The idea that genetic abnormalities contribute significantly to obesity received a renewed thrust with the critical identification of the ob gene and its protein product leptin (Zhang et al. 1994). Leptin is synthesized and secreted by adipose tissue and functions as an afferent satiety signal that likely acts in the brain, particularly within the hypothalamus, to regulate food intake and metabolic rate and ultimately control body fat mass (Caro et al. 1996).

In C57Bl/6J ob/ob mice the ob gene is mutated, resulting in a lack of functional leptin and the development of obesity (Zhang et al. 1994). Repeated daily intraperitoneal injections of 0.6 µmol (0.1 mg) to 60 µmol (10 mg) leptin/kg body weight into these mice decreased their food intake and body weight (Campfield et al. 1995, Halaas et al. 1995, Pelleymounter et al. 1995, Stephens et al. 1995, Weigle et al. 1996). Administration of leptin systemically (6-60 µmol leptin/kg body weight daily) for 3-16 wk (Halaas et al. 1995, Pelleymounter et al. 1995) also effectively lowered food intake and body weights of normal lean mice. To determine whether leptin exerts these effects by acting within the central nervous system, leptin was administered via intracerebroventricular (ICV)⁴ injection. ICV injection of leptin (60 pmol/mouse and 210 pmol/rat) decreased cumulative food intake 7 and 24 h later (Campfield et al. 1995, Schwartz et al. 1996a and 1996b), suggesting that the central nervous system is a target for this action of leptin. However, the effects of ICV-administered leptin have not been extensively studied. To better understand how leptin functions centrally it would be useful to further characterize the time course of ICV-administered leptin on food intake suppression.

Repeated systemic injections of leptin for 3 wk in ob/ob mice also increased metabolic rate (Pelleymounter et al. 1995) on the basis of the volume of oxygen consumed per kilogram body weight. Changes in body composition that occurred during this chronic leptin treatment may have contributed to the altered metabolic rate and confounded interpretation of the data. Collins et al. 1996 reported that norepinephrine turnover was increased in brown adipose tissue of ob/ob mice 2-6 h after intraperitoneal injection of 2.4 nmol leptin. These data imply that leptin may acutely increase the metabolic rate of ob/ob mice.

Recent studies have suggested (Schwartz et al. 1996a, Stephens et al. 1995) that leptin exerts its action by inhibiting the release/synthesis of hypothalamic neuropeptide Y (NPY), a central regulator of energy homeostasis (Leibowitz 1991, White 1993). These data were largely derived after chronic systemic administration of leptin. It has also been suggested that exogenous leptin acutely inhibits the release and/or actions of NPY in ob/ob mice (Smith et al. 1996). It thus seems reasonable to hypothesize that the decreased food intake caused by leptin may be mediated by altered NPY activity, although other factors are also involved because leptin effectively decreases food intake in mice with a knockout of the NPY gene (Erickson et al. 1996).

The reduction in food intake and body weight gain in ob/ob mice after chronic treatment with leptin is analogous to that observed after adrenalectomy. Adrenalectomized (ADX) ob/ob mice that lack endogenous glucocorticoids exhibit an intrinsic reduction in food consumption (Feldkircher et al. 1996, Kim and Romsos 1987, Vander Tuig et al. 1984). The pathways by which chronic therapy with leptin versus ADX of ob/ob mice regulate energy balance are unclear but may converge at some point because both strategies lead to a lowered food intake and body weight. It would be of interest, therefore, to determine if leptin would exert its anorectic actions in ADX ob/ob mice.

The present studies were thus designed to examine effects of a single ICV injection of leptin on the time course of changes in food intake and metabolic rates of both lean and ob/ob mice, to compare the ability of leptin to inhibit NPY-induced food intake in lean versus ob/ob mice, and to determine the effectiveness of ICV-administered leptin in ADX lean and ADX ob/ob mice.

MATERIALS AND METHODS

Experiment 2 was devised to test the efficacy of leptin to modulate NPY-induced feeding. Lean and ob/ob mice were food deprived for 4 h and then injected ICV with either vehicle, 60 pmol leptin, 230 pmol NPY, or a combination of leptin (60 pmol) and NPY (230 pmol). Cumulative food intake was measured every 30 min for 3 h and then 24 h later. These doses of leptin and NPY were chosen on the basis of Experiment 1 above and the ability of 230 pmol NPY to substantially enhance food intake in ob/ob mice (Walker and Romsos 1993). A second trial was conducted in which mice that had been food deprived for 4 h were injected ICV with vehicle or leptin (60 pmol) and then refed for 3 h before ICV injection with vehicle or NPY (230 pmol). This protocol provided a 3-h window for leptin to exert its action before injection of NPY. It also maximized the food intake treatment differences between NPY and vehicle-treated mice, because these mice were relatively satiated at the time NPY was administered.

Experiment 3 examined the effectiveness of leptin in altering food intake and energy expenditure in lean and ob/ob ADX mice. As in Experiment 1, mice were first food deprived for 4 h and then injected ICV with leptin (60 pmol) or vehicle. Cumulative food intake was measured every 30 min for 3 h and then after 24 h. Oxygen consumption was measured at 3 and 24 h.

