Src Homology 3-Domain Growth Factor Receptor-Bound 2-Like (Endophilin) Interacting Protein 1, a Novel Neuronal Protein that Regulates Energy
http://www.100md.com
《内分泌学杂志》
Metabolic Research Unit (J.T., K.W., V.F., L.K.-B., J.M., A.C., S.L., K.B., M.P., R.F., K.W., G.R.C.), School of Exercise and Nutrition Sciences, Deakin University, and ChemGenex Pharmaceuticals Ltd. (K.W., G.R.C.), Waurn Ponds 3217, Victoria, Australia
Department of Medicine (G.J.M., M.W.S.), Harborview Medical Center and University of Washington, Seattle, Washington 98104
Pennington Biomedical Research Center (J.T.), Baton Rouge, Louisiana 70808
Abstract
To identify genes involved in the central regulation of energy balance, we compared hypothalamic mRNA from lean and obese Psammomys obesus, a polygenic model of obesity, using differential display PCR. One mRNA transcript was observed to be elevated in obese, and obese diabetic, P. obesus compared with lean animals and was subsequently found to be increased 4-fold in the hypothalamus of lethal yellow agouti (Ay/a) mice, a murine model of obesity and diabetes. Intracerebroventricular infusion of antisense oligonucleotide targeted to this transcript selectively suppressed its hypothalamic mRNA levels and resulted in loss of body weight in both P. obesus and Sprague Dawley rats. Reductions in body weight were mediated by profoundly reduced food intake without a concomitant reduction in metabolic rate. Yeast two-hybrid screening, and confirmation in mammalian cells by bioluminescence resonance energy transfer analysis, demonstrated that the protein it encodes interacts with endophilins, mediators of synaptic vesicle recycling and receptor endocytosis in the brain. We therefore named this transcript Src homology 3-domain growth factor receptor-bound 2-like (endophilin) interacting protein 1 (SGIP1). SGIP1 encodes a large proline-rich protein that is expressed predominantly in the brain and is highly conserved between species. Together these data suggest that SGIP1 is an important and novel member of the group of neuronal molecules required for the regulation of energy homeostasis.
Introduction
THE HYPOTHALAMUS IS a critical regulator of energy balance, transducing information from both metabolic and humoral signals into adaptive changes in energy intake and expenditure (1, 2). In recent years, our understanding of key molecules involved in the control of food intake and body weight has grown considerably, resulting in the identification of novel obesity-inducing mutations. However, such mutations have been found in only a very small number of obese humans, the largest of these being associated with an estimated 5% of human obesity cases (3). The underlying genetic determinants of common human obesity, therefore, remain poorly understood.
These considerations suggest that additional genes involved in the hypothalamic control of energy homeostasis remain to be discovered. As a first step in identifying these potential genes, we used Psammomys obesus, a unique polygenic animal model of obesity and type 2 diabetes. Previous studies from our laboratory have demonstrated that P. obesus spontaneously develop a wide range of metabolic phenotypes that reflect the range of syndromes observed before (and including) the development of type 2 diabetes (4, 5, 6). To identify novel differentially expressed mRNA transcripts that may regulate this metabolic syndrome, we used differential display PCRs to conduct a large-scale screen of hypothalamic mRNA from lean and obese P. obesus as previously described (7, 8). Here we characterize a novel neuronal transcript, termed SGIP1 [Src homology 3-domain growth factor receptor-bound 2 (GRB2)-like (endophilin) interacting protein 1], which displayed increased mRNA content in the hypothalamus in obesity and after fasting. Furthermore, targeted reduction of hypothalamic SGIP1 mRNA inhibited food intake and decreased body weight, suggesting that this transcript plays an obligatory role in the regulation of energy homeostasis.
Materials and Methods
Experimental animals
Studies in P. obesus and Sprague Dawley rats were conducted in accordance with the regulations and guidelines outlined by the National Health and Medical Research Council of Australia and approved by the Deakin University Ethics Committee. Protocols for Ay/a mice were approved by the Institutional Animal Care and Use Committee of the University of Washington (Seattle, WA) and were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.
P. obesus
A colony of P. obesus is maintained at Deakin University (Geelong, Australia) using the San Poiley outbreeding method (9). Male Sprague Dawley rats (Animal Resource Center, Perth, Australia) were maintained at Deakin University. All animals were fed a diet of standard laboratory chow from which 63% of energy was derived from carbohydrate, 25% from protein, and 12% from fat (Barastoc, Victoria, Australia) and maintained in a temperature-controlled room (22 ± 1 C) with a 12-h light,12-h dark cycle. P. obesus was classified as either lean normal glucose tolerance (nGT), obese impaired glucose tolerance (IGT), or obese diabetic (D2M) according to circulating insulin and blood glucose values and body weight as previously described (4, 6). For differential display PCR (n = 2 per group) and real-time RT-PCR (n = 8 per group), 18-wk-old male lean nGT, obese IGT, and obese D2M P. obesus were separated into two treatment groups: fed and fasted (24 h). Groups were matched for body weight, blood glucose, and plasma insulin. Animals were killed by anesthetic overdose (pentobarbitone, 120 mg/kg) and the hypothalamus rapidly excised, snap frozen in liquid nitrogen, and stored at –80 C.
Agouti (Ay/a) mice
Male wild-type (C57BL6; n = 5) and lethal yellow agouti (Ay/a, n = 5) mice were fed ad libitum. Animals were killed by brief inhalation of CO2 and decapitation. The brain was removed, and a mediobasal hypothalamic wedge (defined caudally by the mamillary bodies, rostrally by the optic chiasm, laterally by the optic tract, and superiorly by the apex of the third ventricle) was dissected, snap frozen, and stored at –80 C for subsequent analysis.
Differential-display PCR
Total hypothalamic RNA was isolated using the RNAzol B method (Tel-Test, Friendswood, TX), treated with 1 U/μl of DNase I (Life Technologies Inc., Gaithersburg, MD), and reverse-transcribed using Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA). Differential-display PCR was performed on hypothalamic cDNA samples using an RNAimage mRNA differential display system (GenHunter, Nashville, TN). SGIP1 was amplified using the A-anchored primer (5'-AAGCT11A-3) and arbitrary primer 38 (5'-CTGCTGCTCCTCCTC-3').
Sequencing
The 5' end of the SGIP1 transcript isolated by differential-display PCR was ascertained using the FirstChoice RLM-RACE kit (Ambion, Austin, TX). PCR products were sequenced using the ABI PRISM BigDye terminator cycle sequencing ready reaction mix and a 373 automated fluorescent DNA sequencer (Applied Biosystems, Foster City, CA).
Real-time RT-PCR
Total hypothalamic RNA was isolated using TRIzol (Invitrogen, Carlsbad, CA) combined with RNeasy mini kit columns (QIAGEN, Chatsworth, CA). RNA quality and quantity were assessed using an Agilent 2100 bioanalyzer and the RNA 6000 Nano LabChip kit (Agilent Technologies, Waldbronn, Germany). Hypothalamic RNA was reverse transcribed using Superscript II RT system (Invitrogen). Oligonucleotide primers for SGIP1 (sense, 5'-TGAAGGCTTCCATAGGCAACA-3'; antisense, 5'-TGGAACGCCTGGGTCTTG-3') were designed using the Primer Express 2.0 software program (Applied Biosystems), and synthesized by Geneworks (Adelaide, Australia). PCR was performed using SYBR Green PCR master mix and an ABI PRISM 7700 sequence detector (Applied Biosystems).
Northern blot analysis
A cDNA probe complementary to SGIP1 mRNA was generated by PCR using primers: sense, 5'-CTTAGCTGGGACCTGGAA-3'; antisense, 5'-GACTCCATGCTTGGGTTAACT-3'. PCR was performed using AmpliTaq Gold master mix (Applied Biosystems) and P. obesus hypothalamic cDNA. Purified DNA probe was radioactively labeled with 32P-dATP (>3000 Ci/mmol; PerkinElmer, Melbourne, Australia) using the Prime-It II random primer labeling kit (Stratagene, La Jolla, CA). Unincorporated radioactivity was removed using the Qiaquick nucleotide removal kit (QIAGEN). SGIP1 DNA probe corresponding to 1 x 106 to 1 x 107 cpm was denatured and used to probe a multiple tissue Northern human RNA blot (BD Biosciences CLONTECH, Palo Alto, CA). The blot was exposed to a phosphor screen overnight and the image captured using the Molecular Dynamics PhosphorImager SI (Sunnyvale, CA).
Intracerebroventricular (ICV) antisense administration
For ICV studies, 16-wk-old male lean P. obesus or 10-wk-old male Sprague Dawley rats were used. A 23G cannula was implanted into the right lateral ventricle on d –7 (2.0 mm posterior; 2.0 mm ventral from the bregma, lowered 2.0 mm below the skull for P. obesus; or 1.5 mm lateral and 0.2 mm posterior from the bregma, and lowered 4.0 mm for Sprague Dawley rats). An Alzet miniosmotic pump (DURECT Corp., Cupertino, CA) was implanted sc into the scapular region and connected to the cannula at d 0. The pumps delivered SGIP1 antisense oligonucleotide (ODN; n = 6–8), jumbled ODN resuspended in sterile 0.9% saline (n = 6–8), or saline alone (n = 6–8) at a rate of 1 μl/h (24 μg/d) for 4 d. SGIP1 antisense ODN (5'-mUmCmAmGmUCCTTCCATmCmAmUmGmG-3') and jumbled ODN (5'-mCmGmCmAmCTTAGCTACmUmUmGmCmU-3'), where m indicates a 2'-O-methyl (2'OMe)-modified base, and all bonds were phosphorothioate linked and synthesized by Integrated DNA Technologies (Coralville, IA).
