Complex regulation of mammalian target of rapamycin complex 1 in the basomedial hypothalamus by leptin and nutritional status

Abstract

The medial basal hypothalamus, including the arcuate nucleus (ARC) and the ventromedial hypothalamic nucleus (VMH), integrates signals of energy status to modulate metabolism and energy balance. Leptin and feeding regulate the mammalian target of rapamycin complex 1 (mTORC1) in the hypothalamus, and hypothalamic mTORC1 contributes to the control of feeding and energy balance. To determine the mechanisms by which leptin modulates mTORC1 in specific hypothalamic neurons, we immunohistochemically assessed the mTORC1-dependent phosphorylation of ribosomal protein S6 (pS6). In addition to confirming the modulation of ARC mTORC1 activity by acute leptin treatment, this analysis revealed the robust activation of mTORC1-dependent ARC pS6 in response to fasting and leptin deficiency in leptin receptor-expressing Agouti-related protein neurons. In contrast, fasting and leptin deficiency suppress VMH mTORC1 signaling. The appropriate regulation of ARC mTORC1 by mutant leptin receptor isoforms correlated with their ability to suppress the activity of Agouti-related protein neurons, suggesting the potential stimulation of mTORC1 by the neuronal activity. Indeed, fasting- and leptin deficiency-induced pS6-immunoreactivity (IR) extensively colocalized with c-Fos-IR in ARC and VMH neurons. Furthermore, ghrelin, which activates orexigenic ARC neurons, increased ARC mTORC1 activity and induced colocalized pS6- and c-Fos-IR. Thus, neuronal activity promotes mTORC1/pS6 in response to signals of energy deficit. In contrast, insulin, which activates mTORC1 via the phosphatidylinositol 3-kinase pathway, increased ARC and VMH pS6-IR in the absence of neuronal activation. The regulation of mTORC1 in the basomedial hypothalamus thus varies by cell and stimulus type, as opposed to responding in a uniform manner to nutritional and hormonal perturbations.

Figure 1

Phosphorylation of S6 in response to fasting and leptin in the MB-ARC and VMH. A, Wild-type (+/+) animals under ad libitum-fed (Ad Lib Fed) or 24-h-fasted conditions treated with leptin (Lep; 5 mg/kg, ip) or vehicle (Veh) for 1 h before the animals were killed. B, Ad libitum-fed Lep^ob/ob mice. C, Overnight (∼16-h) fasted +/+ mice (treated with leptin or vehicle, as indicated) that were processed for immunohistochemical analysis of pS6-IR. Top panels show representative images of immunofluorescent detection of pS6-IR, whereas the graphs below show the fold induction (or total counts) of the number of pS6 cells in the MB-ARC and VMH. Data were quantified and plotted as mean ± sem (n ≥ 3 per condition; bars with different letters were significantly different by ANOVA, P < 0.05). *, P < 0.01; **, P < 0.001 by Student’s t test; P value of trend as indicated in B. 3V, Third ventricle.

Figure 2

S6K1/mTORC1-dependence of pS6-IR in the MB-ARC. A, Wild-type (+/+) or S6K1^−/− (KO) animals under ad libitum-fed or 24-h-fasted conditions were perfused and processed for the immunofluorescent detection of pS6-IR. Top panel shows representative images. Numbers of pS6-IR neurons in the MB-ARC (mean ± sem) for each genotype are presented in the graph (n ≥ 3 per condition). **, P <.001 by Student’s t test. B, Wild-type animals under the indicated ad libitum (Ad Lib)-fed or fasted conditions were treated with icv vehicle or rapamycin (2 μg/μl) overnight and 2 h before the animals were killed and processing for the immunofluorescent detection of pS6-IR. Representative images under each condition are shown. In addition to the MB-ARC pS6-IR that is the focus of this analysis, note some inflammatory induction of pS6-IR in the ependymal cells that line the ventricle, which is also blocked by treatment with rapamycin. 3V, Third ventricle.

