Selective Loss of Leptin Receptors in the Ventromedial Hypothalamic Nucleus Results in Increased Adiposity and a Metabolic Syndrome

Abstract

Leptin, an adipocyte-derived hormone, has emerged as a critical regulator of energy homeostasis. The leptin receptor (Lepr) is expressed in discrete regions of the brain; among the sites of highest expression are several mediobasal hypothalamic nuclei known to play a role in energy homeostasis, including the arcuate nucleus, the ventromedial hypothalamic nucleus (VMH), and the dorsomedial hypothalamic nucleus. Although most studies have focused on leptin’s actions in the arcuate nucleus, the role of Lepr in these other sites has received less attention. To explore the role of leptin signaling in the VMH, we used bacterial artificial chromosome transgenesis to target Cre recombinase to VMH neurons expressing steroidogenic factor 1, thereby inactivating a conditional Lepr allele specifically in steroidogenic factor 1 neurons of the VMH. These knockout (KO) mice, designated Lepr KO^VMH, exhibited obesity, particularly when challenged with a high-fat diet. On a low-fat diet, Lepr KO^VMH mice exhibited significantly increased adipose mass even when their weights were comparable to wild-type littermates. Furthermore, these mice exhibited a metabolic syndrome including hepatic steatosis, dyslipidemia, and hyperleptinemia. Lepr KO^VMH mice were hyperinsulinemic from the age of weaning and eventually developed overt glucose intolerance. These data define nonredundant roles of the Lepr in VMH neurons in energy homeostasis and provide a model system for studying other actions of leptin in the VMH.

OBESITY IS A MAJOR risk factor for a number of chronic conditions, including cancer, diabetes mellitus, and cardiovascular disease (1), and poses an enormous health care burden in developed nations. Obesity reflects an imbalance between energy intake and energy expenditure and results when normal homeostatic mechanisms fail to recognize and/or respond to chronic excess of food intake. Normally, energy homeostasis requires the central integration of afferent signals communicating short- and long-term energy stores and the initiation of compensatory behavioral and physiological responses. A large body of work has revealed that the adipocyte-derived hormone leptin is a key anorexigenic signal that maintains normal energy homeostasis (2,3).

Leptin, the product of the lep gene in mice, circulates in the blood in concentrations proportional to the mass of body adipose stores (4,5). Although severe hyperphagia is perhaps the most prominent effect (6,7), mice deficient in leptin or its receptor (Lepr) exhibit a complex metabolic syndrome, including defective thermoregulation, augmented liver and adipose lipogenesis, and decreased glucose oxidation (8,9,10,11). Moreover, leptin-deficient mice whose food intake was adjusted to match that of wild-type (WT) littermates nonetheless accumulated more fat and body weight than their WT siblings (12). Thus, both severe hyperphagia and decreased energy consumption contribute to the marked obesity of leptin-deficient mice.

Genetic and pharmacological data suggest that leptin acts directly on neural networks to modulate energy homeostasis (13,14). First, central administration of leptin increases energy expenditure and normalizes food intake of leptin-deficient mice (13,15). In the mediobasal hypothalamus, leptin inhibits the expression of the orexigenic peptides neuropeptide Y (NPY) and agouti-related peptide (AgRP) while augmenting the expression of the anorexigenic peptides proopiomelanocortin (POMC, the precursor to α-MSH) and cocaine- and amphetamine-regulated transcript (CART) (16,17,18,19). The colocalization of Lepr with these neuropeptides implies that leptin inhibits feeding by stimulating the release of anorexigenic and inhibiting the release of orexigenic neurotransmitters from first-order neurons of central feeding circuits (20,21,22,23).

Functional Leprs have been identified within several hypothalamic nuclei known to be involved in energy homeostasis, including the arcuate (ARC), ventromedial (VMH), and dorsomedial (DMH) nuclei (24,25). Although the role of leptin in modulating the melanocortin system via its actions on ARC POMC neurons has been extensively studied, other sites of Lepr expression have received substantially less attention (22,26). It is apparent that the ARC cannot explain all aspects of leptin function. For example, specific deletion of Lepr in POMC neurons fails to recapitulate fully the severe obesity of Lepr-deficient mice (27). Thus, to understand the mechanisms by which leptin regulates energy homeostasis, it is necessary to investigate its functions in brain regions outside the ARC.

The VMH has long been implicated as an important regulator of energy homeostasis. Bilateral lesions of the VMH in rats produced a severe hyperphagic obesity syndrome (28,29), although the role of the VMH in this syndrome has subsequently been questioned (30). We previously showed that mice lacking the nuclear receptor steroidogenic factor 1 (SF-1) exhibited marked structural disruption of the VMH and delayed-onset obesity (31,32). Finally, functional Leprs are found in the VMH, particularly the dorsomedial region (33,34,35). To assess the role of VMH Leprs in energy homeostasis, we used the Cre/LoxP strategy to specifically ablate hypothalamic Leprs within SF-1 neurons of the VMH while maintaining expression in the ARC, DMH, and other brain regions. These studies demonstrate important roles of VMH Leprs in energy homeostasis and compartmentalization.

