Disruption of hypothalamic leptin signaling in mice leads to early-onset obesity, but physiological adaptations in mature animals stabilize adiposity levels

Abstract

Distinct populations of leptin-sensing neurons in the hypothalamus, midbrain, and brainstem contribute to the regulation of energy homeostasis. To assess the requirement for leptin signaling in the hypothalamus, we crossed mice with a floxed leptin receptor allele (Leprfl) to mice transgenic for Nkx2.1-Cre, which drives Cre expression in the hypothalamus and not in more caudal brain regions, generating LeprNkx2.1KO mice. From weaning, LeprNkx2.1KO mice exhibited phenotypes similar to those observed in mice with global loss of leptin signaling (Leprdb/db mice), including increased weight gain and adiposity, hyperphagia, cold intolerance, and insulin resistance. However, after 8 weeks of age, LeprNkx2.1KO mice maintained stable adiposity levels, whereas the body fat percentage of Leprdb/db animals continued to escalate. The divergence in the adiposity phenotypes of Leprdb/db and LeprNkx2.1KO mice with age was concomitant with increased rates of linear growth and energy expenditure in LeprNkx2.1KO mice. These data suggest that remaining leptin signals in LeprNkx2.1KO mice mediate physiological adaptations that prevent the escalation of the adiposity phenotype in adult mice. The persistence of severe adiposity in LeprNkx2.1KO mice, however, suggests that compensatory actions of circuits regulating growth and energy expenditure are not sufficient to reverse obesity established at an early age.

Figure 1. The Nkx2.1-Cre driver is expressed broadly in the hypothalamus, but not in the midbrain or brainstem.

X-gal staining in sagittal (A) and coronal (B–E) sections of Nkx2.1-Cre;ROSA26LacZ mice. (A) Sagittal section approximately 200 μm from midline. Labeled lines correspond to approximate locations of coronal sections in B–E: medial preoptic area (B), suprachiasmatic nucleus and paraventricular nucleus (C), ARH (D), and ventral premammillary nucleus (E). Scale bars: 2 mm (A); 500 μm (B–E).

Figure 2. Disruption of leptin signaling is localized to the hypothalamus in Lepr^Nkx2.1KO mice.

(A–D) Immunohistochemical detection of p-STAT3 following overnight fast and subsequent treatment with leptin. p-STAT3 staining was detected in the ARH, VMH, and DMH (A) and the NTS (C) of control animals. p-STAT3 staining was not observed in the ARH, VMH, or DMH (B), but was seen in the NTS (D), of Lepr^Nkx2.1KO animals. White outlines in A and B denote tissue edges. Scale bars: 100 μm (A and B); 50 μm (C and D).

Figure 3. Growth rate and fat deposition are increased in young Lepr^Nkx2.1KO mice.

(A) Weights of male and female Lepr^Nkx2.1KO and control mice; n ≥ 10 for all groups at all points. (B) Naso-anal length over time of male and female Lepr^Nkx2.1KO and control mice; n ≥ 5 for all groups at all points. (C) Adiposity, as measured by NMR, of male Lepr^Nkx2.1KO, control, and Lepr^db/db mice over time; n ≥ 4 for all groups at all points. (D) Adiposity, as measured by NMR, of female Lepr^Nkx2.1KO and control mice over time; n ≥ 5 for all groups at all points. (E) Plasma leptin, as measured by ELISA, at 4, 8, and 12 weeks of male and female Lepr^Nkx2.1KO and control mice; n ≥ 5 for all groups at all points. (A–E) Results are mean ± SEM. *P < 0.01, Lepr^Nkx2.1KO versus control; ^†P < 0.01, Lepr^db/db versus control; ^‡P < 0.01, Lepr^db/db versus Lepr^Nkx2.1KO. P values were calculated between age- and sex-matched groups.

Figure 4. Hyperphagia in Lepr^Nkx2.1KO mice.

(A) Daily food intake of male control, Lepr^Nkx2.1KO, and Lepr^db/db mice at several ages; n ≥ 3 for all groups at all points. (B) Absolute lean body mass of male control, Lepr^Nkx2.1KO, and Lepr^db/db mice over time; n ≥ 4 for all groups at all points. (C) Daily food intake of male control, Lepr^Nkx2.1KO, and Lepr^db/db mice normalized to lean body mass at several ages; n ≥ 3 for all groups at all points. (A–C) Results are mean ± SEM. *P < 0.05, **P < 0.01, Lepr^Nkx2.1KO versus control; ^†P < 0.05, ^††P < 0.01, Lepr^db/db versus control; ^‡P < 0.05, ^‡‡P < 0.01, Lepr^db/db versus Lepr^Nkx2.1KO. P values were calculated between age-matched groups.

Figure 5. Increased energy expenditure in Lepr^Nkx2.1KO vs Lepr^db/db mice.

(A) Daily absolute oxygen consumption (VO₂) of male control, Lepr^Nkx2.1KO, and Lepr^db/db mice at 3 different ages. (B) Daily oxygen consumption normalized to lean body mass of male control, Lepr^Nkx2.1KO, and Lepr^db/db mice at 3 different ages. (C) Locomotor activity, as measured by daily beam breaks, in male control, Lepr^Nkx2.1KO, and Lepr^db/db mice at 3 different ages. (A–C) n ≥ 3 for all groups at all points. Results are mean ± SEM. **P < 0.01, Lepr^Nkx2.1KO versus control; ^†P < 0.05, ^††P < 0.01, Lepr^db/db versus control; ^‡P < 0.05, ^‡‡P < 0.01, Lepr^db/db versus Lepr^Nkx2.1KO. P values were calculated between age-matched groups.

Figure 6. Dysregulation of acute thermogenesis in young Lepr^Nkx2.1KO mice.

(A) Baseline core body temperatures in ad libitum–fed or fasting male control, Lepr^Nkx2.1KO, and Lepr^db/db mice at young and adult time points. (B) Response to acute cold challenge in Lepr^Nkx2.1KO mice at 3, 6, and 12–16- weeks of age and in 3-week-old control mice. (C and D) H&E stain of BAT from 3-week-old control (C) and 3-week-old Lepr^Nkx2.1KO (D) mice. Scale bars: 50 μm. (E) Quantitative PCR analysis of Ucp1 in BAT in control and Lepr^Nkx2.1KO mice at 2 time points. (A, B, and E) n ≥ 4 for all groups at all points. Results are mean ± SEM. **P < 0.01, Lepr^Nkx2.1KO versus control; ^††P < 0.01, Lepr^db/db versus control; ^‡P < 0.05, Lepr^db/db versus Lepr^Nkx2.1KO; ^#P < 0.05, 3-week-old Lepr^Nkx2.1KO versus all other groups; ^ΨP < 0.05 as indicated by brackets.

Figure 7. Activation of the adrenal and thyroid access differs in Lepr^Nkx2.1KO versus control mice.

(A) Morning CORT levels in male control, Lepr^Nkx2.1KO, and Lepr^db/db mice at 8 weeks; n > 3 for all groups. (B) Serum T4 levels in male control and Lepr^Nkx2.1KO mice at 13–15 weeks; n = 8 for all groups. (A and B) Results are mean ± SEM. *P < 0.05, Lepr^Nkx2.1KO versus control; ^††P < 0.01, Lepr^db/db versus control.

Figure 8. Impaired glucose homeostasis in Lepr^Nkx2.1KO mice.

(A) Random-fed whole-blood glucose of Lepr^Nkx2.1KO and control mice at several time points. (B) Fasted whole-blood glucose of Lepr^Nkx2.1KO and control mice at 6 and 10 weeks. (C) Random-fed plasma insulin, as measured by ELISA, of Lepr^Nkx2.1KO and control mice at 4, 8, and 12 weeks. (D) Fasted plasma insulin, as measured by ELISA, of Lepr^Nkx2.1KO and control mice at 6 and 10 weeks. (E) GTT of Lepr^Nkx2.1KO and control mice at 5 weeks of age. 2 mg/kg dextrose was injected i.p. at time 0. (F) Area under the curve calculation of GTT at 5 weeks of age. (A–F) n ≥ 5 for all groups at all points. Results are mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001. P values were calculated between age- and sex-matched groups.