PAPER Obesity is associated with decreasing levels of the circulating soluble leptin receptor in humans

Transcription

1 (2002) 26, ß 2002 Nature Publishing Group All rights reserved /02 $ PAPER Obesity is associated with decreasing levels of the circulating soluble leptin receptor in humans V Ogier 1, O Ziegler 2,LMéjean 1, JP Nicolas 3 and A Stricker-Krongrad 1 * 1 Department of Human Nutrition, Animal Sciences Laboratory, Ecole Nationale Supérieure d Agronomie et des Industries Alimentaires (ENSAIA), Vandoeuvre les Nancy, France; 2 Department of Diabetology, Metabolic and Nutrition Diseases, Center of Clinical Investigation INSERM-Centre Hospitalier Universitaire de Nancy, Hôpital Jeanne d Arc, Dommartin-les-Toul, France; and 3 Faculty of Medicine, Medical and Pediatric Biochemistry Laboratory, Vandoeuvre les Nancy, France OBJECTIVE: Leptin plays a major role in the regulation of body weight. It circulates in both free and bound form. One of the leptin receptor isoforms exists in a circulating soluble form that can bind leptin. In the present study, we measured the soluble leptin receptor (SLR) levels in lean and obese humans. We investigated the relationship between plasma SLR levels, plasma leptin levels and the degree of obesity. We also examined whether SLR concentrations could be modulated by fat mass loss induced by a 3 month weight-reducing diet. SUBJECTS: A total of 112 obese (age y; body mass index (BMI) kg=m 2 ; 23 men and 89 women), 38 overweight (age y; BMI kg=m 2 ; 10 men and 28 women) and 63 lean (age y; BMI kg=m 2 ; 16 men and 47 women) humans. MEASUREMENTS: A direct double monoclonal sandwich enzyme-linked immunosorbent assay (ELISA) was used for the quantitative measurement of the soluble human leptin receptor. Leptin was measured by radioimmunoassay (RIA). Body composition was assessed by biphotonic absorptiometry DEXA (dual energy X-ray absorptiometry). RESULTS: We observed that the SLR is present in human plasma (range ng=ml). SLR levels were lower in obese and overweight than lean subjects ( , , ng=ml, respectively) and were inversely correlated to leptin and percentage of body fat (r ¼ and r ¼ 70.76; respectively; P < ). The ratio of circulating leptin to SLR was strongly related to the percentage of body fat (r ¼ 0.91; P < ). Interestingly a gender difference was observed in SLR levels, which were higher in obese and overweight men than in obese and overweight women. In obese subjects after a 3 month low-calorie diet, SLR levels increased in proportion to the decrease in fat mass. In the gel filtration profile, SLR coeluted exactly with the bound leptin fractions. CONCLUSION: Obesity, in humans is associated with decreasing levels of the circulating soluble leptin receptor (SLR). The relationship of SLR with the degree of adiposity suggests that high SLR levels may enhance leptin action in lean subjects more than in obese subjects. (2002) 26, DOI: =sj=ijo= Keywords: leptin; leptin receptor; obesity; binding protein; cytokine receptor; low-calorie diet; metabolism Introduction Leptin, the ob gene product 1 is implicated in the regulation of food intake and energy balance. 2 4 In rodents, the severe obese phenotype of the ob=ob or db=db mice is induced by a *Correspondence: A Stricker-Krongrad, Metabolic Diseases Physiology and Pharmacology, Millennium Pharmaceuticals Inc., 75 Sidney Street, Cambridge, 02139, MA, USA. krongrad@mpi.com Received 23 July 2001; revised 22 October 2001; accepted 12 November 2001 mutation in the leptin or leptin receptor gene. 1,5 However, in obese humans there are only a few cases of obesity that can be explained by mutation of the leptin gene 6 8 or of the leptin receptor gene. 9,10 Moreover, in obese humans, leptin levels are elevated, 11,12 suggesting that a hallmark of obesity is not leptin deficiency but hyperleptinemia. Leptin achieves its metabolic and endocrine effects by interacting with a receptor which is a member of the class I cytokine receptor family. 5,13 In the cytokine receptor family, the extracellular domain of the hormone receptor is present in the circulation and acts as binding protein to modulate the concentration of the