Materials and Methods

Animal care

All animal experiments were performed under protocols approved by the Institutional Animal Care Committee at University of Texas Southwestern. SF-1/Cre mice were produced as described (36). Mice homozygous for the conditional Lepr allele (Lepr^fl/fl mice) were obtained from Dr. Jeffrey Friedman, Rockefeller University (14). Mice were housed two to four per cage in a temperature-controlled room with a 12-h light, 12-h dark cycle. Fresh water and standard diet [low-fat (LF); Teklad 7001 diet, 2.94 kcal/g, 9.4% kcal from fat; Harlan Teklad, Madison, WI] were available ad libitum. For high-fat (HF) studies, mice were given a 60% kcal fat diet (Research Diets D12492, 5.24 kcal/g, 60% kcal from fat; New Brunswick, NJ). For food intake experiments, 12-wk-old female knockout (KO; designated Lepr KO^VMH) and WT mice were caged individually and sequentially fed LF and HF diets, with food being weighed daily. Littermates were used as controls for all studies. Unless specifically indicated, all studies described here used mice fed a LF diet.

Genotyping

The SF-1/Cre transgene was detected by PCR analysis of tail DNA using the following conditions: 94 C for 45 sec, 55 C for 45 sec, and 72 C for 1 min for a total of 40 cycles (primers 5′-GAGTGAACGAACCTGGTCGAAATCAGTGCG-3′ and 5′-GCATTACCGGCGATGCAACGAGTGATGAG-3′), producing a 408-bp product. Genotyping of the Lepr^fl allele was performed as described (14).

Body weight and composition

Male and female WT and Lepr KO^VMH mice were placed on LF or HF diet beginning at weaning. Body weight was measured every 14 d beginning at 4 wk of age. Naso-anal length was measured with digital calipers at 20 wk of age (Fisher Scientific, Pittsburgh, PA). Body fat was quantified using the Bruker Minispec mq10 nuclear magnetic resonance (NMR) Analyzer (The Woodlands, TX).

Tissue and plasma collection

Twenty-week-old mice were killed by decapitation without anesthesia within 30 sec of handling. Trunk blood was collected in 50-ml conical tubes that had been coated with trace amounts of 7.5% K₃-EDTA (E0270; Sigma-Aldrich, St. Louis, MO). Blood was centrifuged at 2000 × g for 20 min, and plasma was collected, frozen on dry ice, and stored at −80 C until used for analysis. Gonads, adrenals, and liver were simultaneously harvested, weighed, and rapidly frozen in liquid nitrogen.

Liver and plasma triglyceride measurements

Approximately 100 mg liver from 20-wk-old female Lepr KO^VMH and WT mice (LF diet) was homogenized in 4 ml chloroform/methanol (2:1 vol/vol) using a Polytron tissue homogenizer. The homogenates were washed with 1 ml 50 mm NaCl, vortexed, and centrifuged at 1500 × g for 10 min. The resulting organic phase was washed twice with 1 ml 0.36 m CaCl₂/methanol (1:1 vol/vol), vortexed, and again centrifuged at 1500 × g for 10 min. After centrifugation, the organic phase was placed in a fresh tube and the volume brought to 5 ml with chloroform. For determination of cholesterol content, 10 μl 50% Triton X-100 was added to 100 μl liver homogenate and dried under nitrogen. The cholesterol assay was performed using the Roche cholesterol enzymatic reagent (Roche Applied Science, Indianapolis, IN). For determination of triglyceride content, 10 μl 50% Triton X-100 was added to 50 μl liver homogenate and allowed to dry under nitrogen. The triglyceride assay was performed using the Sigma triglycerides enzymatic reagent (Sigma-Aldrich).

Histology

Hematoxylin and eosin (H&E) staining.

To obtain adult tissues, mice were anesthetized and fixed via transcardial perfusion with ice-cold 4% paraformaldehyde. After perfusion, tissues were removed and postfixed overnight in Bouin’s fixative, rinsed twice for 30 min with PBS (pH 7.4), and infiltrated with paraffin following standard protocols. Tissues were stored at 4 C and embedded in 100% paraffin (Paraplast Plus tissue-embedding medium; Fisher) until sectioning. Eight-micrometer sections were cut on a Microm HM 330 rotary microtome (Microm, Heidelberg, Germany) and floated on a 42 C water bath before mounting on Superfrost/Plus microscope slides (Fisher). Slides were then allowed to dry overnight at 37 C. H&E staining was performed following standard protocols. Stains, clarifier, and bluing reagent were purchased from Richard-Allen Scientific (Kalamazoo, MI). After staining procedure, slides were coverslipped using a nonaqueous mounting medium (Richard-Allen Scientific). Images were captured on a Nikon Eclipse e1000m microscope and Nikon DXM1200F digital camera with ACT-1 software (Nikon, Melville, NY).

Phosphorylated signal transducer and activator of transcription-3 (pSTAT3) immunohistochemistry.

