The regulation and activation of ciliary neurotrophic factor signaling proteins in adipocytes.

Ciliary neurotrophic factor (CNTF) is primarily known for its roles as a lesion factor released by the ruptured glial cells that prevent neuronal degeneration. However, CNTF has also been shown to cause weight loss in a variety of rodent models of obesity/type II diabetes, whereas a modified form also causes weight loss in humans. CNTF administration can correct or improve hyperinsulinemia, hyperphagia, and hyperlipidemia associated with these models of obesity. In order to investigate the effects of CNTF on fat cells, we examined the expression of CNTF receptor complex proteins (LIFR, gp130, and CNTFRalpha) during adipocyte differentiation and the effects of CNTF on STAT, Akt, and MAPK activation. We also examined the ability of CNTF to regulate the expression of adipocyte transcription factors and other adipogenic proteins. Our studies clearly demonstrate that the expression of two of the three CNTF receptor complex components, CNTFRalpha and LIFR, decreases during adipocyte differentiation. In contrast, gp130 expression is relatively unaffected by differentiation. In addition, preadipocytes are more sensitive to CNTF treatment than adipocytes, as judged by both STAT 3 and Akt activation. Despite decreased levels of CNTFRalpha expression in fully differentiated 3T3-L1 adipocytes, CNTF treatment of these cells resulted in a time-dependent activation of STAT 3. Chronic treatment of adipocytes resulted in a substantial decrease in fatty-acid synthase and a notable decline in SREBP-1 levels but had no effect on the expression of peroxisome proliferator-activated receptor gamma, acrp30, adipocyte-expressed STAT proteins, or C/EBPalpha. However, CNTF resulted in a significant increase in IRS-1 expression. CNTFRalpha receptor expression was substantially induced in the fat pads of four rodent models of obesity/type II diabetes as compared with lean littermates. Moreover, we demonstrated that CNTF can activate STAT 3 in adipose tissue and skeletal muscle in vivo. In summary, CNTF affects adipocyte gene expression, and the specific receptor for this cytokine is induced in rodent models of obesity/type II diabetes.

Ciliary neurotrophic factor (CNTF) 1 was originally characterized as a trophic factor that supports the survival of embryonic chick ciliary ganglion neurons in vitro (1,2). However, subsequent cloning and sequencing of CNTF revealed that it is unrelated to neurotrophins but is a member of the gp130 cytokine family along with interleukin-6, interleukin-11, LIF, OSM, leptin and CT-1 (3)(4)(5). The actions of CNTF are mediated, in part, by a CNTF-specific receptor (CNTFR␣) that has homology to the interleukin-6R␣ (6). Upon translation, the C terminus of CNTFR␣ is cleaved. Mature CNTFR␣ has no transmembrane or cytosolic domains and is found on the outer surface of the cell membrane where it is attached by a glycosylphosphatidylinositol linkage sensitive to phosphatidylinositol-phospholipase C treatment (7). Initially, CNTFR␣ was described as being distributed predominantly within neural tissues (7) but has since been reported in skeletal muscle, adrenal gland, sciatic nerve, skin, kidney, and testes (8). CNTFR␣ can be cleaved from the cell surface and exist and act in a soluble form. The soluble CNTFR␣ has been detected in the serum and the cerebrospinal fluid and has been shown to initiate signaling in cells not responsive to CNTF alone (6,9). Mice lacking CNTF develop normally and appear to have no visible defects well into adulthood, when they develop minor loss of motor neurons (10). However, mice lacking CNTFR␣ tend to have severe motor neuron defects and die perinatally because they fail to initiate feeding behaviors (11). CNTF signaling is initiated when CNTF binds CNTFR␣, either in its soluble or membrane-bound form (12). Once a CNTF⅐CNTFR␣ complex is formed, two of these heterodimers come together and recruit a gp130 transducer protein, followed by a subsequent recruitment of LIFR protein. The resulting receptor complex is a hexamer of CNTF, CNTFR␣, gp130, and LIFR in a 2:2:1:1 ratio, respectively (13). Within this complex, CNTF and CNTFR␣ make direct contacts with all the complex components (7,12). CNTF⅐CNTFR␣ is considered to be a low affinity binding complex until further bound to gp130 and LIFR (14). Aside from this hexameric, high affinity binding complex, CNTF can bind its receptors and can induce signaling in the absence of CNTFR␣, solely by binding to a gp130:LIFR dimeric receptor (15,16).
