| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pediatrics (B.M.K.), Department of Internal Medicine (B.M.K., A.R.Z.), McDermott Center for Human Growth and Development (B.M.K., J.L.H., T.G. A.R.Z.), The University of Texas Southwestern Medical Center, Dallas, Texas 75390-8591
Address all correspondence and requests for reprints to: Andrew R. Zinn or Bassil M. Kublaoui, Department of Internal Medicine, McDermott Center for Human Growth and Development, The University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, Texas 75390-8591. E-mail: andrew.zinn{at}utsouthwestern.edu or bassil.kublaoui{at}utsouthwestern.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The hypothalamus is a key regulator of energy homeostasis. Leptin secreted by adipocytes activates receptors in the arcuate nucleus, increasing Pomc expression and stimulating release of 
-MSH from neurons that project to the PVN and lateral hypothalamus (5). -MSH binds melanocortin-4 receptors in the PVN and inhibits feeding (6). Leptin also inhibits release of the orexigenic peptides neuropeptide Y and agouti-related peptide (Agrp) from other arcuate nucleus neurons that also project to the PVN and lateral hypothalamus (7, 8). Thus leptin acts on both orexigenic and anorexigenic pathways to reduce feeding and increase energy expenditure.
The Mc4r knockout mouse is obese, hyperphagic, and long (9). It recapitulates the phenotype of the Ay mouse, which ectopically expresses agouti protein, a melanocortin receptor antagonist, in all tissues (10, 11). The phenotypes of Ay and Mc4r knockout mice are also characterized by an abnormal homeostatic response to increased dietary fat (12).
Sim1 heterozygous mice also show hyperphagic obesity, increased linear growth, and sensitivity to a high-fat diet (1), and a girl with severe early-onset obesity and one allele of SIM1 disrupted by a balanced chromosome translocation (3) was clinically similar to children with heterozygous MC4R mutations (13, 14). Within the central nervous system, Sim1 and Mc4r are most abundantly expressed in the PVN (1, 15). The similar phenotypes of Mc4r and Sim1 mutations and expression of both genes in the PVN led us to investigate their anatomic and functional relationship in the regulation of energy balance. Our hypothesis is that regulation of food intake by Mc4r signaling in the PVN is dependent on Sim1.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). Elevated Pomc mRNA with hyperphagia and normoleptinemia suggests resistance to central -MSH action. To test whether this is the case, we injected mice ip with MTII, a potent melanocortin agonist. This route of MTII administration has been shown to suppress feeding via activation of hypothalamic melanocortin receptors, albeit at a higher dose (5–20 mg/kg) than is needed using intracerebroventricular administration (16, 17, 18, 19).
|
, B and C), whereas Sim1 heterozygotes showed a blunted response at both time points (57% and 33%, respectively) (Fig. 1, B and C). Suppression of feeding by MTII waned by 24 h in both groups (Fig. 1D).
Sim1 Heterozygotes Have Normal Energy Expenditure
Previous reports indicated that Mc4r knockout mice show both increased food intake and decreased energy expenditure (20), whereas Sim1 heterozygotes show only hyperphagia (2). We confirmed that energy expenditure was normal in our Sim1 heterozygotes (Fig. 2, A–D). VO2, VCO2, metabolic rate, and activity were indistinguishable from wild-type littermates in either the light or the dark cycle.
|
, A–C).
|
). Injection of hypertonic saline induced PVN c-Fos immunoreactivity equally well in wild-type and Sim1 heterozygous mice (Fig. 4), demonstrating that the defective MTII response of Sim1 heterozygotes was not due to a decrease in the number or general responsiveness of PVN neurons.
|
shows that peripheral MTII administration induced c-Fos immunoreactivity in a subset of eGFP-expressing PVN neurons. Confocal microscopy confirmed that nuclear c-Fos and cytoplasmic eGFP colocalized within cells (data not shown). Some neurons appeared to be positive for c-Fos but negative for eGFP immunoreactivity. However, the anti-eGFP antibody used in this experiment, chosen for compatibility with the c-Fos antibody, showed suboptimal penetration. Immunostaining with a stronger anti-GFP antibody together with an antibody to the neuronal marker Neun confirmed that all neurons in the PVN of the adult mouse express Sim1 (data not shown).
|
, B and C, shows representative images of the PVN of wild-type and Sim1 heterozygous mice. There were no visible differences between the two, nor was there a significant difference in the number of GFP-positive neurons between the two groups (Fig. 5D). There was also no significant difference in PVN area between wild-type and Sim1 heterozygous mice (Fig. 5E).