Twelve-week-old female Lepr KO^VMH and WT mice (LF diet) were given ip injections containing 100 μg recombinant mouse leptin (obtained from the National Hormone and Peptide Program, National Institute of Diabetes and Digestive and Kidney Diseases, Torrance, CA) or PBS as a control. Forty-five minutes thereafter, mice were anesthetized and fixed via transcardial perfusion with PBS followed by 10% neutral buffered formalin. Brains were removed, postfixed in the same fixative overnight, and then cryoprotected in a PBS-buffered 20% sucrose solution for 24 h. Fifty-micrometer sections were cut using a vibrating microtome (Vibratome, St. Louis, MO) and stored in a PBS-buffered, 30% ethylene glycol, 30% glycerol solution at −20 C until further processing.

pSTAT3 immunohistochemistry was performed as described (37). The anti-pSTAT3 antibody (no. 9131, lot 6; Cell Signaling Technology, Danvers, MA) was diluted 1:3000. The biotin-conjugated donkey antirabbit IgG secondary antibody (no. 711-065-152, lot 70285; Jackson ImmunoResearch, West Grove, PA) was diluted 1:1000. An avidin biotinylated-peroxidase complex combined with 3,3′-diaminobenzidine peroxidase substrate was used for detection (Vector Laboratories, Burlingame, CA). Normal donkey serum was purchased from Equitech-Bio, Inc. (Kerrville, TX). All other chemicals were purchased from Sigma-Aldrich. Images were captured as detailed above.

Oil Red O staining.

Twenty-week-old Lepr KO^VMH and WT mice were perfused and livers harvested and processed as described above for mouse brains. After cryoprotection, livers were embedded in OCT compound and stored at −80 C. Ten-micrometer sections were cut using a Leica CM1900 cryostat, placed on Superfrost/Plus microscope slides (Fisher), and allowed to dry at room temperature. Sections were subsequently fixed in formalin and rinsed in three changes of distilled water. For staining, sections were rinsed in absolute propylene glycol for 5 min and then placed in Oil Red O solution (propylene glycol plus 0.5% Oil Red O) at 60 C for 10 min. Sections were then rinsed in 85% propylene glycol for 5 min, rinsed in distilled water, and counterstained in hematoxylin for 2 min. Finally, sections were rinsed in running water for 3 min and coverslipped using an aqueous mounting medium. Images were captured as detailed above.

Plasma hormone measurements

Plasma obtained from mice fed the LF diet was used for all assays. Leptin and insulin levels were determined using ELISA kits from Crystal Chem, Inc. (Downers Grove, IL). Corticosterone, estradiol, and testosterone levels were determined using RIA kits from MP Biomedicals (Solon, OH). All assays were performed with manufacturer’s suggested protocol.

Assessment of female estrous cycle

Female estrous cycles were monitored by microscopic analysis of vaginal smears taken every morning from 12-wk-old female Lepr KO^VMH and WT mice as described (38,39).

Energy expenditure

Indirect calorimetry.

Twelve-week-old female Lepr KO^VMH and WT mice were placed in chambers of an OXYMAX system (Columbus Instruments, Columbus, OH). Food and water were supplied ad libitum. Room temperature was maintained at 22 C with a 12-h light, 12-h dark cycle. Air flow into each chamber was maintained at 0.6 liters/min and exhaust air from each chamber was sampled for 3 min every 39 min. Mice were allowed to adjust to the chamber for 24 h before data collection. O₂ consumption and CO₂ production were measured for 72 h on LF diet and 72 h on HF diet. The respiratory exchange ratio (RER) is the ratio of volume of CO₂ produced (milliliters per kilogram per hour) per volume of O₂ consumed (milliliters per kilogram per hour).

Wheel running.

Mice were placed in individual cages containing a running wheel with a 24-cm diameter. Mice were allowed to adjust to the cages for 5 d before data collection was started. Wheel turns were collected every 30 min for 10 d. Animals were fed LF diet for 5 d before transitioning to HF diet. Each revolution was counted by magnetic switch closures and data were acquired using Vitalview Software from Mini Mitter Co. Inc. (Bend, OR).

Gene expression analyses

Tissues (liver, gonadal fat pads, intrascapular fat pads) from Lepr KO^VMH and WT mice fed LF diet were harvested and stored at −80 C until they were used to prepare total RNA using Trizol (Invitrogen, Carlsbad, CA) according to the manufacturer’s protocol. Transcript levels for acetyl-coenzyme A-carboxylase (ACC), fatty acid synthase (FAS), stearoyl-coenzyme A-desaturase (SCD), Δ-6-desaturase (D6D), sterol regulatory element binding protein (SREBP-1C), peroxisome proliferator-activated receptor α (PPARα), hormone-sensitive lipase (HSL), and uncoupling protein 1 (UCP1) were determined by quantitative PCR using described primers and protocols with an Applied Biosystems (Foster City, CA) Prism 7900HT sequence detection system (40). Primer sequences for all genes are available from the authors upon request.

Glucose tolerance tests

Lepr KO^VMH and WT mice were fasted overnight (14–16 h) but given water ad libitum. The day of the test, the mice were weighed and given a small nick at the end of the tail with a sterile razor blade for blood collection. Blood glucose concentrations were determined by the glucose oxidase method using a OneTouch Ultra glucometer (Lifescan, Inc., Milpitas, CA). A baseline glucose concentration was obtained, after which the mice were injected ip with glucose (1.5 g/kg body weight). Blood glucose levels were sampled from the tail nick at 10, 20, 30, 60, and 120 min after injection. Data are reported as millimoles per liter and were analyzed using a two-way ANOVA with genotype and time as parameters.

Data analysis

Values are reported as means ± sem. Statistical significance was determined by ANOVA, with individual comparisons between means made using Bonferroni’s post hoc analysis or two-tailed unpaired Student’s t test. Analysis was done using GraphPad PRISM software (GraphPad, San Diego, CA).

Results

Generation of Lepr KO^VMH mice

The conditional Lepr allele used in these studies has been used previously to generate liver- and brain-specific KO mice (14). Cre-mediated recombination deletes the first coding exon, including the signal sequence, and thus is predicted to prevent the expression of all splice variants. Transgenic mice expressing Cre recombinase within SF-1 neurons of the VMH were generated as described (Fig. 1A) (36). Cre expression in the hypothalamus of SF-1/Cre transgenic mice, as assessed by analysis of expression of the Cre-dependent ROSA26R reporter gene, paralleled that of endogenous SF-1. The VMH showed intense staining, with a gradient from dorsomedial (highest) to ventrolateral (lowest), whereas adjacent nuclei such as the ARC and DMH were largely unstained (Fig. 1B).

Figure 1.

Cre-mediated deletion of Lepr in the VMH. A, Diagram of the position and orientation of the Cre transgene within the 111-kb mouse SF-1 BAC clone. The BAC contains the SF-1 structural gene, 23 kb of 5′-flanking region, and 88 kb of 3′-flanking region. B, Coronal section through hypothalamus of an adult SF-1/Cre/ROSA26R mouse showing Cre-mediated activation of β-galactosidase expression within the VMH. C, pSTAT3 immunohistochemistry on brain sections from WT mice injected with PBS (top left) or leptin (bottom left) and Lepr KO^VMH mouse injected with leptin (bottom right). The top right shows a schematic of the mediobasal hypothalamus. 3, Third ventricle; LH, lateral hypothalamus. Scale bars, 500 μm (B) and 200 μm (C).

To obtain selective deletion of VMH Leprs, transgenic SF-1/Cre mice were crossed with Lepr^fl/fl mice. SF-1/Cre/Lepr^fl/wt offspring were subsequently crossed with Lepr^fl/fl mice to generate SF-1/Cre/Lepr^fl/fl mice (designated Lepr KO^VMH). Cre-negative, Lepr^fl/fl, and Lepr^fl/wt littermates (designated Lepr WT for these studies) were used as controls in all studies.

To confirm the functional inactivation of Leprs within SF-1 neurons, we examined the ability of ip injections of leptin to stimulate the phosphorylation of STAT3, a known mediator of leptin signaling downstream of the Lepr (37). As shown in Fig. 1C, pSTAT3 was present at comparable levels in the ARC and DMH of leptin-treated Lepr KO^VMH and WT mice. In contrast, markedly decreased pSTAT3 immunoreactivity was seen in the VMH of Lepr KO^VMH mice relative to WT littermates. These studies demonstrate selective ablation of leptin signaling in the VMH but not in closely adjacent hypothalamic nuclei.

Endocrine function of Lepr KO^VMH mice

Although the function of Leprs in the adrenal and gonads is unclear (41,42,43,44), we assessed adrenal and gonadal function of Lepr KO^VMH mice to detect any primary or secondary defects in these endocrine organs. Gonadal weights of both male and female Lepr KO^VMH were comparable to values for WT mice of the same sex; histological inspection of ovaries and testes revealed no obvious morphological abnormalities, and levels of sex steroids in male and female Lepr KO^VMH mice did not differ significantly from levels in sex-matched WT controls (data not shown).

The estrous cycle of female Lepr KO^VMH mice was evaluated by monitoring changes in vaginal cytology. Lepr KO^VMH mice cycled regularly, and there was no significant difference in cycle length. In addition, both male and female Lepr KO^VMH mice had apparently normal fertility, although this was not rigorously quantitated. Thus, although subtle defects may exist, there apparently are no significant perturbations in the reproductive capacity of Lepr KO^VMH mice.

We also assessed adrenal function in Lepr KO^VMH mice. First, Lepr KO^VMH mice appeared healthy and were fully viable, suggesting that they do not have adrenal insufficiency. In addition, morning (nadir) corticosterone levels in Lepr KO^VMH and WT female mice did not differ significantly (Table 1). Finally, the adrenal weights of Lepr KO^VMH mice did not differ significantly from those of WT mice, and histological examination showed normal adrenal zonation (data not shown).

Table 1.

Endocrine function of Lepr KO^VMH mice fed a LF diet

	Lepr WT	Lepr KOVMH
Leptin (ng/ml)	2.4 ± 0.4 (9)	6.3 ± 1.1 (10)a
Insulin (ng/ml)	0.8 ± 0.1 (8)	1.9 ± 0.3 (8)a
Glucose (mmol/liter)	13.4 ± 0.5 (13)	13.3 ± 0.4 (14)
Corticosterone (ng/ml)	7.3 ± 1.2 (12)	14.0 ± 3.5 (14)

For leptin, insulin, glucose, and corticosterone (nadir) measurements, 20-wk-old female Lepr KO^VMH and Lepr WT mice raised on a LF diet were killed by decapitation at 0900 h under free feeding conditions. Results represent the mean ± sem (n). 

^aP < 0.01 vs. Lepr WT in an unpaired, two-tailed Student’s t test. 

Body weight and composition of Lepr KO^VMH mice

To investigate the physiological consequences of specific ablation of VMH Leprs in energy homeostasis, we compared the weights of Lepr KO^VMH mice with those of WT littermates. Mice were weaned at 4 wk of age and placed on either normal (LF) mouse chow (9.4% kcal from fat) or a HF chow (60% kcal from fat) and weighed every 14 d. On LF chow, neither male nor female Lepr KO^VMH mice showed a significant weight increase over their WT littermates. However, both male and female Lepr KO^VMH mice showed marked susceptibility to diet-induced obesity (DIO) (Fig. 2, A and B). By 20 wk of age, female Lepr KO^VMH mice on the HF diet were 20% heavier than WT littermates (40.5 ± 1.8 vs. 33.5 ± 1.5 g, Fig. 2C). Males were less severely affected, with the Lepr KO^VMH males at 20 wk weighing 12% more than WT littermates (52.2 ± 1.9 vs. 46.7 ± 1.3 g). VMH Leprs apparently do not play a role in somatic growth, as there was no difference in the length of male or female mice on either diet (data not shown).

Figure 2.

A and B, Changes in body weight and composition in Lepr KO^VMH mice. The weights of male (A) and female (B) Lepr KO^VMH mice were compared with those of WT littermates on both LF and HF diet; C, 20-wk-old female WT (left) and Lepr KO^VMH (right) littermates fed HF diet; D, body composition of 20-wk-old Lepr KO^VMH and WT mice on LF diet was measured by NMR spectroscopy and normalized to total body mass. Results represent the means ± sem of at least 10 mice. *, P < 0.05; **, P < 0.01; ***, P < 0.001 vs. WT mice on same diet.

To the extent that they are part of a negative feedback loop, one possible consequence of VMH-specific ablation of the Lepr is resistance to certain actions of leptin, possibly leading to a compensatory increase in circulating leptin levels. Accordingly, we measured plasma leptin, insulin, and glucose levels in WT and Lepr KO^VMH mice fed LF diet. As shown in Table 1, 20-wk-old female Lepr KO^VMH mice exhibited a 2- to 3-fold increase in plasma leptin levels relative to WT littermates. In addition, Lepr KO^VMH mice on LF diet were slightly hyperinsulinemic (∼2-fold) while remaining euglycemic. These mice were not obese but did have significantly increased adiposity. Thus, the relative effects of the ablation of Leprs in the VMH vs. secondary effects associated with obesity per se remain to be defined.

NMR analysis demonstrated that the increased weight in Lepr KO^VMH mice entirely reflects increased adiposity. On HF diet, the percent body fat was increased by 15% in male and 21% in female Lepr KO^VMH mice compared with WT littermates (data not shown). Interestingly, although Lepr KO^VMH mice showed no significant increase in total body weight on the LF diet, both males and females on the LF diet exhibited even greater relative increases in percent body fat than on the HF diet, 55 and 30%, respectively (Fig. 2D). Histological evidence supported the NMR body composition data. Examination of white adipose tissue (WAT) from Lepr KO^VMH mice fed a LF diet revealed a marked cellular hypertrophy (Fig. 3, A and B), whereas examination of brown adipose tissue (BAT) showed a significant increase in the size of cellular vacuoles, presumable secondary to excess lipid deposition (Fig. 3, C and D). Thus, despite comparable body weights, Lepr KO^VMH mice on the LF diet demonstrated increased lipid accumulation in adipocytes relative to WT littermates.

Figure 3.

Lepr KO^VMH mice exhibit dyslipidemia on LF diet. A and B, Comparison of H&E-stained WAT from the gonadal fat pad of 20-wk-old female WT (A) and Lepr KO^VMH (B) on LF diet shows marked cellular hypertrophy; C and D, comparison of BAT from the same mice (C, WT; D, Lepr KO^VMH) shows large, clear vacuoles in BAT of Lepr KO^VMH mice; E and F, Oil Red O staining of liver sections from the same mice showing increased staining in Lepr KO^VMH mice relative to WT sections; G and H, hepatic (G) and plasma (H) triglyceride levels from 20-wk-old female mice fed a LF diet. Results represent the means ± sem of at least nine mice. *, P < 0.05 vs. WT mice. Scale bars, 100 μm (A–D) and 50 μm (E and F).

To determine whether the increased adiposity in Lepr KO^VMH mice fed a LF diet was limited to adipocytes or was more generalized, we assessed hepatic lipid content of Lepr KO^VMH mice by both histological and biochemical methods (Fig. 3). By 20 wk of age, the livers of Lepr KO^VMH mice fed a LF diet weighed approximately 50% more than WT controls and contained approximately 50% more triacylglycerol (data not shown, Fig. 3G). This increased lipid content was also readily apparent on histological sections stained with Oil Red O (Fig. 3, E and F). In contrast, there was no difference in liver cholesterol levels (data not shown). Paralleling the increased liver triacylglycerol content, Lepr KO^VMH mice fed a LF diet also exhibited hypertriglyceridemia without concomitant hypercholesterolemia (Fig. 3H, data not shown).

A recent report showed that central leptin administration was associated with significant decreases in hepatic and WAT expression of a number of enzymes involved in de novo fatty acid synthesis, including acetyl-coenzyme A-carboxylase (ACC), FAS, SCD, and D6D along with changes in SREBP in the liver; and PPARα in adipocytes (45). In particular, SCD was previously implicated as a key hepatic enzyme whose expression was down-regulated by leptin as a part of its complex metabolic actions (46). We therefore examined the levels of relevant transcripts in liver and WAT as well as UCP1 in BAT as described in Materials and Methods. Although there was a strong trend toward increased expression (Table 2), hepatic expression of most of these genes did not differ significantly between WT and Lepr KO^VMH on the LF diet (e.g. D6P, FAS, SREBP-1C, and PPARα). In contrast, we did note significantly increased expression of SCD1 in liver and WAT (1.7-fold increase relative to WT mice in liver, 1.5-fold in WAT). These data suggest that, despite their elevated circulating leptin levels, mice with VMH-specific KO of Lepr do show changes in hepatic expression of at least one component of the de novo pathway for fatty acid biosynthesis; these changes are in the opposite direction from those induced by central leptin administration (45). In addition, expression of UCP1 in BAT was significantly decreased in Lepr KO^VMH mice (Table 2).