Figure 3

Fasting induces pS6-IR in LepRb- and AgRP-expressing neurons in the MB-ARC. A, LepRb^EGFP mice were allowed to feed ad libitum (Ad Lib Fed) or were fasted for 24 h before treatment with leptin (Lep; 5 mg/kg, ip) or vehicle (Veh) for 1 h before the animals were killed and processing for the immunofluorescent detection of LepRb/EGFP (green) expression and pS6-IR (red). Top panels show representative images of pS6-IR alone (left) and merged pS6-IR and EGFP-IR images (right). LepRb/EGFP neurons that contain pS6-IR are plotted as mean ± sem (n ≥ 3 per condition; bars with different letters were significantly different by ANOVA, P < 0.05). B, Animals heterozygous for Agrp^LacZ were allowed to feed ad libitum or were fasted for 24 h before treatment with leptin (Lep; 5 mg/kg, ip) or vehicle (Veh) for 1 h before the animals were killed and processing for the immunofluorescent detection of AgRP/LacZ (red)-expression and pS6-IR (green). Top panels show representative images of AgRP (β-gal) alone (top), pS6-IR alone (middle), and merged β-gal and pS6-IR images (bottom). AgRP (β-gal) neurons that contain pS6-IR are plotted as mean ± sem (n ≥ 3 per condition; bars with different letters were significantly different by ANOVA, P < 0.05). 3V, Third ventricle.

Figure 4

pS6-IR in the medial basal hypothalamus of mouse models with specific alterations in LepRb signaling. A, Ad libitum-fed (Ad Lib Fed) wild-type (+/+), db/db, s/s, and l/l mice were perfused and processed for the immunofluorescent detection of pS6-IR. Representative images of pS6-IR from each genotype are shown. B, Ad libitum-fed and 24-h-fasted l/l mice were treated with leptin (Lep) or PBS, as indicated, perfused, and processed for the immunofluorescent detection of pS6-IR. Representative images are shown for each panel. Veh, vehicle; 3V, third ventricle.

Figure 5

pS6-IR localization in activated neurons in the MB-ARC and VMH. A, Wild-type animals were allowed to feed ad libitum or were fasted for 24 h before treatment with leptin (Lep; 5 mg/kg, ip) or vehicle (Veh) for 1 h before the animals were killed and processing for the immunofluorescent detection of c-Fos-IR (red) and pS6-IR (green). Top panels show representative images of merged images at each condition. Smaller panels beneath show digital zoom of each channel and merged images from the boxed area in the larger panels above. Graphs: Total numbers of pS6-IR, c-Fos-IR, and percent double-labeled neurons in the MB-ARC and VMH are plotted as mean ± sem (n ≥ 3 per condition; bars with different letters were significantly different by ANOVA, P < 0.05). B, Wild-type animals under the indicated ad libitum-fed (Ad Lib Fed) or overnight (∼16 h) conditions were treated with icv vehicle or rapamycin ((Rapa) 2 μg/μl) via indwelling catheters overnight and for additional 2 h before the animals were killed and processing for the immunofluorescent detection of pS6-IR (green) and c-Fos-IR (red). Representative images under each condition are shown. In addition to the MB-ARC pS6/c-Fos-IR that is the focus of this analysis, note some inflammatory induction of pS6-IR in the ependymal cells that line the ventricle, which is also blocked by treatment with rapamycin. 3V, Third ventricle.

Figure 6

Ghrelin-induced neuronal firing promotes mTORC1 activation in the MB-ARC. Ad libitum-fed wild-type animals with indwelling icv catheters were treated with vehicle (Veh) or ghrelin (2 μg/μl) for 1 h before the animals were killed and processing for the immunofluorescent analysis of pS6-IR (green) and c-Fos-IR (red). Top panels: Representative merged images. Panels beneath the image from the ghrelin-treated sample represent digital zoom of individual channels and merged images from the boxed area in the larger image. Graph: Total numbers of pS6-IR, c-Fos-IR, and double-labeled neurons are plotted as mean ± sem (n ≥ 3 per condition; *, P ≤ 0.01 by Student’s t test). 3V, Third ventricle.

Figure 7

Insulin induces mTORC1 activation independently of c-Fos in the hypothalamus. A, Ad libitum-fed wild-type animals with indwelling icv catheters were treated with vehicle (Veh) or insulin (Ins) (300 mU/ml) for 1 h before the animals were killed and processing for the immunofluorescent analysis of pS6-IR (green) and c-Fos-IR (red). B, Ad libitum wild-type animals were treated with vehicle (Veh) or insulin (Ins) (400 mU/ml) for 1 h before the animals were killed and processing for the immunofluorescent analysis of pS6-IR (green) and c-Fos-IR (red). Top panels, Representative merged images. Graphs, Total numbers of pS6-IR, c-Fos-IR, and double-labeled neurons, along with percent double-labeled neurons, are plotted as mean ± sem (n ≥ 3 per condition; *, P ≤ 0.01 by Student’s t test). 3V, Third ventricle.