Leptin

Leptin is an adipokine mainly produced by adipocytes in proportion to fat stores⁽Reference Zhang, Proenca and Maffei²⁶⁾. It was originally thought to be only involved in energy intake inhibition and body weight regulation acting at its hypothalamic receptors. This inhibition of appetite takes place mainly through the inhibition of neuropeptide Y (NPY)- and Agouti-related peptide (AgRP)-expressing neurons and the stimulation of pro-opiomelanocortin (POMC)-expressing neurons in the hypothalamic arcuate nucleus⁽Reference Morton, Cummings and Baskin²⁷⁾. However, a significant number of leptin receptor-expressing neurons lie outside the hypothalamic arcuate nucleus, suggesting that other brain regions known to modulate energy balance are involved in leptin's anorectic effect⁽Reference Myers, Munzberg and Leinninger²⁸⁾. Although food intake regulation is a major role of leptin, its receptor is expressed in almost all tissues⁽Reference Tartaglia, Dembski and Weng^29–Reference Frühbeck, Gómez-Ambrosi and Martínez³¹⁾, underlining a high functional pleiotropism involving energy homeostasis, glucose metabolism, reproduction, angiogenesis, immunity, gastrointestinal function, wound healing, bone remodelling and cardiovascular function⁽Reference Frühbeck, Jebb and Prentice^32–Reference Gautron and Elmquist³⁸⁾. Plasma leptin concentrations are increased in obese patients, being strongly correlated with BMI and the percentage of body fat, as well as with leptin mRNA expression in adipose tissue⁽Reference Frühbeck, Gómez-Ambrosi and Muruzábal^23, Reference Gómez-Ambrosi and Frühbeck^39, Reference Frühbeck⁴⁰⁾. The failure of high leptin concentrations to suppress feeding and mediate weight loss in common obesity defines what has been termed leptin resistance⁽Reference Myers, Cowley and Münzberg³⁶⁾.

Leptin-deficient ob/ob mice and leptin receptor-deficient db/db mutants show early-onset obesity, diabetes and reduced energy expenditure⁽Reference Lindström⁴¹⁾. Leptin administration induces a dramatic loss of adipose mass in rodents⁽Reference Pelleymounter, Cullen and Baker^42, Reference Halaas, Gajiwala and Maffei⁴³⁾. This effect is not only mediated by a reduction in energy intake, but also by a direct effect on adipose tissue⁽Reference Halaas, Gajiwala and Maffei⁴³⁾. In this sense, leptin inhibits lipogenesis and stimulates lipolysis in adipocytes through a direct effect⁽Reference Siegrist-Kaiser, Pauli and Juge-Aubry^44, Reference Frühbeck, Aguado and Gómez-Ambrosi⁴⁵⁾ or a centrally mediated action⁽Reference Gallardo, Bonzón-Kulichenko and Fernández-Agulloó⁴⁶⁾ without the release of NEFA, which are intracellularly oxidised⁽Reference Wang, Lee and Unger⁴⁷⁾. Furthermore, leptin replacement in leptin-deficient mice increases energy expenditure⁽Reference Pelleymounter, Cullen and Baker^42, Reference Hwa, Fawzi and Graziano⁴⁸⁾. A few leptin-deficient patients who also exhibit severe early-onset obesity and hormonal alterations have been identified⁽Reference Montague, Farooqi and Whitehead^49, Reference Strobel, Issad and Camoin⁵⁰⁾. Surprisingly, BMR and total energy expenditure were similar to those of age-, sex- and weight-matched controls⁽Reference Farooqi, Jebb and Langmack⁵¹⁾. Leptin replacement in humans reduces body weight mainly at the expense of the fat compartment and reverses metabolic and hormonal alterations⁽Reference Farooqi, Jebb and Langmack^51–Reference Paz-Filho, Wong and Licinio⁵⁴⁾. Interestingly, the weight loss-associated decrease in energy expenditure that takes places after prolonged negative energy balance was less pronounced in leptin-treated patients than in obese leptin-replete controls following a weight-loss programme⁽Reference Farooqi, Jebb and Langmack^51, Reference Farooqi, Matarese and Lord^52, Reference Rosenbaum, Goldsmith and Bloomfield^55, Reference Galgani, Greenway and Caglayan⁵⁶⁾. Therefore, leptin prevents the reduction in metabolic rate that is associated with weight loss⁽Reference Farooqi, Matarese and Lord^52, Reference Rosenbaum, Goldsmith and Bloomfield^55, Reference Galgani, Greenway and Caglayan⁵⁶⁾.

Adiponectin

The adipokine adiponectin is highly expressed in adipose tissue and its circulating concentrations are considerably high, accounting for approximately 0·01 % of total serum protein⁽Reference Kadowaki and Yamauchi^57–Reference Kadowaki, Yamauchi and Kubota⁵⁹⁾. Adiponectin exerts a wide variety of physiological roles through the binding of at least three different receptors called AdipoR1, AdipoR2 and T-cadherin⁽Reference Yamauchi, Kamon and Ito^60, Reference Hug, Wang and Ahmad⁶¹⁾. Serum concentrations of adiponectin are decreased in obese subjects⁽Reference Arita, Kihara and Ouchi^62, Reference Weyer, Funahashi and Tanaka⁶³⁾ and increase with weight loss⁽Reference Yang, Lee and Funahashi⁶⁴⁾. Hypoadiponectinaemia is associated with insulin resistance and patients with T2DM are reported to have decreased concentrations of adiponectin⁽Reference Weyer, Funahashi and Tanaka^63, Reference Hotta, Funahashi and Arita^65, Reference Spranger, Kroke and Mohlig⁶⁶⁾. Patients with CVD also exhibit lower adiponectin concentrations⁽Reference Ouchi, Kihara and Arita⁶⁷⁾. In addition, administration of adiponectin induces glucose-lowering effects and improves insulin sensitivity in rodent models of obesity-associated diabetes⁽Reference Berg, Combs and Du^68, Reference Combs, Wagner and Berger⁶⁹⁾. Moreover, adiponectin also exhibits anti-atherosclerotic properties⁽Reference Matsuzawa, Funahashi and Kihara⁷⁰⁾. Adiponectin-deficient mice as well as mice lacking adiponectin receptors confirm the protective effects of this adipokine in the development of insulin resistance and atherosclerosis⁽Reference Maeda, Shimomura and Kishida^71–Reference Yamauchi, Nio and Maki⁷³⁾.

Although the insulin-sensitising effects of adiponectin are clear, the effects of this adipokine in energy intake and energy expenditure regulation are still a matter of controversy⁽Reference Dridi and Taouis⁷⁴⁾. A lack of effect of intracerebroventricularly administered adiponectin on energy intake has been described. At the same time, weight-reducing actions through an increase in energy expenditure in normal mice have been shown⁽Reference Qi, Takahashi and Hileman⁷⁵⁾. These observations are confirmed in the long term after a 4-week treatment⁽Reference Park, Kim and Kwon⁷⁶⁾. However, other authors have reported that adiponectin increases energy intake through the stimulation of AMP-activated protein kinase (AMPK) in the hypothalamic arcuate nucleus via AdipoR1 with a concomitant decrease in oxygen consumption⁽Reference Kubota, Yano and Kubota⁷⁷⁾. Adiponectin- or AdipoR1-deficient mice show no changes in energy intake under a normal diet⁽Reference Maeda, Shimomura and Kishida^71, Reference Yamauchi, Nio and Maki^73, Reference Frühbeck, Alonso and Marzo⁷⁸⁾ but, when challenged with a high-fat diet, adiponectin knock-out mice show a reduced energy intake accompanied by an increase in oxygen consumption. The absence of lower food consumption in mice lacking adiponectin under a normal diet can be explained by a compensatory increase in the orexigenic pathways in order to avoid excessive weight loss. This reduction in food intake could be detrimental given the increase in energy expenditure observed after adiponectin administration⁽Reference Kubota, Yano and Kubota⁷⁷⁾.

Resistin

Resistin is another adipokine which was initially proposed as a link between increased adiposity and T2DM⁽Reference Steppan, Bailey and Bhat⁷⁹⁾. Resistin has been reported to be highly expressed in adipose tissue in mice. Under physiological circumstances, resistin appears to oppose insulin action in adipocytes and to impair glucose tolerance and insulin sensitivity in mice. Moreover, insulin-stimulated glucose uptake by adipocytes is enhanced by resistin neutralisation and is reduced by resistin treatment⁽Reference Steppan, Bailey and Bhat^79, Reference McTernan, Kusminski and Kumar⁸⁰⁾. Different genetic and dietary models of rodent obesity exhibit increased serum concentrations of resistin⁽Reference Steppan, Bailey and Bhat^79, Reference Gómez-Ambrosi and Frühbeck⁸¹⁾. Moreover, transgenic overexpression of resistin leads to insulin resistance in mice⁽Reference Rangwala, Rich and Rhoades⁸²⁾ and rats⁽Reference Satoh, Nguyen and Miles⁸³⁾. In this sense, ob/ob mice simultaneously lacking resistin exhibit an improved glucose tolerance and insulin sensitivity⁽Reference Qi, Nie and Lee⁸⁴⁾.

The exact role of resistin in human physiology and whether or not resistin is involved in the development of insulin resistance still need to be completely elucidated⁽Reference McTernan, Kusminski and Kumar^80, Reference Kusminski, McTernan and Kumar^85, Reference Schwartz and Lazar⁸⁶⁾. Several groups have described increased concentrations of resistin in obesity⁽Reference Azuma, Katsukawa and Oguchi^87, Reference Degawa-Yamauchi, Bovenkerk and Juliar⁸⁸⁾, while others report no existing differences⁽Reference Silha, Krsek and Skrha^89–Reference Heilbronn, Rood and Janderova⁹¹⁾. However, cross-sectional and prospective epidemiological investigations reveal that human resistin is not significantly involved in the development of insulin resistance⁽Reference Fehmann and Heyn^92–Reference Heidemann, Sun and van Dam⁹⁴⁾ or the metabolic syndrome⁽Reference Utzschneider, Carr and Tong⁹⁵⁾. Rather than a role in the development of insulin resistance, considerable evidence links resistin to inflammation⁽Reference Gómez-Ambrosi and Frühbeck^81, Reference Kusminski, McTernan and Kumar^85, Reference Gómez-Ambrosi and Frühbeck⁹⁶⁾. In this respect, it has been reported that resistin markedly up-regulates pro-inflammatory cytokines⁽Reference Bokarewa, Nagaev and Dahlberg^97, Reference Silswal, Singh and Aruna⁹⁸⁾ and that inflammation up-regulates resistin in human macrophages⁽Reference Kaser, Kaser and Sandhofer^99, Reference Lehrke, Reilly and Millington¹⁰⁰⁾. Furthermore, it has been confirmed that resistin is related to markers of inflammation, being a predictive factor of atherosclerosis⁽Reference Reilly, Lehrke and Wolfe¹⁰¹⁾, and that it also influences pro-inflammatory cytokine release from human adipocytes⁽Reference Kusminski, da Silva and Creely¹⁰²⁾. However, the precise physiological role of resistin in the pathogenesis and perpetuation of inflammation and CVD in humans remains unclear⁽Reference Gómez-Ambrosi and Frühbeck⁸¹⁾.