Table 2.

Hepatocyte and adipocyte gene expression of enzymes in the de novo pathway of fatty acid synthesis and key transcription factors that regulate their expression

	Fold difference
Liver
FAS	1.6 ± 0.4
D6D	1.4 ± 0.3
SCD1	1.7 ± 0.2a
PPARα	1.2 ± 0.2
SREBP1c	1.4 ± 0.2
HSL	1.6 ± 0.5
WAT
SCD1	1.5 ± 0.2a
BAT
UCP1	0.2 ± 0.1a

Gene expression levels were assessed using RNA isolated from the livers, WAT, and BAT of 20-wk-old female Lepr KO^VMH and Lepr WT mice raised on a LF diet. Results represent the mean fold difference of five sibling pairs. Results are given as mean ± sem. HSL, Hormone-sensitive lipase. 

^aP < 0.05 using a two-tailed Student’s t test. 

Food intake and energy expenditure in Lepr KO mice

Twelve-week-old female Lepr KO^VMH mice and WT control mice were housed individually, and their caloric intake on LF chow was measured daily for 7 d, after which they were switched to a HF diet, and daily caloric intake was measured for the next week. Two-way ANOVA analysis showed no significant difference in caloric intake on the LF diet; however, a small, but significant, difference was seen on the HF diet, with Lepr KO^VMH mice eating 1.8 ± 0.06 kcal/d more than WT mice (Fig. 4A). We performed a second experiment looking at long-term caloric intake of Lepr KO^VMH mice on a HF diet. When cumulative caloric intake is measured over an extended period of time, the caloric intake of Lepr KO^VMH mice is significantly increased after several weeks on the HF diet (Fig. 4B).

Figure 4.

Food intake and energy expenditure in Lepr KO^VMH mice. A, Average daily caloric intake of 12-wk-old female Lepr KO^VMH and WT mice, measured for 7 d before and after the transition from LF to HF diet (n = 8 of each genotype); B, cumulative caloric intake of 12-wk-old female Lepr KO^VMH and WT mice over 4 wk after transition to HF diet (n = 8 for each genotype); C, locomotor activity of 12-wk-old Lepr KO^VMH and WT females was measured over five consecutive days on both LF and HF diet and reported as the total number of wheel revolutions (n = 8 for each genotype); D, VO₂ consumption was measured in 12-wk-old Lepr KO^VMH and WT females on LF and HF diets (n = 6 for each genotype). Results are given as the means ± sem. *, P < 0.05; **, P < 0.01 vs. WT mice.

The VMH has been implicated as an important regulator of energy utilization, including spontaneous locomotor activity (47). We therefore used a running wheel assay to evaluate spontaneous locomotor activity. At 12 wk of age, female Lepr KO^VMH mice appeared to have a normal circadian pattern of activity but showed a modest decrease in wheel-running activity on both a LF and HF diet; this decrease did not achieve statistical significance (Fig. 4C). Like WT mice, Lepr KO^VMH mice increased their locomotor activity in response to a HF challenge.

To evaluate the metabolic rate of Lepr KO^VMH mice, 12-wk-old female Lepr KO^VMH mice and WT littermates were analyzed by indirect calorimetry. Mice on LF diet were monitored for 3 d, after which they were switched to HF diet and monitored for at least 3 d. On LF diet, there was no discernable difference in the O₂ consumption of Lepr KO^VMH and WT mice. However, on the HF diet, Lepr KO^VMH mice exhibited a defective adaptive thermogenic response. Although WT mice significantly increased their O₂ consumption by 15%, Lepr KO^VMH mice increased their O₂ consumption by only 8% (Fig. 4D), which did not achieve statistical significance. In response to a HF challenge, WT mice demonstrated a shift from carbohydrate-based to fatty acid-based metabolism, which was measured by a decrease in RER. This substrate response was preserved in Lepr KO^VMH mice, because there was no significant difference in the RER of Lepr KO^VMH and WT mice on either LF or HF diet (data not shown).

Glucose homeostasis in Lepr KO^VMH mice

Ob/ob and db/db mice exhibit significant disturbances in glucose homeostasis (48,49), which prompted us to examine this in our mice. Although no differences in fasting glucose were apparent, glucose tolerance tests (GTTs) performed on 12-wk-old mice fed a LF diet showed significant glucose intolerance in both male and female Lepr KO^VMH mice (Fig. 5, A and B). To determine whether the glucose intolerance of 12-wk-old Lepr KO^VMH mice was a primary effect due to the loss of VMH Leprs or was secondary to obesity, GTTs were performed on 4-wk-old female Lepr KO^VMH mice, before any significant increase in body adiposity is seen. At this age, Lepr KO^VMH females had improved glucose tolerance compared with their WT littermates (Fig. 5C). To examine the basis of this improved glucose tolerance, plasma insulin levels were measured 30 min after a glucose bolus was given to a separate cohort of 4-wk-old Lepr KO^VMH females. Consistent with the results of the GTT, Lepr KO^VMH females had significantly elevated levels of insulin 30 min after a glucose dose (Fig. 5D).

Figure 5.

Glucose homeostasis in Lepr KO^VMH mice. A and B, Twelve-week-old male (A) and female (B) Lepr KO^VMH and WT mice fed a LF diet were fasted for 12–16 h and injected ip with glucose (1.5 mg/g body weight), and whole-blood glucose levels were sampled over 2 h. C, 4-wk-old female Lepr KO^VMH and WT mice fed a LF diet were fasted for 12–16 h and injected ip with glucose (1.5 mg/g body weight), and whole blood was collected over 2 h and glucose levels determined; D, 4-wk-old female Lepr KO^VMH and WT mice were injected ip with glucose (1.5 mg/g body weight), and 30 minutes thereafter, blood was drawn and plasma insulin measured by ELISA; E and F, blood was drawn from the tail vein of 4-wk-old female Lepr KO^VMH and WT mice fed LF diet at 0900 h in a fed state or after a 16-h fast and total glucose and plasma insulin levels were determined. The results represent the means ± sem for at least seven mice in each group. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

To further evaluate glucose homeostasis in Lepr KO^VMH mice, basal glucose and insulin levels were measured in 4-wk-old Lepr KO^VMH and WT mice. In the fed state, 4-wk-old Lepr KO^VMH females were euglycemic despite having significantly (∼2-fold) elevated insulin levels (Fig. 5, E and F). In the fasting state (∼12–16 h), Lepr KO^VMH mice trended toward hypoglycemia and hyperinsulinemia, although neither reached statistical significance. Thus, Leprs in the VMH apparently play a role in glucose homeostasis via regulation of insulin secretion.

Discussion

As discussed above, leptin and its cognate receptor Lepr have emerged as key regulators of appetite and energy homeostasis, and understanding the neuroanatomical basis of leptin signaling has been a major focus of recent research. In particular, the leptin-responsive neural circuitry of the ARC has been extensively studied. Although these circuits have been linked to the control of appetite, manipulation of these circuits cannot explain fully the abnormal energy homeostasis seen in leptin-deficient mice (26,50,51). Using the Cre-LoxP system, we show here that loss of leptin signaling from SF-1-positive neurons of the VMH results in a metabolic syndrome of hyperinsulinemia, increased adiposity, and augmented sensitivity to DIO without severe hyperphagia.

In rodents, an acute caloric challenge results in a subsequent decrease in appetite and an increased metabolic rate (52,53,54). Leptin-responsive neurons of the VMH are responsible, at least in part, for the initiation of these anorexigenic responses to high caloric intake. Thus, mice lacking Leprs in SF-1-positive VMH neurons are susceptible to DIO. This susceptibility results from subtle defects in both food intake and energy expenditure. Although no difference in food intake is seen on a LF diet, Lepr KO^VMH mice eat more than WT mice when switched to a HF diet. Although the difference is small and insignificant on a daily basis, the increased caloric consumption integrated over time becomes statistically significant after several weeks. As well, Lepr KO^VMH mice have a defective adaptive thermogenic response. Although WT mice increase their metabolic rate in response to a HF diet, Lepr KO^VMH mice fail to do so. The failure of Lepr KO^VMH mice to modify food intake and metabolic expenditure in response to a HF challenge results in a positive energy balance, increased adiposity, and obesity.

Using a similar genetic approach, Elmquist, Lowell, and colleagues (55) reported that leptin acutely depolarized and increased the firing rate of neurons within the VMH and that mice lacking Leprs within SF-1 neurons of the VMH were susceptible to DIO. Of considerable importance, the conditional Lepr allele and the SF-1 Cre transgene used in these studies both were derived completely independently; in fact, the conditional Lepr allele used by this group flanks exons 17 and 17′, again producing an apparent complete loss of function following Cre-mediated recombination. Given the use of independently derived transgenic reagents, the close parallels between their results and ours serve as a striking example of the utility of the Cre-loxP strategy to provide valid results regarding tissue-specific effects of gene disruption.

Although we largely confirm the previous finding that Leprs within the VMH mediate resistance to DIO, some differences between the two studies should be noted. In the previous study, increased body weight was seen in Lepr KO^VMH mice fed a LF diet, whereas we observed no increased body weight on the LF diet. Several explanations for this discrepancy may exist. First, the genetic backgrounds may account for some of the observed differences. Although the previous report used mice with a mixed genetic background, mice used here were backcrossed onto a C57BL/6 background for at least five generations and thus may be considered congenic. C57BL/6J mice are themselves highly vulnerable to DIO (56,57), which may explain why the weights of WT mice used in this study exceeded the weights of the KO mice used in the previous study, especially on a HF diet.

Subtle differences in diets may also contribute to the differences in weight gain between the two studies. The previous study used a LF diet containing 12.5% kcal from fat, whereas we used a LF chow containing 9.5% kcal from fat. Given the sensitivity to DIO that these mice display, a 30% relative difference in dietary fat content may account for the discrepant results. Thus, on the LF diet used in this study, no difference in food intake, energy expenditure, or body weight was detectable between Lepr KO^VMH mice and WT mice.