Although CNTF was first identified as a trophic factor in the ciliary ganglion, it was later found to act on other motor neuron populations (17). Hence, it was evaluated as a therapeutic tool in patients suffering from motor neuron diseases (18). Interestingly, during these trials, CNTF administration resulted in unexpected weight loss (19). Additional studies showed that, like leptin, CNTF can activate the same signaling molecules and that CNTFR␣ is co-localized with ObR in the hypothalamic nuclei involved in the regulation of feeding (20). CNTF can also cross the blood-brain barrier in a manner similar to leptin (21). CNTF treatment of leptin-deficient ob/ob mice was found to reduce adiposity, hyperphagia, and hyperinsulinemia associated with this genotype. Leptin administration had the same effect in these animals. However, unlike leptin, CNTF also corrected obesity-related phenotypes in leptin-resistant, ObRdeficient, db/db mice and in mice with diet-induced obesity that are partially resistant to leptin (22). CNTF and its synthetic analog, Axokine, have also been found to suppress NPY gene expression (23) and pCREB in the feeding-relevant brain sites (22). The weight loss caused by CNTF administration is due to the preferential loss of fat (24). It is believed to occur by resetting the hypothalamic weight set point, such that cessation of CNTF treatment does not result in overeating and rebound weight gain (22). Unlike cachectic cytokines, the appetite diminution during CNTF treatment does not appear to be due to stress, inflammatory responses, nausea, or conditioned taste aversion but is possibly due to the modification of NPYergic signaling (22,25).
In this study, we examined the regulation and activation of STATs and proteins by CNTF in adipocytes. The objective of this project was to determine whether CNTF, a cytokine known to result in weight loss, could have effects on peripheral tissues such as white adipose tissue. Our results clearly demonstrate that two of the three CNTF receptors are down-regulated during the adipogenesis of 3T3-L1 cells. However, CNTF administration results in the activation of STAT 3 in both cultured 3T3-L1 adipocytes and in rodent adipose tissue. Also this study provides the first evidence that CNTFR␣ is expressed in adipose tissue and that the expression of this receptor is regulated in four rodent models of obesity/type II diabetes. We also observed that CNTF treatment did not effect the expression of key adipogenic transcription factors such as PPAR␥ and C/EBP␣ but did result in a decrease of fatty-acid synthase (FAS) expression. Also, unlike other cachectic cytokines such as tumor necrosis factor-␣, chronic CNTF treatment did not result in the development of insulin resistance in cultured adipocytes. Moreover, acute CNTF administration resulted in increased GLUT4 expression, whereas chronic CNTF treatment resulted in a substantial increase in IRS-1 expression. Also, acute CNTF treatment resulted in an increase in insulin-induced IRS-1 and Akt activation. In summary, the results of this study demonstrate that both cultured and native adipocytes, as well as skeletal muscle, are responsive to CNTF and that this cytokine may act as an insulin-sensitizer in cultured adipocytes. These studies support our hypothesis that the ability of CNTF to result in weight loss is not solely mediated by the central nervous system.

EXPERIMENTAL PROCEDURES
Materials-Dulbecco's modified Eagle's media (DMEM) was purchased from Invitrogen. Bovine and fetal bovine serum (FBS) were obtained from Sigma and Invitrogen, respectively. Rat CNTF was purchased from Calbiochem. The non-phospho-STAT antibodies were monoclonal IgGs purchased from Transduction Laboratories or polyclonal IgGs from Santa Cruz Biotechnology. A highly phospho-specific polyclonal antibody for STAT 3 (Tyr 705 ) was purchased from BD Biosciences. LIFR and gp130 were rabbit polyclonals from Santa Cruz Biotechnology. PPAR␥ was a mouse monoclonal from Santa Cruz Biotechnology. SREBP-1 and ERK1/ERK2 were rabbit polyclonals from Santa Cruz Biotechnology. Active ERK antibody was a rabbit polyclonal from Promega. C/EBP␣ was a rabbit polyclonal from Dr. Ormond Mac-Dougald (Ann Arbor, MI), and GLUT 4 was a rabbit polyclonal from Dr. Paul Pilch (Boston). CNTFR␣ was a mouse monoclonal purchased from BD Biosciences. IRS-1 polyclonal was a polyclonal obtained from Upstate Biotechology, Inc., and the phospho-specific IRS-1 polyclonal was from BIOSOURCE International. PGNaseF was obtained from New England Biolabs.
Cell Culture-Murine 3T3-L1 preadipocytes were plated and grown to 2 days post-confluence in DMEM with 10% bovine serum. Medium was changed every 48 h. Cells were induced to differentiate by changing the medium to DMEM containing 10% fetal bovine serum, 0.5 mM 3-isobutyl-1-methylxanthine, 1 M dexamethasone, and 1.7 M insulin. After 48 h this medium was replaced with DMEM supplemented with 10% FBS, and cells were maintained in this medium until utilized for experimentation.
Preparation of Whole Cell Extracts-Monolayers of 3T3-L1 preadipocytes or adipocytes were rinsed with phosphate-buffered saline and then harvested in a non-denaturing buffer containing 150 mM NaCl, 10 mM Tris, pH 7.4, 1 mM EGTA, 1 mM EDTA, 1% Triton X-100, 0.5% Nonidet P-40, 1 M phenylmethylsulfonyl fluoride, 1 M pepstatin, 50 trypsin inhibitory milliunits of aprotinin, 10 M leupeptin, and 2 mM sodium vanadate. Samples were extracted for 30 min on ice and centrifuged at 15,000 rpm at 4°C for 15 min. Supernatants containing whole cell extracts were analyzed for protein content using a BCA kit (Pierce) according to the manufacturer's instructions.
Preparation of Nuclear/Cytosolic Extracts-Cell monolayers were rinsed with phosphate-buffered saline and then harvested in a nuclear homogenization buffer (NHB) containing 20 mM Tris, pH 7.4, 10 mM NaCl, and 3 mM MgCl 2 . Nonidet P-40 was added to a final concentration of 0.15%, and cells were homogenized with 16 strokes in a Dounce homogenizer. The homogenates were centrifuged at 1500 rpm for 5 min. Supernatants were saved as cytosolic extract, and the nuclear pellets were resuspended in 0.5 volume of NHB and were centrifuged as before. The pellet of intact nuclei was resuspended again in 0.5 of the original volume of NHB and centrifuged again. A small portion of the nuclei was used for trypan blue staining to examine the integrity of the nuclei. The majority of the pellet (intact nuclei) was resuspended in an extraction buffer containing 20 mM HEPES, pH 7.9, 420 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, and 25% glycerol. Nuclei were extracted for 30 min on ice and then placed at room temperature for 10 min. Two hundred units of DNase I was added to each sample, and tubes were inverted and incubated an additional 10 min at room temperature. Finally, the sample was subjected to centrifugation at 15,000 rpm at 4°C for 30 min. Supernatants containing nuclear extracts were analyzed for protein content.
Gel Electrophoresis and Immunoblotting-Proteins were separated in 5, 7.5, 10, or 12% polyacrylamide (acrylamide from National Diagnostics) gels containing SDS according to Laemmli (26) and transferred to nitrocellulose (Bio-Rad) in 25 mM Tris, 192 mM glycine, and 20% methanol. Following transfer, the membrane was blocked in 4% milk for 1 h at room temperature. Results were visualized with horseradish peroxidase-conjugated secondary antibodies (Sigma) and enhanced chemiluminescence (Pierce). Following chemiluminescence, some results were quantified by scanning film, and densitometry analysis was performed with associated analytical software (Biomax, Eastman Kodak Co.).
Determination of 2-Deoxyglucose-The assay of 2-[ 3 H]deoxyglucose was performed as described previously (27). Prior to the assay, fully differentiated 3T3-L1 adipocytes were serum-deprived for 4 h. Uptake measurements were performed in triplicate under conditions where hexose uptake was linear, and the results were corrected for nonspecific uptake, and absorption was determined by 2-[ 3 H]deoxyglucose uptake in the presence of 5 M cytochalasin B (Sigma). Nonspecific uptake and absorption were always less than 10% of the total uptake.
Animals and Adipose Tissue Isolation-Seven-week-old ob/ϩ and ob/ob mice were purchased from The Jackson Laboratories. Eight-weekold fa/ϩ and fa/fa rats were purchased from Harlan. C57Bl/6J mice were obtained from The Jackson Laboratories at 3-5 weeks of age, and upon receipt were placed on a high fat/high sucrose diet (Research Diets 12331, Surwit diet) or a low fat/high sucrose diet (Research Diets 12329). In each experiment, at least five animals were used for each condition. Twelve-week-old transgenic mice expressing agouti under the control of the ␤-actin promoter were obtained from a colony at the Pennington Biomedical Research Center. Rodents were euthanized by cervical dislocation, and adipose tissue was quickly removed, weighed, and frozen in liquid nitrogen. Frozen fat pads were homogenized in a buffer containing 150 mM NaCl, 10 mM Tris, pH 7.4, 1 mM EGTA, 1 mM EDTA, 1% Triton X-100, 0.5% Nonidet P-40, 1 M phenylmethylsulfonyl fluoride, 1 M pepstatin, 50 trypsin inhibitory milliunits of aprotinin, and 10 M leupeptin, and 2 mM sodium vanadate. Homogenates were centrifuged for 10 min at 5,000 rpm to remove any debris and insoluble material and then analyzed for protein content. All animal studies were carried out with protocols that were reviewed and approved by the institutional animal care and use committee. Adipocytes were isolated from the epididymal fat pads of male C57Bl/6J mice (35g) by collagenase digestion.

RESULTS
The sensitivity of 3T3-L1 cells to cytokine treatment was examined by treating undifferentiated preadipocytes and fully differentiated 3T3-L1 adipocytes with an acute treatment of 0.8 nM CNTF or 0.8 nM LIF. As shown in Fig. 1, immunoblotting of whole cell extracts demonstrated that both preadipocytes and adipocytes express STAT 3. Treatment of both preadipocytes and adipocytes with LIF or CNTF resulted in the rapid activation of STAT 3, as evident by increased tyrosine phosphorylation. However, treatment of preadipocytes resulted in a greater stimulation of STAT 3 activation, relative to adipocytes, despite equivalent expression of STAT 3 protein. In addition, CNTF and LIF treatment caused a robust activation of Akt in preadipocytes, whereas the same treatment of adipocytes did not result in a detectable activation of Akt. The expression of two CNTF receptor complex proteins, LIFR and gp130, was also examined. LIFR was expressed at a substantially higher level in preadipocytes than in adipocytes, whereas the expression of gp130 protein was not differentially expressed in these two cell types.
The expression of CNTF receptor complex proteins was also examined during a time course of adipocyte differentiation. As shown in Fig. 2, the expression of CNTFR␣ protein decreases notably after 15 min of induction of differentiation, and this lower level of expression is maintained for 48 h. However, there were no detectable levels of CNTFR␣ 72 h after the initiation of differentiation in whole cell extracts. Yet we did observe the presence of CNTFR␣ in the media at 72, 96, and 120 h at lower levels (data not shown). As indicated in Fig. 1, the expression of LIFR decreased during adipogenesis, and there was a slight modulation of gp130 expression. The expression of STAT 5A is known to be induced during adipocyte differentiation and is shown as a positive control for adipogenesis.
Because the expression of two CNTF receptor complex proteins was reduced during adipogenesis, we wanted to determine whether these proteins were expressed in adipose tissue and to compare the expression levels to other tissues. Whole cell extracts were isolated from the various tissues indicated in Fig. 3. Western blot analysis revealed lung, stomach, epididymal fat, spleen, heart, brain, testes, and skeletal muscle as tissues expressing both CNTFR␣ and LIFR. All of these tissues had comparable receptor expression levels, except for the brain, which had significantly higher levels of CNTFR␣ expression. Also the molecular weight of CNTFR␣ in stomach and brain was greater than in other tissues. In agreement with our earlier observations (Fig. 2), the expression of CNTFR␣ was abundant in preadipocytes and undetectable in 3T3-L1 adipocytes. We also observed that the expression of CNTFR␣ was upregulated in the epididymal fat pad of an obese Zucker rat as compared with a lean littermate.
To determine whether the altered mobility of CNTFR␣ was due to glycosylation, tissue extracts were incubated with PN-GaseF. As shown in Fig. 4, treatment with PNGaseF resulted in the deglycosylation of CNTFR␣ and LIFR. In particular, the CNTFR␣ bands of larger molecular weights (brain and stomach) co-migrated with the CNTFR␣ from other tissues following digestion, indicating that the size difference between CNTFR␣ in these tissues was due to different glycosylation patterns. Also all LIFR bands migrated at the same molecular weight following PNGaseF treatment.
Although our results demonstrate that fully differentiated 3T3-L1 adipocytes do not express CNTFR␣, it has been demonstrated previously (15,16) that CNTF can signal via gp130 and LIFR in the absence of CNTFR␣. Therefore, we examined the ability of CNTF to activate STATs in a time-dependent manner in 3T3-L1 adipocytes. Serum-deprived fully differentiated 3T3-L1 adipocytes were exposed to CNTF and examined over an 8-h period. Cells were harvested at the times indicated at the top of Fig. 5 and fractionated into cytosolic and nuclear extracts. As shown in Fig. 5A, CNTF administration to 3T3-L1 adipocytes resulted in the nuclear translocation of STAT 3. STAT 3 was present in the nucleus after a 10-or 30-min treatment with CNTF, and the amount of STAT 3 nuclear protein was decreased after a 1-h treatment. After 2 h, there was little STAT 3 present in the nucleus. CNTF treatment did not result in the activation/nuclear translocation of STAT 1 or FIG. 1. The effects of acute CNTF or LIF treatment on 3T3-L1 preadipocytes and adipocytes. Whole cell extracts were prepared from confluent undifferentiated preadipocytes and from fully differentiated 3T3-L1 adipocytes following a 15-min treatment with CNTF (0.8 nM) or LIF (0.8 nM). Whole cell extracts were prepared, and 75 g of each extract was separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis. The detection system was horseradish peroxidase-conjugated secondary antibodies and enhanced chemiluminescence. This is a representative experiment independently performed three times.
FIG. 2. The expression of CNTF receptor complex proteins during adipocyte differentiation. Whole cell extracts were prepared from 3T3-L1 cells at various times following the induction of differentiation. Cells were induced to differentiate at 2 days post-confluence with the addition of a differentiation mixture containing 10% FBS, 0.5 mM 3-isobutyl-1-methylxanthine, 1.0 M dexamethasone, and 1.7 M insulin. After 48 h this medium was replaced with DMEM supplemented with 10% FBS, and cells were maintained in this medium until utilized for experimentation. One hundred g of each extract was separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis. Samples were processed, and results were visualized as described in Fig. 1 legend. This is a representative experiment independently performed three times.
STAT 5B, indicating the specificity of the response. Also CNTF did not effect the distribution of STAT 5A. Unlike other adipocyte-expressed STATs, some STAT 5A is always present in the adipocyte nucleus (28).
The dose-dependent effects of CNTF on 3T3-L1 adipocytes were also examined by treating adipocytes for 10 min with varying concentrations of CNTF. As shown in Fig. 5B, CNTF had no effect on STATs 1 or 5A but resulted in the tyrosine phosphorylation and nuclear translocation of STAT 3. In addition, CNTF treatment resulted in a dose-dependent activation of MAPK (ERKs 1 and 2). To assess the dose effects of CNTF on STAT 3 activation, we compared STAT 3 activation in preadipocytes and adipocytes. As shown in Fig. 5C, CNTF results in a dose-dependent effect on STAT 3 activation in preadipocytes but not in adipocytes.
To characterize further the effects of CNTF, we treated fully differentiated 3T3-L1 adipocytes for a 12-h period, and we isolated whole cell extracts at the times indicated at the top of Fig. 6. As indicated previously, acute CNTF treatment resulted in a time-dependent activation of STAT 3 and MAPK but was unable to activate Akt. A positive control for Akt activation (10 min of treatment of 3T3-L1 adipocytes with 50 nM insulin) is shown in the bottom panel of Fig. 6. Acute CNTF treatment did not affect the expression levels of STATs 1, 3, or 5A. There were also no observable differences in the expression of SREBP-1 protein, as indicated by the levels of the cleaved 67-kDa form of the protein.
Next, we examined the effects of chronic CNTF administration on the expression of adipocyte transcription factors and other adipocyte proteins. Fully differentiated 3T3-L1 adipocytes were exposed to CNTF over a 96-h period. A fresh bolus of CNTF was added to the cells every 24 h. Whole cell extracts were isolated at the times indicated at the top of Fig. 7 and subjected to Western blot analysis. Chronic administration of CNTF did not alter the expression of adipocyte expressed

FIG. 3. Tissue distribution of CNTF receptor complex components in rodents.
Tissue extracts were prepared from 8-week-old Sprague-Dawley rats. Seventy-five g of each extract was separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis. Samples were processed, and results were visualized as described in Fig. 1 legend. Whole cell extracts from 3T3-L1 preadipocytes and adipocytes as well as tissue extracts from the epididymal fat pads from lean and obese Zucker rats were also examined. This is a representative experiment independently performed two times.

FIG. 4. Glycosylation of CNTF receptor complex components in rat tissues.
Tissue extracts were prepared from 8-week-old Sprague-Dawley rats. Eighty g of each extract was incubated with 4 l of PNGaseF (5,000 units/l) as directed by the manufacturer's instruction and then separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis. Samples were processed, and results were visualized as described in Fig. 1 legend. This is a representative experiment independently performed two times.

FIG. 5. Time-and dose-dependent effects of CNTF administration on the phosphorylation and nuclear translocation of STAT proteins in 3T3-L1 cells.
A, cytosolic and nuclear extracts were prepared from fully differentiated 3T3-L1 adipocytes following a treatment with 0.8 nM CNTF for the times indicated at the top of the figure. B, cytosolic and nuclear extracts were prepared from fully differentiated 3T3-L1 adipocytes following a 10-min treatment with CNTF at the doses indicated. C, whole cell extracts were prepared from both preadipocytes and from fully differentiated 3T3-L1 adipocytes following a 10-min treatment with CNTF at the doses shown in the figure. Seventyfive g of each extract was separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis. Samples were processed, and the results were visualized as described in Fig. 1 legend. This is a representative experiment independently performed three times. STATs, PPAR␥, or C/EBP␣. There were no notable differences in the levels of gp130 and LIFR expression, and chronic CNTF treatment was insufficient to induce the expression of CNTFR␣. A positive control of confluent preadipocytes is shown for CNTFR␣ expression. We also observed that CNTF treatment did not alter the expression of acrp 30 in 3T3-L1 adipocytes. Moreover, CNTF had no effect on the expression or secretion of leptin from 3T3-L1 adipocytes (data not shown). Interestingly, CNTF treatment resulted in a decrease of FAS expression and a substantial increase in the expression of IRS-1. Also, the levels of the 67-kDa SREBP-1 protein were slightly decreased by CNTF treatment after 72 h.
Although we did not observe substantial levels of CNTFR␣ in cultured 3T3-L1 adipocytes (Figs. 2, 3, and 5), we were able to detect the expression of CNTFR␣ in rodent adipose tissue from an obese Zucker rat, and the levels of this receptor appeared to be up-regulated in conditions of obesity (Fig. 3). Hence, we examined the expression of CNTF receptors in adipose tissue of additional rodent models of obesity/type II diabetes. Whole cell extracts were prepared from epididymal fat pads of five ob/ob and five ob/ϩ lean littermates. As shown in Fig. 8A, we observed very little expression of CNTFR␣ in the adipose tissue of lean mice, but we observed a substantial increase in the expression of this receptor in three of the five obese insulinresistant ob/ob littermates. In the other two ob/ob mice, there was a modest increase in CNTFR␣ expression. In addition, we observed increased LIFR expression in all five ob/ob mice compared with lean littermates, but we did not observe any substantial changes in gp130 expression. We also examined the expression of these proteins in the epididymal fat pads of fa/ϩ and fa/fa rats. As shown in Fig. 8B, the expression of CNTFR␣ was also substantially up-regulated in this rodent model of obesity/type II diabetes. However, we did not observe an increase in LIFR levels in the fa/fa rats as compared with their lean littermates, although there was a modest increase in gp130 levels in adipose tissue from fa/fa rats. We also examined the expression of these receptors in transgenic mice that over express agouti under the control of the ␤-actin promoter, a condition that causes obesity and type II diabetes (29). We observed a substantial increase in CNTFR␣ levels in the epididymal fat pads of three obese transgenic mice (Tg/ϩ) compared with wild-type lean (ϩ/ϩ) mice. There was also a modest decrease in LIFR and gp130 in the fat pads of mice with agouti-induced obesity. Finally, we examined the expression of CNTF receptors after low fat or high fat feeding in C57B1/6J mice. Seven mice from each condition were analyzed for CNTF receptor expression. The results in Fig. 8D only include three animals per condition. However, this pattern of regulation was observed for all seven animals examined for each condition (data not shown). In C57B1/6J mice, we observed an increase in CNTFR␣ levels with high fat feeding after 12 weeks. A similar pattern was also observed after 7 weeks (data not shown). Overall, there was no modulation of LIFR or gp130 with high fat feeding in the C57Bl/6J mice.
We have shown that cultured adipocytes do not express CNTFR␣, but rodent adipose tissues express detectable levels of the receptor. Therefore, we examined the ability of CNTF to activate STAT 3 in vivo. C57Bl/6J mice were given an intraperitoneal injection of CNTF (33.3 g/kg) or vehicle (saline) control. Fifteen minutes after the injection, the mice were sacrificed, and epididymal adipose tissue, brains, and skeletal muscle were immediately removed and frozen in liquid nitrogen. Whole cell extracts were prepared from these tissues and FIG. 6. The effects of acute CNTF administration on the expression of adipocyte proteins. Whole cell extracts were prepared from fully differentiated 3T3-L1 adipocytes treated with 0.8 nM CNTF for the times shown. Seventy-five g of each extract was separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis. Samples were processed and results were visualized as described in Fig. 1 legend. This is a representative experiment independently performed three times. Seventy-five g of each extract was separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis. Samples were processed, and results were visualized as described in Fig. 1 legend. This is a representative experiment independently performed three times. analyzed for STAT 3 phosphorylation by Western blot analysis. As shown in Fig. 9, acute CNTF treatment resulted in the activation of STAT 3 in epididymal adipose tissue. We were unable to detect STAT 3 phosphorylation in the adipose tissue of five saline-injected mice, but four of the five CNTF injected mice had readily detectable levels of phosphorylated STAT 3. The increase in STAT 3 phosphorylation was not due to increased STAT 3 expression. Also, the expression of LIFR was not changed, and the levels of CNTFR␣ were variable in the 10 mice. The results in Fig. 9B demonstrate constitutive STAT 3 phosphorylation in brain, which was unresponsive to exogenous CNTF. Moreover, we observed an increase in STAT 3 phosphorylation in the skeletal muscle of CNTF-treated animals, as compared with saline controls. As indicated previously (Fig. 2), the levels of CNTFR␣ in the brain are substantially greater than the levels in skeletal muscle.
Our results demonstrate that CNTFR␣ receptor expression was decreased during the adipogenesis of 3T3-L1 cells but expressed in the fat pads of rodents. Therefore, we fractionated epididymal fat pads from C57Bl/6J mice to determine whether the CNTFR␣ receptor was expressed in the stromovascular fraction or in the adipocytes. As shown in Fig. 10, our results clearly demonstrate that CNTFR␣ is expressed highly in the adipocytes, and STAT 3 is expressed at higher levels in the stromovascular portion. We hypothesize that the loss of CNTFR␣ that occurs during differentiation in vitro (Fig. 2) could be an artifact of cell culture because this receptor is expressed in native adipocytes (Fig. 10) and in media obtained from cultured 3T3-L1 adipocytes (data not shown). , and 17-week-old C57B1/6J mice fed a low or high fat diet for 12 weeks (D). In each panel, 75 g of each extract was separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis. Samples were processed and results were visualized as described in Fig. 1 legend.

FIG. 9. In vivo effect of acute CNTF administration in rodents.
Six-week-old C57Bl/6J mice were given an intraperitoneal injection of CNTF (33.3 g/kg) or vehicle (saline) control. Fifteen minutes after the injection, the mice were sacrificed, and epididymal fat pads, brains, and skeletal muscle were immediately removed and frozen in liquid nitrogen. Tissue extracts were analyzed from epididymal fat pads (A) and brain and skeletal muscle (B). In each panel, 75 g of each extract was separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis. Samples were processed, and results were visualized as described in Fig. 1 legend. This is a representative experiment independently performed two times.

FIG. 10. In vivo expression of CNTFR␣ in epididymal fat pads.
Epididymal fat pads were extracted from 6-week-old lean C57Bl/6J mice and fractionated into adipocyte and stromovascular fractions. Seventy-five g of each extract was separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis. Samples were processed, and results were visualized as described in Fig. 1 legend.
Because CNTF administration of ob/ob, db/db, and diet-induced obesity mice has been shown to improve insulin sensitivity in vivo, we examined the ability of CNTF to regulate insulin-sensitive glucose uptake in 3T3-L1 adipocytes. Fully differentiated 3T3-L1 adipocytes were treated for 24 h with CNTF. As shown in Fig. 11A, CNTF treatment resulted in a notable increase (25-50%) in GLUT 4 levels. However, additional treatments of CNTF did not result in a further increase in GLUT 4 levels as chronic CNTF treatment did not substantially increase GLUT 4 mRNA or protein levels (data not shown). Therefore, we examined the ability of CNTF to affect glucose uptake. Fully differentiated adipocytes were treated for 72 h with CNTF. Every 24 h, cells were treated with a fresh bolus of CNTF. Acute insulin treatment (50 nM, 7 min) resulted in a 5-fold increase in insulin-stimulated glucose uptake and was relatively unaffected by chronic CNTF treatment (Fig.  11B). In addition, CNTF had no effect on basal glucose uptake.
Because CNTF treatment resulted in an increase in IRS-1 expression levels (Fig. 7), we examined the ability of this cytokine to induce IRS-1 activation, as judged by tyrosine phosphorylation at residue 896. As shown in Fig. 12, acute insulin treatment (15 min) results in the activation of IRS-1 and Akt, whereas acute CNTF treatment does not. However, CNTF pretreatment (30 min) prior to insulin stimulation resulted in an increased IRS-1 activation (Ͼ20%) and increased Akt phosphorylation (Ͼ25%). The efficacy of the CNTF is demonstrated by the activation of STAT 3.

DISCUSSION
In the light of recent findings demonstrating that CNTF administration results in weight loss and correction of many other obesity/type II diabetes-related symptoms (19,22,24,30), we hypothesized that the effects of this cytokine may not be limited to the central nervous system and that CNTF may also have effects on peripheral tissues such as adipose tissue. Our in vitro studies using 3T3-L1 preadipocytes and adipocytes have shown that CNTF indeed has significant, yet different, effects on these two cell types. In 3T3-L1 adipocytes we observed that CNTF was a potent activator of the Jak/STAT pathway, in particular STAT 3, as well as an activator of a MAPK signaling cascade that resulted in activation of ERKs 1 and 2. In preadipocytes CNTF elicited similar effects but also resulted in the activation of Akt. Our studies revealed that two of the three CNTF receptor components, LIFR and CNTFR␣, were downregulated during the adipogenesis of 3T3-L1 cells. Our study clearly demonstrates that the expression of CNTFR␣ is substantially decreased during the course of adipocyte differentiation. Other studies (16) have shown that CNTFR␣ is downregulated during astrocyte differentiation. A previous investigation had also indicated a decrease in LIFR during adipogenesis (31), but this is the first investigation to demonstrate a decrease in CNTFR␣ during adipogenesis. We hypothesize that decreased expression of CNTF receptors upon differentiation accounts for cultured adipocytes being less sensitive to CNTF treatment than preadipocytes, as judged by STAT 3 or Akt activation.
Although LIFR and CNTFR␣ protein levels are reduced in cultured adipocytes, as compared with preadipocytes, we observed that adipocytes were still responsive to CNTF. It has been demonstrated previously (15,16) that CNTF can induce signaling in the absence of CNTFR␣, solely by binding to a gp130:LIFR dimeric receptor. Acute treatment of CNTF did not alter the expression levels of any STATs or any other adipocyte transcription factors in 3T3-L1 adipocytes. Hence, we examined the chronic effects of CNTF on 3T3-L1 adipocytes, and we observed that this cytokine affected the expression of several adipocyte-enriched proteins, including SREBP-1, FAS, GLUT4, and IRS-1. The reduction in the levels of SREBP-1 and FAS is indicative of decreased biosynthesis of fatty acids that may account for some portion of weight loss and decreased fat mass observed in patients treated with CNTF (Axokine). In agreement with previous findings that this CNTF-induced weight FIG. 11. CNTF does not cause insulin resistance but increases GLUT 4 expression. A, whole cell extracts were prepared from fully differentiated 3T3-L1 adipocytes treated with 0.8 nM CNTF for the times shown. Seventy-five g of each extract was separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis. Samples were processed, and results were visualized as described in Fig. 1 legend. B, fully differentiated 3T3-L1 adipocytes were treated with CNTF for 72 h. A fresh bolus of CNTF was added to the cells every 24 h. Monolayers of adipocytes were used to examine glucose uptake as indicated under "Experimental Procedures." This is a representative experiment independently performed four times.
FIG. 12. The effects of acute CNTF treatment on IRS-1 and Akt activation in 3T3-L1 adipocytes. Whole cell extracts were prepared from fully differentiated 3T3-L1 adipocytes treated with 0.8 nM CNTF or 50 nM insulin for the times indicated in the figure. Seventy-five g of each extract was separated by SDS-PAGE, transferred to nitrocellulose, and subjected to Western blot analysis. Samples were processed, and results were visualized as described in Fig. 1 legend. This is a representative experiment independently performed three times. loss was not due to cachexia or inflammation (22,25), we did not observe any effect of CNTF on PPAR␥ or C/EBP␣, two transcription factors known to be down-regulated by inflammatory cytokines such as tumor necrosis factor-␣ and interferon-␥ (32)(33)(34). Also, unlike the in vitro effects of other cytokines (32,(35)(36)(37), CNTF treatment of 3T3-L1 adipocytes did not result in the onset of insulin resistance (Fig. 10). Moreover, chronic CNTF treatment of these cells actually resulted in an increase in both GLUT4 and IRS-1 protein levels. However, we did not observe any effects of CNTF on basal or insulin-stimulated glucose uptake. Clearly, additional experiments are required to determine whether CNTF can act as an insulin sensitizer. Nonetheless, we have shown that CNTF appears to act synergistically with insulin to increase the level of IRS-1 and Akt phosphorylation in 3T3-L1 adipocytes.
Our results strongly suggest that CNTF affects adipose tissue and skeletal muscle in vivo because an acute intraperitoneal injection of CNTF resulted in STAT 3 activation in both tissues. We also observed that CNTFR␣ is expressed not only in brain and skeletal muscle but also in adipose tissue, spleen, heart, testes, lungs, and stomach. The receptor expression levels, as well as the protein size, vary among these tissues, but our deglycosylation studies clearly demonstrate that they all express CNTFR␣. One of the most important findings we observed was that, in vivo, CNTFR␣ expression was significantly increased in four different rodent models of obesity/type II diabetes, including both genetic and diet-induced obesity. Moreover, we have shown that CNTFR␣ is expressed at higher levels in the adipocytes as compared with the stromovascular portion of the fat pad (Fig. 10). Although we observed an increase in the expression of the LIFR in the ob/ob mice, as compared with lean littermates, the expression of this receptor was not altered in the fa/fa rats or in C57Bl/6J mice with diet-induced obesity.
The results of our study suggest that CNTF and CNTFR␣ may play a role in the regulation of adipocyte metabolism and, perhaps, the control of adipose tissue mass. Our results have led us to hypothesize that CNTF can act as an insulin sensitizer in adipocytes. Therefore, the up-regulation of CNTFR␣ in adipose tissue of obese/type II diabetic rodents could be an adaptive response attempting to increase insulin sensitivity. Interestingly, some studies suggest that CNTFR␣ may not only act as receptor for CNTF but also as a receptor for another unknown CNTF-like factor. For example, mice lacking CNTF develop normally and appear to have no visible defects well into adulthood, when they develop minor loss of motor neurons (10). Yet mice lacking CNTFR␣ tend to have severe motor neuron defects and die perinatally because they fail to initiate feeding behaviors (11). Also the finding that CNTF expression is undetectable in the feeding-relevant brain sites, which express high levels of CNTFR␣ (30), further supports the notion that CNTFR␣ may have additional ligands and/or functions.
In summary, we observed that native as wells as cultured adipocytes are responsive to CNTF treatment. Interestingly, CNTFR␣ is not highly expressed in cultured adipocytes but is readily detectable in rodent adipose tissue and furthermore highly up-regulated in multiple rodent models of obesity/type II diabetes. This is the first demonstration that this receptor is expressed in adipose tissue and that it is highly regulated in obesity/type II diabetes. Current studies are underway to determine the role of CNTFR␣ in adipose tissue function and examine the ability of CNTF to act as insulin sensitizer in fat and muscle.