Central administration of resistin induces a short-term decrease in food intake in rats⁽Reference Tovar, Nogueiras and Tung^103–Reference Cifani, Durocher and Pathak¹⁰⁵⁾, mainly associated with a suppression of the normal fasting-induced increase in the orexigenic neuropeptides NPY and AgRP, and a suppression of the decrease in cocaine and amphetamine-regulated transcript (CART)⁽Reference Vázquez, González and Varela¹⁰⁶⁾. However, other authors have described an up-regulation of hypothalamic NPY after intracerebroventricular administration of resistin in mice without measuring the potential effect on food intake⁽Reference Singhal, Lazar and Ahima¹⁰⁷⁾. Furthermore, leptin-deficient ob/ob mice lacking resistin exhibit higher body weight and body adiposity than ob/ob mice with normal resistin levels⁽Reference Qi, Nie and Lee⁸⁴⁾. This observation has been attributed to a lower metabolic rate, uncovering a crucial role of resistin in the regulation of thermogenesis. In addition to its inhibitory effect on adipogenesis⁽Reference Kim, Lee and Moon¹⁰⁸⁾, the anorexigenic and thermogenic effects of resistin underscore the relevance of this molecule as an interesting regulator of energy homeostasis. It still remains to be elucidated whether resistin acts on the brain or peripheral tissues to control metabolic rate. The complex actions of resistin on energy expenditure and the expression of resistin in several hypothalamic nuclei suggest that resistin may act in an autocrine/paracrine fashion for regulating body weight, deserving further research⁽Reference Nogueiras, Novelle and Vazquez¹⁰⁹⁾.

Acylation-stimulating protein

Acylation-stimulating protein (ASP) is an adipokine predominantly produced by fully differentiated adipocytes⁽Reference Maslowska, Sniderman and Germinario¹¹⁰⁾. ASP results from the cleavage of the complement C3 via factor B and adipsin (factor D) interaction producing C3a, which is rapidly desarginated to give ASP (also known as C3adesArg)⁽Reference Cianflone, Xia and Chen¹¹¹⁾. ASP increases after a fat-containing meal and stimulates TAG synthesis and storage⁽Reference Frühbeck, Gómez-Ambrosi and Muruzábal^23, Reference Yasruel, Cianflone and Sniderman¹¹²⁾. ASP also stimulates translocation of GLUT1 to the cell surface⁽Reference Germinario, Sniderman and Manuel¹¹³⁾. Differences between adipose tissue depots have been observed, with greater ASP binding and action in subcutaneous compared with omental fat, in females compared with males, and in obese compared with non-obese subjects, suggesting that ASP could be a female lipogenic factor⁽Reference Saleh, Al-Wardy and Farhan¹¹⁴⁾. Circulating concentrations of ASP have been reported to be increased in obese and T2DM subjects⁽Reference Cianflone, Xia and Chen¹¹¹⁾.

Mice lacking complement C3 are deficient in ASP. These mice exhibit a striking phenotype, being leaner in spite of having a significantly increased food intake, which is counterbalanced by increased energy expenditure⁽Reference Xia, Stanhope and Digitale^115, Reference Xia, Sniderman and Cianflone¹¹⁶⁾. Similarly, mice lacking the ASP receptor C5L2 showed a phenotype similar to ASP-deficient mice, with delayed postprandial TAG clearance, increased dietary food intake, and increased muscle fatty acid oxidation⁽Reference Paglialunga, Schrauwen and Roy¹¹⁷⁾. Furthermore, exogenously administered ASP increases energy storage while ASP antibody neutralisation increases whole body energy utilisation in mice⁽Reference Paglialunga, Fisette and Munkonda¹¹⁸⁾. Recently, it has been suggested that injections of ASP in the third ventricle inhibit food intake in rats through an increase in POMC expression⁽Reference Roy, Roy and Gauvreau¹¹⁹⁾. Therefore, ASP is pointed out as a potent anabolic hormone that may also be a mediator of energy expenditure. The potential contribution of ASP as a regulator of food intake and energy expenditure in humans remains to be elucidated.

TNF-α

TNF-α is a cytokine involved in the metabolic disturbances of chronic inflammation, playing a major role in pathophysiological processes such as insulin resistance and anorexia⁽Reference Frühbeck, Gómez-Ambrosi and Muruzábal^23, Reference Rodríguez, Catalán and Gómez-Ambrosi¹²⁰⁾. In addition, TNF-α is a potent negative regulator of adipocyte differentiation⁽Reference Cawthorn, Heyd and Hegyi¹²¹⁾. Adipose tissue is both a source of and a target for TNF-α⁽Reference Cawthorn and Sethi¹²²⁾. It has been suggested that TNF-α is a candidate mediator of insulin resistance in obesity, as it is overexpressed in the adipose tissue of obese rodents and humans⁽Reference Hotamisligil¹²³⁾. This cytokine blocks the action of insulin in adipose tissue and skeletal muscle in vitro and in vivo. In this sense, TNF-α-deficient mice exhibit protection from the development of obesity-induced insulin resistance⁽Reference Uysal, Wiesbrock and Marino¹²⁴⁾.

The anorexigenic and cachexic effects of TNF-α are well known⁽Reference Tisdale^125, Reference de Kloet, Pacheco-Lopez and Langhans¹²⁶⁾. This cytokine modulates the expression of neurotransmitters involved in the control of energy homeostasis, favouring anorexia and energy expenditure⁽Reference Amaral, Barbuio and Milanski¹²⁷⁾. Brown adipose tissue (BAT) is a particular form of adipose tissue, functionally and morphologically distinct from white adipose tissue, specialised in dissipating energy in the form of heat⁽Reference Frühbeck, Sesma and Burrell^128, Reference Frühbeck, Becerril and Sáinz¹²⁹⁾. BAT has been suggested to play a role in human energy homeostasis⁽Reference Frühbeck, Becerril and Sáinz^129, Reference Cypess, Lehman and Williams¹³⁰⁾. Several controversial studies suggest that TNF-α may modulate the thermogenic capacity of BAT⁽Reference Cawthorn and Sethi¹²²⁾. Administration of TNF-α increases BAT thermogenic activity⁽Reference Masaki, Yoshimatsu and Chiba¹³¹⁾. This is consistent with the catabolic role of TNF-α. In contrast, other studies suggest that TNF-α decreases the activity of BAT⁽Reference Valladares, Roncero and Benito¹³²⁾. Accordingly, the lack of TNF-α receptor 1 improves the thermoadaptive capacity of obese animals⁽Reference Romanatto, Roman and Arruda¹³³⁾. However, information regarding the role of TNF-α in energy expenditure in humans, beyond its adipose tissue-mobilising effect, is scarce. Disentangling the exact role of TNF-α action in insulin resistance and energy expenditure in humans may provide the basis for the development of novel strategies for treating obesity and the metabolic syndrome.

IL-6

IL-6 is an inflammatory cytokine with pleiotropic effects on a variety of tissues, including stimulation of acute-phase protein synthesis and regulation of glucose and lipid metabolism⁽Reference Frühbeck and Salvador^22, Reference Frühbeck, Gómez-Ambrosi and Muruzábal²³⁾. Adipose tissue produces IL-6, with circulating levels of IL-6 being proportional to adipose mass and the magnitude of insulin resistance⁽Reference Bastard, Maachi and Van Nhieu^134, Reference Tilg and Moschen¹³⁵⁾. Although increased concentrations of IL-6 have been detected in obese subjects, mice lacking IL-6 develop mature-onset obesity, with the obese phenotype being only partly ameliorated by IL-6 replacement⁽Reference Wallenius, Wallenius and Ahren¹³⁶⁾. Interestingly, acute IL-6 treatment has been reported to produce an increase in insulin-stimulated glucose disposal in humans in vivo and to induce fatty acid oxidation, glucose transport, and GLUT4 translocation to the plasma membrane in vitro ⁽Reference Carey, Steinberg and Macaulay¹³⁷⁾. In this sense, IL-6 is considered an autocrine/paracrine regulator of adipocyte function. Its involvement in the development of insulin resistance is still not completely understood⁽Reference Tilg and Moschen¹³⁵⁾.

Similarly to TNF-α, IL-6 has been shown to trigger catabolic effects⁽Reference Tisdale¹²⁵⁾. In this sense, IL-6 exerts anti-obesity effects centrally by increasing energy expenditure⁽Reference Wallenius, Wallenius and Sunter¹³⁸⁾, with mice lacking IL-6 exhibiting adult-onset obesity⁽Reference Wallenius, Wallenius and Ahren^136, Reference Chida, Osaka and Hashimoto¹³⁹⁾ through decreased expression of corticotrophin-releasing hormone (CRH) and oxytocin in the hypothalamus. Due to the fact that CRH is known to stimulate energy expenditure and oxytocin has anorexigenic effects, a reduction in both neuropeptides may be contributing to the obese phenotype of mice lacking IL-6⁽Reference Benrick, Schele and Pinnock¹⁴⁰⁾. In humans, exogenous administration of IL-6 increases energy expenditure and activates the hypothalamic–pituitary–adrenal axis, thereby suggesting that CRH may be mediating this effect, as is observed in mice⁽Reference Stouthard, Romijn and Van der Poll^141, Reference Tsigos, Papanicolaou and Defensor¹⁴²⁾. Moreover, a polymorphism in the IL-6 gene has been shown to influence energy expenditure⁽Reference Kubaszek, Pihlajamaki and Punnonen¹⁴³⁾. It seems clear that IL-6 regulates energy expenditure; however, its exact involvement in the development of obesity in humans and its potential therapeutic utility remain to be fully elucidated. It should be noted that the treatment of obesity with IL-6 is not currently under way due to potential major side effects and the lack of knowledge regarding the relative contribution of different target organs to IL-6-induced thermogenesis⁽Reference Hoene and Weigert¹⁴⁴⁾.

Visfatin

Visfatin, previously identified as nicotinamide phosphoribosyl transferase and colony-enhancing factor of pre-B cells, was originally identified as a modulator of B cell differentiation expressed in lymphocytes, bone marrow, skeletal muscle and liver⁽Reference Ouchi, Parker and Lugus¹⁴⁵⁾. It was named visfatin because it is highly secreted by visceral fat of both mice and humans and its expression levels in serum increase during the development of obesity⁽Reference Ouchi, Parker and Lugus^145, Reference Fukuhara, Matsuda and Nishizawa¹⁴⁶⁾. Visfatin was first reported to have insulin-like activity⁽Reference Fukuhara, Matsuda and Nishizawa¹⁴⁶⁾. However, these findings are currently controversial, with the authors having been forced to retract some of their original conclusions⁽Reference Fukuhara, Matsuda and Nishizawa¹⁴⁷⁾. Surprisingly, plasma visfatin correlates with measures of obesity but not with visceral fat mass or variables of insulin sensitivity in humans. Furthermore, visfatin mRNA expression does not differ between visceral and subcutaneous adipose tissue⁽Reference Berndt, Klöting and Kralisch^148, Reference Catalán, Gómez-Ambrosi and Rodríguez¹⁴⁹⁾. Nicotinamide phosphoribosyl transferase is thought to play an important role in insulin secretion by pancreatic β-cells⁽Reference Revollo, Korner and Mills¹⁵⁰⁾ and appears to also be involved in the regulation of the inflammatory response⁽Reference Moschen, Kaser and Enrich¹⁵¹⁾.

There is little information regarding the effect of visfatin on energy homeostasis. Central administration of visfatin to Sprague–Dawley rats decreased food intake and locomotor activity, and also increased body temperature⁽Reference Park, Jin and Park¹⁵²⁾. These effects resemble those produced by other pro-inflammatory cytokines and they take place via the melanocortin pathway⁽Reference Park, Jin and Park¹⁵²⁾. Undoubtedly, more studies are needed in order to fully understand the real implications of visfatin in glucose metabolism and energy homeostasis⁽Reference Chen, Chung and Chang^93, Reference Arner¹⁵³⁾.

Visceral adipose tissue-derived serpin

Visceral adipose tissue-derived serpin (vaspin) is a member of the serine protease inhibitor family. Vaspin is highly expressed in adipocytes of visceral adipose tissue at the same time that obesity and insulin levels peak in Otsuka Long-Evans Tokushima fatty (OLETF) diabetic obese rats. Administration of vaspin to obese insulin-resistant mice improves glucose tolerance and insulin sensitivity. These findings indicate that vaspin exerts an insulin-sensitising effect in states of obesity⁽Reference Hida, Wada and Eguchi¹⁵⁴⁾. Human vaspin mRNA expression in adipose tissue is not detectable in lean normoglycaemic individuals, but is induced by increased fat mass and decreased insulin sensitivity, which could represent a compensatory mechanism associated with obesity and T2DM⁽Reference Klöting, Berndt and Kralisch¹⁵⁵⁾. However, no differences in the levels of vaspin between individuals with normal glucose tolerance and T2DM have been detected⁽Reference Youn, Klöting and Kratzsch¹⁵⁶⁾. The potential involvement of vaspin in glucose homeostasis certainly requires further investigation⁽Reference Zvonic, Lefevre and Kilroy¹⁵⁷⁾.

Vaspin mRNA in adipose tissue decreases after fasting and its levels are partially recovered after leptin treatment⁽Reference González, Caminos and Vázquez¹⁵⁸⁾. Circulating vaspin concentrations follow a meal-related circadian variation in humans, similar to that seen for ghrelin, suggesting a role for vaspin in the regulation of food intake. Serum vaspin levels exhibit a preprandial rise followed by a rapid decline after meals⁽Reference Jeong, Youn and Kim¹⁵⁹⁾. Both peripheral and central vaspin administration decrease food intake in mice⁽Reference Klöting, Kovacs and Kern¹⁶⁰⁾. Furthermore, intrahypothalamic vaspin administration reduces food intake in rats, decreasing NPY and increasing POMC mRNA expression⁽Reference Brunetti, Di Nisio and Recinella¹⁶¹⁾. Therefore, vaspin exhibits anorexigenic and glucose-lowering effects, suggesting its potential use as a therapeutic tool for the treatment of obesity and related diseases. However, the effect of vaspin on energy expenditure needs to be addressed.

Chemerin

Chemerin, also known as retinoic acid receptor responder 2, is a secreted chemoattractant protein with an important role in adaptive and innate immunity⁽Reference Catalán, Gómez-Ambrosi and Rodríguez^25, Reference Ernst and Sinal¹⁶²⁾. Chemerin has been recently described as an adipokine associated with obesity and the metabolic syndrome⁽Reference Bozaoglu, Bolton and McMillan^163, Reference Goralski, McCarthy and Hanniman¹⁶⁴⁾. Furthermore, chemerin contributes to inflammation by stimulating macrophage adhesion to extracellular matrix proteins and by promoting chemotaxis⁽Reference Hart and Greaves¹⁶⁵⁾. Increased mRNA expression of chemerin has been found in mouse and human adipocytes, while the knockdown of chemerin indicates a major role for this adipokine in regulating adipogenesis and metabolic homeostasis in the adipocyte⁽Reference Goralski, McCarthy and Hanniman¹⁶⁴⁾. Increased circulating levels of chemerin have been found in morbidly obese patients, with a significant decrease after bariatric surgery⁽Reference Sell, Divoux and Poitou^166, Reference Catalán, Gómez-Ambrosi and Rodríguez¹⁶⁷⁾.

Central administration of chemerin does not modify food intake 24 h after treatment in rats⁽Reference Brunetti, Di Nisio and Recinella¹⁶¹⁾. However, chemerin treatment increases both AgRP and POMC mRNA expression in the hypothalamus. This could lead to a null effect, considering that AgRP is orexigenic and POMC is anorexigenic, but it suggests a role of chemerin in the regulation of neuropeptides involved in food intake regulation⁽Reference Brunetti, Di Nisio and Recinella¹⁶¹⁾. The changes in chemerin concentrations after weight loss may merely reflect the reduction in body adiposity, but also a putative role in body weight homeostasis⁽Reference Sell, Divoux and Poitou^166, Reference Catalán, Gómez-Ambrosi and Rodríguez¹⁶⁷⁾.

Omentin

Omentin, also named intelectin or intestinal lactoferrin receptor, is another recently described visceral fat depot-specific adipokine⁽Reference Catalán, Gómez-Ambrosi and Rodríguez^25, Reference Schaffler, Neumeier and Herfarth¹⁶⁸⁾. It is a secreted protein likely to act as both an endocrine factor to modulate systemic metabolism and an autocrine/paracrine factor to regulate adipocyte biology locally⁽Reference Schaffler, Neumeier and Herfarth^168, Reference Yang, Lee and Hu¹⁶⁹⁾. Obesity negatively regulates omentin expression and its release into the circulation, with reduced plasma omentin levels having been observed in obese subjects⁽Reference de Souza Batista, Yang and Lee¹⁷⁰⁾. Omentin increases insulin action by enhancing insulin-mediated glucose transport in isolated adipocytes⁽Reference Schaffler, Neumeier and Herfarth^168, Reference Yang, Lee and Hu¹⁶⁹⁾. Omentin has also been related to inflammation, exerting an anti-inflammatory action and displaying beneficial effects on the metabolic syndrome⁽Reference Tan, Adya and Randeva¹⁷¹⁾.

Little is known regarding the role of omentin in energy homeostasis. Central administration of omentin produces a slight but non-significant increase in food intake in rats⁽Reference Brunetti, Di Nisio and Recinella¹⁶¹⁾. In humans, circulating omentin concentrations change oppositely to what takes place in energy balance, thereby rising after prolonged negative energy balance, as is the case after dietary-induced weight loss⁽Reference Moreno-Navarrete, Catalán and Ortega¹⁷²⁾. Further studies are necessary in order to gain more insight regarding the involvement of omentin in the regulation of appetite and energy expenditure.

Angiontensin II

Murine models of obesity show increased local formation of angiotensin (Ang) II due to elevated secretion of its precursor, angiotensinogen, from adipocytes⁽Reference Frühbeck, Gómez-Ambrosi and Muruzábal^23, Reference Yvan-Charvet and Quignard-Boulange^173, Reference Kalupahana and Moustaid-Moussa¹⁷⁴⁾, with deficiency or overexpression of angiotensinogen affecting body weight regulation⁽Reference Frühbeck and Gómez-Ambrosi¹³⁾. Given the close relationship between Ang II and insulin resistance and the fact that the renin–Ang system is inappropriately activated in obesity⁽Reference Katovich and Pachori^175, Reference Luther and Brown¹⁷⁶⁾, the participation of the adipose tissue–renin–Ang system in the development of insulin resistance and the metabolic syndrome is conceivable in humans, but has to be evaluated in more detail⁽Reference Yvan-Charvet and Quignard-Boulange^173, Reference Engeli, Schling and Gorzelniak¹⁷⁷⁾.

Mice lacking angiotensinogen or any of its two major receptor subtypes, type 1 (AT1R) and type 2, show protection against the development of obesity without notable changes in food intake⁽Reference Massiera, Seydoux and Geloen^178–Reference Yvan-Charvet, Even and Bloch-Faure¹⁸⁰⁾. By contrast, it is clear that these knockout mice exhibit higher energy expenditure, particularly when a high-fat diet is consumed⁽Reference Kouyama, Suganami and Nishida^179, Reference Yvan-Charvet, Even and Bloch-Faure¹⁸⁰⁾. Furthermore, mice deficient in Ang-converting enzyme (ACE), which is responsible for the conversion of Ang I to the bioactive peptide Ang II, also have increased energy expenditure, with reduced fat mass and improved glucose clearance⁽Reference Jayasooriya, Mathai and Walker¹⁸¹⁾. Paradoxically, administration of Ang II to rats by means of subcutaneous osmotic minipumps produces a maintained reduction of food intake with a transient decrease in oxygen consumption⁽Reference Brink, Wellen and Delafontaine^182, Reference Cassis, Helton and English¹⁸³⁾. Furthermore, ACE inhibition with captopril reduces food intake and protects against the development of diet-induced obesity and glucose intolerance in rats⁽Reference de Kloet, Krause and Kim¹⁸⁴⁾. However, Ang II may exert different effects on metabolism, depending on the tissues of action. In this sense, accumulating evidence suggests that increased Ang II activity locally within the brain promotes negative energy balance⁽Reference Porter and Potratz^185, Reference de Kloet, Krause and Scott¹⁸⁶⁾. Taken together, these studies suggest that reduced systemic Ang II signalling protects against diet-induced adipose tissue enlargement by increasing energy expenditure in rodents⁽Reference Yvan-Charvet and Quignard-Boulange¹⁷³⁾. In humans, although ACE inhibitors and AT1R blockers are widely used as antihypertensive agents and are beginning to be used for promoting insulin sensitivity, there is minimal evidence that these agents significantly affect energy homeostasis⁽Reference de Kloet, Krause and Woods¹⁸⁷⁾.

Sex steroids

Adipose tissue expresses different sex steroid-metabolising enzymes that promote the conversion of oestrogens from androgenic precursors, which are produced by the gonads or adrenal glands, thereby regulating the synthesis and bioavailability of endogenous sex steroids, oestrogens and androgens, through different mechanisms⁽Reference Tchernof and Despres^188–Reference Mauvais-Jarvis¹⁹¹⁾. In this sense, in men and postmenopausal women, adipose tissue is the main source of oestrogen synthesis, and circulating levels of oestrogens are directly related to BMI. Sex steroid hormones play a critical role in adipose tissue metabolism, distribution and accretion⁽Reference Tchernof and Despres^188, Reference Mayes and Watson¹⁸⁹⁾. In women, menopause-induced oestrogen deficiency and increased androgenicity are associated with increased visceral obesity and with the subsequent cardiometabolic alterations⁽Reference Tchernof and Despres¹⁸⁸⁾. Moreover, hormone replacement therapy with oestradiol treatment for 1 year decreased intra-abdominal fat⁽Reference Mattiasson, Rendell and Tornquist¹⁹²⁾. Ageing in men is associated with a progressive deficit in androgen production and reduced concentrations of testosterone have been related to increased visceral obesity and the metabolic syndrome⁽Reference Tsai, Boyko and Leonetti^193, Reference Kupelian, Page and Araujo¹⁹⁴⁾. Therefore, treating middle-aged obese men with testosterone reduces abdominal fat⁽Reference Marin, Holmang and Jonsson¹⁹⁵⁾.

In females, oestrogens regulate energy homeostasis via oestrogen receptor (ER)α and ERβ, by reducing food intake⁽Reference Somogyi, Gyorffy and Scalise^18, Reference Brown, Gent and Davis^190, Reference Eckel¹⁹⁶⁾ and adiposity⁽Reference Heine, Taylor and Iwamoto¹⁹⁷⁾, enhancing energy expenditure⁽Reference Musatov, Chen and Pfaff^198, Reference Ropero, Alonso-Magdalena and Quesada¹⁹⁹⁾, and improving insulin sensitivity⁽Reference Ropero, Alonso-Magdalena and Quesada¹⁹⁹⁾. In males, testosterone is converted to oestrogen and controls energy homeostasis via ER and androgen receptors, which share related functions for increasing energy expenditure, reducing fat accumulation⁽Reference Fan, Yanase and Nomura²⁰⁰⁾ and ameliorating glucose homeostasis⁽Reference Mauvais-Jarvis¹⁹¹⁾. It has been recently reported that distinct hypothalamic neurons mediate oestrogenic effects on food intake and energy homeostasis. Food intake is regulated by ERα in POMC neurons while energy expenditure and fat accumulation is controlled by ERα in steroidogenic factor-1 neurons⁽Reference Xu, Nedungadi and Zhu²⁰¹⁾.

Insulin

Insulin was the first hormone to be involved in the control of body weight by the central nervous system⁽Reference Plum, Belgardt and Brüning^202–Reference Brüning, Gautam and Burks²⁰⁴⁾. To date, insulin and leptin are the only hormones that fulfil the criteria to be considered an adiposity signal. Both hormones circulate at concentrations proportional to the amount of body fat and enter the central nervous system where leptin and insulin receptors are expressed by neurons involved in energy homeostasis; administration of either molecule into the brain reduces food intake, whereas its deficiency does the opposite⁽Reference Morton, Cummings and Baskin^27, Reference Schwartz, Woods and Porte²⁰⁵⁾. The breeding of mice with brain-specific insulin receptor deficiency, which translates into an increased food intake and diet-sensitive obesity, demonstrated a critical role for brain insulin signalling in the central regulation of energy disposal and fuel metabolism⁽Reference Brüning, Gautam and Burks²⁰⁴⁾. The central effect of insulin on the reduction of food intake is mainly mediated through an inhibition of NPY/AgRP neurons and the stimulation of POMC neurons⁽Reference Morton, Cummings and Baskin^27, Reference Plum, Belgardt and Brüning²⁰²⁾. Besides these basic homeostatic circuits, food palatability and reward are thought to be major factors involved in the regulation of food intake elicited by insulin and leptin⁽Reference Könner, Klöckener and Brüning²⁰⁶⁾. Although it was initially considered that the effect of insulin was dose dependent, with low doses stimulating thermogenesis and high doses decreasing it⁽Reference Rothwell and Stock²⁰⁷⁾, further studies have shown the thermogenic effect of insulin⁽Reference Menéndez and Atrens^208, Reference Dulloo and Girardier²⁰⁹⁾.

Glucagon

Glucagon is secreted by the α-cells of the pancreatic islets. It has catabolic properties, functioning as a counter-regulatory hormone opposing the actions of insulin. Glucagon maintains blood glucose concentrations during fasting by promoting glycogenolysis and gluconeogenesis as well as by inhibiting glycogenesis and glycolysis in the liver, thereby preventing hypoglycaemia⁽Reference Jiang and Zhang^210, Reference Gómez-Ambrosi, Catalan and Frühbeck²¹¹⁾. Obese patients exhibit increased glucagon levels⁽Reference Kalkhoff, Gossain and Matute^212, Reference Starke, Erhardt and Berger²¹³⁾. Inappropriately elevated concentrations of insulin and glucagon, together with insulin resistance, contribute to the obesity-associated impaired glucose homeostasis⁽Reference Koeslag, Saunders and Terblanche²¹⁴⁾.

Glucagon exerts many extrahepatic actions. It increases lipolysis in adipose tissue and reduces food intake, acting as a satiety factor in the brain⁽Reference Gómez-Ambrosi, Catalan and Frühbeck²¹¹⁾. In humans, pharmacological doses of glucagon decrease the amount of food that is eaten⁽Reference Penick and Hinkle²¹⁵⁾. Although this effect was initially attributed to an indirect action of glucagon via the increase in portal glucose levels⁽Reference Geary, Langhans and Scharrer²¹⁶⁾, it has been clearly demonstrated that glucagon has central actions in the brain to reduce food intake⁽Reference Habegger, Heppner and Geary²¹⁷⁾. Administration of glucagon has been shown to produce weight loss in humans and rats⁽Reference Schulman, Carleton and Whitney^218, Reference Billington, Briggs and Link²¹⁹⁾. This can be explained by the fact that glucagon causes an increase in energy expenditure⁽Reference Davidson, Salter and Best²²⁰⁾ via activation of BAT⁽Reference Billington, Briggs and Link²¹⁹⁾. The effect of glucagon on human BAT remains to be fully clarified. However, this effect of glucagon promoting energy expenditure contrasts with the phenotype observed in glucagon receptor knockout mice that exhibit lower adiposity, despite having normal growth rates, body weight, food intake and resting energy expenditure⁽Reference Gelling, Du and Dichmann²²¹⁾. Similar striking observations are reported for mice lacking synaptotagmin-7, a Ca sensor for insulin and glucagon granule exocytosis, which show normal insulin concentrations but severely reduced glucagon levels. This mouse model exhibits a lean phenotype with increased lipolysis and energy expenditure⁽Reference Lou, Gustavsson and Wang²²²⁾. A potential counter-regulatory increase in GLP-1 in mice lacking glucagon signalling could explain these discrepancies⁽Reference Vuguin and Charron²²³⁾.

Pancreatic polypeptide

PP is a thirty-six-amino acid peptide belonging to the family of peptides including NPY and PYY, which is secreted postprandially under vagal control by pancreatic islet PP cells⁽Reference Field, Chaudhri and Bloom^19, Reference Schwartz, Holst and Fahrenkrug²²⁴⁾. Circulating levels of PP are apparently normal in obese patients⁽Reference Jorde and Burhol^225, Reference Pieramico, Malfertheiner and Nelson²²⁶⁾ but the rapid increase in response to a meal observed in healthy subjects is significantly impaired in obese individuals⁽Reference Lassmann, Vague and Vialettes^227, Reference Holst, Schwartz and Lovgreen²²⁸⁾.

Peripheral administration of PP acutely reduces food intake and gastric emptying and increases energy expenditure in mice, whereas repeated administration during 2 weeks leads to reductions in body weight gain⁽Reference Asakawa, Inui and Yuzuriha²²⁹⁾. Similar effects are observed in transgenic mice overexpressing PP⁽Reference Ueno, Inui and Iwamoto²³⁰⁾. The anorectic effect is mediated through Y4 receptors and is associated with reduced expression of NPY and orexin mRNA in the hypothalamus, being dependent on intact vagal signalling⁽Reference Asakawa, Inui and Yuzuriha²²⁹⁾. In humans, intravenous infusion of PP leads to delayed gastric emptying and reduced cumulative 24 h food intake⁽Reference Batterham, Le Roux and Cohen^231, Reference Schmidt, Näslund and Grybäck²³²⁾. To our knowledge, there are no data reporting the effect of PP on energy expenditure in humans.

Amylin

Amylin, a peptide co-secreted with insulin postprandially by pancreatic β-cells and, therefore, also named islet amyloid polypeptide, was firstly isolated from diabetic human pancreas⁽Reference Cooper, Willis and Clark²³³⁾. Amylin inhibits gastric emptying as well as gastric acid and glucagon secretion⁽Reference Field, Chaudhri and Bloom^19, Reference Cummings and Overduin^234, Reference Lutz²³⁵⁾. Increased circulating concentrations of amylin have been reported in obese rats and humans, suggesting a role for amylin in the pathophysiology of obesity⁽Reference Boyle and Lutz²³⁶⁾.

Central or peripheral administration of amylin decreases meal size and food intake by promoting meal-ending satiation⁽Reference Lutz²³⁵⁾. The anorectic effects of amylin, in contrast to other gut peptides, take place primarily in the area postrema, showing synergy with PYY, CCK and leptin⁽Reference Trevaskis, Parkes and Roth²³⁷⁾. Mechanistic studies in rodents suggest that amylin reduces body weight in a fat-specific manner, preserving lean mass⁽Reference Trevaskis, Parkes and Roth²³⁷⁾. The synthetic amylin analogue pramlintide is prescribed for the treatment of diabetes but also causes mild progressive weight loss in humans⁽Reference Aronne, Fujioka and Aroda²³⁸⁾. Furthermore, pramlintide also exhibits synergic effects with leptin after 20 weeks of treatment in overweight and obese volunteers⁽Reference Roth, Roland and Cole²³⁹⁾. In addition to its role eliciting satiety, amylin appears to influence energy balance by increasing energy expenditure⁽Reference Roth, Hughes and Kendall^240–Reference Osaka, Tsukamoto and Koyama²⁴²⁾. Amylin administration increases lipid utilisation, as indicated by a lower respiratory quotient, reducing adiposity⁽Reference Rushing, Hagan and Seeley²⁴³⁾. Finally, it has been suggested that amylin could prevent the compensatory decrease in energy expenditure that typically takes place during or after weight loss⁽Reference Lutz²³⁵⁾.

Ghrelin

Ghrelin is a twenty-eight-amino acid peptide secreted by oxyntic cells in the stomach fundus. Ghrelin was first characterised as a natural ligand of the growth hormone secretagogue receptor⁽Reference Kojima, Hosoda and Date²⁴⁹⁾. In subsequent studies ghrelin was shown to participate in the complex entero-hypothalamic control of food intake signalling⁽Reference Nakazato, Murakami and Date²⁵⁰⁾. Central or peripheral administration of ghrelin increases food intake and adiposity in rodents⁽Reference Tschöp, Smiley and Heiman^251, Reference Asakawa, Inui and Kaga²⁵²⁾ and humans⁽Reference Wren, Seal and Cohen^253, Reference Rodríguez, Gómez-Ambrosi and Catalán²⁵⁴⁾. In humans, plasma ghrelin concentrations have been shown to rise shortly before and fall quickly after every meal, suggesting a role in meal initiation⁽Reference Cummings, Purnell and Frayo²⁵⁵⁾. Ghrelin concentrations are decreased in human obesity⁽Reference Tschöp, Weyer and Tataranni²⁵⁶⁾, which has been explained as a physiological adaptation to the positive energy balance associated with obesity, and increase in response to diet-induced weight loss⁽Reference Cummings, Weigle and Frayo²⁵⁷⁾. However, many studies have shown that ghrelin concentrations do not increase after surgically induced weight loss following procedures that compromise the functionality of the fundus⁽Reference Frühbeck, Diez Caballero and Gil^258, Reference Frühbeck, Diez-Caballero and Gil²⁵⁹⁾, while other studies have reported an increase in ghrelin levels following bariatric surgery⁽Reference Holdstock, Engström and Öhrvall²⁶⁰⁾.

Chronic central or peripheral administration of ghrelin increases cumulative food intake and decreases energy expenditure, resulting in body weight gain⁽Reference Nakazato, Murakami and Date^250, Reference Tschöp, Smiley and Heiman^251, Reference Wren, Seal and Cohen^253, Reference Wren, Small and Abbott²⁶¹⁾. The orexigenic effect of ghrelin is mediated by the activation of NPY/AgRP, since ghrelin does not stimulate feeding in NPY and AgRP double knockout mice⁽Reference Chen, Trumbauer and Chen²⁶²⁾, and it is also mediated by the inhibition of hypothalamic fatty acid biosynthesis⁽Reference López, Lage and Saha²⁶³⁾. In this sense, the absence of ghrelin or its receptors protects against the development of early-onset obesity⁽Reference Wortley, Del Rincon and Murray^264, Reference Zigman, Nakano and Coppari²⁶⁵⁾ and mice lacking simultaneously ghrelin and its receptor exhibit an increased energy expenditure⁽Reference Pfluger, Kirchner and Günnel²⁶⁶⁾. Therefore, ghrelin signalling inhibition has been suggested as a potential therapeutic tool in obesity treatment, whereas a direct therapeutic application of ghrelin can be contemplated for the treatment of cachexia and anorexia⁽Reference Wren and Bloom²⁴⁸⁾. In this sense, intravenous administration of ghrelin results in weight gain in patients with cardiac cachexia and chronic obstructive pulmonary disease⁽Reference Nagaya, Moriya and Yasumura^267, Reference Nagaya, Itoh and Murakami²⁶⁸⁾.

Glucose-dependent insulinotropic polypeptide

Insulin secretion is higher in response to orally administered than to intravenous glucose administration. This is known as the incretin effect⁽Reference Drucker²⁶⁹⁾. Originally named gastric inhibitory polypeptide, glucose-dependent insulinotropic polypeptide (GIP) was the first incretin identified⁽Reference Brown²⁷⁰⁾. The major stimulus for GIP secretion is nutrient intake. GIP is mainly secreted by K cells in the duodenum and jejunum⁽Reference Drucker^269, Reference Song and Wolfe²⁷¹⁾. Circulating concentrations of GIP are low in the fasting state and rise within minutes following food intake. Moreover, GIP levels increase after a high-fat diet, with postprandial GIP secretion being significantly higher in obese subjects than in age-matched lean individuals⁽Reference Song and Wolfe^271, Reference Roust, Stesin and Go²⁷²⁾. In addition to its role in regulating insulin secretion, GIP stimulates β-cell replication and mass expansion at the same time as it stimulates glucagon secretion⁽Reference Drucker^273, Reference Wideman and Kieffer²⁷⁴⁾. In addition, GIP inhibits gastric acid secretion and gastric emptying, although only at supraphysiological doses⁽Reference Fehmann, Goke and Goke²⁷⁵⁾.

The GIP receptor (GIPR) is present in adipose tissue, regulating adipocyte growth, and there is a large body of biochemical and animal data suggesting that GIP signalling promotes fat accumulation⁽Reference Song and Wolfe^271, Reference Wasada, McCorkle and Harris^276–Reference Althage, Ford and Wang²⁷⁸⁾. Chemical or genetic ablation of GIP signalling or targeted reduction of GIP-secreting cells does not modify food intake⁽Reference Althage, Ford and Wang^278–Reference Hansotia, Maida and Flock²⁸¹⁾. However, the absence of GIP signalling produces a significant increase in energy expenditure, protecting from high-fat diet-induced obesity and insulin resistance⁽Reference Althage, Ford and Wang^278, Reference Miyawaki, Yamada and Ban^279, Reference Hansotia, Maida and Flock²⁸¹⁾. Moreover, peripheral administration of synthetic human GIP reduces energy expenditure in healthy subjects but not in patients with T2DM⁽Reference Daousi, Wilding and Aditya²⁸²⁾. Furthermore, emerging evidence suggests that the rapid resolution of diabetes in morbidly obese patients undergoing bypass surgery is mediated, at least in part, by surgical removal of GIP-secreting cells in the upper small intestine⁽Reference Irwin and Flatt²⁸³⁾. Although inhibiting GIP/GIPR signalling may be beneficial as a treatment for obesity⁽Reference Song and Wolfe²⁷¹⁾, the mechanisms involved in the regulation of food intake and energy expenditure elicited by GIP in humans remain to be fully understood.

Glucagon-like peptide 1

Through action of prohormone convertases, proglucagon is processed to glicentin, oxyntomodulin, intervening peptides 1 and 2, GLP-1 and GLP-2. To date, only GIP and GLP-1 are considered to be incretin hormones in humans, being responsible for as much as 50 % of postprandial insulin secretion⁽Reference Wideman and Kieffer^274, Reference Baggio and Drucker²⁸⁴⁾. GLP-1 is mainly produced by L-cells in the ileum after meals; in addition to its role as an insulinotropic hormone it participates actively in regulating gastric motility, islet β-cell neogenesis, neuronal plasticity and the suppression of plasma glucagon concentrations⁽Reference Kieffer and Habener^285, Reference Drucker²⁸⁶⁾. Although a clear role for GLP-1 in the aetiology of T2DM has not been proved, a common view states that GLP-1 secretion in patients with T2DM is deficient⁽Reference Nauck, Vardarli and Deacon²⁸⁷⁾. Similarly, some controversies exist regarding the involvement of GLP-1 in obesity pathophysiology. It has been suggested that obesity is associated with reduced secretion of GLP-1⁽Reference Ranganath, Beety and Morgan^288–Reference Torekov, Madsbad and Holst²⁹⁰⁾, which is restored to normal levels after weight loss⁽Reference Verdich, Toubro and Buemann²⁹¹⁾, particularly following malabsorptive bariatric surgery⁽Reference Morínigo, Moize and Musri²⁹²⁾.

GLP-1 has an important role in food intake regulation, promoting satiety⁽Reference Turton, O'Shea and Gunn^293–Reference Gutzwiller, Goke and Drewe²⁹⁵⁾ even in obese men⁽Reference Naslund, Barkeling and King^296, Reference Flint, Raben and Ersboll²⁹⁷⁾, acting as a short-term satiation signal, limiting the amount of food eaten and prolonging time between meals⁽Reference Williams, Baskin and Schwartz²⁹⁸⁾. Another important action of this incretin in relation to energy homeostasis is the inhibition of gastric emptying following GLP-1 administration, with the vagus nerve playing an important role⁽Reference Drucker^286, Reference Flint, Raben and Ersboll²⁹⁷⁾. Intravenous administration of GLP-1 increases postprandial energy expenditure via the lower brainstem and the sympathoadrenal system in rats⁽Reference Osaka, Endo and Yamakawa²⁹⁹⁾. Exogenous administration of GLP-1 to humans reduces postprandial thermogenesis, which can be explained by a reduction in meal size⁽Reference Flint, Raben and Ersboll²⁹⁷⁾. However, higher fasting plasma concentrations of GLP-1 are associated with higher resting energy expenditure and fat oxidation rates in humans⁽Reference Pannacciulli, Bunt and Koska³⁰⁰⁾. Data regarding energy expenditure from mice deficient in GLP-1 signalling are conflicting, with some studies finding that loss of function protects against the development of diet-induced obesity by increasing energy expenditure⁽Reference Hansotia, Maida and Flock^281, Reference Knauf, Cani and Ait-Belgnaoui^301, Reference Ayala, Bracy and James³⁰²⁾, while others show that loss of GLP-1 signalling increases fat accumulation⁽Reference Nogueiras, Perez-Tilve and Veyrat-Durebex^303, Reference Barrera, Jones and Herman³⁰⁴⁾. These differences may be related to species-specific differences and effects on locomotor activity⁽Reference Hayes, De Jonghe and Kanoski³⁰⁵⁾.

GLP-1 signalling is a potential target for the treatment of both T2DM and obesity. In this sense, liraglutide, a GLP-1 analogue with a prolonged half-life initially developed for the treatment of T2DM, has shown additional beneficial features for body weight control⁽Reference Vilsboll, Zdravkovic and Le-Thi^306, Reference Astrup, Rossner and Van Gaal³⁰⁷⁾. In this context, activation of the GLP-1 receptor is currently proposed as the most effective drug for treating the metabolic syndrome⁽Reference Day, Ottaway and Patterson³⁰⁸⁾.

GLP-2 has also been involved in the regulation of food intake⁽Reference Tang Christensen, Larsen and Thulesen³⁰⁹⁾. However, deletion of GLP-2 receptor signalling in ob/ob mice impairs the normal islet adaptive response needed for maintaining glucose homeostasis but has no effect on body weight or food intake⁽Reference Bahrami, Longuet and Baggio³¹⁰⁾. GLP-2 has also been associated with gut hypertrophy and intestinal crypt cell proliferation after gastric bypass⁽Reference le Roux, Borg and Wallis³¹¹⁾ but has no effect on energy expenditure⁽Reference Osaka, Endo and Yamakawa²⁹⁹⁾.

Further evidence of the important involvement of incretins in energy homeostasis arises from studies involving dipeptidyl peptidase 4 (DPP-4). DPP-4 is a member of the prolyl oligopeptidase family of peptidases and is the key enzyme responsible for cleaving and inactivating GIP and GLP-1⁽Reference Drucker³¹²⁾. Mice lacking DPP-4 exhibit improved glucose tolerance and insulin sensitivity as well as resistance to diet-induced obesity, which can be explained by reduced food intake and increased energy expenditure⁽Reference Conarello, Li and Ronan³¹³⁾.