The absence of overt obesity in Lepr KO^VMH mice fed a LF diet revealed other underlying metabolic disturbances present in Lepr KO^VMH mice. For example, Lepr KO^VMH mice also had significant expansion of both white and brown adipose compartments, even when their weight did not differ significantly from WT mice. Thus, Lepr KO^VMH mice appear to have a defect in energy partitioning, preferentially directing calories into fat rather than lean body mass.

Consistent with a recent report that central leptin administration down-regulates hepatic and adipocyte expression of key enzymes in the de novo pathway for fatty acid synthesis (45), the Lepr KO^VMH mice exhibited significantly increased hepatic and WAT expression of SCD. Moreover, despite their elevated leptin levels, there was a strong trend toward increased expression of several other components of the fatty acid synthetic pathway in Lepr KO^VMH mice relative to WT controls. Although additional studies are needed, these data suggest that Leprs in the VMH are an important component of central leptin coordination of this key metabolic pathway.

In db/db mice, hyperinsulinemia has been documented from a very early age, before the onset of hyperphagia and obesity (48). Similarly, Lepr KO^VMH mice exhibit hyperinsulinemia as early as weaning, before any detectable difference in body composition. This hyperinsulinemia appears to be an exaggerated postprandial response, because Lepr KO^VMH mice are able to suppress insulin under fasting conditions. Although the mechanisms whereby leptin dampens the postprandial insulin response remain to be determined, the VMH has been shown to make neuronal connections with brain centers known to regulate pancreatic insulin secretion (58). Thus, it may be that leptin signals serve to modulate these neuronal circuits.

Although hyperinsulinemia in 4-wk-old Lepr KO^VMH mice results in improved glucose tolerance, it may contribute to the increased adipose mass, insulin resistance, and glucose intolerance seen in older Lepr KO^VMH mice. Several lines of evidence support this assertion. First, insulin significantly up-regulates the rates of adipocyte and hepatic lipogenesis through both the long-term expression of lipogenic genes and activation of lipogenic enzymes (59). As well, hyperinsulinemic states, especially postprandial hyperinsulinemia, early in life are predictive of future weight gain (60,61). Thus, hyperinsulinemia may stimulate increased adiposity, which then leads to a cycle of insulin resistance and increased insulin secretion in an attempt to maintain euglycemia.

The neuronal circuitry and neuropeptides through which leptin exerts its metabolic effects in the VMH remain to be determined. Because leptin is known to regulate melanocortin signals in the ARC, it is plausible that leptin might function in an analogous manner in the VMH. Thus, it is interesting to note the similarities between the Lepr KO^VMH mice and melanocortin receptor (MCR)-deficient mice, particularly the MC3R KO model. Although MC4R KO mice exhibit an obesity phenotype that includes hyperphagia, increased adiposity and linear growth, and sensitivity to diet-induced obesity (62,63), MC3R KO mice predominantly exhibit a metabolic syndrome and become obese only in response to increased dietary fat (63,64). Interestingly, MC3R KO mice fed a LF diet show dramatically increased fat mass in the face of limited weight gain, identical to the phenotype reported here for the Lepr KO^VMH mice (63,65). Also similar to Lepr KO^VMH mice, MC3R KO mice are normophagic on a LF diet but become mildly hyperphagic when transitioned to a HF diet (66). In addition, both Lepr KO^VMH mice and MC3R KO mice are hyperleptinemic and slightly hyperinsulinemic with normal corticosterone levels (64).

The role of MC3R in the regulation of locomotor movement remains unsettled; one group reported decreased activity only in females (64), whereas a second group reported decreased activity only in males (65). Some decreased locomotor activity was seen in our Lepr KO^VMH mice, but the differences failed to reach statistical significance. Furthermore, the studies reported here were done using 12-wk-old females that already showed significant increases in body fat content. Thus, differences in locomotor activity, if real, may be secondary to the decreased lean body mass to body weight ratio and not a result of differential autonomic output in Lepr KO^VMH mice.

The phenotypes of MC3R KO and Lepr KO^VMH mice apparently differ in some respects. Butler et al. (65,66) reported a significant increase in the RER of MC3R mice transitioned to HF diet, indicating a defect in transitioning to fatty acid oxidation. Using the RER values of MC3R KO and WT mice, Butler showed MC3R mice have a positive fat balance (ratio of fat intake to fat oxidation) when fed a HF diet. In contrast, Lepr KO^VMH mice challenged with a HF diet failed to increase O₂ consumption but nonetheless lowered their RER normally. The RER phenotype in MC3R KO mice is subtle, however, because it was observed by only one of the two groups. Also, the RER effect seems to be both age and background dependent (64,66).

In summary, we show here that VMH neurons contain functionally relevant Leprs that are necessary for appropriate energy homeostasis, confirming previous studies implicating these receptors in the homeostatic response to increased fat intake. Our studies used a completely different SF-1/Cre transgene and conditional Lepr conditional allele than those used by Dhillon et al. (55), providing a striking illustration of the robustness of the Cre-loxP strategy for generating reliable tissue-specific gene disruption. We further demonstrate that leptin signaling in VMH neurons regulates the response to a glucose challenge as well as caloric partitioning between fat and lean body mass. The mechanisms by which leptin action in VMH neurons modulates these responses remain unexplored, and future work is needed to understand more fully the contribution that VMH leptin signaling plays in the regulation of adiposity.