Energy expenditure measurements
Ten-week-old male Sprague Dawley rats had an ICV cannula implanted into the right lateral ventricle on d –10. At d –3, the rats were placed in a whole-body Oxymax equal flow indirect calorimeter (Columbus Instruments, Columbus, OH) for 4–6 h for acclimatization. At d –1 and again at d 3, the animals were placed in the calorimeter for 24 h during which oxygen consumption and carbon dioxide production were measured. Physical activity was measured using an Opto-Varimax miniinfrared animal activity monitor system (Columbus Instruments). At d 0, the rats were implanted with an Alzet pump containing either SGIP1 antisense ODN (n = 4) or jumbled ODN (n = 11) and connected to the ICV cannula as described. To control for the effects of weight loss, seven of the jumbled ODN-treated group were pair fed to the average daily food intake of the SGIP1 antisense group. Estimations of energy production rate, or total energy expenditure (TEE), from oxygen consumption and carbon dioxide production were calculated from previously described equations (10).
Yeast two-hybrid screen
The yeast two-hybrid screen was conducted using the ProQuest two-hybrid system (Life Technologies). Two partial fragments of SGIP1 DNA were cloned into the yeast bait vector pDBLeu and transformed into DH10B cells by electroporation. One fragment consisted of the central proline-rich region of SGIP1 (SGIP1-PR; amino acid 240–499), and the other consisted of the N terminal plus the proline-rich region (SGIP1-NP; amino acid 1–499). Constructs were transformed into the yeast strain MaV203, and the amount of 3-amino-1,2,4-triazole required for suppression of basal HIS3 expression was determined empirically as 25 mM. MaV203 cells harboring the SGIP1 constructs were used in a large-scale transformation with a human brain cDNA expression library (Life Technologies). The 5 x 106 to 1 x 107 transformants were plated onto selective media containing 25 mM 3-amino-1,2,4-triazole but lacking leucine, tryptophan, and histidine. Transformants that induced HIS3 expression were predicted to contain potential interacting proteins. Putative HIS-positive transformants were then tested for induction of two other reporters, URA3 and lacZ. The plasmids isolated from transformants positive for reporter gene expression were sequenced and retransformation performed to confirm the interaction.
Bioluminescence resonance energy transfer (BRET)
The entire human endophilin-3 coding sequence was amplified and cloned into the XhoI/BamHI sites of pGFP2-N2. To generate Rluc-SGIP1 fusions, the entire SGIP1 coding sequence and the SGIP1-NP was amplified and cloned into the XhoI/BamHI sites of pRluc-C2(h). For BRET assays, HEK-293 cells, grown in DMEM supplemented with 10% fetal bovine serum, were transiently transfected with the indicated plasmids in 6-well plates using Lipofectamine 2000 (Invitrogen). Twenty-four hours after transfection, cells were washed twice with PBS containing 0.1 g/liter calcium chloride, 0.1 g/liter magnesium chloride, 1 g/liter D-glucose, and 2 μg/ml aprotinin and then detached and resuspended to a final concentration of 2 x 106 cells/ml. Aliquots of 5 x 104 cells were distributed in a 96-well plate (White OptiPlate, PerkinElmer, Boston, MA), and DeepBlueC coelenterazine substrate was added to a final concentration of 5 μM. Emitted luminescence (410 nm) and fluorescence (515 nm) were detected immediately after addition of the substrate using a Fusion universal microplate analyzer (PerkinElmer). All values are expressed as a BRET ratio, which is defined as the [(emission at 515 nm – emission at 515 nm of nontransfected cells)/(emission at 410 nm – emission at 410 nm of nontransfected cells)].
Statistical methods
For comparisons involving three or more groups, data were assessed for normality by Kolmogorov-Smirnov test, and then one-way ANOVA was performed with the least significant differences or Games-Howell post hoc test for homogeneous or nonhomogeneous data, respectively. Comparisons of between-group mean values were performed using an unpaired Student’s t test for two-group comparisons or paired t test for comparison between two time points of the same group. For the analysis of energy expenditure because energy expenditure needs to be adjusted for weight, analysis of covariance was performed in which the change in TEE from baseline was the dependent variable, treatment group was the independent variable, and weight at baseline was the covariate. To analyze the change in body weight, ANOVA was used with the change in body weight from baseline as response and treatment group as the independent variable. Statistical analyses were performed using Statistical Package for the Social Sciences (SPSS, version 10.1; SPSS Inc., Fullerton, CA) or Statistical Analysis Software (SAS Institute Inc., Cary, NC). Statistical significance was defined as P < 0.05. All values are presented as mean ± SE.
Results
Discovery and characterization of SGIP1
SGIP1 (accession no. AY611625) was one of 27 novel transcripts isolated from approximately 9000 putative transcripts generated by more than 180 differential-display PCRs using hypothalamic mRNA from lean nGT, obese hyperglycemic (IGT), and obese D2M P. obesus (7). Real-time RT-PCR analysis of SGIP1 mRNA content confirmed increased hypothalamic SGIP1 gene expression in obese IGT and D2M animals, compared with lean nGT animals in the ad libitum-fed state (Fig. 1A). Linear relationships between SGIP1 gene expression and body weight (r = 0.674; P < 0.001) and percent body fat (r = 0.548; P < 0.01) were also observed (data not shown). To determine whether SGIP1 mRNA levels were also elevated in the hypothalamus of other mouse models of obesity and diabetes, we studied lethal yellow agouti (Ay/a) mice that develop obesity and diabetes due to a mutation that causes ectopic overexpression of agouti protein and subsequent chronic blockade of neuronal melanocortin receptors (11, 12). SGIP1 mRNA was increased approximately 4-fold in the hypothalamus of ad libitum-fed Ay/a mice relative to wild-type controls (Fig. 1B). Combined with data from P. obesus, these results suggest that increased hypothalamic SGIP1 gene expression may be a common feature in rodent models of obesity and diabetes.
To investigate tissue-specific distribution of SGIP1 gene expression, we measured SGIP1 mRNA in P. obesus brain, liver, adipose tissue, muscle, heart, and spleen using real-time RT-PCR. We found SGIP1 mRNA to be highly expressed in brain, with minimal levels also detected in adipose tissue and spleen (Fig. 1C). This pattern of SGIP1 gene expression is similar to that determined by Northern blot analysis of human tissues, which revealed a specific band of approximately 6 kb in brain RNA that was not detected in other tissues (Fig. 1D). Using in situ hybridization, we found SGIP1 mRNA to be expressed diffusely throughout the brain (data not shown).
A 6317-nucleotide SGIP1 mRNA sequence encoding an 827-amino acid protein was obtained from P. obesus hypothalamic cDNA using 5' rapid amplification of cDNA ends. The human ortholog of SGIP1 (reference sequence NM_032291), the hypothetical transcript DKFZp761D221, is 4572 nucleotides long and predicted to encode an 828-amino acid protein that is 94% identical with P. obesus SGIP1. The human SGIP1 gene is comprised of 27 exons and located on chromosome 1p31.3. SGIP1 is proline rich, with 13% of its amino acid content comprising proline residues that are concentrated in the central portion of the protein. This proline-rich region contains numerous potential SH3- and WW-domain binding sites.
Suppression of SGIP1 expression reduces food intake and body weight
Because hypothalamic SGIP1 expression is elevated in obese animals, we sought to determine (using antisense ODN technology) whether reduced hypothalamic expression of this transcript lowers food intake and body weight. Before ICV infusion into experimental animals, we tested several ODN sequences to verify successful suppression of SGIP1 mRNA in a cell-based system. A 2'OMe chimeric antisense ODN targeted to the initiation codon of SGIP1 consistently suppressed endogenous SGIP1 mRNA by 70–80% after transfection of GT1–7 mouse neuronal cells (supplemental Fig. 1, published as supplemental data on The Endocrine Society’s Journals Online Web site at http://endo.endojournals.org). Infusion of 2'OMe chimeric SGIP1 antisense ODN into the lateral ventricle of lean, male P. obesus at a rate of 24 μg/d for 4 d significantly decreased cumulative food intake, compared with both jumbled ODN (27.4 ± 4.4 vs. 39.0 ± 1.8 g, P = 0.015) and saline-treated animals (27.4 ± 4.4 vs. 46.3 ± 2.7 g, P < 0.001; Fig. 2A). This reduction of food intake was accompanied by a loss of body weight after SGIP1 antisense ODN treatment, compared with both jumbled ODN (–9.2 ± 1.5 vs. –4.0 ± 1.3 g, P = 0.02) and saline-treated controls (–9.2 ± 1.5 vs. –1.1 ± 1.3 g, P = 0.001) at d 4 (Fig. 2B). Whereas cumulative food intake was also slightly decreased by infusion of the jumbled ODN, compared with saline controls, significant differences in intake between these groups were detected only during the first 24 h of infusion.
