| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Genetics and Pediatrics (C.B.K., A.W.X., G.S.B.), Stanford University School of Medicine, Stanford, California 94305; Department of Psychiatry and Behavioral Neurosciences (L.G., F.Y., K.H., R.G.M.), Wayne State University School of Medicine, Detroit, Michigan 48201; and Department of Medicine (G.J.M., M.W.S.), Harborview Medical Center, University of Washington, Seattle, Washington 98104
Address all correspondence and requests for reprints to: Gregory S. Barsh, Departments of Genetics and Pediatrics, Stanford University School of Medicine, Stanford, California 94305. E-mail: gbarsh{at}stanford.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Agrp is normally expressed in one of two neuronal subpopulations in the arcuate nucleus of the hypothalamus that are thought to comprise opposing circuits involved in energy homeostasis, and to directly sense bloodborne nutrients and hormones. The other neuronal subpopulation is marked by expression of pro-opiomelanocortin (Pomc), which gives rise to the anorexigenic hormone 
-MSH and directly opposes the action of Agrp at central melanocortin receptors (reviewed in Refs. 7 and 8). Like Agrp, transcription of Pomc is also modulated by leptin, but in an opposite direction; leptin-deficient animals exhibit decreased levels of Pomc mRNA and increased levels of Agrp mRNA. However, whether provoked by food deprivation or leptin deficiency, the magnitude of the increase in Agrp mRNA tends to be greater, typically 8- to 10-fold, than the magnitude of the decrease in Pomc mRNA, typically 1.5- to 2-fold (3, 5, 9, 10) (reviewed in Ref. 11).
One of the most well-characterized signaling pathways downstream of the leptin receptor involves the signal transducer and activator of transcription 3 (Stat3), which belongs to a family of intracellular signaling molecules that interact directly with cytokine receptors on the cell surface and are phosphorylated as a consequence of ligand binding. Phosphorylation promotes the dimerization and nuclear translocation of Stat proteins, where they directly bind DNA to modulate the transcription of target genes (reviewed in Refs. 12, 13, 14).
Direct evidence for the role of phosphorylated Stat3 (pStat3) in central energy homeostasis comes from studies involving the targeted deletion of Stat3 in mice. Although Stat3 knockout mice die during early embryogenesis (15), mice with a broad neural deletion of Stat3 [generated with a NestinCre transgene (16)] or a partial deletion of Stat3 within the hypothalamus [generated with a RIPCre transgene (17)] develop obesity and have other phenotypes characteristic of Lepob/ob or Leprdb/db. Although demonstrating a critical role for Stat3 in central energy homeostasis, these experiments do not address which neural subtypes require Stat3 or the precise function of Stat3 in those neurons.
Several observations suggest that leptin directly regulates Agrp transcription via Stat3. Many Agrp neurons also express the leptin receptor (18), leptin administration causes a rapid accumulation of pStat3 in Agrp neurons (19), mice with mutated leptin receptors that cannot bind Stat3 are obese with elevated Agrp mRNA (20), and consensus sites for pStat binding exist in the 5' flanking region of Agrp (21). In most contexts, Stat3 (and other Stat proteins) function as transcriptional activators. However, the inverse relationship between leptin receptor occupancy and Agrp mRNA levels implies that pStat3, if it does regulate Agrp in vivo, should act as a transcriptional repressor rather than a transcriptional activator. Indeed, an oft-cited conceptual framework to account for the differential action of leptin on expression of Agrp and Pomc mRNA levels posits that leptin directly activates Stat3 signaling in both neurons, with Stat3 stimulating transcription of Pomc but inhibiting transcription of Agrp (reviewed in Ref. 12). The same model applies to neuropeptide Y (Npy), a potent orexigen that is coexpressed with Agrp in the arcuate nucleus of the hypothalamus (22). Nonetheless, leptin signaling affects other intracellular mediators in addition to Stat3 in these neurons, including phosphoinositide 3-kinase and the MAPK signaling cascades (23, 24, 25). Important questions remain regarding the extent to which each of these pathways contributes to the response of Agrp/Npy and Pomc neurons and the specific role of Stat3 signaling.
