|Home | About | Journals | Submit | Contact Us | Français|
Obesity is a major risk factor in the development of insulin resistance and Type 2 diabetes. Under lean conditions, the adipocyte-derived hormone leptin maintains energy balance by acting on hypothalamic leptin receptors (LRbs) that trigger activation of the JAK2/STAT3 pathway 1–4. Although disruption of LRb-STAT3 signaling promotes obesity in mice, other neuroendocrine features of LRb function such as fertility appear normal, pointing to a requirement for additional regulators in this setting. Here we show that the cAMP and calcium-responsive CREB coactivator TORC1 is required for energy balance and reproduction; TORC1 −/− mice are hyperphagic, obese, and infertile. Indeed, TORC1−/− females are anovulatory, and they have abnormal uterine morphology along with low circulating concentrations of pituitary luteinizing hormone. Hypothalamic TORC1 was highly phosphorylated and inactive in leptin deficient ob/ob mice; and administration of leptin increased amounts of dephosphorylated, nuclear TORC1. Dephosphorylated, active TORC1, in turn, stimulated the expression of CART and KISS1 genes, which encode hypothalamic neuropeptides that mediate leptin effects on satiety and fertility, respectively 5–7. TORC1 over-expression in cultured hypothalamic cells increased CART and KISS1 gene expression, while depletion of TORC1, by RNAi mediated knockdown in vitro or by targeted gene disruption in vivo, decreased it. Leptin potentiated effects of cAMP and calcium activators on TORC1 transcriptional activity over the CART and KISS1 promoters in cells over-expressing LRb; these effects were disrupted by expression of the dominant negative CREB polypeptide A-CREB. As leptin administration also increased recruitment of hypothalamic TORC1 to CART and KISS1 promoters in vivo, our results indicate that the CREB:TORC1 pathway mediates central effects of hormone and nutrient signals on energy balance and fertility.
Sequestered in the cytoplasm through phosphorylation dependent interactions with 14-3-3 proteins, TORCs (TORC1, TORC2, TORC3) shuttle to the nucleus following their dephosphorylation in response to cAMP and calcium signals, where they potentiate cellular gene expression by binding to CREB over relevant genes 8–12. TORC1 contains conserved phosphorylation (Ser151) and ubiquitination (Lys575) sites that have been shown to modulate nuclear shuttling and protein stability in TORC2 13–15 (fig. 1a). By contrast with the ubiquitous pattern of TORC2 expression, however, TORC1 mRNA and protein are detected in the brain but not in other peripheral tissues such as liver, muscle, or adipose 16 (fig. 1a, sup. fig. 1).
Under basal conditions, TORC1 is highly phosphorylated in cultured hypothalamic GT1-7 cells (fig. 1b) 17. Exposure to cAMP or calcium activator triggers TORC1 dephosphorylation and nuclear translocation; phosphorylation-defective S151A mutant TORC1 is constitutively nuclear (sup. fig. 1). Over-expression of wild-type TORC1 potentiates CRE-luciferase reporter activity in cells exposed to the cAMP activator forskolin (FSK) or calcium ionophore (A23187); co-treatment with FSK and A23187 increases reporter activity synergistically (fig. 1b). In line with its constitutive nuclear localization, phosphorylation-defective S151A TORC1 strongly upregulates CRE-luciferase activity even under basal conditions. The effects of TORC1 appear CREB dependent because co-expression of a dominant negative CREB polypeptide, called A-CREB 18, disrupts reporter activity in cells exposed to FSK or A23187.
We evaluated the biological role of TORC1 in maintaining energy balance by insertional mutagenesis of the TORC1 gene with a promoter-less β-galactosidase (β-Geo) gene cassette (fig. 1c). Relative to control littermates, TORC1 mRNA and protein were undetectable in TORC1 −/− mice. Consistent with its regulation by the TORC1 promoter, CNS expression of the β-Geo cassette in TORC1 mutants mirrored that of endogenous TORC1 protein (fig. 1d, sup. fig. 2). In addition to other brain regions, TORC1 expression was prominent in arcuate and ventromedial nuclei of the hypothalamus. Arguing against significant effects of TORC1 on brain development, however, Nissl-stained sections from TORC1 −/− brains appear comparable to wild-type (not shown).