To determine whether the weight-reducing effects of neuronal SGIP1 suppression can be generalized to other rodent species, we conducted a similar experiment in Sprague Dawley rats. Similar to the response of P. obesus, central SGIP1 antisense treatment reduced cumulative food intake, compared with either jumbled ODN (55.0 ± 5.3 vs. 101.5 ± 6.1 g, P < 0.001) or saline-treated control rats (55.0 ± 5.3 vs. 87.1 ± 4.1 g, P = 0.001) over the 4-d study (Fig. 2C), whereas food intake was not different between saline and jumbled ODN-treated rats. This reduction in food intake was again accompanied by a decrease in body weight in SGIP1 antisense-treated animals, compared with jumbled ODN (–36.6 ± 7.5 vs. 13.8 ± 7.6 g, P < 0.001) and saline-treated controls (–36.6 ± 7.5 vs. 1.8 ± 5.6 g, P < 0.001) after 4 d (Fig. 2D). To verify suppression of SGIP1 gene expression, hypothalamic SGIP1 mRNA levels were measured using real-time RT-PCR after 4 d of ICV infusion. Relative to saline-treated controls, hypothalamic SGIP1 mRNA was reduced by 37% by SGIP1 antisense treatment (P = 0.01, Fig. 2E). The effect of SGIP antisense ODNs in extrahypothalamic tissues is not known at this time, and it is possible that effects outside the hypothalamus may contribute to the phenotypes observed in this study. However, the suppression of SGIP1 mRNA in the hypothalamus, along with the well-known role of the hypothalamus in the regulation of energy balance, suggests a key role for SGIP1 in this area.
Collectively, these data suggest that endogenous SGIP1 protein provides a physiological stimulus for food intake. To further investigate this hypothesis, we measured the effect of fasting on SGIP1 mRNA levels in the hypothalamus of Sprague Dawley rats. Our finding that a 48-h fast increased SGIP1 mRNA content in these animals (Fig. 2F) supports the hypothesis that SGIP1 up-regulation contributes to the hypothalamic response to negative energy balance.
Effects of suppression of SGIP1 on energy expenditure
To determine whether increased energy expenditure contributes to weight loss induced by ICV SGIP1 antisense infusion, we used indirect calorimetry and measured physical activity in a separate group of Sprague Dawley rats. These measures were made for 24 h both before and 3 d after implantation of osmotic pumps for ICV infusion of antisense or jumbled ODN. To control for potential effects of reduced body weight and food intake on energy expenditure, a pair-fed group was included that received the jumbled ODN and was provided with food on a daily basis in an amount equal to the average daily food intake of SGIP1 antisense-treated animals. As observed previously suppression of SGIP1 reduced body weight significantly over the 4 d of treatment. Whereas pair-feeding also reduced body weight, ANOVA revealed a significantly greater weight loss among SGIP1 antisense-treated animals than pair-fed controls (Table 1). The reduction in body weight after SGIP1 antisense treatment therefore cannot be explained solely by reduced food intake. Consistent with this observation, SGIP1 antisense-treated animals failed to induce the appropriate reduction in TEE that normally occurs in rodents and humans during energy restriction (Table 1) (13, 14). Thus, although a significant reduction in TEE from baseline was observed after SGIP1 antisense treatment, analysis of covariance revealed that the reduction in TEE in pair-fed animals was significantly greater than the reduction in TEE of SGIP1 antisense-treated animals (Table 1). Our findings therefore suggest that in addition to inhibiting food intake, suppression of SGIP1 in Sprague Dawley rats prevents the reduction of energy expenditure that normally accompanies negative energy balance.
We then determined whether this failure to reduce TEE could be attributed to alterations in either metabolic heat production, a marker of metabolic rate, or physical activity. Whereas there were no differences in metabolic heat production between ad libitum-fed jumbled ODN controls and SGIP1 antisense-treated animals (despite a pronounced difference in body weight), pair-fed jumbled ODN-treated rats demonstrated reduced heat production, particularly at the end of the dark phase and during the majority of the light phase (Fig. 3A). These data suggest that SGIP1 antisense-treated rats exhibited an inappropriately elevated metabolic rate, compared with pair-fed animals, consistent with the estimates of TEE. Whereas jumbled ODN-treated ad libitum-fed animals exhibited the expected increase of physical activity during the dark phase (P < 0.03, Fig. 3B), this response was attenuated in both the SGIP1 antisense-treated and jumbled ODN-treated pair-fed rats. These observations suggest that decreased activity of SGIP1 antisense-treated animals was likely due, at least in part, to reduced food intake and body weight because activity also tended to be reduced in the pair-fed rats. Furthermore, these data indicate that whereas SGIP1 antisense treatment attenuated the effect of weight loss to reduce metabolic heat production and energy expenditure, it did not block its effect to reduce physical activity.
SGIP1 molecular pathway of action
To identify potential interacting partners of SGIP1, and thereby a possible molecular pathway whereby SGIP1 regulates energy balance, we performed a yeast two-hybrid screen of a human brain cDNA library using two truncated fragments of SGIP1 as bait. The two bait fragments used comprised the SGIP1-NP and SGIP1-PR (Fig. 4A). Both the SGIP1-NP and SGIP1-PR bait constructs strongly interacted with human endophilin-3 and the highly similar family member endophilin-1, most likely via the central proline-rich region common to both baits. The interaction between endophilin-3 and SGIP1 was then tested by BRET experiments in HEK-293 liver cells. SGIP1-NP fused to a luciferase reporter strongly interacted with endophilin-3 fused to green fluorescent protein (GFP) indicative of interaction (Fig. 4B). Likewise, full-length SGIP1 also resulted in a significantly increased BRET ratio when cotransfected with endophilin-3. No interaction was observed when SGIP1 constructs were cotransfected with a GFP construct containing ubiquilin or when endophilin-3 constructs were cotransfected with a vector containing only the luciferase reporter (Fig. 4B). The yeast two-hybrid screen data combined with the BRET experiments in mammalian cells confirm an interaction between SGIP1 and endophilin-3.
Discussion
In this report, we present the first experimental evidence for the existence of DKFZp761D221/SGIP1 and identify it as a novel, proline-rich protein predominantly expressed in brain and strongly implicated in the central regulation of energy balance. We found that food intake and body weight were reduced by suppression of hypothalamic SGIP1 mRNA and that hypothalamic expression of this transcript is increased both in rodent models of obesity and in response to fasting. These data collectively suggest a physiological role for this protein in hypothalamic neuronal systems that promote positive energy balance and weight gain. Inhibition of neuronal SGIP1 expression or signaling, therefore, constitutes a novel and potentially useful strategy for obesity treatment.
ICV infusion of 2'OMe-modified antisense ODN (15) targeted to the initiation codon of SGIP1 specifically suppressed endogenous SGIP1 mRNA in rat hypothalamus and resulted in approximately 40% inhibition of food intake and 5–8% reduction in body weight over the 4-d study period. Similar results were observed in P. obesus and are comparable in magnitude with reductions in food intake previously reported in rats after ICV antisense suppression of neuropeptide Y (NPY) (16, 17) or the NPY Y5 receptor (18, 19). To control for nonspecific effects of centrally administered ODN, we included control animals that received ICV infusion of a jumbled sequence ODN of the same oligonucleotide composition and chemistry as the SGIP1 antisense ODN. The lack of effect of the jumbled ODN on food intake and body weight suggests that toxicity or malaise associated with the antisense treatment is unlikely to explain the decreased food intake observed after SGIP1 suppression.
Two observations support the conclusion that in addition to decreasing food intake, depletion of hypothalamic SGIP1 mRNA reduced body weight by inducing an inappropriately high metabolic rate. First, weight loss was reduced in control rats pair fed to the intake of SGIP1 antisense-treated rats. In addition, we found that dietary energy restriction reduced TEE by 20% (P = 0.02), consistent with previous studies in rats (13, 20, 21), primates (22), and humans (14, 23). By comparison, SGIP1 antisense-treated animals failed to reduce their TEE relative to free-feeding, jumbled ODN-treated rats, despite their decreased food intake and body weight and despite showing an appropriate reduction in physical activity. The combination of reduced food intake with the inability to appropriately reduce energy expenditure may therefore explain the potent reduction in body weight seen in SGIP1 antisense-treated rats. Whether the decrease in body weight in SGIP1 antisense-treated animals is due to reduced lean body mass, fat mass, or both is not known and is an important question for future experiments.
Although SGIP1 is an intracellular protein rather than a neuropeptide, increased SGIP1 mRNA expression in the hypothalamus of obese P. obesus, and in fasted Sprague Dawley rats, is a feature shared by orexigenic neuropeptides including NPY, agouti-related protein, and melanin-concentrating hormone (24, 25, 26). Interestingly, SGIP1 mRNA was also increased in the hypothalamus of lethal yellow Ay/a mice that develop obesity as a result of chronic antagonism of hypothalamic melanocortin receptors by ectopically overexpressed agouti protein (11). Increased expression of SGIP1 may therefore contribute to the hyperphagia and obesity of Ay/a mice, and further studies are warranted to investigate this hypothesis.
Our finding that SGIP1 is a large proline-rich protein has implications for its intracellular function. Proline-rich regions are important for protein-protein interactions and are targets for proteins containing SH3 and WW domains (27). We used two truncated proline-rich fragments of SGIP1 as bait in a yeast two-hybrid screen. This analysis, subsequently confirmed by BRET experiments in mammalian cells, revealed a strong interaction between SGIP1 and endophilins, proteins initially isolated as SH3 domain-containing proteins (28, 29). Endophilins are important regulators of clathrin mediated endocytosis (30) and synaptic vesicle recycling (31, 32). They participate in receptor internalization and vesicle recycling by acting as part of a larger complex of proteins including synaptojanin (28, 33), amphiphysin (34), and dynamin (28). The majority of these studies have reported on endophilin-1; however, a recent report has shown an important role for endophilin-3 in the negative regulation of clathrin-mediated endocytosis in brain neurons and may regulate internalization of neurotransmitter receptors (35).
