Leptin action via LepR-b Tyr₁₀₇₇ contributes to the control of energy balance and female reproduction

Abstract

Leptin action in the brain signals the repletion of adipose energy stores, suppressing feeding and permitting energy expenditure on a variety of processes, including reproduction. Leptin binding to its receptor (LepR-b) promotes the tyrosine phosphorylation of three sites on LepR-b, each of which mediates distinct downstream signals. While the signals mediated by LepR-b Tyr₁₁₃₈ and Tyr₉₈₅ control important aspects of energy homeostasis and LepR-b signal attenuation, respectively, the role of the remaining LepR-b phosphorylation site (Tyr₁₀₇₇) in leptin action has not been studied. To examine the function of Tyr₁₀₇₇, we generated a “knock-in” mouse model expressing LepR-b ^F1077, which is mutant for LepR-b Tyr₁₀₇₇. Mice expressing LepR-b ^F1077 demonstrate modestly increased body weight and adiposity. Furthermore, females display impairments in estrous cycling. Our results suggest that signaling by LepR-b Tyr₁₀₇₇ plays a modest role in the control of metabolism by leptin, and is an important link between body adiposity and the reproductive axis.

1. Introduction

Adipose tissue produces the hormone, leptin, in proportion to fat stores to communicate the status of long-term energy reserves to the brain and other organ systems [1,2]. In addition to suppressing food intake, adequate leptin levels permit the expenditure of energy on numerous processes, including growth and reproduction and a variety of parameters that affect metabolic rate. Conversely, lack of leptin action due to null mutations of leptin (e.g., Lep^ob/ob mice) or the leptin receptor (LepR) (e.g., Lepr^db/db mice) results in increased food intake in combination with reduced energy expenditure (and thus obesity) and neuroendocrine dysfunction (including infertility) [3,4]. Many of the effects of leptin are attributable to effects in the CNS, particularly in the hypothalamus [5,6].

Alternative splicing generates several LepR isoforms that possess identical extracellular, transmembrane, and membrane-proximal intracellular domains [7]. The “long” (LepR-b) isoform contains a 300 amino acid intracellular tail and is required to mediate physiologic leptin action [3,6]. LepR-b, like other Type I cytokine receptors, mediates intracellular signaling via an associated Jak family tyrosine kinase [8,9]. Leptin promotes the autophosphorylation and activation of LepR-b-associated Jak2, which phosphorylates three LepR-b tyrosine residues (Tyr₉₈₅, Tyr₁₀₇₇ and Tyr₁₁₃₈) [8–10]. While Jak2 tyrosine kinase activity is required for LepR-b signaling, Jak2 itself is insufficient to mediate most leptin action, as mice expressing a LepR-b mutant (LepR-b^Δ65) that binds and activates Jak2, but which lacks tyrosine phoshorylation sites and other intracellular LepR-b motifs, largely phenocopy LepR-b-deficient Lepr^db/db mice [11]. Thus, LepR-b tyrosine phosphorylation sites are crucial for physiologic leptin action.

Each LepR-b tyrosine phosphorylation site recruits specific SH2 domain-containing effector proteins: Tyr₉₈₅ recruits SHP2 and SOCS3, and attenuates LepR-b signaling, but does not appear to mediate other aspects of metabolic and neuroendocrine control by leptin in vivo [12–15]. Tyr₁₀₇₇ plays a dominant role in activating the latent transcription factor, signal transducer and activator of transcription- (STAT)-5, and Tyr₁₁₃₈ controls STAT3 [10,16]. Mice expressing a LepR-b mutant for Tyr₁₁₃₈ exhibit hyperphagic obesity with decreased energy expenditure that is only slightly less dramatic than that observed in Lepr^db/db mice; dysregulation of the thyroid and adrenal axes in Tyr₁₁₃₈ mutant mice is similar to that observed in Lepr^db/db animals [17–20]. In contrast to the absolute infertility of Lepr^db/db mice, however, reproductive function is relatively preserved in Tyr₁₁₃₈ mutant mice. These results predict that signaling via LepR-b Tyr₁₀₇₇ may modestly impact energy balance (e.g., to account for the slight differences seen in body weight regulation between LepR-b Tyr₁₁₃₈ mutants and Lepr^db/db mice), but that LepR-b Tyr₁₀₇₇ might serve an as important mechanism by which leptin modulates reproductive function.

2. Materials and methods

2.1. Reagents

Leptin was the generous gift of Amylin Pharmaceuticals (San Diego, CA). High fat diet (45% Kcal fat) chow was from Research Diets, NJ. Rabbit anti-pSTAT3 (Y709) was purchased from Cell Signaling (Boston, MA), and donkey serum was from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA). Donkey anti-rabbit Alexa-488 conjugated antibodies were purchased from Molecular Probes, Inc. (Eugene, OR). All other immunohistochemical supplies were purchased from Sigma-Aldrich (St. Louis, MO).

2.2. Animals

Animals were bred in our colony in the Unit for Laboratory Animal Medicine (ULAM) at the University of Michigan and handled in accordance with The University Committee on Use and Care of Animals (UCUCA).

2.3. 2.3 Generation of Lepr^f1077 mice

We modified our previous Lepr targeting vector in the pPNTLox backbone [11,14,17,21] to include the mutation encoding the Tyr₁₀₇₇→Phe₁₀₇₇ substitution. The construct was electroporated into R1 embryonic stem (ES) cells, which were selected as described [11,21]. Clones were analyzed by qPCR and confirmed by Southern blotting [22]; the presence of the substitution mutation was confirmed by a site-specific Taqman PCR allelic discrimination assay. Correctly targeted ES cells clones carrying the desired substitution were used to produce chimeric animals, which were crossed with C57Bl/6J mice to generate germline Lepr^F1077/+ (Lepr^tm5Mgmj) mice. Subsequent genotyping was by qPCR allelic discrimination assay. The resultant Lepr^F1077/+ animals were bred to C57Bl/6J mice for 2 generations before interbreeding of heterozygous (Lepr^F1077/+) animals to produce homozygous (Lepr^F1077/F1077; f/f) and wild-type (Lepr^+/+) mice for study. Animals were backcrossed to C57Bl/6J for≥6 generations prior to interbreeding to produce animals for the analysis of glucose homeostasis. All genotypes studied were the offspring of the same breeders.

2.4. Longitudinal study of f/f mice

Mice were single housed with ad libitum access to food and water. Mice were weighed weekly from 4 to 12 weeks of age. Food consumption was measured weekly by weighing the amount of food left in the cage and subtracting it from the amount present the previous week. Blood was collected in heparinized capillary tubes for glucose measurements using a glucometer and for serum determination of insulin and leptin (ELISA, Crystal Chem, Inc.). Body composition analysis (i.e., fat mass, lean mass and free fluid content) and CLAMS analysis (i.e., oxygen consumption (VO₂), carbon dioxide production (VCO₂), and spontaneous motor activity) were performed on conscious mice at age 14–15 wks by an NMR-based analyzer with a Minispec LF90II (Bruker Optics) and Comprehensive Laboratory Monitoring System (CLAMS, Columbus Instruments), respectively.

For analysis of reproductive function, female mice were checked daily from the time of weaning (at 28 day) for vaginal opening and thereafter for vaginal estrogenization by cellular histology. Mice were checked daily until a second estrous cycle was achieved, or until to 61 day of age.

2.5. Tissue collection

15-wk-old male and female mice received an overdose of pentobarbital (150 mg/kg, IP) and snout to anus length was measured with a micrometer. Gross dissections were performed and tissues including hypothalamus, brown adipose tissue (BAT), liver, white adipose tissue (WAT, gonadal fat pad), and gastrocnemius muscle were collected. Tissue samples were weighed and either snap-frozen on dry ice or fixed in 10% formalin for further analysis of protein lysates and histology, respectively.

2.6. Intravenous glucose tolerance (IVGTT) and microdissection

IVGTTs were performed on a separate cohort of male and female chow or high-fat diet fed 12–14 wk old mice for assessment of glucose stimulated insulin secretion. Mice were anesthetized via i.p. sodium pentobarbital (50–60 mg/kg) and arterial and venous catheters were inserted into the aortic arch and right atrium, respectively, by the University of Michigan Animal Phenotyping Core. The free ends of the catheters were exteriorized at the back of the neck and were fixed subcutaneously following closure of the incision. Following surgery, the catheters were flushed daily with heparin (200 U/ml). After 1 wk of recovery, mice were fasted for 5–6  h and a fasting blood sample was obtained for assessment of basal glucose and insulin. A bolus injection of 50% glucose was then given at 1.5 g/kg-bw via the venous catheter and blood samples were collected T=2, 5, 10, 15, 30, and 60 after the injection for glucose and insulin analysis. Glucose was measured utilizing an Accu-Chek glucometer (Roche, Germany) and plasma insulin was measured using the Linco rat/mouse insulin ELISA kit. Mice were allowed to recover from the IVGTT for 48 h after which time they were anesthetized and brains were microdissected on a rodent coronal brain matrix and hypothalamus and preoptic areas frozen on dry ice. RNA was extracted from frozen tissue using TRIzol (Invitrogen, NY) and converted to cDNA using the iScript cDNA synthesis kit (BioRad, CA). cDNA was analyzed in triplicate via quantitative RT-PCR for Gapdh (control) versus other target genes using a 7500 Real-Time PCR System (Applied Biosystems). Relative mRNA expression was calculated using the 2^−ΔΔdCT method.

2.7. Perfusion and Immunohistochemistry (IHC)

For immunohistochemical analysis, 10 to 12-wk old mice were treated with leptin (5 mg/kg, IP) or PBS and sacrificed 2 h later between 9 a.m. and 12 p.m. Perfusion and IHC procedures were performed essentially as described previously [23]. In brief, mice were deeply anesthetized with an overdose of pentobarbital (150 mg/kg, IP) and transcardially perfused with sterile PBS followed by 4% paraformaldehyde. Brains were removed, post-fixed and cryoprotected before sectioning into 30 μm coronal slices, which were collected into 4 representative series and stored at −20 °C until further use. For immunofluorescence, sections were pretreated in ice-cold methanol, 0.6% glycine and 0.03% SDS, and then blocked with donkey serum and incubated in the primary antibodies (rabbit-anti-pSTAT3 [1:500]). Detection of primary antibodies was done by immunofluorescence with donkey anti-rabbit-Alexa 488 [1:200].

2.8. Statistical analysis

Mean±SEM and differences were analyzed by unpaired Fisher's t-test to determine significant differences between groups (f/f versus +/+). Repeated measures ANOVA was performed with StatView. A log-rank (Mantel-Cox) test was performed for comparison of survival curves. Differences were accepted for p≤0.05.

3. Results

3.1. Generation of Lepr^F1077/F1077 mice

To directly assess the physiologic functions controlled by LepR-b Tyr₁₀₇₇, we generated a “knock-in” mouse model in which the endogenous Lepr gene is replaced by Lepr^F1077- encoding LepR-b ^F1077 (which contains a Tyr₁₀₇₇→Phe₁₀₇₇ substitution in LepR-b) (Fig. 1A and B). Lepr^F1077/+ animals were intercrossed to generate Lepr^F1077/F1077 (f/f) and Lepr^+/+ (+/+) control animals for study.

Fig. 1.

Generation of Lepr^F1077/F1077 mice. (A) Diagram of gene-targeting strategy to generate Lepr^F1077. Sequences from the targeting vector containing the Tyr₁₀₇₇→Phe₁₀₇₇ substitution replaced the endogenous Lepr exon 18b (which encodes the intracellular domain of LepR-b). (B) Diagram of signaling pathways mediated by LepR-b and LepR-b^F1077. (C) STAT3 signaling by LepR-b and LepR-b^F1077in vivo. Control (+/+) and f/f mice were treated with PBS or leptin (5 mg/kg, IP) for 2 h prior to perfusion and the examination to pSTAT3 (white nuclei) in the basomedial hypothalamus by immunofluorescence. 3V-third cerebral ventricle.

To confirm the expected expression pattern and unimpaired activity of LepR-b ^F1077 for signaling pathways other than those mediated by Tyr₁₀₇₇, we examined the stimulation of STAT3 phosphorylation (pSTAT3) by immunofluorescence in the basomedial hypothalamus of f/f and control mice following leptin or PBS treatment (Fig. 1C). As predicted, we detected similar patterns and levels of baseline and leptin-stimulated pSTAT3-immunoreactivity in f/f and +/+ mice, confirming the appropriate distribution and function of LepR-b ^F1077 in f/f mice.

3.2. Lepr^F1077/F1077 mice exhibit modestly increased energy balance despite normal hypothalamic gene expression

We examined body weight and adiposity in male and female control and f/f mice fed a normal chow (NC) or calorically-dense high fat (45%; HFD) diet from 4–12 wks of age (Fig. 2). On a NC diet, the body weight of f/f males was similar to controls although repeated measures ANOVA analysis revealed a statistically significant increase in body weight by genotype for females over the duration of the study (Fig. 2A and B), suggesting a potentially greater role in females for the control of energy balance by LepR-b Tyr₁₀₇₇. While not statistically different at any single age, HFD-challenged animals of both sexes were slightly heavier than controls over the duration of the study by repeated measures ANOVA (Fig. 2C and D). Fat mass was also modestly increased in both male and female NC-fed animals (Fig. 2E and F), although circulating leptin levels were only statistically elevated in female f/f mice (Fig. 2G and H). Although HFD tended to increase fat mass and leptin concentrations in all groups, no significant differences were detected between genotypes in HFD-fed animals (Fig. 2E–H ). No differences were detected in lean mass, snout–anus length, temperature, or other non-adipose tissue parameters between control and f/f mice (Tables 1 and 2).

Fig. 2.

Modestly positive energy balance in female Lepr^F1077/F1077 mice. (A–D) Body weight for males (A and C) and females (B and D) was measured weekly from 4–12 wks for animals fed NC (A and B) and HFD (C and D). E–F. Body adiposity for males (E) and females (F) as determined by fat pad mass at 15 wks of age. (G–H) Plasma leptin levels of males (G) and females (H) on NC and HFD. All data are plotted as mean±SEM, n≥10 per genotype. *p≤0.05 by Student's t-test for (E–H) and by repeated measures ANOVA for B,E,F.

Table 1.

Physiologic parameters for male Lepr^F1077/F1077 mice. At 14 wks of age, male animals of the indicated genotype were subjected to body composition determination. At 15 wks of age, animals were euthanized for the determination of snout to anus (S–A) length, and tissues were weighed. Measurements of serum leptin and insulin are from ad libitum-fed animals at 12 wks of age. Blood glucose measurements are from ad libitum-fed animals of 12 wks of age (or overnight (18-hour) fasted animals of 13 wks of age, where indicated). Temperature was taken in awake freely moving 12 wk old animals. All data are presented as mean±SEM, n≥10 per genotype per condition.

Male	NC		HFD
	+/+	f/f	+/+	f/f
Lean mass (g)	21.2±0.4	20.4±0.3	19.9±0.4	21.1±0.5
S–A length (mm)	92.2±0.7	91.7±0.5	91.1±0.9	91.5±1.4
WAT (g)	0.40±0.02	0.50±0.04⁎	1.07±0.11	1.37±0.18
BAT (g)	0.057±0.003	0.056±0.002	0.087±0.006	0.082±0.010
Liver (g)	0.85±0.03	0.88±0.02	0.70±0.03	0.79±0.05
Muscle (g)	0.138±0.004	0.131±0.002	0.135±0.003	0.130±0.003
Leptin (ng/ml)	2.4±0.4	2.9±0.4	7.1±1.0	11.1±2.4
Insulin (fed) (ng/ml)	1.9±0.3	2.3±0.4	3.3±0.4	3.0±0.5
Glucose (fed) (mmol/L)	6.9±0.2	6.8±0.2	6.8±0.2	7.2±0.3
Glucose (fasted) (mmol/L)	4.1±0.3	4.6±0.2	5.6±0.5	5.5±0.5
Temperature (°C)	37.6±0.2	40.0±0.3	38.4±0.2	38.3±0.2

^⁎p≤0.05 by Student's t-test for f/f compared to +/+ animals on the same diet.

Table 2.

Physiologic parameters for female Lepr^F1077/F1077 mice. At 14 wks of age, female animals of the indicated genotype were subjected to body composition determination. At 15 wks of age, animals were euthanized for the determination of snout to anus (S–A) length, and tissues were weighed. Measurements of serum leptin and insulin are from ad libitum-fed animals at 12 wks of age. Blood glucose measurements are from ad libitum-fed animals of 12 wks of age (or overnight (18-hour) fasted animals of 13 wks of age, where indicated). Temperature was taken in awake freely moving 12 wk old animals. All data are presented as mean±SEM, n≥10 per genotype per condition.

Female	NC		HFD
	+/+	f/f	+/+	f/f
Lean mass (g)	17.7±0.2	17.6±0.5	16.7±0.4	16.6±0.4
S–A length (mm)	88.2±0.6	87.8±0.7	87.8±0.7	87.5±0.7
WAT (g)	0.14±0.01	0.29±0.04⁎	0.30±0.05	0.37±0.07
BAT (g)	0.043±0.002	0.046±0.003	0.052±0.003	0.057±0.006
Liver (g)	0.75±0.02	0.78±0.03	0.55±0.02	0.58±0.03
Muscle (g)	0.109±0.002	0.106±0.003	0.107±0.002	0.108±0.003
Leptin (ng/ml)	1.1±0.1	1.6±0.3⁎	6.0±1.3	4.9±0.8
Insulin (fed) (ng/ml)	1.1±0.1	1.0±0.1	2.1±0.4	1.9±0.4
Glucose (fed) (mmol/L)	6.0±0.2	6.0±0.2	6.5±0.2	6.7±0.2
Glucose (fasted) (mmol/L)	4.4±0.2	3.6±0.4	4.2±0.3	4.4±0.5
Temperature (°C)	37.6±0.3	38.4±0.2	37.9±0.1	37.9±0.1

^⁎p≤0.05 by Student's t-test for f/f compared to +/+ animals on the same diet.

To determine the mechanisms underlying the increased body weight and adiposity of f/f mice, we examined food intake over the course of the study, as well as metabolic rate (Fig. 3). We detected no differences in feeding by genotype in NC-fed animals of either sex, but cumulative food intake was increased by approximately 5–10 % in the HFD-fed f/f mice of both sexes (Fig. 3A and B). While metabolic rate, as determined by oxygen consumption, and total activity did not differ between NC female control and f/f mice, male f/f mice displayed elevated dark and light cycle VO₂ that were independent of activity (which was not altered in f/f mice of either sex) (Fig. 3C, Supplemental Fig. 1). Increased energy expenditure in male f/f animals may contribute to the more modest body weight phenotype displayed by male relative to female f/f mice. Mutation of LepR-b Tyr₁₀₇₇ did not alter respiratory quotient (RQ) in either sex (Supplemental Fig. 1C and D). Thus, f/f animals of both sexes demonstrate slightly positive energy balance relative to controls, likely due to slightly increased food intake, although the differences between genotypes are small enough that they are not detectable under all dietary conditions.

Fig. 3.

Leptin action via Tyr₁₀₇₇ modestly regulates HFD-feeding and energy expenditure. (A–B) Food intake was measured weekly for male (A) and female (B) NC and HFD-fed control and f/f mice, and the total amount of food consumed over the 12-wk study period is shown. (C–F) Male (C and E) and female (D and F) mice were subjected to CLAMS analysis to determine VO₂ corrected for lean body mass (C and D), and total spontaneous motor activity (E and F). Data are shown for dark cycle (Dark), light cycle (Light) and averaged over 24 h (Total). All data are plotted as mean±SEM, n=4–8 per genotype. *p≤0.05 by Student's t-test.

