Leptin: Is It Thermogenic?

Abstract

Animals that lack the hormone leptin become grossly obese, purportedly for 2 reasons: increased food intake and decreased energy expenditure (thermogenesis). This review examines the experimental evidence for the thermogenesis component. Analysis of the data available led us to conclude that the reports indicating hypometabolism in the leptin-deficient ob/ob mice (as well as in the leptin-receptor-deficient db/db mice and fa/fa rats) derive from a misleading calculation artefact resulting from expression of energy expenditure per gram of body weight and not per intact organism. Correspondingly, the body weight-reducing effects of leptin are not augmented by enhanced thermogenesis. Congruent with this, there is no evidence that the ob/ob mouse demonstrates atrophied brown adipose tissue or diminished levels of total UCP1 mRNA or protein when the ob mutation is studied on the inbred C57BL/6 mouse background, but a reduced sympathetic nerve activity is observed. On the outbred "Aston" mouse background, brown adipose tissue atrophy is seen, but whether this is of quantitative significance for the development of obesity has not been demonstrated. We conclude that leptin is not a thermogenic hormone. Rather, leptin has effects on body temperature regulation, by opposing torpor bouts and by shifting thermoregulatory thresholds. The central pathways behind these effects are largely unexplored.

Keywords: ob/ob mouse; body temperature; brown adipose tissue; energy expenditure; leptin; thermogenesis.

Graphical Abstract

Figure 1.

The general picture of leptin action. Leptin, encoded by the ob gene, is expressed in adipocytes (mainly in white adipocytes); its expression is positively correlated with fat mass and regulated in a fasting/feeding-dependent manner. Leptin is secreted and via the circulation reaches its target neurons in hypothalamic nuclei (colored circles, here unspecified) that express the long isoform of the leptin receptor (LepRb, green). Mice carrying the ob mutation lack functional leptin; animals carrying the db mutation (mice) or fa mutation (rats) lack functional LepRb. Leptin action in the hypothalamus activates anorexigenic pathways and is discussed to also trigger mechanisms affecting brown adipose tissue activity and, mainly through this, energy expenditure. The latter two effects are still under debate and are those analyzed in the present review. In the adipocytes: blue circles, nuclei; dark-brown structures mitochondria; yellow circles, lipid droplets; in the neurons: beige circles, leptin.

Figure 2.

Are ob/ob mice hypothermic? (A) ob/ob mice have similar or even higher levels of energy expenditure than wild-type mice under thermoneutral conditions as well as under subthermoneutral conditions, at least on the C57BL/6 background. The limit of their thermoneutral zone, as indicated by the lower critical temperature (LCT), is similar in lean and ob/ob. (B) ob/ob mice have normal body temperature at thermoneutrality and defend a lower, but stable, body temperature when gradually exposed to subthermoneutral temperatures. (C, E) Differences in body temperature resulting from changes in the activation thresholds of thermoregulatory, but not behavioral, effectors. (D, F) Leptin-deficient ob/ob mice develop massive obesity because of excessive food intake, but not from reduced energy expenditure, which is similar or even higher than that in lean mice.

Figure 3.

Brown adipocyte function. Upon activation of the sympathetic nervous system following cold exposure or dietary stimuli, norepinephrine (NE) is secreted from nerve endings and binds to β-adrenoceptors (β-AR) at the plasma membrane of brown adipocytes. This triggers G_s-protein (G_s)-coupled activation of adenylyl cyclase (AC), leading to the generation of cAMP, which will in turn activate protein kinase A (PKA). PKA activates lipolysis of triglycerides (TG), leading to the release of free fatty acids (FFA) that directly activate uncoupling protein 1 (UCP1) by overcoming the purine nucleotide (ATP, GTP, ADP, GDP) inhibition of this protein. The fatty acids may also fuel thermogenesis. Nutrients from the circulation are taken up from circulating triglyceride-rich lipoproteins (TRL; i.e., VLDL and chylomicrons) via the action of lipoprotein lipase (LPL) and the fatty acid translocase “cluster of differentiation 36” (CD36) on endothelial cells (EC), which may also internalize whole TRL particles. Also albumin-bound circulating free fatty acids can be taken up by brown adipose tissue. Another important source of fuel and precursors for de novo lipogenesis is circulating glucose, which is taken up via pathways dependent on glucose transporter 1 and 4 (GLUT1/4). To nourish its exceptionally high oxidative capacity, BAT has also to take up substantial amounts of oxygen. The main product of BAT is heat, which is released and distributed in the body by the circulation.

Figure 4.

Is BAT function affected in ob/ob mice? An apparent defect in brown adipose tissue (BAT) recruitment has repeatedly been discussed to be the cause of the apparent cold intolerance and low energy expenditure. (A) However, despite higher tissue weight, which is mainly due to lipid accumulation, total protein content and DNA content, as well as cytochrome c oxidase activity are unchanged or even higher in ob/ob mice. Blue circles, nuclei; brown structures, mitochondria; yellow circles, lipid droplets. Note that per field/volume unit, the tissue looks atrophied, but the total amount of mitochondria is the same. (B) Uncoupling protein 1 (UCP1) is the only protein capable of performing nonshivering adaptive heat production. Although its mRNA levels normalized to housekeeping gene levels tell little about activity, the levels are a marker of acute sympathetic activation of the tissue, and, in steady-state situation, total tissue UCP1 mRNA levels correlate with the thermogenic capacity. Most reports show unaltered levels of UCP1 mRNA in ob/ob mice, whereas others show slightly reduced levels. Total UCP1 protein content of the tissue, which corresponds to the total thermogenic capacity, is largely unaltered. Purine nucleotides bind to UCP1, thereby inhibiting its activity. GDP binding can be used as a measure of UCP1 content and activation status. Surprisingly, despite unaltered protein levels of UCP1, GDP binding is generally found to be lower in ob/ob mice. This may correspond to lower sympathetic activation, but this is still not understood. (C) The activity of brown adipose tissue is regulated mainly via the sympathetic nervous system. Some reports show altered innervation patterns of BAT in ob/ob mice. Data on norepinephrine content of the tissue are inconclusive. Most reports state lower norepinephrine turnover in BAT of ob/ob mice. The response to pharmacological adrenergic stimulation is largely unaltered in ob/ob mice on the C57BL/6 background, but it is reduced on the Aston background.

Figure 5.

The difference in response to adrenergic stimulation in C57BL/6 and Aston ob/ob. Values are taken from (19). (A) Body weight of ob/ob is higher in both strains, with the Aston mice generally being larger, similarly to most other outbred strains. (B) Metabolism per whole mouse in pentobarbital-anesthetized lean and ob/ob mice of the C57BL/6 and the Aston strain are shown before and after injection of norepinephrine. Basal metabolism is higher in ob/ob mice of each strain, compared with lean controls. Aston mice display higher basal metabolism than C57BL/6 mice, most likely a reflection of their higher body weight (A). Norepinephrine causes responses in all groups, with Aston mice displaying a higher response. B, basal metabolism; NE, norepinephrine-induced metabolism. (C) Quantification of the norepinephrine-induced increase over baseline shows reduced responses in ob/ob mice of the Aston strain as compared with their lean controls, but no difference between lean and ob/ob on the C57BL/6 background.

Figure 6.

Leptin is a pyrexic and not a thermogenic agent. (A) Leptin treatment has no effect on energy expenditure in ob/ob mice under thermoneutral or subthermoneutral temperatures. (B) Leptin treatment acutely increases body temperature in ob/ob mice at subthermoneutral temperatures. (C, E) This effect is, however, not a thermogenic response, but a pyrexic (febrile) increase in body temperature because it leads to a seemingly higher “defended” body temperature through an increase in the activation threshold of heat-conserving, but not behavioral, effectors. The increase in body temperature following leptin treatment is not due to increased thermogenesis, but is mainly due to decreased heat loss. (D, F) Leptin replacement in ob/ob mice reduces food intake and adiposity. Additionally, it acutely increases body temperature in ob/ob mice, mainly by decreasing heat loss from the tail.

Figure 7.

Does BAT play any role in the response to leptin? Data on the effects of leptin replacement on BAT function are less complete than the effects of leptin deficiency. (A) BAT weight is clearly reduced by leptin treatment and also total protein content is slightly decreased. Note that the tissue seems to go from an atrophied to an active state; however, this seems only to represent a transient lipolytic effect, as BAT is not recruited. (B) UCP1 mRNA has been reported to be increased following leptin treatment. This does not, however, translate into higher UCP1 total protein levels. (C) There are signs of increased sympathetic activity following leptin treatment, such as increased norepinephrine (NE) turnover. The response to adrenergic stimulation is, however, unaltered.

Figure 8.

Leptin can prevent starvation-induced hypometabolic responses in lean mice. Although leptin treatment does not seem to induce any thermogenic response in lean or ob/ob mice under standard conditions, under some circumstances it can lead to elevated metabolic rates. Especially upon starvation or food restriction, body temperature and metabolism are usually decreased. Leptin treatment can prevent this drop in body temperature, resulting in higher energy expenditure. This is, however, not a classical thermogenic response (raising energy expenditure to produce heat), but rather an antitorpor effect of leptin (signalling to the brain that food reserves are sufficient to cope with the cold challenge).

Figure 9.

An updated picture of leptin action. Leptin acts in the brain to suppress food intake, and thereby leads to body weight reduction. Through its pyrexic effect, leptin controls body temperature by differentially regulating effector thresholds. However, and in contrast to the commonly accepted view, leptin has no thermogenic effect. It does not activate BAT-dependent thermogenesis, as is evident from the absence of an effect on energy expenditure. Leptin does not lead to thermogenic recruitment of BAT, but there are indications that leptin can affect sympathetic nerve growth and activity, thereby regulating lipolysis. Although leptin thus clearly affects food intake and body temperature thresholds, current experimental evidence neither supports an acute thermogenic effect of leptin, nor an effect on BAT thermogenic capacity. Leptin does, however, seem to affect the sympathetic innervation (and activity) in BAT and WAT, thereby affecting lipolysis and fuel supply, an effect calling for in-depth analysis in future studies.