Leptin and MTII Induce Hypothalamic Sim1 Expression
We asked whether Sim1 expression in the hypothalamus is regulated by leptin or melanocortin signaling. Female C57BL/6 mice, ages 6–8 wk, were injected ip with vehicle or leptin, 5 mg/kg twice daily (b.i.d.), and food intake was measured over 2 d. Leptin-treated mice showed a 31% reduction in food intake (Fig. 6A). Hypothalamic Pomc, Agrp, and Sim1 RNA levels were measured using real-time PCR. As expected, Pomc expression was higher and Agrp expression lower after leptin treatment. Sim1 expression was increased by 50% after leptin treatment. We then asked whether MTII treatment acutely induces Sim1 expression. Female C57BL/6 mice, ages 6–8 wk, were fasted for 24 h before ip injection with either vehicle or MTII, 10 mg/kg (two doses, 2 h apart). MTII treatment led to an 80% increase in hypothalamic Sim1 expression (Fig. 7A). By contrast, oxytocin gene expression was unaffected (Fig. 7B), ruling out nonselective up-regulation of PVN gene expression by MTII.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Elevated hypothalamic Pomc expression despite hyperphagia, normoinsulinemia, and normoleptinemia (1) suggested that Sim1 heterozygous mice are resistant to the anorectic action of melanocortin signaling. The degree of elevation in hypothalamic Pomc expression is similar to that seen after leptin treatment (Fig. 6). This resistance was confirmed by the inability of MTII, a potent melanocortin agonist, to suppress feeding to the same extent in Sim1 heterozygotes vs. wild-type littermates. Although it is conceivable that the Sim1 deficiency may change the pharmacokinetics of CNS access, retention, or half-life of peripherally administered MTII, the coexpression of Sim1 and Mc4r in the same nuclei, and even the same neurons (16), favors an interaction with Mc4r signaling as a mechanism of resistance to the anorectic effect of MTII and the failure of c-Fos activation in Sim1 heterozygotes. More importantly, the fact that MTII induces a normal increase in energy expenditure in Sim1 heterozygotes argues strongly against a nonspecific pharmacokinetic effect. Although MTII activates both Mc3r and Mc4r in vitro, there is ample evidence that its anorectic effect and its effect on energy expenditure is largely, if not fully, mediated by Mc4r: MTII fails to reduce feeding or increase energy expenditure in mice lacking Mc4r (20) but exhibits its full anorectic effect in mice lacking Mc3r (23). In addition, Sim1 is expressed in the hypothalamic nuclei that express Mc4r but not Mc3r in the rat (15, 24, 25) and mouse (22). Finally, the phenotype of Mc3r knockout mice is distinct from that of Sim1 heterozygous, Mc4r knockout, or Ay mice: the Mc3r knockout mouse is hypophagic and has normal body weight with decreased lean mass and increased fat mass, suggesting a defect in energy partitioning (12, 23).
Sim1 heterozygotes and mice deficient in melanocortin signaling (Mc4r knockout and Ay) have similar phenotypes that include hyperphagia, increased length and lean mass, and sensitivity to high fat, but differ with regard to energy expenditure, which is normal in Sim1 heterozygotes but reduced in Mc4r knockout mice (20). In addition, Sim1 heterozygotes, but not Mc4r knockout mice, increase their energy expenditure normally in response to MTII. The basis for this dichotomy appears to be anatomic segregation of neurons where PVN Mc4r neurons mediate Mc4r effects on food intake but not energy expenditure, and Mc4r neurons elsewhere in the CNS mediate Mc4r effects on energy expenditure. The work of Balthasar et al. (16) showed that Mc4 receptors in Sim1 neurons in the PVN, and possibly the amygdala, act to regulate feeding but not energy expenditure. They showed that selective expression of Mc4r in Sim1-positive neurons normalized food intake but not energy expenditure of otherwise null Mc4r mice. Our results demonstrate a correlation between the failure of MTII to activate PVN neurons in Sim1 heterozygotes and resistance to its anorectic effect but not its effect on energy expenditure. The similar anorectic but different metabolic responses to MTII of Sim1 heterozygotes vs. Mc4r knockout mice also suggests that the Mc4r acts in the PVN (and possibly the amygdala) to regulate feeding but elsewhere in the CNS to regulate energy expenditure.
We showed that MTII-mediated c-Fos activation in the PVN colocalized with a subset of Sim1 neurons. Liu et al. (26) previously showed directly that the population of neurons in the PVN activated by MTII express Mc4r. Together with the work of Balthasar et al. (16), who reactivated a loxP-transcriptionally blocked Mc4r allele in the PVN using a Sim1-cre transgenic, and our studies using a Sim1-eGFP transgenic showing that all the neurons in the adult mouse PVN express Sim1, we conclude that Mc4r and Sim1 are coexpressed in a subset of PVN neurons.
That a quantitative deficiency in Sim1 (haploinsufficiency) resulted in a qualitative defect in c-Fos induction suggests that a certain threshold of Sim1 is necessary within PVN neurons for normal activation in response to Mc4r signaling. Possible mechanisms by which Sim1 deficiency can lead to defective PVN melanocortin responsiveness include 1) a subtle developmental phenotype where the number of Sim1 and/or Mc4r neurons is reduced, 2) decreased Mc4r expression, or 3) interference with effectors of the Mc4r signal. The first possibility was suggested by the fact that homozygous Sim1 null mice lack the PVN entirely (4). Our results showing that all PVN neurons in the adult mouse are Sim1 neurons are consistent with previous findings by in situ hybridization (1, 16). Our finding that the number of Sim1 neurons is not diminished in Sim1 heterozygotes argues against hypocellularity as a mechanism of resistance to MTII. Our results are inconsistent with those of Michaud et al. (2), who reported 24% fewer cells in the PVN of Sim1 heterozygotes, with no particular subtype affected. The source of this discrepancy is unclear but may be due to methodological differences. They used hematoxylin staining, which identifies nuclei and therefore both neurons and glial cells, whereas we used GFP immunohistochemistry to specifically count Sim1 neurons.
We previously showed that Sim1 heterozygotes express normal levels of hypothalamic Mc4r (1). Because hypothalamic Mc4r expression is enriched in the PVN (15, 16), this result argues against the second mechanism and is further evidence against a reduction of Mc4r neurons in Sim1 heterozygotes. This leaves the third possibility, interference with melanocortin signaling effector(s), as the most likely explanation for PVN hyporesponsiveness in Sim1 heterozygotes.
Activation of normal numbers of PVN neurons in Sim1 heterozygotes in response to an osmotic stimulus argues against a generalized hyporesponsiveness of these neurons. It is notable that Sim1 heterozygotes have a higher basal c-Fos expression in the PVN. The identity of these basally activated neurons is unclear, and so they may or may not be Mc4r-expressing neurons. The finding that both leptin and MTII induce hypothalamic Sim1 expression, the latter acutely, suggests that Sim1 may act downstream of Mc4r. The fact that MTII induces c-Fos in a substantial subset of Sim1 neurons raises the possibility that Sim1 is an intimate component of hypothalamic melanocortin signaling. In this regard, an increase in hypothalamic Sim1 expression in response to anorectic signals fits with the fact that decreased Sim1 expression is associated with hyperphagia.
In summary, our results suggest that Sim1 itself and/or its transcriptional target(s), as yet unidentified, are part of the Mc4r signaling pathway in a subset of PVN neurons. Figure 8 shows our proposed model whereby activation of Mc4r in PVN neurons leads to induction of Sim1, which is necessary for the inhibition of feeding. The dichotomous effects of Sim1 haploinsufficiency on MTII-induced food intake and energy expenditure agree with the recent finding of Balthasar et al. (16) that Sim1 neurons that express Mc4r in the PVN regulate food intake, and Mc4r neurons elsewhere in the CNS regulate energy expenditure. Further experiments, such as conditional inactivation of Sim1 in adult mice together with identification of genes that are regulated by Sim1, are needed to elucidate the mechanism of its interaction with melanocortin signaling in hypothalamic regulation of feeding.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Real-Time PCR
Real-time PCR was performed as previously described (1). Briefly, hypothalami from fresh brains were dissected with a block (David Kopf Instruments, Tujunga, CA), using the following landmarks: posteriorly, posterior aspect of median eminence; anteriorly, 5 mm anterior to the median eminence; dorsally, the thalamus; laterally, medial to the dentate gyrus. Total RNA was extracted using Tripure reagent (Roche Applied Science, Indianapolis, IN). Quantitative real-time PCR was performed using an MJ Research Opticon instrument (Bio-Rad Laboratories, Inc., Hercules, CA) and the QuantiTect HotStart SYBR green qPCR kit (QIAGEN, Valencia, CA). Pomc, Agrp, Sim1, and Oxytocin measurements were normalized to ß-actin mRNA levels. Primers sequences were 5'-GCC CTC CTG CTT CAG ACC TC-3' and 5'-CGT TGC CAG GAA ACA CGG-3' (Pomc); 5'-GAC GAT GCT CCC CGG GCT GTA TTC-3' and 5'-TCT CTT GCT CTG GGC CTC GTC ACC-3' (b-actin); 5'-GAG GCA GGC AGG TAC TT-3' and 5'-CTG ACC ACA CTA TCT TCA T-3' (Sim1); 5'-TCC CAG AGT TCC CAG GTC TAA GTC-3' and 5'-GCG GTT CTG TGG ATC TAG CAC CTC-3' (Agrp); and 5'-TGG CTT ACT GGC TCT GAC CT-3' and 5'-AGG CAG GTA GTT CTC CTC CTG-3' (Oxytocin). All reactions were subjected to 40 cycles of amplification (denaturation at 94 C for 15 sec, annealing at 53 C for 30 sec, and extension at 72 C for 30 sec). Standard curves were generated using reference cDNA prepared from normal mouse hypothalamus and used to normalize measurements from experiment to experiment. All measurements were made in the exponential phase of the real-time PCR, as described by the manufacturer (Bio-Rad). Reactions were performed in triplicate and the results averaged. The coefficient of variation was less than 15% for each set of measurements.
MTII Treatment
MTII was purchased from Bachem (King of Prussia, PA). Female Sim1 heterozygous mice and wild-type littermates, ages 6–8 wk, were housed individually for 1 wk and handled daily. Five days before experiments they were habituated to daily ip injection using sterile PBS. For measurement of feeding, we used the protocol of Reizes et al. (18). Mice were habituated as above and fed ad libitum. The experiment was performed on 2 separate days (2 d washout) with mice randomly assigned at the beginning to receive PBS or MTII first. On the day of the experiment, food was removed 2.5 h before the onset of the dark cycle (1630 h). At the onset of the dark cycle (1900 h), mice were injected ip with 0.5 ml of PBS or 5 mg/kg MTII in PBS. Cumulative food intake was measured 2, 4, and 24 h after injection. For evaluation of energy expenditure after MTII, we used the protocol of Chen et al. (20) and Balthasar et al. (16). A new cohort of Sim1 heterozygotes and wild-type littermates was habituated to the metabolic cages for 2 d. On d 3 at 1200 h, food was removed. At 1230 h, all mice received 0.2 ml PBS IP, and VO2, VCO2, and metabolic rate were measured for the subsequent 4 h. Food was replaced at 1700 h. On d 4, all mice received 20 mg/kg ip MTII, and VO2, VCO2, and metabolic rate were measured as was done on d 3. For evaluation of c-Fos immunoreactivity, another cohort of Sim1 heterozygotes and wild-type littermates was fasted for 24 h, injected ip with PBS or 10 mg/kg MTII, and killed 3 h later by transcardial perfusion (see below). Induction of c-Fos immunoreactivity in the PVN is a well-established marker of acute melanocortin signaling (5, 27). For evaluation of c-Fos immunoreactivity in Sim1-GFP mice, 6- to 8-wk-old female mice were fasted for 24 h, injected with 10 mg/kg MTII, and killed 3 h later by transcardial perfusion (see below). For all MTII experiments, there was no statistical difference in weight between wild-type and Sim1 heterozygous mice used in this experiment.
Hypertonic Saline Treatment
Hypertonic saline has been shown to be a nonspecific osmotic stimulus of c-Fos immunoreactivity in PVN neurons including magnocellular, parvocellular, vasopressin, oxytocin, and tyrosine hydroxylase neurons (28). Mice were habituated as above. They were injected ip with 0.5ml of 1.5 M NaCl, and water was removed. Brains were harvested 1 h later after transcardial perfusion (see Immunohistochemistry).
Immunohistochemistry
Mice were deeply anesthetized with pentobarbital (7.5 mg/0.15 ml, ip) and transcardially perfused with 10 ml of heparinized saline (10 U/ml, 2 ml/min) followed by 10 ml of phosphate-buffered 4% paraformaldehyde (2 ml/min). Brains were removed, postfixed for 24 h in 4% paraformaldehyde, and then equilibrated in 30% sucrose in PBS for 72 h. Immunohistochemistry was performed as described elsewhere (29). Briefly, brains were coronally sectioned (35 µm) on a freezing microtome and stored in PBS at 4 C. Sections containing the PVN were from Bregma –0.58 mm to –1.22 mm according to Ref. 30 . For colorimetric c-Fos immunohistochemistry, sections were incubated for 16 h at 4 C in rabbit anti-Fos antiserum (Ab-5; 1:3000 dilution; Calbiochem, San Diego, CA), incubated with biotinylated goat antirabbit IgG secondary antiserum (1:600;Vector Laboratories, Burlingame, CA) for 2 h at room temperature, and then incubated in avidin-biotin complex (Vector Laboratories). A black reaction product was produced in cell nuclei using diaminobenzidine with NiSO4.
For fluorescent c-Fos and eGFP double labeling, sections were incubated for 48 h at room temperature in rabbit anti-Fos antiserum (Ab-5; 1:3000 dilution; Calbiochem) and fluorescein-conjugated goat anti-GFP (ab6662; 1:5000 dilution; Abcam, Cambridge, MA). They were then incubated with Cy3-conjugated goat antirabbit IgG secondary antiserum (1:600; Vector Laboratories) for 2 h at room temperature. For fluorescent eGFP labeling, sections were incubated for 72 h in rabbit anti-eGFP antiserum (A-6455; 1:10,000 dilution; Molecular Probes, Eugene, OR) followed by Cy2-conjugated donkey antirabbit IgG secondary antiserum (711-225-152; 1:200 dilution; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) for 2 h at room temperature. Images of sections containing PVN were captured using BQ Nova software Bioquant Image Analysis Corp., Nashville, TN. c-Fos- or eGFP-positive neurons were counted by an investigator blinded to their genotypes. Each side of a section was counted separately, and counts from six sides were averaged for each animal. Area was measured using BQ Nova software and represents the area of one PVN side. Counts of nuclear c-Fos positive neurons were greater than those of cytoplasmic eGFP-positive neurons due to differences in the resolution of superimposed cells in these thick 35-µm sections.
Metabolic Cage Studies
Indirect calorimetry was performed using a 12-cage equal flow CLAMS calorimeter (Columbus Instruments, Columbus, OH). Mice were habituated to the metabolic cages for 2 d before beginning data acquisition. There was no statistical difference in weight between wild-type and Sim1 heterozygous mice used in this experiment.
Leptin Treatment
Leptin was purchased from R&D systems (Minneapolis, MN). C57BL/6 mice (9 wk of age) were housed individually for 1 wk and handled daily. They were habituated, 5 d before experiments, to daily ip injection using sterile PBS. Mice were treated with PBS or leptin 5 mg/kg ip b.i.d. at 0900 h and 1800 h for 2 d (five injections). Food and mice were weighed daily. Hypothalami were isolated 2 h after the last injection as described above.
Data Analysis
Data were analyzed and plotted using Microsoft Excel and GraphPad Prism 4 (GraphPad Software, San Diego, CA). Unless otherwise noted, means were compared using unpaired, two-tailed t tests, with Welch’s correction if F test indicated unequal sample variances. Multiple comparisons were compared using one-way ANOVA with Newman-Keuls multiple comparison post hoc test (Fig. 1, B–D). Differences were considered statistically significant if P < 0.05.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
B.K., J.H., T.G., and A.Z. have nothing to declare.
First Published Online May 25, 2006
Abbreviations: Agrp, Agouti-related peptide; b.i.d., twice daily; CNS, central nervous system; eGFP, enhanced GFP; GFP, green fluorescent protein; MTII, melanotan-2; PVN, paraventricular nucleus; SON, supraoptic nucleus.
Received for publication November 30, 2005. Accepted for publication May 15, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
melanotropin and other proopiomelanocortin peptides in the hypothalamus and limbic system. Proc Natl Acad Sci USA 90:8856–8860This article has been cited by other articles:
 |
|
 |
|
  B. M. Kublaoui, T. Gemelli, K. P. Tolson, Y. Wang, and A. R. Zinn Oxytocin Deficiency Mediates Hyperphagic Obesity of Sim1 Haploinsufficient Mice Mol. Endocrinol., July 1, 2008; 22(7): 1723 - 1734. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. E. Foster-Schubert and D. E. Cummings Emerging Therapeutic Strategies for Obesity Endocr. Rev., December 1, 2006; 27(7): 779 - 793. [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 |
