Leptin, the product of the ob/ob gene (1) is an adipose

Evidence for Leptin Binding to Proteins in Serum of Rodents and Humans: Modulation With Obesity Karen L. Houseknecht, Chris S. Mantzoros, Regina Kuliawat, Edward Hadro, Jeffrey S. Flier, and Barbara B. Kahn Many hormones circulate bound to serum proteins that modulate ligand bioactivity and bioavailability. To understand the biology of leptin action, we investigated the presence of leptin binding proteins in serum. -labeled leptin binds competitively to at least three serum macromolecules with molecular masses of ~85, -176, and -240 kda in rodents and -176 and -240 kda in humans. The ability to bind appears to involve sulfhydryl/disulfide interactions because it is inhibited under reducing conditions. When serum is added to recombinant 125 -leptin, there is a shift in sedimentation of 125 -leptin as analyzed by sucrose gradient centrifugation from approximately S1.9 to approximately S4.3. This shift is markedly attenuated in serum from obese mice {ob/ob, db/db, brown-fat ablated, goldthioglucose treated, high-fat fed) compared with that from nonobese controls. The size distribution of endogenous serum leptin as determined by radioimmunoassay (RA) after sucrose gradient centrifugation is also consistent with saturation of binding in hyperleptinemic obesity. n humans, free leptin increases with BM. Thus, in lean rodents and humans a large proportion of leptin circulates bound to several serum proteins. Free leptin is increased in serum of obese subjects, which may alter leptin bioactivity, transport, and/or clearance. Diabetes 45:1638-1643, 1996 Leptin, the product of the ob/ob gene (1) is an adipose tissue-derived secreted protein that has been implicated in the regulation of food intake and whole-body energy balance. Administration of exogenous leptin to ob/ob mice, which lack the protein, results in reduced food intake, increased energy expenditure, increased physical activity, and normalization of hyperglycemia and hyperinsulinemia (2-5). Leptin effects appear to be exerted, at least in part, at the level of the hypothalamus, since effects are seen in ob/ob mice at far lower doses after intracerebroventricular (CV) than after From the Harvard Thorndike Research Laboratory and Department of Medicine at Harvard Medical School and Beth srael Hospital, Boston, Massachusetts. Address correspondence and reprint requests to Dr. Barbara B. Kahn, Diabetes Unit/Beth srael Hospital, 330 Brookline Ave., Boston, MA 02215. E-mail: bkahn@bih.harvard.edu. Received for publication 3 July 1996 and accepted in revised form 29 August 1996. BSA, bovine serum albumin; GTG, gold-thioglucose; (3ME, p-mercaptoethanol; RA, radioimmunoassay; TBS, Tris-buffered saline; UCP-DTA, uncoupling protein promoter driving diphtheria toxin alpha. peripheral administration (4,6). One probable target is hypothalamic neuropeptide Y expression (6). The expression of leptin and neuropeptide Y appear to be reciprocally regulated (6-8). Circulating leptin concentrations are positively correlated with body lipid and are elevated in many models of rodent and human obesity (9-12). Thus, a hallmark of obesity is not the absence of leptin, but leptin resistance. Leptin resistance could result from downregulation of the leptin receptor, defective binding to or signaling from the receptor, altered pathways downstream of leptin, and/or alterations in the bioavailability or bioactivity of circulating leptin. The gene for the leptin receptor was recently cloned and shows sequence homology to the cytokine receptor family (13-15). Cytokines circulate bound to a number of binding proteins, which include a 2 macroglobulins, extracellular matrix proteins, monospecific binding proteins, and secreted extracellular domains of cytokine receptors (16,17). The leptin receptor is predicted to exist in at least six forms arising from alternative mrna splicing (14,15). One form has a sequence suggesting a soluble receptor that could act as a binding protein for circulating leptin. Potential functions of binding proteins include altering the clearance rate of hormones or cytokines, increasing or decreasing the biological activity of the ligand, and/or providing hormone responsiveness to unresponsive cells. Little is known about the properties of circulating leptin in rodents and humans. Using multiple biochemical approaches, we provide evidence that 2) the majority of leptin circulates bound to protein(s) in serum from mice and humans, 2) there are multiple leptin-binding species, and 3) the phenomenon is influenced, at least in part, by obesity. RESEARCH DESGN AND METHODS Human subjects and animals. Blood samples were collected with informed consent from seven Caucasian women ranging in age from 32 to 66 years. Six subjects had no known medical problems, while a morbidly obese subject had mild NDDM. Subjects fasted overnight before blood sampling unless otherwise indicated. Mouse strains studied included FVB, ob/ob, db/db, and brown-fat-deficient (UCP-DTA, uncoupling protein promoter driving diphtheria toxin alpha [18]). Blood samples were collected by cardiac puncture at time of killing or from the tail vein of live male and female mice. Obesity was induced in male FVB mice (4 weeks old) by injection of goldthioglucose (GTG; 0.5 mg/g body wt; #50833 Fluka, Ronkonkoma, NY) after an overnight fast. GTG-injected mice were killed at 8 weeks of age, and blood was collected. All mice were housed at 21 C with a 12-h light/dark cycle and were fed standard Rodent Chow (Purina #5008) or a high-fat diet (55% of calories from fat, 24% carbohydrate, 21% protein, as previously described [19]) ad libitum. The high-fat diet was initiated at weaning (3 weeks) and continued for 10 weeks. Mice were killed by CO 2 inhalation followed by cervical dislocation. 1638 DABETES, VOL. 45, NOVEMBER 1996

K.L. HOUSEKNECHT AND ASSOCATES A kda 208 144 87 16 (+)BME (-)BME B (+)BME a uj 1 (-)BME 1 1 L ft r i kda 208 87 44 16 rood) 2 E Q. a a> FVB UCP-DTA FVB UCP-DTA Mouse (+) 1.0 n,g leptin Human (+) 1.0 pig leptin n > > O U. Li. O GO CD 5 > > O L LL FG. 1. Effects of reducing agents on the electrophoretic mobility of immunoreactive leptin in lean G4) and obese (2?) mice. Radioligand binding of 125 -leptin to proteins in serum of mice and humans (C). A: plasma or serum samples pooled from two to six lean or obese FVB mice were electrophoresed in the presence (+0ME) or absence (-(ime) of 5% pme and in the presence of SDS (4%) and blotted for leptin as described in METHODS. B: serum from two to four UCP-DTA brown-fat-ablated transgenic mice compared with that from lean FVB littermates. C: sera from two to four normal FVB or ob/ob mice and from nonobese women were electrophoresed in the presence of 4% SDS without PME in 16% (mouse) or 8% (human) Tris/glycine gels and blotted with 125 -labeled recombinant murine or human leptin (± 1.0 ug recombinant murine leptin) as described in METHODS. Data are representative of two experiments for each condition. mmunoblotting and ligand blotting. Serum samples (5 ul) from mice or humans were mixed with loading buffer (450 mmol/1 Tris-HCl, 12% glycerol, 4% SDS, 0.0025% Coomassie blue G, 0.0025% Phenol red) in the absence or presence of p-rnercaptoethanol (pme) (5%). Samples were electrophoresed in 16% tricine gels and transferred to PVDF. Filters were blotted for leptin as previously described (11). For ligand blotting, serum samples were mixed with loading buffer and were electrophoresed under nonreducing conditions in 16 or 8% Tris/glycine gels and transferred to polyvinylidene fluoride (PVDF). Membranes were blocked with 5% bovine serum albumin (BSA; Sigma, radioimmunoassay [RA] grade) for 2 h at room temperature, after which blots were incubated with 125 -labeled mouse or human leptin (Linco Research, St. Louis, MO) in Tris-buffered saline (TBS) (1% BSA) overnight at 4 C. Blots were washed twice in TBS (0.1% Tween-20) and exposed to film. Additionally, human serum was incubated with 125 -leptin for 30 min at 37 C and was then electrophoresed under denaturing (4% SDS) but nonreducing conditions in 8% Tris/glycine gels. Gels were dried and subjected to autoradiography. Sucrose gradient centrifugation. Leptin sedimentation profiles were determined by rate-zonal centrifugation in continuous sucrose gradients (5-20% sucrose in phosphate-buffered saline [ph 7.4]). Sera or standards (lysozyme, S.9; bovine serum albumin, S4.3; fibrinogen, S7.9; catalase, S11.3) were loaded in a final volume of 100 ul. For radioisotope binding studies, serum was incubated in the presence of tracer quantities of 126 Mabeled murine or human leptin (-1,600 cpm/ul) for 30 min at 37 C. To test the specificity of ligand binding, one sample per treatment was simultaneously incubated with 1.0 ug recombinant murine leptin (Eli Lilly, ndianapolis, N). Gradients were centrifuged at 200,000^ for 7 h at 4 C in a TLS-55 swinging-bucket rotor. Fractions were analyzed for 126 -leptin using a gamma counter or analyzed for endogenous leptin by RA (Linco Research). The limit of detection was 0.2 ng/ml, and the intra-assay coefficient of variation was 4.39% for lowlevel controls (2.9 ng/ml) and 5.66% for high-level controls (14.1 ng/ml). nterassay variation was 6.9 and 9% for low and high controls, respectively. RESULTS Ligand blotting and immunoblotting. Electrophoresis of mouse serum or plasma under denaturing and reducing (+(3ME) conditions results in the expected immunoreactive band of 16 kda, which represents free leptin (Fig. LA). However, under nonreducing conditions (- pme), the 16-kDa band is not detectable and three higher molecular mass immunoreactive bands are visible (Fig. LA). A similar phenomenon is observed in serum from obese UCP-DTA mice compared with that from non- DABETES, VOL. 45, NOVEMBER 1996 1639

LEPTN BNDNG PROTENS N RODENT AND HUMAN SERUM s=1.9 s=4.3 s=11.3 1250-1000- E a o 750-500- 250-2000- 1500-1000- 500-6 8 Fraction FG. 2. Sedimentation profiles of 125 -leptin and 1Z5 -leptin/serum protein interactions in mouse serum. 125 -leptin in the presence ( ) or absence ( ) of mouse serum or in the presence of mouse serum and 1.0 ug recombinant murine leptin (K) was subjected to continuous sucrose gradient centrifugation as described in METHODS. Serum samples were pooled from two to six mice. A: normal FVB females, serum leptin 4.03 ng/ml. B: FVB females fasted for 24 h. C: FVB female mice maintained on a high-fat diet for 10 weeks as described in METHODS, serum leptin 5.76 ng/ml. D and E: db/db and ob/ob females; serum leptin values were not determined because of an insufficient sample. F: FVB male mice treated with GTG to induce obesity as described in METHODS, serum leptin 28.5 ng/ml. transgenic lean FVB littermates (Fig. B). These data suggest that a large fraction of leptin in serum and plasma exists in "higher molecular mass" forms that may involve sulfhydryl/disulfide bonds. t is possible that these higher molecular mass species could be multimers of leptin/binding protein complexes. 125 -leptin competitively bound to immobilized proteins in mouse and human serum (Fig. 1C). Radioligand blotting of normal mouse serum revealed specific competitive binding of 125 -leptin to proteins of apparent molecular masses -240, -176, and -85 kda (Fig. 1C). Binding to the 85-kDa band was absent or very low in serum from female ob/ob mice (Fig. 1C) and was somewhat reduced in GTG obese treated mice (not shown). n human serum, leptin bound to proteins with apparent molecular masses of -240 and -176 kda (nonreducing conditions), and binding was completely inhibited by competition with 1 ug recombinant leptin (Fig. 1(7). Similarly, when human serum was preincubated with 125 leptin and then subjected to nonreducing gel electrophoresis, radioactive bands were observed at -240 and -176 kda (not shown). The 85-kDa band observed in mouse samples was either absent or below the level of detection in human samples. Sucrose gradient centrifugation. Sedimentation profiles for 125 -leptin (Figs. 2 and 3) reveal that leptin (molecular mass 16 kda) has an apparent S value of 1.9, comigrating with the lysozyme standard (molecular mass 14.4 kda). ncubation of 125 -leptin with serum from normal female FVB mice results in a redistribution of radiolabeled leptin in the gradients so that the peak fraction of 125 -leptin shifts to approximately S4.3 and comigrates with the BSA standard (molecular mass 66 kda). This increase in the sedimentation rate reflects an increase in the apparent size of 125 -leptin, presumably due to complex formation with other molecules in serum. Using iodinated hormones to assess biological functions always carries the risk of altering the bioactivity or the specificity of binding of the hormone by iodination. n our studies, iodination of leptin does not appear to disrupt its binding to serum protein(s) or to create nonspecific binding, since the shift of the peak to the approximate S4.3 region is completely inhibited with 1 ug cold murine recombinant leptin (Fig. 2A). The fact that 125 -leptin binding demonstrated by an independent technique, ligand binding (Fig. 1C), is also competed with cold leptin provides further evidence that 125 leptin binding is specific. 1640 DABETES, VOL. 45, NOVEMBER 1996

K.L. HOUSEKNECHT AND ASSOCATES 1000n 750-8" 500 H 250- s=1.9 s=4.3 s=11.3 states characterized by significantly elevated circulating leptin. n one additional lean subject with low leptin levels, leptin measurements were at the limit of detection of the assay and therefore are not shown. n lean mice with low circulating leptin (2.5 and 3.5 ng/ml; Fig. 45), the immunoreactive leptin is predominantly found in the bound state, although a portion is free (S-1.9). However, with brown-fat-deficient (UCP-DTA) or high-fat-induced obesity (Fig. 4B), serum leptin levels are elevated (UCP-DTA, leptin = 13.5 ng/ml) and a major portion is found in the free state, again suggesting saturation of the endogenous binding protein(s). This is most clearly seen in the GTG-treated mouse, which is massively obese, has very high serum leptin concentrations (28.5 ng/ml; Fig. 4(7), and the highest peak of free leptin (S-1.9). The profile is shown in relation to obese UCP- DTA mice (note change in scale). 0 2 4 6 8 10 12 Fraction FG. 3. Sedimentation profiles of 125 -leptin and 125 -leptin/serum protein interactions in human serum. 125 -leptin in the presence ( ) or absence ( ) of female human serum or in the presence of human serum and 1.0 ug recombinant murine leptin (X) was subjected to continuous sucrose gradient centrifugation as described in METH- ODS. The data show results for serum from a Caucasian woman in the fed state, age 32 years, BM 27, waist-to-hip ratio 0.72, serum leptin 9.3 ng/ml. This sample was analyzed in duplicate and nearly identical results were obtained. These data are representative of five different women. A similar binding phenomenon is observed when serum from female mice fasted for 24 h is analyzed (Fig. 2E). However, with serum from high-fat-fed obese mice (Fig 2(7), db/db mice, which lack functional leptin receptors (Fig 2D), or ob/ob mice, which lack full-length leptin (Fig 2E), the shift in leptin sedimentation is attenuated. Leptin sedimentation profiles for male mice also illustrate a binding phenomenon that is competitively inhibited with 1 ug cold leptin and is reduced with genetic (UCP-DTA) and high-fat-induced obesity (not shown). Figure 2F shows that the shift is obliterated in males with GTG-induced obesity, the model with the highest serum leptin levels. The leptin binding shift is also observed with human serum (Fig. 3). 125 -leptin specifically interacts with protein(s) to cause a shift toward the approximate S4.3 peak. The data shown are for serum from a woman with a BM of 27. These data are representative of data for five different women. n serum from one morbidly obese woman (BM 38), the shift was reduced. Fasting in women does not alter the ability of 125 -leptin to bind to serum protein(s) (not shown). To determine the sedimentation profile of endogenous leptin in mice and humans, serum was analyzed by sucrose gradient centrifugation and RA (Fig. 4). n women (Fig. 4A), immunoreactive leptin in serum is found both free (S-1.9) and bound (S>1.9). The amount of free leptin increases with BM and appears most affected by the total leptin level (as measured by RA). This is consistent with the notion that the binding proteins are saturated in obese DSCUSSON n the present study, we use multiple experimental approaches to show that a major portion of leptin does not circulate as a 16-kDa monomeric species in normal mice and humans. Ligand binding illustrates that radioactive leptin competitively binds to two (human) or three (mouse) serum proteins (Fig. 1), and immunoblotting indicates that the leptin binding is reversed by reducing agents. Thus, either leptin binds to the binding proteins via sulfhydryl/disulfide bonds or such interactions within the binding proteins critically affect their ability to bind leptin. Sucrose gradient centrifugation, which separates molecular complexes on the basis of size and shape under native conditions, was used to analyze 125 -leptin interactions with serum macromolecules(s). These data provide independent evidence for the leptin binding phenomenon (Figs. 2 and 3). Serum incubation results in a shift of free monomeric leptin from a sedimentation coefficient of -1.9 to a larger complex (S-4.3), and this is blocked by excess cold leptin. We would predict from ligand binding data that the binding complex would be larger than indicated by S value. However, the approximate S4.3 is compatible with leptin binding to the -85- kda protein if the complex has significant asymmetry, since asymmetrical particles sediment more slowly than spherical particles of the same mass and density. The conditions used for sucrose gradient analyses were optimized to resolve the S-4.3 peak because of the large effect seen. t is possible that leptin binding to additional serum proteins could be resolved under different conditions (i.e., larger complexes may be lost in the pellet, fraction 12). The ability of exogenous leptin to form this particular binding complex is attenuated in several models of obesity but not of fasting. This is consistent with the notion that available leptin binding sites are saturated with elevated circulating leptin levels present in obesity. The fact that the greatest diminution of binding was observed in the model with the highest circulating leptin, the GTG-treated mouse, supports this hypothesis. The hypothesis of saturation of serum leptin-binding sites in obesity is further supported by data on endogenous leptin sedimentation profiles (Fig. 4). n nonobese mice and humans, a significant portion of endogenous leptin is bound (S>1.9). With increasing severity of obe- DABETES, VOL. 45, NOVEMBER 1996 1641

LEPTN BNDNG PROTENS N RODENT AND HUMAN SERUM B c A* O) c Q. 10-7.5-2.5-4 6 8 Fraction FG. 4. Sedimentation profiles of endogenous leptin in serum from lean and obese women and male mice. Sera from women (A) and from male mice (B and C) were subjected to continuous sucrose gradient centrifugation as described in METHODS. A: samples are from individual women: BM 37 (, leptin 27.86 ng/ml), BM 38.4 (*, leptin 22.7 ng/ml), BM 33.6 (, leptin 18.3 ng/ml), BM 27 (O, leptin 9.3 ng/ml), BM 20.5 (A, leptin 3.65 ng/ml), BM 18.6 (, leptin 2.21 ng/ml). B: Samples are pooled from two to six male mice: FVB mice (, leptin 3.5 ng/ml; A, leptin 2.5 ng/ml), brown-fat-ablated mice (UCP-DTA,, leptin 13.5 ng/ml), FVB male mice maintained on a high-fat diet for 10 weeks (, serum leptin not determined). C: comparison of serum from brown-fat-ablated male mice ( ) from B with serum from FVB male mice treated with GTG (, serum leptin 28.5 ng/ml) as described in METHODS. sity and increasing circulating leptin levels, leptin "spills over" into the free (S-1.9) pool. The fact that free leptin is detected in serum by RA and not by nonreducing immunoblotting (Fig. 1) most likely reflects the greater sensitivity of the RA. The physiological consequences of increased free leptin are unknown, but free leptin may have more rapid turnover because of proteolytic cleavage or increased clearance. This hypothesis is supported by the observation that the half-life (t l/2 ) of recombinant leptin injected into ob/ob mice is much shorter than that in normal mice (R. Frederich and J.S.F., unpublished observations). One must be cautious in interpreting these data, since it is currently unknown whether binding proteins interfere in the radioimmunoassay. Such interference could have a significant impact on the estimation of leptin levels by RA. One possible role for the binding proteins is to facilitate the transport of leptin across the blood-brain barrier to its hypothalamic (or other) site(s) of action. Precedent for an important role for binding proteins in the transport or uptake of ligands has been demonstrated for other members of this family (16,17). Additionally, for some cytokines and hematopoietic growth factors, association with binding proteins potentiates ligand activity because of biochemical modifications (16,17). These phenomena provide possible explanations for apparent leptin resistance in the setting of increased free leptin. Binding of 125 -leptin to the ~85-kDa protein is severely reduced or absent in serum from ob/ob mice (Fig. 1), which lack leptin (full-length protein). This most likely reflects downregulation of this binding protein similar to the downregulation of growth hormone-binding protein in states of growth hormone deficiency (20). Leptin may regulate the expression or degradation of this ~85-kDa binding protein. The fact that the binding phenomenon, as tested by sucrose gradient centrifugation (Fig. 2E), is not exaggerated in ob/ob serum (because of an absence of competition from endogenous leptin) is consistent with reduced levels of one or more binding proteins. We cannot rule out, however, that a mutant truncated form of leptin binds to and saturates serum binding protein(s). The ~85-kDa protein is the predicted molecular mass of the soluble splice variant of the leptin receptor (15). We are currently in the process of developing antibodies to the leptin receptor to test whether this protein is the soluble form of the receptor. n summary, using several biochemical approaches, we provide the first evidence that leptin circulates in multiple-molecular mass protein complexes in serum in humans and mice and that this binding is modulated by obesity. This binding phenomenon may have a major impact on the biological activity of leptin in normal and pathophysiological states. Future studies will focus on the isolation of these binding proteins, which will allow biochemical and kinetic evaluation of leptin/binding protein interactions. Regulation of these binding proteins may play a critical role in the adipose tissue neuroendocrine axis that regulates food intake and energy balance. 1642 DABETES, VOL. 45, NOVEMBER 1996

K.L. HOUSEKNECHT AND ASSOCATES ACKNOWLEDGMENTS This work was supported by National nstitute of Diabetes and Digestive and Kidney Disease Grants ROl DK- 43051, ROl DK-28082, ROl DK-45874, and P30 DK-46200 and U.S. Department of Agriculture grant 94-37200-0845. K.L.H. was supported by U.S. Department of Agriculture grant 94-0443. C.S.M. is at the Clinical nvestigator Program, Beth srael Hospital-Harvard-MT Health Sciences and Technology, in collaboration with Pfizer. We thank Eli Lilly (ndianapolis, N) for the gift of the recombinant murine leptin and the radioimmunoassay kits (Linco Research), Dr. Bettina Lollmann and Kevin Coughlin for technical assistance, and Dr. Brad Lowell for critically reading the manuscript. REFERENCES 1. Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM: Positional cloning of the mouse obese gene and its human homologue. Nature 372:425-432, 1994 2. Pelleymounter MA, Cullen MJ, Baker MB, Hecht R, Winters D, Boone T, Collins F: Effects of the obese gene product on weight regulation in oblob mice. Science 269:540-543, 1995 3. Halaas JL, Gajiwala KS, Maffei M, Cohen SL, Chait BT, Rabinowitz D, Lallone RL, Burkley SK, Friedman JM: Weight-reducing effects of the plasma protein encoded by the obese gene. Science 269:543-546, 1995 4. Campfield LA, Smith FJ, Guisez Y, Devos R, Burn P: Recombinant mouse OB protein: evidence for a peripheral signal linking adiposity and central neural networks. Science 269:546-549, 1995 5. Weigle DS, Bukowski TR, Foster DC, Holderman S, Kramer JM, Lasser G, Lofton-Day CE, Prunkard DE, Raymond C, Kuyper JL: Recombinant of protein reduces feeding and body weight in the oblob mouse. J Clin nvest 96:2065-2070, 1995 6. Stephens TW, Basinski M, Bristow PK, Bue-Valleskey JM, Burgett SG, Craft K, Hale J, Hoffmann J, Hsiung HM, Kriauciunas A, MacKeller W, Rosteck PR, Schoner B, Smith D, Tinsley FC, Zhang XY, Heiman M: The role of neuropeptide Y in the antiobesity action of the obese gene product. Nature 377:530-532, 1995 7. Sainsbury A, Cusin, Doyle P, Rohner-Jeanrenaud F, Jeanrenaud B: ntracerebroventricular administration of neuropeptide Y to normal rats increases obese gene expression in white adipose tissue. Diabetologia 39:353-356, 1996 8. Schwartz MW, Baskin DG, Bukowski TR, Kuijper Jl, Foster D, Lasser G, Prunkard DE, Porte D, Woods SC, Seeley RJ, Weigle DS: Specificity of leptin action on elevated blood glucose levels and hypothalamic neuropeptide Y gene expression in oblob mice. Diabetes 45:531-535, 1996 9. Considine RV, Considine EL, Williams CJ, Nyce MR, Magosin SA, Bauer TL, Rosato EL, Colberg J, Caro JF: Evidence against either a premature stop codon or the absence of obese gene mrna in human obesity. J Clin nvest 95:2986-2988, 1995 10. Frederich RC, Lollman B, Hamann A, Napolitano-Rosen A, Kahn BB, Lowell BB, Flier JS: Expression of ob mrna and its encoded protein in rodents: impact of nutrition and obesity. J Clin nvest 96:1658-1663, 1995 11. Frederich RC, Hamann A, Anderson S, Lollmann B, Lowell BB, Flier JS: Leptin levels reflect body lipid content in mice: evidence for diet-induced resistance to leptin action. Nature Med 1:1311-1314, 1995 12. Maffei M, Halaas J, Ravussin E, Pratley RE, Lee GH, Zhang Y, Fei H, Kim S, Lallone R, Ranganathan S: Leptin levels in human and rodent: measurement of plasma leptin and ob RNA in obese and weight-reduced subjects. Nature Med 11:1155-1161, 1995 13. Tartaglia LA, Dembski M, Weng X, Deng N, Culpepper J, Devos R, Richards GJ, Campfield LA, Clark FT, Deeds J: dentification and expression cloning of a leptin receptor, OB-R. Cell 83:1263-1271, 1995 14. Chen H, Chariat O, Tartaglia LA, Woolf EA, Weng X, Ellis SJ, Lakey ND, Culpepper J, Moore KJ, Breitbart RE, Duyk GM, Tepper R, Morgenstern JP: Evidence that the diabetes gene encodes the leptin receptor: identification of a mutation in the leptin receptor gene in dbldb mice. Cell 84:491-495, 1996 15. Lee GH, Proenca R, Montez JM, Carroll KM, Darvishzadeh JG, Lee J, Friedman JM: Abnormal splicing of the leptin receptor in diabetic mice. Nature 379:632, 1996 16. Heaney ML, Golde DW: Soluble hormone receptors. Blood 82:1945-1948, 1993 17. Bonner JC, Brody AR: Cytokine-binding proteins in the lung. AmJPhysiol 12:L869-L878, 1995 18. Lowell BB, Susulic VS, Hamann A, Lawitts JA, Himms-Hagen J, Boyer BB, Kozak LP, Flier JS: Development of obesity in transgenic mice after genetic ablation of brown adipose tissue. Nature 366:740-742, 1993 19. Gnudi L, Tozzo E, Shepherd PR, Bliss JL, Kahn BB: High level overexpression of glucose transporter-4 driven by an adipose-specific promoter is maintained in transgenic mice on a high fat diet, but does not prevent impaired glucose tolerance. Endocrinology 135:995-1002, 1995 20. Baumann G: Growth hormone-binding proteins: state of the art. J Endocrinol 141:1-6, 1994 21. Clemmons DR: GF binding proteins and their functions. Mol Reprod Dev 35:368-375, 1993 DABETES, VOL. 45, NOVEMBER 1996 1643