RESULTS

As expected, ob/ob mice (body weight = 35.6 ± 0.9 g) had lower thermogenic activity (oxygen consumption = 107 ± 4 mL/h) than lean mice (body weight = 22.5 ± 0.5 g; oxygen consumption = 137 ± 6 mL/h). Injection of 3 pmol of leptin did not influence oxygen consumption in either lean or ob/ob mice at any of the time points measured (Fig. 2). Whole-body oxygen consumption of ob/ob mice, but not lean mice, increased by 36% 3 h after administration of 60 pmol leptin. Because leptin-treated mice ate substantially less than appropriate controls, lean and ob/ob mice were pair-fed to mice injected with 60 pmol leptin. Pair-fed lean mice consumed 30% less oxygen than leptin-treated lean mice within 3 h, indicating that leptin lowers metabolic efficiency (Fig. 2). Pair-fed ob/ob mice did not alter their already low metabolic rate (Fig. 2).

To avoid the confounding effects of food intake on metabolic rate of lean mice, mice were first deprived of food for 4 h as before and subsequently injected with leptin but not refed. Metabolic rates were unaltered 1 h after leptin but had increased 18% by 3 h (Fig. 3).

We then sought to determine the time course of ICV leptin actions on food intake and metabolic rates within the 3-h period immediately after injection. Leptin (60 pmol)-treated lean and ob/ob mice consumed ~35% less food than their respective controls within 30 min (Fig. 4). There were no differences in oxygen consumption 1 or 2 h after ICV injection of leptin in either lean or ob/ob mice (results not presented).

To determine the effectiveness of leptin in mice with an even stronger drive to eat, lean mice were food deprived for 48 h before injection of leptin and refeeding. Food intake was 62% lower 3 h postinjection and 46% lower 24 h postinjection (ICV) of 60 pmol of leptin than in control mice (Fig. 5).

To further optimize conditions to detect interactions of leptin and NPY on food intake, mice were treated with leptin and fed 3 h before injection of NPY. NPY was thus administered to partially satiated mice. Food intake increased in vehicle-pretreated lean and ob/ob mice injected ICV with 230 pmol of NPY. Pretreatment with leptin, before injection of NPY, attenuated the increase in feeding produced by ICV injection of NPY in ob/ob mice, but not in lean mice (Fig. 7).

DISCUSSION

We directly compared food intake responses of genetically obese ob/ob and lean mice to centrally administered leptin. Our initial approach to ensure that mice would eat within a short time period was to deprive them of food for 4 h before injection of leptin. With the use of this experimental design, we have confirmed previous observations (Campfield et al. 1995, Schwartz et al. 1996 b) that centrally administered leptin is effective in lowering food intake in ob/ob mice and showed that this occurs within 30 min after leptin administration. Although both lean and ob/ob mice treated with 3 pmol leptin tended to consume less food, a significant response was observed only in ob/ob mice, suggesting that ob/ob mice are more sensitive to exogenous leptin than lean mice. This suggestion is consistent with the possibility that the leptin signal transduction pathway is upregulated in ob/ob mice lacking endogenous leptin. However, it should be noted that ob/ob mice treated with 3 pmol leptin ICV still consumed as much food as vehicle-treated lean mice. The higher dose (60 pmol) of leptin lowered food intake in lean and ob/ob mice to the same approximate absolute intake (Fig. 1), or to a slightly lower absolute intake in lean mice than in ob/ob mice (Figs. 4 and 6). Differences in food intake of vehicle-injected lean and ob/ob mice confound comparisons of the relative sensitivity of these mice to leptin. Under the conditions of this study, the absolute food intake of leptin-treated ob/ob mice was not lower than that of lean mice. Leptin may play a primary role in preventing hyperphagia and a secondary role in regulation of normal intake.

NPY is a powerful appetite stimulant (Leibowitz 1991, White 1993) that can activate feeding even in satiated mice. NPY cell bodies in the arcuate nucleus project to the hypothalamic paraventricular nucleus where NPY release vigorously stimulates feeding. The voracious feeding behavior of ob/ob mice has been attributed to elevated release of NPY from paraventricular neurons and increased mRNA content in the arcuate nucleus (Wilding et al. 1993). It has been proposed that leptin suppresses feeding by restraining NPY synthesis and release (Schwartz et al. 1996a and 1996b, Stephens et al. 1995). Although possible, it is unlikely that leptin depresses NPY synthesis sufficiently within 30 min of injection to lower food intake within this time frame. If leptin acutely lowers food intake via NPY-mediated mechanisms, leptin would more likely alter release of NPY or inhibit actions of NPY after release. We thus determined the actions of injected NPY in the presence of leptin. Initially, leptin and NPY were co-administered to mice that had been food deprived for 4 h. Food intake was augmented in lean mice, but not in ob/ob mice. A complicating factor in this study was that control ob/ob mice ate a substantial quantity of food because they had been food deprived. To maximize food intake differences between controls and treatment groups, mice were thus refed and treated with leptin 3 h before NPY administration. Leptin attenuated NPY-induced food intake in ob/ob mice, in agreement with the observation made by Smith et al. 1996. In contrast, a similar dose of leptin was ineffective in preventing NPY-induced feeding in lean mice. The reasons for these phenotype differences are not readily apparent. Injection of leptin caused similar low intakes of food in lean and ob/ob mice, and injection of NPY caused similar high intakes of food in these mice. There were phenotype differences, however, in intake among vehicle-injected mice, as noted earlier. There may be crosstalk between the leptin and NPY signaling systems within the central nervous system, with greater effects observed in ob/ob mice than in lean mice. Administration of a higher dose of leptin, or a lower dose of NPY, may have unmasked expression of these interactions in lean mice. Leptin likely also suppresses food intake by mechanisms independent of NPY because leptin functions in mice lacking NPY (Erickson et al. 1996). Clearly, the complexity of the leptin/food intake regulation system is only beginning to be understood.