TORC1 −/− mice were born at the expected Mendelian frequency, and they were indistinguishable from wild-type controls prior to weaning. Although their linear growth was unimpaired, adult TORC1 −/− mice were infertile; no offspring were obtained from either TORC1 −/− males or TORC1 −/− females mated with wild-type mice (0/6). Anatomically, TORC1 −/− female uteri appeared threadlike in appearance with noticeable thinning of the endometrium (fig. 1e). Although they had comparable numbers of mature follicles, TORC1 −/− ovaries contained no corpora lutea, markers of ovulation. Indeed, circulating concentrations of pituitary luteinizing hormone (LH), a key regulator of ovulation, were down-regulated in TORC1 mutants compared to controls (fig. 1e).
In parallel with these reproductive defects, male and female TORC1 −/−mice also developed persistent obesity beginning at 9 weeks of age on a normal chow diet (fig. 1f); TORC1 +/− heterozygotes had intermediate weights relative to wild-type and TORC1 −/− homozygotes. White adipose mass was increased 2–3 fold in TORC1 mutant mice, whereas other tissues were relatively unaffected (fig. 1f: sup. fig. 3). Taken together, these results indicate that the effects of TORC1 on body weight are specific to white adipose, affect both males and females, and vary with gene dosage.
We performed metabolic studies to determine why TORC1 mutant mice gain more weight. Compared with wild-type littermates, TORC1 −/− animals ate more and they expended less energy at 12 to 14 weeks of age, as determined by physical activity and oxygen consumption monitoring (fig. 2a). Consistent with this disruption, TORC1 −/− mice were hyperglycemic and hypertryglyceridemic at 9 months of age (fig. 2b). Pointing to the development of insulin resistance, circulating levels of insulin were increased in TORC1 +/− mice and to a greater extent in TORC1 −/− homozygotes; they were glucose intolerant by IP glucose-tolerance testing (fig. 2b, c). In line with their obesity, circulating leptin concentrations were also upregulated in TORC1 −/− mice (fig. 2b).
During feeding, increases in circulating concentrations of leptin as well as insulin and glucose promote satiety and fertility, in part through the activation of arcuate neurons in the hypothalamus 4,19–21. Realizing that TORC1 −/− mice are hyperphagic, obese, and infertile, we wondered whether TORC1 is required for the activation of relevant hypothalamic programs in response to feeding signals. Although chronic leptin infusion substantially reduced food intake and body weight in control animals, it had minimal effects on TORC1−/− mice (fig. 2d) 22. Arguing against potential effects on leptin bioavailability, chronic leptin infusion promoted STAT3 phosphorylation a comparable extent in arcuate neurons of wild-type and TORC1 −/− animals (fig. 2e).
Consistent with the ability for leptin to increase hypothalamic STAT3 activity, mRNA amounts for proopiomelanocortin (POMC), neuropeptide Y (NPY), and Agouti Related Peptide (AgRP), regulatory targets of the LRb-STAT3 pathway that encode anorexigenic (POMC) and orexigenic (NPY, AgRP) neuropeptides, were comparable between TORC1 mutants and controls (sup. fig. 4). Indeed, signaling through the downstream melanocortin pathway also appear normal in TORC1 mutants, because intra-peritoneal (IP) administration of the alpha melanocyte stimulating hormone (α-MSH) analog MTII 23 inhibited food intake to the same extent in both wild-type and TORC1 −/− mice (sup fig. 5).
We used leptin deficient ob/ob mice to determine whether TORC1 activity is disrupted in obesity. Supporting this idea, ob/ob mice had increased amounts of phosphorylated, inactive TORC1 in the hypothalamus (fig. 2f). IP leptin injection increased amounts of dephosphorylated, nuclear TORC1 protein in arcuate cells of ob/ob mice (fig. 2g). Consistent with a parallel role for nutrient signaling, IP glucose administration also promoted the accumulation of dephosphorylated TORC1 in the hypothalamus (sup fig. 6). Correspondingly, TORC1 was nuclear-localized in arcuate cells during ad libitum feeding but remained cytoplasmic in other regions of the CNS (sup fig. 6). Taken together, these results indicate that hormone and nutrient signals modulate hypothalamic TORC1 activity under lean conditions, and that TORC1 activity is disrupted in obesity.