In conclusion, we provide evidence to show that SGIP1 is a novel brain-specific component of the neuronal molecular machinery that regulates energy homeostasis. Although the mechanism whereby SGIP1 favors increased food intake and body weight awaits further study, our data support a model in which SGIP1 functions as an endocytic protein that affects signaling by receptors in neuronal systems involved in energy homeostasis via its interaction with endophilin.
Acknowledgments
The authors thank Dr. Eric Ravussin and Haiyan Yang for assistance with interpretation and analysis of indirect calorimetry measurements and Drs. David Segal and Andrew Butler for helpful discussions and reading of the manuscript.
Footnotes
This work was supported by funding from ChemGenex Pharmaceuticals Ltd.; National Institutes of Health Grants DK52989, DK12829, DK68304, NS32273 (to M.W.S.); and the Clinical Nutrition Research Unit at the University of Washington.
Abbreviations: BRET, Bioluminescence resonance energy transfer; D2M, diabetic; GFP, green fluorescent protein; GRB2, growth factor receptor-bound 2; ICV, intracerebroventricular; IGT, impaired glucose tolerance; nGT, normal glucose tolerance; NPY, neuropeptide Y; ODN, oligonucleotide; 2'OMe, 2'-O-methyl; SGIP1, Src homology 3-domain GRB2-like (endophilin) interacting protein 1; SGIP1-NP, N terminal plus the proline-rich region of SGIP1; SGIP1-PR, proline-rich region of SGIP1; TEE, total energy expenditure.
References
Barsh GS, Schwartz MW 2002 Genetic approaches to studying energy balance: perception and integration. Nat Rev Genet 3:589–600
Schwartz MW, Woods SC, Porte Jr D, Seeley RJ, Baskin DG 2000 Central nervous system control of food intake. Nature 404:661–671
Farooqi IS, Keogh JM, Yeo GS, Lank EJ, Cheetham T, O’Rahilly S 2003 Clinical spectrum of obesity and mutations in the melanocortin 4 receptor gene. N Engl J Med 348:1085–1095
Barnett M, Collier GR, Collier FM, Zimmet P, O’Dea K 1994 A cross-sectional and short-term longitudinal characterisation of NIDDM in Psammomys obesus. Diabetologia 37:671–676
Collier GR, Walder K, Lewandowski P, Sanigorski A, Zimmet P 1997 Leptin and the development of obesity and diabetes in Psammomys obesus. Obes Res 5:455–458
Walder KR, Fahey RP, Morton GJ, Zimmet PZ, Collier GR 2000 Characterization of obesity phenotypes in Psammomys obesus (Israeli sand rats). Int J Exp Diabetes Res 1:177–184
Trevaskis J, McMillan JS, Windmill K, Walder K, Collier GR 2004 Identification of hypothalamic genes implicated in the development of obesity in Psammomys obesus using differential display PCR. Comp Biochem Physiol B Biochem Mol Biol 137:65–73
Collier GR, McMillan JS, Windmill K, Walder K, Tenne-Brown J, de Silva A, Trevaskis J, Jones S, Morton GJ, Lee S, Augert G, Civitarese A, Zimmet PZ 2000 Beacon: a novel gene involved in the regulation of energy balance. Diabetes 49:1766–1771
Poiley S 1960 A systematic method of breeder rotation for non-inbred laboratory animal colonies. Proc Animal Care Panel 10:159–166
Ferrannini E 1988 The theoretical bases of indirect calorimetry: a review. Metabolism 37:287–301
Klebig ML, Wilkinson JE, Geisler JG, Woychik RP 1995 Ectopic expression of the agouti gene in transgenic mice causes obesity, features of type II diabetes, and yellow fur. Proc Natl Acad Sci USA 92:4728–4732
Lu D, Willard D, Patel IR, Kadwell S, Overton L, Kost T, Luther M, Chen W, Woychik RP, Wilkison WO, Cone RD 1994 Agouti protein is an antagonist of the melanocyte-stimulating-hormone receptor. Nature 371:799–802
Even PC, Nicolaidis S 1993 Adaptive changes in energy expenditure during mild and severe feed restriction in the rat. Br J Nutr 70:421–431
Ravussin E, Burnand B, Schutz Y, Jequier E 1985 Energy expenditure before and during energy restriction in obese patients. Am J Clin Nutr 41:753–759
Dean NM, Butler M, Monia M, Manoharan M 2001 Pharmacology of 2'-O-(2-methoxy)ethyl-modified antisense oligonucleotides. In: Crooke S, ed. Antisense drug technology: principles, strategies and applications. New York: Marcel Dekker, Inc.; 319–338
Akabayashi A, Wahlestedt C, Alexander JT, Leibowitz SF 1994 Specific inhibition of endogenous neuropeptide Y synthesis in arcuate nucleus by antisense oligonucleotides suppresses feeding behavior and insulin secretion. Brain Res Mol Brain Res 21:55–61
Hulsey MG, Pless CM, White BD, Martin RJ 1995 ICV administration of anti-NPY antisense oligonucleotide: effects on feeding behavior, body weight, peptide content and peptide release. Regul Pept 59:207–214
Tang-Christensen M, Kristensen P, Stidsen CE, Brand CL, Larsen PJ 1998 Central administration of Y5 receptor antisense decreases spontaneous food intake and attenuates feeding in response to exogenous neuropeptide Y. J Endocrinol 159:307–312
Schaffhauser AO, Stricker-Krongrad A, Brunner L, Cumin F, Gerald C, Whitebread S, Criscione L, Hofbauer KG 1997 Inhibition of food intake by neuropeptide Y Y5 receptor antisense oligodeoxynucleotides. Diabetes 46:1792–1798
Gonzales-Pacheco DM, Buss WC, Koehler KM, Woodside WF, Alpert SS 1993 Energy restriction reduces metabolic rate in adult male Fisher-344 rats. J Nutr 123:90–97
Keesey RE, Corbett SW 1990 Adjustments in daily energy expenditure to caloric restriction and weight loss by adult obese and lean Zucker rats. Int J Obes 14:1079–1084
DeLany JP, Hansen BC, Bodkin NL, Hannah J, Bray GA 1999 Long-term calorie restriction reduces energy expenditure in aging monkeys. J Gerontol A Biol Sci Med Sci 54:B5–B11; discussion B12–B13
Hirsch J, Hudgins LC, Leibel RL, Rosenbaum M 1998 Diet composition and energy balance in humans. Am J Clin Nutr 67:551S–555S
Mizuno TM, Makimura H, Silverstein J, Roberts JL, Lopingco T, Mobbs CV 1999 Fasting regulates hypothalamic neuropeptide Y, agouti-related peptide, and proopiomelanocortin in diabetic mice independent of changes in leptin or insulin. Endocrinology 140:4551–4557
Sahu A, Kalra PS, Kalra SP 1988 Food deprivation and ingestion induce reciprocal changes in neuropeptide Y concentrations in the paraventricular nucleus. Peptides 9:83–86
Tritos NA, Mastaitis JW, Kokkotou E, Maratos-Flier E 2001 Characterization of melanin concentrating hormone and preproorexin expression in the murine hypothalamus. Brain Res 895:160–166
Zarrinpar A, Bhattacharyya RP, Lim WA 2003 The structure and function of proline recognition domains. Sci STKE 2003:RE8
Ringstad N, Nemoto Y, De Camilli P 1997 The SH3p4/Sh3p8/SH3p13 protein family: binding partners for synaptojanin and dynamin via a Grb2-like Src homology 3 domain. Proc Natl Acad Sci USA 94:8569–8574
Giachino C, Lantelme E, Lanzetti L, Saccone S, Bella Valle G, Migone N 1997 A novel SH3-containing human gene family preferentially expressed in the central nervous system. Genomics 41:427–434
Schuske KR, Richmond JE, Matthies DS, Davis WS, Runz S, Rube DA, van der Bliek AM, Jorgensen EM 2003 Endophilin is required for synaptic vesicle endocytosis by localizing synaptojanin. Neuron 40:749–762
Ringstad N, Gad H, Low P, Di Paolo G, Brodin L, Shupliakov O, De Camilli P 1999 Endophilin/SH3p4 is required for the transition from early to late stages in clathrin-mediated synaptic vesicle endocytosis. Neuron 24:143–154
Schmidt A, Wolde M, Thiele C, Fest W, Kratzin H, Podtelejnikov AV, Witke W, Huttner WB, Soling HD 1999 Endophilin I mediates synaptic vesicle formation by transfer of arachidonate to lysophosphatidic acid. Nature 401:133–141
Verstreken P, Koh TW, Schulze KL, Zhai RG, Hiesinger PR, Zhou Y, Mehta SQ, Cao Y, Roos J, Bellen HJ 2003 Synaptojanin is recruited by endophilin to promote synaptic vesicle uncoating. Neuron 40:733–748
Micheva KD, Ramjaun AR, Kay BK, McPherson PS 1997 SH3 domain-dependent interactions of endophilin with amphiphysin. FEBS Lett 414:308–312
Sugiura H, Iwata K, Matsuoka M, Hayashi H, Takemiya T, Yasuda S, Ichikawa M, Yamauchi T, Mehlen P, Haga T, Yamagata K 2004 Inhibitory role of endophilin 3 in receptor-mediated endocytosis. J Biol Chem 279:23343–23348(James Trevaskis, Ken Wald)
Department of Medicine (G.J.M., M.W.S.), Harborview Medical Center and University of Washington, Seattle, Washington 98104
Pennington Biomedical Research Center (J.T.), Baton Rouge, Louisiana 70808
Abstract
To identify genes involved in the central regulation of energy balance, we compared hypothalamic mRNA from lean and obese Psammomys obesus, a polygenic model of obesity, using differential display PCR. One mRNA transcript was observed to be elevated in obese, and obese diabetic, P. obesus compared with lean animals and was subsequently found to be increased 4-fold in the hypothalamus of lethal yellow agouti (Ay/a) mice, a murine model of obesity and diabetes. Intracerebroventricular infusion of antisense oligonucleotide targeted to this transcript selectively suppressed its hypothalamic mRNA levels and resulted in loss of body weight in both P. obesus and Sprague Dawley rats. Reductions in body weight were mediated by profoundly reduced food intake without a concomitant reduction in metabolic rate. Yeast two-hybrid screening, and confirmation in mammalian cells by bioluminescence resonance energy transfer analysis, demonstrated that the protein it encodes interacts with endophilins, mediators of synaptic vesicle recycling and receptor endocytosis in the brain. We therefore named this transcript Src homology 3-domain growth factor receptor-bound 2-like (endophilin) interacting protein 1 (SGIP1). SGIP1 encodes a large proline-rich protein that is expressed predominantly in the brain and is highly conserved between species. Together these data suggest that SGIP1 is an important and novel member of the group of neuronal molecules required for the regulation of energy homeostasis.