Understanding the precise mechanism by which leptin modulates Agrp transcription in vivo would have both basic and clinical implications. However, these mechanisms are challenging to study for several reasons. The arcuate nucleus of the hypothalamus is part of a circuit in which there exists the potential for cross talk and feedback, Agrp/Npy neurons represent a small fraction of the hypothalamus and cannot be distinguished anatomically, and there is no bona fide system for studying regulation of Agrp transcription in cultured cells.
We have previously identified a 40-kb region upstream of the Agrp locus that recapitulates both the cell type-specific and fasting-responsive nature of Agrp expression in transgenic mice (26). Here, we describe how an extension of this observation has been applied in two different ways to study the potential role of Stat3 in Agrp/Npy neurons. Using recently developed genomic tools and resources for comparative sequence analysis, we identify several noncoding motifs that have been under strong evolutionary constraint during the last 200 million years, including two putative Stat binding sites in the Agrp promoter. By studying an Agrp transgene engineered to prevent Stat binding, we find that these sites are not required for cell type-specific expression in the arcuate nucleus of the hypothalamus but, surprisingly, are required for fasting-induced increases in Agrp expression. To investigate whether this unexpected finding can be explained solely by loss of pStat3, we use a Cre transgene driven by Agrp regulatory elements to remove activatable Stat3 from Agrp/Npy neurons, and determine the consequences on regulation of Agrp and Npy mRNA. Our findings reveal a previously unappreciated complexity to the system in which the ability of Agrp mRNA to respond to altered metabolic state depends on the Stat3 binding sites of the Agrp promoter but not on the presence of activatable Stat3.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). Genome sequence from many additional vertebrates has recently become available, such that the strength of comparative analysis is limited only by phylogenetic scope. For example, brain administration of Agrp mimetics or antagonists perturbs food intake in fish (28) and in avians (29). In goldfish and zebrafish, Agrp is expressed specifically in the hypothalamus, and up-regulated in response to fasting (28, 30); however, in chickens, Agrp is expressed throughout the body and (at least in broiler chickens) fails to respond to leptin treatment (31, 32). Thus, Agrp early in vertebrate evolution probably exhibited the same function and pattern of expression as currently observed in rodents; however, certain aspects of transcriptional regulation may have been lost in some vertebrates. These observations should be reflected in the conservation pattern of cis-acting elements across different phyla.
|
), and cat, chimpanzee, cow, and tenrec. Conservation is also apparent for untranslated regions (exon 1 and part of exon 4) as well as a 760-bp 5' flanking area corresponding to the proximal conserved region (Fig. 1A). Similar comparisons of more distantly related species (including opossum, platypus, chicken, and pufferfish) did not identify any regions of high sequence similarity outside of Agrp exon 4, even when the parameters for defining conservation were relaxed (Fig. 1A). Thus, despite the functional similarities of Agrp among vertebrates, genomic sequences comprising and surrounding the Agrp transcriptional unit have diverged to the extent that regional conservation (i.e. >75% over a 100-bp window) between distantly related vertebrate species is no longer detectable.
Within the 760-bp proximal conserved region are two separate peaks (labeled A and B in Fig. 1), and within each of these, we noticed two putative Stat binding sites that are almost perfectly conserved (Fig. 1B; supplemental Fig. 1, published as supplemental data on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org). Identical sites are also found upstream of Agrp in opossum and platypus, which is especially notable given the overall lack of similarity for Agrp exons 1–3 between opossum or platypus and eutherian mammal sequence. Nonetheless, the two putative Stat sites in opossum and platypus sequence are found the same distance apart from each other, and the same distance from Agrp exon 4 as in other mammalian sequences. This level of evolutionary constraint is extremely high—greater, for example, than protein coding regions in exons 2 and 3—and suggests that the region is a target for one or more trans-acting factors that regulates Agrp expression. We also identified a potential Stat binding site in the chicken Agrp upstream sequences; however, its significance is less certain because only a single site has been identified (Fig. 1B).