We performed gene profiling studies to identify hypothalamic genes that contribute to the metabolic and reproductive phenotypes of TORC1 mutant mice. This analysis revealed that mRNAs for the neuropeptide genes Cocaine and Amphetamine Regulated Transcript (CART) and KISS1 were down-regulated in TORC1 −/− animals. CART and KISS1 have been found to mediate effects of LRb signaling on feeding and fertility 6,7,24–27. Indeed, CART is co-expressed with POMC in arcuate neurons, where it inhibits food intake in response to leptin 27, while KISS1 expression in the arcuate promotes reproductive function by stimulating the secretion of hypothalamic gonadotropin releasing hormone (GnRH) 28,29. Similar to TORC1 −/− animals, mice with a knockout of KISS1 have low circulating concentrations of LH, exhibit abnormal uterine morphology, and are infertile 30. We confirmed that CART and KISS1 genes are down-regulated in TORC1 −/− mice by Q-PCR and in situ hybridization analysis (fig. 3a,b). As well, hypothalamic staining for kisspeptin, a cleavage product of the KISS1 precursor, was dramatically reduced in arcuate neurons of TORC1 −/− mice (fig. 3b). Importantly, TORC1 driven β-gal mRNA was co-expressed with CART and KISS1 neuropeptides in arcuate cells by dual immunohistochemistry and in situ hybridization (fig. 3c).
Realizing that CART and KISS1 promoters contain CREB binding sites (TGACG/CGTCA) that are conserved between mouse, rat, and human homologs, we considered that TORC1 may regulate both genes via a direct mechanism. Supporting this idea, CREB has been shown to promote CART gene expression in response to cAMP 31–33, although a similar role for KISS1 regulation has not been established. In keeping with its effects on TORC1 dephosphorylation, A23187 treatment increased endogenous mRNA amounts for CART and KISS1 in GT1-7 cells; this induction was blocked in cells depleted of TORC1 by RNAi-mediated knockdown (fig. 3d). Exposure to A23187 or FSK also increased CART and KISS1 reporter activities in transient assays (fig. 4a,b); over-expression of wild-type TORC1, and to a greater extent phosphorylation-defective (S151A) TORC1, enhanced transcription from both promoters. Consistent with the role of CREB in promoting TORC1 recruitment, expression of dominant negative ACREB inhibitor blocked induction of CART and KISS1 reporters by FSK and A23187 (fig. 4a,b).
We performed chromatin immunoprecipitation assays (ChIPs) to determine whether TORC1 and CREB regulate CART and KISS1 genes directly. In line with its constitutive nuclear localization, CREB occupied CART and KISS1 genes in GT1-7 cells comparably under basal conditions and following exposure to FSK or A23187 (fig. 4c). TORC1 occupancy over the CART and KISS1 genes was low under basal conditions - when TORC1 is sequestered in the cytoplasm -and increased following exposure to cAMP or calcium activator - when dephosphorylated TORC1 shuttles to the nucleus and binds to CREB. Consistent with its effect on amounts of nuclear TORC1 protein in the hypothalamus, leptin administration IP also increased TORC1 recruitment to CART and KISS1 promoters in ob/ob mice, while CREB occupancy over both genes was constitutive (fig. 4d). Taken together, these results indicate that CREB and TORC1 regulate hypothalamic CART and KISS1 gene expression through a direct mechanism.
Based on their importance for transcriptional induction in response to cAMP and calcium, we wondered whether TORC1 and CREB are also required for effects of leptin on neuropeptide gene expression. Exposure to leptin increased CART and KISS1 reporter activities synergistically with FSK in cells co-transfected with a leptin receptor (LRb) expression vector; these effects were augmented by over-expression of TORC1 (fig. 4e, f). Similar to its effects on cAMP and calcium signaling, ACREB inhibitor blocked induction of both promoters in cells treated with leptin.