Leptin increased metabolic rate in lean and ob/ob mice within 3 h after administration, demonstrating that effects of leptin on metabolic rate do not require prerequisite changes in body composition. Hwa et al. (1996) also recently reported that metabolic rates of ob/ob mice were increased during a 22-h period after ICV injection of leptin. The mechanism responsible for increasing metabolic rate was not explored. However, we suggest that leptin activates diet-induced thermogenesis via central nervous stimulation of brown adipose tissue. Consistent with this suggestion, ob/ob mice, which are deficient in leptin, are well known to have a deficit in diet-induced thermogenesis (Knehans and Romsos 1983). Metabolic rates of vehicle-treated ob/ob mice were low and not further lowered by acute food restriction (Fig. 2). Injection of leptin ICV increased metabolic rates in these mice (Fig. 2), likely via enhanced release of norepinephrine from sympathetic nerve terminals innervating brown adipose tissue (Collins et al. 1996). On the basis of the hypothesis that leptin activates or sustains diet-induced thermogenesis, fed lean mice with a fully functional diet-induced thermogenic system would not be expected to increase metabolic rate in response to leptin. However, when these lean mice are deprived of food or food restricted, diet-induced thermogenesis by sympathetic nervous system stimulation of brown adipose tissue would be lowered (Knehans and Romsos 1983). It was under these conditions that leptin increased metabolic rates in lean mice (Figs. 2 and 3). NPY increases food intake and reduces diet-induced thermogenesis (Billington et al. 1991, Egawa et al. 1991, Walker and Romsos 1993). Thus leptin might regulate metabolic rate by interfering with actions of NPY. Alternatively, leptin might, for example, stimulate actions of corticotropin-releasing hormone, a neuropeptide that inhibits food intake and stimulates metabolic rates (Arase et al. 1989, Holt and York 1989, Rothwell 1990).

The hyperphagia and deficit in diet-induced thermogenesis characteristic of ob/ob mice are corrected by removal of glucocorticoids via adrenalectomy (Feldkircher et al. 1996, Kim and Romsos 1987, Vander Tuig et al. 1984), a response parallel to that observed after administration of leptin to intact ob/ob mice. We thus determined if leptin would function in ADX mice. Leptin effectively lowered food intake in ADX lean and ADX ob/ob mice. Our data thus indicate that leptin functions within the central nervous system to regulate food intake even in the absence of glucocorticoids. Conversely, glucocorticoids function to stimulate appetite in the absence of leptin (i.e., in intact ob/ob mice). Leptin and glucocorticoids may thus act as counterregulatory hormones in the control of food intake, possibly by influences on NPY or corticotropin-releasing hormone (Zakrzewska et al. 1997). In addition, the possibility that glucocorticoids modulate feeding by regulating synthesis/release of leptin at the adipocyte cannot be eliminated. Indeed we (Mistry et al. 1996) and others (Vos et al. 1995) have demonstrated that ob gene expression is influenced by glucocorticoids.

ADX corrects the deficit in diet-induced thermogenesis in ob/ob mice by activating sympathetic nervous stimulation of brown adipose tissue thermogenesis (Vander Tuig et al. 1984). On the basis of our observations in nonadrenalectomized mice in which leptin increased metabolic rate only in situations in which diet-induced thermogenesis was low, we would predict that leptin would not stimulate metabolic rate in fed ADX ob/ob mice. This prediction was supported by the results obtained (Fig. 9).

High-affinity receptors for leptin are present in the hypothalamus (Ghilardi et al. 1996, Lee et al. 1996). One receptor isoform associates with members of the janus kinase (JAK) family to activate cytoplasmic transcription factors called signal transducers and activators of transcription (STAT) (Ghilardi et al. 1996). This would provide a potential mechanism whereby leptin could regulate gene transcription and exert long-term effects on food intake and metabolic rate. Leptin also has been reported to rapidly (i.e., within minutes) modulate intracellular calcium concentrations within isolated arcuate neurons from the hypothalamus (Glaum et al. 1996). This would provide a potential mechanism to explain the relatively rapid initial effects of leptin on food intake observed in this study. Coordination of these multiple leptin signaling pathways within the central nervous system provides a powerful mechanism to regulate body fat mass.

ACKNOWLEDGMENTS