Introduction
THE HYPOTHALAMUS IS a critical regulator of energy balance, transducing information from both metabolic and humoral signals into adaptive changes in energy intake and expenditure (1, 2). In recent years, our understanding of key molecules involved in the control of food intake and body weight has grown considerably, resulting in the identification of novel obesity-inducing mutations. However, such mutations have been found in only a very small number of obese humans, the largest of these being associated with an estimated 5% of human obesity cases (3). The underlying genetic determinants of common human obesity, therefore, remain poorly understood.
These considerations suggest that additional genes involved in the hypothalamic control of energy homeostasis remain to be discovered. As a first step in identifying these potential genes, we used Psammomys obesus, a unique polygenic animal model of obesity and type 2 diabetes. Previous studies from our laboratory have demonstrated that P. obesus spontaneously develop a wide range of metabolic phenotypes that reflect the range of syndromes observed before (and including) the development of type 2 diabetes (4, 5, 6). To identify novel differentially expressed mRNA transcripts that may regulate this metabolic syndrome, we used differential display PCRs to conduct a large-scale screen of hypothalamic mRNA from lean and obese P. obesus as previously described (7, 8). Here we characterize a novel neuronal transcript, termed SGIP1 [Src homology 3-domain growth factor receptor-bound 2 (GRB2)-like (endophilin) interacting protein 1], which displayed increased mRNA content in the hypothalamus in obesity and after fasting. Furthermore, targeted reduction of hypothalamic SGIP1 mRNA inhibited food intake and decreased body weight, suggesting that this transcript plays an obligatory role in the regulation of energy homeostasis.
Materials and Methods
Experimental animals
Studies in P. obesus and Sprague Dawley rats were conducted in accordance with the regulations and guidelines outlined by the National Health and Medical Research Council of Australia and approved by the Deakin University Ethics Committee. Protocols for Ay/a mice were approved by the Institutional Animal Care and Use Committee of the University of Washington (Seattle, WA) and were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.
P. obesus
A colony of P. obesus is maintained at Deakin University (Geelong, Australia) using the San Poiley outbreeding method (9). Male Sprague Dawley rats (Animal Resource Center, Perth, Australia) were maintained at Deakin University. All animals were fed a diet of standard laboratory chow from which 63% of energy was derived from carbohydrate, 25% from protein, and 12% from fat (Barastoc, Victoria, Australia) and maintained in a temperature-controlled room (22 ± 1 C) with a 12-h light,12-h dark cycle. P. obesus was classified as either lean normal glucose tolerance (nGT), obese impaired glucose tolerance (IGT), or obese diabetic (D2M) according to circulating insulin and blood glucose values and body weight as previously described (4, 6). For differential display PCR (n = 2 per group) and real-time RT-PCR (n = 8 per group), 18-wk-old male lean nGT, obese IGT, and obese D2M P. obesus were separated into two treatment groups: fed and fasted (24 h). Groups were matched for body weight, blood glucose, and plasma insulin. Animals were killed by anesthetic overdose (pentobarbitone, 120 mg/kg) and the hypothalamus rapidly excised, snap frozen in liquid nitrogen, and stored at –80 C.
Agouti (Ay/a) mice
Male wild-type (C57BL6; n = 5) and lethal yellow agouti (Ay/a, n = 5) mice were fed ad libitum. Animals were killed by brief inhalation of CO2 and decapitation. The brain was removed, and a mediobasal hypothalamic wedge (defined caudally by the mamillary bodies, rostrally by the optic chiasm, laterally by the optic tract, and superiorly by the apex of the third ventricle) was dissected, snap frozen, and stored at –80 C for subsequent analysis.
Differential-display PCR
Total hypothalamic RNA was isolated using the RNAzol B method (Tel-Test, Friendswood, TX), treated with 1 U/μl of DNase I (Life Technologies Inc., Gaithersburg, MD), and reverse-transcribed using Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA). Differential-display PCR was performed on hypothalamic cDNA samples using an RNAimage mRNA differential display system (GenHunter, Nashville, TN). SGIP1 was amplified using the A-anchored primer (5'-AAGCT11A-3) and arbitrary primer 38 (5'-CTGCTGCTCCTCCTC-3').
Sequencing
The 5' end of the SGIP1 transcript isolated by differential-display PCR was ascertained using the FirstChoice RLM-RACE kit (Ambion, Austin, TX). PCR products were sequenced using the ABI PRISM BigDye terminator cycle sequencing ready reaction mix and a 373 automated fluorescent DNA sequencer (Applied Biosystems, Foster City, CA).
Real-time RT-PCR
Total hypothalamic RNA was isolated using TRIzol (Invitrogen, Carlsbad, CA) combined with RNeasy mini kit columns (QIAGEN, Chatsworth, CA). RNA quality and quantity were assessed using an Agilent 2100 bioanalyzer and the RNA 6000 Nano LabChip kit (Agilent Technologies, Waldbronn, Germany). Hypothalamic RNA was reverse transcribed using Superscript II RT system (Invitrogen). Oligonucleotide primers for SGIP1 (sense, 5'-TGAAGGCTTCCATAGGCAACA-3'; antisense, 5'-TGGAACGCCTGGGTCTTG-3') were designed using the Primer Express 2.0 software program (Applied Biosystems), and synthesized by Geneworks (Adelaide, Australia). PCR was performed using SYBR Green PCR master mix and an ABI PRISM 7700 sequence detector (Applied Biosystems).
Northern blot analysis
A cDNA probe complementary to SGIP1 mRNA was generated by PCR using primers: sense, 5'-CTTAGCTGGGACCTGGAA-3'; antisense, 5'-GACTCCATGCTTGGGTTAACT-3'. PCR was performed using AmpliTaq Gold master mix (Applied Biosystems) and P. obesus hypothalamic cDNA. Purified DNA probe was radioactively labeled with 32P-dATP (>3000 Ci/mmol; PerkinElmer, Melbourne, Australia) using the Prime-It II random primer labeling kit (Stratagene, La Jolla, CA). Unincorporated radioactivity was removed using the Qiaquick nucleotide removal kit (QIAGEN). SGIP1 DNA probe corresponding to 1 x 106 to 1 x 107 cpm was denatured and used to probe a multiple tissue Northern human RNA blot (BD Biosciences CLONTECH, Palo Alto, CA). The blot was exposed to a phosphor screen overnight and the image captured using the Molecular Dynamics PhosphorImager SI (Sunnyvale, CA).
Intracerebroventricular (ICV) antisense administration
For ICV studies, 16-wk-old male lean P. obesus or 10-wk-old male Sprague Dawley rats were used. A 23G cannula was implanted into the right lateral ventricle on d –7 (2.0 mm posterior; 2.0 mm ventral from the bregma, lowered 2.0 mm below the skull for P. obesus; or 1.5 mm lateral and 0.2 mm posterior from the bregma, and lowered 4.0 mm for Sprague Dawley rats). An Alzet miniosmotic pump (DURECT Corp., Cupertino, CA) was implanted sc into the scapular region and connected to the cannula at d 0. The pumps delivered SGIP1 antisense oligonucleotide (ODN; n = 6–8), jumbled ODN resuspended in sterile 0.9% saline (n = 6–8), or saline alone (n = 6–8) at a rate of 1 μl/h (24 μg/d) for 4 d. SGIP1 antisense ODN (5'-mUmCmAmGmUCCTTCCATmCmAmUmGmG-3') and jumbled ODN (5'-mCmGmCmAmCTTAGCTACmUmUmGmCmU-3'), where m indicates a 2'-O-methyl (2'OMe)-modified base, and all bonds were phosphorothioate linked and synthesized by Integrated DNA Technologies (Coralville, IA).
Energy expenditure measurements
Ten-week-old male Sprague Dawley rats had an ICV cannula implanted into the right lateral ventricle on d –10. At d –3, the rats were placed in a whole-body Oxymax equal flow indirect calorimeter (Columbus Instruments, Columbus, OH) for 4–6 h for acclimatization. At d –1 and again at d 3, the animals were placed in the calorimeter for 24 h during which oxygen consumption and carbon dioxide production were measured. Physical activity was measured using an Opto-Varimax miniinfrared animal activity monitor system (Columbus Instruments). At d 0, the rats were implanted with an Alzet pump containing either SGIP1 antisense ODN (n = 4) or jumbled ODN (n = 11) and connected to the ICV cannula as described. To control for the effects of weight loss, seven of the jumbled ODN-treated group were pair fed to the average daily food intake of the SGIP1 antisense group. Estimations of energy production rate, or total energy expenditure (TEE), from oxygen consumption and carbon dioxide production were calculated from previously described equations (10).
Yeast two-hybrid screen
The yeast two-hybrid screen was conducted using the ProQuest two-hybrid system (Life Technologies). Two partial fragments of SGIP1 DNA were cloned into the yeast bait vector pDBLeu and transformed into DH10B cells by electroporation. One fragment consisted of the central proline-rich region of SGIP1 (SGIP1-PR; amino acid 240–499), and the other consisted of the N terminal plus the proline-rich region (SGIP1-NP; amino acid 1–499). Constructs were transformed into the yeast strain MaV203, and the amount of 3-amino-1,2,4-triazole required for suppression of basal HIS3 expression was determined empirically as 25 mM. MaV203 cells harboring the SGIP1 constructs were used in a large-scale transformation with a human brain cDNA expression library (Life Technologies). The 5 x 106 to 1 x 107 transformants were plated onto selective media containing 25 mM 3-amino-1,2,4-triazole but lacking leucine, tryptophan, and histidine. Transformants that induced HIS3 expression were predicted to contain potential interacting proteins. Putative HIS-positive transformants were then tested for induction of two other reporters, URA3 and lacZ. The plasmids isolated from transformants positive for reporter gene expression were sequenced and retransformation performed to confirm the interaction.
Bioluminescence resonance energy transfer (BRET)
The entire human endophilin-3 coding sequence was amplified and cloned into the XhoI/BamHI sites of pGFP2-N2. To generate Rluc-SGIP1 fusions, the entire SGIP1 coding sequence and the SGIP1-NP was amplified and cloned into the XhoI/BamHI sites of pRluc-C2(h). For BRET assays, HEK-293 cells, grown in DMEM supplemented with 10% fetal bovine serum, were transiently transfected with the indicated plasmids in 6-well plates using Lipofectamine 2000 (Invitrogen). Twenty-four hours after transfection, cells were washed twice with PBS containing 0.1 g/liter calcium chloride, 0.1 g/liter magnesium chloride, 1 g/liter D-glucose, and 2 μg/ml aprotinin and then detached and resuspended to a final concentration of 2 x 106 cells/ml. Aliquots of 5 x 104 cells were distributed in a 96-well plate (White OptiPlate, PerkinElmer, Boston, MA), and DeepBlueC coelenterazine substrate was added to a final concentration of 5 μM. Emitted luminescence (410 nm) and fluorescence (515 nm) were detected immediately after addition of the substrate using a Fusion universal microplate analyzer (PerkinElmer). All values are expressed as a BRET ratio, which is defined as the [(emission at 515 nm – emission at 515 nm of nontransfected cells)/(emission at 410 nm – emission at 410 nm of nontransfected cells)].
Statistical methods
For comparisons involving three or more groups, data were assessed for normality by Kolmogorov-Smirnov test, and then one-way ANOVA was performed with the least significant differences or Games-Howell post hoc test for homogeneous or nonhomogeneous data, respectively. Comparisons of between-group mean values were performed using an unpaired Student’s t test for two-group comparisons or paired t test for comparison between two time points of the same group. For the analysis of energy expenditure because energy expenditure needs to be adjusted for weight, analysis of covariance was performed in which the change in TEE from baseline was the dependent variable, treatment group was the independent variable, and weight at baseline was the covariate. To analyze the change in body weight, ANOVA was used with the change in body weight from baseline as response and treatment group as the independent variable. Statistical analyses were performed using Statistical Package for the Social Sciences (SPSS, version 10.1; SPSS Inc., Fullerton, CA) or Statistical Analysis Software (SAS Institute Inc., Cary, NC). Statistical significance was defined as P < 0.05. All values are presented as mean ± SE.
Results
Discovery and characterization of SGIP1
SGIP1 (accession no. AY611625) was one of 27 novel transcripts isolated from approximately 9000 putative transcripts generated by more than 180 differential-display PCRs using hypothalamic mRNA from lean nGT, obese hyperglycemic (IGT), and obese D2M P. obesus (7). Real-time RT-PCR analysis of SGIP1 mRNA content confirmed increased hypothalamic SGIP1 gene expression in obese IGT and D2M animals, compared with lean nGT animals in the ad libitum-fed state (Fig. 1A). Linear relationships between SGIP1 gene expression and body weight (r = 0.674; P < 0.001) and percent body fat (r = 0.548; P < 0.01) were also observed (data not shown). To determine whether SGIP1 mRNA levels were also elevated in the hypothalamus of other mouse models of obesity and diabetes, we studied lethal yellow agouti (Ay/a) mice that develop obesity and diabetes due to a mutation that causes ectopic overexpression of agouti protein and subsequent chronic blockade of neuronal melanocortin receptors (11, 12). SGIP1 mRNA was increased approximately 4-fold in the hypothalamus of ad libitum-fed Ay/a mice relative to wild-type controls (Fig. 1B). Combined with data from P. obesus, these results suggest that increased hypothalamic SGIP1 gene expression may be a common feature in rodent models of obesity and diabetes.
To investigate tissue-specific distribution of SGIP1 gene expression, we measured SGIP1 mRNA in P. obesus brain, liver, adipose tissue, muscle, heart, and spleen using real-time RT-PCR. We found SGIP1 mRNA to be highly expressed in brain, with minimal levels also detected in adipose tissue and spleen (Fig. 1C). This pattern of SGIP1 gene expression is similar to that determined by Northern blot analysis of human tissues, which revealed a specific band of approximately 6 kb in brain RNA that was not detected in other tissues (Fig. 1D). Using in situ hybridization, we found SGIP1 mRNA to be expressed diffusely throughout the brain (data not shown).
A 6317-nucleotide SGIP1 mRNA sequence encoding an 827-amino acid protein was obtained from P. obesus hypothalamic cDNA using 5' rapid amplification of cDNA ends. The human ortholog of SGIP1 (reference sequence NM_032291), the hypothetical transcript DKFZp761D221, is 4572 nucleotides long and predicted to encode an 828-amino acid protein that is 94% identical with P. obesus SGIP1. The human SGIP1 gene is comprised of 27 exons and located on chromosome 1p31.3. SGIP1 is proline rich, with 13% of its amino acid content comprising proline residues that are concentrated in the central portion of the protein. This proline-rich region contains numerous potential SH3- and WW-domain binding sites.
Suppression of SGIP1 expression reduces food intake and body weight
Because hypothalamic SGIP1 expression is elevated in obese animals, we sought to determine (using antisense ODN technology) whether reduced hypothalamic expression of this transcript lowers food intake and body weight. Before ICV infusion into experimental animals, we tested several ODN sequences to verify successful suppression of SGIP1 mRNA in a cell-based system. A 2'OMe chimeric antisense ODN targeted to the initiation codon of SGIP1 consistently suppressed endogenous SGIP1 mRNA by 70–80% after transfection of GT1–7 mouse neuronal cells (supplemental Fig. 1, published as supplemental data on The Endocrine Society’s Journals Online Web site at http://endo.endojournals.org). Infusion of 2'OMe chimeric SGIP1 antisense ODN into the lateral ventricle of lean, male P. obesus at a rate of 24 μg/d for 4 d significantly decreased cumulative food intake, compared with both jumbled ODN (27.4 ± 4.4 vs. 39.0 ± 1.8 g, P = 0.015) and saline-treated animals (27.4 ± 4.4 vs. 46.3 ± 2.7 g, P < 0.001; Fig. 2A). This reduction of food intake was accompanied by a loss of body weight after SGIP1 antisense ODN treatment, compared with both jumbled ODN (–9.2 ± 1.5 vs. –4.0 ± 1.3 g, P = 0.02) and saline-treated controls (–9.2 ± 1.5 vs. –1.1 ± 1.3 g, P = 0.001) at d 4 (Fig. 2B). Whereas cumulative food intake was also slightly decreased by infusion of the jumbled ODN, compared with saline controls, significant differences in intake between these groups were detected only during the first 24 h of infusion.
