Ribosomal S6K1 in POMC and AgRP Neurons Regulates Glucose Homeostasis but Not Feeding Behavior in Mice

Summary Hypothalamic ribosomal S6K1 has been suggested as a point of convergence for hormonal and nutrient signals in the regulation of feeding behavior, bodyweight, and glucose metabolism. However, the long-term effects of manipulating hypothalamic S6K1 signaling on energy homeostasis and the cellular mechanisms underlying these roles are unclear. We therefore inactivated S6K1 in pro-opiomelanocortin (POMC) and agouti-related protein (AgRP) neurons, key regulators of energy homeostasis, but in contrast to the current view, we found no evidence that S6K1 regulates food intake and bodyweight. In contrast, S6K1 signaling in POMC neurons regulated hepatic glucose production and peripheral lipid metabolism and modulated neuronal excitability. S6K1 signaling in AgRP neurons regulated skeletal muscle insulin sensitivity and was required for glucose sensing by these neurons. Our findings suggest that S6K1 signaling is not a general integrator of energy homeostasis in the mediobasal hypothalamus but has distinct roles in the regulation of glucose homeostasis by POMC and AgRP neurons.


Genotyping
For detection of Cre-mediated excision of exons 3 and 4 of Rps6kb1 in the hypothalami of AgRPS6K1KO and POMCS6K1KO mice, genomic DNA was isolated from the hypothalamus, cortex, liver and skeletal muscle as previously described . The generation of a ~500 bp DNA product following PCR with primers TCCACCCACCAGTAAAGAGC and CCTCAGTCTCCTGAGTGTTAAGG is indicative of the floxed allele, whereas deletion is denoted by a ~450 bp band generated by primers TCCACCCACCAGTAAAGAGC and ACAAGAGGGCCAGTTGATGG.

Metabolic and food intake studies
Studies were performed in the animal's home cage unless indicated. Bodyweights from group-housed mice were measured weekly at 9-10am until 26 weeks of age. Glucose (i.p. 2 g/kg) and insulin (i.p. 0.75 U/kg) -tolerance tests were performed at 8 and 26 weeks of age, as previously described Choudhury et al., 2005). At 34 weeks of age, fed mice were weighed prior to EchoMRI analysis of body composition. Fasted tail blood was analyzed for serum leptin (Millipore) and insulin (Crystal Chem.) by ELISA and for FFA (WAKO Chem.) and triglycerides (Sigma Aldrich). Serum corticosterone levels were determined by ELISA (Immunodiagnostics Systems) from fed mice. For food intake studies, mice were group housed until 10 weeks of age and then singly housed. Mice were allowed to acclimatize for 2 weeks and periodically fasted overnight to acclimatize them. Ad-libitum food intake was measured over 3 consecutive days and for 24 h following an overnight fast.
Food intake was measured from singly housed mice injected with either vehicle or leptin (i.p. 0.3 mg/kg) at 9am and again at 6pm for 3 consecutive days. Food intake was also measured in fed mice for 8-24 hours following the injection of ghrelin (i.p. 5 mg/kg) or melanotan-II (MTII, i.p. 2 mg/kg) at 9am. Stainless steel cannulae (Plastics One Inc.) were stereotaxically implanted into the lateral cerebral ventricle so that the cannula tip was 0.4 mm posterior to bregma, 1.0 mm lateral to the midline and -2.1 mm ventral to the surface of the skull. Post-surgery, the mice were singly housed, left for one week to recover and then habituated to overnight fasts for 2 weeks. Following an overnight fast, mice were injected with 0.5 µl of artificial cerebrospinal fluid (aCSF) or leucine (i.c.v. 2.2 µg) and food intake monitored over a 24 h period. All injections were performed with a 31-gauge stainless steel injector which projected 0.5 mm below the tip of the cannula. Following the infusion, the injector was left in place for an additional 30 s to allow the drugs to diffuse away from the cannula tip.
Treatments, with either vehicle or drug, were crossed-over following a 1 week wash-out period. Correct cannula placement was confirmed at the end of the study by an increase in food intake after i.c.v. administration of NPY (1 µg). Mice were singly housed and habituated in CLAMS cages (Columbus Instruments) for 1-2 days prior to assessment of locomotion and energy expenditure over the subsequent 2 days.

Automated food intake monitoring
An episodic food intake monitoring apparatus (BioDAQ, Research Diets, Inc.) was used to assess feeding patterns in singly housed mice on normal chow. Food intake was measured and averaged from 5 consecutive days to obtain 24 h food intake kinetics. The BioDAQ system weighs the hopper with food (± 10 mg) every second and uses an algorithm to determine feeding bouts (changes in stable weight before and after a bout). Meals consist of one or more bouts separated by an inter-meal interval of 300 s with a minimum meal size of 20 mg.

Quantitative RT-PCR analysis
Tissues were lysed and homogenized in TRIzol reagent (Ambion) and total RNA was isolated using the RNeasy mini kit (Qiagen). First-strand cDNA was generated using Taqman reverse transcription reagents (Applied Biosystems) and qPCR was performed using Taqman universal PCR mastermix in a 7900HT real-time PCR system (Applied Biosystems). mRNA quantities were normalized to Hprt, Ppia or Gapdh after determination by the comparative Ct method. Primers used were: Abcc8 (Mm00803450), Agrp .

Hypothalamic immunohistochemistry
Immunohistochemistry was performed as previously described Choudhury et al., 2005). Fasted mice were injected with leptin (i.p. 5 mg/kg) and perfused with paraformaldehyde (4% w/v). Arcuate sections were incubated with rabbit anti-POMC precursor (1:1000; Phoenix Pharmaceuticals Inc.) and detection performed using a secondary antibody coupled to Alexa-Fluor-488. After extensive washing, slices were incubated with a rabbit anti-pSTAT3 (Tyr705) antibody (1.1000, Cell Signaling) which was detected using an ABC detection kit (Vector labs). Fluorescent images were taken with an epifluorescence microscope fitted with a color digital camera. Mice expressing GFP (POMC-GFP mouse) or YFP (AgRPCre-YFP) were fixed and arcuate sections incubated with a rabbit p70 S6 kinase (49D7, Cell Signaling) primary antibody (1:200) followed by a secondary antibody coupled to Alexa-Fluor-594. POMC neuronal measurements were counted from labeled cells using ImageJ software.