Leptin modulates energy balance, at least in part, by promoting the expression of anorectic peptides such as those derived from proopiomelanocortin (POMC) and inhibiting the production of the orexigenic peptides (including neuropeptide Y (NPY) and by agouti-related peptide (AgRP)) in the arcuate nucleus of the hypothalamus (ARC) [24,25]. To examine whether altered expression of these hypothalamic neuropeptides might contribute to the modest defects in energy balance in f/f animals, we examined hypothalamic mRNA expression for these peptides in male and female NC and HFD fed control and f/f mice (Supplemental Fig. 2). The expression of Pomc, Npy, and Agrp were similar in f/f mice and controls, regardless of diet, suggesting that LEPR Tyr₁₀₇₇ does not modulate energy balance by impacting the expression of these neuropeptides.

3.3. Minimal alteration in glucose homeostasis in Lepr^F1077/F1077 mice

Despite the minimal effect of LepR-b Tyr₁₀₇₇ mutation on body weight, we examined glucose homeostasis in f/f mice to determine whether Tyr₁₀₇₇ might contribute to glucose homeostasis independently of energy balance. Neither glucose nor insulin concentrations differed by genotype in fed or fasted animals of either sex on NC or HFD (Tables 1 and 2). During an intravenous glucose tolerance test, glucose clearance and insulin secretion were similar between control and f/f mice of both sexes on either NC or HFD (Fig. 4A, B, E, F; Supplemental Fig. 2); area under the curve for both glucose and insulin were not different by genotype during this test (Fig. 4C, D, G, H). Male HFD f/f mice did trend towards overall impaired insulin secretion following glucose administration, however, and exhibited increased blood insulin levels relative to control mice 15 min following glucose injection (Fig. 4E)- consistent with modestly decreased insulin sensitivity resulting from slightly increased adiposity. Overall these data suggest that LepR-b Tyr₁₀₇₇ does not play a significant direct role in the regulation of glucose homeostasis and/or insulin secretion.

Fig. 4.

Preserved glucose stimulated insulin secretion in Lepr^F1077/F1077 mice. Glucose stimulated insulin secretion was assessed by intravenous glucose tolerance test (IVGTT) in males and females on HFD. (A and B). Plasma glucose for males (A) and females (B) during the IVGTT. (C and D). Area under the curve (AUC) for plasma glucose for males (C) and females (D) for the duration of the test. (E and F). Plasma insulin levels in male (E) and (F) female mice during the 60 min post-glucose bolus. (G and H). Area under the curve (AUC) for plasma insulin for male (G) and female (H) mice for the duration of the test. *p≤0.05 by Student's t-test and indicates time points in which groups statistically differed. All data are plotted as mean±SEM, n≥8 per genotype.

3.4. Impaired reproductive function in Lepr^F1077/F1077 females

To determine the potential contribution of LepR-b Tyr₁₀₇₇ to the control of reproduction by leptin, we examined estrus cycling in f/f and control females by daily monitoring of vaginal cytology (Fig. 5). We observed no significant difference in the time of vaginal opening (30.4±0.3 day for +/+ and 32±0.4 day for f/f) or in the time of onset of first estrus (45.7±1.3 day for +/+ and 44.2±1.4 day for f/f), suggesting that signaling by LepR-b Tyr₁₀₇₇ is not required for the neuroendocrine events that lead up to and initiate puberty. In contrast, continued monitoring of these animals revealed that f/f females tended to progress to a second estrus less frequently than controls (p=0.07) and those f/f animals that achieved a second estrus had inter-cycle intervals almost twice as long as controls (Fig. 5A and B). Thus, while LepR-b Tyr₁₀₇₇ is not required for the onset of fertility, it is required for ongoing appropriate function of the reproductive system. Hypothalamic expression of the mRNAs encoding kisspeptin (Kiss) and tachykinin-2/neurokinin B (Tac2), both of which are important for the control of reproduction [26,27], were not different between adult f/f and control animals (Supplemental Fig. 4), consistent with the essential lack of LepR-b expression in kisspeptin-containing hypothalamic neurons [28,29], and suggesting that the effect of LepR-b Tyr₁₀₇₇ on reproduction must be mediated independently of effects on the expression of these neuropeptides.

Fig. 5.