2 active ligand in the extracellular milieu Leptin, like most hormones circulates in humans both free and bound to macromolecules The estimation of the percentage of bound and free leptin has indicated that a significantly greater proportion of total leptin circulates in the bound form in lean compared with obese subjects. 18 Interestingly, one of the isoforms of the leptin receptor is a putative soluble leptin receptor that could act as a binding protein. 13,25 In the present study, we developed an immuno-assay for the direct quantitative measurement of the soluble leptin receptor (SLR) in human plasma and we measured SLR levels in lean and obese male and female subjects. We postulated that differences in the levels of SLR in plasma between obese and lean patients might explain their differences in the repartition of free and bound leptin. For this reason, we examined the relationship between plasma soluble leptin receptor levels, plasma leptin levels and the degree of obesity. We also investigated whether SLR concentrations could be modulated by fat mass loss induced by a 3 month weightreducing diet. Methods Patients The GenOlor I is a case control study designed to evaluate the interaction between Genetic factors and ENvironmental determinants of Obesity in LORraine. This study conducted since 1997 in France has been approved by the Ethics Committee of Nancy and all the subjects of the study gave their informed consent. At this time, 213 patients were characterized at the Department of Diabetology, Metabolic and Nutrition Diseases (38 overweight and 112 obese patients) and to the Center of Clinical Investigation (63 lean patients; CHU- INSERM). The obese patients were included on the basis of a BMI (body mass index) ranging from 30 to 45 kg=m 2. The obese individuals were included in the study only if there was at least one obese individual in their first-degree family (siblings, mother, father or children). The lean subjects were included on the basis of a BMI less than 30 kg=m 2. Only individuals having less than one obese subject in their firstdegree family or less than two obese subjects in their seconddegree family were included in this study. Obese individuals with the same degree of adiposity (fat mass) were divided into two sub-groups: obese with high leptin concentration (leptin=fat mass (FM) > 0.66 ng=ml=kg for the men and > 1.16 ng=ml=kg for the women); and obese with low leptin concentration (leptin=fm < 0.26 ng=ml=kg for the men and < 0.60 ng=ml=kg for the women). These limits corresponded to the patient distribution around the average leptin=fm 1 s.d. The lean individuals were divided in two subgroups: lean (BMI < 25 kg=m 2 ) and overweight individuals (25 BMI < 30 kg=m 2 ). During 3 months, seven obese female subjects (age 40 8 y; weight kg) and one overweight female subject (age 39 y; weight 74 kg) followed a low-calorie diet (corresponding to 30% less than their recommended daily calorie value). Obesity-related phenotype BMI was calculated as body weight (kg) divided by the height squared (m 2 ). Body composition was assessed by biphotonic absorptiometry DEXA (dual energy X-ray absorptiometry) using the LUNAR DPX-IQ (Lunar Corporation, Madison WI, USA). Information about age at obesity onset, dieting, exercise habits and medical history were obtained from questionnaires. Plasma samples were obtained after an 8 h fast, in patients with weight stable for 1 month, and 5 days after the end of menstruation for the women. Leptin radioimmunoassay (RIA) Leptin was measured in plasma samples using reagent supplied by Linco Research Inc. (St Louis, MO, USA). Plasma for each subject was analyzed in triplicate using a human leptin polyclonal antibody produced in rabbit. The tracer was human leptin radio-labeled by iode 125 (specific activity < 3 mci). The radioactivity was counted in a gamma counter COBRA II Auto Gamma (Packard Instrument Co. Inc. Meriden, CT, USA). The detection limit of the assay was 0.5 ng=ml and the upper limit of linearity was 100 ng=ml. Soluble human leptin receptor enzyme-linked immunosorbent assay (ELISA) A direct double monoclonal sandwich ELISA was used for the quantitative measurement of the soluble human leptin receptor. Microtitration wells were coated by Biovendor Laboratory Medecine (Czech Republic) with a monoclonal anti-human leptin soluble receptor antibody. This antibody was produced in a mouse immunized with a purified recombinant human leptin receptor extracellular domain. A homodimeric chimeric protein, resulting from the expression in a mouse myeloma cell line of a fusion between a cdna sequence encoding the extracellular domain of the leptin receptor and the Fc region of human IgG (R&D Systems, Minneapolis, MN, USA), was used as standard (4 200 ng=ml). In COS7 cells transiently transfected with the human leptin receptor cdna, this chimeric protein was able to displace human leptin (dissociation constant (K d ) 0.7 nm) with an apparent inhibitory concentration of 50% (IC 50 )of 2 nm. The direct protein protein interaction of human leptin with the chimeric protein was studied using surface plasmon resonance (Biacore) and indicated a K d of 3 nm. Each plasma or standard was diluted 1:3 and incubated in duplicate (100 ml each) on the micro-titration plate (1 h, 20 C, under gentle shaking). The wells were washed six times (350 ml per well). The second monoclonal antibody (100 ml) coupled to horseradish peroxidase was used in the direct sandwich detection of the bound antigen (1 h, 20 C, under gentle shaking). After six washes, hydrogen peroxide 497

3 498 and tetramethylbenzidine were used as horseradish peroxidase substrates. The reaction was stopped after 10 min with an acidic solution of 0.2 M H 2 SO 4. Absorbance was measured spectrophotometrically at 450 nm in a Microplate Reader Benchmark (Bio Rad laboratories, Hercules, CA, USA). The calibration curve was constructed by plotting the absorbance of the standards vs log of the known concentrations using a four parameters function (Microplate Manager 4.0, Bio Rad). The intra-assay coefficients of variation (n ¼ 12) were 3.1 and 2.8% at the concentrations of 27.0 and 58.3 ng=ml, respectively. The inter-assay coefficients of variation (n ¼ 8) were 8.3 and 9.6% at the concentrations of 33.5 and 63.3 ng=ml, respectively. The least detectable concentration was 0.8 ng=ml. Spiking recoveries of samples with different amounts of the recombinant human leptin receptor ranged from 90 to 110% of the spiked amount. Linearity was measured by dilution of the recombinant human leptin receptor (1:1 to 1:8 at two different concentrations) and ranged from 97 to 106% of the initial amount. The effect of human leptin on the soluble human leptin receptor assay were measured by incubating increasing concentrations (4 200 ng=ml) of recombinant human leptin receptor with different concentrations of leptin (0, 10, 40 and 80 ng=ml). No interference of exogenously added human leptin with detection of the soluble human leptin receptor was observed. followed by Fisher protected least significant difference (PLSD) tests. Statistical significance was taken as P < Multivariate correlation analyses were made using linear and non-linear parametric models and total (r) and partial (r) correlation coefficients. Results Circulating leptin levels As expected, human plasma leptin concentrations measured by RIA were correlated with BMI (r ¼ 0.80; P < for the Gel filtration Plasma samples (1.6 ml) from one lean women (age 39 y; BMI 20.8 kg=m 2 ; leptin 14.5 ng=ml) and one obese women (age 38 y; BMI 36.2 kg=m 2 ; leptin 38.7 ng=ml) were incubated with 650 ml of 125 I leptin ( cpm, Linco) for 24 h at 4 C. These plasma were size-fractionated by gel filtration on a Superdex 200 (Amersham Pharmacia Biotech, Saclay, France) column (1.660) on a FPLC (Fast Protein Liquid Chromatography) system with 25 mm phosphate-buffered saline, ph 7.4 used for elution. The non-specific background (12%) was determined by fractionated 125 I leptin ( cpm) in the absence of plasma. Two other plasma samples, from a lean women (age 26 y; BMI: 18.7 kg=m 2 ; leptin 6.9 ng=ml; SLR 46.9 ng=ml) and an obese women (age 49 y; BMI 39.9 kg=m 2 ; leptin 35.8 ng=ml; SLR 34.8 ng=ml) were fractionated in the same conditions and leptin and SLR levels were measured in the eluate fractions (1 ml) by a sensitive human leptin RIA (Linco Research Inc., St Louis, MO, USA) and by the soluble human leptin receptor ELISA. Statistical analyses Data are expressed as mean standard deviation. Normality of data distribution was verified using skewness and kurtosis coefficients. Data analyses were performed using the Stat- View 5 software (ABACUS, CA, USA). Group comparisons (when more than two groups) were made using analysis of variance (ANOVA) and covariance analysis (ANCOVA) and Figure 1 (A) Correlation between plasma leptin levels and percentage body fat (BF) in men (filled circle, thick line; n ¼ 49; leptin ¼ %BF ; r 2 ¼ 0.81; P < ) and in women (open circle, thin line; n ¼ 164; leptin ¼ %BF r 2 ¼ 0.76; P < ). (B) Negative correlation between soluble leptin receptor (SLR) levels and percentage body fat (BF) in all groups (n ¼ 213; SLR ¼ %BF ; r 2 ¼ 0.57; P < ).

4 men and r ¼ 0.75; P < for the women; data not shown). The correlation increased to r ¼ 0.90 for the men and to 0.87 for the women when leptin concentrations were compared with percentage of body fat measured by DEXA (Figure 1A). As shown in Table 1, leptin levels were significantly higher in the obese group ( ng=ml) than in the overweight group ( ng=ml) and the lean group ( ng=ml). The three groups were significantly different when the ratio of leptin to percentage body fat was compared (obese vs overweight, P < ; obese vs lean, P < and overweight vs lean, P < 0.01), but this difference existed only for the obese vs lean groups (P < 0.01) when we compared the ratio of leptin per unit of fat mass. Sex differences were observed: if mean BMI values were the same in women and in men (P ¼ 0.58), plasma leptin concentrations were significantly higher in women ( ng=ml) than in men ( ng=ml, P < ). Even if the percentage of body fat was higher in women than in men (41 11 vs 27 10%), the gender difference observed remained after adjustment of leptin levels on percentage of body fat. As observed in Figure 1A, for the same percentage of body fat a wider distribution of leptin levels was observed in the obese group (leptin levels ranged from 5.2 to 82.4 ng=ml). For this reason, the obese group was partitioned into two subgroups: obese with high leptin concentration (leptin=fm > 0.66 ng=ml=kg for the men and > 1.16 ng=ml=kg for the women) and obese with low leptin concentration (leptin=fm < 0.26 ng=ml=kg for the men and < 0.60 ng=ml=kg for the women), with similar percentage of body fat (P ¼ 0.73). Circulating soluble human leptin receptor levels We measured the circulating SLR in the plasma using a sensitive and specific direct sandwich immuno-assay using monoclonal antibodies against the human leptin receptor and a recombinant chimeric human leptin receptor IgG fusion protein. The chimeric protein demonstrated an affinity constant to leptin similar to this of leptin to the leptin receptor (0.7 vs 3 nm, respectively). Recovery and linearity analyses were performed using the recombinant chimeric protein spiked into human blood samples. The average recovery for different concentrations was 96% or greater when compared with controls. When fixed amounts were added and serially diluted, the sample dilution values closely paralleled the recombinant standard dilution values. Together, these data indicated that this immuno-assay measured the SLR without significant interference from potential binding proteins present in the plasma. Further analysis of human leptin receptor assay interactions was made in the presence of exogenously added leptin (0, 10, 40 and 80 ng=ml). We found SLR concentrations between 88 and 101% of the initial concentrations for the three concentrations of leptin added. Altogether, these data suggested that the present human leptin receptor immuno-assay measures total human soluble leptin receptor protein levels. When measured in the plasma of the 213 patients, circulating SLR levels ranged from 10 to 100 ng=ml (Table 1). Circulating SLR levels were higher in men ( ng=ml) than in women ( ng=ml). In contrast to the changes observed for leptin, circulating SLR levels were significantly higher in the lean group ( ng=ml) than in the overweight group ( ng=ml) and in the obese group ( ng=ml; Table 1). Accordingly, the ratio of circulating SLR level per unit of fat mass was decreased in the obese group ( ng=ml=kg) when compared with the lean group ( ng=ml=kg). The difference in SLR levels is significant even after adjustment on fat mass. Interestingly, pronounced sex difference were observed in overweight and obese but not in lean individuals: circulating SLR levels were significantly lower in obese ( ng=ml) and overweight ( ng=ml) women when compared with obese ( ng=ml) and overweight men ( ng=ml; P < ). Although total body fat was higher in lean women than in lean men (15 5 vs 11 4 kg, respectively; P < 0.01), no differences were found in circulating SLR levels between lean women ( ng=ml) and 499 Table 1 Biological data for all the patients Patient Age (y) BMI (kg=m 2 ) Fat mass (kg) Body fat (%) Leptin (ng=ml) SLR (ng=ml) Leptin=SLR Obese n ¼ 23 male A 43 8 A 45 7 A A A A n ¼ 89 female (18 50) ( ) (21 60) (24 57) ( ) ( ) ( ) Overweight n ¼ 10 male B 27 7 B B B B B n ¼ 28 female (19 48) ( ) (10 38) (12 46) ( ) ( ) ( ) Lean n ¼ 16 male C 14 5 C 24 7 C C C C n ¼ 47 female (18 50) ( ) (4 27) (7 42) ( ) ( ) ( ) Means s.d. Different superscript letters indicate significant differences (ANOVA) at P < , except for the overweight and lean groups which were significantly different at P < and P < 0.01 for leptin and leptin=soluble leptin receptor (SLR) levels, respectively.