Our results indicate that TORC1 is activated by hormonal and nutrient signals in the hypothalamus, where it promotes energy balance and fertility by enhancing CREB activity over relevant neuropeptide genes. Similar to leptin deficent ob/ob mice, TORC1 −/− females have abnormal uterine morphology and low circulating LH levels 34,35. By contrast with ob/ob animals, however, TORC1 mutant mice are only moderately obese, potentially reflecting compensatory effects of other TORC family members. Consistent with this idea, TORC2 is also expressed in the hypothalamus where it undergoes nuclear shuttling in response to feeding stimuli 36.
In addition to its effects on JAK2/STAT3 signaling, leptin has also been reported to modulate cation channel activity 37,38 and to inhibit the activity of the energy sensing Ser/Thr kinase AMPK 21. Based on the ability for calcium and AMPK pathways to regulate TORC1 activity, we imagine that these pathways may also mediate effects of leptin on TORC1 in the hypothalamus.
The importance of TORC1 in energy balance appears to be evolutionarily conserved; Drosophila TORC, the single fly homolog of mammalian TORCs, is also expressed primarily in the brain where it regulates energy consumption as well as glucose and lipid homeostasis 39. Drosophila TORC and mammalian TORC1 are regulated through phosphorylation by Salt Inducible Kinases (SIKs) and other members of the AMPK family 10,15,39. Indeed, knockdown of Drosophila SIK2 in neurons promotes starvation resistance and improves energy balance, suggesting that this Ser/Thr kinase also contributes to effects of hormonal and nutrient signals on hypothalamic TORC1 activity.
Obesity risk in humans has a strong genetic component, which is thought to involve heterozygous loss-of-function mutations in genes that, individually, may display only modest phenotypic changes 40,41. The presence of hyperphagia, increased adiposity, and insulin resistance even in heterozygous TORC1 +/− mice suggests that mutations in the TORC1 gene may also promote the development of obesity in humans. Future epidemiological studies of TORC1 gene mutations in affected populations should provide further insight in this regard.
All animals were housed in a temperature-controlled environment under a 12h light-dark cycle with free access to water and a standard rodent chow diet (Lab Diet 5001), unless otherwise specified. All animal studies were approved by the Salk Institute Institutional Animal Care and Use Committee. TORC1 mutant mice were generated by insertional mutagenesis. Mouse embryonic stem (ES) cells containing an insertional gene-trap in the CRTC1 locus (XK522; 129/Ola mouse strain) were obtained from BayGenomics 42,43. ES cells were injected into C57BL/6 blastocysts to generate chimeric mice, which were backcrossed with C57BL/6 mice (Harlan). Prior to intercrossing, heterozygous mice were backcrossed with C57BL/6 mice for three successive generations. The heterozygous progeny were intercrossed to obtain homozygous, heterozygous and wild-type littermate animals.
Male C57BL/6J, BKS.Cg-m +/+ Leprdb/J, and B6.V LepOb mice were obtained from Jackson Laboratories. All mice were allowed to adapt to their environment for at least 1 week prior to use in studies.
Genomic DNA from tail biopsies was prepared as described previously 44. Insertion of the pGT0lxf cassette was verified by sequence analysis of PCR-amplified genomic fragments. The following primers were used: A, 5′-GCATCCCTAGCTCTCACTCAGTTAC-3′; and B, 5′-GCGCGTACATCGGGCAAATAA-3′. Genotyping of TORC1 mutant mice was determined by PCR for the wild-type allele and the mutant allele. The wild-type allele was amplified using primers A and C (5′-ATTCCTCATATACCTCTCTTCTGGTGC-3′). The mutant allele was amplified using primers A and D (5′-GCATGAATCAACTTTGGAGACATGCG-3′).
Mice were individually housed for at least 3 days prior to the measurement of food intake or calorimetry. Daily food intake of ad libitum mice was determined by measuring the weight of food pellets on consecutive days. Locomotor activity, oxygen consumption and carbon dioxide production were simultaneously measured with a Comprehensive Lab Animal Monitoring System (Columbus), as previously described 44. For nocturnal food intake studies, food was removed from individually housed mice 1h prior to the onset of the dark cycle. Saline, leptin (2μg/g), and MTII (2μg/g) were administered by intraperitoneal injection 30 min prior to the onset of the dark cycle. Food was replaced at the onset of the dark cycle and food intake was measured after 90 min.
For whole blood and plasma measurements, mouse blood was collected from the tail vein into EDTA-coated capillary tubes (StatSpin). Blood glucose and triglyceride levels were measured with a OneTouch Ultra glucometer (LifeScan) and a CardioChek PA analyzer, respectively. Plasma concentrations of insulin (Mercodia) and leptin (Alpco) were measured by standard immunoassay methods according to the manufacturer’s protocol. For serum measurements, blood was collected by cardiac puncture from anesthetized (350 mg/kg chloral hydrate, i.p.) mice. Serum levels of luteinizing hormone (LH) and follicle-stimulating hormone were determined at the University of Virginia Center for Research in Reproduction Ligand Assay and Analysis Core.
Mice were fasted overnight for 16h and then injected with glucose (2mg/g; i.p.). Blood glucose levels were determined prior to the test and at 15, 30, 60, and 120 min following the injection.
Osmotic minipumps (0.25μL/h, Alzet) were aseptically filled with 0.45 μm-filtered sterile phosphate-buffered saline (PBS) or leptin (1.2mg/mL). Minipumps were primed by incubation at 37°C in sterile 0.9% saline for 8 to 16h. Male 7–8 week old TORC1 −/− mice and wild-type littermates were anesthetized under isoflurane and the minipump was implanted under the dorsal skin between the scapulae. Body weight and food intake were measured daily for 10 days, starting after the implantation of the minipumps. After 10 days of infusion, mice were anesthetized and transcardially perfused for brain immunohistochemistry, as described below.
To optimize cellular labeling for CART and KISS1, mice were injected ICV with 5μL of sterile colchicine (2mg/mL), as described previously 45. Mice were anesthetized with ketamine/xylazine and mounted in a stereotaxic frame. A 33 guage cannula was placed into the lateral cerebral ventricle using the coordinates: +0.3mm anterior-posterior to bregma, +0.8mm lateral to the midline, −2.5mm dorsal-ventral. Colchicine was injected over a 15min period through the cannula using a hamilton syringe. Following injection the cannula was removed and the surgical incision was closed with wound clips. Mice were anesthetized and transcardially perfused for brain immunohistochemistry 36h after colchicine treatment.
Expression plasmid for the leptin receptor (LRb) was a gift from Dr. Martin G. Myers Jr. The luciferase reporter plasmid containing sequence from the mouse CART promoter spanning −641 to +30 was kindly provided by Dr. Michael J. Kuhar 31. The KISS1 luciferase reporter was generated by cloning the human KISS1 promoter spanning −1132 to +1, into pGL2 (Promega). The EVX-1 luciferase reporter and TORC1 expression constructs have been described previously 8. The S151A TORC1 expression construct was generated by site-directed mutagenesis. Lentiviruses encoding U6 promoter-driven interfering RNAs directed against the TORC1 sequence, 5′-GGTCCCTGCCCAACGTGAAC-3′, were generated as described previously 12. Forskolin (Sigma), A23187 (Calbiochem), leptin (R and D Systems), and MTII (Bachem) were purchased from the respective manufacturers.
Mice were anesthetized (350 mg/kg chloral hydrate, i.p.) and transcardially perfused with 4% paraformaldehyde in 0.1M sodium borate buffer at 4°C, as described previously46. Brains were post-fixed for 2h at 4°C and then incubated in 0.05 M potassium-PBS (K-PBS) containing 15% (w/v) sucrose 4°C for 12–16 hr. Brains were sectioned on a sliding microtome (25 μM to 30 μM), collected in equally-spaced series and stored in cryoprotectant (20% glycerol and 30% ethylene glycol in 0.1 M phosphate buffer) at −20°C. Immunohistochemistry was conducted on free-floating sections by the avidin-biotin-complex method using the chromogen, diamino-benzidine (Vector Labs). For P-STAT3 (Y705) immunohistochemistry (IHC), sections were pre-treated with 1% H2O2 in 1% NaOH for 10 min, 0.3% glycine in K-PBS for 10 min, and 0.03% SDS in K-PBS for 10 min 47. For TORC1, CART, and KISS1 IHC, sections were pre-treated with 0.3% H2O2 for 10 min and 0.3% glycine, 0.3% Triton X-100 in K-PBS for 10 min. Brain sections were blocked in a K-PBS solution containing 1% Probumin (Millipore), 1% normal donkey serum (Jackson ImmunoResearch) and 0.03% Triton-X-100. Sections were incubated overnight at 4°C with primary antibody diluted in blocking solution. Tissue sections were washed and then incubated with Biotin-SP-conjugated Donkey Anti-Rabbit IgG (1:500, Jackson ImmunoResearch) in blocking solution for 1 hr. Sections were washed and incubated with Vectastain ABC (Vector Labs) for 1 hr. Staining was developed using a Nickel-enhanced diamino-benzidine reaction (Vector Labs). Brain sections were mounted on gelatin-subbed slides, dried, dehydrated and mounted with DPX (Electron Microscopy), unless the sections were also being processed for in situ hybridization. In this case, brain sections were mounted on SuperFrost Plus slides (Brain Research Laboratories) and processed for in situ hybridization, as described below. The primary antibodies used were P-STAT3 Y705 (1:1000, #9145, Cell Signaling Technologies), KISS1 (1:1000, #9754, Chemicon), CART (1:10000, #6838, PBL-Salk Institute), and TORC1 (1:2500, #6938, PBL-Salk Institute). For TORC1 IHC and IF the antibody was pre-adsorbed against the immunogen carrier β-thryoglobulin and 0.5 mM TORC2 peptide (MESPSTSL), which was synthesized at the Salk Institute Peptide Synthesis Core Facility.
Immunofluorescence was conducted on brain sections or cells fixed with 4% paraformaldehyde in phosphate-buffered saline. Tissue sections or cells were processed as described for immunohistochemistry except the samples were not incubated in hydrogen peroxide and the antigen was visualized using a rhodamine red x-conjugated donkey anti-rabbit antibody (Jackson Immunoresearch).
Methods for probe synthesis, hybridization and autoradiography as described previously 46. In situ hybridization was conducted using 35S-labelled cRNA probes generated from a rat melanin concentrating hormone (MCH) cDNA 48, a partial β-galactosidase (β-Gal) cDNA (681bp), and full-length mouse CART cDNA. The partial β-Gal cDNA was PCR amplified and cloned from the gene-trap cassette isolated from TORC1 −/− mice. The full-length mouse CART cDNA was cloned from hypothalamic mouse RNA. Brain sections were mounted onto poly-L-lysine coated slides, dried under vacuum, and postfixed with 4% paraformaldehyde for 30 min at 4°C. Tissue sections were digested with proteinase K (10μg/mL) for 30 min at 37°C, acetylated and dehydrated prior to hybridization. Labeled probes (1–3×109 dpm/μg) were diluted in hybridization solution, applied to sections and allowed to hybridize overnight at 60°C. Sections were subsequently treated with ribonuclease A (20μg/ml) for 30 min at 37°C, washed with 0.1× SSC buffer (65–85°C), dehydrated and exposed to x-ray films (β-max; Eastman Kodak) for 1–2 d. Slides were then coated with Kodak NTB-2 liquid emulsion, and exposed at 4°C for 3–4 weeks.
The mouse hypothalamic cell line GT1-7 was a gift from Dr. Pam Mellon (UCSD). GT1-7 cells, 293T cells and GH3 cells were cultured in DMEM (Mediatech) containing 10% fetal bovine serum (HyClone), 100 μg/ml penicillin-streptomycin and 1mM pyruvate. Transient transfections of GT1-7 cells were performed using Fugene HD (Roche). HEK293T cells and GH3 cells were transfected by using Lipofectamine 2000 (Invitrogen). For luciferase assays, cells were seeded in 24-well plates and transfected 16h later. Cells were treated for 4h, as indicated, 16 to 48h post-transfection. In reporter assays employing LRb, cells were transfected and treated under serum-free conditions. Cell lysates were prepared and the activities of luciferase and β-galactosidase were determined, as described previously 8.
Hypothalamic tissue was dissected as described previously 46 with a slight modification. Brains were rapidly removed from anesthetized animals. The brain was sliced coronally with a razor blade to generate a block spanning from the optic chiasm to the rostral end of the mammillary bodies. For RNA isolation or chIP assays the brain slice was immersed in RNA later or ice-cold PBS, respectively. The hypothalamic area was dissected by making two cuts at either end of the optic chiasm and a cut above the third ventricle. These cuts also generated two amygdala-enriched regions, lateral to the hypothalamus, and a cortex-enriched region dorsal to the external capsule. The dissected brain tissue was either snap-frozen for protein extraction or immersed in RNAlater for RNA extractions.
RNA was isolated from cells using an RNeasy kit (Qiagen) according to the manufacturer’s protocol. For tissue, animals were anesthetized (350 μg/kg chloral hydrate, i.p.) and the relevant tissue or brain region was rapidly dissected out and snap-frozen in liquid N2 or immersed in RNA later for 12 to 16h at 4°C. Total RNA was isolated from the tissue using Trizol reagent (Invitrogen) according to the manufacturer’s protocol. Total RNA was then treated with DNase (RNase-Free; Roche) and post-cleaned using an RNeasy kit (Qiagen). 1 μg of total RNA was reverse-transcribed to cDNA using Superscript III (Invitrogen) and random hexamers. cDNAs from samples were amplified and detected using SYBR Green I reagent (Roche) and a LightCycler 480 Instrument (Roche), respectively. Quantification of mRNA levels was performed using the LightCycler 480 Software (Roche) using standard curves and normalizing to β-Actin or L32 expression.
Male 12-week old wild-type and TORC1 −/− mice (3 of each genotype) were fasted overnight for 15 to 16h and fed a normal chow diet for 6h. Total RNA was isolated from hypothalamic tissue, as described above. Gene-profiling experiments were conducted with an Affymetrix Mouse Genome 430 array and analyzed, as described previously 49.
GT1-7 cells or GH3 cells were plated in 15-cm plates, grown to 80–90% confluency and treated as indicated. For animal experiments, hypothalamic tissue was dissected from mice after the indicated manipulations. Cells or hypothalamic tissue were crosslinked with 1% v/v formaldehyde for 10 min at 4°C, and quenched with 0.125M glycine for 5 min. Chromatin was prepared and immunoprecipitations performed, as described previously 49. Immunoprecipitations were conducted using antibodies for CREB (#244, PBL-Salk Institute) and TORC1 (#6937E1, PBL-Salk Institute). Rabbit IgG (Santa Cruz) was also utilized as a negative control. Following crosslink reversal, and proteinase K treatment, DNA was isolated by two rounds of phenol-chloroform extractions and ethanol precipitation. DNA was treated with RNase (DNase-free; Roche) and purified with a QIAquick PCR purification kit (Qiagen). Occupancy on target promoters was determined by PCR or quantitative real-time PCR.
Cells or tissue were snap-frozen in liquid N2 at the end of the experiment. Cells and tissue were lysed in a modified RIPA buffer containing 50 mM Tris pH 7.5 at 4°C, 150 mM NaCl, 1 mM EDTA, 50 mM NaF, 5 mM sodium pyrophosphate, 10 mM β-glycerophosphate, 1% NP-40, 1 mM Na3VO4, 0.25% sodium deoxycholate, 0.1% SDS and protease inhibitors (Sigma). The lysate was cleared by centrifugation and the protein concentration of the cleared lysate was determined by the method of Bradford. Western blots for TORC1, TORC2, CREB and P-CREB Ser133 were performed using antibodies 6939E1, 6865, 244, 5322, respectively (Salk Institute). Antibodies for P-(Ser151) TORC1 and HSP90 were obtained from Cell Signaling and Santa Cruz, respectively.
P-STAT3 staining and nuclear and cytoplasmic TORC1 staining in hypothalamic sections were quantified by an independent observer in a blinded manner. Five to eight matched arcuate-containing sections from whole-brain series were counted from independent experiments.
Data presented are means ±SEM. Statistical analyses were performed using SigmaStat (Systat). Statistical differences for one factor between two groups or more than two groups were determined using an unpaired Student’s t-test or an analysis of variance (ANOVA) with a post-hoc test, respectively. Statistical differences for two factors between more than two groups were determined using a two-way ANOVA with a post-hoc test. Values of P<0.05 were considered statistically significant.