Disrupted estrous cycling in Lepr^F1077/F1077 females. Vaginal histology was examined daily in female +/+ and f/f mice from 28 day until 61 day of age (or until a second estrous cycle was observed). (A) Kaplan–Meier plot of time to attainment of second estrus cycle (VE2) for female +/+ and f/f mice revealed no statistically significant differences between groups as a whole (p=0.07); however, 84% of +/+ mice attained VE2 while only 58% of f/f mice did. (B) Time between VE1 and VE2 for animals that attained VE2 by the end of the study. Data are plotted as mean±SEM; *p≤0.05 by Student's t-test, n≥12 per genotype.

4. Discussion

The role for the LepR-b tyrosine phosphorylation site, Tyr₁₀₇₇, in leptin action has not previously been examined. Our data reveal that signaling via LepR-b Tyr₁₀₇₇ contributes to the control of energy balance, primarily by modulating food intake, and modulates the female reproductive axis following the onset of puberty. Although LepR-b Tyr₁₀₇₇ clearly plays a role in the control of energy balance, the contribution of this signaling pathway pales in comparison to that of Tyr₁₁₃₈, mutation of which doubles food intake, diminishes energy expenditure, and increases adiposity and leptin levels by approximately 10-fold [17–20]. Since the obesity of animals mutant for LepR-b Tyr₁₁₃₈ is slightly less than that of Lepr^db/db animals and Tyr₉₈₅ mutant animals are slightly lean [14,17–20], Tyr₁₀₇₇-mediated signaling likely contributes to the difference in energy balance between animals mutant for LepR-b Tyr₁₁₃₈ and Lepr^db/db animals. Mice mutant for all three intracellular tyrosine residues on LepR-b (LepR-b^3F) are not different than mice mutant solely for Tyr₁₁₃₈ in terms of energy balance and metabolism, however, and both of these animal models remain less impaired than Lepr^db/db animals [20]. Since the metabolic phenotype of mice containing a LepR-b deleted for the majority of the intracellular domain (LepR-b^Δ65) closely resembles that of Lepr^db/db animals [11], it is possible that a non-phosphorylated LepR-b motif, rather than Tyr₁₀₇₇, represents the major non-Tyr ₁₁₃₈ metabolic signal mediated by LepR-b . While signaling by LepR-b Tyr₁₀₇₇ modestly impacts feeding, it does not appear to directly impact parameters of glucose homeostasis, and the normal to increased energy expenditure in f/f animals suggests no direct role for Tyr₁₀₇₇ in the augmentation of metabolic rate by leptin.

Relative to the anovulatory infertility of Lepr^db/db mice on most genetic backgrounds, reproductive function is only modestly impacted in mice mutant for Tyr₁₁₃₈: While female Tyr₁₁₃₈-mutant animals exhibit slightly delayed onset of estrus, they ovulate and can reproduce [17–20]. The impaired progression of f/f mice to regular cyclicity and the increased interval between cycles suggests an important role for LepR-b Tyr₁₀₇₇ in the control of the reproductive axis by leptin. Indeed, while mice expressing LepR-b ^3F exhibit parameters of energy balance and metabolism similar to mice mutant only for Tyr₁₁₃₈ [20], reproduction in the LepR-b ^3F animals is severely impaired relative to Tyr₁₁₃₈ mutant mice; LepR-b ^3F mice demonstrate absolute infertility similar to Lepr^db/db animals and mice expressing LepR-b ^Δ65, consistent with the requirement for a LepR-b tyrosine phosphorylation site other than Tyr₁₁₃₈ in the control of reproduction by LepR-b . The impaired reproductive function of f/f mice, coupled with the lack of detectable changes to reproduction in mice mutant for LepR-b Tyr₉₈₅, suggest that Tyr₁₀₇₇ represents the site crucial for reproduction. Since f/f mice exhibit only minor increases in adiposity in the absence of other detectable metabolic or gene expression changes, the reproductive derangements in these animals likely result from direct effects of Tyr₁₀₇₇ signaling on components of the reproductive axis. In contrast, the incomplete fertility of mice mutant for LepR-b Tyr₁₁₃₈ could be secondary to the obesity, diabetes, and altered endocrine milieu (e.g., hypothyroidism, hypercortisolism) of these animals [17–20], rather than resulting from a direct effect of Tyr₁₁₃₈ signaling on the reproductive system.

Since Tyr₁₀₇₇ directly mediates the majority of STAT5 signaling by LepR-b [10,16], we postulate that the phenotype of the f/f animals stems from impaired LepR-b→STAT5 signaling. Unfortunately, leptin-induced phosphorylation of STAT5 does not appear as robust as phosphorylation of STAT3 and therefore, we were unable to assess the degree of impaired STAT5 signaling via immunoprecipitation and Westen blot in the f/f animals. Indeed, deletion of STAT5 from the CNS (Stat5^{fl/fl Nestin-cre} animals) produces mild obesity that manifests primarily in older animals [30]. Since young Stat5^{fl/fl Nestin-cre} animals were not studied for their endocrine phenotype, it is not possible to determine whether the impaired thermoregulation, and the hyperleptinemic and hyperinsulinemic phenotypes of these animals were primary and present before the onset of their obesity, or rather if these components of their phenotype were secondary to their increased adiposity. Therefore, we cannot directly compare to the metabolic phenotype of Stat5^{fl/fl Nestin-cre} to our LepR-b Tyr₁₀₇₇ mutated f/f animals. Additionally, STAT5 was deleted throughout the CNS in Stat5^{fl/fl Nestin-cre} animals, and some of their phenotype may stem from impaired signaling by other cytokines, such as GM-CSF, in non-LepR-b neurons [30,31]. Furthermore, no measures were taken to assess reproductive maturation or function in Stat5^{fl/fl Nestin-cre} mice, so we are unable to address whether the reproductive phenotype seen in female f/f mice is directly due to loss of LepR-b→Stat5 signaling. In the future, it will be important to directly compare the roles for STAT5 and Tyr₁₀₇₇ in LepR-b signaling in both metabolic and reproductive control.

The LepR-b-expressing cell type(s) and neural mechanisms responsible for mediating the effects of Tyr₁₀₇₇ and/or STAT5 on the control of energy balance and reproduction also remain unclear. As for Stat5^{fl/fl Nestin-cre} animals, which displayed no detectable alterations in the known direct neuropeptide transcriptional targets of leptin action [30], similar analysis in f/f mice revealed no alteration in the expression of hypothalamic neuropeptides known to be involved in energy balance (such as Pomc, Npy, and Agrp). Neither were we able to detect changes in the mRNA expression of known ARC mediators of reproductive control (such as Kiss, pDyn, and Tac2) in the f/f animals (Supplemental Fig. 4 and data not shown). Indeed, we and others have shown little overlap between LepR-b-expressing and kisspeptin-containing neurons, and leptin minimally, if at all, contributes to the control of transcription in these neurons under fed or fasted conditions [28,29,32] (and our data not shown). Along these lines, deletion of LepRbs from ARC specific POMC and AgRP cell populations does not impair fertility [33,34]. These data suggest that the LepR-b cells responsible for Tyr₁₀₇₇ mediated metabolic and reproductive effects likely reside outside of the classic ARC circuits where the soma containing these regulatory peptides reside. Leptin is known to be permissive of the GnRH surge in rodents, and acyclic db/db mice display impaired GnRH release and pituitary dysfunction. Although LepRbs are not directly present on GnRH neurons [29], it is possible that leptin-induced phosphorylation of STAT5 may be the signal, which indirectly controls leptin's permissive action on the GnRH surge and subsequent cyclicity. The anatomical target for LepR-b- Tyr₁₀₇₇ signaling, and ultimately leptin's overall reproductive effects may be the ventral premammilary nucleus (PMv); PMv LepR-b neurons have been shown to mediate leptin action on the reproductive system [28,35–38] and deletion of LepR-b from Nos1-expressing neurons, the majority of which lie in the PMv, interferes with female reproductive function [39]. Unfortunately, transcriptional targets for leptin have not been identified in the PMv. Identifying leptin-regulated PMv transcripts and determining roles for specific leptin signals, such as Tyr₁₀₇₇, in their regulation will be crucial for our long-term understanding of the mechanisms by which leptin modulates both energy homeostasis and the reproductive axis.

Conflict of interest

None declared.

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.molmet.2012.05.001.

Appendix A. Supplementary materials