5 500 men ( ng=ml; P ¼ 0.61). Circulating SLR levels were negatively correlated with the percentage of body fat (r ¼ 70.76; P < ; Figure 1B). Consequently, leptin levels and circulating levels of the SLR were negatively correlated (r ¼ 70.74; P < ; Figure 2A). Leptin is an independent predictor of SLR levels as after neutralization of the effect of percentage body fat, the correlation between SLR and leptin levels is still significant (r ¼ 70.21; P < 0.01). Interestingly, the ratio of leptin to SLR levels was strongly correlated to the percentage body fat (r ¼ 0.91; P < ; Figure 2B). Biometric parameters for the two obese groups (obese with high or low leptin levels) are reported in Table 2. For the two groups the fat mass was not different (P ¼ 0.73), but leptin levels were 3-fold higher in the obese high leptin ( ng=ml) than in the obese low leptin ( ng=ml) group. In these two groups we observed a significant difference in circulating SLR levels (P < 0.05). The obese high leptin group ( ng=ml) had lower SLR levels than the obese low leptin group ( ng=ml). The SLR levels of the patients on a low-calorie diet for 3 months were increased when compared with their initial value. Thus, we could observe a rise in SLR levels when weight loss and fat mass loss occurred (Table 3). Although the number of patients was small, the gain in SLR was positively correlated to the loss of fat mass (r ¼ 0.76; P < 0.05; Figure 3). Separation of bound and free leptin When 125 I leptin was incubated with human plasma and fractionated on a gel filtration column, two peaks of radioactivity were observed (Figure 4A). Peak I eluted at the void volume of the column and represents bound leptin. Peak II corresponds to unbound or free leptin since it eluted at the same position as native 125 I leptin. In order to estimate the real quantity of bound and free leptin we measured immunoreactive leptin in fractions of plasma of lean and obese subjects. We found that 37% of the total leptin was bound to macromolecules in lean and only 15% in the obese subjects (Figure 4B). A higher proportion of leptin circulated in bound form in lean compared with obese subjects. Having found that plasma levels of SLR were higher in lean than in obese individuals, we measured the SLR in the fractionated plasma of lean and obese subjects. We found that the SLR coeluted exactly with the bound leptin fractions. This observation is another indication that the SLR might be one of the leptin binding proteins. Also, we observed that the areas under the curve of SLR were higher in lean than in obese subjects. Figure 2 (A) Negative correlation between soluble leptin receptor (SLR) levels and plasma leptin levels in all groups (n ¼ 213; SLR ¼ leptin ; r 2 ¼ 0.56; P < ). (B) Correlation between plasma leptin-to-slr ratio and percentage body fat (BF) in the all groups (n ¼ 213; leptin=slr ¼ % BF ; r 2 ¼ 0.83; P < ). Table 2 Biological data for the obese high leptin and the obese low leptin groups Patient Fat mass (kg) Leptin (ng=ml) SLR (ng=ml) Leptin=SLR Obese high leptin n ¼ 4 male { * { n ¼ 10 female (32 56) ( ) ( ) ( ) Obese low leptin n ¼ 4 male n ¼ 8 female (31 56) ( ) ( ) ( ) Means s.d. Obese high leptin: leptin=fat mass (FM) > 0.66ng=ml=kg for the men and > 1.16 ng=ml=kg for the women. Obese low leptin: leptin=fm < 0.26 ng=ml=kg for the men and < 0.60 ng=ml=kg for the women). *P < 0.05; { P < 0.01; { P <

6 Table 3 diet Biological data for the seven obese and one overweight patients before and after 3 months on a low-calorie Body weight (kg) BMI (kg=m 2 ) Fat mass (kg) Leptin (ng=ml) SLR (ng=ml) 501 Before weight loss (n ¼ 8) * (74 108) ( ) (32 53) ( ) ( ) After weight loss (n ¼ 8) (67 100) ( ) (26 48) ( ) ( ) Low-calorie diet corresponded to 30% less than their recommended daily calorie intake (mean s.d.); *P < Figure 3 Correlation between soluble leptin receptor (SLR) gain and fat mass (FM) loss in subjects after a 3 month low-calorie diet (n ¼ 8; SLR gain ¼ 2.43 FM loss70.04; r 2 ¼ 0.58; P < 0.05). Discussion To our knowledge, this is the first study to measure direct human soluble leptin receptor concentrations in a human population of obese, overweight and lean subjects. The present results support the presence of the SLR in human plasma, as observed in another study measuring SLR levels in normal and diabetic pregnancies by RIA. 26 Interestingly we observed that, although leptin levels were increased in women compared with men, plasma SLR concentrations were not different in lean men and women, but were increased in obese and overweight men compared to obese and overweight women. Previous measurement of leptin binding activity (LBA) in human serum have detected a high affinity binding protein which affinity is compatible with the soluble human leptin receptor. 22 However, no difference in LBA between sexes in this human population was observed. These discrepancies can be explained by two facts. Firstly, we measured the total soluble leptin receptor Figure 4 (A) Gel filtration elution profiles of 125 I leptin alone (open circle) or after incubation with plasma from either an obese individual (filled triangle) or a lean individual (open triangle). Peak I corresponds to bound leptin and peak II corresponds to free leptin. (B) Gel filtration elution profiles of immunoreactive leptin in plasma sample from one obese subject (filled circle) and one lean subject (open circle) and soluble leptin receptor (SLR) in plasma sample from one obese subject (filled triangle) and one lean subject (open triangle).

7 502 concentrations and Quinton et al 22 measured leptin binding activity (only a fraction of which might be the soluble leptin receptor). Secondly, we observed sex differences in the obese and overweight subjects but not in lean subjects in our population, and their population was a mixture of lean and overweight individuals (BMI < 30 kg=m 2 ). Therefore, sex differences in SLR levels in obese and overweight patients could be explained by differences in fat mass. However, no differences in SLR levels were observed between lean women and lean men despite their different fat masses, indicating that SLR levels might be modulated differently in women than in men. In the present study using gel filtration studies, we also demonstrated that leptin circulates in human plasma in a free form and in a bound form, in agreement with all the other published studies. By immuno-assay measurement in the gel filtration fractions, we reported that the SLR coeluted with the bound fraction of leptin. Recent studies have demonstrated that a relationship exists between percentage of free leptin, bound leptin and fat mass in humans. These studies indicate the existence of a negative correlation between percentage of bound leptin and percentage body fat. However, there were no differences between lean and obese patients regarding the absolute levels of bound leptin. 18 In the present study, the direct measurement of the concentrations of the soluble leptin receptor in lean, overweight and obese humans allowed us to establish if such a relationship between SLR, leptin and fat mass might exist. Interestingly, SLR levels were higher in lean than overweight and obese patients. In contrast to SLR levels, leptin levels were about 1.7- and 3.5-fold higher in obese patients than in the overweight and lean subjects, respectively. Therefore, the negative correlation observed between SLR and leptin levels was not unexpected. We have also observed that the total SLR levels were negatively correlated to percentage body fat and this is in good agreement with the negative correlation founded between the percentage of bound leptin and the percentage of body fat reported previously. 18 On the other hand, observation that obese patients with high leptin concentrations have less SLR than obese patients with low leptin concentrations indicated that a close link exists between SLR and leptin concentrations for a same degree of obesity. To further strengthen the link between SLR and change in adiposity, we measured SLR levels in obese and overweight patients before and after a 3 month low-calorie diet. Interestingly, when fat mass decreases, SLR levels increase. Moreover, we found that the SLR increase after weight loss positively correlated with the loss of fat mass. This result clearly indicates that the levels of SLR are closely related to the degree of adiposity and that experimental decrease in adiposity is reflected accordingly in changes in SLR levels. Thus the reduction of the SLR levels may be secondary to the development of obesity since it is reversed by fat mass loss. In the present study, low SLR levels were associated with high leptin levels and this may be an epiphenomenom that reflects the down-regulation of leptin receptors by high leptin levels. Indeed, a down-regulation of leptin receptor isoforms by leptin in vitro has been demonstrated. Exposure to leptin induces a down-regulation of the accessible leptin binding sites in CHO cell lines expressing both OB-Ra and OB-Rb. 27 Similarly, the expression levels of the OB-Rb isoform are negatively regulated by leptin in neuroblastoma cells in a dose-dependant manner. 28 On the other hand, recent in vivo reports indicate that high leptin levels are associated with high SLR levels, leading to the idea of a stabilization of leptin by the soluble form of its receptor. In animals, Gavrilova et al demonstrated that during late-stage mouse pregnancy, extreme hyperleptinemia is attributable to the binding of leptin to a secreted form of the leptin receptor made by the placenta, leading to the hypothesis that it decreases leptin clearance. 20 Huang et al demonstrated that in Zucker diabetic fatty rats, the high level of circulating leptin is correlated with an increased leptin expression in adipose tissue and a stabilization of leptin by its binding to the soluble receptor. In both cases, the soluble leptin receptor levels were up-regulated by 20- fold. 29 In humans, a mutation in the leptin receptor gene leads to a truncated receptor that lacks transmembrane and cytoplasmic domains and is secreted into the circulation. This soluble receptor binds the majority of leptin and increases bound and total leptin levels. 24 Whether or not the SLR stabilizes leptin is not addressed in our study, but we have observed that high leptin levels were not associated with high SLR levels. Therefore, in the absence of leptin signaling (in Zucker diabetic fatty rats or in humans with a mutated leptin receptor gene), this down-regulation may not occur. The existence or the absence of a functional leptin receptor signalling pathway would explain why, in some cases, high leptin levels are associated with low or high SLR levels, respectively. In conclusion we have observed that the soluble leptin receptor is present in human plasma and that soluble leptin receptor levels are inversely correlated to leptin levels and the degree of obesity. Moreover, we have also observed that, during weight loss, soluble leptin receptor levels increased proportionally to the decrease of fat mass. These results suggest that high levels of soluble leptin receptor may enhance leptin action in lean subjects more than in obese subjects. Acknowledgements The technical assistance of Francine Michel is gratefully acknowledged. The GenOlor study was financially supported by Servier. Virginie Ogier is a recipient of a grant from the Fondation pour la Recherche Médicale. 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 1994; 372: Pelleymounter MA, Cullen MJ, Baker MB, Hecht R, Winters D, Boone T, Collins F. Effects of the obese gene product on body weight regulation in ob=ob mice. Science 1995; 269:

8 3 Halaas JL, Gajiwala KS, Maffei M, Cohen SL, Chait BT, Rabinowitz D, Lallone RL, Burley SK, Friedman JM. Weight-reducing effects of the plasma protein encoded by the obese gene. Science 1995; 269: 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 1995; 269: Tartaglia LA, Dembski M, Weng X, Deng N, Culpepper J, Devos R, Richards GJ, Campfield LA, Clark FT, Deeds J et al. Identification and expression cloning of a leptin receptor, OB-R. Cell 1995; 83: Considine RV, Considine EL, Williams CJ, Nyce MR, Zhang P, Opentanova I, Ohannesian JP, Kolaczynski JW, Bauer TL, Moore JH, Caro JF. Mutation screening and identification of a sequence variation in the human ob gene coding region. Biochem Biophys Res Commun 1996; 220: Montague CT, Farooqi IS, Whitehead JP, Soos MA, Rau H, Wareham NJ, Sewter CP, Digby JE, Mohammed SN, Hurst JA, Cheetham CH, Earley AR, Barnett AH, Prins JB, O Rahilly S. Congenital leptin deficiency is associated with severe early-onset obesity in humans. Nature 1997; 387: Strobel A, Issad T, Camoin L, Ozata M, Strosberg AD. A leptin missense mutation associated with hypogonadism and morbid obesity. Nature Genet 1998; 18: Considine RV, Considine EL, Williams CJ, Hyde TM, Caro JF. The hypothalamic leptin receptor in humans: identification of incidental sequence polymorphisms and absence of the db=db mouse and fa=fa rat mutations. Diabetes 1996; 45: Clement K, Vaisse C, Lahlou N, Cabrol S, Pelloux V, Cassuto D, Gourmelen M, Dina C, Chambaz J, Lacorte JM, Basdevant A, Bougneres P, Lebouc Y, Froguel P, Guy-Grand B. A mutation in the human leptin receptor gene causes obesity and pituitary dysfunction. Nature 1998; 392: Maffei M, Halaas J, Ravussin E, Pratley RE, Lee GH, Zhang Y, Fei H, Kim S, Lallone R, Ranganathan S et al. Leptin levels in human and rodent: measurement of plasma leptin and ob RNA in obese and weight-reduced subjects. Nature Med 1995; 1: Considine RV, Sinha MK, Heiman ML, Kriauciunas A, Stephens TW, Nyce MR, Ohannesian JP, Marco CC, McKee LJ, Bauer TL et al. Serum immunoreactive-leptin concentrations in normalweight and obese humans. New Engl J Med 1996; 334: Lee GH, Proenca R, Montez JM, Carroll KM, Darvishzadeh JG, Lee JI, Friedman JM. Abnormal splicing of the leptin receptor in diabetic mice. Nature 1996; 379: Baumann G, Amburn KD, Buchanan TA. The effect of circulating growth hormone-binding protein on metabolic clearance, distribution, and degradation of human growth hormone. J Clin Endocrinol Metab 1987; 64: Baumann G, Shaw MA, Buchanan TA. In vivo kinetics of a covalent growth hormone-binding protein complex. Metabolism 1989; 38: Heaney ML, Golde DW. Soluble hormone receptors. Blood 1993; 82: Muller-Newen G, Kohne C, Heinrich PC. Soluble receptors for cytokines and growth factors. Int Arch Allergy Immunol 1996; 111: Sinha MK, Opentanova I, Ohannesian JP, Kolaczynski JW, Heiman ML, Hale J, Becker GW, Bowsher RR, Stephens TW, Caro JF. Evidence of free and bound leptin in human circulation. Studies in lean and obese subjects and during short-term fasting. J Clin Invest 1996; 98: Diamond FB Jr, Eichler DC, Duckett G, Jorgensen EV, Shulman D, Root AW. Demonstration of a leptin binding factor in human serum. Biochem Biophys Res Commun 1997; 233: Gavrilova O, Barr V, Marcus-Samuels B, Reitman M. Hyperleptinemia of pregnancy associated with the appearance of a circulating form of the leptin receptor. J Biol Chem 1997; 272: Eichler DC, Root AW, Duckett G, Moore KL, Diamond FB. A spuncolumn assay for determination of leptin binding in serum. Anal Biochem 1999; 267: Quinton ND, Smith RF, Clayton PE, Gill MS, Shalet S, Justice SK, Simon SA, Walters S, Postel-Vinay MC, Blakemore AI, Ross RJ. Leptin binding activity changes with age: the link between leptin and puberty. J Clin Endocrinol Metab 1999; 84: McConway MG, Johnson D, Kelly A, Griffin D, Smith J, Wallace AM. Differences in circulating concentrations of total, free and bound leptin relate to gender and body composition in adult humans. Ann Clin Biochem 2000; 37: Lahlou N, Clement K, Carel JC, Vaisse C, Lotton C, Le Bihan Y, Basdevant A, Lebouc Y, Froguel P, Roger M, Guy-Grand B. Soluble leptin receptor in serum of subjects with complete resistance to leptin: relation to fat mass. Diabetes 2000; 49: Liu C, Liu XJ, Barry G, Ling N, Maki RA, De Souza EB. Expression and characterization of a putative high affinity human soluble leptin receptor. Endocrinology 1997; 138: Lewandowski K, Horn R, O Callaghan CJ, Dunlop D, Medley GF, O Hare P, Brabant G. Free leptin, bound leptin, and soluble leptin receptor in normal and diabetic pregnancies. J Clin Endocrinol Metab 1999; 84: Uotani S, Bjorbaek C, Tornoe J, Flier JS. Functional properties of leptin receptor isoforms: internalization and degradation of leptin and ligand-induced receptor downregulation. Diabetes 1999; 48: Hikita M, Bujo H, Hirayama S, Takahashi K, Morisaki N, Saito Y. Differential regulation of leptin receptor expression by insulin and leptin in neuroblastoma cells. Biochem Biophys Res Commun 2000; 271: Huang L, Wang Z, Li C. Modulation of circulating leptin levels by its soluble receptor. J Biol Chem 2001; 276: