Opposite effects of a high-fat diet and calorie restriction on ciliary neurotrophic factor signaling in the mouse hypothalamus

Abstract

In the mouse hypothalamus, ciliary neurotrophic factor (CNTF) is mainly expressed by ependymal cells and tanycytes of the ependymal layer covering the third ventricle. Since exogenously administered CNTF causes reduced food intake and weight loss, we tested whether endogenous CNTF might be involved in energy balance regulation. We thus evaluated CNTF production and responsiveness in the hypothalamus of mice fed a high-fat diet (HFD), of ob/ob obese mice, and of mice fed a calorie restriction (CR) regimen. RT-PCR showed that CNTF mRNA increased significantly in HFD mice and decreased significantly in CR animals. Western blotting confirmed that CNTF expression was higher in HFD mice and reduced in CR mice, but high interindividual variability blunted the significance of these differences. By immunohistochemistry, hypothalamic tuberal and mammillary region tanycytes stained strongly for CNTF in HFD mice, whereas CR mice exhibited markedly reduced staining. RT-PCR and Western blotting disclosed that changes in CNTF expression were paralleled by changes in the expression of its specific receptor, CNTF receptor α (CNTFRα). Injection of recombinant CNTF and detection of phospho-signal transducer and activator of transcription 3 (P-STAT3) showed that CNTF responsiveness by the ependymal layer, mainly by tanycytes, was higher in HFD than CR mice. In addition, in HFD mice CNTF administration induced distinctive STAT3 signaling in a large neuron population located in the dorsomedial and ventromedial nuclei, perifornical area and mammillary body. The hypothalamic expression of CNTF and CNTFRα did not change in the hyperphagic, leptin-deficient ob/ob obese mice; accordingly, P-STAT3 immunoreactivity in CNTF-treated ob/ob mice was confined to ependymal layer and arcuate neurons. Collectively, these data suggest that hypothalamic CNTF is involved in controlling the energy balance and that CNTF signaling plays a role in HFD obese mice at specific sites.

Keywords: CNTF receptor; STAT3; ependyma; fasting; median eminence; obesity; tanycytes; third ventricle.

Figure 1

(A) trends of the body weight of mice fed an HFD diet, a low-fat diet (Ctrl), or a CR regimen for 12 weeks. RT-PCR analysis of NPY (B) and POMC (C) expression in the hypothalamus of HFD, control (Ctrl) and CR mice. Mean ± SEM. ^*P < 0.05.

Figure 2

RT-PCR (A and C) and Western blot (B and D) analyses of CNTF (A and B) and CNTFRα (C and D) expression in the hypothalamus of HFD, control (Ctrl) and CR mice. For Western blots the densitometric analysis was normalized to β-tubulin expression. In (B) recombinant CNTF (P, 30 ng) loaded together with hypothalamic protein extracts served as a positive control. Mean ± SEM. ^*P < 0.05.

Figure 3

Immunohistochemical detection of CNTF in the mouse tuberal hypothalamus. Some CNTF-stained tanycytes on the ventrolateral wall of the third ventricle (3V) in control mouse (A). In an HFD mouse (B), strongly increased CNTF staining of tanycytes also results in more (inset of B, arrow) or less (inset of B, arrowheads) intense staining of their long projections. In contrast, tanycytes in CR mice (C) exhibit very low-level staining. The insets of (A,B) are the enlargements of the corresponding framed areas. ME, median eminence. Bar: (A–C) 400 μm; inset of (A) 40 μm; inset of (B) 300 μm.

Figure 4

CNTF/P-STAT3 double-labeling of coronal sections of the tuberal hypothalamus from CNTF-treated control (A–C), HFD (D–F) and CR (G–I) mice. Compared with controls (A,B) the greater CNTF expression seen in tanycytes of HFD mice (D) is matched by increased CNTF responsiveness (E). In CR mice both CNTF expression (G) and responsiveness (H) are lower than in control mice and, especially, HFD mice. Bar: 100 μm.

Figure 5

Percentage of CNTF-producing (A) and CNTF-responsive (B) tuberal and mammillary tanycytes in HFD, control (Ctrl) and CR mice. Morphometric evaluation performed by confocal microscopy in double-stained sections where tanycytes were labeled by GFAP and CNTF responsiveness was evaluated by P-STAT3 immunohistochemistry in CNTF-treated mice. Mean ± SEM. ^*P < 0.05.

Figure 6

P-STAT3 immunohistochemistry in the tuberal region of CNTF-treated mice. In a control mouse (A), P-STAT3 staining depicts ependymal cells (ep) and neurons located in the arcuate nucleus (Arc) and median eminence (ME). In an HFD mouse (B), P-STAT3 immunoreactivity increases at these sites, but responsiveness to CNTF is also detectable in the dorsomedial hypothalamus (DM) and, to a lesser extent, the ventromedial nucleus (VMH). CNTF responsiveness is blunted in a CR mouse (C). The inset of (C) shows a representative Nissl-stained coronal section of the tuberal region of the hypothalamus. (D) is the enlargement of the area framed in (B), where P-STAT3 immunoreactivity is found not only in cell nuclei but also in varicose projections (arrowheads). In (E), double-staining immunofluorescence and confocal microscopy show that in the DM of a CNTF-treated HFD mouse P-STAT3-positive cell nuclei are contained in HuC/D-positive neuronal cell bodies. 3V, third ventricle; v, blood vessel. Bar: (A–C) 100 μm; (D) 20 μm; (E) 15 μm; inset of (C) 400 μm.

Figure 7

RT-PCR analysis of CNTF (A) and CNTFRα (B) in the hypothalamus of wild type (wt), ob/+ and ob/ob mice. Mean ± SEM.