To determine whether the weight-reducing effects of neuronal SGIP1 suppression can be generalized to other rodent species, we conducted a similar experiment in Sprague Dawley rats. Similar to the response of P. obesus, central SGIP1 antisense treatment reduced cumulative food intake, compared with either jumbled ODN (55.0 ± 5.3 vs. 101.5 ± 6.1 g, P < 0.001) or saline-treated control rats (55.0 ± 5.3 vs. 87.1 ± 4.1 g, P = 0.001) over the 4-d study (Fig. 2C), whereas food intake was not different between saline and jumbled ODN-treated rats. This reduction in food intake was again accompanied by a decrease in body weight in SGIP1 antisense-treated animals, compared with jumbled ODN (–36.6 ± 7.5 vs. 13.8 ± 7.6 g, P < 0.001) and saline-treated controls (–36.6 ± 7.5 vs. 1.8 ± 5.6 g, P < 0.001) after 4 d (Fig. 2D). To verify suppression of SGIP1 gene expression, hypothalamic SGIP1 mRNA levels were measured using real-time RT-PCR after 4 d of ICV infusion. Relative to saline-treated controls, hypothalamic SGIP1 mRNA was reduced by 37% by SGIP1 antisense treatment (P = 0.01, Fig. 2E). The effect of SGIP antisense ODNs in extrahypothalamic tissues is not known at this time, and it is possible that effects outside the hypothalamus may contribute to the phenotypes observed in this study. However, the suppression of SGIP1 mRNA in the hypothalamus, along with the well-known role of the hypothalamus in the regulation of energy balance, suggests a key role for SGIP1 in this area.
Collectively, these data suggest that endogenous SGIP1 protein provides a physiological stimulus for food intake. To further investigate this hypothesis, we measured the effect of fasting on SGIP1 mRNA levels in the hypothalamus of Sprague Dawley rats. Our finding that a 48-h fast increased SGIP1 mRNA content in these animals (Fig. 2F) supports the hypothesis that SGIP1 up-regulation contributes to the hypothalamic response to negative energy balance.
Effects of suppression of SGIP1 on energy expenditure
To determine whether increased energy expenditure contributes to weight loss induced by ICV SGIP1 antisense infusion, we used indirect calorimetry and measured physical activity in a separate group of Sprague Dawley rats. These measures were made for 24 h both before and 3 d after implantation of osmotic pumps for ICV infusion of antisense or jumbled ODN. To control for potential effects of reduced body weight and food intake on energy expenditure, a pair-fed group was included that received the jumbled ODN and was provided with food on a daily basis in an amount equal to the average daily food intake of SGIP1 antisense-treated animals. As observed previously suppression of SGIP1 reduced body weight significantly over the 4 d of treatment. Whereas pair-feeding also reduced body weight, ANOVA revealed a significantly greater weight loss among SGIP1 antisense-treated animals than pair-fed controls (Table 1). The reduction in body weight after SGIP1 antisense treatment therefore cannot be explained solely by reduced food intake. Consistent with this observation, SGIP1 antisense-treated animals failed to induce the appropriate reduction in TEE that normally occurs in rodents and humans during energy restriction (Table 1) (13, 14). Thus, although a significant reduction in TEE from baseline was observed after SGIP1 antisense treatment, analysis of covariance revealed that the reduction in TEE in pair-fed animals was significantly greater than the reduction in TEE of SGIP1 antisense-treated animals (Table 1). Our findings therefore suggest that in addition to inhibiting food intake, suppression of SGIP1 in Sprague Dawley rats prevents the reduction of energy expenditure that normally accompanies negative energy balance.
We then determined whether this failure to reduce TEE could be attributed to alterations in either metabolic heat production, a marker of metabolic rate, or physical activity. Whereas there were no differences in metabolic heat production between ad libitum-fed jumbled ODN controls and SGIP1 antisense-treated animals (despite a pronounced difference in body weight), pair-fed jumbled ODN-treated rats demonstrated reduced heat production, particularly at the end of the dark phase and during the majority of the light phase (Fig. 3A). These data suggest that SGIP1 antisense-treated rats exhibited an inappropriately elevated metabolic rate, compared with pair-fed animals, consistent with the estimates of TEE. Whereas jumbled ODN-treated ad libitum-fed animals exhibited the expected increase of physical activity during the dark phase (P < 0.03, Fig. 3B), this response was attenuated in both the SGIP1 antisense-treated and jumbled ODN-treated pair-fed rats. These observations suggest that decreased activity of SGIP1 antisense-treated animals was likely due, at least in part, to reduced food intake and body weight because activity also tended to be reduced in the pair-fed rats. Furthermore, these data indicate that whereas SGIP1 antisense treatment attenuated the effect of weight loss to reduce metabolic heat production and energy expenditure, it did not block its effect to reduce physical activity.
SGIP1 molecular pathway of action
To identify potential interacting partners of SGIP1, and thereby a possible molecular pathway whereby SGIP1 regulates energy balance, we performed a yeast two-hybrid screen of a human brain cDNA library using two truncated fragments of SGIP1 as bait. The two bait fragments used comprised the SGIP1-NP and SGIP1-PR (Fig. 4A). Both the SGIP1-NP and SGIP1-PR bait constructs strongly interacted with human endophilin-3 and the highly similar family member endophilin-1, most likely via the central proline-rich region common to both baits. The interaction between endophilin-3 and SGIP1 was then tested by BRET experiments in HEK-293 liver cells. SGIP1-NP fused to a luciferase reporter strongly interacted with endophilin-3 fused to green fluorescent protein (GFP) indicative of interaction (Fig. 4B). Likewise, full-length SGIP1 also resulted in a significantly increased BRET ratio when cotransfected with endophilin-3. No interaction was observed when SGIP1 constructs were cotransfected with a GFP construct containing ubiquilin or when endophilin-3 constructs were cotransfected with a vector containing only the luciferase reporter (Fig. 4B). The yeast two-hybrid screen data combined with the BRET experiments in mammalian cells confirm an interaction between SGIP1 and endophilin-3.
Discussion
In this report, we present the first experimental evidence for the existence of DKFZp761D221/SGIP1 and identify it as a novel, proline-rich protein predominantly expressed in brain and strongly implicated in the central regulation of energy balance. We found that food intake and body weight were reduced by suppression of hypothalamic SGIP1 mRNA and that hypothalamic expression of this transcript is increased both in rodent models of obesity and in response to fasting. These data collectively suggest a physiological role for this protein in hypothalamic neuronal systems that promote positive energy balance and weight gain. Inhibition of neuronal SGIP1 expression or signaling, therefore, constitutes a novel and potentially useful strategy for obesity treatment.
ICV infusion of 2'OMe-modified antisense ODN (15) targeted to the initiation codon of SGIP1 specifically suppressed endogenous SGIP1 mRNA in rat hypothalamus and resulted in approximately 40% inhibition of food intake and 5–8% reduction in body weight over the 4-d study period. Similar results were observed in P. obesus and are comparable in magnitude with reductions in food intake previously reported in rats after ICV antisense suppression of neuropeptide Y (NPY) (16, 17) or the NPY Y5 receptor (18, 19). To control for nonspecific effects of centrally administered ODN, we included control animals that received ICV infusion of a jumbled sequence ODN of the same oligonucleotide composition and chemistry as the SGIP1 antisense ODN. The lack of effect of the jumbled ODN on food intake and body weight suggests that toxicity or malaise associated with the antisense treatment is unlikely to explain the decreased food intake observed after SGIP1 suppression.
Two observations support the conclusion that in addition to decreasing food intake, depletion of hypothalamic SGIP1 mRNA reduced body weight by inducing an inappropriately high metabolic rate. First, weight loss was reduced in control rats pair fed to the intake of SGIP1 antisense-treated rats. In addition, we found that dietary energy restriction reduced TEE by 20% (P = 0.02), consistent with previous studies in rats (13, 20, 21), primates (22), and humans (14, 23). By comparison, SGIP1 antisense-treated animals failed to reduce their TEE relative to free-feeding, jumbled ODN-treated rats, despite their decreased food intake and body weight and despite showing an appropriate reduction in physical activity. The combination of reduced food intake with the inability to appropriately reduce energy expenditure may therefore explain the potent reduction in body weight seen in SGIP1 antisense-treated rats. Whether the decrease in body weight in SGIP1 antisense-treated animals is due to reduced lean body mass, fat mass, or both is not known and is an important question for future experiments.
Although SGIP1 is an intracellular protein rather than a neuropeptide, increased SGIP1 mRNA expression in the hypothalamus of obese P. obesus, and in fasted Sprague Dawley rats, is a feature shared by orexigenic neuropeptides including NPY, agouti-related protein, and melanin-concentrating hormone (24, 25, 26). Interestingly, SGIP1 mRNA was also increased in the hypothalamus of lethal yellow Ay/a mice that develop obesity as a result of chronic antagonism of hypothalamic melanocortin receptors by ectopically overexpressed agouti protein (11). Increased expression of SGIP1 may therefore contribute to the hyperphagia and obesity of Ay/a mice, and further studies are warranted to investigate this hypothesis.
Our finding that SGIP1 is a large proline-rich protein has implications for its intracellular function. Proline-rich regions are important for protein-protein interactions and are targets for proteins containing SH3 and WW domains (27). We used two truncated proline-rich fragments of SGIP1 as bait in a yeast two-hybrid screen. This analysis, subsequently confirmed by BRET experiments in mammalian cells, revealed a strong interaction between SGIP1 and endophilins, proteins initially isolated as SH3 domain-containing proteins (28, 29). Endophilins are important regulators of clathrin mediated endocytosis (30) and synaptic vesicle recycling (31, 32). They participate in receptor internalization and vesicle recycling by acting as part of a larger complex of proteins including synaptojanin (28, 33), amphiphysin (34), and dynamin (28). The majority of these studies have reported on endophilin-1; however, a recent report has shown an important role for endophilin-3 in the negative regulation of clathrin-mediated endocytosis in brain neurons and may regulate internalization of neurotransmitter receptors (35).
In conclusion, we provide evidence to show that SGIP1 is a novel brain-specific component of the neuronal molecular machinery that regulates energy homeostasis. Although the mechanism whereby SGIP1 favors increased food intake and body weight awaits further study, our data support a model in which SGIP1 functions as an endocytic protein that affects signaling by receptors in neuronal systems involved in energy homeostasis via its interaction with endophilin.
Acknowledgments
The authors thank Dr. Eric Ravussin and Haiyan Yang for assistance with interpretation and analysis of indirect calorimetry measurements and Drs. David Segal and Andrew Butler for helpful discussions and reading of the manuscript.
Footnotes
This work was supported by funding from ChemGenex Pharmaceuticals Ltd.; National Institutes of Health Grants DK52989, DK12829, DK68304, NS32273 (to M.W.S.); and the Clinical Nutrition Research Unit at the University of Washington.
Abbreviations: BRET, Bioluminescence resonance energy transfer; D2M, diabetic; GFP, green fluorescent protein; GRB2, growth factor receptor-bound 2; ICV, intracerebroventricular; IGT, impaired glucose tolerance; nGT, normal glucose tolerance; NPY, neuropeptide Y; ODN, oligonucleotide; 2'OMe, 2'-O-methyl; SGIP1, Src homology 3-domain GRB2-like (endophilin) interacting protein 1; SGIP1-NP, N terminal plus the proline-rich region of SGIP1; SGIP1-PR, proline-rich region of SGIP1; TEE, total energy expenditure.
References
Barsh GS, Schwartz MW 2002 Genetic approaches to studying energy balance: perception and integration. Nat Rev Genet 3:589–600
Schwartz MW, Woods SC, Porte Jr D, Seeley RJ, Baskin DG 2000 Central nervous system control of food intake. Nature 404:661–671
Farooqi IS, Keogh JM, Yeo GS, Lank EJ, Cheetham T, O’Rahilly S 2003 Clinical spectrum of obesity and mutations in the melanocortin 4 receptor gene. N Engl J Med 348:1085–1095
Barnett M, Collier GR, Collier FM, Zimmet P, O’Dea K 1994 A cross-sectional and short-term longitudinal characterisation of NIDDM in Psammomys obesus. Diabetologia 37:671–676
Collier GR, Walder K, Lewandowski P, Sanigorski A, Zimmet P 1997 Leptin and the development of obesity and diabetes in Psammomys obesus. Obes Res 5:455–458
Walder KR, Fahey RP, Morton GJ, Zimmet PZ, Collier GR 2000 Characterization of obesity phenotypes in Psammomys obesus (Israeli sand rats). Int J Exp Diabetes Res 1:177–184
Trevaskis J, McMillan JS, Windmill K, Walder K, Collier GR 2004 Identification of hypothalamic genes implicated in the development of obesity in Psammomys obesus using differential display PCR. Comp Biochem Physiol B Biochem Mol Biol 137:65–73
Collier GR, McMillan JS, Windmill K, Walder K, Tenne-Brown J, de Silva A, Trevaskis J, Jones S, Morton GJ, Lee S, Augert G, Civitarese A, Zimmet PZ 2000 Beacon: a novel gene involved in the regulation of energy balance. Diabetes 49:1766–1771
Poiley S 1960 A systematic method of breeder rotation for non-inbred laboratory animal colonies. Proc Animal Care Panel 10:159–166
Ferrannini E 1988 The theoretical bases of indirect calorimetry: a review. Metabolism 37:287–301
Klebig ML, Wilkinson JE, Geisler JG, Woychik RP 1995 Ectopic expression of the agouti gene in transgenic mice causes obesity, features of type II diabetes, and yellow fur. Proc Natl Acad Sci USA 92:4728–4732
Lu D, Willard D, Patel IR, Kadwell S, Overton L, Kost T, Luther M, Chen W, Woychik RP, Wilkison WO, Cone RD 1994 Agouti protein is an antagonist of the melanocyte-stimulating-hormone receptor. Nature 371:799–802
Even PC, Nicolaidis S 1993 Adaptive changes in energy expenditure during mild and severe feed restriction in the rat. Br J Nutr 70:421–431
Ravussin E, Burnand B, Schutz Y, Jequier E 1985 Energy expenditure before and during energy restriction in obese patients. Am J Clin Nutr 41:753–759
Dean NM, Butler M, Monia M, Manoharan M 2001 Pharmacology of 2'-O-(2-methoxy)ethyl-modified antisense oligonucleotides. In: Crooke S, ed. Antisense drug technology: principles, strategies and applications. New York: Marcel Dekker, Inc.; 319–338
Akabayashi A, Wahlestedt C, Alexander JT, Leibowitz SF 1994 Specific inhibition of endogenous neuropeptide Y synthesis in arcuate nucleus by antisense oligonucleotides suppresses feeding behavior and insulin secretion. Brain Res Mol Brain Res 21:55–61
Hulsey MG, Pless CM, White BD, Martin RJ 1995 ICV administration of anti-NPY antisense oligonucleotide: effects on feeding behavior, body weight, peptide content and peptide release. Regul Pept 59:207–214
Tang-Christensen M, Kristensen P, Stidsen CE, Brand CL, Larsen PJ 1998 Central administration of Y5 receptor antisense decreases spontaneous food intake and attenuates feeding in response to exogenous neuropeptide Y. J Endocrinol 159:307–312
Schaffhauser AO, Stricker-Krongrad A, Brunner L, Cumin F, Gerald C, Whitebread S, Criscione L, Hofbauer KG 1997 Inhibition of food intake by neuropeptide Y Y5 receptor antisense oligodeoxynucleotides. Diabetes 46:1792–1798
Gonzales-Pacheco DM, Buss WC, Koehler KM, Woodside WF, Alpert SS 1993 Energy restriction reduces metabolic rate in adult male Fisher-344 rats. J Nutr 123:90–97
Keesey RE, Corbett SW 1990 Adjustments in daily energy expenditure to caloric restriction and weight loss by adult obese and lean Zucker rats. Int J Obes 14:1079–1084
DeLany JP, Hansen BC, Bodkin NL, Hannah J, Bray GA 1999 Long-term calorie restriction reduces energy expenditure in aging monkeys. J Gerontol A Biol Sci Med Sci 54:B5–B11; discussion B12–B13
Hirsch J, Hudgins LC, Leibel RL, Rosenbaum M 1998 Diet composition and energy balance in humans. Am J Clin Nutr 67:551S–555S
Mizuno TM, Makimura H, Silverstein J, Roberts JL, Lopingco T, Mobbs CV 1999 Fasting regulates hypothalamic neuropeptide Y, agouti-related peptide, and proopiomelanocortin in diabetic mice independent of changes in leptin or insulin. Endocrinology 140:4551–4557
Sahu A, Kalra PS, Kalra SP 1988 Food deprivation and ingestion induce reciprocal changes in neuropeptide Y concentrations in the paraventricular nucleus. Peptides 9:83–86
Tritos NA, Mastaitis JW, Kokkotou E, Maratos-Flier E 2001 Characterization of melanin concentrating hormone and preproorexin expression in the murine hypothalamus. Brain Res 895:160–166
Zarrinpar A, Bhattacharyya RP, Lim WA 2003 The structure and function of proline recognition domains. Sci STKE 2003:RE8
Ringstad N, Nemoto Y, De Camilli P 1997 The SH3p4/Sh3p8/SH3p13 protein family: binding partners for synaptojanin and dynamin via a Grb2-like Src homology 3 domain. Proc Natl Acad Sci USA 94:8569–8574
Giachino C, Lantelme E, Lanzetti L, Saccone S, Bella Valle G, Migone N 1997 A novel SH3-containing human gene family preferentially expressed in the central nervous system. Genomics 41:427–434
Schuske KR, Richmond JE, Matthies DS, Davis WS, Runz S, Rube DA, van der Bliek AM, Jorgensen EM 2003 Endophilin is required for synaptic vesicle endocytosis by localizing synaptojanin. Neuron 40:749–762
Ringstad N, Gad H, Low P, Di Paolo G, Brodin L, Shupliakov O, De Camilli P 1999 Endophilin/SH3p4 is required for the transition from early to late stages in clathrin-mediated synaptic vesicle endocytosis. Neuron 24:143–154
Schmidt A, Wolde M, Thiele C, Fest W, Kratzin H, Podtelejnikov AV, Witke W, Huttner WB, Soling HD 1999 Endophilin I mediates synaptic vesicle formation by transfer of arachidonate to lysophosphatidic acid. Nature 401:133–141
Verstreken P, Koh TW, Schulze KL, Zhai RG, Hiesinger PR, Zhou Y, Mehta SQ, Cao Y, Roos J, Bellen HJ 2003 Synaptojanin is recruited by endophilin to promote synaptic vesicle uncoating. Neuron 40:733–748
Micheva KD, Ramjaun AR, Kay BK, McPherson PS 1997 SH3 domain-dependent interactions of endophilin with amphiphysin. FEBS Lett 414:308–312
Sugiura H, Iwata K, Matsuoka M, Hayashi H, Takemiya T, Yasuda S, Ichikawa M, Yamauchi T, Mehlen P, Haga T, Yamagata K 2004 Inhibitory role of endophilin 3 in receptor-mediated endocytosis. J Biol Chem 279:23343–23348(James Trevaskis, Ken Wald)