leptin mouse experiment

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 13 October 2020

Leptin alters energy intake and fat mass but not energy expenditure in lean subjects

• Pavlina Chrysafi 1   na1 ,
• Nikolaos Perakakis   ORCID: orcid.org/0000-0002-2319-6603 1   na1 ,
• Olivia M. Farr 1 ,
• Konstantinos Stefanakis 1 ,
• Natia Peradze 1 ,
• Aleix Sala-Vila 2 , 3 &
• Christos S. Mantzoros   ORCID: orcid.org/0000-0003-3755-8158 1  

Nature Communications volume  11 , Article number:  5145 ( 2020 ) Cite this article

12k Accesses

51 Citations

37 Altmetric

Metrics details

• Endocrine system and metabolic diseases

Based on studies in mice, leptin was expected to decrease body weight in obese individuals. However, the majority of the obese are hyperleptinemic and do not respond to leptin treatment, suggesting the presence of leptin tolerance and questioning the role of leptin as regulator of energy balance in humans. We thus performed detailed novel measurements and analyses of samples and data from our clinical trials biobank to investigate leptin effects on mechanisms of weight regulation in lean normo- and mildly hypo-leptinemic individuals without genetic disorders. We demonstrate that short-term leptin administration alters food intake during refeeding after fasting, whereas long-term leptin treatment reduces fat mass and body weight, and transiently alters circulating free fatty acids in lean mildly hypoleptinemic individuals. Leptin levels before treatment initiation and leptin dose do not predict the observed weight loss in lean individuals suggesting a saturable effect of leptin. In contrast to data from animal studies, leptin treatment does not affect energy expenditure, lipid utilization, SNS activity, heart rate, blood pressure or lean body mass.

Similar content being viewed by others

The role of leptin and low testosterone in obesity

Obesity-induced and weight-loss-induced physiological factors affecting weight regain

Metabolic surgery-induced changes of the growth hormone system relate to improved adipose tissue function

Introduction.

Leptin, the prototypical adipokine, circulates at levels proportional to fat mass 1 , 2 and responds to acute changes in energy intake 3 , 4 , 5 .

On the basis of experiments in ob/ob and lean mice, leptin was thought to effectively cause weight loss by regulating appetite, energy expenditure, sympathetic nervous system (SNS) activity, lipolysis, and lipid–carbohydrate utilization 6 , 7 , 8 . In contrast, leptin administration has been far less effective in animal models of obesity with leptin excess 9 . In humans, with the exception of severe leptin deficiencies due to leptin mutations (congenital leptin deficiency, CLD) or lipodystrophies (generalized (GL) or partial lipodystrophies (PL)) 10 , 11 , 12 , 13 , the majority of studies in overweight or obese populations with hyperleptinemia showed minimal if any effects of leptin treatment on weight or body composition 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 . Recent research efforts focus therefore on the identification of threshold leptin concentrations in the blood, below which a treatment with leptin may be effective by leading to significant weight loss among people with obesity 23 . Additionally, due to the poor efficacy of leptin treatments on weight regulation in obesity, it has been questioned whether leptin does act to reduce body weight in humans, emphasizing the need to investigate the physiology of leptin and its effects on metabolic outcomes in lean individuals who may be more likely to respond to leptin administration 24 , 25 .

We have previously reported no effect of short-term leptin treatment on weight loss during acute 72-h fasting (studies 1 and 2) (Fig.  1 for study design) 26 , 27 . In contrast, we observed a significant reduction of weight in lean chronic and partial hypoleptinemic women due to strenuous exercise (studies 3 and 4) treated with leptin from 8 and up to 36 weeks (Fig.  1 for study designs) 28 , 29 . Here, we perform new measurements and analyze data from our previous studies 26 , 27 , 28 , 29 , 30 to (1) assess whether circulating concentrations of leptin before treatment initiation are associated with the weight loss observed in our study subjects, (2) investigate whether any effects of leptin either in the short term during fasting and/or in the fed state are dose-dependent, by employing physiological vs supraphysiological vs pharmacological doses of leptin, and explore whether such effects may differ between lean men, lean women, and obese men, (3) investigate the trajectories of weight and fat mass changes in relation to leptin levels during long-term leptin treatment and after its termination, and (4) determine the potential underlying mechanistic pathways through which leptin affects the physiology of energy homeostasis in lean subjects by testing energy intake (measured caloric intake using ad libitum feeding), energy expenditure (physical activity and resting metabolic rate (RMR) measurements), SNS outputs (heart rate (HR), blood pressure (BP), catecholamine levels, and the renin–angiotensin–aldosterone system (RAAS)), fuel utilization, and metabolite–lipid–lipoprotein profile.

Study 1: Eight healthy lean men and seven healthy lean women were studied under three separate Clinical Research Center (CRC)-based conditions for 72 h: one under isocaloric fed state conditions (normoleptinemia) and two during complete fasting state conditions (induced hypoleptinemia) scheduled in a random order and in a double-blind fashion with administration of physiologic replacement leptin doses (fasting + leptin) or placebo (fasting + placebo). Study 2: Five lean men, five men with obesity, and five lean women participated in three fed-normoleptinemic and three fasting-induced hypoleptinemic studies, which were conducted in the CRC, with leptin administration at three different doses (Dose A = 0.01 mg/kg, Dose B = 0.1 mg/kg, Dose C = 0.3 mg/kg). Study 3: Open-label long-term leptin treatment in mildly hypoleptinemic women. Study 4: Placebo-controlled long-term leptin treatment in mildly hypoleptinemic women.

Baseline leptin levels do not predict weight-loss magnitude

In study 1 (three interventions: fed—untreated, fasting—treated with leptin, and fasting—treated with placebo) and study 2 (six interventions: fed and fasting state treated with physiological, supraphysiological, or pharmacological doses of leptin) (see Fig.  1 for study designs), lean men had consistently lower levels of leptin before each intervention compared to lean women, whereas obese men had similar leptin levels to lean women (study 2) (Fig.  2a, b , left and Supplementary Figs.  2 and 3 ). In both studies and across all interventions, the % body weight change was similar between lean men and women (Fig.  2a, b , middle) and lower in obese men (Fig.  2b , middle). In study 1, when men and women were investigated both together and separately, there was no association or trend (all P values > 0.2) between leptin levels in the blood at baseline (before treatment initiation) (Fig.  2a , right) and % weight loss after 72-h fasting treated with placebo or leptin. Similarly, in study 2, although the sample size per group was small ( n  = 5), no strong association of leptin levels at baseline with % weight loss was observed for obese and lean men, when values from all doses were combined, both before (Fig.  2b , right) and after adjusting for intraindividual variability (i.e., for contribution of three values in the correlation by each subject) (Supplementary Fig.  1a , left and middle). For lean women, there was an association of lower % weight loss with higher baseline leptin values (with no clear cut-off levels) before (Fig.  2b , right) but not after adjusting for intraindividual variability (Supplementary Fig.  1a , right). Finally, the % weight loss was not affected by leptin dose in any of the three groups (lean men, lean women, and obese men) (Supplementary Fig.  1b ).

a Cross-over study of lean subjects during 72-h fed state, fasting+placebo and fasting+leptin (study 1; n  = 13). Left: baseline leptin levels in each admission. Center: % weight change at the end of each admission. Right: correlation of baseline leptin with % weight change at the end of each admission. Numbers above bars correspond to subject ID. P values of unpaired t test between lean men (LM) vs lean women (LW) and of correlations are reported; R , correlation coefficient. b Cross-over study of LM, LW, and obese men (OM) in 72-h fasting treated with escalating leptin doses (study 2; n  = 15). Left: baseline leptin levels in each admission, Center: % weight change at the end of each admission. Right: correlation of baseline leptin with % weight change at the end of each admission. Numbers above bars correspond to subject ID. P values from one-way ANOVA, from post hoc Bonferroni test between LM vs LW vs OW and from correlations are reported. c Open-label (study 3; n  = 7) and placebo-controlled long-term leptin treatment study (study 4; n  = 19 (leptin = 10; placebo = 9)) in women with mild hypoleptinemia. Left: correlation of baseline leptin with % weight change after 8 weeks of leptin treatment. Subjects of study 3 were combined with leptin-treated subjects of study 4 in one analysis. Center and right: changes of leptin, body weight, and fat mass from baseline (Δ = change from baseline at each timepoint). In study 4, dashed lines correspond to the washout period after 36 weeks of study. In study 3, P values ( P ) for time effect (i.e., days of treatment) and in study 4, P values of G (group: leptin or placebo), T (time: weeks of treatment), and G * T interaction of mixed models adjusted for baseline are reported. By P  < 0.05 (study 3) and G * T  < 0.05 (study 4), post hoc Bonferroni test was performed (only significant results are reported). One, two, or three asterisks indicate P  < 0.05, <0.01, or <0.001 for the specific timepoint vs baseline in study 3 and for leptin vs placebo in the specific timepoint in study 4. Correlations were performed with Pearson’s or Spearman’s correlation test. Data are presented as means ± SEMs. Exact P values : b , left: leptin at baseline dose 0.3 mg/kg = 0.003 × 10 −1 ; leptin at baseline dose 0.3 mg/kg LM vs LW = 0.003 × 10 −1 . c Center : leptin = 0.003 × 10 −1 ; weight = 0.001 × 10 −3 ; fat mass = 0.002 × 10 −5 ; leptin post hoc test at 8 1/2 weeks = 0.031 and at 13 weeks = 0.003 × 10 −1 . Weight post hoc test at 6 1/2 weeks = 0.002, at 8 1/2 weeks = 0.001 × 10 −1 , at 11 weeks = 0.001 × 10 −1 , and at 13 weeks = 0.001 × 10 −1 . Fat mass post hoc test at 6 1/2 weeks = 0.002, at 8 1/2 weeks = 0.002 × 10 −2 , at 11 weeks = 0.005 × 10 −3 , and at 13 weeks = 0.004 × 10 −4 . c , Right : leptin G  = 0.005 × 10 −2 ; T  = 0.007 × 10 −5 ; G * T  = 0.002 × 10 −5 ; body weight T  = 0.007 × 10 −5 ; fat mass T  = 0.009 × 10 −5 ; G * T  = 0.001 × 10 −2 ; leptin post hoc test at 4 weeks = 0.039; at 8 weeks = 0.001; at 12 weeks = 0.004 × 10 −1 ; at 16 weeks = 0.005 × 10 −3 ; at 20 weeks = 0.003 × 10 −2 ; at 24 weeks = 0.001 × 10 −2 ; at 28 weeks = 0.002 × 10 −3 ; at 32 weeks = 0.004 × 10 −3 ; at 36 weeks = 0.002 × 10 −3 . Fat mass post hoc test at 24 weeks = 0.002 × 10 −1 ; at 36 weeks = 0.001 × 10 −1 .

In studies 3 and 4 (long-term leptin replacement in mildly hypoleptinemic women, see Fig.  1 for study design), women had generally low body fat % (mean with min, max % for study 3 = 22.7 (17.7, 28), study 4 = 22.4 (14.3, 29.9)) but they all had stable weight in the last 6 months prior to study participation. Additionally, none of the women in study 3 and only three women all from the placebo group in study 4 were underweight (BMIs 17.6, 18.1, and 18.4 kg/m 2 ). Leptin levels of the women in studies 3 and 4 were low (mean ± SE, study 3 = 3.4 ± 0.7, study 4 = 4.3 ± 0.4 ng/ml), but generally higher than the leptin levels of patients with generalized lipodystrophy (1.3 ± 0.3 ng/ml) 31 , thus representing a model of milder, acquired partial leptin deficiency.

Baseline leptin levels before treatment initiation spanned in both studies between 1.5 and 8 ng/ml. In agreement with our observations in the short-term fasting studies, baseline leptin levels were not associated with % weight loss during the first 8 weeks of leptin treatment (Fig.  2c , left), during which women both in studies 3 and 4 were treated with the same dose of leptin (0.08 mg/kg/day). In contrast, weight and fat mass decreased in parallel in response to the increasing levels of leptin in both studies (Fig.  2c , middle and right). In study 4, the weight changes were stabilized on week 12, when leptin dose adjustments per study protocol occurred in order to prevent further weight loss, whereas interestingly fat mass loss continued up to the 36th week (Fig.  2c , right). Of note, both in studies 3 and 4, no lean mass loss was observed (Supplementary Fig.  1c, d ). Weight and fat mass partially reverted toward baseline after discontinuation of leptin treatment and return of leptin to pretreatment levels (Fig.  2c , right).

Altogether, across multiple studies, baseline leptin did not predict the % weight loss observed after leptin administration in lean populations. Lean men and women demonstrate similar % weight loss during acute fasting, which is independent from leptin dose. Obese men show less % weight loss compared to lean individuals during acute fasting, which is, similar to lean population, independent from baseline leptin levels and leptin dose. During long-term leptin treatment, the increasing leptin levels in women with partial acquired hypoleptinemia are associated with a parallel reduction of fat mass and consequently weight, which are both reversible after treatment termination. Next, we aimed to investigate how leptin regulates body weight in lean individuals by investigating the effects of leptin on parameters related to energy intake, energy expenditure, lipolysis, and lipid utilization.

Leptin affects energy intake but not energy expenditure

The lack of a weight-regulatory effect by leptin treatment during short-term fasting in lean individuals indicates no impact of leptin on energy balance in the short term, and when any potential effects of leptin on energy intake are experimentally controlled with the imposed fasting. To assess now the effects of leptin on energy intake in this experimental setting, an ad libitum meal was offered at the end of the 72-h fasting treated with placebo or leptin, as well as at the end of a 72-h isocaloric fed state. In this meal, higher caloric intake was observed after fasting treated with placebo compared to fed state, but this was partially reduced (−17.3%) and was closer to fed-state levels after fasting treated with leptin (Fig.  3a , left). Furthermore, leptin levels directly before meal intake correlated negatively, curvilinearly, and strongly with calorie intake ( r  = −0.644, P  < 0.001, Fig.  3a right), with an inflection point approximately at 10 ng/ml, when all values from the three interventions were combined (unadjusted), as well as when we have adjusted for treatment with placebo or leptin during fasting (after adjustment for intra-subject variability, r  = −0.569, P  = 0.042, Supplementary Fig.  1e ).

a Energy intake after 72-h in fed state, fasting+placebo, or fasting + leptin (study 1, n  = 13). P values from repeated measure ANOVA and post hoc Bonferroni test are reported. R, correlation coefficient. b Expected (based on leptin-induced caloric deficit in study 1) vs observed fat mass loss during long-term leptin treatment in studies 3 and 4. As per the protocol, in study 4, if a subject lost >5% of baseline weight, the dose was reduced by 0.04 mg/kg (red arrows). P values of G (group: observed or expected fat mass change), T (days/weeks of treatment), and G * T interaction of mixed models are reported. By G * T  < 0.05, post hoc Bonferroni test was performed: two asterisks indicate P  < 0.01 for observed vs expected fat mass change at the specific timepoint. c Energy expenditure during 72-h fed state, fasting+ placebo, and fasting + leptin (study 1, n  = 13). P values of G (group: fed, fasting + placebo, fasting + leptin), T (days of treatment), and G * T interaction of mixed models adjusted for baseline are reported. Post hoc Bonferroni test was performed between the estimated means of the three groups and between the three groups at each timepoint. Three asterisks indicate P  < 0.001 for fed vs fasting + placebo (red) and fed vs fasting + leptin (blue). d Energy expenditure during 72-h fasting treated with escalating leptin doses (study 2, n  = 15). P values of D (dose: 0.01, 0.1, and 0.3 mg/kg/d), T (hours of fasting), and D * T interaction of mixed models adjusted for baseline are reported. e Energy expenditure during open-label long-term leptin treatment in mildly hypoleptinemic women (study 3, n  = 7). P values of paired t test (RMR and body temperature) and of time effect of mixed models adjusted for baseline (exercise score) are reported. No post hoc test was performed since P  > 0.05. f Energy expenditure during placebo-controlled long-term leptin treatment in mildly hypoleptinemic women (study 4, n  = 19 (leptin = 10; placebo = 9)). P values of G (group: placebo or leptin), T (weeks of study), and G * T interaction of mixed models adjusted for baseline are reported. No post hoc test was performed since G * T  > 0.05. All P values are two-sided. For post hoc Bonferroni test, only significant results are reported. Data are demonstrated as means ± SEMs. Exact P values: a Correlation of food intake with leptin prior to meal = 0.002 × 10 −2 . b Study 4 : post hoc test at 12 weeks = 0.002; at 24 weeks = 0.001; at 36 weeks = 0.003. c Day 3 fed vs Pl = 0.004 × 10 −3 ; fed vs Le = 0.003 × 10 −1 . d Temperature T  = 0.002 × 10 −12 .

In the long-term leptin studies, we have not performed an ad libitum meal to assess food intake. Nevertheless, presuming that the ~18% caloric deficit during a test meal that we detected with leptin treatment compared to placebo during short-term fasting in women of study 1 persists with long-term leptin administration in women of studies 3 and 4, we have projected the expected fat mass loss due to reduced energy intake and compared it with the observed (true) fat mass loss during treatment. For these projections, we have used information about the average food intake of the subjects in studies 3 and 4 that were collected by self-report questionnaires at screening. In study 3, where leptin dose was increased in women with no ovulation after 2 months of treatment from 0.08 mg/kg/day to 0.2 mg/kg/day and was not adjusted according to weight changes, the expected fat mass loss due to reduced energy intake is almost identical to the observed one (Fig.  3b , left). In study 4, total fat mass was assessed for the first time after 3 months of leptin treatment in a stable dose, which was much lower in the third month compared to the dose of study 3. Additionally, in study 4, the increase of leptin dose after the third month of treatment was much smaller compared to study 3, whereas in many participants, a dose reduction was necessary in order to prevent further weight loss per study protocol. Given the expected plateauing of leptin effects with time, as with most weight loss treatments, a smaller fat mass loss is observed than the one expected based on the ~18% daily caloric deficit created by leptin, but subjects in the leptin treatment group still lost 1.3 kg of fat at 12 weeks, 3.0 kg at 24 weeks, and 3.8 kg of fat up to 36 weeks of leptin treatment (Fig.  3b , right).

Next, we assessed whether markers of energy expenditure are affected by leptin treatment. RMR and/or body temperature did not change: (a) by short-term leptin treatment compared to placebo in lean people during acute fasting (Fig.  3c , study 1), (b) by escalating doses of leptin during acute fasting or in the fed state in lean men, lean women, or obese men (Fig.  3d and Supplementary Figs.  2 , 3 , study 2) 32 , and (c) by long-term leptin treatment in lean mildly hypoleptinemic women (Fig.  3e, f , studies 3 and 4). Additionally, respiratory rate, which is a parameter that is included in most models of estimation of total energy expenditure based on respiratory function 33 , was not affected by escalating doses of leptin during acute fasting or in fed state (Fig.  3d and Supplementary Figs.  2 , 3 , study 2). Finally, physical activity in lean mildly hypoleptinemic women as assessed with metabolic equivalents*time was also not altered with leptin administration (Fig.  3e, f , studies 3 and 4).

Leptin has no effect on markers of SNS activity

Several animal studies have demonstrated regulatory effects of leptin on SNS activity, which may also lead to changes in energy expenditure. We have thus analyzed data from several markers of SNS activity in our studies. During short-term fasting, leptin replacement did not change HR, systolic BP (SBP), and diastolic BP (DBP) (Fig.  4a ). Similarly, no differences in HR, SBP, and DBP were observed during fasting or fed state between the different doses of leptin, when lean men, lean women, and obese men were investigated together (Fig.  4b ) or when we have compared the values between them (Supplementary Figs.  2 and 3 ). Similar to the results in the short-term fasting studies, long-term leptin treatment in hypoleptinemic women did not affect HR, SBP, or DBP (Fig.  4c, d ).

a Seventy-two hours fed state or fasting + leptin or fasting + placebo (study 1, n  = 13). P values of G (group: fed or fasting + placebo or fasting + leptin), T (time: days of study), and G * T interaction of mixed models adjusted for baseline are reported. b Seventy-two hours fasting treated with escalating leptin doses (study 2, n  = 15). P values of D (dose: 0.01 or 0.1 or 0.3 mg/kg/d), T (time: hours of fasting), and D * T interaction of mixed models adjusted for baseline are reported. c Open-label long-term leptin treatment in mildly hypoleptinemic women (study 3, n  = 7). P value ( P ) of paired t test is reported. d Placebo-controlled long-term leptin treatment in mildly hypoleptinemic women (study 4, n  = 19 (leptin = 10; placebo = 9)). P values of G (group: placebo or leptin), T (time: weeks of study), and G * T interaction of mixed models adjusted for baseline are reported. No post hoc Bonnferroni test was performed since G * T  > 0.05 (studies 1 and 4) and D * T  > 0.05 (study 2). All P values are two-sided. Data are demonstrated as means ± SEMs. Exact P values: b HR T  = 0.001 × 10 −17 ; SBP T  = 0.003 × 10 −1 ; DBP T  = 0.002 × 10 −2 ; MBP T  = 0.002 × 10 −2 .

Higher SNS activity may also be resulting by increased adrenal function. During short-term fasting, aldosterone levels increased compared to fed state but independently from treatment (leptin or placebo) (Fig.  5a , left). Similarly, 24-h urine cortisol and catecholamines collected at the second day of the study were generally higher after fasting compared to fed state (Fig.  5a , right) but nonsignificantly different between leptin and placebo group (−8.5% for cortisol and +8.5% for norepinephrine in leptin treatment compared to placebo). In study 2, plasma aldosterone, renin, urine epinephrine, and norepinephrine increased during fasting (compared to baseline levels) at the same magnitude in all leptin doses (physiological, supraphysiological, or pharmacological) (Fig.  5b and Supplementary Fig.  2 ). Long-term leptin treatment in mildly hypoleptinemic women in study 3 resulted in a small, early coordinated decrease in aldosterone and renin during the first 15 days of treatment with return to baseline for both hormones at day 45 of treatment (Fig.  5c ). In contrast to study 3, no decrease in aldosterone or renin was observed in study 4, and this was extended to a lack of changes in urine catecholamines and blood cortisol (Fig.  5d ). Altogether, no robust evidence of significant changes on markers of SNS activity with leptin treatment was observed in our study populations.

a Seventy-two hours fed state or fasting+leptin or fasting + placebo (study 1, n  = 13). Left: blood aldosterone and cortisol at the start and completion of the study. Right: 24-h urine cortisol, epinephrine, and norepinephrine collected at the last day of the study. P values of G (group: fed or fasting + placebo or fasting + leptin), T (time: days of study), and G * T interaction of mixed models, adjusted for baseline are reported. For urine catecholamines, P values were calculated with repeated measure ANOVA, since only the group factor existed. By G * T  < 0.05 (blood aldosterone and cortisol) and by G  < 0.05 (urine catecholamines), post hoc Bonferroni test was performed between the estimated means of the three groups and between the three groups at each timepoint. Three asterisks indicate P  < 0.001 for fed vs fasting + placebo (red) and for fed vs fasting + leptin (blue) at the particular timepoint. b Seventy-two hours of fasting treated with escalating leptin doses (study 2, n  = 15). P values of D (dose: 0.01 or 0.1 or 0.3 mg/kg/d), T (time: days of fasting), and D * T interaction of mixed models adjusted for baseline are reported. No post hoc Bonnferroni test was performed since D * T  > 0.05. c Open-label long-term leptin treatment in mildly hypoleptinemic women (study 3, n  = 7). P value ( P ) of time effect (i.e., days of study) of mixed models adjusted for baseline is reported. By P  < 0.05 post hoc Bonferroni’s test for each timepoint compared to baseline was additionally performed and two asterisks indicate P  < 0.01 for the specific timepoint vs 0 (baseline). d Placebo-controlled long-term leptin treatment in mildly hypoleptinemic women (study 4, n  = 19 (leptin = 10; placebo = 9)). P values of G (group: placebo or leptin), T (time: weeks of study), and G * T interaction of mixed models adjusted for baseline are reported. No post hoc Bonnferroni test was performed since G * T  > 0.05. For Bonferroni post hoc tests, only significant results are reported. All P values are two-sided. Data are demonstrated as means ± SEMs. Exact P values : a aldosterone T  = 0.001 × 10 −3 ; day 3 post hoc test for fed vs Pl = 0.004 × 10 −2 ; fed vs Le = 0.006 × 10 −2 . b Aldosterone T  = 0.001 × 10 −15 ; renin T  = 0.005 × 10 −17 ; urine epinephrine T  = 0.004 × 10 −1 . c Aldosterone post hoc test at 15 days = 0.005.

Long-term leptin transiently increases free fatty acids

Studies have suggested a differential role for leptin on lipid metabolism depending on energy status, with low leptin levels signaling the shift from carbohydrate to increased lipolysis and lipid utilization during starvation and with leptin treatment inducing lipolysis by stimulating SNS activity in nonfasting conditions 8 , 34 , 35 .

In our lean population (study 1), respiratory quotient indicated a reduction in the utilization of carbohydrates and an increase in the utilization of lipids during short-term fasting, which was not affected by leptin treatment (Supplementary Fig.  4a ). Similarly, in studies 3 and 4, no robust changes in macronutrient utilization during long-term leptin treatment were observed (Supplementary Fig.  4b, c ). To further investigate the above finding, we have performed a metabolite–lipid–lipoprotein analysis. In study 1, 68 lipoproteins, lipids, and metabolites were significantly different between the three admissions (Fig.  6a, b ). In a sparse partial least-squares discriminant analysis (sPLS-DA) between the two fasting conditions (leptin or placebo-treated), component 1 consisting of ten parameters discriminated progressively between the different days of fasting but not between placebo or leptin (in Fig.  6c , faint colors (days 0–1 of fasting) are gathered at the right (area of positive values for component 1) and bright colors (days 2–3 of fasting) at the left (area of negative values for component 1), whereas the circles (leptin) and the squares (placebo) are equally distributed from right to left). Component 1 included classic milestones of metabolic adaptation during starvation, such as amino acids and ketone bodies that their concentrations change with fasting but independent from treatment (placebo or leptin) (Supplementary Fig.  5a, b ). This shows that blocking hypoleptinemia does not prevent the shift from carbohydrates to lipid utilization and ketone formation during starvation in lean humans. Similarly, in study 2, sPLS-DA demonstrated changes with time during fasting (component 2 consisting of ketone bodies, amino acids, and fatty acids) but not with leptin dose (Supplementary Fig.  6 ). In line with the above findings, in study 4, long-term leptin administration did not induce any significant changes in amino acids, ketone bodies, or lipoproteins compared to placebo (fatty acids were not assessed in this study, apart from free fatty acids (FFA)), which is indicated by the lack of distinct clusters in sPLS-DA and the lack of significantly different parameters in one-way ANOVA (Fig.  6e ).

a – d Effects on metabolite and lipid metabolism of 72-h fed state or fasting treated with leptin or placebo (study 1, n  = 13). a Evaluation of metabolites, lipids, and lipoproteins with one-way ANOVA between the three admissions. Red dots indicate parameters significantly different and blue dots parameters not significantly different between groups (fed vs fasting + placebo vs fasting + leptin) with a preset false discovery rate of P  < 2.15 × 10 −4 (total 68 parameters significant). b Heatmap of the 68 significant parameters according to one-way ANOVA for the three admissions. c sPLS-DA analysis of fasting+leptin vs fasting + placebo: symbols indicate the measurement of component 1 in relation to measurement of component 3 for one subject/on one treatment/on one day of fasting: Blue circles correspond to leptin and red squares to placebo. Increasing color intensity indicates more time (days) of fasting. d The ten parameters that compose components 1 and 3 and their level of contribution (loading) to the component. e Effects on metabolite and lipid metabolism of long-term leptin treatment in mildly hypoleptinemic women (placebo-controlled study 4, n  = 19 (leptin = 10; placebo = 9)). Left: sPLS-DA analysis of metabolites and lipoproteins in placebo vs leptin. Symbols indicate the measurement of component 1 in relation to component 2 for one subject/on one treatment/on one day of study: blue circles correspond to leptin and red squares to placebo. Increasing color intensity of symbol indicates more time (weeks) of study. Large oval-colored shapes indicate 95% confidence interval for each group. The observed major overlap between groups suggests no significant differences between placebo and leptin. Right: Evaluation of metabolites and lipoproteins with one-way ANOVA in placebo and leptin-treated subjects for up to 36 weeks. Each dot represents a parameter (blue dot = nonsignificant parameter, red = preset color for significant parameters but no such parameter was detected). NMR-based metabolomics were used to quantify amino acids, metabolites, and lipids bound to lipoproteins. GC/MS-EI was used to quantify fatty acid methyl esters. Le_0, Le_1 etc. indicate day 0 (baseline), 1 etc. of fasting + leptin. Pl_0, Pl_1 etc. indicate day 0 (baseline), 1 etc. of fasting + placebo. For metabolite nomenclature, see Supplementary Data 2.

Regarding the lipid profile, when we assessed explicitly the concentrations of FFA (Fig.  7b ), as well as of triglycerides, phospholipids, and sphingomyelins bound to different size lipoproteins (Supplementary Data  1 and Supplementary Fig.  5c ), no difference was observed between placebo and leptin in acute fasting of study 1. Interestingly, though component 3 consisting of ten parameters (Fig.  6d ), mainly fatty acids, was able to discriminate 7 subjects (brown squares clustering together at lower half of the score plot—in the area of negative values of component 3) on day 3 of placebo treatment (Fig.  6c ), thus indicating that the concentrations of these fatty acids are probably significantly different between placebo and leptin treatment on day 3. Indeed, the elevated concentrations of fatty acids (both free and bound to lipoproteins) are reduced partially to baseline level at the third day of placebo but not at the third day of leptin treatment (Fig.  7a ). This shows that blocking hypoleptinemia not only does not prevent the fasting-induced changes in lipid metabolism, but it may slightly stimulate them.

a , b Seventy-two hours fed state or fasting+leptin or fasting+placebo (study 1, n  = 13). a Blood concentrations of fatty acid profile from start and till completion of the study as ratios of the baseline (0 day). GC/MS-EI was used to quantify fatty acid methyl esters in whole plasma. b Blood-free fatty acids (FFA) from start and till completion of the study; mixed model was performed (for FFA adjusted for baseline). P values of G (group: fed or fasting + placebo or fasting + leptin), T (time: days of study), and G * T interaction of mixed models are reported. In panels a and b by G * T  < 0.05, post hoc Bonferroni test was performed between the estimated means of the three groups and between the three groups at each timepoint. One, two, or three asterisks indicate P  < 0.05, <0.01, or <0.001 for fed vs fasting + placebo (red) and for fed vs fasting + leptin (blue). One, two, or three hash signs indicate P  < 0.05, <0.01, or <0.001 for fasting + leptin vs fasting + placebo in the Bonferroni post hoc t test. c Open-label long-term leptin treatment in mildly hypoleptinemic women (study 3, n  = 7). Blood FFA concentrations. P value of time effect (i.e., days of study) of mixed models adjusted for baseline is reported. Two asterisks indicate P  < 0.01 for the specific timepoint vs 0 (baseline) in the Bonferroni post hoc t test (performed by P  < 0.05). d Placebo-controlled long-term leptin treatment in mildly hypoleptinemic women (study 4, n  = 19 (leptin = 10; placebo = 9)). P values of G (group: placebo or leptin), T (time: weeks of study), and G * T interaction of mixed models adjusted for baseline are reported. By G * T  < 0.05, post hoc Bonferroni test was additionally performed between the two groups at each timepoint. One or two asterisks indicate P  < 0.05, or <0.01 for leptin vs placebo for the specific timepoint. For Bonferroni post hoc tests, only significant results are reported. All P values are two-sided. Data are demonstrated as means ± SEMs. Exact P values. a Total FA (ratio) G  = 0.008 × 10 −4 ; T  = 0.002 × 10 −1 ; fed vs Le = 0.004 × 10 −1 ; day 2 fed vs Pl = 0.008 & fed vs Le = 0.002; day 3 fed vs Le = 0.001 × 10 −1 & Le vs Pl = 0.002. SFA (ratio) G  = 0.002 × 10 −4 ; T  = 0.003 × 10 −2 ; fed vs Le = 0.008 × 10 −5 ; day 2 fed vs Pl = 0.003 & fed vs Le = 0.002; day 3 fed vs Le = 0.001 × 10 −1 & Le vs Pl = 0.001. MUFA (ratio) G  = 0.002 × 10 −6 ; fed vs Le = 0.008 × 10 −7 ; day 1 fed vs Le = 0.021; day 2 fed vs Pl = 0.001 & fed vs Le = 0.001 × 10 −1 ; day 3 fed vs Le = 0.001 × 10 −1 & Le vs Pl = 0.036. PUFA (ratio) T  = 0.005 × 10 −3 . C14:0 (ratio) G  = 0.009 × 10 −2 ; Le vs Pl = 0.004 × 10 −1 ; day 3 fed vs Le = 0.002 & Le vs Pl = 0.002. C16:0 (ratio) G  = 0.004 × 10 −5 ; T  = 0.002 × 10 −2 ; fed vs Le = 0.002 × 10 −5 ; day 2 fed vs Pl = 0.008 × 10 −1 & fed vs Le = 0.008 × 10 −1 ; day 3 fed vs Le = 0.001 × 10 −1 & Le vs Pl = 0.002. C16:1 (ratio) G  = 0.002 × 10 −6 ; T  = 0.004 × 10 −2 ; fed vs Le = 0.007 × 10 −7 ; day 1 fed vs Pl = 0.012; day 2 fed vs Pl = 0.009 & fed vs Le = 0.001 × 10 −1 ; day 3 fed vs Le = 0.001 × 10 −1 & Le vs Pl = 0.014. C18:0 (ratio) day 3 fed vs Le = 0.001 × 10 −1 & Le vs Pl = 0.008. C18:1n9cis (ratio) G  = 0.001 × 10 −6 ; fed vs Le = 0.006 × 10 −7 ; day 1 fed vs Le = 0.022; day 2 fed vs Pl = 0.001 & fed vs Le = 0.001 × 10 −1 ; day 3 fed vs Le = 0.001 × 10 −1 & Le vs Pl = 0.039. C18:2n6cis (ratio) T  = 0.001 × 10 −1 ; C20:4n6 (ratio) G  = 0.002 × 10 −1 ; T  = 0.002 × 10 −1 ; fed vs Le = 0.001 × 10 −1 ; day 2 fed vs Pl = 0.022 & fed vs Le = 0.004; day 3 fed vs Le = 0.001 × 10 −1 & Le vs Pl = 0.008 × 10 −1 . C20:5n3 (ratio) day 2 fed vs Le = 0.025; day 3 fed vs Le = 0.004 & Le vs Pl = 0.038. C20:3n6 (ratio) T  = 0.006 × 10 −12 ; C22:0 (ratio) T  = 0.005 × 10 −6 ; C22:6n3 (ratio) T  = 0.001 × 10 −3 ; day 2 fed vs Pl = 0.004 & fed vs Le = 0.004; day 3 fed vs Le = 0.003 × 10 −1 & Le vs Pl = 0.011. b G  = 0.004 × 10 −11 ; T  = 0.005 × 10 −13 ; G * T  = 0.004 × 10 −3 ; fed vs Le = 0.005 × 10 −10 ; fed vs Pl = 0.003 × 10 −6 ; day 2 fed vs Le 0.002 × 10 −7 & fed vs Pl = 0.004 × 10 −5 ; day 3 fed vs Le = 0.008 × 10 −9 & fed vs Pl = 0.001 × 10 −10 . c Post hoc test at 15 days = 0.002. d Post hoc test at 8 weeks = 0.011 and at 16 weeks = 0.002.

In agreement with the mild stimulatory role of leptin on lipid catabolism during fasting, long-term leptin treatment in fed state led to a transient increase of the circulating levels of FFA. In study 3, FFA was increased at day 15 of treatment and returned to baseline later (Fig.  7c ). In study 4, FFA, similar to study 3, was increased significantly in the leptin compared to the placebo group up to 20 weeks ( P  = 0.002 for treatment adjusted for baseline), but the significance of this change disappears when timepoints through 36 weeks are included (Fig.  7d ). In examining relationships between FFA and the hypothalamic–pituitary–peripheral axes previously measured, we found a negative correlation of FFA with aldosterone for study 3 ( r  = −0.536 and P value = 0.047 adjusted for multiple timepoints/subjects), which was not confirmed in study 4 where both the changes in FFA and in aldosterone are milder (Supplementary Table  1 ). Finally, we did not find any association of FFA with thyroid-stimulating hormone (TSH), free triiodothyronine (T3), free thyroxine (T4), adrenocorticotropic hormone (ACTH), cortisol, renin, growth hormone-binding protein (GHBP), or insulin-like growth factor 1 (IGF-1) in the leptin group (Supplementary Table  1 ).

Altogether, leptin treatment does not induce a major shift from carbohydrate to lipid utilization, but it may affect fatty acid profile, either by maintaining very high fatty acid levels during short-term treatment in acute fasting or by transiently increasing FFA during long-term treatment. Importantly, these findings justify a more in-depth lipidomic analysis in the future, that will include lipid subgroups that were not assessed in our current study.

We investigated herein the effects of short- and long-term leptin treatment on the regulation of body weight and composition in lean normo- and mildly hypoleptinemic individuals in four clinical studies and observed important differences compared to the reported effects of leptin in animal models (ob/ob or lean) and in human studies with people in hypoleptinemic–lipodystrophic or obese–hyperleptinemic state. Our results support the hypothesis that leptin demonstrates differential effects on energy regulation, depending on the metabolic context and energy balance 25 , 36 , as reflected by leptin levels, with a progressive loss of function from conditions of energy and leptin deficiency to conditions of energy and leptin excess.

Regarding body weight, short-term leptin treatment does not further induce the weight loss observed during acute fasting in lean individuals. The lack of effect of leptin can be both due to the short duration of treatment and due to the abolishment of the effects of leptin on energy intake through the imposed complete fasting. Long-term leptin treatment in lean mildly hypoleptinemic women led to 4–4.5% of body weight loss (exclusively fat mass), which is far less compared to the weight loss observed in CLD 11 , modestly less than the weight loss observed in severe hypoleptinemia due to GL (~5.5%) 10 , and significantly more compared to the neutral weight effects observed in hyperleptinemic obesity. However, blood leptin levels at baseline, which ranged between 1.5 and 8 ng/ml, did not correlate with the % weight loss due to leptin treatment. This suggests that the response to leptin treatment in terms of weight loss may not depend linearly on the leptin blood concentrations, but may dependent on the metabolic context (i.e., energy status, presence of genetic mutations). Thus, defining strict thresholds in leptin blood concentrations as reliable predictors of weight loss with leptin treatment in obese populations may prove to be challenging and demands further studies in large populations with a wide range of leptin levels and different metabolic phenotypes.

Second, animal studies (mainly in ob/ob mice and in lean rodents) demonstrated that leptin administration prevents the expected reduction in energy expenditure due to low-energy intake 37 , potentially by acting on hypothalamic nuclei, upregulating SNS activity, and adrenal hormone secretion toward thermogenesis and increased HR and BP, and by increasing physical activity 38 , 39 , 40 , 41 , 42 , 43 , 44 . In humans, such effects are only modest, if any, as observed in CLD, GL, or PL 11 , 44 , 45 , 46 , 47 . In overweight/obese–hyperleptinemic people, leptin treatment does not affect energy expenditure 19 , 22 , 45 , apart from a reported improvement in non-resting energy expenditure with leptin replacement after stabilization to reduced body weight with diet, in a study of sequential study design 46 , 47  and even if present this does not necessarily translate to better body weight sustainment 22 , 46 , 48 . We now show that leptin treatment does not increase resting energy expenditure, does not stimulate physical activity, and does not affect markers of SNS activity (HR, BP, cortisol, and catecholamine production) in the lean normoleptinemic and partial hypoleptinemic individuals of our studies, supporting the rather marginal, if any, effects of leptin on energy expenditure in humans.

Third, both stimulatory and inhibitory effects of leptin on lipolysis and lipid utilization have been reported based on the metabolic context (starvation or not), magnitude, and type of leptin deficiency (CLD, GL, and PL) and leptin dose. In rodents, starvation leads to hypoleptinemia and increased white adipose tissue (WAT) lipolysis via activation of the hypothalamic–pituitary–adrenal (HPA) axis 35 , 49 . Both WAT lipolysis and the activation of HPA are suppressed after physiologic leptin replacement, but stimulated after supraphysiologic leptin treatment 35 . In lean humans, a correlation between decreasing leptin levels and an increase in cortisol, FFA, and ketones during starvation was recently reported 34 , which suggested an anti-lipolytic role for leptin. In our study, administration of leptin in lean individuals does not attenuate the amino acid surge or the robust increase observed in circulating FFA and ketone bodies with fasting (even in very high leptin doses) and does not significantly affect cortisol or catecholamine levels. On the contrary, we observe higher concentrations of total fatty acids during the third day of leptin treatment compared to placebo, which supports a stimulatory, if any, and not an inhibitory effect of leptin on lipid catabolism. This is in agreement with observations in nonfasting conditions. Specifically, leptin stimulates lipid utilization in ob/ob mice 50 and lipolysis in lean mice fed ad libitum possibly through activation of sympathetic neurons innervating adipocytes. In humans, similar to weight regulation, there is a progressive loss of the lipocatabolic effects of leptin from conditions of leptin deficiency to leptin excess. Consequently, in people with CLD, leptin replacement stimulates lipid catabolism (lipolysis and oxidation) as indicated by increases in ketone bodies, FFA, and acylcarnitines 51 . In people with GL or PL, leptin treatment has a modest effect on lipid catabolism, since it does not affect FFA and ketone body concentrations, but increases acylcarnitines and by-products of branched-chain amino acids and protein degradation 52 . Similarly, in our studies, leptin has a modest lipocatabolic effect in lean mildly hypoleptinemic women, as it is associated with a transient increase in FFA but no changes in ketone bodies or amino acid concentrations. Importantly, the increase in FFA was not associated with alterations in hypothalamic–pituitary function and specifically with thyroid hormones or IGF-1, which are known to have lipolytic effects 53 , 54 . It was only associated with the reduction observed in aldosterone levels in study 3, which was not verified in study 4, where dose adjustments were performed to prevent too much body weight loss. Given that aldosterone has rather lipolytic properties 55 , the transient, increased serum FFA may have an inhibitory effect on aldosterone secretion as a part of a compensatory mechanism. Furthermore, even though we see some small changes in FFA with leptin, these are not proportional to the changes in fat mass, suggesting that they may account for some minor but not all of leptin’s effects on body weight/fat mass that are probably due to the decreasing energy intake.