Peptide YY

PYY is a thirty-six-amino acid gut hormone so called after the tyrosine residues at each terminus of the peptide that belongs to the NPY family⁽Reference Tatemoto and Mutt^314, Reference Karra and Batterham³¹⁵⁾. It is secreted mainly from specialised enteroendocrine cells, called L-cells, of the distal gut, with the highest production being in the ileum and colon. Two main endogenous forms of PYY exist, PYY_1–36 and PYY_3–36. PYY_3–36 is the dominant circulating form of the peptide both in the fasted and fed states, accounting for 60 % of postprandial circulating PYY⁽Reference Grandt, Schimiczek and Beglinger^316, Reference Kirchner, Tong and Tschöp³¹⁷⁾. Circulating concentrations of PYY rise within 15 min after nutrient ingestion. PYY_1–36 has specificity for Y1 and Y5 receptors, increasing food intake. However, PYY_3–36 binds preferentially to Y2 receptors, thereby stimulating anorectic pathways⁽Reference Zac-Varghese, De Silva and Bloom³¹⁸⁾. Lower levels of PYY_3–36 have been reported in obese individuals, suggesting that this gut hormone has a role in the pathophysiology of obesity⁽Reference Batterham, Cohen and Ellis³¹⁹⁾.

Peripheral injection of PYY_3–36 has been shown to reduce food intake and to induce a negative energy balance in mice and rats⁽Reference Batterham, Cowley and Small³²⁰⁾, monkeys⁽Reference Moran, Smedh and Kinzig³²¹⁾ and humans⁽Reference Batterham, Cowley and Small³²⁰⁾, even in obese patients⁽Reference Batterham, Cohen and Ellis³¹⁹⁾. This occurs through modulation of different cortical and hypothalamic brain areas⁽Reference Batterham, Ffytche and Rosenthal³²²⁾. However, these anorexigenic effects of PYY_3–36 have not been confirmed by others⁽Reference Tschöp, Castañeda and Joost³²³⁾. Furthermore, despite the effects of PYY_3–36 on food intake inhibition, mice lacking Pyy do not exhibit a clear phenotype, showing normal feeding behaviour, growth and energy expenditure⁽Reference Schonhoff, Baggio and Ratineau^324, Reference Wortley, Garcia and Okamoto³²⁵⁾, or even obesity⁽Reference Boey, Lin and Karl^326, Reference Batterham, Heffron and Kapoor³²⁷⁾. Intravenous administration of PYY_3–36 increases lipolysis and energy expenditure in humans⁽Reference Sloth, Holst and Flint³²⁸⁾, with total PYY being significantly correlated with postprandial energy expenditure⁽Reference Doucet, Laviolette and Imbeault³²⁹⁾. However, this association has not been unequivocally found⁽Reference Guo, Ma and Enriori³³⁰⁾. The extent of PYY_3–36 involvement in the regulation of energy homeostasis and the underlying mechanisms mediating the effects of PYY_3–36 on energy expenditure in humans are still not fully understood.

Cholecystokinin

CCK is secreted mainly by I-cells in the proximal small intestine in response to lipids and proteins in the meal⁽Reference Field, Chaudhri and Bloom^19, Reference Cummings and Overduin^234, Reference Raybould³³¹⁾. The predominant circulating forms of CCK in rodents include CCK octapeptide (CCK-8) and CCK-22, whereas larger molecular forms are also present in human plasma⁽Reference Liddle, Goldfine and Rosen³³²⁾. CCK is involved in modulating intestinal motility, stimulating pancreatic enzyme secretion, enhancing gallbladder contraction and regulating meal size but not meal frequency⁽Reference Lo, Samuelson and Chambers³³³⁾. A total of two CCK receptors have been cloned so far: CCK1R and CCK2R. Selective CCK1R antagonists block the anorectic effect of CCK, whereas selective antagonism of CCK2R has no effect on food intake⁽Reference Cummings and Overduin^234, Reference Crawley and Corwin^334–Reference Lo, King and Samuelson³³⁶⁾.

The satiating effect of CCK was first described more than three decades ago⁽Reference Gibbs, Young and Smith³³⁷⁾, with vagotomy suppressing the anorectic effects of peripheral CCK⁽Reference Smith, Jerome and Cushin³³⁸⁾. In humans, intravenous infusion of CCK induces a dose-dependent suppression of food intake⁽Reference Lieverse, Jansen and Masclee³³⁹⁾. Administration of CCK before the start of a meal does not delay the onset of eating, but rather reduces the amount of food consumed once eating begins⁽Reference Woods¹²⁾. However, long-term CCK1R stimulation failed to produce significant weight loss in obese patients due to the rapid development of tolerance⁽Reference Crawley and Beinfeld^340, Reference Jordan, Greenway and Leiter³⁴¹⁾, thereby questioning the potential of CCK as an anti-obesity target⁽Reference Cummings and Overduin²³⁴⁾.

Rats deficient in CCK1R show increased meal size and obesity⁽Reference Moran, Katz and Plata-Salaman³⁴²⁾. However, mice lacking CCK or CCK1R exhibit a normal food intake and body weight, apparently indicating that CCK is not essential for the long-term maintenance of body weight⁽Reference Lo, Samuelson and Chambers^333, Reference Kopin, Mathes and McBride³³⁵⁾. Interestingly, CCK knockout mice fed on a high-fat diet develop protection against obesity despite having a normal food intake, probably through decreased lipid absorption and increased energy expenditure⁽Reference Lo, King and Samuelson³³⁶⁾.

Oxyntomodulin

Oxyntomodulin is another cleavage product of proglucagon secreted by intestinal L-cells after meals in proportion to the energy content of foods⁽Reference Bataille, Tatemoto and Gespach^343, Reference Le Quellec, Kervran and Blache³⁴⁴⁾. It was named oxyntomodulin after its inhibitory action on the oxyntic glands of the stomach⁽Reference Dubrasquet, Bataille and Gespach³⁴⁵⁾. Oxyntomodulin inhibits gastric acid secretion and pancreatic enzyme secretion⁽Reference Schjoldager, Mortensen and Myhre³⁴⁶⁾. Although no oxyntomodulin receptor has been identified yet, it appears that the actions of oxyntomodulin are mediated via the GLP-1 receptor⁽Reference Schepp, Dehne and Riedel³⁴⁷⁾, since the anorectic effect of oxyntomodulin is abolished in GLP-1 receptor-deficient mice⁽Reference Baggio, Huang and Brown³⁴⁸⁾. However, GLP-1 and oxyntomodulin appear to activate different hypothalamic pathways⁽Reference Parkinson, Chaudhri and Kuo³⁴⁹⁾ and, therefore, a separate unidentified oxyntomodulin receptor may exist⁽Reference Karra and Batterham³¹⁵⁾.

Central or peripherally administered oxyntomodulin inhibits food intake in fasted and non-fasted rats⁽Reference Dakin, Gunn and Small^350, Reference Dakin, Small and Batterham³⁵¹⁾. However, the anorectic effect in mice is only observed after intracerebroventricular administration⁽Reference Baggio, Huang and Brown³⁴⁸⁾. Intravenous administration of oxyntomodulin suppresses appetite and reduces food intake in humans⁽Reference Cohen, Ellis and Le Roux³⁵²⁾. Furthermore, subcutaneous injections of oxyntomodulin resulted in weight loss and a change in the levels of adipokines consistent with a loss of body fat over a 4-week period in overweight and obese subjects⁽Reference Wynne, Park and Small³⁵³⁾. Central administration of oxyntomodulin increases energy expenditure and causes a disproportionate reduction in body weight compared with pair-fed rats⁽Reference Dakin, Small and Batterham^351, Reference Dakin, Small and Park³⁵⁴⁾. In humans, 4 d subcutaneous self-administration of pre-prandial oxyntomodulin three times per d promotes a negative energy balance, increasing energy expenditure while reducing energy intake⁽Reference Wynne and Bloom³⁵⁵⁾. However, the acute thermogenic effect of oxyntomodulin observed in rats and humans has not been reproduced in mice⁽Reference Baggio, Huang and Brown³⁴⁸⁾. Further studies are needed in order to investigate whether the effect of oxyntomodulin on energy expenditure in humans is maintained in the long term, but data presented above support the role of oxyntomodulin as a potential anti-obesity tool.

Uroguanylin

Guanylin and uroguanylin have been well-known key paracrine players in intestinal ion and water balance for over 20 years, acting as endogenous ligands of guanylyl cyclase (GUCY) 2C and increasing cyclic guanosine monophosphate (cGMP) production⁽Reference Potter³⁵⁶⁾. They are secreted by intestinal epithelial cells as prohormones, requiring proteolytic enzymic conversion into active hormones in the target tissue⁽Reference Seeley and Tschöp³⁵⁷⁾. Physiological functions for these molecules include the modulation of epithelial cell balance in the intestinal epithelium and the regulation of Na balance through actions on the kidney⁽Reference Carrithers, Ott and Hill³⁵⁸⁾. Recently, Valentino et al. ⁽Reference Valentino, Lin and Snook¹⁶⁾ revealed a new endocrine role for uroguanylin in energy homeostasis. The uroguanylin precursor, prouroguanylin, is secreted into the circulation after meals in both mice and humans; it can then be cleaved to uroguanylin in the hypothalamus to activate GUCY2C for decreasing food intake⁽Reference Valentino, Lin and Snook¹⁶⁾. Deletion of GUCY2C in mice disrupts appetite regulation specifically by impairing satiation, producing hyperphagia associated with obesity and glucose intolerance⁽Reference Valentino, Lin and Snook¹⁶⁾. No changes in cold-induced thermogenesis assessed by core body temperature were observed⁽Reference Valentino, Lin and Snook¹⁶⁾. However, the role of uroguanylin in the modulation of energy expenditure needs to be addressed in both rodents and humans. Furthermore, it has been suggested that uroguanylin could exert a direct effect on adipose tissue, regulating lipolysis⁽Reference Frühbeck³⁵⁹⁾, given the fact that cGMP is a second messenger known to be involved in the lipolytic effect of natriuretic peptides, which are closely related to uroguanylin⁽Reference Lafontan, Moro and Berlan³⁶⁰⁾. The uroguanylin–GUCY2C endocrine axis may offer a novel therapeutic target for regulating food intake and a weapon against obesity⁽Reference Valentino, Lin and Snook^16, Reference Frühbeck³⁵⁹⁾.

Fibroblast growth factor 19

The family of fibroblast growth factors (FGF) regulates a plethora of processes including organ development, the maintaining of bile acid homeostasis and the activation of hepatic protein and glycogen synthesis⁽Reference Beenken and Mohammadi^361, Reference Kir, Beddow and Samuel³⁶²⁾. FGF19 is expressed in the distal small intestine, with the concentration of circulating FGF19 increasing in response to feeding. Transgenic mice expressing human FGF19 exhibit an increased metabolic rate and decreased adiposity despite having increased food intake with an increase in fatty acid oxidation⁽Reference Tomlinson, Fu and John^363, Reference Fu, John and Adams³⁶⁴⁾. However, in addition to its metabolic actions FGF19 also has proliferative effects, with transgenic mice developing hepatocellular carcinoma within 1 year, thereby rendering FGF19, a priori, unsuitable as a candidate for combating obesity⁽Reference Wu, Ge and Lemon³⁶⁵⁾.

Insulin-like growth factor system

Members of the insulin-like growth factor (IGF) system are functionally related to insulin. The IGF regulatory system consists of IGF (IGF-I and IGF-II), type I and type II IGF receptors, and IGF-binding proteins (IGFBP-1–6)⁽Reference Juul^367, Reference Kawai and Rosen³⁶⁸⁾. IGF are ubiquitously expressed, although the main source of circulating IGF-I is the liver. They exert actions in almost all tissues and are among the major regulators of growth⁽Reference Kawai and Rosen^368, Reference Baker, Liu and Robertson³⁶⁹⁾. While insulin is a short-term regulator of glucose homeostasis, IGF have been suggested to exert long-term regulation of glucose homeostasis⁽Reference Sandhu, Heald and Gibson^370–Reference Clemmons³⁷²⁾. Insulin and IGF-I show cross-reactivity at the receptor level. After ligand binding-induced autophosphorylation, insulin receptor and IGF-I receptor catalyse the phosphorylation of cellular proteins such as members of the insulin receptor substrate family⁽Reference Saltiel and Kahn³⁷³⁾. Other functions in which IGF play a critical role are the regulation of growth, neuroprotection, tumorigenesis and longevity⁽Reference Clemmons^372, Reference Holzenberger, Dupont and Ducos^374, Reference Sanchez-Alavez, Osborn and Tabarean³⁷⁵⁾. Adipose tissue levels of IGF-I have been shown to be higher in both rodent and human obesity⁽Reference Frystyk, Vestbo and Skjaerbaek³⁷⁶⁾, although the IGF-I-induced signalling cascade is impaired in obese mice⁽Reference Le Marchand-Brustel, Heydrick and Jullien³⁷⁷⁾.

IGF-I treatment by osmotic minipumps at adult age reduces hyperphagia, obesity, hyperinsulinaemia, hyperleptinaemia and hypertension in rats programmed to develop the metabolic syndrome by fetal programming⁽Reference Vickers, Ikenasio and Breier³⁷⁸⁾. However, IGF-I administration does not exhibit anorectic effects in sheep⁽Reference Foster, Ames and Emery³⁷⁹⁾. Singularly, another study reported that central injection of IGF-II, but not IGF-I, reduces short-term food intake in rats⁽Reference Lauterio, Marson and Daughaday³⁸⁰⁾. It has been recently reported that IGF-I may play an important role in thermogenesis⁽Reference Sanchez-Alavez, Osborn and Tabarean³⁷⁵⁾. Administration of IGF-I to the preoptic area, a hypothalamic region involved in the control of thermoregulation, produces hyperthermia involving activation of BAT in mice⁽Reference Sanchez-Alavez, Osborn and Tabarean³⁷⁵⁾. This thermogenic effect was accompanied by a switch from glycolysis to fatty acid oxidation and appears to be dependent of the insulin receptor, since it is absent in mice lacking the neuronal insulin receptor. These findings suggest a more important role of the IGF system in energy expenditure than previously thought⁽Reference Sanchez-Alavez, Osborn and Tabarean³⁷⁵⁾.

Although IGFBP are generally thought to inhibit the action of IGF through high-affinity binding which prevents interaction with IGF receptors, IGFBP can potentially either inhibit or enhance IGF actions⁽Reference Kawai and Rosen^368, Reference Firth and Baxter³⁸¹⁾. Overexpression of IGFBP2 by adenovirus prevents weight gain and hyperglycaemia in diet-induced obese mice⁽Reference Wheatcroft, Kearney and Shah³⁸²⁾. Moreover, it reverses diabetes at the same time as reducing food intake and inhibits body weight gain in insulin-resistant ob/ob mice by unexplored mechanisms⁽Reference Hedbacker, Birsoy and Wysocki³⁸³⁾.

Fibroblast growth factor 21

FGF21 is a pleiotropic hormone-like protein that has emerged as a major regulator of energy homeostasis⁽Reference Kharitonenkov and Larsen³⁸⁴⁾. Production of FGF21 takes place mainly in the liver⁽Reference Tyynismaa, Raivio and Hakkarainen³⁸⁵⁾ and is regulated by PPARα⁽Reference Badman, Pissios and Kennedy³⁸⁶⁾. FGF21 has been shown to be a major regulator of hepatic lipid metabolism in ketotic states, being up-regulated during fasting⁽Reference Badman, Pissios and Kennedy^386, Reference Galman, Lundasen and Kharitonenkov³⁸⁷⁾. FGF21 transgenic mice are resistant to diet-induced obesity, with FGF21 administration reducing serum glucose and TAG levels in obese and diabetic ob/ob and db/db mice⁽Reference Kharitonenkov, Shiyanova and Koester³⁸⁸⁾. In humans, FGF21 correlates with BMI and may be a novel biomarker for fatty liver⁽Reference Dushay, Chui and Gopalakrishnan³⁸⁹⁾. Since the expected beneficial effects of endogenous FGF21 for improving glucose tolerance and reducing TAG levels are absent in obese mice and men, obesity has been proposed to be a FGF21-resistant state⁽Reference Fisher, Chui and Antonellis³⁹⁰⁾.

Intraperitoneal injections or central administration of FGF21 increases energy expenditure, improves insulin sensitivity and reverses hepatic steatosis in diet-induced obese mice⁽Reference Xu, Lloyd and Hale³⁹¹⁾ and rats⁽Reference Sarruf, Thaler and Morton³⁹²⁾. The thermogenic effect may be related to the activation of BAT, since it has been reported that this adipose depot may be a source of FGF in response to cold, exhibiting an autocrine role in the stimulation of thermogenesis⁽Reference Hondares, Iglesias and Giralt³⁹³⁾. Although the wide interindividual variation in serum FGF21 observed in humans raises some doubts regarding its therapeutic relevance⁽Reference Kharitonenkov and Larsen³⁸⁴⁾, the reported metabolic effects of FGF21 highlight the need for more research in order to assess the use of FGF21 for treating metabolic diseases.

Sex hormone-binding globulin

Sex hormone-binding globulin (SHBG) transports androgens and oestrogens in blood and regulates their access to target tissues⁽Reference Hammond³⁹⁴⁾. SHBG is mainly produced by hepatocytes and its secretion fluctuates, being primarily influenced by metabolic and hormonal factors⁽Reference Hammond³⁹⁴⁾. Obesity results in reduced hepatic synthesis and blood concentrations of SHBG, with blood levels of SHBG correlating negatively with energy expenditure in postmenopausal women⁽Reference Svendsen, Hassager and Christiansen³⁹⁵⁾ and with fat accumulation in men⁽Reference Abate, Haffner and Garg³⁹⁶⁾. Moreover, low circulating concentrations of SHBG are a strong predictor of the risk of T2DM in men and women⁽Reference Ding, Song and Manson³⁹⁷⁾ as well as of the metabolic syndrome in non-obese men⁽Reference Kupelian, Page and Araujo¹⁹⁴⁾. This finding has been related to the suppressive effect of insulin on SHBG. Furthermore, the combined effect of increased levels of sex hormones previously mentioned, together with the reduced concentrations of SHBG, leads to an increase in the bioavailable androgens and oestrogens which may promote cellular proliferation and inhibit apoptosis in target cells, thereby being involved in the increased risk of cancer associated with obesity⁽Reference Calle and Kaaks³⁹⁸⁾.

Myostatin

Myostatin is a secreted member of the transforming growth factor-β (TGF-β) family that acts as a negative regulator of skeletal muscle growth by signalling through activin receptors⁽Reference Lee⁴⁰²⁾. During adulthood, the myostatin protein is produced by skeletal muscle, circulates in the blood, and limits muscle mass⁽Reference Lee⁴⁰²⁾. Myostatin is expressed almost exclusively in skeletal muscle, although detectable levels of myostatin mRNA are also present in adipose tissue. Myostatin overexpression in mice induces a dramatic loss of muscle and adipose tissue mass with normal food intake⁽Reference Zimmers, Davies and Koniaris⁴⁰³⁾. The loss of fat is concordant with the capacity of myostatin to block adipogenesis⁽Reference Rebbapragada, Benchabane and Wrana⁴⁰⁴⁾. As can be expected, mice lacking myostatin exhibit increased muscle mass but, surprisingly, show reduced adiposity⁽Reference McPherron and Lee⁴⁰⁵⁾. This finding can be explained by an increased fatty acid oxidation in peripheral tissues through the stimulation of enzymes involved in lipolysis and in mitochondrial fatty acid oxidation⁽Reference Zhang, McFarlane and Lokireddy⁴⁰⁶⁾. The decreased fat mass of myostatin-null mice can be further explained by a concomitant stimulation of thermogenesis through the activation of BAT⁽Reference Zhang, McFarlane and Lokireddy⁴⁰⁶⁾. Inhibition of myostatin signalling either in skeletal muscle or adipose tissue evidenced that body fat loss is an indirect result of metabolic changes in skeletal muscle⁽Reference Guo, Jou and Chanturiya⁴⁰⁷⁾, apparently mediated by increased energy expenditure and leptin sensitivity⁽Reference Choi, Yablonka-Reuveni and Kaiyala¹⁷⁾. Interestingly, skeletal muscle myostatin protein levels and plasma concentrations were higher in extremely obese women, with the former being correlated with the severity of insulin resistance⁽Reference Hittel, Berggren and Shearer⁴⁰⁸⁾. Leptin replacement increases muscle mass of ob/ob mice, and this effect is associated with a leptin-induced reduction in the skeletal muscle expression of myostatin⁽Reference Sáinz, Rodríguez and Catalán⁴⁰⁹⁾. However, leptin administration to hypoleptinaemic women did not decrease serum levels of myostatin, suggesting that leptin is not probably involved in the regulation of the myostatin axis in humans, although expression of myostatin in skeletal muscle was not measured⁽Reference Brinkoetter, Magkos and Vamvini⁴¹⁰⁾. Recently, the association of myostatin gene polymorphisms with obesity in humans has been reported, although the pathophysiological mechanisms remain to be elucidated⁽Reference Pan, Ping and Zhu⁴¹¹⁾.

Renin

As mentioned previously, an overactive renin–Ang system has been involved in the development of obesity-associated co-morbidities as well as in energy homeostasis⁽Reference Kalupahana and Moustaid-Moussa¹⁷⁴⁾. Renin catalyses the rate-limiting step of Ang II production⁽Reference Kalupahana and Moustaid-Moussa¹⁷⁴⁾. Mice lacking renin exhibit lower blood pressure and undetectable plasma levels of renin, Ang I and Ang II⁽Reference Takahashi, Lopez and Cowhig⁴¹²⁾. Unexpectedly, these mice are resistant to diet-induced obesity via an increased metabolic rate and partly through a gastrointestinal loss of dietary fat, but not from increased locomotor activity or reduced food intake⁽Reference Takahashi, Li and Hua¹⁴⁾. Some, but not all, of the observed alterations were reversed after Ang II administration. Furthermore, it has been reported that transgenic rodents overexpressing renin eat significantly more after 24 h than controls⁽Reference Szczepanska-Sadowska, Paczwa and Dobruch⁴¹³⁾ and develop obesity⁽Reference Uehara, Tsuchida and Kanno⁴¹⁴⁾ by mechanisms not related to Ang II⁽Reference Gratze, Boschmann and Dechend⁴¹⁵⁾. These findings suggest that renin inhibitors may be a therapeutic tool against obesity, insulin resistance and their cardiometabolic co-morbidities.

Atrial natriuretic peptides

ANP and brain natriuretic peptide (BNP) are synthesised in the heart and they are considered to exert their predominant effects in lowering blood pressure, controlling blood volume and reducing heart overgrowth in pathological conditions⁽Reference Beleigoli, Diniz and Ribeiro⁴¹⁷⁾. Another related peptide, C-type natriuretic peptide (CNP) is expressed mainly in the central nervous system but also in the vascular endothelial cells and chondrocytes⁽Reference Pandey⁴¹⁸⁾. ANP and BNP preferentially bind to GUCY-A, promoting the production of cGMP and the activation of protein kinase G⁽Reference Kuhn⁴¹⁹⁾. CNP is the physiological ligand for GUCY-B.

All members of the system (natriuretic peptides and their receptors) are expressed in adipose tissue, while their expression levels are altered in obesity⁽Reference Beleigoli, Diniz and Ribeiro^417, Reference Sarzani, Dessi-Fulgheri and Paci^420–Reference Moro and Smith⁴²³⁾. ANP and BNP, but not CNP, have been reported to induce potent lipolytic effects in human adipocytes similar to those exerted by the β-adrenoceptor agonist isoproterenol⁽Reference Lafontan, Moro and Berlan^360, Reference Birkenfeld, Boschmann and Moro⁴²⁴⁾. Moreover, ANP inhibits human visceral adipocyte growth in culture at physiological concentrations⁽Reference Sarzani, Marcucci and Salvi⁴²⁵⁾. ANP availability is decreased in obesity, with BMI being inversely correlated to circulating ANP and BNP concentrations⁽Reference Wang, Larson and Levy⁴²⁶⁾.

Besides their physiological role as lipid-mobilising agents, natriuretic peptides are also involved in the regulation of food intake and energy expenditure. CNP suppresses oxygen consumption in BAT in mice by attenuating the sympathetic nervous system activity, possibly under the control of the hypothalamus⁽Reference Inuzuka, Tamura and Yamada⁴²⁷⁾. Furthermore, it has been shown that natriuretic peptides can promote muscle mitochondrial biogenesis and fat oxidation, preventing the development of obesity and insulin resistance in mice⁽Reference Miyashita, Itoh and Tsujimoto⁴²⁸⁾. In this sense, intravenously administered ANP induces postprandial lipid oxidation in humans and increases energy expenditure⁽Reference Birkenfeld, Budziarek and Boschmann⁴²⁹⁾. These findings suggest that natriuretic peptides may represent a promising therapeutic tool for combating obesity and T2DM.

Thyroid hormones

The thyroid gland produces the parental form of thyroid hormone, thyroxine (T4), and lower amounts of the major active form of thyroid hormone, triiodothyronine (T3)⁽Reference Baxter and Webb⁴³¹⁾. T3 is produced by deiodination of T4 in target cells by specific deiodinases. Secretion of T4 by the thyroid gland is stimulated by thyroid-stimulating hormone secreted by the pituitary gland⁽Reference Yen⁴³²⁾. Thyroid hormones bind to thyroid hormone receptor (THR)α and THRβ, which are members of the nuclear hormone receptor family⁽Reference Zhang and Lazar⁴³³⁾. Thyroid hormones affect numerous cellular processes that are relevant for energy homeostasis⁽Reference Kim⁴³⁰⁾. Thyroid-stimulating hormone is usually moderately increased in obesity, which is a consequence rather than a cause of obesity⁽Reference Reinehr⁴³⁴⁾. Alterations in thyroid hormones affect body weight. Hypothyroidism is frequently associated with a modest weight gain, decreased metabolic rate and cold intolerance, whereas hyperthyroidism is related to weight loss despite increased appetite and elevated metabolic rate⁽Reference Reinehr^434, Reference Silva⁴³⁵⁾.

Hyperthyroidism in humans and rodents causes increased food intake but reduced body weight compared with euthyroid controls due to increased energy expenditure. The increase in oxygen consumption and body temperature is accompanied by enhanced fatty acid oxidation⁽Reference Baxter and Webb⁴³¹⁾. Evidence of the critical role of thyroid hormones on energy homeostasis arises from genetic mouse models lacking THR⁽Reference Alkemade^436, Reference Barros and Gustafsson⁴³⁷⁾. Mice lacking THRα⁽Reference Park, Zhao and Glidewell-Kenney⁴³⁸⁾ or THRβ⁽Reference Foryst-Ludwig, Clemenz and Hohmann⁴³⁹⁾ exhibit reduced thermogenesis as well as other metabolic alterations. Although the involvement of thyroid hormones in energy homeostasis is critical, the physiological mechanisms explaining this effect remain elusive. The thermogenic effect of thyroid hormones has been related to accelerated ATP turnover and reduced efficiency of ATP synthesis as well as to changes in the efficiency of metabolic processes downstream from the mitochondria ‘futile and substrate cycles’⁽Reference Kim^430, Reference Silva⁴³⁵⁾. Peripheral administration of T3 increases food intake but also energy expenditure⁽Reference Dillo⁴⁴⁰⁾. An increase in hypothalamic AMPK may be mediating the orexigenic effect of T3⁽Reference Ishii, Kamegai and Tamura⁴⁴¹⁾. Similarly, it has been recently described that besides the critical role of T3 stimulating thermogenesis in skeletal muscle⁽Reference Kim⁴³⁰⁾, thyroid hormone-induced modulation of AMPK activity and lipid metabolism in the hypothalamus and subsequent thermogenic activation of BAT is a major regulator of whole-body energy homeostasis⁽Reference López, Varela and Vázquez⁴⁴²⁾. Although the information available makes the thyroid system an interesting field for the development of therapeutic drugs in the fight against obesity, available data regarding effectiveness of thyroid hormone therapy for treating obesity are inconclusive⁽Reference Kaptein, Beale and Chan⁴⁴³⁾.

Osteopontin

Osteopontin, also known as secreted phosphoprotein-1, bone sialoprotein-1 and early T lymphocyte activation (Eta-1) is a phosphoprotein expressed by a wide variety of cell types, such as osteoclasts, macrophages, hepatocytes and vascular smooth muscle cells, among others⁽Reference Scatena, Liaw and Giachelli⁴⁴⁷⁾. Osteopontin exerts important actions on bone turnover, serving as attachment for osteoclasts activating resorption⁽Reference Scatena, Liaw and Giachelli^447, Reference Reinholt, Hultenby and Oldberg⁴⁴⁸⁾. In addition to bone remodelling, osteopontin is also involved in several pathophysiological processes including inflammation, immunity, neoplastic transformation, progression of metastases, wound healing and cardiovascular function⁽Reference Scatena, Liaw and Giachelli⁴⁴⁷⁾. Osteopontin has been shown to be also produced by adipocytes and to play an important role in obesity and obesity-associated insulin resistance⁽Reference Gómez-Ambrosi and Frühbeck^39, Reference Gómez-Ambrosi, Catalán and Ramírez^449–Reference Chapman, Miles and Ofrecio⁴⁵³⁾. Expression of osteopontin is dramatically increased in adipose tissue from obese individuals⁽Reference Gómez-Ambrosi, Catalán and Ramírez^449, Reference Pietiläinen, Naukkarinen and Rissanen^454, Reference Hurtado Del Pozo, Calvo and Vesperinas-Garcia⁴⁵⁵⁾, suggesting the important role that this protein has in the molecular alterations that take place during adipose tissue enlargement. Moreover, osteopontin has been suggested to play a pivotal role linking obesity to insulin resistance development by promoting inflammation and the accumulation of macrophages in adipose tissue⁽Reference Nomiyama, Perez-Tilve and Ogawa^450, Reference Chapman, Miles and Ofrecio^453, Reference Kiefer, Zeyda and Gollinger⁴⁵⁶⁾. Higher levels of expression of osteopontin have been shown to be related to macrophage accumulation in adipose tissue and to liver steatosis in morbidly obese subjects⁽Reference Bertola, Deveaux and Bonnafous⁴⁵⁷⁾, which promote fibrosis progression in non-alcoholic steatohepatitis⁽Reference Syn, Choi and Liaskou⁴⁵⁸⁾. Furthermore, it has been recently shown that osteopontin expression is increased in omental adipose tissue of colon cancer patients, suggesting a potential role of osteopontin linking increased inflammation in visceral adiposity with neoplastic processes⁽Reference Catalán, Gómez-Ambrosi and Rodríguez⁴⁵⁹⁾.

Osteocalcin

Osteocalcin is a non-collagenous protein marker of osteoblastic activity thought to play a role in mineralisation and Ca homeostasis⁽Reference Calvo, Eyre and Gundberg⁴⁶⁰⁾. Osteocalcin is secreted mainly by osteoblasts and its levels decrease in malnutrition, starvation and anorexia nervosa. Osteocalcin has been traditionally considered as a biological marker of bone formation⁽Reference Calvo, Eyre and Gundberg⁴⁶⁰⁾ but now has also been shown to play a role in the regulation of metabolism and in the development of CVD⁽Reference Kanazawa, Yamaguchi and Yamamoto^461–Reference Ducy⁴⁶³⁾.

The regulation of bone remodelling by leptin led to the hypothesis that bone exerts a role in the feedback control of energy homeostasis⁽Reference Lee, Sowa and Hinoi⁴⁶⁴⁾. In this sense, mice lacking the osteoblast-secreted molecule osteocalcin exhibit an increased adiposity and insulin resistance⁽Reference Lee, Sowa and Hinoi^464, Reference Martin⁴⁶⁵⁾. Osteocalcin is able to improve glucose tolerance in vivo through the stimulation of the expression of insulin and β-cell proliferation and the induction of the expression of adiponectin and genes involved in energy expenditure in adipocytes⁽Reference Lee, Sowa and Hinoi^464, Reference Ferron, Hinoi and Karsenty⁴⁶⁶⁾. In this sense, administration of osteocalcin improves glucose metabolism and prevents the development of T2DM in mice⁽Reference Ferron, Hinoi and Karsenty^466, Reference Ferron, McKee and Levine⁴⁶⁷⁾.

Serum osteocalcin levels are positively associated with insulin sensitivity and secretion in non-diabetic subjects as well as in patients with type 2 diabetes⁽Reference Fernández-Real and Ricart^462, Reference Kanazawa, Yamaguchi and Tada⁴⁶⁸⁾. Interestingly, osteocalcin concentrations are reduced in obese individuals and increase after weight loss in parallel to the reduction in visceral fat mass⁽Reference Fernández-Real, Izquierdo and Ortega⁴⁶⁹⁾. Finally, a recent work suggests that further molecules secreted by bone, yet to be identified, affect energy metabolism⁽Reference Yoshikawa, Kode and Xu⁴⁷⁰⁾.