An in Vivo Assay for the Evolutionarily Conserved Stat Binding Sites
We mutated each of the putative Stat sites in the context of a BAC reporter construct that efficiently drives arcuate specific and fasting-responsive lacZ expression in transgenic mice (26). Using homologous recombination in Escherichia coli, we altered the BAC construct such that one of the sites was changed from TTCCAGGAA to TTCCAGAGG, and the other from TTCCTGGAA to TTCCTCGAG (Fig. 1C); these alterations were chosen based on previous in vitro studies of Stat binding specificity and surveys of physiological Stat target sites (33, 34).
Multiple transgenic lines were established from independent founders injected with either the control BAC reporter construct (BAC172ßgeo) or the one carrying mutated Stat sites (BAC172ßgeoMUT). We observed reporter expression in the arcuate nucleus of both control and mutated transgenic lines (7 of 9 and 5 of 14, respectively), indicating that the putative Stat binding elements are not necessary to drive Agrp expression in the arcuate nucleus (Fig. 2, A and B).
|
). Because each line represents a different integration site, and because some integration sites are subject to position effects associated with transgene silencing, an accurate comparison of the control and mutated constructs required that multiple animals be examined from each of several different lines. Figure 2, D and E, depicts results for every animal we analyzed, with individual transgenic lines organized in columns. Although there were some animals in every transgenic line in which no staining was detectable, the overall level is clearly greater in lines carrying the control construct, with five of the nine control lines exhibiting mean levels of staining that were higher than each of the lines carrying the mutated construct, and an overall difference across all animals carrying the control vs. the mutated construct that is highly significant (P < 0.0001) (Fig. 2F).
Because the comparison between transgenic lines carrying the control or the mutated construct was carried out on animals that had been fasted for 48 h, and because transgenic lines generated with the control construct exhibit increased expression upon fasting that mimics normal Agrp regulation, we suspected that the putative Stat sites (or at least the five nucleotides that we altered) might be required for increased Agrp expression normally provoked by fasting. We tested this idea by comparing levels of X-gal staining between animals that were free-feeding or fasted for 48 h. For each of the five transgenic lines generated from the mutated construct that exhibited any staining at all (Fig. 2E), there was no difference between animals that were free-fed or those that were fasted (Fig. 2G). As an alternative approach, we also counted the number of X-gal-positive neurons on the surface of a coronal brain slice for one transgenic line that produced very consistent low-level X-gal staining, and observed no difference between slices from fasted vs. free-fed animals (Fig. 2H). In contrast, all six of the control transgenic lines tested showed increased levels of X-gal staining in fasted compared with free-fed animals (26) (data not shown). Thus, evolutionarily conserved Stat binding sites are required for fasting-induced up-regulation of Agrp, but the mutated BAC transgene behaves opposite to what would be predicted were pStat3 acting as a transcriptional repressor.
Agrp Neuron-Specific Disruption of Stat3
To investigate whether the unexpected finding described above can be explained by loss of activatable Stat3, we used the Cre-loxP system to conditionally disrupt the Stat3 gene specifically in Agrp neurons. Animals carrying a Cre transgene controlled by Agrp regulatory elements, Tg.AgrpCre, were crossed with animals carrying Stat3flox (35), a gene-targeted Stat3 allele with loxP sites that flank exon 22, which encodes a tyrosine residue whose phosphorylation is critical for Stat3 dimerization and nuclear translocation (Fig. 3A). Among backcross progeny, we generated both Tg.AgrpCre/+; Stat3flox/flox and Tg.AgrpCre/+; Stat3flox/– animals (hereafter referred to as Agrp-Stat3 mutant mice) and littermate controls,+/+; Stat3flox/flox, or +/+; Stat3flox/– (these genotypes did not exhibit any intergroup differences and are considered as a single control group in what follows).
|
, upper panel).
We then carried out similar double-labeling experiments in control and Agrp-Stat3 mutant mice to examine the pattern of total Stat3 immunostaining and Npy mRNA hybridization signal. In control mice, Stat3 immunostaining appears in both Npy-expressing neurons and a surrounding, mostly lateral, population of neurons in the arcuate nucleus (Fig. 3B, middle panel). In Agrp-Stat3 mutant mice, Stat3 immunostaining persisted in this latter population but was completely eliminated from Npy-expressing neurons (Fig. 3B, lower panel). Finally, we examined the pattern of hypothalamic pStat3 immunostaining 45 min after ip injection of leptin. Control animals exhibited robust immunostaining in response to leptin in several regions of the hypothalamus, similar to what has been reported previously by Munzberg et al. (36, 37) (Fig. 3C). In Agrp-Stat3 mutant mice, leptin-induced pStat3 immunostaining was unaffected in most regions of the hypothalamus but was eliminated from the medial region of the arcuate nucleus (Fig. 3C). Taken together, these observations indicate that mutant animals specifically lack the ability to activate Stat3 in Agrp neurons.
Agrp-Stat3 mutant mice exhibited a mild obesity phenotype, which will be presented in detail elsewhere (Gong, L., and R. G. MacKenzie, manuscript in preparation). With regard to the focus of the current paper—the effect of Stat3 on expression of orexigenic neuropeptides in Agrp/Npy neurons—we measured hypothalamic Agrp and Npy mRNA levels in control and Agrp-Stat3 mutant mice under both free-feeding conditions and after food deprivation. A 48-h fast caused a significant elevation of Agrp and Npy mRNA in both Agrp-Stat3 mutant mice and nontransgenic littermate control mice (Fig. 4A). Notably, Agrp mRNA levels in mutant mice were no different from those in control mice (Fig. 4, A and B). In one experiment, Npy mRNA levels during free feeding were slightly elevated in Agrp-Stat3 mutant mice, but this result was not confirmed in a subsequent experiment (Fig. 4B).
|
).
Interpretation of energy homeostasis circuitry based on neuropeptide mRNA levels can be complicated by secondary effects; for example, hyperleptinemia in diet-induced obesity is expected to cause decreased levels of Agrp mRNA and therefore counteract inputs that would otherwise result in more elevated Agrp mRNA levels. However, this explanation is unlikely to apply to the results depicted in Fig. 4, because the mild obesity that develops in Agrp-Stat3 mutant mice is not apparent until several months of age, and there is no difference between mutant and control Agrp mRNA levels in both old (Expt I, Fig. 4A) and young (Expt II, Fig. 4A) animals.
In summary, our findings indicate that transcriptional regulation of Agrp mRNA between the fed and fasted states cannot be explained by a high basal level in the fasted state with repression mediated by pStat3 in the fed state (Fig. 4D, left panel). Instead, our results suggest a new model (Fig. 4D, right panel), in which a low level of Agrp mRNA represents basal activity during the fed state, which is then activated by factors in addition to Stat3 during the fasted state.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Of course, Stat3 is only one of several downstream mediators of leptin signaling in the CNS; additional tyrosine phosphorylation sites in the leptin receptor activate the MAPK cascade, and leptin has a rapid effect on cell membrane potential (42) and phosphatidylinositol 3-kinase activity (43), events that are probably independent of Stat3 phosphorylation. Nonetheless, most attempts to study regulation of gene expression downstream of leptin signaling have focused on Stat3, in which relatively small (1–2 kb) regions of 5' flanking sequence from the Pomc (37), Npy (44, 45), or the TRH gene (46, 47) have been fused to a reporter gene and tested for enhancer activity in cultured cells.
The BAC-based lacZ reporter strategy applied here is similar in concept to these approaches but less efficient, because it depends on the construction and analysis of multiple transgenic lines for each data point. On the other hand, the BAC-based transgenic approach allows potential regulatory regions to be investigated in vivo, and in the context of a genomic DNA fragment sufficiently large to recapitulate the normal pattern of endogenous gene expression. This is especially important in the case of Agrp neurons, where Stat3 has been postulated to inhibit, rather than stimulate Agrp and Npy expression. For example, previous studies of Agrp transcription identified an approximately 500-bp minimal promoter that directs expression of a reporter gene in GT1–7 cells (which are derived from a gonadotrophic releasing hormone-expressing cell line), but deletion of the putative Stat sites had little effect on reporter gene expression (21). In the case of Npy, reporter gene studies identified leptin-responsive Stat binding sites in an approximately 500-bp promoter region, but the effect of leptin on reporter gene expression in PC12 or neuroblastoma cells was opposite to that which occurs in hypothalamic neurons in the arcuate nucleus (44, 45).
Within the approximately 42.5 kb of DNA upstream of Agrp required for normal reporter gene expression in the arcuate nucleus of the hypothalamus, our results demonstrate that five evolutionarily conserved nucleotides are required for transcriptional activation induced by fasting. Although the mutations we engineered were chosen because they disrupt Stat protein-DNA interactions, the underlying mechanism must involve an additional transcription factor or transcription factor complex, because removal of activatable Stat3 from Agrp neurons did not alter basal or fasting-induced Agrp mRNA levels (Fig. 4, A and B). Likely candidates for members of this complex in addition to Stat3 include members of the Foxo transcription factor family and/or the glucocorticoid receptor. Consensus binding sites for Foxo proteins lie 7–13 nucleotides away from each of the putative Stat sites in the Agrp promoter, and recent studies from Kitamura et al. (48) suggest that Foxo1 interacts with Stat3 to regulate Agrp expression. In addition, a previous study by Makimura et al. (49) has shown that transcriptional activation of Agrp induced by fasting depends on the presence of glucocorticoids. Thus, a plausible explanation for the observation that Stat binding elements but not activatable Stat3 are required for fasting-induced stimulation of Agrp mRNA posits that one or more proteins in addition to, or instead of, Stat3, form a transcriptional activation complex at the Stat binding site under conditions of energy deficit (Fig. 4D, right). Although molecular characteristics of this complex remain to be fully characterized, our results suggest that protein-DNA interactions at this site have been conserved for more than 200 million years of evolution.
Bates et al. (20) have shown that Agrp mRNA levels are elevated in mice carrying a leptin receptor mutation that specifically prevents leptin activation of Stat3 (Leprs/s). Apparent discordance between our results and those of Bates et al. (20) suggests that the elevated Agrp mRNA levels in Leprs/s mice are not due to impaired leptin-Stat3 signaling in Agrp neurons but, instead, are a consequence of impaired leptin-Stat3 signaling in other cell types. Nonetheless, our results do not preclude the possibility that pStat3 might function as a transcriptional repressor of Agrp under certain circumstances. For example, in the transition from a fasted to a fed state as leptin levels rise, the putative transcriptional activation complex referred to in Fig. 4D may dissociate from the Stat binding site at the same time that new pStat3 dimers enter the nucleus. Both events would likely contribute to decreased transcription of Agrp, and the loss of one or the other might affect the kinetics of the process, such that in the absence of pStat3, the rate of Agrp transcriptional initiation declines more slowly than normal. Thus, a snapshot of Agrp mRNA levels provided by quantitative RT-PCR in fed and fasted animals might not capture a change in total amount of Agrp mRNA integrated over many weeks. Such a role may be uncovered by evaluating neuropeptide gene expression in control and Agrp-Stat3 mutant mice after leptin injection or in more complex paradigms.
The Stat group of transcription factors were initially recognized as transcriptional activators whose nuclear migration and DNA binding is triggered by phosphorylation and dimerization, but it has gradually become apparent that there are a number of molecular variations on this theme (14), and the action of a particular Stat protein will depend on both the target gene and the cellular context. As shown here, the ability to test how specific modifications to DNA target sites or members of DNA binding protein complexes affect transcriptional regulation in vivo will be helpful in integrating molecular models for Stat action with their effects in whole animals.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
BAC Transgene Mutagenesis
Five base pairs located within two putative Stat binding sites upstream of Agrp on BAC172ßgeo (26) were mutated using homologous recombination in bacteria (53). BAC172ßgeoMUT was generated by a two-step recombination strategy. First, a 224-bp region containing the Stat binding sites was replaced with a tetracycline resistance (TetR) cassette. The TetR cassette was then replaced with a mutated version of the original 224-bp region through a second round of recombination and selection against the TetR marker. This strategy specifically alters 5 bp without the net insertion or deletion of any sequence.
The TetR cassette was amplified from pFRT-Tet-17 by PCR using primers HATetRF1 (AGCTGTATGCCCACCGGGGGAACAGTGTTTTCTGCTCCCTTGGTTTCCAGAGATCTATGATTCCCTTTGTCAACAG) and HATetRR1 (GTCCCCACAGATGTTCACTTAGTCTCCCTAGGACATGTTGTTTCCTTTGCAAGCTTATGATGATGATGTGCTTAAAAAC). Each primer has 50 nucleotides identical with the targeting sites of BAC172ßgeo, providing homology arms for recombination; the remaining 3' sequence serves as an annealing primer for amplification of the TetR cassette. Recombination between the amplified TetR fragment and BAC172ßgeo, resulting in BAC172ßgeoStatTet, was carried out as described previously (53), and correctly recombined BACs were verified by PCR using primers that flank the targeting site and thus generate a larger PCR product upon the insertion of the TetR cassette.
A 230-bp fragment of the Agrp promoter in which the two Stat sites were mutated as described in Results was constructed by PCR-directed mutagenesis using primers StatStuI (TTCCAGAGGCCTTAGGTAGAAAGGGGG) and StatXhoI (CTTTGCCTCGAGGAACTTTGTCAAGGGC); this fragment was cloned and then used as the template for PCR amplification with primers HAStatDelFor (AGCTGTATGCCCACCGGGGGAACAGTGTTTTCTGCTCCCTTGGTTTCCAGAGG) and HAStatDelRev (GTCCCCACAGATGTTCACTTAGTCTCCCTAGGACATGTTGTTTCCTTTGCCTCG). Recombination between the mutated targeting cassette and BAC172ßgeoStatTet gave rise to BAC172ßgeoMUT; correctly targeted BAC molecules were identified by restriction digestion after PCR amplification of the targeted region and direct sequencing of PCR products across the targeting site. We also compared the restriction digest patterns of BAC172ßgeo and BAC172ßgeoMUT to insure that no rearrangement of the BAC occurred during the recombineering process.
Mouse Genetics
We have previously described the construction and applications of an AgrpCre transgene engineered by BAC recombineering. For the work described here, a different AgrpCre transgene was constructed at Wayne State University (R. G. MacKenzie) using the same strategy, but with a cDNA of codon-improved Cre (iCre) recombinase (54). The Cre cDNA was inserted at the initiator ATG of the Agrp coding sequence located in exon 2. The modified BAC was then linearized and transgenic animals constructed by the University of Michigan Transgenic Core Facility. Eutopic expression of functional iCre was determined by immunostaining for Cre and expression of EGFP using the reporter strain Gt(ROSA)26Sortm2Sho (55). Crossing Tg.AgrpCre mice with mice in which exon 22 of the Stat3 gene was flanked by loxP sites (35) yielded Tg.AgrpCre/Stat3flox/flox and allowed for the deletion of activatable Stat3 specifically from Agrp neurons. Occasionally, spurious germline recombination in Tg.AgrpCre/Stat3flox/flox background resulted in offspring carrying a single recombined allele of the Stat3 gene (Stat3flox/–) and these animals were included as indicated (see Results). As described by Takeda et al. (35), small amounts of an abnormal Stat3 protein, Stat3
, are produced by the recombined Stat3 allele; Stat3 cannot be phosphorylated and therefore cannot be activated by Jak-Stat receptors. All animal work was carried out under Association for Assessment and Accreditation of Laboratory Animal Care-approved institutional guidelines.
Histology, in Situ Hybridization, and Immunohistochemistry
For studies of Agrp reporter gene expression using the modified BAC, BAC transgenic mice were either free-fed on standard mouse chow or fasted for approximately 36 h before reporter expression analysis (food was removed from cage setups between 1600 and 1800 h). After CO2 euthanasia, the brain was immediately dissected and 1-mm coronal brain slices were cut through the hypothalamic region using a mouse brain matrix (ASI instruments, Warren, MI). Brain slices were fixed in PBS containing 4% paraformaldehyde (PFA), washed three times for 30 min in PBS containing 0.1% Triton X-100, and incubated for 24 h at 37 C in X-gal staining solution (5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, 1 mg/ml X-gal, and 2 mM MgCl2 in PBS). Brain slices were viewed under a dissecting microscope, assigned a semiquantitative score from 0–5 (Fig. 4C), and captured with Axiocam imaging software.
For studies of combined mRNA in situ hybridization and immunohistochemistry, free-feeding Tg.AgrpCre/Stat3flox/flox and Stat3flox/flox animals were euthanized and transcardially perfused with saline followed by 4% PFA. Brains were excised and placed in ice-cold 4% PFA for 4 h, and then in 20% sucrose at 4 C overnight. Brains were embedded, frozen in OCT; 8-µm cryostat sections were stored at –80 C. For hybridization, sections were dried for 1 h at room temperature, rinsed three times (5 min each) with PBS, treated with proteinase K (0.25 µg/ml PBS) for 10 min, and then rinsed briefly with (2 mg/ml) glycine before being placed into 4% PFA. After three more PBS rinses (5 min each), sections were treated with 0.1 M triethanolamine and 0.25% acetic anhydride for 10 min, washed with 2x standard saline citrate (SSC) three times (2 min each), and then dehydrated through an alcohol series. Hybridization was carried out with 50–100 ng of digoxigenin-labeled cRNA probe overnight at 55 C, after which sections were washed two times with 2x SSC (1 min each), and then incubated with RNase (0.2 mg/ml) for 1 h at 37 C before being washed again in progressively decreasing concentrations of SSC (2x, 1x, 0.5x, 0.1x; 5 min each), followed by a final 1-h wash in 0.1x SSC at 67 C. After cooling to room temperature, sections were transferred to Tris-buffered saline, treated with 0.25% Triton X-100 and 1% hydrogen peroxide for 1 h, followed by 1% blocking reagent. Sections were incubated overnight at 4 C with rabbit anti-Cre (1:8000) or rabbit anti-Stat3 (1:1000) together with sheep anti-digoxigenin (1:200) coupled to alkaline phosphatase. Sections were then incubated for 1 h with a goat antirabbit antibody (1:100) coupled to horseradish peroxidase. After washing, Cre or Stat3 was visualized by Alexa 488-tyramide amplification, followed by detection of Npy cRNA with alkaline phosphatase histochemistry.
A similar approach was used to examine pStat3, except control and mutant mice were fasted overnight, euthanized 45 min after ip injection of 5 mg/kg mouse leptin (R&D Systems, Minneapolis, MN), and 30-µm floating sections prepared. pStat3 was detected using a rabbit anti-pStat3 primary antibody (Cell Signaling Technology, Beverly, MA) at 1:2000 dilution and an Alexa 568-coupled goat antirabbit secondary antibody (Molecular Probes, Eugene, OR). Fluorophores were visualized using an IX70 inverted Olympus microscope (Melville, NY) with mercury arc illumination and a standard filter cube. Images were captured using a high-sensitivity, variable integration time, color charge-coupled device video camera (Hitachi, Tokyo, Japan).
Measurement of Neuropeptide mRNA Levels
Tg.AgrpCre/Stat3flox/flox and Stat3flox/flox littermate controls were euthanized by cervical dislocation, and RNA was extracted from hypothalamic wedges using RNAeasy columns (Qiagen, Valencia, CA). Relative mRNA expression was determined by quantitative RT-PCR using the 2–
CT method (56) for which ß-actin served as the reference gene. Oligonucleotide primers were as follows: Npy, CTGACCCTCGCTCTATCTCTG and AGTATCTGGCCATGTCCTCTG; Pomc, CCCAAGGACAAGCGTTACGG and GTGCGCGTTCTTGATGATGG; Agrp, TTGTGTTCTGCTGTTGGCACT and AGCAAAAGGCATTGAAGAAGC; Actb, CAACGAGCGGTTCCGATG and CACTGTGTTGGCATAGAGG.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The authors have nothing to declare.
First Published Online May 18, 2006
Abbreviations: Agrp, Agouti-related protein; CNS, central nervous system; Npy, neuropeptide Y; PFA, paraformaldehyde; POMC, pro-opiomelanocortin; pStat, phosphorylated signal transducer and activator of transcription; SSC, standard saline citrate; Stat, signal transducer and activator of transcription; TetR, tetracycline resistance.
Received for publication March 6, 2006. Accepted for publication May 12, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
C(T) method. Methods 25:402–408[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  D. L. Morris and L. Rui Recent advances in understanding leptin signaling and leptin resistance Am J Physiol Endocrinol Metab, December 1, 2009; 297(6): E1247 - E1259. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Balthasar Feeding signals to the hungry mind Exp Physiol, August 1, 2009; 94(8): 857 - 866. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Gong, F. Yao, K. Hockman, H. H. Heng, G. J. Morton, K. Takeda, S. Akira, M. J. Low, M. Rubinstein, and R. G. MacKenzie Signal Transducer and Activator of Transcription-3 Is Required in Hypothalamic Agouti-Related Protein/Neuropeptide Y Neurons for Normal Energy Homeostasis Endocrinology, July 1, 2008; 149(7): 3346 - 3354. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. van de Wall, R. Leshan, A. W. Xu, N. Balthasar, R. Coppari, S. M. Liu, Y. H. Jo, R. G. MacKenzie, D. B. Allison, N. J. Dun, et al. Collective and Individual Functions of Leptin Receptor Modulated Neurons Controlling Metabolism and Ingestion Endocrinology, April 1, 2008; 149(4): 1773 - 1785. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. L. Piper, E. K. Unger, M. G. Myers Jr., and A. W. Xu Specific Physiological Roles for Signal Transducer and Activator of Transcription 3 in Leptin Receptor-Expressing Neurons Mol. Endocrinol., March 1, 2008; 22(3): 751 - 759. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. W. Xu, L. Ste-Marie, C. B. Kaelin, and G. S. Barsh Inactivation of Signal Transducer and Activator of Transcription 3 in Proopiomelanocortin (Pomc) Neurons Causes Decreased Pomc Expression, Mild Obesity, and Defects in Compensatory Refeeding Endocrinology, January 1, 2007; 148(1): 72 - 80. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
0.01, two-tailed Mann-Whitney test). G, X-gal staining scores from brain slices of BAC172ßgeoMUT transgenic animals that were either fasted for 48 h or free-fed. H, X-gal-positive cells on the surface of coronal slices from BAC172ßgeoMUT line N transgenic animals that were either free-fed (n = 5) or fasted for 48 h (n = 5).