Regarding energy intake, in ob/ob mice and lean rodents, leptin replacement decreases caloric intake 9 , 37 , 56 . Similarly, in severe hypoleptinemic populations with CLD or GL, leptin decreases robust food intake by affecting hedonic and homeostatic nervous centers that control satiety and hunger feeling 31 , 57 , 58 . In lean women with mild acquired hypoleptinemia, leptin administration reduces salience, attention, and rewarding value of food 59 . In obese subjects after weight loss, which can be characterized as a condition of relative leptin deficiency, it affects brain activity and increases satiation 47 , 48 . In contrast, in obese–hyperleptinemic men, studied at their usual weight, leptin administration has minimal effects on appetite regulation 60 . In our studies, leptin administration in normoleptinemic lean subjects during short-term fasting partially prevents the increase of food intake at refeeding. Thus, it is plausible to expect similar effects on energy intake in partially hypoleptinemic individuals under long-term leptin treatment, considering the similar impacts on body weight in our longer-term trials. Indeed, the projection curve for the expected fat mass loss due to reduced energy intake almost overlaps the real curve of fat mass loss observed in study 3, where leptin dose was not adjusted based on body weight changes.

In summary, we present herein that one of the main metabolic effects of leptin in lean subjects is the regulation of energy intake, an effect that is saturable as leptin increases to within physiological levels at least during refeeding after food deprivation. This can be translated into weight loss, mainly due to fat mass loss, in the long term in subjects with chronic mild hypoleptinemia. Additionally, leptin treatment may lead to a transient increase in circulating FFA, without affecting energy expenditure and SNS activity. Although the effects of leptin on weight regulation, energy intake, and lipid catabolism are progressively lost with progression from conditions of energy and leptin deficiency to conditions of energy and leptin excess, the response to leptin treatment in terms of weight loss may not depend linearly on the leptin blood concentrations prior to treatment initiation. In the future, larger and longer studies of leptin administration to lean individuals in physiologic and supraphysiologic doses, as well as in the subset of obese patients with low endogenous leptin levels and/or obese subjects with induced hypoleptinemia, are needed to fully elucidate physiology and potential therapeutic utility of leptin in obesity 25 , 61 .

This study has some limitations. In short-term fasting studies, we measured RMR but not total or non-resting energy expenditure due to lack of metabolic chambers. Additionally, no weighted buffet meals to assess energy intake were performed longitudinally under long-term leptin replacement and this remains to be studied in detail in the future. Physical activity was calculated using daily self-report diaries as a surrogate of exercise-induced energy expenditure and this is a validated method. Our metabolite–lipid–lipoprotein analysis, although lege artis, did not include all circulating lipids or metabolites, and did not describe lipid subgroups and individual lipid species that should be the focus of more in-depth studies in the future. Additionally, whether the increase we observed in FFA in studies 3 and 4 is related to an upregulation of lipolysis, reduced lipogenesis or changes in re-esterification could not be addressed with certainty in the context of the current experimental setting. We also acknowledge that the sample size, especially in study 2, may have been small, resulting in increased type II error for some parameters. We have tried to address this by performing both combined for groups or doses/group analyses, as well as separate analyses for each group/dose. Finally, conclusions about SNS activity derive from catecholamines, HR and BP levels, and not from pharmacological blockade that may be able to detect very small differences or heart rate variability measures, which we have reported in the past 62 .

We utilized data and specimens from our previous studies to perform new measurements and analyses 26 , 27 , 28 , 29 , 30 (Supplementary Table  2 ). The primary outcomes of our analysis were (a) correlations of baseline leptin levels with % of weight change in all four clinical studies–interventions (Fig.  1 for study design), (b) differences in % weight changes between lean men, lean women, and obese in escalating leptin doses (study 2), and (c) weight, fat mass, and FFA changes in relation to leptin levels during long-term leptin treatment and after its termination (studies 3 and 4). The secondary outcomes were changes in energy expenditure (i.e., RMR and physical activity), energy intake, SNS activity (i.e., HR, BP, body temperature, and serum/urine catecholamines), and metabolic profile (i.e., lipoproteins, amino acids, fatty acids, and ketone bodies) in all four clinical studies–interventions (Fig.  1 for study design).

Study approval

The human studies were approved by the Institutional Review Board of the General Clinical Research Center (GCRC) of the Beth Israel Deaconess Medical Center and were performed under an investigator-held IND. Written informed consent was obtained from all participants prior to inclusion in the study.

Study 1: short-term mechanistic study

Eight healthy lean men (age = 23.3 ± 1.2 yr; BMI = 23.7 ± 0.6 kg/m 2 ) and seven healthy lean women (age = 22.4 ± 1.2 yr; BMI = 21.7 ± 2.2 kg/m 2 ) with regular menstrual cycles and not on oral contraceptives for at least 6 months were studied under three separate Clinical Research Center (CRC)-based conditions for 72 h: one under isocaloric fed-state conditions (normoleptinemia) and two during complete fasting-state conditions (induced hypoleptinemia) scheduled in a random order and in a double-blind fashion with administration of physiologic replacement of leptin doses (fasting + leptin) or placebo (fasting + placebo) 26 , 27 . The interval between admissions was at least 8 weeks to allow recovery of hematocrit, leptin levels, and body weight. Each subject completed three studies (i.e., fed, fasting + placebo, and fasting + leptin) with the following exceptions: two males withdrew before completing the fasting + leptin study and one female did not complete the fasting + placebo study. We excluded the two males from the analysis as their corresponding data were insufficient, but we included the female since she had completed 2/3 studies (fed and fasting + leptin), so that in total, findings from 13 subjects (6 males and 7 females) were analyzed. During each fed or fasting study, subjects were admitted to the CRC the evening before study day 0. The isocaloric fed state consisted of four standardized meals per day: breakfast (20% of daily calories) at 8:00, lunch (35% of daily calories) at 13:00, dinner (35% of daily calories) at 18:00, and a snack (10% of daily calories) at 22:00. During the fasting state, only a standardized volume of calorie-free fluids, electrolytes (NaCl (500 mg) and KCL (40 meq)), and vitamin supplements was allowed. Ad libitum feeding was allowed starting at 13:00 on the third study day and meals were weighed to obtain accurate measures of the calories ingested. Body composition (bioelectric impedance analysis; RJL Systems, Clinton Township, MI), RMR (DeltaTrac II Metabolic Monitor; SensorMedics), and morning vital signs (HR, BP) were assessed at the beginning and end of each study. Blood samples were obtained at 8:00–8:30 am on days 0, 1, 2, and 3. Urine collection was performed on day 2. The doses of leptin were 0.01 mg/kg given at 8 am and every 6 h on day 0 and 0.025 mg/kg at 8 am and every 6 h on days 1 and 2 for males and 0.02 mg/kg given at 8 am and every 6 h on day 0 and 0.05 mg/kg given at 8 am and every 6 h on days 1 and 2 for females (Amylin, Inc., San Diego, CA; previously known as r-metHuLeptin, provided by Amgen, Inc., Thousand Oaks, CA) administered subcutaneously (Fig.  1 for study design and Supplementary Fig.  7 for flow diagram). Males and females were administered a single dose of 0.025 mg/kg and 0.05 mg/kg, respectively, at 8 am on day 3. The results from these studies had previously been reported separately for men and women but are combined herein [ClinicalTrials.gov Study 1: NCT00140231].

Study 2: short-term leptin dose escalation study

Five lean men (age = 22.2 ± 0.9 yr; BMI = 22.0 ± 0.5 kg/m 2 ), five men with obesity (age = 23.4 ± 1.5 yr; BMI = 32.0 ± 1.0 kg/m 2 ), and five lean women (age = 20.4 ± 0.7 yr; BMI = 21.9 ± 0.7 kg/m 2 ) participated in 3 fed–normoleptinemic and 3 fasting-induced hypoleptinemic studies, which were conducted in the CRC, with leptin administration at three different doses (dose A = 0.01 mg/kg, dose B = 0.1 mg/kg, and dose C = 0.3 mg/kg) (protocol previously published 30 ) (Fig.  1 for study design and Supplementary Fig.  7 for flow diagram). One normoleptinemic (fed) and one hypoleptinemic study (fasting) were performed at each of the three different doses of leptin, resulting in six visits in total. Leptin (metreleptin, supplied by Amgen, Inc., Thousand Oaks, CA) was administered once daily at 8:00 am subcutaneously. For males, fed studies were performed after the completion of all three fasting studies. For females, the first day of each fasting study was scheduled during the beginning of each follicular phase, and thus fed studies were conducted either in between or after the fasting studies.

The duration of each fasting-induced hypoleptinemic visit was 72 h. Subjects were admitted to the CRC the night before the first study day and received a standardized 748-kcal snack at 22:00. After that, subjects fasted until 22:00 of day 3 when they received a standardized 225-kcal snack. Only a standardized volume of calorie-free fluids, electrolytes (NaCl (500 mg) and KCL (40 meq)), and vitamin supplements were allowed from the beginning until the end of fasting on day 3. The interval between admissions was no less than 2 weeks.

The duration of each fed/normoleptinemic study was 24 h. Subjects were admitted to the CRC the night before the study day. On the study day, participants received an isocaloric diet with breakfast at 7:00 (20% of daily calories), lunch at 14:00 (35% of daily calories), dinner at 18:00 (35% of daily calories), and a snack at 22:00 (10% of daily calories), during which subjects received an isocaloric diet. Each admission was separated by 1–12 weeks. Eight subjects received the 0.01 mg/kg/day and the 0.1 mg/kg/day doses on consecutive days, since the 0.01 mg/kg/day dose was not expected to alter leptin levels 24 h later.

Vital signs, including HR, BP, body temperature, and respiratory rate, were measured at 7:00, 14:00, and 18:00–20:00 of each study day (e.g., 3 days for the hypoleptinemic state and 1 day for the normoleptinemic state). Body weight was measured on the morning of each study day, prior to blood sampling and prior to breakfast regarding the fed admissions, with the same scale in CRC and with subjects dressed in a standard hospital gown. Leptin was administered at 8:00 every morning. Leptin levels were measured at +30 min, +1 h, + 2 h, +3 h, +4 h, +5 h, +6 h, +8 h, +10 h, +12 h, +18 h, and +24 h after each dose (presented at the beginning and completion of treatment). Serum samples obtained in the early morning (prior to dose administration), noon (after leptin’s peak), and evening of each study day were used for renin, aldosterone, FFA, and lipoprotein/metabolite (nuclear magnetic resonance (NMR)-based metabolomics) measurements. For fatty acids with gas chromatography electron ionization mass spectrometry (GC/MS-EI), serum samples were examined on day 1 at 8:00 (before leptin administration), and on days 1, 2, and 3 at 14:00 (close to the serum peak of leptin). Finally, urine catecholamines were measured at baseline and on day 3 of each admission at the fasting state. Renin, aldosterone, and urine catecholamine measurements were not available in the fed state [ClinicalTrials.gov Study 2: NCT00140205].

Study 3: long-term leptin replacement study

Eight lean women (age = 24.8 ± 5.4 years; body mass index (BMI) = 20.5 ± 2.0 kg/m 2 ) with acquired hypoleptinemia due to hypothalamic amenorrhea (HA) secondary to strenuous exercise for at least 6 months were studied. All subjects were otherwise healthy, without active eating disorders, with stable weight (inclusion criteria: within ± 15% of ideal body weight for ≥6 months) and were not taking any medications, including estrogen, for at least 3 months. Finally, all participants had normal prolactin and thyrotropin levels, ratios of luteinizing hormone (LH) to follicle-stimulating hormone (FSH) of more than 1.5, and no signs of hirsutism or acne. Subjects received leptin (0.08 mg/kg/day, self-injected subcutaneously twice daily as 40% in the morning and 60% in the evening) initially for 2 months (protocol previously published 29 ). Subjects who had not ovulated in the first 2 months continued with a third month of treatment at an increased dose of 0.2 mg/kg/day (with the same administration schedule as above). Ovulation was determined with one or more of the following: a 2 mm per day growth of the dominant follicle from its preovulatory size (≥18 mm in length), with subsequent collapse or internal echo appearance in the pelvic ultrasonography (performed weekly); serum or urinary LH surge; >4 ng per ml increase in progesterone levels. Blood samples were obtained weekly and body composition was determined with dual-energy X-ray absorptiometry (DEXA) every other week, starting 1 month before initiation of leptin treatment (baseline month, where measurements were performed at the beginning and end of the month). RMR was measured (DeltraTrac II Metabolic Monitor, SensorMedics) during the baseline month and after 15 days of leptin treatment. Morning vital signs (HR, BP, and temperature) were obtained in the morning during baseline month and after 15 days of leptin treatment. Daily exercise records were obtained. Physical activity was calculated as the weekly sum of the product of Mets (metabolic equivalent) *Duration (hours) for each activity type. Metabolic equivalent values used were according to the 2011 Compendium of Physical Activities (Supplementary Table  3 ) 63 . Even though eight females were initially enrolled, one subject withdrew after 1 month for reasons unrelated to the study, and thus, the results are derived from the remaining seven subjects (Fig.  1 for study design and Supplementary Fig.  7 for flow diagram).

Study 4: confirmatory placebo-controlled study

Twenty females, between 18 and 35 years old with hypoleptinemia due to secondary HA for ≥6 months coincident with strenuous exercise and/or low body weight (within ± 15% of ideal body weight for ≥6 months at the time of screening), were studied. All subjects were otherwise healthy, without active eating disorders or other psychiatric disease and were not taking any medications that could affect hormone or bone mass measurements (i.e., glucocorticoids, antiseizure medications, thyroid hormones, or estrogens) for at least 3 months. None of the subjects had hyperprolactinemia, hypo- or hyperthyroidism, Cushing’s syndrome, congenital adrenal hyperplasia, or primary ovarian failure. Subjects were randomized with a 1:1 allocation to receive either metreleptin or placebo for 36 weeks 28 . Randomization tables were produced by the Harvard Catalyst biostatisticians with SAS and delivered directly to the Research Pharmacy for use such that study staff that recruited subjects (medical doctors, care providers) as well as the participants would remain blinded. Primary and secondary outcomes of the study were the difference between the placebo- and leptin-treated group for bone mineral content, bone markers, and bone mineral density, as well as reproductive outcomes from baseline to 36 weeks. Metreleptin was self-injected subcutaneously once daily at a dose of 0.08 mg/kg/day for 12 weeks, and subjects who had begun menstruating remained on this dose until the completion of the study. The dose for subjects who had not menstruated at week 12 was increased to 0.12 mg/kg/day. If a subject lost >5% of her baseline weight, the dose was reduced by 0.04 mg/kg. Fasting blood samples were collected every 4 weeks, along with fasting vital signs (HR, BP, and temperature) and body weight measurements. Body composition and RMR were measured every 12 weeks with DEXA and Sensormedics Vmax Encore equipment (VIASYS Respiratory Care Inc.), respectively. Physical activity was calculated as described for study 3. Among the 20 participants who were enrolled in the study, 11 were assigned randomly to receive metreleptin (age = 26.6 ± 1.4 years; BMI = 20.9 ± 0.6 kg/m 2 ) and 9 to receive placebo (age = 25.4 ± 1.2 years; BMI = 19.8 ± 0.7 kg/m 2 ). One participant in the metreleptin-treated group withdrew from the study because she developed injection-site reactions soon after the baseline visit, leaving 10 in the metreleptin group and 9 in the placebo group (Fig.  1 for study design and Supplementary Fig.  7 for flow diagram) [ClinicalTrials.gov Study 4: NCT00130117].

Biochemical analysis

FFA (intra-assay variability: 1.5%, sensitivity: 0.01–4.00 mEq/L NEFA) was measured using commercially available enzymatic colorimetric assay from Fujifilm Wako Diagnostics U.S.A Corporation (Mountain View, CA, USA). Levels were measured with an automated immunoassay system (Immulite 1000, Siemens, Deerfield, IL). All samples were run in duplicates within the same run for a given subject and were repeated if the coefficient of variation for any sample was N15%. Aldosterone (intra-assay CV 1.4–3.4%, inter-assay CV 9.5–12.1%, and sensitivity: 22.4 pg/mL) and renin (intra-assay CV 1.7–5.3%, inter-assay CV 4.0–5.5%, and sensitivity: 14.8 pg/mL) were measured using commercially available immunoassays (R&D Systems, Inc., Minneapolis, MN, USA). Levels were measured with an automated immunoassay system (Immulite 1000, Siemens, Deerfield, IL). Similar to the FFA analysis, all samples were run in duplicates within the same run for a given subject and were repeated if coefficient of variation for any sample was >15%.

NMR-based metabolomics

High-throughput proton NMR metabolomics (Nightingale Health Ltd, Helsinki, Finland) was used to quantify circulating metabolites and lipids within lipoprotein particles. This is a targeted metabolomics approach where all metabolic measures are of known identity and therefore are in level 1 identification level according to Summer et al. 64 . The method leads to simultaneous quantification of lipoprotein subclasses with lipid concentrations, fatty acids, amino acids, ketone bodies, and metabolites related to gluconeogenesis (Nightingale Health biomarker quantification library 2016). Details of the experimentation and proton NMR spectrometer characteristics have been described previously 65 , 66 , 67 . In brief, serum samples are stored in a freezer at −80 °C. Before preparation, frozen samples are slowly thawed at +4 °C overnight and then mixed gently and centrifuged at 3400× g to remove possible precipitate. Aliquots of each sample (100 μl) are added to 100 μl of sodium phosphate buffer (75 mM Na 2 HPO 4 in 80%/20% H 2 O/D 2 O, pH 7.4, including also 0.08% sodium 3-(trimethylsilyl)propionate-2,2,3,3-d4 and 0.04% sodium azide) automatically with a Gilson Liquid Handler 215 to 3-mm outer-diameter SampleJet NMR tubes. The resulting solution is then mixed by aspirating three times. The prepared samples are stored in 96-tube racks that are inserted into one of the five well-plate positions in the SampleJet TM (Bruker BioSpin GmbH, Germany) sample changer. The latter is placed on top of the superconducting magnet inside which the NMR probehead is located. The sample changer includes a cooling unit, which maintains the temperature of samples waiting to be measured at +6 °C, and a preheating unit, which warms up the sample just before measurement. The sample is then kept idle inside the NMR probehead to achieve temperature stabilization at 36.95 °C. Thus, the measurement temperature is constant. Two NMR spectra are acquired from each serum sample using a Bruker AVANCE III spectrometer operating at 500.36 MHz. The first spectrum includes overlapping resonances arising mainly from different lipid molecules in various lipoprotein particles. The second spectrum, acquired with spectrometer settings using a T2-relaxation-filtered pulse sequence to suppress most of the broad macromolecule and lipoprotein lipid signals, leading to enhanced detection of rapidly tumbling smaller solutes. Representative spectra are illustrated in reviews of the Nightingale NMR metabolomics method 65 . Data processing includes the Fourier transformations to NMR spectra, automated phase correction, overall signal check for missing/extra peaks, background control, baseline removal, and spectral area-specific signal alignments, and comparisons with the spectra of the two quality control samples. The NMR metabolomics (Nightingale Health Ltd., Helsinki, Finland) method has also been used and described in refs. 68 , 69 , 70 , 71 , 72 , 73 . The effects of these experimental aspects on the metabolic biomarker concentrations are best reflected in the coefficients of variation (CV) for the measurements, and representative CVs have been published previously in the supplement of refs. 73 , 74 in which mean CV (%) was 4.5% and 5.0%, respectively. The relation of quantified biomarker data to annotated NMR spectral data has recently been discussed extensively in ref. 72 and additional spectral data have been made publicly available in ( https://www.ebi.ac.uk/metabolights/MTBLS974 ). The lipoprotein lipid measures were quantified using gel permeation high-performance liquid chromatography (GP HPLC) as calibration reference, as described previously 65 , 66 , 75 . The GP HPLC assay captures the cholesterol, triglyceride, and phospholipid levels in lipoprotein subclasses, which in turn are accurately reproduced in the Nightingale NMR platform in a high-throughput manner ( https://pubmed.ncbi.nlm.nih.gov/32359769/ ). The quantified metabolite measures have also been compared with alternatively analytical methods for measuring the same metabolites, including two commercial mass-spectrometric platforms, showing good consistency 76 . Numerous published large epidemiological studies have used the Nightingale Health platform (see https://nightingalehealth.com/publications for an overview) and the platform is currently being used to measure all 500,000 samples from the UK Biobank 77 . For this study, and all prior publications using the Nightingale platform, we used the quantified biomarker measures in absolute concentration units or ratios provided directly from the commercial metabolomics platform, and no raw spectral data were used in epidemiological analyses.

Quantification of serum fatty acids

Fatty acid methyl esters (FAMEs) from whole plasma were determined as follows. In a chloroform-resistant Eppendorf, 30 µL of plasma were spiked with 65 µL of Internal standard (ISTD) nonadecanoic acid (C19:0) (Merck) 100 µg/mL solution (6.5 µg). Spiked plasma was extracted with chloroform–methanol (2:1, v/v). Organic phase was transferred into a screw-cap test tube and evaporated to dryness under N2 at 37 °C. Plasma fatty acids were hydrolized and methylated following an adaptation of the method described by Agren et al. 78 . Briefly, 100 µL of n-toluene and 500 µL of boron trifluoride–methanol solution (14%) were added to the tube, which was capped and placed into a block heather (100 °C) for 60 min. After cooling, 500 µL of distilled water and 500 µL of n-hexane were added. After shaking for 1 min, the tubes were centrifuged for 5 min at 2200 × g at room temperature to separate the layers. The hexane layer was placed into a test tube and evaporated to dryness under N 2 at 30 °C. The extracts were reconstituted with 100 µL of n-hexane and transferred into an automatic injector vial equipped with a glass insert of 300 µL.

FAMEs were analyzed by GC/MS-EI using an Agilent 6890 N GC equipped with an Agilent 7683 autosampler, and an Agilent 5973 N mass spectrometry detector. FAMES was separated with a J&W DB-FastFAME capillary column (30 m × 0.2 mm × 0.25-μm film thickness) (Agilent). The injector temperature was set at 250 °C, and 1-μL injections were made (split ratio 25:1). GC was run using an optimized temperature program as follows: the temperature program started at 50 °C, held for 0.5 min, increased to 194 °C at a rate of 25 °C/min, held for 1 min, and increased to 245 °C at the rate of 5 °C/min, held for 3 min. Helium was used as a carrier gas (14 psi, constant pressure mode). FAMEs were detected using selected ion monitoring (SIM) mode. Several m/z ions common to saturated, monounsaturated, and polyunsaturated FAMES were monitored (see detailed information in Supplementary Table  4 ). All data were quantified by integrating the area under the curve of each metabolite using MassHunter Quant (Agilent Technologies). Nine mixtures of FAME external calibration standards were prepared by dilution in hexane of certified FAME reference material mix (Supelco 37 Component FAME Mix, Merck) and kept at −20 °C until analysis. About 30 µL of each mixture was added to a tube, was spiked with 650 µL of ISTD C19:0 methyl ester 100 µg/mL solution (6.5 µg), evaporated to dryness under N 2 at 30 °C, reconstituted with 100 µL of hexane, and transferred into an automatic injector vial equipped with a glass insert of 300 µL. The equivalents of C19:0 added to the samples, as free fatty acid ISTD, were the same as the amount of C19:0–methyl ester added to the external calibrators. The concentration of FAMES in the samples was calculated by linear regression of the peak area ratio relative to that of the internal standard.

Statistics and reproducibility

Statistical analysis was performed with SPSS v19.0 (SPSS, Inc., Chicago, IL) for Windows, with GraphPad prism 7 (GraphPad Software Inc., La Jolla, CA), and with MetaboanalystR 79 . The results are presented in figures and tables as mean ± SEM. In study 3 ( n  = 7), individual data points are shown. Mixed model adjusted for baseline was used for all parameters (unless otherwise specified). In fatty acids deriving from GC/MS-EI, where values were normalized to baseline (thus, ratios were created), mixed model without adjustment for baseline was used. Compound symmetry was selected as repeated covariance type. Factors were time (corresponded to days or hours of fasting for study 1 and study 2, days of treatment for study 3, and weeks of treatment for study 4) and group (fed, fasting treated with placebo, and fasting treated with leptin for study 1; physiologic (0.01 mg/kg), supraphysiologic (0.1 mg/kg), pharmacologic (0.3 mg/kg) for study 2, and leptin or placebo for study 4). In study 2, two additional analyses were performed that included as factors time (corresponded to hours of fasting or to hours after leptin administration in fed state) and group (lean men, lean women, and obese men) for each leptin dose separately. When only the group factor- existed p value was calculated with unpaired t test (e.g., Fig.  2a leptin at baseline and % weight change), with one-way ANOVA (Fig.  2b leptin at baseline and % weight change), and with repeated measure ANOVA (Fig.  3a food intake and Fig.  5a urine catecholamines). When only the time factor existed and only two timepoints were available for analysis, paired t test was used (study 3: Fig.  3e RMR and temperature, Fig.  4c and Supplementary Fig.  4b ). Post hoc Bonferroni corrections were performed between the total means of the groups, as well as between the means of the groups in the individual timepoints when the time*group P value < 0.05. When only the group factor existed, post hoc Bonferroni corrections were performed when P  < 0.05. When only the time factor existed, post hoc Bonferroni’s test comparing the timepoint vs baseline was performed when P  < 0.05. Curve estimation was performed in SPSS to choose the best fit and ANOVA to calculate the corresponding P values for the logarithmic curve to show relationships between variables. One outlier according to the ROUT method (GraphPad) with a Q = 0.1% was detected and removed from the urine cortisol measurements in study 1 and from the FFA measurements of study 4. Pearson’s or Spearman’s (when data were not normally distributed) correlation tests were used for correlations between variables. For the correlations, an additional analysis by using repeated measure correlation (rmcorr package for R) was performed, in order to adjust for the fact that some subjects have contributed in the correlation more than one point. This was the case for (a) study 2, correlation of baseline leptin levels with % of weight loss (Supplementary Fig.  1a , adjustment was performed since each subject has contributed three points corresponding to three different leptin doses), (b) study 1, correlation of food intake to baseline leptin levels in an ad libitum meal intake after fasting (Supplementary Fig.  1e , adjustment was performed for fasting since each subject has contributed two points, corresponding to leptin or placebo treatment; values were logarithmically transformed since they were not linear), and (c) study 3 and Study 4, correlations of hormones with FFA (Supplementary Table  1 , adjustment was performed, since each subject has multiple FFA and hormonal measurements in different timepoints).

For study 1, measurements from NMR and GC/MS-EI were analyzed together both for all three admissions (Fig.  6a —fed, fasting + leptin, fasting + placebo) as well as only for the two fasting admissions (Fig.  6c, d —fasting + leptin, fasting + placebo). For study 2 and study 4, measurements from NMR (but not from GC/MS-EI) were available and analyzed (Supplementary Fig.  6 for study 2 and Fig.  6e for study 4). Three parameters (i.e., pyruvate, C18:1trans, C24:0) and five timepoints (i.e., in day 1 two subjects treated with leptin and one with placebo, in day 2 one treated with placebo, and in day 3 one treated with leptin) had more than 50% missing values (due to low or no detection, lack of quantification due to sample irregularities, or rejection by automatic quality controls) and were excluded from the analysis in study 1. Similarly, four parameters (pyruvate, glycerol, phenylalanine, and C18:1trans) in study 2 and one parameter (pyruvate) in study 4 had more than 50% missing values and were excluded. In all other cases, the missing values (5,4% of the whole data set) were replaced by half of the minimum value in the original data set to create a processed data set. Subsequently, the data from the processed data set were mean-centered and divided by the standard deviation of each variable to create a normalized–scaled data set. The means ± SEM of the processed data set for study 1 for all three admissions is provided in Supplementary Data  1 , as well as for selected (most important) parameters of study 1 and study 2 in Supplementary Fig.  5 and Supplementary Fig.  6b–d , respectively. In Fig.  6a , the normalized–scaled data set for all three admissions was used to perform a one-way ANOVA with an adjusted FDR to the number of measurements ( 0.05/n) , i.e., 2.15 × 10 −4 . This identified 68 significantly different parameters between the three admissions, which were then used to create the heatmap demonstrated in Fig.  6b . In study 1, 2, and 4, sparse partial least-squares discriminant analysis (sPLS-DA) was performed with a setup of five components, each consisting of a maximum of 10 variables and with a fivefold CV validation. Similarly, the normalized–scaled data set for the two admissions (fasting + placebo vs fasting + leptin) was used to perform the sPLS-DA analysis and present the score plot and parameters involved in components 1 and 3 in Fig.  6c, d (study 1). Similarly, the normalized–scaled data set was used to perform the one-way ANOVA, the sPLS-DA, and present the score plot in study 4 (demonstrated in Fig.  6e ). Finally, the normalized–scaled data set that included the measurements at start and completion of each admission was used to perform the sPLS-DA with score plot and component 2 in Study 2 (demonstrated in Supplementary Fig.  6a ).

In study 1, for the five timepoints excluded from combined NMR and GC/MS-EI sPLS-DA and one-way ANOVA analysis in Fig.  6 due to >50% missing values, FAME values were though available. Thus, an additional analysis of FAMEs was performed that included these 5 timepoints. In this analysis, additionally the FAME ratio was calculated, to adjust for baseline differences due to lipid extraction—lipid volume loaded as well as due to expected physiological variability between three admissions performed in different timepoints. Missing baseline measurement of an admission (i.e., fed or fasting + leptin or fasting + placebo) was replaced with the average of the baseline values of the other two admissions of the same subject. Ratios were then calculated and analyzed with mixed models without adjustment for baseline and are presented in Fig.  7 .

Regarding study reproducibility, all studies have been performed in the past. Independent repeat of any part of the experiment—studies were neither in our aims nor possible due to restrictions in the administration of leptin in humans with the exception of people with congenital leptin deficiency or generalized lipodystrophy where an FDA approval exists. Thus, we have (a) used existing data from older measurements (see Supplementary Table  2 ), (b) we have performed new measurements in already-collected and appropriately stored samples from the above studies. New hormonal measurements (i.e., renin and aldosterone) and FFA measurements of the samples were performed in duplicate and were repeated if coefficient of variation for any sample was >15%. The value deriving from the repetition was considered the valid one if the coefficient of variation was <15%. The metabolite–lipoprotein–fatty acid measurements were performed by state-of-the-art methods described above, with de-identified samples and blinded personnel regarding the different groups.

Additional resources

Clinical trial registry numbers are the following: study 1: NCT00140231; study 2: NCT00140205; study 4: NCT00130117 available in ClinicalTrials.gov. Flow diagrams of all four studies are available in the Supplementary Material (Supplementary Fig.  7 ).

Reporting summary

Further information on research design is available in the  Nature Research Reporting Summary linked to this article.

Data availability

Any other data that support the findings of this study are available from the corresponding author upon reasonable request. A list of all the identified metabolites along with relevant identifying information is available in Supplementary Data  2 . Source data are provided with this paper.

Frederich, R. C. et al. Leptin levels reflect body lipid content in mice: evidence for diet-induced resistance to leptin action. Nat. Med. 1 , 1311–1314 (1995).

Article   CAS   PubMed   Google Scholar  

Considine, R. V. et al. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N. Engl. J. Med. 334 , 292–295 (1996).

Boden, G., Chen, X., Mozzoli, M. & Ryan, I. Effect of fasting on serum leptin in normal human subjects. J. Clin. Endocrinol. Metab. 81 , 3419–3423 (1996).

CAS   PubMed   Google Scholar  

Wu, B. et al. Effect of repeated fasting/refeeding on body weight control and energy balance regulation in rats. Wei Sheng Yan Jiu 39 , 601–605 (2010).

PubMed   Google Scholar  

Audi, L. et al. Leptin in relation to resumption of menses in women with anorexia nervosa. Mol. Psychiatry 3 , 544–547 (1998).

Halaas, J. L. et al. Weight-reducing effects of the plasma protein encoded by the obese gene. Science 269 , 543–546 (1995).

Article   ADS   CAS   PubMed   Google Scholar  

Singh, A. et al. Leptin-mediated changes in hepatic mitochondrial metabolism, structure, and protein levels. Proc. Natl Acad. Sci. USA 106 , 13100–13105 (2009).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Zeng, W. et al. Sympathetic neuro-adipose connections mediate leptin-driven lipolysis. Cell 163 , 84–94 (2015).

Halaas, J. L. et al. Physiological response to long-term peripheral and central leptin infusion in lean and obese mice. Proc. Natl Acad. Sci. USA 94 , 8878–8883 (1997).

Brown, R. J. et al. Long-term effectiveness and safety of metreleptin in the treatment of patients with generalized lipodystrophy. Endocrine 60 , 479–489 (2018).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Farooqi, I. S. et al. Beneficial effects of leptin on obesity, T cell hyporesponsiveness, and neuroendocrine/metabolic dysfunction of human congenital leptin deficiency. J. Clin. Investig. 110 , 1093–1103 (2002).

Lee, J. H., Chan, J. L., Sourlas, E., Raptopoulos, V. & Mantzoros, C. S. Recombinant methionyl human leptin therapy in replacement doses improves insulin resistance and metabolic profile in patients with lipoatrophy and metabolic syndrome induced by the highly active antiretroviral therapy. J. Clin. Endocrinol. Metab. 91 , 2605–2611 (2006).

Oral, E. A. et al. Long-term effectiveness and safety of metreleptin in the treatment of patients with partial lipodystrophy. Endocrine 64 , 500–511 (2019).

Heymsfield, S. B. et al. Recombinant leptin for weight loss in obese and lean adults: a randomized, controlled, dose-escalation trial. J. Am. Med. Assoc. 282 , 1568–1575 (1999).

Article   CAS   Google Scholar  

Mittendorfer, B. et al. Recombinant human leptin treatment does not improve insulin action in obese subjects with type 2 diabetes. Diabetes 60 , 1474–1477 (2011).

Moon, H. S. et al. Efficacy of metreleptin in obese patients with type 2 diabetes: cellular and molecular pathways underlying leptin tolerance. Diabetes 60 , 1647–1656 (2011).

Shetty, G. K. et al. Leptin administration to overweight and obese subjects for 6 months increases free leptin concentrations but does not alter circulating hormones of the thyroid and IGF axes during weight loss induced by a mild hypocaloric diet. Eur. J. Endocrinol. 165 , 249–254 (2011).

Fogteloo, A. J., Pijl, H., Frolich, M., McCamish, M. & Meinders, A. E. Effects of recombinant human leptin treatment as an adjunct of moderate energy restriction on body weight, resting energy expenditure and energy intake in obese humans. Diabetes, Nutr. Metab. 16 , 109–114 (2003).

CAS   Google Scholar  

Hukshorn, C. J. et al. Weekly subcutaneous pegylated recombinant native human leptin (PEG-OB) administration in obese men. J. Clin. Endocrinol. Metab. 85 , 4003–4009 (2000).

Hukshorn, C. J. et al. The effect of pegylated recombinant human leptin (PEG-OB) on weight loss and inflammatory status in obese subjects. Int. J. Obes. Relat. Metab. Disord.: J. Int. Assoc. Study Obes. 26 , 504–509 (2002).

Zelissen, P. M. et al. Effect of three treatment schedules of recombinant methionyl human leptin on body weight in obese adults: a randomized, placebo-controlled trial. Diabetes, Obes. Metab. 7 , 755–761 (2005).

Lejeune, M. P., Hukshorn, C. J., Saris, W. H. & Westerterp-Plantenga, M. S. Effect of dietary restraint during and following pegylated recombinant leptin (PEG-OB) treatment of overweight men. Int. J. Obes. Relat. Metab. Disord.: J. Int. Assoc. Study Obes. 27 , 1494–1499 (2003).

DePaoli, A., Long, A., Fine, G. M., Stewart, M. & O’Rahilly, S. Efficacy of metreleptin for weight loss in overweight and obese adults with low leptin levels. Diabetes 67 , 296 (2018).

Article   Google Scholar  

Flier, J. S. & Maratos-Flier, E. Leptin’s physiologic role: does the emperor of energy balance have no clothes? Cell Metab. 26 , 24–26 (2017).

Friedman, J. M. Leptin and the endocrine control of energy balance. Nat. Metab. 1 , 754–764 (2019).

Chan, J. L., Heist, K., DePaoli, A. M., Veldhuis, J. D. & Mantzoros, C. S. The role of falling leptin levels in the neuroendocrine and metabolic adaptation to short-term starvation in healthy men. J. Clin. Investig. 111 , 1409–1421 (2003).

Chan, J. L. et al. Differential regulation of metabolic, neuroendocrine, and immune function by leptin in humans. Proc. Natl Acad. Sci. USA 103 , 8481–8486 (2006).

Chou, S. H. et al. Leptin is an effective treatment for hypothalamic amenorrhea. Proc. Natl Acad. Sci. USA 108 , 6585–6590 (2011).

Welt, C. K. et al. Recombinant human leptin in women with hypothalamic amenorrhea. N. Engl. J. Med. 351 , 987–997 (2004).

Chan, J. L., Wong, S. L. & Mantzoros, C. S. Pharmacokinetics of subcutaneous recombinant methionyl human leptin administration in healthy subjects in the fed and fasting states: regulation by gender and adiposity. Clin. Pharmacokinetics 47 , 753–764 (2008).

Oral, E. A. et al. Leptin-replacement therapy for lipodystrophy. N. Engl. J. Med. 346 , 570–578 (2002).

Mackowiak, P. A., Wasserman, S. S. & Levine, M. M. A critical appraisal of 98.6 degrees F, the upper limit of the normal body temperature, and other legacies of Carl Reinhold August Wunderlich. J. Am. Med. Assoc. 268 , 1578–1580 (1992).

Gilgen-Ammann, R. et al. Energy expenditure estimation from respiration variables. Sci. Rep. 7 , 15995 (2017).

Article   ADS   PubMed   PubMed Central   CAS   Google Scholar  

Steinhauser, M. L. et al. The circulating metabolome of human starvation. JCI insight 3 , e121434 (2018).

Article   PubMed Central   Google Scholar  

Perry, R. J. et al. Leptin mediates a glucose-fatty acid cycle to maintain glucose homeostasis in starvation. Cell 172 , 234–248 (2018).

Rosenbaum, M. & Leibel, R. L. Physiological responses to leptin levels in lipodystrophy: a model for other hypoleptinemias? J. Clin. Investig. 128 , 3237–3239 (2018).

Article   PubMed   PubMed Central   Google Scholar  

Pelleymounter, M. A. et al. Effects of the obese gene product on body weight regulation in ob/ob mice. Science 269 , 540–543 (1995).

Tang, G. B., Tang, X. F., Li, K. & Wang, D. H. Intracerebroventricular administration of leptin increase physical activity but has no effect on thermogenesis in cold-acclimated rats. Sci. Rep. 5 , 11189 (2015).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Pandit, R., Beerens, S. & Adan, R. A. H. Role of leptin in energy expenditure: the hypothalamic perspective. Am. J. Physiol. Regul. Integr. Comp. Physiol. 312 , R938–R947 (2017).

Correia, M. L., Morgan, D. A., Sivitz, W. I., Mark, A. L. & Haynes, W. G. Leptin acts in the central nervous system to produce dose-dependent changes in arterial pressure. Hypertension 37 , 936–942 (2001).

Machleidt, F. et al. Experimental hyperleptinemia acutely increases vasoconstrictory sympathetic nerve activity in healthy humans. J. Clin. Endocrinol. Metab. 98 , E491–E496 (2013).

Morton, G. J. et al. Identification of a physiological role for leptin in the regulation of ambulatory activity and wheel running in mice. Am. J. Physiol. Endocrinol. Metab. 300 , E392–E401 (2011).

Satoh, N. et al. Sympathetic activation of leptin via the ventromedial hypothalamus: leptin-induced increase in catecholamine secretion. Diabetes 48 , 1787–1793 (1999).

Shibuya, I. et al. Regulation of catecholamine synthesis by leptin. Ann. N. Y Acad. Sci. 971 , 522–527 (2002).

Korner, J. et al. Randomized double-blind placebo-controlled study of leptin administration after gastric bypass. Obesity 21 , 951–956 (2013).

Kissileff, H. R. et al. Leptin reverses declines in satiation in weight-reduced obese humans. Am. J. Clin. Nutr. 95 , 309–317 (2012).

Rosenbaum, M. et al. Low-dose leptin reverses skeletal muscle, autonomic, and neuroendocrine adaptations to maintenance of reduced weight. J. Clin. Investig. 115 , 3579–3586 (2005).

Rosenbaum, M., Sy, M., Pavlovich, K., Leibel, R. L. & Hirsch, J. Leptin reverses weight loss-induced changes in regional neural activity responses to visual food stimuli. J. Clin. Investig. 118 , 2583–2591 (2008).

CAS   PubMed   PubMed Central   Google Scholar  

Ahima, R. S. et al. Role of leptin in the neuroendocrine response to fasting. Nature 382 , 250–252 (1996).

Hwa, J. J. et al. Leptin increases energy expenditure and selectively promotes fat metabolism in ob/ob mice. Am. J. Physiol. 272 , R1204–R1209 (1997).

Lawler, K. et al. Leptin-mediated changes in the human metabolome. J. Clin. Endocrinol. Metab. 105 , dgaa251 (2020).

Grewal, S. et al. Metabolomic analysis of the effects of leptin replacement therapy in patients with lipodystrophy. J. Endocr. Soc. 4 , bvz022 (2020).

Article   PubMed   CAS   Google Scholar  

Clemmons, D. R. Metabolic actions of insulin-like growth factor-I in normal physiology and diabetes. Endocrinol. Metab. Clin. North Am. 41 , 425–443 (2012).

Kaklamani, V. G. et al. Dietary fat and carbohydrates are independently associated with circulating insulin-like growth factor 1 and insulin-like growth factor-binding protein 3 concentrations in healthy adults. J. Clin. Oncol. 17 , 3291–3298 (1999).

Kirsten, R. et al. Effects of aldosterone on lipid metabolism and renal oxygen consumption in the rat. Pflug. Arch. 368 , 189–194 (1977).

Shimomura, I., Hammer, R. E., Ikemoto, S., Brown, M. S. & Goldstein, J. L. Leptin reverses insulin resistance and diabetes mellitus in mice with congenital lipodystrophy. Nature 401 , 73–76 (1999).

Puschel, J. et al. Beneficial effects of leptin substitution on impaired eating behavior in lipodystrophy are sustained beyond 150 weeks of treatment. Cytokine 113 , 400–404 (2019).

