Important roles of brain-specific carnitine palmitoyltransferase and ceramide metabolism in leptin hypothalamic control of feeding

Abstract

Brain-specific carnitine palmitoyltransferase-1 (CPT-1c) is implicated in CNS control of food intake. In this article, we explore the role of hypothalamic CPT-1c in leptin's anorexigenic actions. We first show that adenoviral overexpression of CPT-1c in hypothalamic arcuate nucleus of rats increases food intake and concomitantly up-regulates orexigenic neuropeptide Y (NPY) and Bsx (a transcription factor of NPY). Then, we demonstrate that this overexpression antagonizes the anorectic actions induced by central leptin or compound cerulenin (an inhibitor of fatty acid synthase). The overexpression of CPT-1c also blocks leptin-induced down-regulations of NPY and Bsx. Furthermore, the anorectic actions of central leptin or cerulenin are impaired in mice with brain CPT-1c deleted. Both anorectic effects require elevated levels of hypothalamic arcuate nucleus (Arc) malonyl-CoA, a fatty acid-metabolism intermediate that has emerged as a mediator in hypothalamic control of food intake. Thus, these data suggest that CPT-1c is implicated in malonyl-CoA action in leptin's hypothalamic anorectic signaling pathways. Moreover, ceramide metabolism appears to play a role in leptin's central control of feeding. Leptin treatment decreases Arc ceramide levels, with the decrease being important in leptin-induced anorectic actions and down-regulations of NPY and Bsx. Of interest, our data indicate that leptin impacts ceramide metabolism through malonyl-CoA and CPT-1c, and ceramide de novo biosynthesis acts downstream of both malonyl-CoA and CPT-1c in mediating their effects on feeding and expressions of NPY and Bsx. In summary, we provide insights into the important roles of malonyl-CoA, CPT-1c, and ceramide metabolism in leptin's hypothalamic signaling pathways.

Fig. 1.

CPT-1c modulates the anorectic actions of leptin and cerulenin. The adenovirus that encodes human CPT-1c and coexpresses GFP (Ade-CPT-1c) was delivered into the Arc of rats. The adenoviral vector expressing GFP alone was used as the null control (Ade-null). After 4 d following the delivery of the viruses, the rats were killed. (A) CPT-1c and CPT-1a protein levels in the Arc were quantitated (n = 10). Two representative blots of CPT-1c, GFP and actin from each treated group are shown. The Arc mRNA levels of human (h)-CPT-1c (n = 12), rat-CPT-1a (n = 4–5), and CPT-2 (n = 6) were quantitated and the mRNA levels of (h)-CPT-1c are shown. Rat-CPT-1a: Ade-null, 100 ± 4.4%, Ade-(h)-CPT-1c, 94 ± 4.9%; P > 0.05. CPT-2: Ade-null, 100 ± 6.8%, Ade-(h)-CPT-1c, 96 ± 12%; P > 0.05. Arc levels of long-chain acylcarnitines (palmitoylcarnitine, stearoylcarnitine, and oleoylcarnitine) were measured (n = 5). (B) After 4 d following the delivery of the viruses, the rats were fasted over a dark cycle (12 h). Then, some rats were provided with food and the food intakes at 24 h following food presentation are shown (n = 9–13). The other rats were killed after the fast. The mRNA levels of NPY, AgRP, POMC, and Bsx were measured (NPY: n = 12; AgRP and POMC: n = 10; Bsx: n = 4–6). (C) After 4 d following the delivery of the viruses, the rats received an intra-Arc infusion of leptin (1 μg) or cerulenin (40 μg) at 2 h before the onset of the dark cycle and food intakes and body weights were monitored (n = 5). The percent inhibitions of food intake at 24 h following the treatments are shown. The changes in 24-h body weight are also presented. (D) The rats with Arc overexpression of CPT-1c received an Arc infusion of leptin as described in C and were killed after the dark cycle (12 h). The mRNA levels of NPY and Bsx mRNA were quantitated. The percent decreases in NPY or Bsx mRNA level by leptin (relative to PBS) are shown (NPY: n = 12; Bsx: n = 11–12). (E) The mice with brain CPT-1c deleted (CPT-1c knockout) or their wild-type littermates were fasted over a dark cycle (12 h). Then, the mice received an intracerebroventricular injection of leptin (4 μg in PBS) or cerulenin (125 μg in 25% DMSO). For both treatments (leptin, n = 4–5; cerulenin, n = 4–5), the percent inhibitions (relative to the vehicle) of food intake at 24 h following the injections are shown. The changes in 24-h body weight are also shown. WT, wild-type; KO, CPT-1c knockout.

Fig. 2.

Arc ceramide metabolism is required in leptin's anorectic actions. (A) CPT-1c overexpression: the Ade-CPT-1c or Ade-GFP (null) was delivered into the Arc of rats. The rats were killed following 8-h food deprivation and ceramide levels in the Arc were measured (n = 7). CPT-1c deletion: The CPT-1c knockout mice and their wild-type littermates were killed under ad libitum-fed conditions. The ceramide levels in the mediobasal hypothalamus encompassing the Arc were measured (n = 5–6). (B) Before the dark cycle, leptin (1 μg in PBS) was infused into the Arc of rats. Ceramide levels (at 8 h following the infusion) were measured (n = 8–12). (C) The rats with Arc overexpression of CPT-1c or GFP (null), under ad libitum-fed conditions, received an intra-Arc infusion of leptin (1 μg in PBS) before the dark cycle. Arc ceramide levels (at 8 h following the infusion) in these rats were measured. The percent decreases in ceramide level by leptin (relative to PBS) are shown (n = 8–9) (note: the CPT-1c overexpression per se under ad libitum-fed conditions has no significant effect on the ceramide level.). (D) The Ade-MCD or Ade-GFP (null) was delivered into the Arc of rats. After 4 d following the delivery of the viruses, some rats were killed after a period (8 h) of food deprivation. Arc ceramide levels in these rats were measured (n = 5). The other rats, under ad libitum-fed conditions, received an intra-Arc infusion of leptin (1 μg in PBS) before the dark cycle. Arc ceramide levels (at 8 h following the infusion) were measured. The percent decreases in ceramide level by leptin (relative to PBS) are shown (n = 8–9) (note: the MCD overexpression of per se under ad libitum-fed conditions has no significant effect on the ceramide level). (E and F) At 4 h before the dark onset, N-Hexanoyl-d-sphingosine (2.5 μg in DMSO), a cell-permeable analog of natural ceramide (Cer-analog), was infused into the Arc. Two hours later (i.e., 2 h before the dark onset), the rats received an intra-Arc infusion of leptin (1 μg in PBS). (E) Some rats received food at the onset of the dark cycle and the food intakes and body weights at 24 h following leptin infusion are shown (n = 6). (F) The other rats were fasted over the dark cycle (12 h) and were then killed. The mRNA levels of NPY and Bsx in these rats were quantitated (n = 5–6). (G) Food was removed before onset of the dark cycle and the rats received one intra-Arc infusion of myriocin (4 μg in DMSO). Six hours later, these rats received the other intra-Arc infusion of myriocin (4 μg in DMSO). After another 2-h period for recovery, some rats were killed for ceramide assay (n = 5) (note: to this end, these rats had been fasted for 8 h). The other rats were presented with food. The food intakes and body weights at 24 h after the first injection are shown (n = 12–13). (H) The rats received the double myriocin treatment as described in G. After 12 h following the first infusion, the rats were killed. The mRNA levels of Arc neuropeptides (n = 6–9) and Bsx (n = 6–8) in these rats were measured.

Fig. 3.

Ceramide de novo biosynthesis mediates malonyl-CoA and CPT-1c actions on food intake and body weight. (A and B) The rats with Arc overexpression of MCD (n = 8), CPT-1c (n = 12), or GFP (null, n = 16) received a single intra-Arc infusion of myriocin (4 μg in DMSO) before the dark cycle and food was removed. Eight hours later, (A) some rats were killed and were then used for ceramide assay; (B) the other rats were provided with food and the food intakes at 24 h following the infusion are shown. (MCD: n = 5–6; CPT-1c: n = 7–10). (C and D) The rats with Arc overexpression of MCD, CPT-1c, or GFP (null) received myriocin treatment as described above. The rats were killed after the dark cycle (12 h) during which period no food was present. The mRNA levels of NPY, AgRP (C) (n = 5–7) and Bsx (D) (n = 5–6) in these rats were quantitated.