradian changes in either group.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Effect of single daily subcutaneous injections of 4 mg/kg Mod-OB for 8 consecutive days on nocturnal food intake. Data are means ± SE. *§# Significant differences between Mod-OB and vehicle group: * P < 0.05; § P < 0.01; # P < 0.001.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Effect of single daily subcutaneous injections of 4 mg/kg Mod-OB for 8 consecutive days on nocturnal meal size. Data are means ± SE. *§ Significant differences between Mod-OB and vehicle group: * P  0.05; § P < 0.01.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Effect of single daily injections of 4 mg/kg Mod-OB for 8 consecutive days on nocturnal meal duration. Data are means ± SE. * Significant difference between Mod-OB and vehicle group, P < 0.05.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Effect of single daily injections of 4 mg/kg Mod-OB for 8 consecutive days on nocturnal meal frequency. Data are means ± SE. No significant differences.

There was no sign of tolerance to the repeated administration of Mod-OB in any of the dependent variables.

Rats ate less during the diurnal period. Although Mod-OB tended to decrease diurnal food intake, no significant effects were detected. Mod-OB reduced diurnal meal size on 2 days: 2.5 ± 0.3 vs. 4.0 ± 0.5 g in the control group [t(22) = 2.57, P < 0.05] on experimental day 9 and 2.8 ± 0.3 vs. 4.4 ± 0.5 g [t(22) = 3.10, P < 0.01] on experimental day 13. There were no significant diurnal effects on any of the other meal parameters analyzed.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The results demonstrate that chronic subcutaneous administration of Mod-OB, a modified human recombinant OB protein with a biological half-life >48 h, leads to sustained decreases in food intake and body weight in adult male Sprague-Dawley rats. This is consistent with previous observations of the effects of repeated intracerebroventricular and intraperitoneal administration of OB in genetically obese mice and their lean littermates (4, 16, 29), indicating that these effects of OB can also be observed in lean rats and with subcutaneous administration. These data all support the hypothesis that OB plays a role in the long-term physiological control of food intake and energy balance. The important new finding in the present study is that a selective reduction in the size of meals, in particular of nocturnal meals, accounts for the inhibitory effect of chronically administered Mod-OB on spontaneous food intake. The strategy of repeated subcutaneous injections of the long-acting OB preparation that we employed provides a good model of the sustained release of endogenous OB from enlarged adipose tissue. Therefore, our findings suggest that endogenous OB inhibits feeding by decreasing meal size.

The present finding that the primary effect of chronically administered Mod-OB is a decrease in meal size complements what is known about pathologies of the endogenous OB signaling system. That is, the hyperphagia of both ob/ob obese mice and fa/fa obese rats are due to dramatic increases in meal size, with meal frequency unaffected or even reduced (1, 17, 41). This functional similarity suggests that exogenous OB affects the same biological mechanisms that are normally affected by physiological variations of endogenous OB in intact animals. It also suggests that genetic lesions in the OB signaling system elicit a permanent increase in meal size in addition to, or despite, other physiological disturbances resulting from the mutation or compensatory responses to them. In the accompanying paper, Flynn et al. (12) report that a decrease in meal size also occurs in response to acute intracerebroventricular administration of normal recombinant OB in rats. Together with our data, this suggests that there is no fundamental difference between the modes of action of acute central and chronic peripheral OB treatment. In another study, however, acute paraventricular hypothalamic or intravenous administration of OB reduced meal frequency rather than meal size (3). The reasons for this discrepancy are unclear. However, both of these studies involved a single bolus administration of OB and only short-term (<24 h) measurements. Thus these acute findings and the results of the present chronic study may not be directly comparable.

Our observation of a delayed onset of the reduction in meal size together with the persistence of the inhibitory effects of Mod-OB for 3 days postinjection appear to be consistent with the extended half-life of Mod-OB and the subcutaneous route of injection. During the treatment with Mod-OB, the decrease in daily food intake was only ~13-20%. This, however, is consistent with previous reports of the modest potency of OB's effects on food intake and body weight in normal mice, which contrasts with the much more potent effects of OB in genetically obese ob/ob mice, which lack biologically active OB. Several possible mechanisms for this variable sensitivity of the OB system have been proposed (5).

Chronic subcutaneous Mod-OB treatment resulted in a selective reduction of spontaneous meal size, with no decrease in meal frequency. This indicates that OB decreases food intake by affecting the mechanisms controlling meal size. It also demonstrates that chronic peripheral OB administration does not interfere with the animals' ability to initiate meals. The reduction of meal size by Mod-OB was accompanied by a nonsignificant increase in meal frequency, which may have been a secondary compensatory response to the reduction in meal size.

We do not believe that the decreases in meal size observed here resulted from an aversive effect of Mod-OB. Unmodified OB protein does not support a conditioned taste aversion (42). Furthermore, no aversive or nonspecific effects were observed in a larger number of rats that received prolonged Mod-OB treatment. Here, the mean eating rate during meals tended to decrease in response to Mod-OB. An effect of hypophagic agents on eating rate has at times been suspected to reflect a nonspecific side effect rather than physiological satiety (43). It should be noted in this context, however, that meal size is usually a more sensible and less variable parameter than meal duration (22). Furthermore, the microstructure of eating during meals was not recorded in the present study. Thus whether Mod-OB affected local eating rate during meals, intrameal pauses between bursts of eating, or other microstructural parameters remains to be determined. Further evidence for the specificity of Mod-OB's action on feeding in female rats is presented in the accompanying paper (9).

Because basal insulin level is another strong candidate for a centrally acting lipostatic feedback signal in the control of food intake and energy balance (34), it is interesting to note that the effects of chronic Mod-OB treatment on meal patterns that we report here are slightly different from those produced by chronic insulin administration. Continuous intracerebroventricular infusion of 2 mU insulin/day reduced daily food intake and nocturnal meal size by amounts similar to those in the present study, but, in contrast to the effects of chronic Mod-OB, insulin also reduced nocturnal meal frequency (30). Thus it is unclear whether the feeding effects of chronic insulin and OB are mediated by the same mechanisms. The compensatory hypophagia in response to an experimentally induced increase in body weight was also mainly due to a reduction of nocturnal meal size (15). These findings support the hypothesis that body fat influences feeding mainly through changes in meal size. In contrast, several metabolic manipulations affected food intake by changing meal frequency rather than meal size (20).

Mod-OB appeared to increase energy expenditure as well as to decrease food intake. The maximum body weight difference between the Mod-OB group and the control group was 24 g. When this occurred (day 14), the mean cumulative reduction in metabolizable energy intake was ~620 kJ (51 g × 12.2 kJ/g). If the difference in body weight represented solely a decrease in the accretion of adipose tissue, then its energy equivalent would be ~700 kJ (24 g × 29 kJ/g) (14). Because the metabolic efficiency of chow for fat deposition is ~71% (31), a hypophagia of ~1,000 kJ would be required for such a decrease in adipose tissue accretion. As this is considerably larger than the observed hypophagia [1,000 > 620 ± 74 kJ, 1-sample t-test: t(11) = 5.11, P < 0.001], Mod-OB probably stimulated energy expenditure in addition to reducing food intake. This conclusion is in line with results of pair-feeding studies that indicate that a stimulation of energy expenditure contributes considerably to the effect of OB on body weight (16, 23). This increase was probably metabolism rather than physical activity, because Mod-OB did not stimulate running wheel activity in female rats under otherwise similar conditions (9).

The food intake suppression by Mod-OB was particularly prominent during the dark phase, whereas exogenous OB's stimulatory effect on energy expenditure in juvenile rats was evident mainly during the early light period (39). The different circadian patterns of OB's hypophagic and hypermetabolic effects suggest that these effects are not mediated by identical mechanisms. Perhaps OB reduces feeding more efficiently when energy balance is positive and food intake is usually high (during the dark) and increases energy expenditure particularly well when basal metabolism is comparatively low (during the light). The nocturnal increase in OB gene expression in rats, which is probably related to the higher nocturnal food intake (32), may thus help to limit the nocturnal energy deposition. On the other hand, plasma OB concentration in humans is also elevated during the night (36), whereas humans eat more during the day.

Although experimentally induced body weight decreases of ~10% have been shown to stimulate compensatory hyperphagia (10), Mod-OB-treated rats did not increase their food intake above the control level after an ~7% body weight loss during the injection period. It requires further investigation to determine whether this ~7% body weight loss was insufficient to trigger a compensatory response or whether there are other reasons for the absence of compensatory hyperphagia after chronic Mod-OB administration.

Perspectives