Schlogl, H. et al. Leptin substitution in patients with lipodystrophy: neural correlates for long-term success in the normalization of eating behavior. Diabetes 65 , 2179–2186 (2016).

Farr, O. M. et al. Leptin therapy alters appetite and neural responses to food stimuli in brain areas of leptin-sensitive subjects without altering brain structure. J. Clin. Endocrinol. Metab. 99 , E2529–E2538 (2014).

Westerterp-Plantenga, M. S., Saris, W. H., Hukshorn, C. J. & Campfield, L. A. Effects of weekly administration of pegylated recombinant human OB protein on appetite profile and energy metabolism in obese men. Am. J. Clin. Nutr. 74 , 426–434 (2001).

Dallner, O. S. et al. Dysregulation of a long noncoding RNA reduces leptin leading to a leptin-responsive form of obesity. Nat. Med. 25 , 507–516 (2019).

Chan, J. L., Mietus, J. E., Raciti, P. M., Goldberger, A. L. & Mantzoros, C. S. Short-term fasting-induced autonomic activation and changes in catecholamine levels are not mediated by changes in leptin levels in healthy humans. Clin. Endocrinol. 66 , 49–57 (2007).

Ainsworth, B. E. et al. 2011 Compendium of physical activities: a second update of codes and MET values. Med. Sci. sports Exerc. 43 , 1575–1581 (2011).

Article   PubMed   Google Scholar  

Sumner, L. W. et al. Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI). Metabolomics 3 , 211–221 (2007).

Soininen, P., Kangas, A. J., Wurtz, P., Suna, T. & Ala-Korpela, M. Quantitative serum nuclear magnetic resonance metabolomics in cardiovascular epidemiology and genetics. Circ. Cardiovasc. Genet. 8 , 192–206 (2015).

Soininen, P. et al. High-throughput serum NMR metabonomics for cost-effective holistic studies on systemic metabolism. Analyst 134 , 1781–1785 (2009).

Inouye, M. et al. Metabonomic, transcriptomic, and genomic variation of a population cohort. Mol. Syst. Biol. 6 , 441 (2010).

Zuber, V., Colijn, J. M., Klaver, C. & Burgess, S. Selecting likely causal risk factors from high-throughput experiments using multivariable Mendelian randomization. Nat. Commun. 11 , 29 (2020).

Hagenbeek, F. A. et al. Heritability estimates for 361 blood metabolites across 40 genome-wide association studies. Nat. Commun. 11 , 39 (2020).

Vojinovic, D. et al. Relationship between gut microbiota and circulating metabolites in population-based cohorts. Nat. Commun. 10 , 5813 (2019).

Gallois, A. et al. A comprehensive study of metabolite genetics reveals strong pleiotropy and heterogeneity across time and context. Nat. Commun. 10 , 4788 (2019).

Deelen, J. et al. A metabolic profile of all-cause mortality risk identified in an observational study of 44,168 individuals. Nat. Commun. 10 , 3346 (2019).

Kettunen, J. et al. Genome-wide study for circulating metabolites identifies 62 loci and reveals novel systemic effects of LPA. Nat. Commun. 7 , 11122 (2016).

Holmes, M. V. et al. Lipids, lipoproteins, and metabolites and risk of myocardial infarction and stroke. J. Am. Coll. Cardiol. 71 , 620–632 (2018).

Wurtz, P. & Soininen, P. Reply to: “Methodological issues regarding: “a third of nonfasting plasma cholesterol is in remnant lipoproteins: lipoprotein subclass profiling in 9293 individuals””. Atherosclerosis 302 , 59–61 (2020).

Liu, J. et al. Integration of epidemiologic, pharmacologic, genetic and gut microbiome data in a drug-metabolite atlas. Nat. Med. 26 , 110–117 (2020).

Julkunen, H., Cichonska, A., Slagboom, P. E. & Würtz, P. Blood biomarker score identifies individuals at high risk for severe COVID-19 a decade prior to diagnosis: metabolic profiling of 105,000 adults in the UK Biobank. Preprint at https://doi.org/10.1101/2020.07.02.20143685 (2020).

Agren, J. J., Julkunen, A. & Penttila, I. Rapid separation of serum lipids for fatty acid analysis by a single aminopropyl column. J. Lipid Res. 33 , 1871–1876 (1992).

Chong, J. & Xia, J. MetaboAnalystR: an R package for flexible and reproducible analysis of metabolomics data. Bioinformatics 34 , 4313–4314 (2018).

Download references

Acknowledgements

This study was funded by NIH K24DK081913. Ni.P. was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation)—389891681 (PE 2431/2-1). A.S.V. is the recipient of the Instituto de Salud Carlos III Miguel Servet fellowship (grant CP II 17/00029). P.C. received a 2018 Summer Research Fellowship (SRF) funded by the Endocrine Society to support participation in the research project.

Author information

These authors contributed equally: Pavlina Chrysafi, Nikolaos Perakakis.

Authors and Affiliations

Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, 02215, USA

Pavlina Chrysafi, Nikolaos Perakakis, Olivia M. Farr, Konstantinos Stefanakis, Natia Peradze & Christos S. Mantzoros

Barcelonaβeta Brain Research Center, Pasqual Maragall Foundation, Barcelona, Spain

Aleix Sala-Vila

Institut Hospital del Mar d’Investigacions Mèdiques (IMIM), Barcelona, Spain

You can also search for this author in PubMed   Google Scholar

Contributions

C.S.M. designed the experiment. P.C., N.i.P., O.M.F., N.a.P., A.S.V., K.S., and C.S.M. conducted the experiments and acquired the data. P.C. and N.i.P. analyzed the data. P.C. and N.i.P. wrote the paper with input from all of the other authors.

Corresponding author

Correspondence to Christos S. Mantzoros .

Ethics declarations

Competing interests.

C.S.M. is advisor of Ansh Labs LLC, and consultant to Novo Nordisk and grant recipient through BIDMC and has served as an advisor to Aegerion and Visiting Professor to Regeneron. The remaining authors declare no competing interests.

Additional information

Peer review information Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information, peer review file, description of additional supplementary files, supplementary data 1, supplementary data 2, reporting summary, source data, source data, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Chrysafi, P., Perakakis, N., Farr, O.M. et al. Leptin alters energy intake and fat mass but not energy expenditure in lean subjects. Nat Commun 11 , 5145 (2020). https://doi.org/10.1038/s41467-020-18885-9

Download citation

Received : 24 May 2020

Accepted : 18 September 2020

Published : 13 October 2020

DOI : https://doi.org/10.1038/s41467-020-18885-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Total and h-specific gdf-15 levels increase in caloric deprivation independently of leptin in humans.

• Pavlina Chrysafi
• Laura Valenzuela-Vallejo
• Christos S. Mantzoros

Nature Communications (2024)

Leptin/obR signaling exacerbates obesity-related neutrophilic airway inflammation through inflammatory M1 macrophages

• Rongjun Wan
• Chengping Hu

Molecular Medicine (2023)

Lipid profile after omega-3 supplementation in neonates with intrauterine growth retardation: a randomized controlled trial

• Mai Elsheikh
• Doaa El Amrousy
• Amany El-Barky

Pediatric Research (2023)

The association of appetite and hormones (leptin, ghrelin, and Insulin) with resting metabolic rate in overweight/ obese women: a case–control study

• Sara Hajishizari
• Hossein Imani
• Khadijeh Mirzaei

BMC Nutrition (2022)

The ambiguous role of obesity in oncology by promoting cancer but boosting antitumor immunotherapy

• José Antônio Fagundes Assumpção
• Gabriel Pasquarelli-do-Nascimento
• Kelly Grace Magalhães

Journal of Biomedical Science (2022)

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

• Open access
• Published: 28 June 2021

Novel insights into the genetically obese ( ob/ob ) and diabetic ( db/db ) mice: two sides of the same coin

• Francesco Suriano 1 ,
• Sara Vieira-Silva 2 ,
• Gwen Falony 2 ,
• Martin Roumain 3 ,
• Adrien Paquot 3 ,
• Rudy Pelicaen 1 ,
• Marion Régnier 1 ,
• Nathalie M. Delzenne 1 ,
• Jeroen Raes 2 ,
• Giulio G. Muccioli 3 ,
• Matthias Van Hul 1 &
• Patrice D. Cani   ORCID: orcid.org/0000-0003-2040-2448 1  

Microbiome volume  9 , Article number:  147 ( 2021 ) Cite this article

26k Accesses

109 Citations

14 Altmetric

Metrics details

Leptin-deficient ob/ob mice and leptin receptor-deficient db/db mice are commonly used mice models mimicking the conditions of obesity and type 2 diabetes development. However, although ob/ob and db/db mice are similarly gaining weight and developing massive obesity, db/db mice are more diabetic than ob/ob mice. It remains still unclear why targeting the same pathway—leptin signaling—leads to the development of two different phenotypes. Given that gut microbes dialogue with the host via different metabolites (e.g., short-chain fatty acids) but also contribute to the regulation of bile acids metabolism, we investigated whether inflammatory markers, bacterial components, bile acids, short-chain fatty acids, and gut microbes could contribute to explain the specific phenotype discriminating the onset of an obese and/or a diabetic state in ob/ob and db/db mice.

Six-week-old ob/ob and db/db mice were followed for 7 weeks; they had comparable body weight, fat mass, and lean mass gain, confirming their severely obese status. However, as expected, the glucose metabolism and the glucose-induced insulin secretion were significantly different between ob/ob and db/db mice. Strikingly, the fat distribution was different, with db/db mice having more subcutaneous and ob/ob mice having more epididymal fat. In addition, liver steatosis was more pronounced in the ob/ob mice than in db/db mice. We also found very distinct inflammatory profiles between ob/ob and db/db mice, with a more pronounced inflammatory tone in the liver for ob/ob mice as compared to a higher inflammatory tone in the (subcutaneous) adipose tissue for db/db mice. When analyzing the gut microbiota composition, we found that the quantity of 19 microbial taxa was in some way affected by the genotype. Furthermore, we also show that serum LPS concentration, hepatic bile acid content, and cecal short-chain fatty acid profiles were differently affected by the two genotypes.

Taken together, our results elucidate potential mechanisms implicated in the development of an obese or a diabetic state in two genetic models characterized by an altered leptin signaling. We propose that these differences could be linked to specific inflammatory tones, serum LPS concentration, bile acid metabolism, short-chain fatty acid profile, and gut microbiota composition.

Video abstract.

Over the past 40 years, obesity has reached epidemic proportions and has become a huge public health and economic issue since it is a major contributor to several metabolic comorbidities, including insulin resistance, type 2 diabetes (T2D), and liver diseases [ 1 , 2 , 3 ]. Obesity is characterized by an imbalance between energy intake and energy expenditure [ 4 , 5 ], although its prevalence within individuals varies with behavior, genetic, environmental, and physiological factors [ 6 ]. It is well established that obesity is associated with a state of chronic, low-grade inflammation distinguished by the production of several inflammatory cytokines and adipokines [ 7 ]. In the last two decades, the gut microbiota has emerged as a fundamental environmental factor modulating whole-body metabolism by influencing energy balance, glucose metabolism, gut barrier function, and low-grade inflammation among others [ 8 ]. Numerous metabolic functions can be traced back to microbial metabolites, of which the short-chain fatty acids (SCFAs) are the most studied and have been associated with several metabolic functions [ 9 , 10 ]. Moreover, the gut microbiota has been shown to modulate the bile acid (BA) profile, mainly by metabolizing primary BA into secondary BA, thus increasing their chemical diversity. These molecules are also known for regulating several host metabolic processes [ 11 ].

Obesity is a risk factor in which several organs and systems are involved. Among these, the liver and adipose tissue play a central role. Contrary to the metabolic function of the liver, the adipose tissue has the capacity to store and release energy under the form of lipids as well as the ability to act as an active endocrine organ capable of synthesizing a wide variety of biologically active compounds (i.e., adipokines) that are involved in the regulation of several metabolic pathways [ 12 ]. The best-known adipokine is leptin, which is mainly produced by mature adipocytes. Besides its role in satiety, leptin plays an important role in the regulation of energy homeostasis, lipid and glucose metabolism, and the immune response via the cognate leptin receptor (ObR) [ 13 , 14 ]. Alterations in leptin signaling are closely associated with metabolic diseases, such as obesity and T2D [ 15 ]. The genetic leptin-deficient ob/ob mice and the leptin-resistant db/db mice are therefore widely used as animal models to study obesity and related metabolic disorders [ 16 , 17 , 18 , 19 ]. Ob/ob mice are characterized by a mutation of the obese ( ob ) gene encoding leptin, whereas the db/db mice have a mutation of the diabetes ( db ) gene encoding for the ObR [ 20 ]. Both mouse models have defective leptin signaling with a complete lack of leptin production in ob/ob mice and an overexpression of circulating leptin in db/db mice to which they cannot respond due to a complete deficiency of the long isoform of the leptin receptor (ObRb) [ 15 ]. Despite a different underlying molecular mechanism at the base of the leptin deficiency (ligand versus receptor), both models show a similar phenotype in regard to hyperphagia, hypometabolism, and obesity, but manifest different impairments in glucose metabolism [ 20 , 21 ]. Indeed, the ob/ob mice develop obesity and mild insulin resistance, while the db/db mice develop obesity and diabetes. These differences are not yet fully understood as many mechanistic details associating leptin signaling with the development of an obese and a diabetic state remain poorly investigated. Recent studies using both genetic models have identified novel markers of obesity and T2D [ 18 ], as well as a different gut microbiota composition across different ages that were closely linked to fluctuations in blood glucose [ 22 ]. However, identification of novel mediators and a better understanding of the different metabolic pathways associated with the leptin signaling could result in the development of new potential therapeutic strategies to tackle obesity and its related metabolic disorders. This study aimed at explaining why despite having the same fat mass and the same body weight, the onset of metabolic complications observed in both ob/ob and db/db mice matched by age and sex and fed an identical diet for 7 weeks were different. To explore this hypothesis, we have characterized inflammatory markers, bacterial components, BA, SCFAs, and gut microbes.

Mice and experimental design

Male homozygous ob/ob mice (B6.V-Lepob/ob/JRj) were used as a leptin-deficient obese model, and their lean littermates served as controls (CT ob); (n = 9–10 per group). Male homozygous db/db mice (BKS-Lepr/db/db/JOrlRj), functionally deficient for the long-form leptin receptor, were used as a hyperleptinemic obese type 2 diabetic model, and their lean littermates served as controls (CT db); (n = 9–10 per group). Mice were purchased at the same time and from the same supplier (Janvier Laboratories, Le Genest-Saint-Isle, France) at the age of 6 weeks. Mice were housed in a specific pathogen- and opportunistic-free (SOPF) controlled environment (room temperature of 22 ± 2 °C, humidity 55 ± 10%, 12 h daylight cycle, lights off at 6 p.m.) in groups of two mice per cage, with free access to sterile food and sterile water. Upon delivery, mice underwent an acclimation period of one week, during which they were fed a standard diet containing 10% calories from fat (D12450Ji; Research Diet; New Brunswick, NJ, USA) and were then kept ad libitum on the same diet for 7 weeks. Milli-Q water filtered by a Millipak® Express 40 with a 0.22-μm membrane filter (Merck Millipore, Burlington, Massachusetts, USA) was autoclaved and provided ad libitum. All mouse experiments were approved by and performed in accordance with the guideline of the local ethics committee (Ethics committee of the Université catholique de Louvain for Animal Experiments specifically approved this study that received the agreement number 2017/UCL/MD/005). Housing conditions were specified by the Belgian Law of 29 May 2013, regarding the protection of laboratory animals (agreement number LA1230314).

Measurements during the study

Body weight, food, and water intake were recorded three times per week. Body composition was assessed weekly by using 7.5-MHz time domain-nuclear magnetic resonance (TD-NMR) (LF50 Minispec; Bruker; Rheinstetten, Germany).

Oral glucose tolerance test and insulin resistance index

In the 6th week of the experiment, mice were fasted for 6 h and given an oral glucose load (1 g glucose per kg body weight). Blood glucose was measured 30 min before oral glucose load (− 30 min) and 15, 30, 60, 90, and 120 min after oral glucose load. Blood glucose was determined with a glucose meter (Accu Check, Roche, Basel, Switzerland) on blood samples collected from the tip of the tail vein.

Plasma insulin concentration was determined on blood samples using an ELISA kit (Mercodia, Uppsala, Sweden) according to the manufacturer’s instructions. Insulin resistance index was determined by multiplying the area under the curve of both blood glucose (− 30 to 120 min) and plasma insulin (− 30 and 15 min) obtained following the oral glucose tolerance test.

Collection of fecal material

For microbial composition analysis, freshly defecated feces were collected after the acclimation period (day 0), after 3 weeks (day 21), and after 6 weeks (day 42) and kept on dry ice before storage at − 80 °C. In order to determine the fecal energy contents, fecal samples were collected in the 5th week of the experiment during a 24-h period after mice were transferred to clean cages. The samples were dried overnight at 60 °C and weighted to assess the amount of feces secreted per day. Then energy content was measured on a C1 calorimeter from IKA (Germany). Per cage containing two animals, one mean value was considered for analysis.

Tissue sampling

At the end of the experimental period and after 6 h of fasting, mice were anesthetized with isoflurane (Forene, Abbott, Queenborough, Kent, UK). Portal vein blood was collected in a lipopolysaccharide (LPS) free tube, while vena cava blood was collected in EDTA-containing tubes. After centrifugation (12 000× g for 5 min) serum and plasma were aliquoted and immediately immersed in liquid nitrogen before storage at − 80 °C for further analysis. Liver, brown and white adipose tissues (subcutaneous, epididymal, and visceral), muscles (soleus, gastrocnemius, tibialis, and vastus lateralis), and cecal content were precisely dissected, weighed, and immediately snap-frozen in liquid nitrogen and stored at − 80 °C for further analysis.

Histological analysis and immunohistochemistry

A portion of the liver and subcutaneous adipose tissue (SAT) were fixed in 4% paraformaldehyde solution for 24 h at room temperature. Samples were then immersed in ethanol 100% for 24 h before processing for paraffin embedding and preparation of 5-μm tissue sections. Adipocyte size was determined on H&E stained sections and macrophage infiltration was quantified after immunostaining with F4/80 antibody (Ab6640, Abcam, Cambridge, UK). Images were captured at × 20 magnification and obtained using a SNC400 slide scanner and digital Image Hub software 561 (Leica Biosystems, Wetzlar, Germany). Analyses were performed using ImageJ (version 1.48r, National Institutes of Health, Bethesda, Maryland, USA) in a blinded manner. Crown-like structures (CLSs) were counted both in the hepatic and adipose tissue as an indicator of immune cell recruitment and inflammation and were expressed as the number of CLSs per field. A minimum of 5 high-magnification fields were analyzed per mouse.

RNA preparation and real-time qPCR analysis

Total RNA was prepared from collected tissues using TriPure reagent (Roche). Quantification and integrity analysis of total RNA was performed by running 1 μl of each sample on an Agilent 2100 Bioanalyzer (Agilent RNA 6000 Nano Kit, Agilent, Santa Clara, CA, USA). cDNA was prepared by reverse transcription of 1 μg total RNA using a Reverse Transcription System Kit (Promega, Madison, Wisconsin, USA). Real-time PCR was performed with the CFX96 Real-time PCR system and CFX manager 3.1 software (BioRad, Hercules, California, USA) using GoTaq qPCR Master Mix (Promega, Madison, Wisconsin, USA) for detection, according to the manufacturer’s instructions. RPL19 RNA was chosen as the housekeeping gene, and data were analyzed according to the 2 −ΔΔCT method. The identity and purity of the amplified product were assessed by melting curve analysis at the end of amplification. The primer sequences for the targeted mouse genes are presented in the Additional file  1 : Table S1.

Biochemical analyses

Total lipids were measured after extraction with chloroform-methanol according to a modified Folch method [ 23 ] as previously described [ 24 ]. Triglyceride and cholesterol concentrations were measured using a kit coupling an enzymatic reaction and spectrophotometric detection of the final product (Diasys Diagnostic and systems, Holzheim, Germany). All analyses and samples were run in duplicate.

Lipopolysaccharides assay

LPS levels were measured in serum collected from the portal vein of ob/ob , db/db , and their respective lean littermates using a competitive inhibition enzyme immunoassay (Cloud-Clone Corp, Houston, TX). Samples were diluted (1:10) with the Charles River Endosafe dispersing agent (Charleston, South Carolina, USA) to disperse endotoxin molecules during sample preparation, and heated 15 min at 70 °C to inactivate nonspecific inhibitors of endotoxin. Samples displaying hemolysis were excluded from the analysis according to the manufacturer’s instructions. The endotoxin concentration was determined spectrophotometrically at 450 nm and calculated from the standard curve of known amounts of Escherichia coli endotoxin. All determinations were performed in duplicate.

Bile acid and short-chain fatty acid quantification

Bile acids and SCFAs were quantified using an HPLC-MS adapted method, as previously described [ 25 ]. Briefly, for BA analysis, liver tissue was homogenized in ice-cold distilled water and proteins precipitated using acetone (in the presence of 7 deuterated internal standards). Next, samples were centrifuged, supernatants recovered, and evaporated to dryness. Chromatographic separation was achieved using an Ascentis Express C-18 column (100 × 4.6 mm, 2.7 μm) (Sigma-Aldrich) and a gradient of water and acetonitrile in the presence of formic acid. For ionization, an ESI probe operating in negative mode was used.

For SCFAs analysis, the cecal content (50–60 mg wet material) was homogenized in water followed by sonication in an ice water bath. Acetonitrile was used for protein precipitation (in the presence of valproic acid as internal standard). Following centrifugation, the supernatant was recovered and a derivatization step (using 3-nitrophenylhydrazine in the presence of EDC and pyridine) performed. Samples were purified using liquid-liquid extraction to remove the remaining reagents. After evaporation, the final residue was analyzed using an LTQ Orbitrap XL mass spectrometer coupled to an Accela HPLC system (ThermoFisher Scientific). A Hypersil GOLD PFP (100 × 2.1 mm; 1.9 μm) column using a gradient of water-acetonitrile-acetic acid and acetonitrile-acetic acid allowed separating the different isomers. For ionization, an APCI probe was used in positive mode. Calibration curves were prepared using the same conditions to determine sample content. Xcalibur® software was used for data analysis. For each cecal content, an aliquot was freeze-dried to determine a dry residue that was used for data normalization.

For both types of analytes, calibration curves were prepared using the same conditions to determine sample content. Xcalibur® software was used for data analysis.

Microbial load measurement

Microbial load measurement by flow cytometry was determined in the fecal samples of both ob/ob and db/db mice and their littermate counterparts. Briefly, 20 mg frozen (− 80 °C) aliquots were dissolved in physiological solution to a total volume of 100 ml (8.5 g × l −1 NaCl; VWR International). Subsequently, the slurry was diluted 500 times. Samples were filtered using a sterile syringe filter (pore size of 5 μm; Sartorius Stedim Biotech). Next, 1 ml of the microbial cell suspension obtained was stained with 1 μl SYBR Green I (1:100 dilution in dimethylsulfoxide; shaded for 15 min of incubation at 37 °C; 10,000 concentrate, Thermo Fisher Scientific). The flow cytometry analysis was performed using a C6 Accuri flow cytometer (BD Biosciences) based on a previously published study [ 26 ]. Fluorescence events were monitored using the FL1 533/30-nm and FL3 > 670-nm optical detectors. In addition, forward- and sideward-scattered light was collected. The BD Accuri CFlow software was used to gate and separate the microbial fluorescence events on the FL1/FL3 density plot from background events. A threshold value of 2,000 was applied on the FL1 channel. The gated fluorescence events were evaluated on the forward and sideward density plot, as to exclude remaining background events. Instrument and gating settings were kept identical for all samples (fixed staining/gating strategy) [ 26 ]. On the basis of the exact weight of the aliquots analyzed, cell counts were converted to microbial loads per gram of fecal material.

Fecal microbiota sequencing

Fecal DNA extraction and microbiota profiling by 16S rRNA gene sequencing were performed as described previously [ 27 ]. Briefly, DNA was extracted from frozen fecal pellets using the MoBio PowerMicrobiome RNA isolation kit with the addition of 10 min incubation at 90 °C after the initial vortex step. The V4 region of the 16S rRNA gene was amplified with primer pair 515F/806R. Samples were processed for multiplex sequencing with dual-index barcoding. Sequencing was performed on the Illumina MiSeq platform (San Diego, California, USA), to generate paired-end reads of 250 bases in length in each direction. After de-multiplexing using LotuS (version 1.565) [ 28 ], fastq sequencing files were pre-processed using the DADA2 pipeline (R package version 1.6.0) [ 29 ], for trimming, quality control, merging of pairs, and taxonomic annotation using the SILVA (version 132n) database [ 30 ]. With one sample failing sequencing quality control (N < 500 reads after QC), 112 fecal sequencing profiles were obtained.

Deriving quantitative microbiota profiles

The quantitative microbiome profiling (QMP) matrix was built as described previously [ 31 ] by combining sequencing data and microbial load assessment by flow cytometry. A script is available at https://github.com/raeslab/QMP/blob/master/QMP.R . In short, samples were downsized to even sampling depth, defined as the ratio between sampling size (16S rRNA gene copy number corrected sequencing depth) and microbial load (average total cell count per gram of frozen fecal material). 16S rRNA gene copy number corrections were based on the ribosomal RNA operon copy number database rrnDB [ 32 ]. The copy number corrected sequencing depth of each sample was rarefied to the level necessary to equate the minimum observed sampling depth in the cohort (original sampling depth range = [4e−8,7e−7]). The minimum rarefaction level was 609 cnv-corrected reads (approx. 2500 non-corrected reads). The obtained rarefied-to-even-sampling-depth genus-level matrix was then converted into numbers of cells per gram. From an input of 112 samples with 101 genera (observed with minimum 1 read), with a 17-fold difference in original sampling depth, the obtained QMP matrix had a final size of 112 samples and 94 observed genera characterized at a final sampling depth of 4.11e−08 cnv-corrected reads per cell in a gram of sample. Zero values in the microbiota matrix are therefore interpretable as non-detectable genera at the final sampling depth.

Statistical analysis

Metabolic parameter correlation analysis.

Principal component analysis (PCA) of the metabolic parameters measured in the figures (i.e., Figs.  1 , 2 , 3 , 4 ,  5 , and S 2 ) of the present study was performed using the R package “psych” (version 2.0.12) [ 33 ]. Missing data (2%) was imputed using the median metabolic parameter value to be able to compute the component scores. Three principal components were extracted, following results obtained by parallel analysis (scree plot). The PCA was performed without rotation. The loadings matrix of the PCA was investigated manually to identify contrasting signs of the correlations of the variables with the principal components.

Different phenotype features between ob/ob and db/db mice . ( a ) ∆ (Delta) of the body weight (starting at day 0) and final body weight (g). ( b ) ∆ of the fat mass (starting at day 3) and final fat mass (g) measured by time-domain nuclear magnetic resonance (TD-NMR). ( c ) ∆ of the lean mass (starting at day 3) and final lean mass (g) measured by time-domain nuclear magnetic resonance (TD-NMR). ( d ) Adipose tissues (SAT: subcutaneous; EAT: epididymal; VAT: visceral; BAT: brown) weight (g). ( e ) Muscles (SOL: soleus; GAS: gastrocnemius; TA: tibialis; VL: vastus lateralis) weight (g). ( f ) Size of the adipocytes in the subcutaneous adipose tissue (SAT). Scale bar, 100 μm; magnification, × 20. ( g ) Morphology of the liver, SAT, and cecum. ( h ) Plasma glucose (mg/dL) profile after 1 g/kg glucose oral challenge in freely moving mice and ( i ) the mean area under the curve (AUC) measured between 0 and 120 min after glucose loading. ( j ) Plasma insulin (μg/L) measured 30 min before and 15 min after glucose loading. ( k ) Insulin resistance index determined by multiplying the AUC of blood glucose by the AUC of insulin. Green: CT ob lean mice, red: ob/ob mice, blue CT db lean mice, and violet: db/db mice. Data are presented as the mean ± s.e.m, ** P < 0.01, **** P < 0.0001 (n = 8–10). Data were analyzed using two-way ANOVA followed by Tukey’s post hoc test for ( a–c) and ( h) and according to one-way ANOVA followed by Tukey’s post hoc test for ( d – f ) and ( i – k )

Different hepatic features between ob/ob and db/db mice. ( a ) Liver weight at necropsy (g); Total lipid content (mg lipids/mg tissue); Liver triglycerides (nmol/mg tissue); Liver cholesterol (nmol/mg tissue) measured using a spectrophotometer. ( b ) mRNA expression of liver lipid metabolism markers measured by RT-qPCR. ( c ) mRNA expression of liver immune cells markers measured by RT-qPCR. ( d ) Representative pictures of staining for F4/80 in the liver. Scale bar, 100 μm; magnification, × 20. Arrowheads point to crown-like structures. ( e ) mRNA expression of liver receptors and inflammatory cytokines markers measured by RT-qPCR. ( f ) mRNA expression of liver fibrosis markers measured by RT-qPCR. Green: CT ob lean mice, red: ob/ob mice, blue CT db lean mice, and violet: db/db mice. Data are presented as the mean ± s.e.m, * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001 (n = 7–10). For the mRNA expression, relative units were calculated versus the mean of the CT ob mice values set at 1. Data were analyzed by one-way ANOVA followed by Tukey’s post hoc test

Different serum LPS concentration, hepatic bile acid content, and bile acid metabolism between ob/ob and db/db mice. ( a ) Serum LPS concentration (ng/mL) measured by competitive inhibition enzyme immunoassay. ( b ) Liver bile acid content (pmol/5mg tissue) quantified by HPLC-MS. ( c ) mRNA expression of liver bile acid synthesis and conjugation markers measured by RT-qPCR. ( d ) mRNA expression of liver bile acid export markers measured by RT-qPCR. ( e ) mRNA expression of liver bile acid reabsorption markers measured by RT-qPCR. ( f ) mRNA expression of ileal bile acid reabsorption markers measured by RT-qPCR. Dashed black line: CT lean mice, green: CT ob lean mice, red: ob/ob mice, blue: CT db lean mice, and violet: db/db mice. Data are presented as the mean ± s.e.m, * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001(n = 8–10). For the mRNA expression, relative units were calculated versus the mean of the CT ob mice values set at 1. Data were analyzed by one-way ANOVA followed by Tukey’s post hoc test. CA, cholic acid; CDCA, chenodeoxycholic acid; DCA, deoxycholic acid; MCA, muricholic acid; T, taurine; UDCA, ursodeoxycholic Acid. a, alpha; b, beta; o, omega conjugated species

Different subcutaneous adipose tissue features between ob/ob and db/db mice. ( a ) mRNA expression of SAT immune cells markers measured by RT-qPCR. ( b ) Representative pictures of F4/80 staining in SAT. Scale bar, 100 μm; magnification, × 20. Arrowheads point to crown-like structures. ( c ) mRNA expression of SAT receptors and inflammatory cytokines markers measured by RT-qPCR. ( d ) mRNA expression of SAT lipid metabolism and adipogenesis markers measured by RT-qPCR. Green: CT ob lean mice, red: ob/ob mice, blue CT db lean mice, and violet: db/db mice. Data are presented as the mean ± s.e.m, * P < 0.05, ** P < 0.01, *** P < 0.001 (n = 8–10). For the mRNA expression, relative units were calculated versus the mean of the CT ob mice values set at 1. Data were analyzed by one-way ANOVA followed by Tukey’s post hoc test

Different short-chain fatty acids profile between ob/ob and db/db mice. ( a ) Cecum weight (g); Cecal content weight (g); Cecal tissue weight (g). ( b ) Amount of acetic acid, butyric acid, and propionic acid in the cecal content (nmol/mg of dry cecal content) measured by liquid chromatography-mass spectrometry (UPLC-MS). ( c ) Amount of isobutyric acid, 2-methylbutyric acid, valeric acid, isovaleric acid, and hexanoic acid in the cecal content (nmol/mg of dry cecal content) measured by liquid chromatography-mass spectrometry (UPLC-MS). ( d ) Principal component analysis (PCA) score plot of mice based on all measured metabolic parameters. Green: CT ob lean mice, red: ob/ob mice, blue CT db lean mice, and violet: db/db mice. Data are presented as the mean ± s.e.m, * P < 0.05, **** P < 0.0001 (n = 7–10). Data were analyzed by one-way ANOVA followed by Tukey’s post hoc test for ( a–c )

Metabolic and fecal data association to genotype

The data are presented as the means ± s.e.m (standard error of mean). The statistical significance of difference for the metabolic parameters was evaluated by one-way or two-way ANOVA followed by Tukey’s post hoc multiple comparison test, while for the microbial load and the bacterial genera abundances, non-parametric equivalents: Kruskal-Wallis test with Dunn’s multiple comparison test, were used. For the metabolic parameters, only statistically significant differences between ob/ob and db/db mice were reported. The data with a superscript symbol ( # CT ob vs CT db; * ob/ob vs db/db) are significantly different ( #, * P < 0.05; ##, ** P < 0.01; ###, *** P < 0.001; ####, **** P < 0.0001). All the analyses were performed using GraphPad Prism version 8.00 for Windows (GraphPad Software). The presence of outliers was assessed using the Grubbs test.

Partitioning of microbiota variation according to genotype and sampling day

Visualization of fecal microbiota profile variation was performed by principal coordinates analysis (PCoA) using Bray-Curtis dissimilarity between genus-level quantitative microbiota profiles using the R package vegan [ 34 ]. Visualization (arrows) of the direction and degree of association of mouse genotypes on microbiota composition was performed by post hoc fit on the PCoA (R package vegan envfit function). The explanatory power of mouse genotype and day of sampling, on microbial community genus-level QMP variation, was estimated by permutational multivariate analysis of variance (Adonis test, R package vegan adonis2 function).

Taxa-metabolic parameters associations

Correlations between single taxa quantitative abundances (genera) and metabolic parameters were assessed by non-parametric Spearman correlation, excluding taxa with less than 15% prevalence in the dataset.

All tests were subjected to multiple testing corrections (Benjamini-Hochberg method) whenever applicable.

Different phenotypic features between ob/ob and db/db mice

After 7 weeks of follow-up, both ob/ob and db/db mice gradually gained the same body weight while feeding ad libitum on normal diet, thereby confirming the obesogenic effect of impaired leptin-signaling (Fig.  1 a). Body composition analysis using NMR showed a similar increase in fat mass (Fig.  1 b) and a lower lean mass (Fig.  1 c) in both ob/ob and db/db mice. Interestingly, despite having similar total fat mass gain, at the end of the experiment, we found that both ob/ob and db/db mice had a different fat mass distribution of various fat depots. Both epididymal adipose tissue (EAT) and brown adipose tissue (BAT) showed significantly higher weight in ob/ob mice (23.7% and 24.7%, respectively) (Fig.  1 d), whereas subcutaneous adipose tissue (SAT) was 22.9% heavier in db/db mice compared with ob/ob mice (Fig.  1 d). No differences were observed for the visceral adipose tissue (VAT) mass when comparing ob/ob and db/db mice (Fig.  1 d). Among the different types of muscles, the soleus (SOL) mass was the only one to have a significant 20.6% reduction in db/db mice compared with ob/ob mice (Fig.  1 e). The increase in fat mass was associated with larger adipocytes in both mutant mice (Fig.  1 f). During the necropsy, we also found that the morphology of different tissues (i.e., liver, adipose tissues, and cecum) in term of size, shape, and color was similar between the two control lean groups, while it was different between ob/ob and db/db mice (Fig.  1 g). Despite their equal body weight and fat mass gain, db/db mice had an enhanced food and water intake throughout the duration of the experiment (Additional file  2 : Fig. S1a-b). Measurement of body temperature showed a markedly lower temperature (− 1.2 °C) in db/db mice when compared to ob/ob mice, indicating a different energy metabolism (Additional file  2 : Fig. S1c). Conversely, calculating the energy excretion (i.e. amount of feces secreted in 24h multiplied by the fecal energy content measured by bomb calorimeter) revealed that db/db mice had a lower energy uptake compared to ob/ob mice (Additional file  2 : Fig. S1d-f).

Different glucose and insulin profile between ob/ob and db/db mice

The blood glucose profile and the glucose-induced insulin secretion were significantly different between ob/ob and db/db mice. At basal levels and after the oral glucose load, fasted db/db mice exhibited a more pronounced impaired glucose tolerance, which was maintained throughout the duration of the oral glucose tolerance test (OGTT) as indicated by a 64.5% increase in the area under the curve (Fig.  1 h, i), and a 73.9% reduction in plasma insulin levels compared with fasted ob/ob mice (Fig.  1 j). Contrarily to an impaired insulin secretion in db/db mice, ob/ob mice produced significantly more insulin in response to oral glucose administration, suggesting an insulin resistance state (Fig.  1 j). Overall, both models developed insulin resistance to a similar degree as evidenced by the calculation of the insulin resistance index (Fig.  1 k).

Different lipid and inflammatory hepatic profile between ob/ob and db/db mice

We found that ob/ob mice had a significant 25.1% increase in the liver weight (Fig.  2 a) and displayed more severe hepatic steatosis compared to db/db mice. Hepatic lipid accumulation was confirmed by a 59.8% increase in total hepatic lipid contents and was mainly due to strongly increased hepatic levels of triglycerides and cholesterol (33.8% and 57.9%, respectively) (Fig.  2 a). In order to understand the underlying mechanism of the development of hepatic steatosis, we analyzed a large panel of genes involved in lipid metabolism (Fig.  2 b). In ob/ob mice, we observed a significantly higher mRNA expression of a marker linked to fatty acid uptake and storage (i.e., cluster of differentiation 36, encoded by Cd36 ). Consistent with their higher lipid and cholesterol accumulation, ob/ob mice displayed increased lipid synthesis markers (i.e., acetyl-CoA carboxylase alpha, encoded by Acaca ; fatty acid synthase, encoded by Fasn ; 3-hydroxy-3-methylglutaryl-CoA reductase, encoded by Hmgcr ; and peroxisome proliferator-activated receptor gamma, encoded by Pparg ) as compared to db/db mice, strongly suggesting a different hepatic lipid metabolism between the two mutant groups. The mRNA expression of two key genes associated with fatty acid oxidation (i.e., carnitine palmitoyltransferase 1A, encoded by Cpt1a ; and peroxisome proliferator-activated receptor alpha, encoded by Ppara ) was not significantly changed in either ob/ob or db/db mice, suggesting no changes in the fatty oxidation pathway (Fig.  2 b).

To further investigate whether hepatic lipid steatosis was also associated with hepatic inflammation, we measured the mRNA expression of several markers associated with recruitment/infiltration of various types of the immune cell population (i.e., C-C motif chemokine ligand 2, encoded by Ccl2 ; adhesion G-protein-coupled receptor E1, encoded by Adgre1 ; integrin subunit alpha X, encoded by Itgax ; cluster of differentiation 68, encoded by Cd68 ; and cluster of differentiation 163, encoded by Cd163 ). In the ob/ob mice, we observed a significant upregulation of the mRNA expression of Ccl2 (a chemokine that regulates migration and infiltration of monocytes/macrophages), Adgre1 (a marker reflecting the total number of mature macrophages), Itgax (a marker of dendritic cells), and Cd68 (a marker of monocytes/macrophages), while a reduction of the expression of Cd163 (a marker of anti-inflammatory monocyte/macrophages), barely failed to attain statistical significance ( P = 0.060) in ob/ob mice compared to db/db and lean mice (Fig.  2 c). We next confirmed an 84.1% increase of macrophage infiltration in the liver of ob/ob mice compared to db/db mice by performing a F4/80 immunostaining and counting crown-like structures (CLSs, i.e. macrophages surrounding dead or dying hepatocytes with large lipid droplets) on hepatic slices (Fig.  2 d). Consistently with the higher immune cell recruitment, the mRNA expression of key receptors involved in the recognition of pathogen-associated molecules patterns of Gram-negative bacteria (i.e., cluster differentiation 14, encoded by Cd14 ; toll-like receptor 4, encoded by Tlr4 ; toll-like receptor 2, encoded by Tlr2 ; NLR family pyrin domain containing 3, encoded by Nlrp3 ), and of pro-inflammatory cytokines (i.e., tumor necrosis factor alpha, encoded by Tnf ; interleukin 1 beta, encoded by Il1b ) were significantly upregulated in ob/ob mice compared to db/db mice (Fig.  2 e), while no changes in the mRNA expression of toll-like receptor 5 (encoded by Tlr5 ) were observed (Fig.  2 e). These results suggest a severe liver inflammation associated with massive recruitment of immune cells in ob/ob mice. Given that chronic liver inflammation leads to fibrosis [ 35 ], we also investigated the expression of fibrosis-related genes (i.e., collagen type I alpha 1 chain, encoded by Col1a1 ; and transforming growth factor beta, encoded by Tgfb1 ). The expression of both genes was significantly increased in the ob/ob mice compared to the db/db mice (Fig.  2 f). Altogether, these results highlight a different hepatic profile in terms of steatosis, inflammation, and fibrosis between ob/ob and db/db mice.

Different bile acid metabolism and bile acid profile between ob/ob and db/db mice

Hepatic inflammation can be triggered by several stimuli. Gut-derived endotoxin such as lipopolysaccharides (LPS, components of Gram-negative bacteria outer membrane) can reach the liver via the portal circulation and promote the release of large amounts of proinflammatory mediators via its receptor, TLR4 [ 7 ]. Additionally, cholestasis, i.e., a decrease in bile flow due to impaired secretion by hepatocytes or to obstruction of bile flow through the bile ducts, can lead to accumulation of bile acids in the liver and thereby contribute to inflammation [ 36 ]. For this reason, we measured the serum LPS concentration and the BA content in the liver of both ob/ob and db/db mice, and their respective lean littermates. Strikingly, we found a significant 32.5% increase of serum LPS concentration in the db/db mice compared to the ob/ob mice (Fig.  3 a), and consistent with our hypothesis, the amount of cholic acid (CA), a major primary free BA, was 94.5% significantly increased in the liver of ob/ob mice compared to the db/db mice. Conversely, there were no significant variations in the content of taurocholic acid (TCA), taurochenodeoxycholic acid (TCDCA), taurodeoxycholic acid (TDCA), tauroursodeoxycholic acid (TUDCA), tauro-alpha-beta muricholic acid (T(a+b) MCA) and tauro-omega muricholic acid (ToMCA) in the liver of both ob/ob and db/db mice (Fig.  3 b).

Given that the BA profile is regulated by several mechanisms, we measured a large panel of markers associated with BA metabolism (i.e., synthesis, transport, and pool size) [ 37 ]. In the ob/ob mice, we observed a significant downregulation in the mRNA expression of markers involved in the classical (neutral) and the alternative (acidic) bile acid synthesis as well as in the CA production (i.e., cytochrome P450, family 8, subfamily B, member 1, encoded by Cyp8b1 ; and cytochrome P450, family 27, subfamily A, member 1, encoded by Cyp27a1 ), (Fig.  3 c), while the mRNA expression of a rate-limiting enzyme of BA synthesis (i.e., cytochrome P450, family 7, subfamily A, member 1, encoded by Cyp7a1 ) tended to be decreased in ob/ob mice (Fig.  3 c). Following BA synthesis, primary BAs are conjugated to taurine in mice by the enzymes bile acid CoA ligase (BAL) and bile acid CoA:amino acid N-acyltransferase (BAT) in order to increase their solubility for biliary secretion. Both enzymes are under the regulation of the hepatocyte nuclear factor 4 alpha (HNF4α) [ 38 ]. We observed in ob/ob mice a significant downregulation in the mRNA expression of Slc27a5 (coding for BAL), and of Hnf4a (coding for HNF4α), while no changes in Baat (coding for BAT) occurred (Fig.  3 c). These results suggest an impaired BA synthesis and conjugation in the ob/ob mice. We also measured several markers involved in either cholesterol, phospholipids transports, or BA reabsorption. We found that Abcg5/8 (coding for cholesterol transporters ATP binding cassette, subfamily G, member 5, and 8) were significantly downregulated in the ob/ob mice compared to the db/db mice, whereas Abcb4 mRNA (coding for the phospholipid transporter MDR2) was significantly increased in the ob/ob mice. The Abcb11 mRNA (coding for bile salt export pump of hepatocytes BSEP) remained unaffected when comparing ob/ob and db/db (Fig.  3 d), whereas the expression of Slc51b (coding for the transcellular transport of bile acids OSTβ), was significantly increased in ob/ob mice (Fig.  3 d). The majority of the conjugated primary BAs are reabsorbed in the distal ileum and shuttled from the enterocytes into the portal circulation, where they are taken up by the hepatocytes and re-secreted into bile. In order to investigate the enterohepatic circulation, we measured the expression of several transporters implicated in this path. We found that the hepatic expression of Slc10a1 (coding for the sodium (Na + ) taurocholate cotransporting polypeptide NTCP)) and Oatp1b2 (coding for the organic anion transporter OATP1B2) was significantly downregulated in the ob/ob mice compared to the db/db mice (Fig.  3 e), whereas the ileal expression of Slc10a2 (coding for the apical sodium-dependent bile salt transporter ASBT), Fabp6 (coding for the bile acid-binding protein IBABP), and Scl51b were not significantly affected in either ob/ob or db/db mice (Fig.  3 f). Slc51a mRNA (coding for the transcellular transport of bile acids OSTα) was the only marker to be slightly increased ( P = 0.066) in the db/db mice compared to ob/ob mice (Fig.  3 f).

Altogether, these results highlight an impaired BA metabolism associated with a different bile acid content between ob/ob and db/db . We hypothesized that not only LPS but also the hepatic BA accumulation may be the trigger of the changes observed above, thereby impairing the normal BA metabolism as well as the normal enterohepatic circulation of the BA.

Different inflammatory profile in the subcutaneous adipose tissue between ob/ob and db/db mice

Body fat distribution and adipose tissue dysfunction are key factors involved in the development of obesity and its related metabolic disorders [ 39 ]. Because the metabolic, endocrine, and inflammatory profile of adipose tissue is depot dependent [ 40 ], we extensively characterized crucial markers related to the recruitment/infiltration of various types of immune cells, inflammation, and lipid metabolism, in two different and representative adipose tissue depots (SAT and VAT). Intriguingly, and in contrast to that observed in the liver, we found that the mRNA expression of Ccl2 , Adgre1 , and Cd68 , was significantly increased in the SAT of db/db mice compared to the ob/ob mice, while no differences in the mRNA expression of Itgax (upregulated both in ob/ob and db/db mice) and Cd163 were observed (Fig.  4 a). The same tendencies for these markers were observed in the VAT of db/db mice (Additional file  3 : Fig. S2a-b). To further confirm the increased macrophage infiltration into the SAT, immunohistochemical F4/80 staining showed that db/db mice presented a 34.5% increase in the number of CLSs compared to the ob/ob mice (Fig.  4 b). CLSs formed by proinflammatory macrophages are found around large dying adipocytes during a state of obesity and have been associated with inflammation and insulin resistance both in mice and humans [ 41 , 42 , 43 , 44 ]. Along with the increased number of immune cells, the mRNA expression of Il1b and Ifng (coding for interferon gamma), two important proinflammatory cytokines, was significantly increased in the db / db mice compared to the ob/ob mice, while no significant changes in the expression of Tlr4 , Tlr2 (Fig.  4 c) occurred. Interestingly, the mRNA expression of Tlr5 , a key receptor involved in the recognition of pathogens-associated molecular patterns from Gram-positive bacteria (i.e., flagellin) was significantly increased in the ob/ob compared to the db/db mice (Fig.  4 c). However, its increased expression was not associated with inflammation in the SAT of ob/ob mice. Additionally, the expression of Ptgs2 (coding for prostaglandin-endoperoxidase synthase 2), a rate-limiting enzyme for prostaglandin production, which is implicated primarily in the regulation of inflammation in the white adipose tissue, was significantly increased in the db/db mice compared to the ob/ob mice (Fig.  4 c). In the VAT, the expression of Il6 , a major proinflammatory cytokine, was the only marker to be significantly increased in the db/db mice, while no significant differences were observed in the expression of other markers (i.e., Tlr4 , Tlr2 , Tlr5 , Il1b ) between ob/ob and db/db mice (Additional file  3 : Fig. S2a-b). It is well established that proinflammatory cytokines play a crucial role in the regulation of adipogenesis, thereby influencing the formation of new adipocytes [ 45 ]. For that reason, we used quantitative PCR to determine the mRNA expression of key master regulators of the adipogenesis such as Pparg and Cebpa (coding for CCAAT enhancer-binding protein alpha), and fundamental markers involved in lipid synthesis (i.e., Acaca , Fasn ) . We observed that Cebpa was significantly reduced in the db/db mice compared to the ob/ob mice, while the other markers tended to be downregulated to a greater extent in the db/db than in the ob/ob mice (Fig.  4 d). No significant changes were observed for Cpt1a and Ppara mRNA expression between ob/ob and db/db mice, suggesting no changes in the lipid oxidation (Fig.  4 d). These results mainly suggest an impaired adipocyte differentiation in the db/db mice.

Different short-chain fatty acids and gut microbiota profile between ob/ob and db/db mice

Changes in gut bacteria-derived metabolites and gut microbiota composition could also participate in the different effects described above. SCFAs are the most abundant bacterial metabolites present in the gastrointestinal tract, which are involved in the regulation of several metabolic pathways [ 10 ]. In the present study, the amount of SCFAs was analyzed in the cecal content. Despite changes in the morphology of the cecum, there were no significant differences in the cecum weight, cecal content weight, and cecal tissue weight between ob/ob and db/db mice (Fig.  5 a). On the other hand, we found that the amount of acetic acid, butyric acid (Fig.  5 b), isobutyric acid, and hexanoic acid (Fig.  5 c) was significantly decreased in the db/db mice compared to the ob/ob mice (36.4%, 36.9%, 40.7%, and 84%, respectively). No significant differences in the amount of propionic acid (Fig.  5 b), 2-methylbutyric acid, valeric acid, and isovaleric acid between ob/ob and db/db mice were observed (Fig.  5 c). Furthermore, when taking into consideration all the metabolic parameters, the principal component analysis (PCA) showed that the two control groups clustered together, while there is a clear separation between the two mutant groups (Fig.  5 d), strongly emphasizing their metabolic diversity. PCA resulted in three principal components, explaining respectively 38%, 15%, and 7% of the total variance in the data set. The first principal component was correlated with overall weight-related metabolic parameters, explaining the difference between the control groups and experimental groups. For the second principal component (PC2), which explained the difference between the ob/ob and db/db experimental groups, the liver and SAT gene expressions had contrasting loadings. This indicates that the two mutant models can be differentiated based on their metabolic parameter profile and that inflammation of the liver (for ob/ob ) and inflammation of SAT (for db/db ) explains this differentiation. Moreover, cecal content of SCFAs had a positive loading for PC2, explaining its lower abundance in the db/db model.

Given that ob/ob and db/db were fed the same control diet for the full experiment, these results suggest that the different SCFA profiles are not diet-related but could reflect a different gut microbiota profile between ob/ob and db/db . To that end, we first determined the total microbial cell count in fecal samples collected on three different days (day 0, day 21, day 42) using flow cytometry. We found no difference in the feces total microbial density between ob/ob and db/db mice in the three different days as well as for the lean littermate groups (Fig.  6 a). Second, we combined amplicon sequencing (16S rRNA gene) with experimentally measured microbial loads to obtain quantitative microbiota profiles for both ob/ob and db/db mice and their respective littermates using fresh feces collected during the same days as the microbial load. We also investigated microbiota alpha-diversity, and there was no significant difference in richness observed between days (Kruskal-Wallis P = 0.49) or mice groups ( P = 0.12). Microbiota genus-level compositional variation, as visualized in a principal coordinates analysis (PCoA; Bray-Curtis dissimilarity; Fig.  6 b), revealed a distinct clustering between the ob/ob and the db/db groups (permutational analysis of variance Adonis test; R 2 = 0.248, P = 1e−05, N = 53) as well as between the two control groups (Adonis test; R 2 = 0.261, P = 1e−05, N = 59) across sampling days. These four mice groups explained 29.5% of overall fecal microbiota variation, while sampling day added 7.1% explained variance within groups (Adonis test [groups + days]; P = 1e−05, N = 112). When looking at the gut microbiota composition, we observed specific taxa differences between mice groups. Despite a distinct gut microbiota composition between the mice groups already at day 0 (Adonis test; R 2 = 0.354, P = 1e−05, N = 37), we identified several taxa that shift in abundance by day 42 in both ob/ob and db/db mice as well as between the two control groups (Fig.  6 c). We found that the quantity of 19 genera was significantly ( Clostridium _ sensu _ stricto _1, Dubosiella , Escherichia / Shigella , Faecalibaculum , Klebsiella , Muribaculum, and Turicibacter ) (Fig.  6 c and Additional file  4 : Table S2), or tended (i.e., A2, Bacteroides, Lachnospiraceae, Lachnoclostridium, Lactobacillus, Lactococcus, Lachnospiraceae_ FCS020, Marvinbryantia , Ruminoclostridium , Ruminoclostridium 5, Shuttlerworthia , and Tyzzerella ) (Additional file  5 : Fig. S3) to be affected by either the ob/ob or the db/db genotype or by both. Surprisingly, we also observed that the quantity of 11 other genera was significantly different between the two control groups ( Bilophila , Clostridium _ sensu _ stricto _1, Dubosiella , Lachnospiraceae _NK4A136_group, Lachnospiraceae _UCG.006, Olsenella , Rikenellaceae _RC9_gut group, Turicibacter ) (Fig.  6 c and Additional file  4 : Table S2), or tended to be (i.e., Akkermansia muciniphila , Parabacteroides , and Ruminococcaceae _UCG_014) (Additional file  5 : Fig S3). Altogether, these results highlight a different gut microbiota profile and composition not only between the two mutant mice, but also between their respective controls, although displaying the same lean and non-diabetic phenotype. Given the important role in the cross-talk between gut microbes and host, we then sought to correlate the bacterial genera with various metabolic parameters (Additional file  6 : Table S3). In particular, we identified Akkermansia muciniphila and Shuttleworthia as the two genera to be the most negatively ( A. muciniphila ) and positively ( Shuttleworthia ) correlated with body weight, glucose profile, lipid metabolism, bile acid metabolism, and liver and adipose tissue inflammation.

Similar fecal microbial load but different quantitative gut microbiota profiles among the four genotype groups. ( a ) Microbial load (cells/g of feces) at day 0, day 21, and day 42 measured by flow cytometry (n = 8–10). ( b ) Genus-level fecal microbiome community variation, represented by principal coordinates analysis (Bray-Curtis dissimilarity PCoA) (n = 112). Arrows correspond to a post hoc fit of the mouse groups on the PCoA. ( c ) Genera displaying significant quantitative abundance differences between mouse genotypes at day 42 (n = 7–10). Genera with a prevalence across samples lower than 15% were excluded. Data are presented as the mean ± s.e.m, #, * P < 0.05, ## P < 0.01, ###, *** P < 0.001, ####, **** P <0.0001. Green: CT ob lean mice, red: ob/ob mice, blue CT db lean mice, and violet: db/db mice. Data were analyzed by the Kruskal-Wallis test with Dunn’s multiple comparison test for ( a ) and ( c )

Ob/ob and db/db mice are widely used as animal models to investigate the pathogenesis of metabolic diseases such as obesity and T2D. However, although both animal models rely on the disruption of the leptin signaling pathway by targeting the ligand ( ob/ob ) or the receptor ( db/db ), and both models are characterized by hyperphagia, massive obesity, and fat mass gain, they are discrepant for glucose metabolism. So far, the origin of these phenotypical differences is unknown. To this aim, in the present study, we extensively characterized these mice. Although both ob/ob and db/db mice had equivalent evolutions in terms of body weight and fat mass gain, we found they had quite distinctive metabolic features, thereby decoupling the observed metabolic features from the obese phenotype. Besides being diabetic, db/db mice had higher food intake, and therefore a lower feeding efficiency, than ob/ob mice. This is likely explained by several mechanisms, such as the loss of glucose in the urine during polyuria, the higher energy excretion in the feces, and the lower body temperature. In agreement with our study, Giesbertz et al. have previously shown that despite the same body weight, ob/ob and db/db mice had a different metabolite profiling in plasma and tissues [ 18 ]. However, the authors did not further investigate the origins of these differences. In the present study, we discovered that several important features such as the inflammatory tone in different tissues, the gut microbiota composition, bacterial components (i.e., LPS), bacteria-derived metabolites, as well as different bioactive lipids (i.e., bile acids) allowed discriminating the db/db from the ob/ob mice. Therefore, our data further explain the difference between the two phenotypes and have led to the identification of novel markers.

Ob/ob mice develop an altered hepatic lipid metabolism, with a higher hepatic steatosis and inflammatory tone characterized by a marked increase in immune cell infiltration. We have explored several mechanisms that could account for this phenotype.

We and others have previously demonstrated in ob/ob mice that the inflammatory phenotype observed in the adipose tissue as well as liver dysfunction is closely linked to the gut microbiota, since its depletion using antibiotics lowers endotoxemia-induced inflammation and related metabolic disorders [ 46 , 47 ]. A previous study in db/db mice fed with a standard chow diet also showed that the leakage of gut microbiota-derived LPS into the portal blood is a well-established mechanism of metabolic endotoxemia that promotes liver damage [ 16 ]. These findings were in contrast with our study, in which db/db mice were protected from liver damage. Differences in experimental procedures (i.e., different diet composition, ages, duration of the experiment) may explain the discrepancies between the studies. However, bacterial components such as LPS are not the only cause of liver damage. Other factors, such as the BA, are also involved in the regulation of innate immunity and liver function [ 48 ], and cholestasis, which is an impaired bile flow leading to accumulation of bile acids in the liver, can also promote liver inflammation. In our study, we observed that the hepatic level of cholesterol, the precursor for BA synthesis, was significantly increased in ob/ob mice. Strikingly, cholic acid (CA) levels were 94.5% higher in ob/ob than in db/db mice, whereas the other BA were comparable between both genotypes. As a matter of fact, the expression of main enzymes involved in the classical pathway of the BA synthesis ( Cyp7a1 , Cyp8b1 , Cyp27a1 ) was downregulated in ob/ob mice and all other markers were pointing towards a lower BA conjugation, higher BA excretion, and lower BA reabsorption. The downregulation of those markers could be interpreted as a protecting mechanism of the liver from the toxic effect of bile acid accumulation. Additionally, we observed that the hepatic Slc51b expression, a basolateral organic solute transporter that mediates bile acid efflux, was significantly increased in ob/ob mice. Given the significant role exerted by the enterohepatic circulation in the regulation of the BA synthesis [ 49 ], we found that the expression of transporters in the ileum regulating the reabsorption of bile acids ( Slc10a2 , Fabp6 , Slc51a , Slc51b ) was unchanged in both mutant groups. Altogether, these data are in accordance with human and animal studies showing that during cholestasis, an alteration of the bile acid transporters occurs and is characterized by a downregulation of the uptake systems (NTCP, and OTAPs) and upregulation of basolateral bile acid export systems (OSTβ) (reviewed in [ 50 ]). Bile acid signaling in the liver and in the intestine is now considered a potential target for the treatment of obesity and non-alcoholic fatty liver disease (NAFLD) [ 51 ]. The role of bile acid in inducing liver injury is mainly evidenced by the use of bile acid sequestrants, whose use reversed liver injury and prevented the progression of steatosis, inflammation, and fibrosis in mice fed a Western diet-induced non-alcoholic steatohepatitis (NASH) mouse model [ 52 ]. Furthermore, given the bidirectional link between bile acids and gut microbiota composition, we cannot exclude that a disruption of the bacterial gut community may affect bile acid synthesis in the liver. A previous study in mice has shown that the gut microbiota not only regulates secondary bile acid metabolism but also inhibits bile acid synthesis in the liver by alleviating farnesoid X receptor (FXR) inhibition in the ileum [ 53 ]. Hence, we may not exclude the role of the gut microbiota as an explanation of our results as further discussed below.

Unlike the relatively low inflammation observed in the liver of db/db mice compared to ob/ob mice, we found that db/db mice had a higher inflammatory tone in the adipose tissue than ob/ob mice. Several potential mechanisms have emerged as the main trigger in the onset of adipose tissue inflammation, including gut-derived substances, dietary component, metabolites, and adipocyte death (reviewed in [ 54 ]). Despite no change in the expression of the TLRs (i.e., TLR4, and TLR2), we may speculate that the downregulation in the expression of fundamental markers associated with adipocyte differentiation ( Pparg , Cebpa ) , may explain adipocyte death, recruitment of immune cells, and production of proinflammatory cytokines, thereby triggering adipose tissue inflammation and insulin resistance in db/db mice. We have previously shown in vivo and in vitro that LPS acts as a master switch to control adipose tissue metabolism and its plasticity during obesity [ 55 ]. However, SCFAs, whose concentrations were reduced in the cecal content of db/db mice, could also be involved. Several studies in vitro and in vivo have shown their effects on immunity, inflammation, and adipose tissue expansion [ 56 , 57 , 58 ]. Here, we found that the concentration of SCFAs in the cecal content was not significantly increased in ob/ob . This observation is not in line with a previous study in ob/ob mice having shown that changes in gut microbiota composition were associated with an increased concentration of SCFAs (i.e., butyrate, and acetate) in the cecal content and less energy content in the stool of the mutant mice [ 59 ]. Contrary to these findings, we found a higher energy excretion in the feces of both ob/ob and db/db mice compared to their respective control groups. Therefore, in our context, it is unlikely that the SCFAs account for the differences in obese phenotypes. Intriguingly, we observed a significant increase in the amount of hexanoic acid in the cecal content of the ob/ob mice compared to the db/db mice. So far, there are no studies describing its role in the onset of obesity development as well as in the regulation of liver and adipose tissue function and metabolism, and further studies are needed to confirm its function. Certain SCFAs, such as acetate, have been shown to modulate appetite in mice [ 60 ]. This could explain the higher food intake observed in db/db mice. Given the important role of the gut microbiota in all the metabolic functions mentioned above, we decided to study the overall microbial community in depth using a recently developed method combining amplicon sequencing and flow cytometry: quantitative microbiome profiling (QMP). Microbial load, defined as the total number of bacteria in a given quantity of sample, was proposed as a main driver of microbiota alteration as shown in a cohort of patients with inflammatory bowel disease [ 31 ]. Here, we did not observe significant differences in the microbial load between ob/ob and db/db mice over the three different time points, thereby excluding this factor as a major driver of the phenotype. By doing QMP, we demonstrate that some genera are more present in the ob/ob mice compared to the db/db mice, and vice versa, and we discovered new genera that may be implicated in the onset of these pathological conditions.

In the present study, we identified that the quantity of Clostriudium_sensu_stricto _1, Dubosiella , Faecalibaculum , Turicibacter (Gram-positive bacteria of the phylum Firmicutes), and Muribaculum (Gram-negative bacteria of the phylum Bacteroidetes) was significantly higher in ob/ob mice when compared to the db/db mice. A recent human study has shown that Clostridium_sensu_stricto _1 is positively correlated with indicators of body weight and serum lipids [ 61 ], while Faecalibaculum and Muribaculum are two recently identified bacteria that have been isolated from the feces and the intestine of murine models respectively [ 62 , 63 ]. So far, there are no studies describing the relationship between Faecalibaculum in the context of obesity and related metabolic disorders, while there is one recent study showing a higher proportion of OTUs most closely related to Muribaculum species in BA fed mice [ 64 ], and another recent one showing a lower proportion of Muribaculum intestinalis in mice fed with high-fat diet, high-glucose diet, and high-fructose diet [ 65 ]. Consistent with our observations, data from other studies observed a higher abundance of Lactobacillus in ob/ob mice [ 66 ]. The increase in Lactobacillus was unexpected as this genus is usually considered a “beneficial bacterium.” However, several studies have already linked Lactobacillus spp . with obesity [ 67 , 68 , 69 ]. It cannot be excluded that differences in the abundance of this bacterial taxa may also reflect the distinct food intake and energy excreted in the feces observed between ob/ob and db/db mice. Moreover, we found positive correlations between Lactobacillus and the hepatic lipid content, bile acid metabolism, and inflammation markers, thereby suggesting that the role of Lactobacillus spp. needs further investigation in studies designed specifically for this purpose. Conversely to the ob/ob , in the db/db mice, we identified a higher quantity of certain Gram-negative bacteria such as Bacteroides (member of the phylum Bacteroidetes ) , Escherichia/Shigella , Klebsiella (member of the phylum Proteobacteria), Lachnospiraceae (member of the phylum Firmicutes), and Gram-positive bacteria such as Lactococcus . A recent study in obese individuals with and without T2D showed that the participants with T2D, compared with participants in the obese non-diabetic group, displayed different microbial signatures with higher Proteobacteria members (that is, Escherichia and Shigella ) in the plasma and mesenteric adipose tissue. This observation also corroborates data showing higher abundance of Escherichia and Shigella in the feces of dysglycemic individuals compared with normoglycemic individuals [ 70 ]. Other recent human studies highlighted the presence of bacteria and bacterial DNA, mainly from Proteobacteria and Firmicutes, in several adipose tissues in obesity and T2D, thereby suggesting a critical role of bacteria in promoting and sustaining local adipose tissue subclinical inflammation and therefore affecting the different metabolic disorders linked to obesity [ 70 , 71 ]. Klebsiella , another member of the Proteobacteria phylum, was also found to be enriched in obese children [ 72 ], and members of the Lachnospiraceae family have also been associated with T2D [ 73 ]. Along with our previous studies, we observed a lower quantity of Akkermansia muciniphila in ob/ob and even lower in db/db mice. This observation has also been confirmed in humans [ 74 , 75 ]. Our group was the first to describe the ability of this bacterium to delay development of diet-induced obesity and insulin resistance in mice, namely via the modulation of the energy homeostasis and restoration of the gut barrier function [ 75 ]. More recently, in humans, we confirmed in a placebo-controlled study in overweight/obese insulin-resistant volunteers that supplementation with A. muciniphila could prevent the worsening of several metabolic parameters [ 76 ]. In addition to the different gut microbiota profiles between ob/ob and db/db , we also identified genera that differed between the two control groups. For example, a higher quantity of A. muciniphila in CT db mice, and higher quantity of Dubosiella and Olsenella in CT ob mice, among others.

Dubosiella has been recently isolated from the murine intestine and associated with protection from adiposity in mice [ 77 ]. Studies in both mice and humans have also described the association between increased physical activity and microbiome changes as well as SCFAs production [ 78 ], thus we may not rule out that the distinct microbiota profile between ob/ob and db/db mice and their lean counterparts may reflect a different locomotor activity that occurred over the duration of the experiment.

As shown in Fig.  6 b and Fig.  5 d despite a different microbiota composition, the two control groups clustered together when taking into consideration all the metabolic parameters, suggesting that the increase in certain beneficial bacteria plays an important role in the modulation of the metabolic function. Taking this together, we propose that the divergent shifts in gut microbial community contribute to the development of the two complex phenotypes, although further studies are needed to determine whether the associated microbial taxa have a causal effect on body weight, glucose profile, and inflammation. However, the reason for changes in the gut microbiota still remains unclear, despite unchanged genetic background and diet. Furthermore, the difference in the microbiota composition and bile acid profile are likely contributing to the different hepatic phenotypes observed between mice. We may not rule out that divergences in food intake and immune system activation could also have contributed to shape the gut microbiota composition. We also acknowledge that having used only male mice is a limitation of the present study. Indeed, the use of mice of both sexes would have provided additional metabolic information and further elucidate gender-related dissimilarities in the overall gut microbiota composition of genetically obese and diabetic mice.

Our results support that the unique metabolic features differentiating ob/ob and db/db mice are explained in part by severe differences in their gut microbiota compositions, gut bacterial components like the LPS, and gut-derived metabolites such as SCFAs, as well as in their bile acid profiles (Fig.  7 ). We also described a different inflammatory tone at two different biological sites, with the liver being more affected in ob/ob mice and the adipose tissue in db/db mice, thereby emphasizing that the development of obesity and diabetes is more organ-dysfunction (i.e., liver and adipose tissue) related. These findings further underscore the differences between the two mutant strains and emphasize that these are not interchangeable experimental models (Fig.  7 ). By discovering their specificities, connecting important biological markers, and identifying new bacteria, we open innovative opportunities for functional studies in the context of obesity and related metabolic disorders such as diabetes, liver injury, and adipose tissue inflammation.

Graphical abstract. This figure summarizes the major differences observed between the two different models. Each specificity related to the organ of body fluid are depicted by a pictogram of the organ

Availability of data and materials

All data generated or analyzed during this study are included in this published article and its supplementary information files. The raw amplicon sequencing data analyzed in this study have been deposited in the European Nucleotide Archive (ENA) at EMBL-EBI under accession number PRJEB44809 ( https://www.ebi.ac.uk/ena/browser/view/PRJEB44809 ). The processed quantitative microbiota matrix is provided as Additional file  7 : Table S4.

Abbreviations

Type 2 diabetes

• Short-chain fatty acids

Quantitative microbial profiling

Subcutaneous adipose tissue

Visceral adipose tissue

Crown-like structures

Oral glucose tolerance test

Toll-like receptors

• Lipopolysaccharides

Cholic Acid

Collaboration, N.C.D.R.F. Trends in adult body-mass index in 200 countries from 1975 to 2014: a pooled analysis of 1698 population-based measurement studies with 19.2 million participants. Lancet. 2016;387(10026):1377–96.

Article   Google Scholar  

Grieve E, Fenwick E, Yang HC, Lean M. The disproportionate economic burden associated with severe and complicated obesity: a systematic review. Obes Rev. 2013;14(11):883–94. https://doi.org/10.1111/obr.12059 .

Article   CAS   PubMed   Google Scholar  

O'Neill S, O'Driscoll L. Metabolic syndrome: a closer look at the growing epidemic and its associated pathologies. Obes Rev. 2015;16(1):1–12. https://doi.org/10.1111/obr.12229 .

Bäckhed F, et al. The gut microbiota as an environmental factor that regulates fat storage. Proc Natl Acad Sci USA. 2004;101(44):15718–23. https://doi.org/10.1073/pnas.0407076101 .

Article   CAS   PubMed   PubMed Central   Google Scholar  

Backhed F, et al. Mechanisms underlying the resistance to diet-induced obesity in germ-free mice. Proc Natl Acad Sci U S A. 2007;104(3):979–84. https://doi.org/10.1073/pnas.0605374104 .

Bluher M. Obesity: global epidemiology and pathogenesis. Nat Rev Endocrinol. 2019;15(5):288–98. https://doi.org/10.1038/s41574-019-0176-8 .

Article   PubMed   Google Scholar  

Cani PD, Amar J, Iglesias MA, Poggi M, Knauf C, Bastelica D, et al. Metabolic endotoxemia initiates obesity and insulin resistance. Diabetes. 2007;56(7):1761–72. https://doi.org/10.2337/db06-1491 .

Cani PD, Osto M, Geurts L, Everard A. Involvement of gut microbiota in the development of low-grade inflammation and type 2 diabetes associated with obesity. Gut Microbes. 2012;3(4):279–88. https://doi.org/10.4161/gmic.19625 .

Article   PubMed   PubMed Central   Google Scholar  

Cani PD, van Hul M, Lefort C, Depommier C, Rastelli M, Everard A. Microbial regulation of organismal energy homeostasis. Nat Metab. 2019;1(1):34–46. https://doi.org/10.1038/s42255-018-0017-4 .

Canfora EE, Jocken JW, Blaak EE. Short-chain fatty acids in control of body weight and insulin sensitivity. Nat Rev Endocrinol. 2015;11(10):577–91. https://doi.org/10.1038/nrendo.2015.128 .

Wahlstrom A, et al. Intestinal Crosstalk between Bile Acids and Microbiota and Its Impact on Host Metabolism. Cell Metab. 2016;24(1):41–50. https://doi.org/10.1016/j.cmet.2016.05.005 .

Suriano F, Van Hul M, Cani PD. Gut microbiota and regulation of myokine-adipokine function. Curr Opin Pharmacol. 2020;52:9–17. https://doi.org/10.1016/j.coph.2020.03.006 .

Friedman JM, Halaas JL. Leptin and the regulation of body weight in mammals. Nature. 1998;395(6704):763–70. https://doi.org/10.1038/27376 .

Wang B, Chandrasekera PC, Pippin JJ. Leptin- and leptin receptor-deficient rodent models: relevance for human type 2 diabetes. Curr Diabetes Rev. 2014;10(2):131–45. https://doi.org/10.2174/1573399810666140508121012 .

Wauman J, Zabeau L, Tavernier J. The Leptin Receptor Complex: Heavier Than Expected? Front Endocrinol (Lausanne). 2017;8:30.

Brun P, Castagliuolo I, Leo VD, Buda A, Pinzani M, Palù G, et al. Increased intestinal permeability in obese mice: new evidence in the pathogenesis of nonalcoholic steatohepatitis. Am J Physiol Gastrointest Liver Physiol. 2007;292(2):G518–25. https://doi.org/10.1152/ajpgi.00024.2006 .

Everard A, Lazarevic V, Derrien M, Girard M, Muccioli GG, Neyrinck AM, et al. Responses of gut microbiota and glucose and lipid metabolism to prebiotics in genetic obese and diet-induced leptin-resistant mice. Diabetes. 2011;60(11):2775–86. https://doi.org/10.2337/db11-0227 .

Giesbertz P, Padberg I, Rein D, Ecker J, Höfle AS, Spanier B, et al. Metabolite profiling in plasma and tissues of ob/ob and db/db mice identifies novel markers of obesity and type 2 diabetes. Diabetologia. 2015;58(9):2133–43. https://doi.org/10.1007/s00125-015-3656-y .

Geurts L, et al. Altered gut microbiota and endocannabinoid system tone in obese and diabetic leptin-resistant mice: impact on apelin regulation in adipose tissue. Front Microbiol. 2011;2:149.

Coleman DL. Obese and diabetes: two mutant genes causing diabetes-obesity syndromes in mice. Diabetologia. 1978;14(3):141–8. https://doi.org/10.1007/BF00429772 .

Coleman DL, Hummel KP. The influence of genetic background on the expression of the obese (Ob) gene in the mouse. Diabetologia. 1973;9(4):287–93. https://doi.org/10.1007/BF01221856 .

Yang M, et al. Gut Microbiota Composition and Structure of the Ob/Ob and Db/Db Mice. Int J Endocrinol. 2019;2019:1394097.

Folch J, Lees M, Sloane Stanley GH. A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem. 1957;226(1):497–509. https://doi.org/10.1016/S0021-9258(18)64849-5 .

Everard A, Plovier H, Rastelli M, van Hul M, de Wouters d’Oplinter A, Geurts L, et al. Intestinal epithelial N-acylphosphatidylethanolamine phospholipase D links dietary fat to metabolic adaptations in obesity and steatosis. Nat Commun. 2019;10(1):457. https://doi.org/10.1038/s41467-018-08051-7 .

Lefort C, et al. Hepatic NAPE-PLD Is a Key Regulator of Liver Lipid Metabolism. Cells. 2020;9(5):1247.  https://doi.org/10.3390/cells9051247 .

Prest EI, Hammes F, Kötzsch S, van Loosdrecht MCM, Vrouwenvelder JS. Monitoring microbiological changes in drinking water systems using a fast and reproducible flow cytometric method. Water Res. 2013;47(19):7131–42. https://doi.org/10.1016/j.watres.2013.07.051 .

Vieira-Silva S, Sabino J, Valles-Colomer M, Falony G, Kathagen G, Caenepeel C, et al. Quantitative microbiome profiling disentangles inflammation- and bile duct obstruction-associated microbiota alterations across PSC/IBD diagnoses. Nat Microbiol. 2019;4(11):1826–31. https://doi.org/10.1038/s41564-019-0483-9 .

Hildebrand F, Tadeo R, Voigt A, Bork P, Raes J. LotuS: an efficient and user-friendly OTU processing pipeline. Microbiome. 2014;2(1):30. https://doi.org/10.1186/2049-2618-2-30 .

Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13(7):581–3. https://doi.org/10.1038/nmeth.3869 .

Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 2013;41(Database issue):D590–6. https://doi.org/10.1093/nar/gks1219 .

Vandeputte D, Kathagen G, D’hoe K, Vieira-Silva S, Valles-Colomer M, Sabino J, et al. Quantitative microbiome profiling links gut community variation to microbial load. Nature. 2017;551(7681):507–11. https://doi.org/10.1038/nature24460 .

Stoddard SF, Smith BJ, Hein R, Roller BRK, Schmidt TM. rrnDB: improved tools for interpreting rRNA gene abundance in bacteria and archaea and a new foundation for future development. Nucleic Acids Res. 2015;43(Database issue):D593–8. https://doi.org/10.1093/nar/gku1201 .

Revelle W. Psych: Procedures for Psychological, Psychometric, and Personality Research. R package version 2.0.12; 2020.

Google Scholar  

Oksanen, J., et al., Vegan: Community Ecology Package. R package version 2.5-7. 2020.

Koyama Y, Brenner DA. Liver inflammation and fibrosis. J Clin Invest. 2017;127(1):55–64. https://doi.org/10.1172/JCI88881 .

Fickert P, Wagner M. Biliary bile acids in hepatobiliary injury - What is the link? J Hepatol. 2017;67(3):619–31. https://doi.org/10.1016/j.jhep.2017.04.026 .

Sarenac TM, Mikov M. Bile Acid Synthesis: From Nature to the Chemical Modification and Synthesis and Their Applications as Drugs and Nutrients. Front Pharmacol. 2018;9:939. https://doi.org/10.3389/fphar.2018.00939 .

Inoue Y, Yu AM, Inoue J, Gonzalez FJ. Hepatocyte nuclear factor 4alpha is a central regulator of bile acid conjugation. J Biol Chem. 2004;279(4):2480–9. https://doi.org/10.1074/jbc.M311015200 .

Longo M, et al. Adipose Tissue Dysfunction as Determinant of Obesity-Associated Metabolic Complications. Int J Mol Sci. 2019;20(9):2358. https://doi.org/10.3390/ijms20092358 .

Chait A, den Hartigh LJ. Adipose Tissue Distribution, Inflammation and Its Metabolic Consequences, Including Diabetes and Cardiovascular Disease. Front Cardiovasc Med. 2020;7:22.

Lumeng CN, DelProposto JB, Westcott DJ, Saltiel AR. Phenotypic switching of adipose tissue macrophages with obesity is generated by spatiotemporal differences in macrophage subtypes. Diabetes. 2008;57(12):3239–46. https://doi.org/10.2337/db08-0872 .

Strissel KJ, Stancheva Z, Miyoshi H, Perfield JW, DeFuria J, Jick Z, et al. Adipocyte death, adipose tissue remodeling, and obesity complications. Diabetes. 2007;56(12):2910–8. https://doi.org/10.2337/db07-0767 .

Cinti S, Mitchell G, Barbatelli G, Murano I, Ceresi E, Faloia E, et al. Adipocyte death defines macrophage localization and function in adipose tissue of obese mice and humans. J Lipid Res. 2005;46(11):2347–55. https://doi.org/10.1194/jlr.M500294-JLR200 .

Apovian CM, Bigornia S, Mott M, Meyers MR, Ulloor J, Gagua M, et al. Adipose macrophage infiltration is associated with insulin resistance and vascular endothelial dysfunction in obese subjects. Arterioscler Thromb Vasc Biol. 2008;28(9):1654–9. https://doi.org/10.1161/ATVBAHA.108.170316 .

Jiang N, Li Y, Shu T, Wang J. Cytokines and inflammation in adipogenesis: an updated review. Front Med. 2019;13(3):314–29. https://doi.org/10.1007/s11684-018-0625-0 .

Cani PD, Bibiloni R, Knauf C, Waget A, Neyrinck AM, Delzenne NM, et al. Changes in gut microbiota control metabolic endotoxemia-induced inflammation in high-fat diet-induced obesity and diabetes in mice. Diabetes. 2008;57(6):1470–81. https://doi.org/10.2337/db07-1403 .

Membrez M, Blancher F, Jaquet M, Bibiloni R, Cani PD, Burcelin RG, et al. Gut microbiota modulation with norfloxacin and ampicillin enhances glucose tolerance in mice. FASEB J. 2008;22(7):2416–26. https://doi.org/10.1096/fj.07-102723 .

Fiorucci S, Biagioli M, Zampella A, Distrutti E. Bile Acids Activated Receptors Regulate Innate Immunity. Front Immunol. 2018;9:1853. https://doi.org/10.3389/fimmu.2018.01853 .

Chiang JYL. Bile acid metabolism and signaling in liver disease and therapy. Liver Res. 2017;1(1):3–9. https://doi.org/10.1016/j.livres.2017.05.001 .

Halilbasic E, Claudel T, Trauner M. Bile acid transporters and regulatory nuclear receptors in the liver and beyond. J Hepatol. 2013;58(1):155–68. https://doi.org/10.1016/j.jhep.2012.08.002 .

Arab JP, Karpen SJ, Dawson PA, Arrese M, Trauner M. Bile acids and nonalcoholic fatty liver disease: Molecular insights and therapeutic perspectives. Hepatology. 2017;65(1):350–62. https://doi.org/10.1002/hep.28709 .

Takahashi S, Luo Y, Ranjit S, Xie C, Libby AE, Orlicky DJ, et al. Bile acid sequestration reverses liver injury and prevents progression of nonalcoholic steatohepatitis in Western diet-fed mice. J Biol Chem. 2020;295(14):4733–47. https://doi.org/10.1074/jbc.RA119.011913 .

Sayin SI, Wahlström A, Felin J, Jäntti S, Marschall HU, Bamberg K, et al. Gut microbiota regulates bile acid metabolism by reducing the levels of tauro-beta-muricholic acid, a naturally occurring FXR antagonist. Cell Metab. 2013;17(2):225–35. https://doi.org/10.1016/j.cmet.2013.01.003 .

Reilly SM, Saltiel AR. Adapting to obesity with adipose tissue inflammation. Nat Rev Endocrinol. 2017;13(11):633–43. https://doi.org/10.1038/nrendo.2017.90 .

Muccioli GG, Naslain D, Bäckhed F, Reigstad CS, Lambert DM, Delzenne NM, et al. The endocannabinoid system links gut microbiota to adipogenesis. Mol Syst Biol. 2010;6(1):392. https://doi.org/10.1038/msb.2010.46 .

Hernandez MAG, et al. The Short-Chain Fatty Acid Acetate in Body Weight Control and Insulin Sensitivity. Nutrients. 2019;11(8):1943. https://doi.org/10.3390/nu11081943 .

Wang X, He G, Peng Y, Zhong W, Wang Y, Zhang B. Sodium butyrate alleviates adipocyte inflammation by inhibiting NLRP3 pathway. Sci Rep. 2015;5(1):12676. https://doi.org/10.1038/srep12676 .

Al-Lahham S, Rezaee F. Propionic acid counteracts the inflammation of human subcutaneous adipose tissue: a new avenue for drug development. Daru. 2019;27(2):645–52. https://doi.org/10.1007/s40199-019-00294-z .

Turnbaugh PJ, Ley RE, Mahowald MA, Magrini V, Mardis ER, Gordon JI. An obesity-associated gut microbiome with increased capacity for energy harvest. Nature. 2006;444(7122):1027–31. https://doi.org/10.1038/nature05414 .

Frost G, Sleeth ML, Sahuri-Arisoylu M, Lizarbe B, Cerdan S, Brody L, et al. The short-chain fatty acid acetate reduces appetite via a central homeostatic mechanism. Nat Commun. 2014;5(1):3611. https://doi.org/10.1038/ncomms4611 .

Zeng Q, Li D, He Y, Li Y, Yang Z, Zhao X, et al. Discrepant gut microbiota markers for the classification of obesity-related metabolic abnormalities. Sci Rep. 2019;9(1):13424. https://doi.org/10.1038/s41598-019-49462-w .

Lim S, Chang DH, Ahn S, Kim BC. Whole genome sequencing of “Faecalibaculum rodentium” ALO17, isolated from C57BL/6J laboratory mouse feces. Gut Pathog. 2016;8(1):3. https://doi.org/10.1186/s13099-016-0087-3 .

Lagkouvardos I, Pukall R, Abt B, Foesel BU, Meier-Kolthoff JP, Kumar N, et al. The Mouse Intestinal Bacterial Collection (miBC) provides host-specific insight into cultured diversity and functional potential of the gut microbiota. Nat Microbiol. 2016;1(10):16131. https://doi.org/10.1038/nmicrobiol.2016.131 .

Just S, Mondot S, Ecker J, Wegner K, Rath E, Gau L, et al. The gut microbiota drives the impact of bile acids and fat source in diet on mouse metabolism. Microbiome. 2018;6(1):134. https://doi.org/10.1186/s40168-018-0510-8 .

Do MH, et al. High-Glucose or -Fructose Diet Cause Changes of the Gut Microbiota and Metabolic Disorders in Mice without Body Weight Change. Nutrients. 2018;10(6):761. https://doi.org/10.3390/nu10060761 .

Kashani A, Brejnrod AD, Jin C, Kern T, Madsen AN, Holm LA, et al. Impaired glucose metabolism and altered gut microbiome despite calorie restriction of ob/ob mice. Animal Microbiome. 2019;1(1):11. https://doi.org/10.1186/s42523-019-0007-1 .

Armougom F, Henry M, Vialettes B, Raccah D, Raoult D. Monitoring bacterial community of human gut microbiota reveals an increase in Lactobacillus in obese patients and Methanogens in anorexic patients. Plos One. 2009;4(9):e7125. https://doi.org/10.1371/journal.pone.0007125 .

Stsepetova J, et al. Diversity and metabolic impact of intestinal Lactobacillus species in healthy adults and the elderly. Br J Nutr. 2011;105(8):1235–44. https://doi.org/10.1017/S0007114510004770 .

Ignacio A, et al. Correlation between body mass index and faecal microbiota from children. Clin Microbiol Infect. 2016;22(3):258.e1–8.

Article   CAS   Google Scholar  

Anhe FF, et al. Type 2 diabetes influences bacterial tissue compartmentalisation in human obesity. Nat Metab. 2020;2(3):233–42. https://doi.org/10.1038/s42255-020-0178-9 .

Massier L, Chakaroun R, Tabei S, Crane A, Didt KD, Fallmann J, et al. Adipose tissue derived bacteria are associated with inflammation in obesity and type 2 diabetes. Gut. 2020;69(10):1796–806. https://doi.org/10.1136/gutjnl-2019-320118 .

Hou YP, et al. Human Gut Microbiota Associated with Obesity in Chinese Children and Adolescents. Biomed Res Int. 2017;2017:7585989.

Qin J, Li Y, Cai Z, Li S, Zhu J, Zhang F, et al. A metagenome-wide association study of gut microbiota in type 2 diabetes. Nature. 2012;490(7418):55–60. https://doi.org/10.1038/nature11450 .

Dao MC, Everard A, Aron-Wisnewsky J, Sokolovska N, Prifti E, Verger EO, et al. Akkermansia muciniphila and improved metabolic health during a dietary intervention in obesity: relationship with gut microbiome richness and ecology. Gut. 2016;65(3):426–36. https://doi.org/10.1136/gutjnl-2014-308778 .

Cani PD, de Vos WM. Next-Generation Beneficial Microbes: The Case of Akkermansia muciniphila. Front Microbiol. 2017;8:1765. https://doi.org/10.3389/fmicb.2017.01765 .

Depommier C, Everard A, Druart C, Plovier H, van Hul M, Vieira-Silva S, et al. Supplementation with Akkermansia muciniphila in overweight and obese human volunteers: a proof-of-concept exploratory study. Nat Med. 2019;25(7):1096–103. https://doi.org/10.1038/s41591-019-0495-2 .

Cox LM, Sohn J, Tyrrell KL, Citron DM, Lawson PA, Patel NB, et al. Description of two novel members of the family Erysipelotrichaceae: Ileibacteriumvalens gen. nov., sp. nov. and Dubosiella newyorkensis, gen. nov., sp. nov., from the murine intestine, and emendation to the description of Faecalibacterium rodentium. Int J Syst Evol Microbiol. 2017;67(5):1247–54. https://doi.org/10.1099/ijsem.0.001793 .

Mailing LJ, Allen JM, Buford TW, Fields CJ, Woods JA. Exercise and the Gut Microbiome: A Review of the Evidence, Potential Mechanisms, and Implications for Human Health. Exerc Sport Sci Rev. 2019;47(2):75–85. https://doi.org/10.1249/JES.0000000000000183 .

Download references

Acknowledgements

We thank, A. Barrois, A. Puel, S. Genten, H. Danthinne, B. Es Saadi, L. Gesche, R. M. Goebbels (at UCLouvain, Université catholique de Louvain) for their excellent technical support and assistance. We thank C. Bouzin from the IREC imagery platform (2IP) from the Institut de Recherche Expérimentale et Clinique (IREC) for their excellent help.

PDC is a senior research associate at FRS-FNRS (Fonds de la Recherche Scientifique), Belgium. He is supported by the Fonds de la Recherche Scientifique (FNRS, FRFS-WELBIO: WELBIO-CR-2019C-02R, and EOS programme no.30770923).

Author information

Authors and affiliations.

Metabolism and Nutrition Research group, Louvain Drug Research Institute (LDRI), Walloon Excellence in Life Sciences and BIOtechnology (WELBIO), UCLouvain, Université catholique de Louvain, Av. E. Mounier, 73 B1.73.11, 1200, Brussels, Belgium

Francesco Suriano, Rudy Pelicaen, Marion Régnier, Nathalie M. Delzenne, Matthias Van Hul & Patrice D. Cani

Department of Microbiology and Immunology, Rega Institute for Medical Research, VIB Center for Microbiology, KU Leuven, University of Leuven, Leuven, Belgium

Sara Vieira-Silva, Gwen Falony & Jeroen Raes

Bioanalysis and Pharmacology of Bioactive Lipids Research Group, Louvain Drug Research Institute (LDRI), UCLouvain, Université catholique de Louvain, Brussels, Belgium

Martin Roumain, Adrien Paquot & Giulio G. Muccioli

You can also search for this author in PubMed   Google Scholar

Contributions

FS, MVH, and PDC conceived and designed the study. FS performed the experiments and the data analysis. FS, MVH, and PDC performed the interpretation. SVS and GF prepared the samples for sequencing and conducted the sequencing. SVS and GF performed the bioinformatics and statistical analysis for the gut microbiota. MRO and AP prepared the samples for the BA and SCFAs analysis and conducted the experiment. RP performed the PCA and one part of the statistical analysis. MRE counted the CLSs. NMD, JR, and GGM contributed to financial resources and critically revising the manuscript. FS, MVH, and PDC wrote the paper. All authors read and approved the final version before submission.

Corresponding author

Correspondence to Patrice D. Cani .

Ethics declarations

Ethics approval and consent to participate.

Not applicable.

Consent for publication

Competing interests.

PDC is an inventor of patent applications dealing with the use of Akkermansia muciniphila and its components in the context of obesity and related disorders. PDC is co-founder of A-Mansia Biotech SA. The other authors declare no conflict of interest.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: table s1..

RT-qPCR primer sequences for the targeted mouse genes.

Additional file 2: Fig. S1.

Different food intake and water intake profile, body temperature, feces production and energy excreted by feces in ob/ob and db/db mice. ( a ) Food intake evolution (g/mouse/day) measured for the entire experiment (n = 4-5). ( b ) Water intake evolution (mL/mouse/day) measured for the entire experiment (n = 4-5). ( c ) Body temperature (°C) (n = 9-10). ( d ) Feces produced per day (mg/mouse) (n = 4-5). ( e ) Caloric content (cal/g of feces) in 24h feces collected (n = 4-5). ( f ) Energy excreted by feces (cal/g of feces/24h) (n = 4-5). Green: CT ob lean mice, red: ob/ob mice, blue CT db lean mice, and violet: db/db mice. Data are presented as the mean ± s.e.m, * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001. Data were analyzed by one-way ANOVA followed by Tukey’s post hoc test.

Additional file 3: Fig. S2.

Similar visceral adipose tissue features between ob/ob and db/db mice. ( a ) mRNA expression of VAT immune cells markers measured by RT-qPCR. ( b ) mRNA expression of VAT receptors and inflammatory cytokines markers measured by RT-qPCR. Green: CT ob lean mice, red: ob/ob mice, blue CT db lean mice, and violet: db/db mice. Data are presented as the mean ± s.e.m., ** P < 0.01 (n = 8-10). For the mRNA expression, relative units were calculated versus the mean of the CT ob mice values set at 1. Data were analyzed by one-way ANOVA followed by Tukey’s post hoc test.

Additional file 4: Table S2.

Genera displaying significant quantitative abundance differences between mouse genotypes at day 42 (n = 37, Kruskal-Wallis and post-hoc Dunn test). Genera with a prevalence across samples lower than 15% were excluded. Multiple testing correction was performed (BH method).

Additional file 5: Fig. S3.

Different quantitative gut microbiota profiles among the four genotype groups. Green: CT ob lean mice, red: ob/ob mice, blue CT db lean mice, and violet: db/db mice. Data are presented as the mean ± s.e.m, (n = 7–10). Genera with a prevalence across samples lower than 15% were excluded. Data were analyzed by Kruskal-Wallis test with Dunn’s multiple comparison test.

Additional file 6: Table S3.

Taxa-metabolic parameters associations. Spearman correlation between bacterial genera and selected metabolic parameters. Genera whose prevalence was less than 15% of the samples were excluded. Multiple testing correction was performed (Benjamini-Hochberg method).

Additional file 7: Table S4.

Processed quantitative microbiota matrix of day 0, 21, 42.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Suriano, F., Vieira-Silva, S., Falony, G. et al. Novel insights into the genetically obese ( ob/ob ) and diabetic ( db/db ) mice: two sides of the same coin. Microbiome 9 , 147 (2021). https://doi.org/10.1186/s40168-021-01097-8

Download citation

Received : 29 October 2020

Accepted : 19 May 2021

Published : 28 June 2021

DOI : https://doi.org/10.1186/s40168-021-01097-8

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Leptin-deficiency
• Leptin-receptor deficiency
• Liver inflammation
• Adipose tissue inflammation
• Mouse gut microbiota
• Quantitative microbiota profile

ISSN: 2049-2618

• General enquiries: [email protected]

• History & Society
• Science & Tech
• Biographies
• Animals & Nature
• Geography & Travel
• Arts & Culture
• Games & Quizzes
• On This Day
• One Good Fact
• New Articles
• Lifestyles & Social Issues
• Philosophy & Religion
• Politics, Law & Government
• World History
• Health & Medicine
• Browse Biographies
• Birds, Reptiles & Other Vertebrates
• Bugs, Mollusks & Other Invertebrates
• Environment
• Fossils & Geologic Time
• Entertainment & Pop Culture
• Sports & Recreation
• Visual Arts
• Demystified
• Image Galleries
• Infographics
• Top Questions
• Britannica Kids
• Saving Earth
• Space Next 50
• Student Center

How was the leptin protein in mice discovered and how has it benefited diabetes and obesity treatment in humans?

The Obese ( ob/ob ) Mouse and the Discovery of Leptin

Cite this chapter.

• V. Daniel Castracane 5 &
• Michael C. Henson 6  

Part of the book series: Endocrine Updates ((ENDO,volume 25))

2408 Accesses

7 Citations

4 Altmetric

Early theories describing appetite regulation and energy expenditure suggested the lipostatic theory, which hypothesized that some peripheral signal, probably from adipose tissue, would feedback to central satiety centers to modulate food intake and body weight. However, the experimental techniques needed to validate this hypothesis were lacking at that time. Subsequently, two strains of obese mutants, the ob/ob mouse and later the db/db mouse, were discovered 50 and 40 years ago, respectively, and proved invaluable to studying the regulation of food intake, energy expenditure and obesity. Prior to the development of today’s more sophisticated techniques for studying biological and biochemical processes, the use of parabiosis, the surgical attachment of two animals with a shared blood supply, provided valuable insights into the obesity. Information gained from studies of these strains of mice, especially in parabiotic studies with normal counter parts, provided evidence for a humoral factor involved in appetite regulation and initiated the search for its identity. Friedman et al discovered leptin in 1994, and demonstrated that this hormone, the product of the obese (ob) gene, was produced in white adipose tissue and served as the peripheral signal to the central nervous system of nutritional status. After leptin’s discovery, the obese mouse model continued to play an invaluable role in the validation of leptin as the missing factor in the ob/ob mouse and served as a principal model to delineate the many facets of leptin physiology.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Subscribe and save.

• Get 10 units per month
• Download Article/Chapter or eBook
• 1 Unit = 1 Article or 1 Chapter
• Cancel anytime
• Available as PDF
• Read on any device
• Instant download
• Own it forever
• Compact, lightweight edition
• Dispatched in 3 to 5 business days
• Free shipping worldwide - see info
• Durable hardcover edition

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Unable to display preview.  Download preview PDF.

Similar content being viewed by others

Genetically Obese Animals

Discovery of Leptin and Elucidation of Leptin Gene Expression

Y. Zhang, R. Proenca, M. Maffei, M. Barone, L. Leopold and J. M. Friedman, Positional cloning of the mouse obese gene and its human homolog. Nature 372:425–432 (1994).

Article   PubMed   CAS   Google Scholar  

M. Ingalls, M. M. Dickie and G. D. Snell, Obese, a new mutation in the house mouse. J. Hererfiry 41:317–318 (1950).

CAS   Google Scholar  

D. L. Coleman, Diabetes-obesity syndromes in mice. Diabetes 32, Suppl. 1:1–6 (1982).

Google Scholar  

D. L. Coleman, and K. P. Hummel, The influence of genetic background on the expression of the obese (Ob) gene in the mouse. Diabetologia 9, 287–293 (1973).

K. P. Hummel, M. M. Dickie, and D. L. Coleman, Diabetes, a new mutation in the mouse. Science 153, 1127–1128(1966).

M. X. Zarrow, J. M. Yochim, and J. L. McCarthy, Experimental Endocrinology. A sourcebook of basic techniques. Academic Press, New York (1964).

J. C. Finerty, Parabiosis in physiologic studies, Physiol Rev 32:277–302 (1952).

PubMed   CAS   Google Scholar  

D. L. Coleman, Effects of parabiosis of obese with diabetic and normal mice. Diabetologia 9, 294–298 (1973).

G. C. Kennedy, The role of depot fat in the hypothalamic control of food intake in the rat. Proc. R. Soc. Land. B. Biol. Sci. 140, 578–596 (1953).

Article   CAS   Google Scholar  

G. R. Hervey, The effects of lesions in the hypothalamus in parabiotic rats. J. Physiol. 145, 336–352(1959).

D. L. Coleman, Lessons from studies with genetic forms of diabetes in the mouse. Metabolism , 162–164(1983)

G. R. Hervey, Regulation of energy balance. Nature 222, 629–631 (1969).

M. A. Pelleymounter, M. J. Cullen, M. B. Baker, R. Hecht, D. Winters, T. Boone, and F. Collins, Effects of the obese gene product on body weight regulation in ob/ob mice Science 269, 540–543 (1995).

D. S. Weigle, T. R. Bukowski, D. C. Foster, S. Holderman, J. M. Kramer, G. Lasser, C. E. Lofton-Day, D. E. Prunkard, C. Raymond and J. L. Kuijper, Recombinant ob protein reduces feeding and body weights in the ob/ob mouse. J. Clin. Invest. 96, 2065–2070 (1995).

J. L. Halaas, K. S. Gajiwala, M. Maffei, S. L. Cohen, B. T. Chait, D. Rabinowitz, R. L. Lallone, S. K. Burley, and J. M. Friedman, Weight-reducing effects of the plasma protein encoded by the obese gene. Science 269, 543–546 (1995).

L. A. Campfield, F. J. Smith, Y. Guisez, R. Devos, and P. Bum, Recombinant mouse OB protein: Evidence for a peripheral signal linking adiposity and central neural networks. Science 269, 546–549 (1995).

S. Kebanov, CM. Astle, O. DeSimone, V. Ablamunits, and D. E. Harrison, Adipos tissue transplantation protects ob/ob mice from obesity, normalizes insulin sensitivity and restores fertility. J. Endocrinol. 186, 203–211 (2005).

O. Gavrilova, B. Marcus-Samuels, D. Graham, J. K. Kim, G. I. Shulman, A. L. Castle, C. Vinson, M. Ekhaus, and M. L. Reitman, Surgical implantation of adipose tissue reverses diabetes in lipoatrophic mice, J. Clin. Invest . 105, 271–278 (2000).

C. Colombo, J. J. Cutson, T. Yamauchi, C. Vinson, T. Kadowaki, O. Gavrilova, and M. L. Reitman, Transplantation of adipose tissue lacking leptin is unable to reverse the metabolic abnormalities associated with lipoatrophy. Diabetes 51, 2727–2733 (2002).

D. J. Good, Using obese models in research: Special considerations for lACUC members, animal care technicians and researchers. Lab Animal 34, 30–37 (2005).

Article   PubMed   Google Scholar  

S. M. Turner, P. A. Linfoot, R. A. Neese, and M. K. Hellerstein, Sources of plasma glucose and liver glycogen in fasted ob/ob mice. Acta. Diabetol . 42, 187–193 (2005).

M. O. Olatinwo, G. K. Bhat, CD. Stah, and D. R. Mann, Impact of gonadotropin administration on folliculogenesis in prepubertal ob/ob mouse. Mol. Cell. Endocrinol. 245, 121–127 (2005).

F. Dong, X. Zhang, X. Yang, LB. Esberg, H. Yang, Z. Zhang, B. Culver, and J. Ren, Impaired cardiac contractile function in ventricular myocytes from leptin-deficient ob/ob obese mice, J. Endocrinol. 188, 25–36 (2006).

R. Anzawa, M. Bernard, S. Tamareille, D. Baetz, S. Confort-Gouny, J. P. Gascard, P. Cozzone and D. Feuvray, Intracellular sodium increase and susceptibility to ischaemia in hearts from type 2 diabetic db/db mice. Diabetologia 20, 1–9 (2006).

Download references

Author information

Authors and affiliations.

Texas Tech University School of Medicine, Lubbock, TX

V. Daniel Castracane

Department of Biological Sciences, Purdue University Calumet, Hammond, IN

Michael C. Henson

You can also search for this author in PubMed   Google Scholar

Editor information

Editors and affiliations.

Texas Tech University School of Medicine, Lubbock, Texas, USA

V. Daniel Castracane PhD ( Professor Emeritus and Director of Laboratory Operations ) ( Professor Emeritus and Director of Laboratory Operations )

Foundation for Blood Research, P. O. Box 190, Scarborough, ME, USA

Department of Biological Sciences, Purdue University Calumet, Hammond, Indiana, USA

Michael C. Henson PhD ( Professor and Head ) ( Professor and Head )

Dept. of Obstetrics and Gynecology, School of Medicine, Tulane University Health Sciences, New Orleans, LA, USA

Rights and permissions

Reprints and permissions

Copyright information

© 2006 Springer Science+Business Media, LLC

About this chapter

Castracane, V.D., Henson, M.C. (2006). The Obese ( ob/ob ) Mouse and the Discovery of Leptin. In: Castracane, V.D., Henson, M.C. (eds) Leptin. Endocrine Updates, vol 25. Springer, Boston, MA. https://doi.org/10.1007/978-0-387-31416-7_1

Download citation

DOI : https://doi.org/10.1007/978-0-387-31416-7_1

Publisher Name : Springer, Boston, MA

Print ISBN : 978-0-387-31415-0

Online ISBN : 978-0-387-31416-7

eBook Packages : Medicine Medicine (R0)

Share this chapter

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Publish with us

Policies and ethics

• Find a journal
• Track your research

• Search Menu
• Sign in through your institution
• Volume 165, Issue 10, October 2024 (In Progress)
• Volume 165, Issue 9, September 2024
• Advance Articles
• Basic Science Collections
• Thematic Issues
• Clinical Practice Guidelines
• Endocrine Reviews
• Endocrinology
• Journal of the Endocrine Society
• The Journal of Clinical Endocrinology & Metabolism
• JCEM Case Reports
• Molecular Endocrinology (Archives)
• Endocrine Society Journals
• Author Guidelines
• Submission Site
• Open Access
• Why Publish with the Endocrine Society?
• Advertising & Corporate Services
• Reprints, ePrints, Supplements
• About Endocrinology
• Editorial Board
• Author Resources
• Reviewer Resources
• Rights & Permissions
• Member Access
• Terms and Conditions
• Journals on Oxford Academic
• Books on Oxford Academic

Article Contents

Materials and methods, acknowledgments.

• < Previous

A Leptin Dose-Response Study in Obese ( ob/ob ) and Lean (+/?) Mice

• Article contents
• Figures & tables
• Supplementary Data

Ruth B. S. Harris, Jun Zhou, Stephen M. Redmann, Gennady N. Smagin, Steven R. Smith, Erin Rodgers, Jeffrey J. Zachwieja, A Leptin Dose-Response Study in Obese ( ob/ob ) and Lean (+/?) Mice, Endocrinology , Volume 139, Issue 1, 1 January 1998, Pages 8–19, https://doi.org/10.1210/endo.139.1.5675

• Permissions Icon Permissions

This experiment determined the amount of leptin required to correct different abnormalities in leptin-deficient ob/ob mice. Baseline food intakes and body weights of lean (+/?) and obese ( ob/ob ) C57Bl/6J <ob> mice were recorded for 7 days. An Alzet miniosmotic pump was placed in the peritoneal cavity of each mouse and delivered 0, 1, 2, 5, 10, or 42 μg/day human leptin for 7 days. In ob/ob mice, 2 μg leptin/day reduced food intake and body weight, and increased hypothalamic and brain stem serotonin concentrations. All fat pads were reduced 35–40% by 10 μg leptin/day, and liver weight, lipid, and glycogen decreased. Serum insulin and glucose were reduced in all leptin-treated ob/ob mice, and levels were normalized by 10 μg/day leptin. Low rectal temperatures of ob/ob mice were corrected by 10 and 42 μg/day leptin. These doses also increased brown adipose tissue uncoupling protein expression. The only responses in lean mice were a transient reduction in food intake and weight loss with 10 or 42 μg/day leptin. This study shows enhanced leptin sensitivity in ob/ob mice and suggests that increased temperature and sympathetic activity are indirect responses to high concentrations of protein.

GENETICALLY obese ob/ob mice have a single gene mutation that results in a syndrome of obesity that includes diabetes, infertility, hypothyroidism, hypercorticoidism, low sympathetic activity, and impaired thermoregulation ( 1 ). Parabiosis studies with ob/ob mice indicated that the mutation causes a deficiency in a circulating lipostatic factor ( 2 ). In 1994, Zhang et al. ( 3 ) identified the product of the gene mutation in ob/ob mice that was responsible for their obesity and was also the presumed circulating factor. This protein, leptin, has the structure of a long chain helical cytokine ( 4 ) and is expressed in adipose tissue in proportion to adipocyte size ( 5 , 6 ).

Many investigators have established that administration of leptin to genetically obese ob/ob mice reduces food intake and body weight ( 7 – 10 ). The suppression of food intake is mediated by a hypothalamic splice variant of the leptin receptor, OB-Rb, which has a long intracellular domain ( 11 ). OB-Rb is mutated in genetically obese db/db mice ( 12 ), and as expected, they are unresponsive to the hypophagic effects of both peripheral and central administration of leptin ( 7 – 10 ). Obese humans have high circulating concentrations of leptin, but are not responsive to its effects on food intake ( 13 ). As there is a concentration gradient between serum and cerebrospinal fluid leptin ( 14 , 15 ), it has been hypothesized that a rate-limited transport system prevents peripheral leptin from activating the central receptor that suppresses food intake. The limitation on transport may be due either to binding proteins in the circulation ( 16 ) or to a specific transport protein ( 17 ). It has been proposed that one of the leptin receptor subtypes, with a short intracellular domain, may function as a transport protein and regulate leptin uptake into the brain ( 12 ).

In addition to suppressing food intake, peripheral administration of leptin to ob/ob mice corrects infertility ( 18 ), reverses hyperglycemia and hyperinsulinemia ( 9 ), and increases body temperature and metabolic rate ( 9 ). In lean mice, leptin has minimal effects on food intake, but causes the loss of body fat, presumably due to a leptin-induced increase in energy expenditure ( 19 ). Leptin has been shown to increase norepinephrine (NE) turnover in brown, but not white, adipose tissue, suggesting that the metabolic effect of leptin is attributable to activation of the sympathetic nervous system ( 20 ). This would be consistent with leptin-deficient ob/ob mice having low sympathetic tone compared with lean controls ( 1 ).

Leptin has a circulating half-life of approximately 30 min, is released in a pulsatile manner from adipose tissue, and demonstrates a circadian rhythm in circulating levels with a nocturnal elevation in concentration ( 13 , 21 ). The majority of studies investigating the effects of leptin on physiological parameters have administered the protein in one or two daily injections in doses ranging from 3–250μ g/day ( 7 – 10 ). We have found that this method of protein administration causes excessive, intermittent elevations of the serum leptin concentration ( 22 ). In this experiment recombinant human leptin was infused for 7 days at doses ranging from 0–42 μg/day from Alzet miniosmotic pumps placed in the peritoneal cavity of the mice. This method of administration provided constant delivery of protein (0–1.75μ g/h), but did not mimic the diurnal changes in leptin release. The objectives of the study were 2-fold. The first was to determine which of the physiological responses to leptin in ob/ob mice occurred with low doses of constantly infused leptin and which were induced only with large doses of protein. This would allow separation of physiological from potentially pharmacological responses. The second objective was to determine which of the responses observed in ob/ob mice were also apparent in lean mice that already have normal circulating concentrations of leptin and do not respond to the satiety effects of peripherally administered protein.

Five-week-old female C57B1/6J lep <ob>, ob/ob , and lean (+/?) mice were purchased from Jackson Laboratory (Bar Harbor, ME) and housed individually with continuous free access to chow (Purina mouse chow 5015, Ralston Purina, St. Louis, MO) and water. Room temperature was maintained at 26 C, and lights were on for 12 h/day from 0600 h. Body weights and food intakes were recorded daily at 0700 h, and rectal temperatures of mice were measured four times during the study at 0900 h, twice before and twice after pump placement, using a thermistor probe (Thermistor thermometer model 8110–20, Cole Palmer Instrument Co., Chicago, IL).

After 4 days of baseline measurements of body weight and food intake, lean and ob/ob mice were divided into six weight-matched groups, and an Alzet pump (model 1007D, Alza Corp., Palo Alto, CA) was placed in the peritoneal cavity of each mouse. The pumps delivered 0, 1, 2, 5, 10, or 42 μg human recombinant leptin/day in a total volume of 12 μl, and PBS was used as a diluent. Leptin was a gift from ZymoGenetics Corp. (Seattle, WA). After pump placement, measurements of daily food intake and body weight were continued, and rectal temperatures were measured after 2 and 4 days of infusion.

All mice were decapitated in the morning of day 7 of leptin infusion. Trunk blood was collected for serum analysis of insulin (Rat RIA kit, Linco Research, St. Louis, MO), corticosterone (RIA kit, ICN Radiochemical, Irvine, CA), glucose (Sigma Diagnostic Kit 510. Sigma Chemical Co., St. Louis, MO), and human leptin (Human Leptin RIA kit, Linco Research). The carcass, liver, pancreas, ovaries, uterus, spleen, adrenals, heart, kidney, and inguinal, perirenal, retroperitoneal, mesenteric, gonadal, and intrascapular brown fat were dissected and weighed. Tissues smaller than 250 mg were weighed on a microbalance (Cahn C-31 microbalance, Cahn Instruments, Cerritos, CA). The liver, hypothalamus, brown fat, and gonadal fat were snap-frozen. All procedures were approved by the Pennington Biomedical Research Center Institutional Animal Care and Use Committee.

TRIzol reagent (Life Technologies, Grand Island, NY) was used to extract total RNA from gonadal fat for measurement of leptin expression by Northern blot analysis, as described previously ( 6 ), and from brown fat for measurement of uncoupling protein (UCP) messenger RNA (mRNA) by Northern blot analysis as described previously ( 23 ), using a complementary DNA probe generously provided by Dr. Daniel Riquier. The hypothalamus, brain stem, and cortex were dissected and snap-frozen in liquid nitrogen for analysis of monoamines by HPLC, as described previously ( 24 ). Liver tissue was frozen for subsequent analysis of glycogen by the method of Lo et al. ( 25 ) and for lipid content by chloroform-methanol extraction. The triceps surae muscle from mice in the 0, 2, and 10 μg leptin/day groups was frozen for enzyme analysis. A 5% homogenate was made of the muscle samples in Tris buffer (175 m m KCl, 2 m m EDTA, and 10 m m Tris-HCl, pH 7.4). Citrate synthase was used as a general marker for oxidative capacity and was assayed as described by Srere ( 26 ). β-Hydroxyacyl coenzyme A dehydrogenase (HOAD) activity was used as a marker for fatty acid oxidation potential and was determined as described by Askew et al. ( 27 ). Hexokinase activity was used a marker for glucose utilization potential and was assayed according to the method of Uyeda and Racker ( 28 ).

Small pieces of frozen liver from lean and ob/ob mice treated with either 0 or 10 μg/day leptin were homogenized in Krebs bicarbonate buffer, pH 7.5, containing protease inhibitors (10μ m leupeptin, 2 U/ml aprotinin, and 1 μ m phenylmethylsulfonylfluoride). A crude membrane fraction was prepared by centrifuging the homogenate for 10 min at 3,000 × g and then recentrifuging the supernatant at 11,000 × g for 20 min. Samples (40 μg) of both the resulting supernatant and pellet were separated by SDS-PAGE in a 9% acrylamide gel in Tris glycine buffer (25 m m Tris, 192 m m glycine, and 0.1% SDS, pH 8.3). The proteins were transferred to a polyvinylidene difluoride membrane (Boehringer Mannheim, Mannheim, Germany) in 25 m m Tris, 192 m m glycine, and 20% methanol. Leptin receptor was detected by Western blot using a polyclonal rabbit antimouse OB-R antibody, a gift from Affinity BioReagents (Golden, CO). The blot was developed using a chemiluminescence system (BM Chemiluminescence Blotting Substrate, Boehringer Mannheim) according to the manufacturer’s directions using the first antibody at a 1:4,000 dilution and an antirabbit IgG POD second antibody (Boehringer Mannheim).

Statistical analysis

For each genotype, the response variables food intake and body weight were separately modeled as a repeated measures ANOVA over the course of the experiment. To provide overall tests of the dosage effect of leptin on each variable, a profile analysis was effected by testing the appropriate contrasts corresponding to the parallel, coincident, and level profiles hypotheses for days 4–7 of infusion after the mice had recovered from the surgery. Comparisons of the effects of different leptin dosages on the response measures on specific days of infusion were obtained by testing the appropriate contrasts corresponding the hypothesis of interest; where appropriate, P values reported for these contrasts have been adjusted for multiple comparisons by Bonferroni’s method. Where possible, tests were conducted using Satterthwaite’s approximation to determine the appropriate degrees of freedom ( 29 ).

Rectal temperatures were analyzed by repeated measures ANOVA, with day as the repeated measure. Organ weights, leptin mRNA, UCP mRNA, liver glycogen, liver lipid, muscle enzymes, serum measurements, and brain neurotransmitter concentrations were analyzed by two-way ANOVA to determine whether there were genotype effects and by one-way ANOVA with post-hoc Duncan’s multiple range test to determine treatment effects within each genotype. The one-way ANOVA was performed even when the two-way ANOVA did not show an interaction between genotype and leptin, as the value for some of the parameters, such as body fat and serum hormones, were so much greater in obese than in lean mice that variance within ob/ob groups masked substantial treatment effects in lean animals. The SAS System for Windows (release 6.12, SAS Institute, Cary, NC) was used for computations.

Food intakes of ob/ob and lean mice are shown in Fig. 1 . The dramatic reduction in food intake on the first day of leptin infusion was partially due to surgery. During subsequent days of infusion, 1 μg leptin/day tended to reduce food intake compared with that of control obese mice, but repeated measures analysis showed that the difference was significant only on days 1 and 5 of infusion (adjusted P < 0.003 and P < 0.03, respectively). All other leptin doses (2, 5, 10, and 42 μg/day) suppressed food intake in the obese mice from the first day of infusion, and intake remained below that of controls until the end of the experiment, although the intake of mice receiving 2 and 5 μg/day gradually increased during the 7 days of infusion. A maximal effect was reached with 10 μg leptin/day, at which dose food intake remained at approximately 20% of the control intake. In lean mice, food intake was transiently suppressed on the first 2 days of leptin infusion in mice receiving 10 or 42 μg leptin/day. Lower doses of leptin had no significant effect on food intake. Daily body weights of the mice are shown in Fig. 2 . ob/ob mice receiving 0 or 1 μg/day leptin gained weight during the 7 days of infusion, and there was no significant difference between the body weights of the two groups. The body weights of mice receiving 2 or 5 μg/day plateaued at a lower level than their preinfusion weight, whereas mice receiving 10 or 42 μg/day constantly lost weight. Lean mice receiving 42 μg leptin/day weighed significantly less than controls from day 2 of infusion. For mice receiving 10 μg/day leptin, the weight difference was significant from day 3. The 5 μg/day group weighed significantly less ( P < 0.02) than controls on days 4, 5, and 6 of leptin infusion, but not on day 7, and mice receiving 2 μg/day weighed significantly less than controls on days 5 and 6 of infusion.

Daily food intakes of lean and obese mice. Data are the mean ± sem for groups of six mice. An Alzet pump delivering 0, 1, 2, 5, 10, or 42 μg/day human recombinant leptin was placed in the peritoneal cavities of the mice on the day indicated by the arrow .

Daily body weights of lean and obese mice. Data are the mean ± sem for groups of six mice. An Alzet pump delivering 0, 1, 2, 5, 10, or 42 μg/day human recombinant leptin was placed in the peritoneal cavities of the mice on the day indicated by the arrow . Statistical analysis revealed significant differences in body weights of ob/ob mice receiving 2μ g leptin/day, or more, compared with those of controls. In lean mice, the animals receiving 10 or 42 μg/day leptin weighed significantly less than controls.

Rectal temperatures of the mice are shown in Fig. 3 . Repeated measurements of temperature caused a progressive increase in temperature of all animals. Repeated measures two-way ANOVA showed a significant effect of genotype ( P < 0.00001) and no significant effect of leptin dose, but a significant effect of day ( P < 0.0001) on temperatures of the mice. There were also significant interactions between genotype and day ( P < 0.0001) and among genotype, leptin dose, and day ( P < 0.05). Before leptin infusion, the temperatures of all obese mice were significantly lower than those of lean animals. By the second and fourth days of infusion, the temperatures of ob/ob mice receiving 10 or 42μ g/day leptin were no longer different from those of lean animals. There was no effect of leptin on the temperatures of lean mice. Results from measurements of UCP mRNA expression are shown in Fig. 4 . Two-way ANOVA showed no significant effect of genotype or leptin on UCP expression. However, when a one-way ANOVA was performed within the obese genotype, UCP expression was significantly elevated in ob/ob mice receiving 10 and 42μ g leptin/day compared with that in other obese mice.

Rectal temperatures of lean and obese mice measured twice before leptin infusion and twice during leptin infusion. Data are the mean ± sem for groups of six mice. Obese mice had significantly lower temperatures than lean mice, except in the 10 and 42 μg/day groups on the second and fourth days of leptin infusion. Superscripts indicate a significant difference between treatment groups within the obese genotype on the fourth day of infusion. There were no significant differences within lean or obese genotypes on any other day.

Expression of UCP mRNA intrascapular brown fat from lean and obese mice infused for 7 days with increasing amounts of human recombinant leptin. Data are the mean ± sem for groups of four to six mice. Superscripts indicate significant differences in expression between groups of obese mice. There was no effect of leptin on UCP expression in lean mice.

The weights of different adipose depots are shown in Table 1 . In obese mice, all fat pad weights were reduced with a leptin dose of 2 μg/day or more. A maximal effect was obtained with 10 μg/day. In the lean mice, the only significant change in fat pad weight was the retroperitoneal fat of mice in the 10μ g/day group. The other fat pads also tended to be reduced at leptin doses of 10 and 42 μg/day, but differences did not reach statistical significance. Gonadal fat leptin mRNA expression is shown in Fig. 5 . Two-way ANOVA showed a significant effect of both genotype ( P < 0.0001) and leptin dose ( P < 0.0004) and a significant interaction between the two variables ( P < 0.0013). Post-hoc Duncan’s multiple range test demonstrated that leptin expression was substantially higher in adipose tissue from ob/ob mice than in that from lean mice. There was no effect of leptin administration on leptin expression in lean mice, but the two highest doses of leptin caused a significant reduction in expression in ob/ob mice.

Expression of leptin mRNA measured in gonadal fat from lean and obese mice after 7 days of infusion with increasing amounts of leptin. Data are the mean ± sem for groups of four to six mice. Leptin expression was measured by Northern blot analysis and expressed as a ratio to 28S ribosomal RNA. Superscripts indicate significant differences in levels of expression in fat from obese mice. There was no effect of leptin on expression in tissue from lean mice.

The effect of leptin infusion on the weight of fat depots in lean and ob/ob mice

Data are the mean ± sem for groups of six mice infused with PBS or leptin for 7 days. The statistical summary represents a two-way ANOVA, demonstrating genotype effects. Statistically significant differences within genotype were determined by one-way ANOVA and subsequent Duncan’s multiple range test; different superscript letters indicate significant differences between treatment groups within a genotype. Rp, Retroperitoneal. Total is the sum of the weights of the five dissected fat pads.

Organ weights are shown in Table 2 . Two-way ANOVA showed genotype effects on the weights of kidneys, adrenals, and ovaries. There was no statistically significant effect of any dose of leptin on the weight of any of the organs measured in either obese or lean mice. Liver composition is shown in Table 3 . Livers of ob/ob mice were significantly larger than those in lean mice and contained substantially more lipid and glycogen. In contrast to the other organs, leptin significantly reduced liver weight in obese mice starting at the 1 μg/day dose and reaching a maximal effect with 10 μg/day. Liver lipid was reduced at all doses of leptin, and liver glycogen content was decreased by 2 μg/day. None of the doses of leptin had any effect on liver weight, liver lipid, or liver glycogen content in lean mice. A Western blot of the short form leptin receptors is shown in Fig. 6 . There was no detectable long form (OB-Rb) receptor, and there was no obvious effect of leptin on the amount of short form receptor present in livers of lean or ob/ob mice, although the results suggest that leptin receptors were expressed at higher levels per unit protein in livers of lean mice than in those of obese mice.

Liver short form leptin receptor detected by Western blot using a polyclonal antibody raised to the extracellular membrane portion of the receptor. The top blot is liver from ob/ob mice, and the bottom blot is liver from lean mice infused with either 0 or 10 μg/day leptin. Tissue was homogenized and centrifuged at 3,000 × g for 10 min The supernatant was centrifuged at 11,000 × g for 20 min. Both the pellet (P) and the supernatant(s) were analyzed for receptor. Forty micrograms of protein were loaded in each lane. The 100-kDa band represents the receptors with a short intracellular domain (OB-Ra, OB-Rc, OB-Rd, and OB-Rf). There was no effect of leptin treatment on receptor in lean or obese mice.

Organ weights of lean and obese mice treated with leptin

Data are the mean ± sem for six mice infused with PBS or leptin for 7 days. The statistical summary represents a two-way ANOVA, testing for a genotype effect. There were no significant differences between treatment groups within a genotype for any organ, analyzed by one-way ANOVA.

Liver composition of lean and ob/ob mice treated with leptin

Data are the mean ± sem for six mice infused with PBS or leptin for 7 days. The statistical summary represents a two-way ANOVA, testing for genotype effects. Different superscript letters indicate significant differences between treatment groups within a genotype, determined by one-way ANOVA and subsequent Duncan’s multiple range test.

Triceps surae muscle hexokinase, citrate synthase, and HOAD activity are shown in Table 4 . Enzyme activity was only measured in tissue from three treatment groups per genotype. HOAD activity was significantly higher in ob/ob than in lean mice, but genotype had no effect on citrate synthase or hexokinase activity. There was no effect of leptin on enzyme activity in ob/ob mice, but both hexokinase and HOAD activities were increased in lean mice treated with 2 μg leptin/day. This increase was reversed by 10 μg leptin/day.

Muscle enzyme activity in lean and ob/ob mice infused with leptin for 7 days

Data are the mean ± sem for groups of four to six mice. Enzyme activity was measured in the triceps muscles collected from animals that had been infused with 0, 2, or 10 μg/day recombinant human leptin. The statistical summary represents a two-way ANOVA used to detect genotype effects. Different superscript letters indicate statistically significant differences between treatment groups within a genotype, determined by one-way ANOVA and post-hoc Duncan’s multiple range test.

The results of serum analysis are shown in Table 5 . In both lean and obese mice, serum leptin tended to increase with increased levels of infusion. A statistically significant difference was detected between controls and animals given 10 or 42 μg/day. There was no difference in leptin concentrations measured in mice infused with 10 or 42 μg/day leptin, suggesting that the pumps were unable to deliver the highest concentrations of protein efficiently. There was a significant genotype effect on serum insulin, glucose, and corticosterone, all of which were elevated in ob/ob mice compared with lean mice. Hyperinsulinemia in obese mice was improved significantly with leptin doses as low as 2 μg/day, and with 10 μg leptin/day, the serum insulin level was similar to that in lean mice. There was no change in the serum insulin concentration in any group of lean animals, and serum glucose was significantly reduced only by 42 μg leptin/day. There was a trend for increasing doses of leptin to suppress serum corticosterone in obese mice, but it did not reach statistical significance ( P < 0.09). In lean mice, there was no effect of leptin on corticosterone ( P < 0.8).

The effect of leptin infusion on serum hormone and glucose levels in lean and obese mice

Data are the mean ± sem for six mice infused with PBS or leptin for 7 days. The statistical summary represents a two-way ANOVA, testing for genotype effects. Different superscript letters indicate significant differences within a genotype, determined by one-way ANOVA and subsequent Duncan’s multiple range test.

The results of HPLC analysis of monoamines and their metabolites in brain stem, hypothalamus, and frontal cortex are shown in Tables 6 and 7 . Table 6 shows concentrations of serotonin (5-HT), its metabolite 5-HIAA, and the ratio of the two as an index of 5-HT metabolism. Two-way ANOVA indicated a significant genotype effect on 5-HT and 5-hydroxyinoleacetic acid (5-HIAA) concentrations in the hypothalamus and brain stem, with both compounds being present at higher concentrations in obese than in lean animals. 5-HT metabolism was different between lean and obese mice only in the brain stem. There were no effects of genotype on concentrations in the frontal cortex. Within genotypes, leptin administration caused significant elevations in 5-HT metabolism in obese, but not lean, mice. Leptin doses of 10 or 42 μg/day caused significant increases in 5-HIAA concentrations, compared with control values, in both the brain stem and hypothalamus of ob/ob mice. Leptin had no effect on the concentrations of any other neurotransmitter measured in the three brain areas dissected. Therefore, Table 7 summarizes the effect of genotype on the concentrations of monoamines within the three brain regions. In the brain stem, there was a significant genotype effect on the concentrations of all monoamines and metabolites measured, except for 3-methoxy, 4-hydroxyphenylethylene glycol (MHPG), MHPG/NE, and homovanillic acid (HVA)/dopamine (DA). In the hypothalamus there was no significant genotype effect on 3,4-dihydroxyphenylacetic acid (DOPAC) or DA concentrations or on the DOPAC/DA ratio; however, there were significant differences in the MHPG, MHPG/NE, and HVA/DA contents of tissue from lean and ob/ob mice. In the frontal cortex, there was a genotype effect on HVA, DOPAC/DA, and HVA/DA concentrations. MHPG was below detectable levels in this tissue.

Serotonin and 5-HIAA (nanograms per mg tissue) in brain stem, hypothalamus, and cortex of lean and ob/ob mice treated with leptin

Data are the mean ± sem for six mice infused with PBS or leptin for 7 days. The statistical summary represents a two-way ANOVA, testing for effects of genotype and leptin dose. Different superscript letters indicate statistically significant differences between treatment groups within a genotype, determined by one-way ANOVA and subsequent Duncan’s multiple range test.

Brain catecholamines in lean and ob/ob mice

Data are the mean ± sem for groups of 36 lean or ob/ob mice. The significance is for a genotype effect on catecholamine concentration determined by two-way ANOVA in which genotype and leptin dose were independent variables. There was no significant effect of leptin or of an interaction between genotype and leptin on any of the parameters; therefore, the means for the two genotypes are the average from all animals in the experiment.

In this experiment, lean and genetically obese ob/ob mice were infused with increasing doses of recombinant human leptin for 7 days to define genotypic differences in response and to determine which responses to leptin were observed at low concentrations of protein and which required large amounts of protein. Analysis of serum human leptin by RIA indicated that we achieved our goal of causing progressive increases in circulating leptin concentrations in both lean and obese mice up to the 10 μg/day dose. As similar circulating concentrations were found in animals infused with 10 and 42 μg/day leptin, it is possible that leptin precipitated from the highly concentrated solution used to fill the 42-μg pumps or that some of the protein bound to the internal walls of the pump. Alternatively, mice may have a mechanism for clearing excessive amounts of protein from the circulation. Detection of leptin in serum from ob/ob mice treated with PBS indicated that the RIA was not specific for leptin, but it did permit comparison of relative concentrations of protein in serum from the different treatment groups. The lowest infusion dose of 1 μg/day did not increase detectable leptin beyond that measured in mice receiving PBS.

All doses of leptin greater than 1 μg/day caused significant reductions in the food intake and body weight of ob/ob mice. Low doses of leptin (2 and 5 μg/day = 0.08 and 0.2 μg/h) caused an initial drop in food intake that was partially reversed by the end of the experiment and resulted in stabilization at a reduced body weight. In contrast, the two highest doses of leptin caused a stable reduction in food intake and continuous weight loss in obese mice. Brown adipose UCP levels and rectal temperature were also elevated with high doses of leptin, as discussed below, and it is likely that the associated elevation of energy expenditure contributed to weight loss in these mice. In lean mice, only the two highest doses of leptin caused significant changes in food intake and body weight. These results demonstrate that ob/ob mice have an increased sensitivity to the energy balance effects of leptin compared with that in lean mice. Although the intake of lean mice returned to control levels, there was no evidence of compensatory hyperphagia, suggesting that when the mice became resistant to the feeding effects of leptin, they were also insensitive to the existing reduction in body weight.

The ability of lean mice to develop resistance to the satiety aspects of leptin suggest that these animals are a more appropriate model than ob/ob mice for investigating leptin activity in humans. Obese humans and mice made obese by dietary means have elevated circulating concentrations of leptin but maintain a normal food intake ( 30 – 32 ). Central infusion of leptin into dietary obese mice inhibits intake ( 32 ), indicating that the resistance to peripheral leptin is caused by a failure to transport the protein to the hypothalamic long form, OB-Rb, receptor, which is responsible for the hypophagic effects of leptin ( 12 ). A similar rate-limiting transport system appears to be present in humans, as a concentration gradient in leptin is maintained between blood and cerebrospinal fluid ( 14 , 15 ). The adaptive mechanism that limits leptin transport to the brain may be due to circulating binding proteins that limit the amount of free protein available for transport ( 16 ) or to a rate-controlled transport system at the blood-brain barrier ( 14 ). In ob/ob mice, this adaptive mechanism is either absent or substantially inhibited. The gradual increase in food intake of ob/ob mice given 2 or 5 μg/day leptin suggest that they have a limited ability to prevent leptin from reaching central receptors that control food intake.

Weight loss in ob/ob mice was accompanied by a reduction in body fat content, and there was no obvious site-specific response. In obese mice, increasing doses of leptin caused progressive loss of hepatic lipid and glycogen stores; however, by the end of this experiment, liver lipid and glycogen levels in mice given the two highest doses of leptin were still twice those found in lean mice. Despite the substantial change in liver composition and metabolism, there was no change in the level of expression of hepatic short form leptin receptors (OB-Ra, OB-Rc, OB-Rd, and OB-Rf), which have been shown to have signaling capability ( 33 , 34 ). As reported previously ( 35 ), the long form of the leptin receptor, OB-Rb, was not detectable in the liver. If the change in liver composition had been directly mediated by leptin, a change in receptor number may have been expected as liver energy stores declined. The design of this experiment did not allow us to determine which responses in the mice were direct effects of leptin and which were secondary to the state of negative energy balance induced by leptin.

In lean mice, all fat pads were also reduced by 30–50% in animals treated with 10 μg/day leptin compared with those in controls; however, the difference was not statistically significant due to the relatively small size of the pads even in control animals. This observation confirms the hypothesis of Levin et al. ( 19 ) that leptin has metabolic effects, independent of those associated with hypophagia, that lead to a loss of body fat in both lean and ob/ob mice. As leptin had no significant effect on the hepatic lipid or glycogen content of lean mice, these results also suggest that the metabolic effect is tissue specific, diverting nutrients from adipose tissue to other tissues that have a higher metabolic rate. This concept was supported by the measurement of muscle enzyme activity. In ob/ob mice, there was no significant effect of leptin treatment on any of the three enzymes measured, suggesting that glucose and fatty acid metabolism in these tissues was not substantially changed, even with 10 μg/day leptin. In lean mice, hexokinase and HOAD activities increased with a low dose of leptin and then decreased with the higher dose. These changes were small, and significance was due to the absence of variance in some of the treatment groups, but they may be representative of a shift in nutrient utilization from glucose to fatty acid oxidation, consistent with fatty acids being mobilized from adipose tissue.

Measurements of serum insulin and glucose showed that 2 μg/day human leptin caused a significant reduction in basal serum insulin and glucose levels in ob/ob mice, confirming previous reports that leptin improves glucose clearance in ob/ob mice ( 9 , 10 ). The improvement cannot be entirely attributed to the reduction in food intake and body fat content of the mice, as Pellymounter et al. ( 9 ) found a reduction in serum insulin with a dose of leptin that did not change the body weights of ob/ob mice. Emillson et al. ( 36 ) have shown that leptin directly inhibits glucose-stimulated insulin release from pancreatic β-cells; however, the changes in ob/ob mice indicate an improvement in tissue insulin responsiveness, leading to a reduced requirement for insulin, rather than an inhibition of insulin release in the absence of a change in glucose uptake. If this had been the case, leptin-treated mice would have had lower insulin, but higher serum glucose, concentrations than controls. ob/ob mice given 10 μg/day leptin had serum insulin and glucose concentrations equivalent to those in lean controls. In these animals, which had a significantly reduced food intake and were mobilizing body fat, it is likely that a combination of factors, including direct effects of leptin in tissue insulin sensitivity and a reduced glucose load, contributed to the drop in serum insulin and glucose concentrations.

In vitro studies have shown that leptin induces insulin resistance in HepG2 cells ( 37 ), a hepatocellular carcinoma cell line, and rat adipocytes ( 38 ). These observations are not consistent with improved insulin responsiveness in leptin-treated ob/ob mice. There are a number of possible explanations for the discrepancy. The first is that we measured basal insulin and glucose levels when the animals were in a nonfed state. It is possible that leptin changes glucose-stimulated insulin release and insulin responsiveness, which would not have been detected in this experiment. Another explanation is that ob/ob mice are abnormal in their response to leptin. This would not be too surprising, as ob/ob mice have an increased sensitivity to leptin and are the only animals that remain hypophagic in response to peripherally administered leptin ( 7 , 9 , 32 , 39 ). In addition, there was no effect of leptin on basal serum insulin concentrations in lean mice, and noninsulin-dependent diabetes is associated with both elevated leptin and insulin levels ( 40 , 41 ). Finally, in vitro studies may not be representative of in vivo responses due to the absence of appropriate feedback systems, compensatory mechanisms, and secondary responses to leptin treatment. Studies of in vivo glucose utilization in leptin-treated animals are required to confirm the relevance of in vitro studies to the whole animal response to leptin. In addition to reducing serum insulin and glucose, 2 μg/day leptin appeared to reverse the hypercorticoidism of ob/ob mice, although the 68% drop in the average corticosterone concentration was not statistically significant. This change in circulating corticosterone may also have contributed to the improved insulin status of ob/ob mice, as glucocorticoids inhibit glucose uptake ( 42 ).

Others have speculated that the metabolic effects of leptin are associated with activation of the sympathetic nervous system ( 20 ). Sympathetic tone is reduced in ob/ob mice that are leptin deficient ( 1 ), and leptin treatment has been reported to increase norepinephrine turnover in brown, but not white, adipose tissue ( 20 ). In this experiment, brown adipose tissue UCP expression was used as an indirect index of sympathetic activity and was increased only in ob/ob mice infused with 10 or 42 μg/day leptin. These results demonstrate that much higher doses of leptin are required to increase UCP expression and sympathetic activity than are required to change food intake and cause weight loss in either lean or ob/ob mice. Changes in rectal temperatures of ob/ob mice paralleled the increase in UCP expression. It is well established that thermoregulation is impaired in both ob/ob mice, which are deficient in leptin, and db/db mice, which have a mutated long form leptin receptor ( 1 ). As the temperatures of ob/ob mice receiving 10 or 42μ g/day leptin were not different from those of lean mice, it appears that high doses of leptin mediate either an increase in heat production or a reduction in heat loss. Both of these responses would be consistent with increased sympathetic activity, its associated activation of brown adipose tissue, and vasoconstriction in the skin ( 43 ). However, it is well established that cytokines cause a febrile response ( 44 ), although this is usually a transient effect in conditions of trauma or infection ( 45 ), and leptin has been shown to stimulate macrophage cytokine production and phagocytic activity in vitro ( 46 ). Therefore, it is possible that the effect on body temperature was also partially due to a leptin-induced increase in concentrations of inflammatory cytokines. We did not measure any cytokines other than leptin in this study due to a limitation in the amount of serum available.

Others ( 18 ) have reported that 14 days of leptin injections caused a significant increase in ovarian and uterine weights in ob/ob mice, but in this experiment we did not find any significant effect of leptin on the weights of reproductive organs in either lean or ob/ob mice. The uterine weights in ob/ob mice tended to increase with leptin treatment, but the difference did not reach statistical significance due to the large variability within each group. There was no indication of leptin having any effect on the ovaries of ob/ob mice, which were significantly smaller than those of lean animals. The difference in these results and those reported previously ( 18 ) may be due to the short duration of this experiment, which involved only 7 days of leptin infusion, compared with the 5-day reproductive cycle of mice. Either this period of leptin infusion was not long enough to stimulate the growth of reproductive tissue, or early stages of cell development, which would have been detected by histological examination of the tissue, were not apparent as a significant change in tissue weight.

Measurements of catecholamines and monoamines in several brain areas of lean and ob/ob mice demonstrated many genotypic differences. Although NE was elevated in both the hypothalamus and brain stem of ob/ob mice compared with levels in lean mice, NE metabolism (MHPG/NE) was lower in the hypothalamus of ob/ob mice than lean mice, confirming previous reports of reduced NE synthesis and metabolism in these animals ( 47 ). The failure of leptin to reverse this defect demonstrates that it is not a direct result of leptin deficiency and is not associated with the hyperphagia, hyperinsulinemia, or hypothermia. DA metabolism, measured as DOPAC/DA or HVA/DA, was elevated in ob/ob mice compared with lean animals. The lack of site specificity and the failure of leptin to correct this difference suggest that it is also unrelated to the energy balance aspects of leptin deficiency. 5-HT and its metabolism, indicated by the 5-HIAA/5-HT ratio, was elevated in ob/ob mice and responded to leptin treatment. Doses of leptin as low as 2 μg/day leptin increased the 5-HIAA concentration and 5-HT metabolism in the hypothalamus and brain stem, but not the cortex, of ob/ob , but not lean, mice. The site specificity of this response suggests an association between the reduced food intake in leptin-treated ob/ob mice and 5-HT, which is known to suppress food intake ( 48 , 49 ). It is possible that the satiety effect of leptin in ob/ob mice is mediated in part by modulation of this neurotransmitter, whereas the absence of a change in 5-HT metabolism in leptin-treated lean mice is consistent with their recovery of a normal food intake by the end of the experiment. To date, the majority of studies investigating leptin-sensitive central control of food intake have focused on neuropeptide Y, which is elevated in the hypothalamus of genetically obese mice and rats and is down-regulated by leptin ( 38 , 50 ). However, neuropeptide Y knockout mice are responsive to leptin ( 51 ), indicating some redundancy in the mechanisms that mediate leptin-induced hypophagia. The results of this experiment demonstrate a correlation between 5-HT metabolism and food intake in ob/ob mice, but further investigation is needed to establish the true relationship among leptin, 5-HT, and food intake.

The second objective of this study was to determine genotypic differences in response to leptin in lean and ob/ob mice. The most obvious difference was an exaggerated sensitivity to leptin in ob/ob mice compared with that in lean animals. In ob/ob mice, several changes were observed with 2 μg/day leptin, whereas only the two highest doses of leptin caused reliable changes in lean mice, consistent with other reports of no response in lean animals to doses of leptin that produce significant changes in food intake, body weight, serum insulin, and body temperature in ob/ob mice ( 9 ). Potential explanations for the increased sensitivity of ob/ob mice to the protein include a difference in circulating concentrations of protein; a decreased amount of binding protein in ob/ob mice, resulting in an increased amount of bioavailable protein in the circulation; an increased number of leptin receptors in ob/ob mice; or a failure of obese mice to down-regulate the receptor in response to continuous agonism by leptin. There was no obvious difference in circulating concentrations of leptin in lean and ob/ob mice given the same dose of protein, indicating that increased responsiveness in ob/ob mice could not be attributed to elevated levels of circulating protein. We were not able to determine the proportion of circulating leptin that was bound to either binding proteins ( 16 ) or soluble receptor ( 11 ), and the possibility of an increased amount of free leptin in ob/ob mice cannot be excluded. In this experiment we only measured short form receptors present in the liver. There was no effect of leptin treatment on the amount of receptor present in either lean or obese mice, and the Western blots suggested that more receptor was present in tissue from lean than ob/ob mice.

In summary, the results of this experiment, in which a large number of variables were measured in lean and obese mice treated with increasing amounts of leptin, demonstrated that ob/ob mice were more responsive to leptin than were lean animals; they showed reduced food intake, body weight, and serum insulin and glucose levels and increased hypothalamic and brain stem serotonin metabolism when given 2 μg/day leptin. The small amount of protein needed to initiate these changes suggests that they are all primary physiological responses to leptin. As human and mouse leptin have 84% homology ( 3 ), it is likely that even lower concentrations of murine recombinant leptin would be required to initiate a response in ob/ob mice. Correction of hypothermia and increased expression of brown fat UCP in ob/ob mice required relatively large doses of protein, which suggests that these effects are representative of responses to a pharmacological dose of leptin. In contrast to ob/ob mice, the only response observed in lean mice was a transient reduction in food intake and a reduction in body weight of mice given the two highest doses of leptin. The absence of leptin in ob/ob mice during growth and development may cause them to be especially sensitive to exogenous protein and result in a failure to adapt to the protein, such that mechanisms that prevent a continued effect of leptin on food intake in lean mice are absent or minimal in ob/ob mice. In this experiment we did not determine which of the changes in leptin-treated ob/ob mice were a direct response to the protein and which were secondary to the state of negative energy balance that resulted from a sustained inhibition of food intake. The exaggerated sensitivity of ob/ob mice to leptin indicates that other animal models, such as dietary obese mice, are more appropriate when considering the effect of leptin on physiological and biochemical parameters in vivo .

The authors thank Joseph Kuijper (ZymoGenetics, Seattle, WA) for providing the human recombinant leptin used in this study, Dr. Daniel Ricquier (Centre de Recherche sur l’Endocrinologicole Moleculaire et le Development, Meudon, France) for providing the complementary DNA probe for UCP, and Phillip Schwartz (Affinity BioReagents, Golden, CO) for providing leptin receptor antibody.

Bray GA , York DA 1979 Hypothalamic and genetic obesity in experimental animals: an autonomic and endocrine hypothesis. Physiol Rev 59 : 719 – 809

Google Scholar

Coleman DL 1973 Effects of parabiosis of obese with diabetes and normal mice. Diabetologia 9 : 294 – 298

Zhang Y , Proenca R , Maffei M , Barone M , Leopold L , Friedman JM 1994 Positional cloning of the mouse obese gene and its human homologue. Nature 372 : 425 – 432

Zhang F , Basinski MB , Beals JM , Briggs SL , Churgay LM , Clawson DK , DiMarchi RD , Furman TC , Hale JE , Hsiung HM , Schoner BE , Smith DP , Zhang XY , Wery J-P , Schevitz RW 1997 Crystal structure of the obese protein leptin-E100. Nature 387 : 206 – 209

Hamilton BS , Paglia D , Kwan AY , Deitel M 1995 Increased obese mRNA expression in omental fat cells from massively obese humans. Nat Med 1 : 953 – 956

Harris RB , Ramsay TG , Smith SR , Bruch RC 1996 Early and late stimulation of ob mRNA expression in meal-fed and overfed rats. J Clin Invest 97 : 2020 – 2026

Campfield LA , Smith FJ , Guisez Y , Devos R , Burn P 1995 Recombinant mouse OB protein: evidence for a peripheral signal linking adiposity and central neural networks. Science 269 : 546 – 549

Halaas JL , Gajiwala KS , Maffei M , Cohen SL , Chait BT , Rabinowitz D , Lallone RL , Burley SK , Friedman JM 1995 Weight-reducing effects of the plasma protein encoded by the obese gene. Science 269 : 543 – 546

Pelleymounter MA , Cullen MJ , Baker MB , Hecht R , Winters D , Boone T , Collins F 1995 Effects of the obese gene product on body weight regulation in ob/ob mice. Science 269 : 540 – 543

Weigle DS , Bukowski TR , Foster DC , Holderman S , Kramer JM , Lasser G , Lofton-Day CE , Prunkard DE , Raymond C , Kuijper JL 1995 Recombinant ob protein reduces feeding and body weight in the ob/ob mouse. J Clin Invest 96 : 2065 – 2070

Chen H , Charlat O , Tartaglia LA , Woolf EA , Weng X , Ellis SJ , Lakey ND , Culpepper J , Moore KJ , Breitbart RE , Duyk GM , Tepper RI , Morgenstern JP 1996 Evidence that the diabetes gene encodes the leptin receptor: identification of a mutation in the leptin receptor gene in db/db mice. Cell 84 : 491 – 495

Lee GH , Proenca R , Montez JM , Carroll KM , Darvishzadeh JG , Lee JI , Friedman JM 1996 Abnormal splicing of the leptin receptor in diabetic mice. Nature 379 : 632 – 635

Caro JF , Sinha MK , Kolaczynski JW , Zhang PL , Considine RV 1996 Leptin: the tale of an obesity gene. Diabetes 45 : 1455 – 1462

Schwartz MW , Peskind E , Raskind M , Boyko EJ , Porte Jr D 1996 Cerebrospinal fluid leptin levels: relationship to plasma levels and to adiposity in humans. Nat Med 2 : 589 – 593

Caro JF , Kolaczynski JW , Nyce MR , Ohannesian JP , Opentanova I , Goldman WH , Lynn RB , Zhang PL , Sinha MK , Considine RV 1996 Decreased cerebrospinal-fluid/serum leptin ratio in obesity: a possible mechanism for leptin resistance. Lancet 348 : 159 – 161

Houseknecht KL , Mantzoros CS , Kuliawat R , Hadro E , Flier JS , Kahn BB 1996 Evidence for leptin binding to proteins in serum of rodents and humans with obesity. Diabetes 45 : 1638 – 1643

Banks WA , Kastin AJ , Huang W , Jaspan JB , Maness LM 1996 Leptin enters the brain by a saturable system independent of insulin. Peptides 17 : 305 – 311

Barash IA , Cheung CC , Weigle DS , Ren H , Kabigting EB , Kuijper JL , Clifton DK , Steiner RA 1996 Leptin is a metabolic signal to the reproductive system. Endocrinology 137 : 3144 – 3147

Levin N , Nelson C , Gurney A , Vandlen R , de Sauvage F 1996 Decreased food intake does not completely account for adiposity reduction after ob protein infusion. Proc Natl Acad Sci USA 93 : 1726 – 1730

Collins S , Kuhn CM , Petro AE , Swick AG , Chrunyk BA , Surwit RS 1996 Role of leptin in fat regulation. Nature 380 : 677

Licinio J , Mantzoros C , Negrao AB , Cizza G , Wong M-L , Bongiorno PB , Chrousos GP , Karp B , Allen C , Flier JS , Gold PW 1997 Human leptin levels are pulsatile and inversely related to pituitary-adrenal function. Nat Med 3 : 575 – 579

Harris RBS , Zhou J , Weigle DS , Kuijper JL 1997 Recombinant leptin exchanges between parabiosed mice but does not reach equilibrium Am J Physiol 272 : R1800 – R1808

Rybkin II , Zhou Y , Voulaufova J , Smagin GN , Ryan DH , Harris RBS 1997 The effect of restraint stress on food intake and body weight is determined by time of day Am J Physiol 273 : R1612 – R1622

Youngblood BD , Zhou J , Smagin GN , Ryan DH , Harris RBS 1997 Sleep deprivation by the flower pot technique and spatial reference memory. Physiol Behav 61 : 249 – 256

Lo S , Russel JC , Taylor AW 1970 Determination of glycogen in small tissue samples J Appl Physiol l28 : 234 – 236

Srere PA 1969 Citrate synthase. Methods Enzymol 13 : 3 – 5

Askew EW , Dohm GL , Huston RL 1975 Fatty acid and ketone body metabolism in the rat: response to diet and exercise. J Nutr 105 : 1422 – 1432

Uyeda F , Racker E 1965 Regulatory mechanisms in carbohydrate metabolism. VII. Hexokinse and phosphofructokinase. J Biol Chem 240 : 4682 – 4688

Littell RC , Milliken GA , Stroup WW , Wolfinger RD 1996 SAS System for Mixed Models . SAS Institute , Cary

Google Preview

Maffei M , Halaas J , Ravussin E , Pratley RE , Lee GH , Zhang Y , Fei H , Kim S , Lallone R , Ranganathan S 1995 Leptin levels in human and rodent: measurement of plasma leptin and ob RNA in obese and weight-reduced subjects. Nat Med 1 : 1155 – 1161

Considine RV , Sinha MK , Heiman ML , Kriauciunas A , Stephens TW , Nyce MR , Ohannesian JP , Marco CC , McKee LJ , Bauer TL 1996 Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N Engl J Med 334 : 292 – 295

Heek MV , Compton DS , France CF , Tedesco RP , Fawzi AB , Graziano MP , Sybertz EJ , Strader DC , Davis Jr HR 1997 Diet-induced obese mice develop peripheral, but not central, resistance to leptin. J Clin Invest 99 : 385 – 390

Murakami T , Yamashita T , Iida M , Kuwajima M 1997 A short form of leptin receptor performs signal transduction. Biochem Biophys Res Commun 231 : 26 – 29

White DW , Kuropatwinski KK , Devos R , Baumann H , Tartaglia LA 1997 Leptin receptor (OB-R) signaling. Cytoplasmic domain mutational analysis and evidence for receptor homo-oligomerization. J Biol Chem 272 : 4065 – 4071

Wang MY , Zhou YT , Newgard CB , Unger RH 1996 A novel leptin receptor isoform in rat. FEBS Lett 392 : 87 – 90

Emilsson V , Liu Y-L , Cawthorne MA , Morton NM , Davenport M 1997 Expression of the functional leptin receptor mRNA in pancreatic islets and direct inhibitory action of leptin on insulin secretion. Diabetes 46 : 313 – 316

Cohen B , Novick D , Rubinstein M 1996 Modulation of insulin activities by leptin. Science 272 : 1185 – 1188

Muller G , Ertl J , Gerl M , Preibisch G 1997 Leptin impairs metabolic actions of insulin in isolated rat adipocytes. J Biol Chem 272 : 10585 – 10593

Cusin I , Rohner-Jeanrenaud F , Stricker-Krongrad A , Jeanrenaud B 1996 The weight-reducing effect of an intracerebroventricular bolus injection of leptin in genetically obese fa/fa rats. Reduced sensitivity compared with lean animals. Diabetes 45 : 1446 – 1450

Sinha MK , Ohannesian JP , Heiman ML , Kriauciunas A , Stephens TW , Magosin S , Marco C , Caro JF 1996 Nocturnal rise of leptin in lean, obese, and non-insulin-dependent diabetes mellitus subjects. J Clin Invest 97 : 1344 – 1347

Hotta K , Gustafson TA , Ortmeyer HK , Bodkin NL , Nicolson MA , Hansen BC 1996 Regulation of obese (ob) mRNA and plasma leptin levels in rhesus monkeys. Effects of insulin, body weight, and non-insulin-dependent diabetes mellitus. J Biol Chem 271 : 25327 – 25331

Paquot N , Schneiter P , Jequier E , Tappy L 1995 Effects of glucocorticoids and sympathomimetic agents on basal and insulin-stimulated glucose metabolism. Clin Physiol 15 : 231 – 240

Foster DO , Frydman ML 1978 Nonshivering thermogenesis in the rat. II. Measurements of blood flow with microspheres point to brown adipose tissue as the dominant site of the calorigenesis induced by noradrenaline. Can J Physiol Pharmacol 56 : 110 – 122

Mackowiak PA , Boulant JA 1996 Fever’s glass ceiling. Clin Infect Dis 22 : 525 – 536

Lennie TA , McCarthy DO , Keesey RE 1995 Body energy status and the metabolic response to acute inflammation Am J Physiol 269 : R1024 – R1031

Gainsford T , Wilson TA , Metcalf D , Handman E , MacFarlane C , Ng A , Nicola NA , Alexander WA , Hilton DJ 1996 Leptin can induce proliferation, differentiation and functional activation of hemopoietic cells. Proc Natl Acad Sci USA 93 : 14564 – 14568

Johnston JL , Romsos DR , Bergen WG 1986 Reduced brain norepinephrine metabolism in obese ( ob/ob ) mice is not normalized by tyrosine supplementation. J Nutr 116 : 435 – 445

Gill K , Amit ZA 1987 Effects of serotonin uptake blockade on food, water and ethanol consumption in rats. Alcohol Clin Exp Res 11 : 444 – 449

Dumont C , Laurent J , Grandam A , Bossier JR 1981 Anorectic properties of a new long acting serotonin uptake inhibitor. Life Sci 28 : 1939 – 1945

Schwartz MW , Baskin DG , Bukowski TR , Kuijper JL , Foster D , Lasser G , Prunkard DE , Porte Jr D , Woods SC , Seeley RJ , Weigle DS 1996 Specificity of leptin action on elevated blood glucose levels and hypothalamic neuropeptide Y gene expression in ob/ob mice. Diabetes 45 : 531 – 535

Erickson JC , Clegg KE , Palmiter RD 1996 Sensitivity to leptin and susceptibility to seizures of mice lacking neuropeptide Y. Nature 381 : 415 – 421

Email alerts

Related articles in pubmed, citing articles via.

• About the Endocrine Society
• Recommend to Your Librarian
• Advertising and Corporate Services
• Journals Career Network

Affiliations

• Online ISSN 1945-7170
• Print ISSN 0013-7227
• Copyright © 2024 Endocrine Society
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings
• My Bibliography
• Collections
• Citation manager

Save citation to file

Email citation, add to collections.

• Create a new collection
• Add to an existing collection

Add to My Bibliography

Your saved search, create a file for external citation management software, your rss feed.

• Search in PubMed
• Search in NLM Catalog
• Add to Search

Determination of the half-life of circulating leptin in the mouse

Affiliations.

• 1 Columbia University Institute of Human Nutrition, New York, NY, USA.
• 2 Columbia University Department of Pediatrics, Division of Molecular Genetics, New York, NY, USA.
• 3 Naomi Berrie Diabetes Center, New York, NY, USA.
• 4 New York Obesity Research Center, New York, NY, USA.
• PMID: 28025576
• PMCID: PMC5340585
• DOI: 10.1038/ijo.2016.238

Background: The adipokine hormone, leptin, is a major component of body weight homeostasis. Numerous studies have been performed administering recombinant mouse leptin as an experimental reagent; however, the half-life of circulating leptin following exogenous administration of recombinant mouse leptin has not been carefully evaluated.

Methods: Exogenous leptin was administered (3 mg leptin per kg body weight) to 10-week-old fasted non-obese male mice and plasma was serially collected at seven time points; plasma leptin concentration was measured by enzyme-linked immunosorbent assay at each time point to estimate the circulating half-life of mouse leptin.

Results: Under the physiological circumstances tested, the half-life of mouse leptin was 40.2 (±2.2) min. Circulating leptin concentrations up to 1 h following exogenous leptin administration were 170-fold higher than endogenous levels at fasting.

Conclusions: The half-life of mouse leptin was determined to be 40.2 min. These results should be useful in planning and interpreting experiments employing exogenous leptin. The unphysiological elevations in circulating leptin resulting from widely used dosing regimens for exogenous leptin are likely to confound inferences regarding some aspects of the hormone's clinical biology.

PubMed Disclaimer

Conflict of interest statement

Conflict of Interest: The authors have no conflicts of interest.

Figure 1. The half life of mouse…

Figure 1. The half life of mouse recombinant leptin is (40.2 +/−2.2) minutes

Figure 2. Commonly used recombinant leptin injection…

Figure 2. Commonly used recombinant leptin injection concentrations are supraphysiological

Plasma leptin concentrations were measured…

Figure 3. Body weight/compositions, blood glucose, food…

Figure 3. Body weight/compositions, blood glucose, food intake, and plasma leptin concentrations of wild type…

Similar articles

• The effects of age on uptake, serum half-life and bioavailability of mouse [D-Leu-4]OB3, a synthetic peptide amide with leptin-like activity, in male C57BL/6J mice. Lee DW, Leinung MC, Grasso P. Lee DW, et al. Regul Pept. 2008 Oct 9;150(1-3):62-5. doi: 10.1016/j.regpep.2008.03.008. Epub 2008 Mar 26. Regul Pept. 2008. PMID: 18455249
• Pharmacokinetics of human leptin in mice and rhesus monkeys. Ahrén B, Baldwin RM, Havel PJ. Ahrén B, et al. Int J Obes Relat Metab Disord. 2000 Dec;24(12):1579-85. doi: 10.1038/sj.ijo.0801447. Int J Obes Relat Metab Disord. 2000. PMID: 11126209
• Intranasal administration of mouse [D-Leu-4]OB3, a synthetic peptide amide with leptin-like activity, enhances total uptake and bioavailability in Swiss Webster mice when compared to intraperitoneal, subcutaneous, and intramuscular delivery systems. Novakovic ZM, Leinung MC, Lee DW, Grasso P. Novakovic ZM, et al. Regul Pept. 2009 Apr 10;154(1-3):107-11. doi: 10.1016/j.regpep.2009.01.002. Epub 2009 Jan 22. Regul Pept. 2009. PMID: 19344673
• PASylation of Murine Leptin Leads to Extended Plasma Half-Life and Enhanced in Vivo Efficacy. Morath V, Bolze F, Schlapschy M, Schneider S, Sedlmayer F, Seyfarth K, Klingenspor M, Skerra A. Morath V, et al. Mol Pharm. 2015 May 4;12(5):1431-42. doi: 10.1021/mp5007147. Epub 2015 Apr 10. Mol Pharm. 2015. PMID: 25811325
• Leptin corrects host defense defects after acute starvation in murine pneumococcal pneumonia. Mancuso P, Huffnagle GB, Olszewski MA, Phipps J, Peters-Golden M. Mancuso P, et al. Am J Respir Crit Care Med. 2006 Jan 15;173(2):212-8. doi: 10.1164/rccm.200506-909OC. Epub 2005 Oct 6. Am J Respir Crit Care Med. 2006. PMID: 16210671
• Cholesterol reduction by immunization with a PCSK9 mimic. Zhang B, Chuang GY, Biju A, Biner D, Cheng J, Wang Y, Bao S, Chao CW, Lei H, Liu T, Nazzari AF, Yang Y, Zhou T, Chen SJ, Chen X, Kong WP, Ou L, Parchment DK, Sarfo EK, SiMa H, Todd JP, Wang S, Woodward RA, Cheng C, Rawi R, Mascola JR, Kwong PD. Zhang B, et al. Cell Rep. 2024 Jun 25;43(6):114285. doi: 10.1016/j.celrep.2024.114285. Epub 2024 May 30. Cell Rep. 2024. PMID: 38819987 Free PMC article.
• Size matters: the biochemical logic of ligand type in endocrine crosstalk. Lone JB, Long JZ, Svensson KJ. Lone JB, et al. Life Metab. 2024 Feb;3(1):load048. doi: 10.1093/lifemeta/load048. Epub 2023 Dec 8. Life Metab. 2024. PMID: 38425548 Free PMC article.
• Receptor-mediated drug delivery of bispecific therapeutic antibodies through the blood-brain barrier. Pardridge WM. Pardridge WM. Front Drug Deliv. 2023;3:1227816. doi: 10.3389/fddev.2023.1227816. Epub 2023 Jul 10. Front Drug Deliv. 2023. PMID: 37583474 Free PMC article.
• Rules for body fat interventions based on an operating point mechanism. Bar A, Karin O, Mayo A, Ben-Zvi D, Alon U. Bar A, et al. iScience. 2023 Jan 25;26(2):106047. doi: 10.1016/j.isci.2023.106047. eCollection 2023 Feb 17. iScience. 2023. PMID: 36818281 Free PMC article.
• TET3 epigenetically controls feeding and stress response behaviors via AGRP neurons. Xie D, Stutz B, Li F, Chen F, Lv H, Sestan-Pesa M, Catarino J, Gu J, Zhao H, Stoddard CE, Carmichael GG, Shanabrough M, Taylor HS, Liu ZW, Gao XB, Horvath TL, Huang Y. Xie D, et al. J Clin Invest. 2022 Oct 3;132(19):e162365. doi: 10.1172/JCI162365. J Clin Invest. 2022. PMID: 36189793 Free PMC article.
• Myers MG, Leibel RL. Lessons from Rodent Models of Obesity. In: De Groot LJB-PP, Chrousos G, et al., editors. Endotext. South Dartmouth: MDText.com , Inc.; 2015.
• Rosenbaum M, Leibel RL. 20 years of leptin: role of leptin in energy homeostasis in humans. J Endocrinol. 2014;223(1):T83–96. - PMC - PubMed
• Montague CT, Farooqi IS, Whitehead JP, Soos MA, Rau H, Wareham NJ, et al. Congenital leptin deficiency is associated with severe early-onset obesity in humans. Nature. 1997;387:903–8. - PubMed
• Clement K, Vaisse C, Lahlou N, Cabrolk S, Pelloux V, Cassuto D, et al. A mutation in the human leptin receptor gene causes obesity and pituitary dysfunction. Nature. 1998;392:398–401. - PubMed
• Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM. Positional cloning of the mouse obese gene and its human homolog. Nature. 1994;372:425–32. - PubMed

Publication types

• Search in MeSH

Related information

• PubChem Compound (MeSH Keyword)

Grants and funding

• P30 DK026687/DK/NIDDK NIH HHS/United States
• R01 DK052431/DK/NIDDK NIH HHS/United States

LinkOut - more resources

Full text sources.

• Europe PubMed Central
• Nature Publishing Group
• Ovid Technologies, Inc.
• PubMed Central

Other Literature Sources

• scite Smart Citations

• Citation Manager

NCBI Literature Resources

MeSH PMC Bookshelf Disclaimer

The PubMed wordmark and PubMed logo are registered trademarks of the U.S. Department of Health and Human Services (HHS). Unauthorized use of these marks is strictly prohibited.

• Research Article
• Neuroscience

POMC neurons expressing leptin receptors coordinate metabolic responses to fasting via suppression of leptin levels

• Alexandre Caron
• Heather M Dungan Lemko
• Carlos M Castorena
• Teppei Fujikawa
• Caleb C Lord
• Newaz Ahmed
• Charlotte E Lee
• William L Holland

• University of Texas Southwestern Medical Center, United States ;
• Howard Community College, United States ;
• UT Health San Antonio, United States ;
• Open access
• Copyright information

Share this article

Cite this article.

• Joel K Elmquist
• Copy to clipboard
• Download BibTeX
• Download .RIS

Introduction

Materials and methods, article and author information.

Leptin is critical for energy balance, glucose homeostasis, and for metabolic and neuroendocrine adaptations to starvation. A prevalent model predicts that leptin’s actions are mediated through pro-opiomelanocortin (POMC) neurons that express leptin receptors (LEPRs). However, previous studies have used prenatal genetic manipulations, which may be subject to developmental compensation. Here, we tested the direct contribution of POMC neurons expressing LEPRs in regulating energy balance, glucose homeostasis and leptin secretion during fasting using a spatiotemporally controlled Lepr expression mouse model. We report a dissociation between leptin’s effects on glucose homeostasis versus energy balance in POMC neurons. We show that these neurons are dispensable for regulating food intake, but are required for coordinating hepatic glucose production and for the fasting-induced fall in leptin levels, independent of changes in fat mass. We also identify a role for sympathetic nervous system regulation of the inhibitory adrenergic receptor (ADRA2A) in regulating leptin production. Collectively, our findings highlight a previously unrecognized role of POMC neurons in regulating leptin levels.

Pro-opiomelanocortin (POMC) neurons of the arcuate nucleus of the hypothalamus (ARC) are critical regulators of energy balance and glucose homeostasis ( Mercer et al., 2013 ; Gautron et al., 2015 ). These neurons consist of a heterogeneous population with respect to neurotransmitters used and the receptors expressed ( Hentges et al., 2009 ; Williams et al., 2010 ; Lam et al., 2017 ). Electrophysiology and immunohistochemistry studies have established that ~30% of hypothalamic POMC neurons are responsive to leptin ( Cheung et al., 1997 ; Ernst et al., 2009 ; Williams et al., 2010 ). Given the role of POMC neurons and leptin in metabolism, a conventional model indicates that a subset of POMC cells that expresses the leptin receptor (LEPR) are mediating the metabolic actions of leptin ( Cheung et al., 1997 ; Balthasar et al., 2005 ). This idea was supported by early observations that prenatal manipulations of LEPR-expressing POMC neurons mildly affect body weight ( Münzberg et al., 2003 ; Balthasar et al., 2004 ; Huo et al., 2009 ; Berglund et al., 2012 ; Huang et al., 2012 ; Mercer et al., 2013 ). However, POMC neurons share developmental origins with other cell types, including subsets of NPY/AgRP neurons ( Padilla et al., 2010 ; Lam et al., 2017 ). As such, it is possible that developmental compensation, or Lepr deletion from non-POMC neurons, are behind the phenotypes observed with conventional transgenic models ( Bouret et al., 2004 ; Lam et al., 2017 ). In addition, although it was repeatedly suggested that leptin’s anorexigenic effects act through non-ARC POMC neurons ( Myers et al., 2009 ; Berglund et al., 2012 ; Berglund et al., 2013 ), the direct contribution of LEPR-expressing POMC neurons on glucose homeostasis has been difficult to dissect due to inevitable alterations of fat mass resulting from prenatal deletions. As such, dissociating the pathways involved in leptin’s and melanocortin’s effects on adiposity versus glucose homeostasis is key for the development of anti-obesity and anti-diabetes therapies.

The activity and expression of POMC is highly dependent on energy status ( Mizuno et al., 1998 ). During obesity, there is an energy surplus and POMC levels are elevated ( Schwartz et al., 1997 ; Cowley et al., 2001 ). Inversely, during a state of negative energy balance, such as fasting, POMC expression is decreased ( Mizuno et al., 1998 ). Because POMC deficiency causes severe obesity, tremendous efforts have been made to understand a causative role of the POMC neurons in the pathophysiology of both syndromic and diet-induced obesity ( Krude et al., 1998 ; Enriori et al., 2007 ). However, relatively little is known about the function of these neurons in the context of low energy levels, despite early suggestions that the effect of fasting to reduce POMC is physiologically relevant ( Mizuno et al., 1998 ). In addition, fasting leads to a rapid fall in circulating leptin levels that is out of proportion to the loss in fat mass ( Becker et al., 1995 ; Moinat et al., 1995 ; Saladin et al., 1995 ; Ahima et al., 1996 ; Flier, 1998 ; Ahima et al., 1999 ). Despite early suggestions that the fall in leptin represent a central physiologic response to fasting required for metabolic adaptations to low energy states, the mechanisms behind fasting-induced reductions in leptin are unknown ( Ahima et al., 1996 ; Flier, 1998 ; Ahima et al., 1999 ; Flier and Maratos-Flier, 2017 ). Paradoxically, LEPR-null animals do not experience a decrease in leptin levels with fasting, suggesting that LEPRs themselves are required for the starvation-induced fall in leptin ( Hardie et al., 1996 ). Together, these observations indicate that neurons expressing LEPRs might play a role in repressing plasma leptin levels during starvation. However, the actual contribution of LEPR-expressing POMC neurons in regulating leptin secretion is unknown.

One way the CNS may regulate leptin is through altering activity of adrenergic receptors expressed by adipocytes. Acute activation of the sympathetic nervous system reduces leptin gene expression and leptin production through a β3-adrenoceptor (ADRB3)-dependent mechanism ( Moinat et al., 1995 ; Gettys et al., 1996 ; Giacobino, 1996 ; Mantzoros et al., 1996 ; Trayhurn et al., 1996 ; Deng et al., 1997 ; Trayhurn et al., 1998 ; Caron et al., 2018 ). In addition, forcing the expression of human α2-adrenoreceptor (ADRA2) in mouse adipose tissue results in elevated leptin ( Valet et al., 2000 ), suggesting that the ADRA2/ADRB3 balance in adipocytes is critical for leptin regulation. These observations suggest that leptin could regulate its own expression through a negative feedback loop from the brain to the adipose tissue. However, the central pathways and the mechanisms underlying these actions are yet to be fully characterized.

Here, we report that a subset of POMC neurons that express LEPRs directly control glucose homeostasis and are necessary to regulate leptin synthesis, independent of changes in fat mass. We used a tamoxifen-inducible Pomc CreERt2 transgenic mouse model to generate mice in which Lepr expression is spatiotemporally-controlled in a neuron-specific fashion. Within one week of deleting LEPRs from POMC neurons in adult mice, hepatic glucose production was impaired, while body weight, food intake, and energy expenditure were unaltered. In addition, mice with adult deletion of LEPRs in POMC neurons showed an impairment in the fasting-induced fall in leptin levels. We also identified an important role for adipose tissue ADRA2A in regulating leptin synthesis. Our results support a model predicting that LEPR-expressing POMC neurons coordinate metabolic responses to fasting via suppression of leptin levels.

LEPR-expressing POMC neurons are required for normal liver insulin sensitivity in adult mice

The use of conventional prenatal Pomc Cre models was key in deciphering the contribution of many receptors and pathways in glucose and energy homeostasis ( Hill et al., 2010 ;  Xu et al., 2010 ;  Berglund et al., 2012 ; Caron et al., 2016 ). However, it is now appreciated that prenatal manipulations may lead to compensatory events during development ( Padilla et al., 2010 ; Bouret et al., 2004 )). Importantly, there is a subpopulation of cells that express Pomc Cre during development, but do not express POMC in adults ( Padilla et al., 2010 ). To circumvent these issues, we used a tamoxifen-inducible Pomc CreERt2 transgenic mouse model ( Berglund et al., 2013 ) to generate Pomc CreERt2 :: Lepr flox/flox mice in which Lepr expression is spatiotemporally controlled in a neuron-specific fashion. We first assessed the impact of adult deletion of LEPR-expressing POMC neurons on glucose homeostasis. Fed and fasting glycemia were not different before, or one week after, injection of tamoxifen, indicating that the drug per se, did not impair glucose levels ( Figure 1A ). However, adult ablation of LEPRs from POMC neurons resulted in significantly higher fasting glycemia as early as two weeks post-deletion, while fed glycemia was greater at three weeks ( Figure 1A ). This effect was sustained for the entire experimental period. Fed and fasting insulin and glucagon levels were not different between groups ( Figure 1B–C ). Although no changes in glycemia were detectable in the first week, insulin response was already substantially impaired, as assessed by an insulin tolerance test ( Figure 1D–E ). We did not observe any difference in glycemia following a glucagon stimulation test ( Figure 1—figure supplement 2 ).

LEPR-expressing POMC neurons are required for normal liver insulin sensitivity in adult mice.

( A ) Fed and fasting (16 hr) glucose one week before, and every week for four weeks after, Pomc CreERt2 :: Lepr flox/flox and littermate controls were injected with the last dose of tamoxifen (n = 12). ( B ) Fed and fasting (48 hr) insulin four weeks after tamoxifen was given (n = 4–6). ( C ) Fed and fasting (48 hr) glucagon four weeks after tamoxifen was given (n = 4–6). ( D ) Glucose excursion during an insulin tolerance test (ITT) only one week following the last injection of tamoxifen (n = 5–6). ( E ) Area under the curve for the ITT shown in B (n = 5–6). ( F ) Glucose infusion rate (GIR) needed to maintain euglycemia (119.3 ± 3.9 vs 122.0 ± 8.2 mg/dl) during an hyperinsulinemic-euglycemic clamp performed only one week following the last injection of tamoxifen (n = 6). ( G ) Glucose disposal (Rd) during the same hyperinsulinemic-euglycemic clamp (n = 6). ( H ) Basal and clamped hepatic glucose production (HPG) (n = 6). ( I ) Basal and clamped lipolysis rate as assessed by measuring free fatty acid (FFA) levels (n = 6). The data are expressed as the mean ± SEM. ***p<0.001, **p<0.01 and *p<0.05 versus littermate controls.

We further explored the impact of deleting LEPRs in adult POMC neurons on systemic glucose metabolism by performing hypersulinemic-euglycemic clamp assays one week after the deletion in an independent cohort of animals. The glucose infusion rate needed to maintain euglycemia (119.3 ± 3.9 vs 122.0 ± 8.2 mg/dl) was significantly decreased in knock-out animals ( Figure 1F ), further demonstrating whole-body insulin resistance. Importantly, glucose disposal was unaltered, but insulin-induced suppression of hepatic glucose production was drastically impaired in the clamped state ( Figure 1G–H ). Moreover, the ability of insulin to suppress lipolysis during the clamped state was unaltered, suggesting that insulin resistance occurred specifically in the liver ( Figure 1I ). Deletion of LEPRs in POMC neurons in adult mice did not affect fed or fasting levels of NEFA and triglycerides (data not shown), again suggesting that impaired liver insulin sensitivity, but presumably not impaired insulin secretion, contributes to systemic insulin resistance. Altogether, these data demonstrate that LEPR-expressing POMC neurons directly regulate liver metabolism in adult mice. This is in agreement with previous findings ( Hill et al., 2010 ; Berglund et al., 2012 ) in which LEPRs were deleted during development. We found that insulin resistance can be detected one week post-deletion ( Figure 1D–I ), however blood glucose levels did not rise until two weeks post-deletion ( Figure 1A ). These findings suggest that deletion of LEPRs in adult POMC neurons impairs liver insulin sensitivity, and the resulting hepatic insulin resistance leads to the development of hyperglycemia.

LEPR-expressing POMC neurons are dispensable for the regulation of energy balance in adult mice

It has generally been assumed that LEPR-expressing POMC neurons are important for feeding and weight regulation ( Cheung et al., 1997 ; Balthasar et al., 2005 ), despite evidence that other subsets of POMC neurons are more likely to regulate energy balance ( Huo et al., 2009 ;  Berglund et al., 2013 ). Because prenatal deletion of Lepr in POMC neurons impairs body weight and fat mass, the direct contribution of these neurons in regulating glucose homeostasis has always been hard to dissect. Here, we show that deleting LEPRs from POMC neurons in adult mice does not affect body weight or body composition ( Figure 2A–C ). Four weeks following the deletion, we evaluated food intake and energy expenditure using metabolic cages. We observed that food intake was unchanged in mice lacking LEPRs in adult POMC neurons ( Figure 2D ). Moreover, oxygen consumption, respiratory exchange ratio (VCO 2 /VO 2 ) and physical activity were all unaltered ( Figure 2E–G ). These results suggest that LEPR-expressing POMC neurons regulate liver insulin sensitivity independently of changes in body weight.

LEPR-expressing POMC neurons are dispensable for the regulation of energy balance in adult mice.

( A ) Body weight before, and up to four weeks after, Pomc CreERt2 :: Lepr flox/flox and littermate controls were injected with tamoxifen (n = 12). ( B ) Fat mass and C) Lean mass as assessed by nuclear magnetic resonance (NMR) four weeks following tamoxifen administration (n = 12). ( D ) Daily food intake, ( E ) Oxygen consumption (VO 2 ), ( F ) Respiratory exchange ratio (RER), and G) locomotor activity in CaloSys Calorimetry System cages four weeks after the administration of tamoxifen (n = 5). Summary graphs showing average data for light (ZT0-ZT12) and dark (ZT12-ZD24) cycles are presented under each diurnal graph. The data are expressed as the mean ± SEM.

Fasting reduces Pomc mRNA expression in the ARC ( Mizuno et al., 1998 ), and this reduction contributes to the promotion of hunger ( Mercer et al., 2013 ). We found that adult deletions of LEPRs in POMC neurons did not affect fed or fasting levels of Pomc mRNA ( Figure 3A ). Another population of hypothalamic neurons that regulate energy balance and glucose homeostasis are the orexigenic neuropeptide Y (NPY)/agouti related peptide (AgRP) neurons ( Schwartz et al., 2000 ; Morton et al., 2006 ). During fasting, the activity of these neurons increases, which promotes food-seeking and eating behaviors ( Takahashi and Cone, 2005 ). Moreover, leptin inhibits NPY/AgRP neurons and fasting relieves this inhibition ( Schwartz et al., 1996 ). Interestingly, mice with adult deletions of LEPRs in POMC neurons had blunted mRNA levels of Npy and Agrp in response to starvation ( Figure 3B–C ). This suggests that despite normal food intake in unrestrained conditions ( Figure 2 ), fasting-induced hyperphagia might be impaired in mice lacking LEPR in adult POMC neurons. However, we found that mice consumed the same amount of food when access to laboratory chow was restored after a 48 hr fast ( Figure 4A ). Interestingly, feeding-induced hyperglycemia was higher in mice lacking LEPRs in adult POMC neurons ( Figure 4B ). Together, these results reinforce the idea that LEPR-expressing POMC neurons are dispensable for the regulation of energy balance in adult mice. Moreover, these data further demonstrate impaired glucose homeostasis when LEPRs are deleted from adult POMC neurons. At this point, it remains unclear whether manipulating LEPR-expressing POMC neurons results in dysfunction of NPY/AgRP neurons or if the receptors themselves are critical for the fasting response.

Deletion of LEPRs in adult POMC neurons impairs fasting-induced expression of orexigenic neuropeptides in the mediobasal hypothalamus.

( A ) Pomc , ( B ) Agrp and ( C ) Npy mRNA expression in mediobasal hypothalamus of fed and fasted (48 hr) Pomc CreERt2 :: Lepr flox/flox and littermate control mice four weeks after tamoxifen was given (n = 8–14). The data are expressed as the mean ± SEM. ***p<0.001 versus littermate controls.

Deletion of LEPRs in adult POMC neurons impairs postprandial glycemia.

( A ) Food intake and ( B ) Blood glucose up to six hours after food access was restored to 48 hr fasted Pomc CreERt2 :: Lepr flox/flox and littermate control mice, four weeks after tamoxifen was given (n = 8). The data are expressed as the mean ± SEM. **p<0.01 and *p<0.05 versus littermate controls.

LEPR-expressing POMC neurons are required for the fasting-induced fall in leptin levels independent of changes in fat mass in adult mice

Fasting leads to a rapid fall in circulating leptin levels, despite no initial changes in fat mass ( Becker et al., 1995 ; Moinat et al., 1995 ; Saladin et al., 1995 ; Ahima et al., 1996 ; Flier, 1998 ; Ahima et al., 1999 ). However, this regulation in leptin levels is blunted in LEPR-null animals ( Hardie et al., 1996 ), suggesting that LEPRs per se are required for the starvation-induced fall in leptin. In order to better understand the potential contribution of LEPR-expressing POMC neurons in regulating leptin production, we compared the impact of 48 hr of fasting in prenatal and adult models. In contrast to prenatal deletions ( Figure 5A–B ), deleting LEPRs in POMC neurons in adult mice did not affect fasting-induced decreases in body weight or fat-mass loss ( Figure 5E–F ). Consistent with previous reports ( Moinat et al., 1995 ; Trayhurn et al., 1995 ; Ahima et al., 1996 ; Hardie et al., 1996 ), fasting induced a robust fall in both circulating leptin and visceral adipose Lep mRNA levels in wild-type littermate controls ( Figure 5C–D and G–H ). Strikingly, this effect was prevented in mice with either prenatal ( Figure 5C–D ) or adult ( Figure 5G–H ) deletions of LEPRs in POMC neurons. Although modest, expression of Lep in visceral adipose tissue was significantly higher, in fed mice lacking LEPRs in adult POMC neurons ( Figure 5D and H ), suggesting that the deletion may affect leptin regulation even in the fed state. Collectively, these results indicate that LEPR-expressing POMC neurons are required for the starvation-induced fall in leptin, independent of changes in fat mass. Preventing fasting-induced falls in leptin might explain the blunted response observed in Agrp and Npy expression ( Figure 3B–C ).

LEPR-expressing POMC neurons are required for the fasting-induced fall in leptin levels, independent of changes in fat.

( A ) Weight loss, and ( B ) fat-mass loss after a 48 hr fast in mice with constitutive (prenatal) deletion of LEPRs in POMC neurons and littermate controls (n = 7–10). ( C ) Plasma leptin levels, and D) visceral adipose tissue Lep mRNA expression in fed or fasted (48 hr) mice with constitutive deletion of LEPRs in POMC neurons and littermate controls (n = 7–14). ( E ) Weight loss, and ( F ) fat-mass loss after a 48 hr fast in Pomc CreERt2 :: Lepr flox/flox and littermate control mice four weeks after tamoxifen was given (n = 12). ( G ) Plasma leptin levels, and ( H ) visceral adipose tissue Lep mRNA expression in fed or fasted (48 hr) in Pomc CreERt2 :: Lepr flox/flox and littermate control mice four weeks after tamoxifen was given (n = 6–13). The data are expressed as the mean ± SEM. ***p<0.001, **p<0.01 and *p<0.05 versus littermate controls.

Gi-coupled alpha-2A adrenergic receptors (ADRA2A) regulate leptin synthesis

Given that adult deletions of LEPRs in POMC neurons are sufficient to prevent the fasting-induced fall in circulating leptin levels, we next sought to determine how these neurons regulate leptin production in adipose tissue. One way the CNS may negatively regulate leptin is through the activation of ADRB3 ( Moinat et al., 1995 ; Collins and Surwit, 1996 ; Gettys et al., 1996 ; Giacobino, 1996 ; Mantzoros et al., 1996 ; Trayhurn et al., 1996 ; Trayhurn et al., 1998 ; Evans et al., 1999 ). In addition, overexpression of ADRA2 in mouse adipose tissue increases leptin levels ( Valet et al., 2000 ), suggesting that the ADRA2/ADRB3 balance in adipocytes is critical for regulation of leptin. We first investigated the expression of the nine identified adrenergic receptors in visceral adipose tissue ( Figure 6A ). The expression of most of the adrenergic receptors was unchanged in mice with adult deletions of LEPRs in POMC neurons compared to wild-type littermates. However, the fasting-induced decrease in Adra2a mRNA expression was not only prevented, but reversed following the deletion of LEPRs in adult POMC neurons ( Figure 6A ). Using an independent cohort, we found that this observation was not only reproducible, but also specific to visceral adipose tissue ( Figure 6B–C ). This result is in line with the fact that visceral but not subcutaneous adipose tissue is the primary source of leptin in rodents ( Trayhurn et al., 1995 ). These findings suggest that ADRA2A may be a candidate for mediating the starvation-induced fall in leptin.

Deletion of LEPRs in adult POMC neurons impairs visceral adipose tissue expression of Adra2a with fasting.

( A ) Expression of the nine adrenergic receptors in fed of fasted (24 hr – 48 hr) Pomc CreERt2 :: Lepr flox/flox and littermate control mice four weeks after tamoxifen was given (n = 8–14). ( B ) Comparison of the expression of Adra2a and ( C ) Adrb3 in epidydimal (eWAT) versus inguinal (iWAT) adipose tissue in an independent cohort of fed of fasted (48 hr) Pomc CreERt2 :: Lepr flox/flox and littermate control mice four weeks after tamoxifen was given (n = 5–6). The data are expressed as the mean ± SEM. ***p<0.001 and *p<0.05 versus littermate controls.

The function of ADRA2A in adipocyte physiology and pathophysiology is well known ( Lafontan and Berlan, 1995 ; Garg et al., 2016 ). However, its role in leptin synthesis has never been investigated. To functionally validate a role for ADRA2 in regulating leptin expression and production, C57BL/6J mice were intraperitoneally injected with the ADRA2 agonist clonidine, and visceral adipose tissue was collected 1 hr later. Strikingly, clonidine increased Lep mRNA expression by six fold ( Figure 7A ). In another cohort of C57BL/6J mice, we also observed that clonidine rapidly increased plasma leptin levels ( Figure 7B ). We next sought to evaluate whether clonidine treatment altered leptin production in mice with adult deletions of LEPRs in POMC neurons. Because clonidine affects every ADRA2, including those express in the CNS, we performed the experiment using adipose tissue explants from mice that were fed or fasted for 48 hr prior to the euthanasia. In fed animals, we found higher leptin release in knock-out animals ( Figure 7C ), consistent with the higher expression of Lep mRNA observed in visceral adipose tissue ( Figure 5D,H ). Furthermore, in the fasted condition, clonidine was effective at inducing leptin release only in adipose tissue explants from mice with LEPRs deleted in adult POMC neurons ( Figure 7C ). These explant studies indicate that this effect is adipose tissue-autonomous and not mediated through central effects. These results are in line with the observation that Adra2a mRNA expression increases with fasting in visceral adipose tissue of knock-out animals ( Figure 6A ). Clonidine was ineffective in subcutaneous adipose tissue ( Figure 7D ), again suggesting that the regulation of leptin production is specific to visceral fat. Together, these results suggest a role for ADRA2 as critical regulator of both leptin expression and production. In addition, these data suggest that ablation of LEPRs in adult POMC neurons prevents the starvation-induced fall in leptin by increasing ADRA2A activity in visceral white adipose tissue.

Pharmacological activation of ADRA2 stimulates leptin production.

( A ) Visceral adipose tissue Lep mRNA expression one hour following an intraperitoneal (1 mg/kg) injection of the ADRA2 agonist clonidine (n = 10–12). ( B ) Plasma leptin levels up to two hours following the administration of clonidine in an independent cohort (n = 4–7). ( C ) Leptin release from epidydimal (eWAT) and ( D ) inguinal (iWAT) adipose tissue explants from fed and fasted (48 hr) Pomc CreERt2 :: Lepr flox/flox and littermate control mice following the addition of clonidine (1 µM) (n = 6). This experiment was performed four weeks after tamoxifen was given. The data are expressed as the mean ± SEM. ***p<0.001, **p<0.01, and *p<0.05 versus littermate controls.

Leptin signaling in POMC neurons has been predicted to be key in regulating energy balance and glucose homeostasis ( Münzberg et al., 2003 ; Balthasar et al., 2004 ; Kievit et al., 2006 ; Huo et al., 2009 ; Berglund et al., 2012 ; Huang et al., 2012 ; Mercer et al., 2013 ). Our current findings dissociate the effects of LEPR-expressing POMC neurons on glucose homeostasis and changes in energy balance. In addition, our results suggest that POMC neurons are key regulators of leptin levels. This is interesting as one of the questions in leptin biology is the mechanism behind starvation-induced falls in leptin ( Friedman, 2016 ; Beshel et al., 2017 ). Although it may appear paradoxical that a subset of LEPR-expressing POMC cells controls leptin synthesis, previous studies have suggested that LEPRs are required for the starvation-induced fall in leptin ( MacDougald et al., 1995 ; Hardie et al., 1996 ; Commins et al., 2000 ). This supports previous models that falling leptin is required to activate neuroendocrine responses ( Ahima et al., 1996 ; Ahima et al., 1999 ). We also identify a role for ADRA2A in regulating leptin levels during starvation. This is in agreement with a report in which expression of human ADRA2A in adipocytes resulted in elevated leptin levels ( Valet et al., 2000 ). Collectively, our study highlights a previously unrecognized role of POMC neurons in the regulation of leptin levels and provides a new framework for the understanding of leptin action and regulation in the context of changing states of energy balance.

The current study highlights the ongoing importance of developing more refined transgenic tools, including adult-inducible models. Here, we used a tamoxifen-inducible Pomc CreERt2 transgenic mouse model to generate mice in which Lepr expression is spatiotemporally controlled in a neuron-specific fashion. Recent findings have demonstrated a need for the development of such a tool. First, the central melanocortin pathways are developmentally plastic, and as such compensations might affect the resulting phenotype, inherently limiting the conclusions that can be drawn ( Bouret et al., 2004 ; Wu et al., 2009 ; Padilla et al., 2010 ; Bouret et al., 2012 ; Wu et al., 2012 ). In addition, POMC neurons share developmental origin with other cell types, including their NPY/AgRP counterparts ( Padilla et al., 2010 ). For instance, over 25% of POMC-positive neurons were shown to express high levels of Agrp ( Lam et al., 2017 ). Likewise, we recently developed an Agrp CreERt2 transgenic mouse model to better study the role of AgRP neurons in ghrelin response ( Wang et al., 2014 ). These inducible tools will allow us to revisit fundamental beliefs about the central melanocortin system.

The canonical effect of leptin action in the brain is to regulate energy balance ( Millington, 2007 ; Mercer et al., 2013 ). Despite early evidence that ablating LEPRs only in POMC neurons results in moderate changes in body weight ( Balthasar et al., 2004 ), leptin action on POMC neurons in the ARC is considered a prototypical site of action in the control of food intake and energy expenditure. We and others have previously proposed that leptin directly acts on POMC neurons to regulate glucose homeostasis ( Huo et al., 2009 ; Berglund et al., 2012 ). There is also evidence that subpopulations of POMC neurons that do not express LEPRs may regulate food intake ( Xu et al., 2008 ; Williams et al., 2010 ; Berglund et al., 2013 ; Campbell et al., 2017 ). It is also possible that the mild obesity observed in previous studies is the consequence of Lepr deletion from a proportion of AgRP neurons. Our data indicate that the effects of leptin on energy balance are not through direct actions on POMC neurons.

Here we show that action of leptin on POMC neurons regulates glucose homeostasis independent of its effects on energy balance. Specifically, removing LEPRs from POMC neurons in adult mice resulted in insulin resistance and impaired hepatic glucose production within one week following deletion. This was followed by sustained hyperglycemia, independent of changes in insulin and glucagon levels, in glucose disposal, or in the ability of insulin to suppress lipolysis. Although food intake was unaltered both in ad libitum or refeeding conditions, postprandial glycemia was impaired in mice lacking LEPRs in adult POMC neurons. Together, this suggests that altering leptin signaling in POMC neurons results in rapid-onset hepatic insulin resistance ( Brown and Goldstein, 2008 ). This specific effect is consistent with many reports showing direct consequences in the liver following genetic manipulations in POMC neurons ( Hill et al., 2010 ; Xu et al., 2010 ; Berglund et al., 2012 ; Berglund et al., 2013 ; Shi et al., 2013 ; Williams et al., 2014 ; Caron et al., 2016 ). It was also recently shown that POMC neurons are important for hepatic parasympathic nerve activity in response to leptin ( Bell et al., 2018 ). A recent study also stresses the importance of insulin signaling in POMC neurons in regulating adipose tissue lipolysis and the development of liver steatosis ( Shin et al., 2017 ). However, whether POMC neurons regulate glucose and lipid hepatic metabolism directly through the autonomic nervous system, or indirectly by altering metabolic hormone requires further investigation. It is nevertheless clear from our study that LEPR-expressing POMC neurons play a pivotal role in liver metabolism, independently of changes in energy balance.

Our data also highlight an unexpected role for LEPR-expressing POMC neurons in regulating the fasting-induced fall in leptin. We show that the ability of fasting to suppress leptin is impaired in transgenic mouse models with either prenatal or adult deletion of LEPRs in POMC neurons. Although there is a general consensus that leptin levels are tightly correlated to adiposity ( Frederich et al., 1995 ; Considine et al., 1996 ), our data suggest that this fasting-dependent regulation is independent of changes in body weight or fat mass. Moreover, this effect appears specific to visceral adipose tissue, which is in line with the fact that leptin is predominantly secreted from visceral white adipocytes in rodents ( Trayhurn et al., 1995 ).

However, one important question still remains. In a particular, how does LEPR signalling in POMC neurons regulate adipocyte leptin secretion during fasting? One speculation is that the deletion of leptin receptors reduces POMC activity and renders the neurons less effective at activating downstream targets. Another possibility is more provocative. In particular, we propose that LEPR-expressing POMC neurons are part of a regulatory loop that is important for adaptative responses to fasting. Fasting rapidly alters key metabolic signals and decreases the circulating peripheral hormones (such as insulin) which are required to maintain normal leptin levels ( Saladin et al., 1995 ; D'souza et al., 2017 ). These changes are all sensed by POMC neurons. However, drops in leptin trigger neuroendocrine responses that promote survival, including the inhibition of the sexual and thyroid axes and activation of the stress axis ( Ahima et al., 1996 ; Ahima et al., 1999 ). These survival responses are extreme and safeguards may have evolved to ensure that they are not initiated too quickly. LEPR-expressing POMC neurons might represent such a ‘gatekeeper’ to control the inhibition of leptin production. Thus, removing LEPRs from POMC neurons would prevent their ability to sense small fluctuations in leptin levels ultimately blunting the ability to fully suppress leptin levels.

In support of this model, we observed that fasting-induced expression of Npy and Agrp in the mediobasal hypothalamus was impaired in Pomc CreERt2 :: Lepr flox/flox mice, suggesting that the falling leptin might be an important signal activating NPY/AgRP neurons during starvation ( Bi et al., 2003 ). Although this impaired response might be a direct consequence of elevated fasting-leptin levels, we did not observe any differences in food intake. Importantly, this does not invalidate the role of these neurons in regulating re-feeding behavior after a fast. However, these results indicate that preventing the normal fall in leptin levels during fasting have major repercussions, not only on the neuroendocrine system ( Ahima et al., 1996 ), but also on behavioral, metabolic and neuronal responses.

Mechanistically, we show that visceral adipose tissue expression of Adra2a , which normally decreases with fasting, is actually increasing in mice lacking fasted mice lacking LEPRs in adult POMC neurons. Interestingly, the expression of Adra2a is not altered in subcutaneuous adipose tissue, further supporting visceral-dependent effect. It is noteworthy that the sympathetic regulation differs between different fat depots, both in terms of innervation and outflow ( Brito et al., 2007 ; Brito et al., 2008 ; Nguyen et al., 2017 ). These findings also add another layer of complexity to the way the brain regulates peripheral tissues through the activation of GPCRs. Our pharmacological experiments also support the notion that ADRA2 are important for leptin regulation. ADRB3 is well-known to negatively regulate leptin though a cAMP-dependent mechanism ( Moinat et al., 1995 ; Gettys et al., 1996 ; Giacobino, 1996 ; Mantzoros et al., 1996 ; Slieker et al., 1996 ; Trayhurn et al., 1996 ; Deng et al., 1997 ; Trayhurn et al., 1998 ; Caron et al., 2018 ). Because ADRB3 is Gs-coupled, we hypothesize that Gi-coupled ADRA2 might have the opposite action on leptin synthesis. Treating mice with an ADRA2 agonist is sufficient to increase both circulating leptin and mRNA levels in visceral fat. We also found that this regulation is tissue-autonomous, as clonidine effectively affected leptin release only in visceral adipose tissue explants from mice lacking LEPRs in adult POMC neurons. From a translational point of view, the observation that ADRA2A activation stimulates leptin production is meaningful. Human adipocytes express high levels of ADRA2A but few or no ADRB3, while murine adipocytes show high levels of ADRB3 and very low number of ADRA2 ( Lafontan and Berlan, 1993 ; Lafontan and Berlan, 1995 ). By creating mice that have a human-like pattern of adrenoreceptors, researchers previously established that the ADRA2/ADRB2 balance in adipocytes is critical for regulating fat mass ( Valet et al., 2000 ). Increasing the ADRA2/ADRB3 balance in adipose tissue resulted in increased circulating levels of leptin, suggesting that this balance is also important for regulating leptin production. However, because these mice were obese, the direct contribution of the ADRA2/ADRB3 balance was hard to define. Here, we show that despite no changes in body weight, the ADRA2/ADRB3 balance in adipocyte is still important for leptin regulation.

In conclusion, our study indicates that a subset POMC neurons that express LEPRs directly controls glucose homeostasis and is necessary to control leptin synthesis, independently of changes in fat mass. We also identified an important role for adipose tissue ADRA2A in regulating leptin synthesis. From a conceptual standpoint, our results predict that leptin regulates its own expression through a negative feedback loop between POMC neurons and adipose tissue.

Animal work described in this manuscript has been approved and conducted under the oversight of the UT Southwestern Institutional Animal Care and Use Committee (IACUC). Male mice were housed at an ambient temperature of 23 ± 1˚C and maintained on a 12 hr light/dark cycle (lights on 0700–1900) and fed with normal mouse chow diet (Harlan, Teklad Global 16% Protein Rodent Diet 2016; 12% kcal from fat, 3 kcal/g).

Pomc Cre (RRID: IMSR_JAX:005965 ) mice ( Balthasar et al., 2004 ) and Pomc CreERt2 (RRID:MGI:5569339) mice ( Berglund et al., 2013 ) were crossed with Lepr flox/flox (RRID: MGI:3511747 ) mice ( McMinn et al., 2004 ) to generate mice with constitutive deletion of LEPRs in POMC neurons ( Pomc Cre :: Lepr flox/flox ) and adult deletion of LEPRs in POMC neurons ( Pomc CreERt2 :: Lepr flox/flox ) respectively. Mice were maintained on a C57Bl/6J (RRID: IMSR_JAX:000664 ) background at UT Southwestern Medical Center. Adult ablation was induced by tamoxifen. Tamoxifen (0.15 mg/g; Sigma-Aldrich, T5648) was suspended in corn oil (Sigma-Aldrich, C8267) and was administered intraperitoneally (three injections every 48 hr for 5 days) to 10–12 week-old Pomc CreERt2 :: Lepr flox/flox and Pomc CreERt2 :: Lepr +/+ (littermate control) mice. Fasting experiments were performed from 0800 to 0800 (48 h) or from 1600 to 0800 (16 h). The efficiency of the recombination following tamoxifen was performed by crossing Pomc CreERt2 mice with Ai14(RCL-tdT)-D mice (RRID: IMSR_JAX:007914 ) mice. Validation of the mouse model is presented in Figure 1—figure supplement 1 .

Immunohistochemistry and validation of the inducible mice

Immunohistochemistry was performed to visualize phospho-Stat3 (Tyr705, Cell Signaling Technology Cat# 9131, RRID: AB_331586 ), β-endorphin (Phoenix Pharmaceuticals Cat# H-022–33, RRID: AB_2314007 ), as well as the fluorescent reporter tdTomato (Santa Cruz Biotechnology Cat# sc-33354, RRID: AB_639922 ) in the brain and pituitary ( Scott et al., 2009 ; Williams et al., 2010 ). For leptin-induced Stat3 activation experiments, mice were fasted for 16 hr (1600 to 0800) and injected i.p. with mouse recombinant leptin (5 mg/kg; National Hormone and Peptide Program, AFP1783). Mice were anesthetized 45 min later using an i.p. injection of chloral hydrate (350 mg/kg) and then perfused transcardially with 0.9% saline followed by 10% neutral buffered formalin.

Assessment of insulin sensitivity and glucose levels

Blood samples were collected from the tail vein and glucose was measured using a glucometer (Bayer’s Contour Blood Glucose Monitoring System; Leverkusen, Germany). For insulin tolerance test (ITT), mice were fasted for 4 hr and then administered insulin by intraperitoneal injection (0.75 U/kg body weight, human insulin, Eli Lilly).

Hyperinsulinemic-euglycemic clamps

Hyperinsulinemic-euglycemic clamps were performed on conscious, unrestrained mice as previously described ( Holland et al., 2011 ). Euglycemia was maintained by variable infusion of 20% dextrose. Steady state was achieved 80 min after initiating hyperinsulinemia and maintained for 40 min. Additional blood samples were taken before initiating hyperinsulinemia and at the end of the clamp for analysis of insulin and free fatty acids.

Glucagon stimulation test

Glucagon stimulation test was performed in mice fasted for one hour (0800 to 0900). Briefly, human recombinant glucagon (120 µg/kg i.p.) was given and blood glucose monitored every 10 min for one hour.

Assessment of leptin, insulin and glucagon levels

Blood was collected in EDTA tubes. Plasma was isolated by centrifugation (4000 g x 10 min at 4°C) and was stored at −80°C for further biochemical analyses. Plasma leptin (Mouse/Rat Leptin ELISA, ALPCO, 22-LEPMS-E01), insulin (Mouse Ultrasensitive Insulin ELISA, ALPCO, 80-INSMSU-E01), and glucagon (Mercodia Glucagon ELISA, 10-1281-01) were measured following manufacturer recommendations.

Assessment of body composition

Fat mass and lean mass were assessed by nuclear magnetic resonance (NMR) spectroscopy using a nuclear magnetic resonance (NMR) spectroscopy (Bruker Minispec mq10 NMR 0.23T/10MHz).

Metabolic cages studies

A combined indirect calorimetry system (CaloSys Calorimetry System, TSE Systems Inc.) was used for all metabolic studies. Experimental animals were acclimated for 5 days in a metabolic chamber with food and water. Oxygen consumption (VO2), carbon dioxide production (VCO2), respiratory exchange ration (RER) and food intake were measured after acclimation. Locomotion was measured using a multi-dimensional infrared light beam system.

Quantitative real-time PCR

Total mRNA was isolated from visceral (epidymal) and subcutaneous (inguinal) white adipose tissues using the RNeasy Lipid Tissue Mini Kit (Qiagen, 74104). Total mRNA was isolated from liver using RNA STAT-60 reagent (Tel-Test, Inc). The RNA concentrations were estimated from absorbance at 260 nm. cDNA synthesis was performed using a High Capacity cDNA Kit (Applied Biosystems). mRNA extraction and cDNA synthesis were performed following the manufacturer’s instructions. cDNA was diluted in DNase-free water before quantification by real-time PCR. mRNA transcript levels were measured in duplicate samples using a ABI 7900 HT Sequence Detection System (Applied Biosystems). The relative amounts of all mRNAs were calculated using the ΔΔCT assay. Primers for 18 s (Hs99999901_s1), Adra1a (Mm00442668_m1), Adra1b (Mm00431685_m1), Adra1d (Mm01328600_m1), Adra2a (Mm00845383_s1), Adra2b (Mm00477390_s1), Adra2c (Mm00431686_s1), Adrb1 (Mm00431701_s1), Adrb2 (Mm02524224_s1), Adrb3 (Mm02601819_g1), Agrp (Mm00475829_g1), Lep (Mm00434759_m1), Npy (Mm00445771_m1) and Pomc (Mm00435874_m1) were purchased from Applied Biosystems.

Pharmacological activation of ADRA2 in vivo

The ADRA2 agonist clonidine hydrochloride (Sigma-Aldrich, St. Louis, MO, US; C7897) was administered intraperitoneally (1 mg/kg) to 10–12 week-old C57BL/6J mice following 4 hr of fasting. Two independent cohorts were used to evaluate Lep RNA expression and circulating leptin levels.

Ex vivo leptin release assay

Pomc CreERt2 :: Lepr flox/flox and Pomc CreERt2 :: Lepr +/+ (littermate control) mice were fasted for 48 hr and ~10–20 mg of visceral (epidydimal) and subcutaneous (inguinal) white adipose tissues were cultured in 96 wells plate containing 0.200 ml of Krebs-Ringer Bicarbonate Buffer containing 5 mM glucose and 4% fatty acid-free BSA, as described ( Caviglia et al., 2011 ). Tissues were subsequently treated either with or without 1 µM clonidine hydrochloride (Sigma-Aldrich, St. Louis, MO, US; C7897) for basal and clonidine conditions respectively, and leptin release was measured by ELISA and corrected to tissue weight.

Statistical analysis

Data are expressed as mean ± SEM. Comparison between two experimental conditions were analyzed by Student’s unpaired t test. Two-way ANOVA followed by Bonferroni post hoc test was used to compare more than two experimental conditions. All statistical tests were performed using GraphPad Prism (version 7.0), and p<0.05 was considered statistically significant.

• Elmquist JK
• Google Scholar
• Prabakaran D
• Mantzoros C
• Maratos-Flier E
• Balthasar N
• McGovern RA
• Dalgaard LT
• Funahashi H
• Christiansen LM
• Edelstein E
• Aschkenasi C
• Brichard SM
• Berglund ED
• Williams KW
• Robinson BM
• Bartness TJ
• Goldstein JL
• Campbell JN
• Lyubetskaya A
• Verstegen AMJ
• McCarroll SA
• Mouchiroud M
• Caviglia JM
• Brasaemle DL
• Considine RV
• Kriauciunas A
• Stephens TW
• Ohannesian JP
• Rubinstein M
• Assimacopoulos-Jeannet F
• Giacobino JP
• Tonelli-Lemos L
• Wunderlich CM
• Kleinridders A
• Kloppenburg P
• Wunderlich FT
• Frederich RC
• Harkness PJ
• Otero-Corchon V
• Richardson JA
• Rutkowski JM
• Brozinick JT
• Birnbaum MJ
• Zabolotny JM
• Yoshimura A
• Biebermann H
• Polex-Wolf J
• Nicole Kohnke S
• Rimmington D
• O'Rahilly S
• MacDougald OA
• Mantzoros CS
• Dragatsis I
• Millington GW
• Kleopoulos SP
• Cummings DE
• Schwartz MW
• Hollenberg AN
• Leinninger GM
• Guerre-Millo M
• Campfield LA
• Sternson SM
• Friedman JM
• Bue-Valleskey J
• Takahashi KA
• Zingaretti MC
• Osborne-Lawrence S
• Margatho LO
• Palmiter RD
• Anderson JG

Author details

Contribution, contributed equally with, competing interests.

• Division of Hypothalamic Research, Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, United States
• Department of Neuroscience, University of Texas Southwestern Medical Center, Dallas, United States
• Department of Pharmacology, University of Texas Southwestern Medical Center, Dallas, United States

For correspondence

National institute of diabetes and digestive and kidney diseases (r37dk053301), canadian diabetes association (nod_pf-3-15-4756-ac), american heart association (14sdg17950008), national institute of diabetes and digestive and kidney diseases (k01dk11164401), national institute of diabetes and digestive and kidney diseases (r01dk114036), american heart association (16sdg27260001).

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank the Mouse Metabolic Phenotyping Core at UT Southwestern Medical Center at Dallas. This work was supported by the NIH (R37DK053301 to JKE, R01DK114036 to CL, K01DK11164401 to CMC) and by the American Heart Association (14SDG17950008 to TF, 16SDG27260001 to CL). AC is a Canadian Diabetes Association fellow.

Animal experimentation: Animal work described in this manuscript has been approved and conducted under the oversight of the UT Southwestern Institutional Animal Care and Use Committee (IACUC, APN 2015-101301 and APN 2015-101263).

© 2018, Caron et al.

This article is distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use and redistribution provided that the original author and source are credited.

• 4,759 views
• 705 downloads
• 79 citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Download links

Downloads (link to download the article as pdf).

• Article PDF
• Figures PDF

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools), categories and tags.

• Glucose homeostasis

Research organism

Be the first to read new articles from elife.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Malays J Med Sci
• v.8(2); 2001 Jul

The Unfolding Tale of Leptin

The discovery of leptin, a product of the ob gene, in 1994 ( 1 ) has raised hopes of many an obese population throughout the world for a probable solution to their fatness. Its discovery has seemingly provided the missing link in the regulation of body weight and energy balance and has added an extra impetus to research in obesity. Within a short period of time, research on leptin has spread to numerous laboratories throughout the world and beyond obesity, so much so it is now becoming increasingly difficult trying to keep pace with the rate at which new knowledge on the physiology of leptin is being generated, let alone making sense of the data that is being published.

As has been the case with a lot of other discoveries, the discovery and identification of leptin, for some reason, took a long time in coming. It began way back in 1950 with the appearance of a recessive mutant colony of house mice kept in captivity. These mice, which have now been found to lack the ob gene ( ob/ob ), had hyperphagia, decreased energy expenditure and early onset obesity ( 2 ). In 1953, the lipostatic theory of weight control was proposed ( 3 ) where it was hypothesized that a circulating factor produced by the adipocytes interacted with the hypothalamus to regulate food intake, body weight and overall long-term energy balance. The ingenious experiments of Hervey ( 4 ) in 1958, in which he produced obesity in rats after lesioning the ventromedial hypothalamus, and parabiotic experiments of Hausberger ( 5 ) a year later confirmed the probable presence of this circulating factor. However it was some 40 years later that this factor, which we now refer to as leptin, was eventually detected and characterised. Through positional cloning, the mouse ob gene was discovered and has been shown to encode a 4.5 kilobase mRNA transcript with a highly conserved 167 amino acid chain ( 1 ). The mouse and human ob genes have been localized to chromosome 6 and 7q31.3 respectively ( 1 , 6 , 7 ).

It is now some six years since the discovery of leptin and the obese saddled with fat however are still awaiting delivery of that promised “fat-melting” hormone. The euphoria and the enthusiasm that was raised by the discovery of leptin has now settled and an aura of realism is beginning to prevail. Whilst the discovery of leptin has in a way provided a logical direction to research in obesity, there is still a long way to go before we can say for certain how significant the discovery of leptin has been in the understanding and control of obesity. From early on it became evident that leptin deficiency may not by itself totally explain obesity. Whilst parabiosis between ob/ob mice and normal mice leads to weight reduction in the former ( 5 ), similar experiments between normal mice and another obese strain of mice ( db/db ) leads to starvation and death in the normal mice ( 8 ). Harris et al ( 9 ) found the same with fa/fa obese rats and their normal counterparts. Evidently these obese mice and rats had, in their blood, a surfeit of an appetite-inhibiting circulating factor, presumably leptin, that it starved to death the normal mice and rats when their circulations were connected respectively. The fact that it was not affecting the obese mice or rats themselves indicated that there was present an element of resistance in these animals. These observations when interpreted on the basis of what we know about leptin today, suggest that obesity may not just be due to the absence or deficiency of leptin per se but due more to the presence of leptin resistance. That this may indeed be the case is further indicated by observations that in the general population plasma leptin levels correlate positively with fat mass i.e. the higher the fat mass the higher the plasma leptin levels ( 10 ). Moreover, a change in leptin responsiveness in a model of diet-induced obesity in the mouse, where 5–10 times more leptin was required to achieve the equivalent weight loss of that produced in the ob/ob mouse ( 11 ), substantiates further this possibility.

To date two leptin receptor isoforms have been identified i.e. the long form receptor ( obRb ) which has a 340 amino acid intracellular domain and the short form receptor ( obRa ) with a 34 amino acid intracellular domain. The long form leptin receptor belongs to the class I cytokine family and its mRNAs have been identified in the arcuate nucleus, dorsomedial nucleus, ventromedial nucleus and ventral premamillary nuclei. The short form leptin receptor mRNAs have been identified in the choroid plexus, vascular endothelium, liver, lung, gonads and kidney. The short form receptors are involved in the transport of leptin across membranes, particularly the blood brain barrier and the kidney. Incidently, leptin is mainly cleared in the kidney and hence in chronic renal failure leptin levels rise and probably contribute to the loss of appetite and weight loss in these patients. The biological effects of leptin however are thought to result from its binding to the long form receptors followed by the activation of a JAK-STAT signaling pathway (Janus Kinase - signal transducer and activator of transcription) which then alter the expression in the target cell.

In the regulation of energy balance, leptin first crosses the blood brain barrier aided by its binding to the short form receptor, and then binds to the long form receptors in the hypothalamic nuclei. In the hypothalamus, the long-form of leptin receptor is co-expressed with neurons producing Neuropeptide Y (NPY), Agouti-related peptide (AgRP), proopiomelanocortin (POMC) and Cocaine and amphetamine regulated transcript (CART). NPY and AgRP both stimulate feeding behaviour whereas POMC and CART suppress feeding. Leptin binding to the NPY and AgRP neurons suppresses NPY and AgRP release whereas its binding to POMC and CART containing neurons increases the release of POMC (precursor of a MSH) and CART. The net effect is a suppression of appetite and feeding, increased autonomic activity and thermogenesis. (For a more detailed review readers are referred to ref 12 , 13 )

In obesity, in addition to leptin deficiency, leptin resistance is hypothesized to occur at various levels in the leptin pathway. It may be a result of (a) the presence of leptin antibodies, or (b) increased leptin binding proteins, or (c) defective transporter system at the blood brain barrier, or (d) defective receptors, or (e) defective intracellular signaling, or even (f) resetting of leptinstat in the hypothalamus. Whilst naturally occurring antibodies or increased leptin binding proteins in obesity have not been demonstrated so far, there is however evidence for the absence of leptin in some obese mice ( 2 ) and in a rare form of human obesity ( 14 ), mutations in the leptin receptor ( 15 ), impaired leptin signal transduction ( 16 , 17 ) and even a possibility of decreased leptin transport across the blood brain barrier ( 18 ). Apart from leptin deficiency, which is inherited as an autosomal recessive trait, it is uncertain whether the rest of the abnormalities observed are present at birth or occur later on in life. Clearly, there is still a need of much work to fully understand the nature of these abnormalities and how we could overcome them.

Although there has been disappointment that leptin deficiency was not the answer to the common obesity, since its discovery however, leptin has very quickly become a subject of study beyond obesity. Numerous possible neuroendocrine roles for leptin are being explored and the adipose tissue is seemingly functioning as part of the endocrine system.

Leptin is believed to permissively activate the hypothalamic-pituitary-gonadal axis during puberty ( 19 , 20 , 21 , 22 ) as mutations of ob and db genes result in hypothalamic hypogonadism in humans ( 23 ). In normal children leptin levels increase before puberty and reach their peak at the onset of puberty ( 24 ) after which they begin to decline in boys but continue to increase in girls with the levels depending upon the fat mass. Administration of leptin to pre-pubertal mice and non-human primates accelerates puberty ( 25 ). Besides, leptin stimulates GnRH release from hypothalamic explants ( 26 ). Low leptin and absent diurnal leptin rhythm occur with exercise-induced amenorrhea ( 27 ). On the other hand, there is also evidence that LH and FSH responses to GnRH administration in young girls correlate negatively with body mass index and circulating leptin ( 28 ). These observations however do not exclude the permissive role for leptin in puberty as chronic hyperleptinaemia in transgenic skinny mice accelerates puberty and late onset hypothalamic hypogonadism ( 29 ) suggesting dual effects of leptin. Leptin in the required concentrations stimulates puberty but in excessive amounts leads to infertility.

Leptin is now considered an important link between obesity and infertility. It affects reproduction and fertility not only through its action on the hypothalamic-pituitary-gonadal axis but also through direct ovarian action. Follicular fluid leptin concentration has been shown to be a promising marker of assisted reproduction treatment success in normal women. The role for leptin in the follicular fluid is unclear but it has been reported that a lower follicular leptin concentration favoured a successful outcome of assisted reproduction in women with polycystic ovarion syndrome ( 30 ). In addition, IGF-1 augmentation of FSH-stimulated estradiol production by the ovarion granulosa cells is inhibited by leptin ( 31 ) and this may be one of the possible mechanism of infertility in obese women. A functional deficiency in the long form leptin receptor in the endometrium has also been linked to subfertility in some women ( 32 ). Increased leptin levels in serum and peritoneal fluid of patients with pelvic endometriosis have also been reported ( 33 ).

Leptin is now known to be produced in the placenta ( 34 ) where it may serve an autocrine/paracine role in human implantation and placentation ( 35 ). Matrix metalloproteinases are necessary during cytotrophoblast invasion and leptin has been shown to activate these proteases ( 36 ). Impaired placentation has been repeatedly observed in preeclampsia. Interestingly however, both serum free leptin levels ( 37 ) and placenta leptin levels ( 38 ) are raised in women with preeclampsia. Placental leptin release is augmented during advanced labour but it is absent during cesarian section ( 39 ). Leptin levels in the placental tissue also correlate positively to placental weight ( 40 ). Levels of leptin in the cord blood are positively correlated with body weight and fat mass of the newborn ( 41 ). In twin pregnancy, placental and cord blood leptin levels have been found to be lower ( 42 ) or higher ( 43 ) in the growth retarded infant when compared to its normal-sized twin. The amount of leptin mRNA in placentae from insulin treated diabetic women are higher when compared to normal placentae ( 44 ). Umbilical cord plasma concentration of leptin has also been reported to be higher in infants of diabetic women ( 45 ). Although considerable information exists implicating leptin in a number of gestational situations, the information however is still insufficient and at times conflicting for a cohesive interpretation of the physiological role for leptin in reproduction. Nevertheless, some involvements are emerging and there is a clear need for more work.

A significant relationship has been reported between leptinaemia and plasma renin activity in women with essential hypertension ( 46 ). The overall effect of leptin on arterial pressure however remains unclear. Chronic systemic administration of leptin increases arterial pressure and heart rate in conscious animals ( 47 ). Intravenous bolus administration of pharmacologic doses of synthetic murine leptin to denervated SHR however causes natriuresis ( 48 ). In another study leptin has been shown to have vasodilating properties, probably mediated through endothelium dependant hyperpolarizing factor ( 49 ). In addition to these, vitrous leptin has been reported to be elevated in proliferative diabetic retinopathy and retinal detachment ( 50 ). Higher plasma leptin levels have also been observed in cases of advanced proliferative or nonproliferative diabetic retinopathy ( 51 ). Leptin has also been postulated to have thrombotic tendency ( 52 ) and has been found to stimulate the proliferation of haematopoietic stem cells ( 53 ). Leptin evidently also enhances wound re-epithelialization ( 54 ), induces an angiogenic response or has angiogenic activity ( 55 ). Apart from that, leptin is evidently also involved in linking nutritional state to T cell function, as leptin replacement reverses the immunosuppressive affects of acute starvation in mice ( 56 ). In a recent report it was observed that patients who survived an acute septic episode had three times higher levels of leptin compared to non-survivors ( 57 ). Surgical stress is associated with increased serum leptin levels ( 58 ).

Clearly the discovery of leptin has not only further stimulated research in the study of obesity but also the role of leptin in a number of other areas of physiology and medicine. Little is known about the regulation of its synthesis or release although it appears that leptin is produced constitutively and released from the adipocyte continuously. Its interaction with insulin also needs further scrutiny. The influence of age and diet on leptin sensitivity or resistance are other areas that need to be researched. For the moment however, the mounting evidence which may still be somewhat hazy, patchy and at times contradictory, suggests that leptin may have a greater endocrine role than previously envisaged. Leptin functioning as a link between the adipose tissue and energy balance, reproduction, fertility and a number of other physiological functions, suggests that it may be time to re-look at our view of adipose tissue as purely an energy store and insulating tissue. Adipose tissue an endocrine organ? No doubt it may be a wee bit too early to press for this but the possibility nevertheless exists. How many of it’s hypothesized roles will eventually be confirmed or whether it shall become another “full of promise but failed to deliver” case is left to be seen Its discovery has nevertheless brought with it a very fertile area of investigation that should keep many scientists busy over the coming decade or so.

• Consulting Editors
• For authors
• Publication ethics
• Publication alerts by email
• Advertising

Advertisement

Leptin physiology and pathophysiology: knowns and unknowns 30 years after its discovery

Jeffrey s. flier 1 and rexford s. ahima 2.

1 Department of Medicine and Neurobiology, Harvard Medical School, Boston, Massachusetts, USA.

2 Division of Endocrinology, Diabetes, and Metabolism, Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.

Address correspondence to: Jeffrey S. Flier, Division of Endocrinology, Beth Israel Deaconess Medical Center, Center for Life Sciences, 7th Floor, 330 Brookline Avenue, Boston, Massachusetts 02215, USA. Phone: 617.735.3343; Email: [email protected] .

Find articles by Flier, J. in: JCI | PubMed | Google Scholar

Find articles by Ahima, R. in: JCI | PubMed | Google Scholar

Published January 2, 2024 - More info

The cloning of the ob gene in 1994 and reports in 1995 that administration of the encoded protein leptin reversed obesity of ob/ob mice that lacked it were major breakthroughs in endocrinology and metabolism ( 1 , 2 ). The name leptin was derived from leptos, the Greek root for thin, because leptin was initially considered to be a signal from adipose tissue to the brain; rising levels of leptin acted through a negative feedback mechanism to limit obesity by reducing food intake and increasing energy expenditure.

The discovery of leptin did not occur in a vacuum. Before 1994, substantial research in rodents and humans provided evidence of homeostatically regulated responses that acted to resist body weight induced by overfeeding or underfeeding ( 3 ). Experiments employing hypothalamic lesions as well as a joining of the circulation between rodents (parabiosis) suggested that one or more circulating factors informed the brain of energy stores ( 4 ). However, the identity of such factors was unknown, as was their tissue of origin, mode of regulation, and how they effected physiologic responses.

The discovery that leptin was an adipose-derived hormone that informed brain targets of energy stores was a foundational centerpiece of modern metabolic science. But emerging leptin biology has produced many surprises, and important questions remain unanswered. This Viewpoint is an assessment of current understanding, stressing insights needed to advance the field.

Complete absence of leptin or its inability to signal, caused by loss-of-function mutations of the ligand or its receptor in both rodents and humans, respectively, produces profound obesity. Consistent with classic endocrine logic, leptin replacement reverses obesity due to leptin mutations but does not correct obesity caused by mutations in its receptor. Contrary to the initial idea of leptin being an “antiobesity hormone,” avoidance of obesity is not leptin’s dominant physiologic role. Leptin expression and circulating levels increase and reflect the degree of adiposity in diet-induced obese and several mouse obesity models, but hyperleptinemia clearly doesn’t prevent obesity ( 5 ). Hyperleptinemia proportional to obesity was also observed in humans. Thus, leptin resistance appears to be present in most cases of obesity, perhaps analogous to insulin resistance in type 2 diabetes.

Obesity treatment is an enormous unmet medical need for which leptin initially appeared a logical answer, and development of leptin for treatment of obesity was aggressively pursued. However, unlike the benefit of supplemental insulin treatment in individuals with diabetes with insulin resistance, leptin treatment had little or no effect to reduce obesity in the general population, and thus the strategy was quickly abandoned ( 6 ).

Simultaneously, another distinct physiologic role for leptin was emerging: the ability of falling leptin levels to signal the starved state to the brain. The rapid fall of leptin expression and levels with starvation was first observed in 1995 in mice ( 7 ). In 1996, falling leptin was shown to be necessary and sufficient to cause the neuroendocrine adaptation to starvation — including the suppression of reproduction ( 8 ). Leptin administration to prepubertal female mice has a permissive action in accelerating the onset of puberty, further highlighting the important linkage between leptin and neuroendocrine function, and repletion of leptin restored menstrual cycles in women with hypoleptinemic amenorrhea ( 9 , 10 ). Taken together, it is now established that leptin is an adipocyte-derived hormonal signal to the brain that drives the transition between the physiology of starvation and energy sufficiency.

So, what is leptin’s role in “common obesity”? Why is hyperleptinemia unable to prevent weight gain in most people? This topic has generated substantial confusion and, of late, has been largely ignored. We address some of the reasons here, beginning with our understanding of obesity and leptin in rodents. C57BL/6 mice fed a diet high in fat and sugar develop obesity (called diet-induced obesity [DIO]); these mice are now widely seen as the best (albeit imperfect) rodent model for common human obesity. Leptin expression and levels rise as obesity develops in DIO mice, and leptin administration has little (or no) effect to reverse obesity, suggesting “resistance” to endogenous and exogenous leptin ( 11 ).

It is important to stress that this leptin resistance in DIO mice is only partial, as obesity in these mice is much less severe than obesity in db/db mice completely lacking leptin receptor. Some leptin signals are still being sensed in DIO mice — including the brain signal to maintain reproductive competence.

What causes this partial leptin resistance? A key early approach was to search for an antagonist of leptin signaling in hypothalamic target cells after acute leptin administration. The first and best studied such molecule was suppressor of cytokine signaling 3 (SOCS3), an intracellular inhibitor of leptin signaling shown to be acutely induced by leptin in hypothalamic neurons and found to be increased in DIO mice ( 12 ). A role for SOCS3 in leptin resistance is also supported by genetic models, where haploinsufficiency of SOCS3 confers protection against DIO, as does SOCS3 knockout in leptin target neurons, such as those expressing pro-opiomelanocortin (POMC) ( 13 , 14 ). Inaccessibility of the hypothalamus has made it impossible to determine whether SOCS3 expression in key target cells influences susceptibility to obesity in humans. Although SOCS3 is an attractive target for treatment of obesity, its general inhibition would likely have adverse effects because SOCS3 inhibits signaling by many cytokines.

Whatever its intracellular mechanism, does hyperleptinemia itself cause leptin resistance? In one mouse model of DIO, hyperleptinemia was found to be required for leptin resistance, an idea further supported by evidence that lowering leptin levels in DIO mice with an anti-leptin antibody improves energy homeostasis ( 15 , 16 ). Further complexity regarding the role of leptin in energy feedback was uncovered in a model of mouse obesity caused by forced overfeeding, with leptin levels “clamped” at normal levels and unable to rise. When forced overfeeding is ended, spontaneous feeding was suppressed (reflecting negative feedback) to the same extent in mice without hyperleptinemia as in hyperleptinemic DIO mice ( 17 ). This finding suggested a signal other than leptin must have suppressed feeding in response to obesity. The identity of such a factor, its actions, and possible role in obesity pathogenesis remain unknown at present.

Importantly, sensitivity to DIO varies widely across mouse strains, with some, like male C57BL/6 mice, being highly sensitive to obesity and others being resistant ( 18 , 19 ). The role of leptin, leptin resistance, and/or other factors in accounting for genetic differences in DIO susceptibility should be an important subject of future research.

The discovery of leptin stimulated an explosive expansion of research on the neural circuits that regulate energy homeostasis in response to leptin and other factors. Neurons in the arcuate nucleus expressing POMC and agouti-related peptide (AgRP)/neuropeptide Y (NPY) respond directly to leptin. Activating and inhibitory signals from POMC and AgRP/NPY neurons, respectively, converge on downstream neurons expressing melanocortin 4 receptors, whose activation suppresses appetite ( 20 ). The complexity of central neural circuitry and how it integrates leptin and other signals is a subject of intense research.

Apart from the severe genetic disruptions of the leptin pathway that led to its discovery, we know surprisingly little 30 years later about the role of leptin action and resistance in human obesity. This knowledge gap is highlighted by comparing insights into leptin biology with knowledge about the physiology and biochemistry of insulin action in type 2 diabetes. Innumerable studies have quantified the effects of insulin infusions at varying doses on insulin action (glucose uptake, hepatic glucose production, lipolysis, etc.) and signaling in target tissues, such as fat and muscle, which led to the emergence of major insights for the field. In stark contrast, over 30 years, virtually no such studies have been performed with leptin in lean and obese humans.

Which studies might be of greatest interest? Are individuals who remain lean (and healthy) without dieting or medications a subset with exquisite sensitivity to rising leptin, consistent with the initial hypothesis that leptin serves as an antiobesity “adipostatic” signal ( Figure 1 )? Similarly, are individuals with obesity with modest hyperleptinemia a subgroup that might be responsive to leptin therapy, with appetite suppression and reduction in fat mass in response to exogenous leptin ( Figure 1 )? These important questions could be answered by infusing leptin into such individuals (and relevant controls) and measuring the effects on hunger and food ingestion, among other responses.

Patterns of leptin levels in response to changes in feeding and adiposity. Circulating leptin levels fall with starvation and rise to prior levels with refeeding. It is possible, but not yet demonstrated, that some lean individuals respond briskly to rising leptin to prevent obesity or respond to another as-yet undiscovered signal to prevent obesity. It is also possible, but not yet demonstrated, that some individuals with obesity with relatively low leptin levels for their degree of obesity might respond to exogenous leptin with weight loss. Most individuals with obesity have high leptin levels to which they are partially resistant, causing them to be unresponsive to exogenous leptin. This “leptin resistance” may be caused by desensitization induced by hyperleptinemia.

Why haven’t such obvious human experiments been conducted? There are several possible answers. Perhaps, following the failure of leptin trials, investigators and the pharmaceutical industry lost interest in the topic and moved on to more appealing and tractable questions. This disinterest might have been furthered by the lack of easily quantifiable leptin responses, compared with insulin, where glucose is pertinent and easily measured. Inaccessibility of critical leptin target tissues for biochemical analysis is another obvious impediment. On the other hand, despite these issues, key insights into leptin physiology and resistance and their role in obesity would have been seen as important discoveries.

Perhaps the most likely explanation for the limited progress toward understanding leptin biology in human obesity is the unavailability of leptin for human clinical investigations. Three companies developed leptin analogs for potential use in obesity, and the failure of studies of these analogs to produce sufficient clinical benefit caused these efforts to be terminated. Along the way, internal studies and analyses that might have been done were never published, and requests from investigators for the hormone were typically denied. Rights to the best-studied analog, metreleptin, were passed to progressively smaller companies. Today, the sole use of leptin is for treatment of exceptionally rare cases of total leptin deficiency as well as rare lipodystrophies, where reversal of hypoleptinemia has beneficial metabolic effects ( 21 ).

Obesity is a major health crisis facing the US and other nations. Regrettably, research on one of the most important metabolic discoveries of the past 50 years has been limited by a lack of availability of recombinant leptin for human research. Scientific leaders and funders like NIH should understand the challenges and make a commitment to solving this problem. Only then might the great potential for leptin’s discovery to illuminate the pathophysiology and treatment of obesity be fulfilled.

Conflict of interest: JSF serves on the board of directors of Scholar Rock Corporation, a biopharmaceutical company, and has equity and income from that role. JSF’s spouse is an employee of Alnylam Corporation.

Copyright: © 2024, Flier et al. This is an open access article published under the terms of the Creative Commons Attribution 4.0 International License.

Reference information: J Clin Invest . 2024;134(1):e174595. https://doi.org/10.1172/JCI174595.

• Zhang Y, et al. Positional cloning of the mouse obese gene and its human homologue. Nature . 1994;372(6505):425–432. View this article via: CrossRef PubMed Google Scholar
• Halaas JL, et al. Weight-reducing effects of the plasma protein encoded by the obese gene. Science . 1995;269(5223):543–546. View this article via: CrossRef PubMed Google Scholar
• Sims EA, Horton ES. Endocrine and metabolic adaptation to obesity and starvation. Am J Clin Nutr . 1968;21(12):1455–1470. View this article via: CrossRef PubMed Google Scholar
• Hervey GR. The effects of lesions in the hypothalamus in parabiotic rats. J Physiol . 1959;145(2):336–352. View this article via: CrossRef PubMed Google Scholar
• Frederich RC, et al. Leptin levels reflect body lipid content in mice: evidence for diet-induced resistance to leptin action. Nat Med . 1995;1(12):1311–1314. View this article via: CrossRef PubMed Google Scholar
• Heymsfield SB, et al. Recombinant leptin for weight loss in obese and lean adults: a randomized, controlled, dose-escalation trial. JAMA . 1999;282(16):1568–1575. View this article via: CrossRef PubMed Google Scholar
• Frederich RC, et al. Expression of ob mRNA and its encoded protein in rodents. Impact of nutrition and obesity. J Clin Invest . 1995;96(3):1658–1663. View this article via: JCI CrossRef PubMed Google Scholar
• Ahima RS, et al. Role of leptin in the neuroendocrine response to fasting. Nature . 1996;382(6588):250–252. View this article via: CrossRef PubMed Google Scholar
• Ahima RS, et al. Leptin accelerates the onset of puberty in normal female mice. J Clin Invest . 1997;99(3):391–395. View this article via: JCI CrossRef PubMed Google Scholar
• Welt CK, et al. Recombinant human leptin in women with hypothalamic amenorrhea. N Engl J Med . 2004;351(10):987–997. View this article via: CrossRef PubMed Google Scholar
• El-Haschimi K, et al. Two defects contribute to hypothalamic leptin resistance in mice with diet-induced obesity. J Clin Invest . 2000;105(12):1827–1832. View this article via: JCI CrossRef PubMed Google Scholar
• Bjorbaek C. Identification of SOCS-3 as a potential mediator of central leptin resistance. Mol Cell . 1998;1(4):619–625. View this article via: CrossRef PubMed Google Scholar
• Howard JK, et al. Enhanced leptin sensitivity and attenuation of diet-induced obesity in mice with haploinsufficiency of Socs3. Nat Med . 2004;10(7):734–738. View this article via: CrossRef PubMed Google Scholar
• Kievit P, et al. Enhanced leptin sensitivity and improved glucose homeostasis in mice lacking suppressor of cytokine signaling-3 in POMC-expressing cells. Cell Metab . 2006;4(2):123–132. View this article via: CrossRef PubMed Google Scholar
• Knight ZA, et al. Hyperleptinemia is required for the development of leptin resistance. PLoS One . 2010;5(6):e11376. View this article via: CrossRef PubMed Google Scholar
• Zhao S, et al. Partial leptin reduction as an insulin sensitization and weight loss strategy. Cell Metab . 2019;30(4):706–719. View this article via: CrossRef PubMed Google Scholar
• Ravussin Y, et al. Evidence for a non-leptin system that defends against weight gain in overfeeding. Cell metab . 2018;28(2):289–299. View this article via: CrossRef PubMed Google Scholar
• West DB, et al. Dietary obesity in nine inbred mouse strains. Am J Physiol . 1992;262(6 pt 2):1025–1032. View this article via: PubMed CrossRef Google Scholar
• Takahashi N, et al. Divergent effects of leptin in mice susceptible or resistant to obesity. Horm Metab Res . 2002;34(11-12):691–697. View this article via: CrossRef PubMed Google Scholar
• Andermann ML, Lowell BB. Toward a wiring diagram understanding of appetite control. Neuron . 2017;95(4):757–778. View this article via: CrossRef PubMed Google Scholar
• Meehan CA, et al. Metreleptin for injection to treat the complications of leptin deficiency in patients with congenital or acquired generalized lipodystrophy. Expert Rev Clin Pharmacol . 2016;9(1):59–68. View this article via: CrossRef PubMed Google Scholar
• Version 1 (January 2, 2024): Electronic publication

Article tools

• Download citation information
• Send a comment
• Terms of use
• Standard abbreviations
• Need help? Email the journal

Related Collection:

Jci's 100th anniversary.

Solid organ transplantation: solid but not yet spectacular Laurence A. Turka

The GLP-1 journey: from discovery science to therapeutic impact Daniel J. Drucker

CAR T cells for hematological malignancies Barbara Savoldo et al.

Clinical investigation of hypoxia-inducible factors: getting there Gregg L. Semenza

Pancreatic β cell function versus insulin resistance: application of the hyperbolic law of glucose tolerance Richard N. Bergman

Nitric oxide in vascular biology: elegance in complexity Joseph Loscalzo

How the JCI’s most-cited paper sparked the field of lipoprotein research Michael S. Brown et al.

SGLT2 inhibitors: cardiorenal metabolic drugs for the ages Ralph A. DeFronzo

G protein–coupled receptors: from radioligand binding to cellular signaling Howard A. Rockman et al.

Bisphosphonates for osteoporosis: from bench to clinic Teresita Bellido

Pandemics past, present, and future: progress and persistent risks Arturo Casadevall

Major breakthroughs in hematopoietic stem cell transplantation and future challenges in clinical implementation Leslie S. Kean et al.

Life-saving effect of pulmonary surfactant in premature babies J. Usha Raj et al.

Advancing chemokine research: the molecular function of CXCL8 Yiran Hou et al.

Navigating an enigma: the continuing journey of autoimmunity discoveries Mariana J. Kaplan

Defining and targeting mechanisms of eosinophilic inflammation in a new era of severe asthma treatment Joshua A. Boyce et al.

The other pandemic: lessons from 40 years of HIV research Mary E. Klotman et al.

A half-century of VEGFA: from theory to practice Susan E. Quaggin

Two decades of advances in preeclampsia research: molecular mechanisms and translational studies S. Ananth Karumanchi

The forgotten pandemic: how understanding cholera illuminated mechanisms of chloride channels in multiple diseases Qais Al-Awqati

• Article usage
• Citations to this article
• Discovery of leptin
• Uncovering the role of leptin in obesity
• Remaining gaps in knowledge
• Version history

Sign up for email alerts

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Author Biographies
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

Cold exposure rejuvenates the metabolic phenotype of panx1 −/− mice.

1. Introduction

2. materials and methods, 2.2. body temperature and composition measurements, 2.3. oral glucose tolerance test (ogtt) and insulin tolerance test (itt), 2.4. tissue collection, 2.5. biochemical analysis of serum, 2.6. histology and image analysis, 2.7. mitochondrial isolation, 2.8. rna extraction, reverse transcription, and quantitative rt-pcr, 2.9. protein extraction and western blotting, 2.10. statistical analysis, 3. results and discussion, 3.1. panx1 deletion enhances glucose over fat metabolism in 14-week-old mice, 3.2. the metabolic phenotype of panx1 −/− mice normalizes between 14 and 20 weeks of age, 3.3. cold exposure evokes different metabolic changes in panx1 −/− mice and wt mice, 3.4. cold-induced wat morphological changes are amplified in panx1 −/− mice, 3.5. panx1 −/− mice do not increase expression of mitochondrial oxidation genes upon cold exposure, 4. conclusions, supplementary materials, author contributions, institutional review board statement, informed consent statement, data availability statement, acknowledgments, conflicts of interest.

• Rutkowski, J.M.; Stern, J.H.; Scherer, P.E. The cell biology of fat expansion. J. Cell Biol. 2015 , 208 , 501–512. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Spiegelman, B.M.; Flier, J.S. Obesity and the regulation of energy balance. Cell 2001 , 104 , 531–543. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Wu, J.; Bostrom, P.; Sparks, L.M.; Ye, L.; Choi, J.H.; Giang, A.H.; Khandekar, M.; Virtanen, K.A.; Nuutila, P.; Schaart, G.; et al. Beige adipocytes are a distinct type of thermogenic fat cell in mouse and human. Cell 2012 , 150 , 366–376. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Li, Y.; Fromme, T.; Schweizer, S.; Schottl, T.; Klingenspor, M. Taking control over intracellular fatty acid levels is essential for the analysis of thermogenic function in cultured primary brown and brite/beige adipocytes. EMBO Rep. 2014 , 15 , 1069–1076. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ye, L.; Wu, J.; Cohen, P.; Kazak, L.; Khandekar, M.J.; Jedrychowski, M.P.; Zeng, X.; Gygi, S.P.; Spiegelman, B.M. Fat cells directly sense temperature to activate thermogenesis. Proc. Natl. Acad. Sci. USA 2013 , 110 , 12480–12485. [ Google Scholar ] [ CrossRef ]
• Cheung, S.Y.; Huang, Y.; Kwan, H.Y.; Chung, H.Y.; Yao, X. Activation of transient receptor potential vanilloid 3 channel suppresses adipogenesis. Endocrinology 2015 , 156 , 2074–2086. [ Google Scholar ] [ CrossRef ]
• Zhang, L.L.; Yan Liu, D.; Ma, L.Q.; Luo, Z.D.; Cao, T.B.; Zhong, J.; Yan, Z.C.; Wang, L.J.; Zhao, Z.G.; Zhu, S.J.; et al. Activation of transient receptor potential vanilloid type-1 channel prevents adipogenesis and obesity. Circ. Res. 2007 , 100 , 1063–1070. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Rossato, M.; Granzotto, M.; Macchi, V.; Porzionato, A.; Petrelli, L.; Calcagno, A.; Vencato, J.; De Stefani, D.; Silvestrin, V.; Rizzuto, R.; et al. Human white adipocytes express the cold receptor TRPM8 which activation induces UCP1 expression, mitochondrial activation and heat production. Mol. Cell. Endocrinol. 2014 , 383 , 137–146. [ Google Scholar ] [ CrossRef ]
• Zhu, Y.; Gao, Y.; Tao, C.; Shao, M.; Zhao, S.; Huang, W.; Yao, T.; Johnson, J.A.; Liu, T.; Cypess, A.M.; et al. Connexin 43 Mediates White Adipose Tissue Beiging by Facilitating the Propagation of Sympathetic Neuronal Signals. Cell Metab. 2016 , 24 , 420–433. [ Google Scholar ] [ CrossRef ]
• Lee, S.Y.; Fontana, F.; Sugatani, T.; Portales Castillo, I.; Leanza, G.; Coler-Reilly, A.; Civitelli, R. Connexin43 in mesenchymal lineage cells regulates body adiposity and energy metabolism in mice. JCI Insight 2024 , 9 , e170016. [ Google Scholar ] [ CrossRef ]
• Adamson, S.E.; Meher, A.K.; Chiu, Y.H.; Sandilos, J.K.; Oberholtzer, N.P.; Walker, N.N.; Hargett, S.R.; Seaman, S.A.; Peirce-Cottler, S.M.; Isakson, B.E.; et al. Pannexin 1 is required for full activation of insulin-stimulated glucose uptake in adipocytes. Mol. Metab. 2015 , 4 , 610–618. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Tozzi, M.; Hansen, J.B.; Novak, I. Pannexin-1 mediated ATP release in adipocytes is sensitive to glucose and insulin and modulates lipolysis and macrophage migration. Acta Physiol. 2020 , 228 , e13360. [ Google Scholar ] [ CrossRef ]
• Esseltine, J.L.; Laird, D.W. Next-Generation Connexin and Pannexin Cell Biology. Trends Cell Biol. 2016 , 26 , 944–955. [ Google Scholar ] [ CrossRef ]
• Lee, V.R.; Barr, K.J.; Kelly, J.J.; Johnston, D.; Brown, C.F.C.; Robb, K.P.; Sayedyahossein, S.; Huang, K.; Gros, R.; Flynn, L.E.; et al. Pannexin 1 regulates adipose stromal cell differentiation and fat accumulation. Sci. Rep. 2018 , 8 , 16166. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Molica, F.; Meens, M.J.; Dubrot, J.; Ehrlich, A.; Roth, C.L.; Morel, S.; Pelli, G.; Vinet, L.; Braunersreuther, V.; Ratib, O.; et al. Pannexin1 links lymphatic function to lipid metabolism and atherosclerosis. Sci. Rep. 2017 , 7 , 13706. [ Google Scholar ] [ CrossRef ]
• Senthivinayagam, S.; Serbulea, V.; Upchurch, C.M.; Polanowska-Grabowska, R.; Mendu, S.K.; Sahu, S.; Jayaguru, P.; Aylor, K.W.; Chordia, M.D.; Steinberg, L.; et al. Adaptive thermogenesis in brown adipose tissue involves activation of pannexin-1 channels. Mol. Metab. 2021 , 44 , 101130. [ Google Scholar ] [ CrossRef ]
• Bargiotas, P.; Krenz, A.; Hormuzdi, S.G.; Ridder, D.A.; Herb, A.; Barakat, W.; Penuela, S.; von Engelhardt, J.; Monyer, H.; Schwaninger, M. Pannexins in ischemia-induced neurodegeneration. Proc. Natl. Acad. Sci. USA 2011 , 108 , 20772–20777. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Aguilar-Perez, A.; Pacheco-Costa, R.; Atkinson, E.G.; Deosthale, P.; Davis, H.M.; Essex, A.L.; Dilley, J.E.; Gomez, L.; Rupert, J.E.; Zimmers, T.A.; et al. Age- and sex-dependent role of osteocytic pannexin1 on bone and muscle mass and strength. Sci. Rep. 2019 , 9 , 13903. [ Google Scholar ] [ CrossRef ]
• Ehrlich, A.; Pelli, G.; Pick, R.; Clochard, L.; Molica, F.; Kwak, B.R. Pannexin1 deletion in lymphatic endothelium affects lymphatic function in a sex-dependent manner. Physiol. Rep. 2024 , 12 , e16170. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Freitas-Andrade, M.; Bechberger, J.F.; MacVicar, B.A.; Viau, V.; Naus, C.C. Pannexin1 knockout and blockade reduces ischemic stroke injury in female, but not in male mice. Oncotarget 2017 , 8 , 36973–36983. [ Google Scholar ] [ CrossRef ]
• Galarraga, M.; Campion, J.; Munoz-Barrutia, A.; Boque, N.; Moreno, H.; Martinez, J.A.; Milagro, F.; Ortiz-de-Solorzano, C. Adiposoft: Automated software for the analysis of white adipose tissue cellularity in histological sections. J. Lipid Res. 2012 , 53 , 2791–2796. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Taylor, D.; Gottlieb, R.A. Parkin-mediated mitophagy is downregulated in browning of white adipose tissue. Obesity 2017 , 25 , 704–712. [ Google Scholar ] [ CrossRef ]
• Morel, S.; Braunersreuther, V.; Chanson, M.; Bouis, D.; Rochemont, V.; Foglia, B.; Pelli, G.; Sutter, E.; Pinsky, D.J.; Mach, F.; et al. Endothelial Cx40 limits myocardial ischaemia/reperfusion injury in mice. Cardiovasc. Res. 2014 , 102 , 329–337. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Chevalier, C.; Stojanovic, O.; Colin, D.J.; Suarez-Zamorano, N.; Tarallo, V.; Veyrat-Durebex, C.; Rigo, D.; Fabbiano, S.; Stevanovic, A.; Hagemann, S.; et al. Gut Microbiota Orchestrates Energy Homeostasis during Cold. Cell 2015 , 163 , 1360–1374. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Kharitonenkov, A.; Shiyanova, T.L.; Koester, A.; Ford, A.M.; Micanovic, R.; Galbreath, E.J.; Sandusky, G.E.; Hammond, L.J.; Moyers, J.S.; Owens, R.A.; et al. FGF-21 as a novel metabolic regulator. J. Clin. Investig. 2005 , 115 , 1627–1635. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Choi, H.M.; Kim, H.R.; Kim, E.K.; Byun, Y.S.; Won, Y.S.; Yoon, W.K.; Kim, H.C.; Kang, J.G.; Nam, K.H. An age-dependent alteration of the respiratory exchange ratio in the db/db mouse. Lab. Anim. Res. 2015 , 31 , 1–6. [ Google Scholar ] [ CrossRef ]
• Muta, O.; Odaka, M.; Fujii, Y.; Fushimi, T.; Sato, H.; Osakabe, N. Difference in endocrine and behavior between short-term single- and paired-housing mice in metabolic cage. Neurosci. Lett. 2023 , 806 , 137246. [ Google Scholar ] [ CrossRef ]
• Charron, M.J.; Gorovits, N.; Laidlaw, J.S.; Ranalletta, M.; Katz, E.B. Use of GLUT-4 null mice to study skeletal muscle glucose uptake. Clin. Exp. Pharmacol. Physiol. 2005 , 32 , 308–313. [ Google Scholar ] [ CrossRef ]
• Graham, T.E.; Kahn, B.B. Tissue-specific alterations of glucose transport and molecular mechanisms of intertissue communication in obesity and type 2 diabetes. Horm. Metab. Res. 2007 , 39 , 717–721. [ Google Scholar ] [ CrossRef ]
• Nikami, H.; Shimizu, Y.; Endoh, D.; Yano, H.; Saito, M. Cold exposure increases glucose utilization and glucose transporter expression in brown adipose tissue. Biochem. Biophys. Res. Commun. 1992 , 185 , 1078–1082. [ Google Scholar ] [ CrossRef ]
• Shepherd, P.R.; Gnudi, L.; Tozzo, E.; Yang, H.; Leach, F.; Kahn, B.B. Adipose cell hyperplasia and enhanced glucose disposal in transgenic mice overexpressing GLUT4 selectively in adipose tissue. J. Biol. Chem. 1993 , 268 , 22243–22246. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Santoro, A.; Kahn, B.B. Adipocyte Regulation of Insulin Sensitivity and the Risk of Type 2 Diabetes. N. Engl. J. Med. 2023 , 388 , 2071–2085. [ Google Scholar ] [ CrossRef ]
• Barbatelli, G.; Murano, I.; Madsen, L.; Hao, Q.; Jimenez, M.; Kristiansen, K.; Giacobino, J.P.; De Matteis, R.; Cinti, S. The emergence of cold-induced brown adipocytes in mouse white fat depots is determined predominantly by white to brown adipocyte transdifferentiation. Am. J. Physiol. Endocrinol. Metab. 2010 , 298 , E1244–E1253. [ Google Scholar ] [ CrossRef ]
• Roth, C.L.; Molica, F.; Kwak, B.R. Browning of White Adipose Tissue as a Therapeutic Tool in the Fight against Atherosclerosis. Metabolites 2021 , 11 , 319. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Rusiecka, O.M.; Molica, F.; Nielsen, M.S.; Tollance, A.; Morel, S.; Frieden, M.; Chanson, M.; Boengler, K.; Kwak, B.R. Mitochondrial pannexin1 controls cardiac sensitivity to ischaemia/reperfusion injury. Cardiovasc. Res. 2023 , 119 , 2342–2354. [ Google Scholar ] [ CrossRef ]
• Calderon-Dominguez, M.; Sebastian, D.; Fucho, R.; Weber, M.; Mir, J.F.; Garcia-Casarrubios, E.; Obregon, M.J.; Zorzano, A.; Valverde, A.M.; Serra, D.; et al. Carnitine Palmitoyltransferase 1 Increases Lipolysis, UCP1 Protein Expression and Mitochondrial Activity in Brown Adipocytes. PLoS ONE 2016 , 11 , e0159399. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Lee, K.; Jin, H.; Chei, S.; Lee, J.Y.; Oh, H.J.; Lee, B.Y. Dietary Silk Peptide Prevents High-Fat Diet-Induced Obesity and Promotes Adipose Browning by Activating AMP-Activated Protein Kinase in Mice. Nutrients 2020 , 12 , 201. [ Google Scholar ] [ CrossRef ]
• Shin, H.; Ma, Y.; Chanturiya, T.; Cao, Q.; Wang, Y.; Kadegowda, A.K.G.; Jackson, R.; Rumore, D.; Xue, B.; Shi, H.; et al. Lipolysis in Brown Adipocytes Is Not Essential for Cold-Induced Thermogenesis in Mice. Cell Metab. 2017 , 26 , 764–777. [ Google Scholar ] [ CrossRef ]
• Berry, D.C.; Jiang, Y.; Arpke, R.W.; Close, E.L.; Uchida, A.; Reading, D.; Berglund, E.D.; Kyba, M.; Graff, J.M. Cellular Aging Contributes to Failure of Cold-Induced Beige Adipocyte Formation in Old Mice and Humans. Cell Metab. 2017 , 25 , 481. [ Google Scholar ] [ CrossRef ]
• Holman, C.D.; Sakers, A.P.; Calhoun, R.P.; Cheng, L.; Fein, E.C.; Jacobs, C.; Tsai, L.; Rosen, E.D.; Seale, P. Aging impairs cold-induced beige adipogenesis and adipocyte metabolic reprogramming. Elife 2024 , 12 , RP87756. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Silva, G.D.N.; Amato, A.A. Thermogenic adipose tissue aging: Mechanisms and implications. Front. Cell Dev. Biol. 2022 , 10 , 955612. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Saito, M. Brown adipose tissue as a regulator of energy expenditure and body fat in humans. Diabetes Metab. J. 2013 , 37 , 22–29. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Molica, F.; Quercioli, A.; Montecucco, F.; Schindler, T.H.; Kwak, B.R.; Morel, S. A Genetic Polymorphism in the Pannexin1 Gene Predisposes for The Development of Endothelial Dysfunction with Increasing BMI. Biomolecules 2020 , 10 , 208. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ikeda, K.; Kang, Q.; Yoneshiro, T.; Camporez, J.P.; Maki, H.; Homma, M.; Shinoda, K.; Chen, Y.; Lu, X.; Maretich, P.; et al. UCP1-independent signaling involving SERCA2b-mediated calcium cycling regulates beige fat thermogenesis and systemic glucose homeostasis. Nat. Med. 2017 , 23 , 1454–1465. [ Google Scholar ] [ CrossRef ]
• Kazak, L.; Chouchani, E.T.; Jedrychowski, M.P.; Erickson, B.K.; Shinoda, K.; Cohen, P.; Vetrivelan, R.; Lu, G.Z.; Laznik-Bogoslavski, D.; Hasenfuss, S.C.; et al. A creatine-driven substrate cycle enhances energy expenditure and thermogenesis in beige fat. Cell 2015 , 163 , 643–655. [ Google Scholar ] [ CrossRef ] [ PubMed ]

Click here to enlarge figure

Share and Cite

Molica, F.; Ehrlich, A.; Pelli, G.; Rusiecka, O.M.; Montessuit, C.; Chanson, M.; Kwak, B.R. Cold Exposure Rejuvenates the Metabolic Phenotype of Panx1 −/− Mice. Biomolecules 2024 , 14 , 1058. https://doi.org/10.3390/biom14091058

Molica F, Ehrlich A, Pelli G, Rusiecka OM, Montessuit C, Chanson M, Kwak BR. Cold Exposure Rejuvenates the Metabolic Phenotype of Panx1 −/− Mice. Biomolecules . 2024; 14(9):1058. https://doi.org/10.3390/biom14091058

Molica, Filippo, Avigail Ehrlich, Graziano Pelli, Olga M. Rusiecka, Christophe Montessuit, Marc Chanson, and Brenda R. Kwak. 2024. "Cold Exposure Rejuvenates the Metabolic Phenotype of Panx1 −/− Mice" Biomolecules 14, no. 9: 1058. https://doi.org/10.3390/biom14091058

Article Metrics

Article access statistics, supplementary material.

ZIP-Document (ZIP, 3545 KiB)

Further Information

Mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals