The gut hormone secretin triggers a gut–brown fat–brain axis in the control of food intake

What is the topic of this review? Brown fat's role in meal‐associated thermogenesis and the related consequences for energy balance regulation with a focus on the gut hormone secretin, which has been identified as the endocrine molecular mediator of meal‐associated brown fat thermogenesis. What advances does it highlight? The finding of the secretin‐induced gut–brown fat–brain axis creates new opportunities to manipulate brown fat and thereby energy balance in a natural way while living in a thermoneutral environment. The role of brown fat as a mere catabolic heater organ needs to be revised and more attention should be directed towards the regulatory role of brown fat beyond energy expenditure.

inner mitochondrial membrane of brown adipocytes. In 2009, seminal reports on metabolically active brown fat in adult humans largely stimulated the interest in this thermogenic tissue (Cypess et al., 2009;Saito et al., 2009;van Marken Lichtenbelt et al., 2009;Virtanen et al., 2009). It was found that acute cold exposure activates human brown fat. Cold-induced activation of brown fat is elicited by the binding of noradrenaline to -adrenergic receptors, which triggers the canonical cAMP-protein kinase A (PKA)-lipolysis pathway. Therefore, sympathomimetic drugs were evaluated early on for pharmacological activation of human brown fat. Unfortunately, neither non-specific sympathomimetics nor selective agonists for the relatively fat-specific 3 -adrenergic receptor (AR) can stimulate human brown fat without marked effects on the cardiovascular system Cypess et al., 2015;Vosselman et al., 2012). Thus, druggable nonadrenergic receptors in brown fat that activate this heating organ are in demand to attenuate body fat accumulation by increasing resting energy expenditure in humans.
Other than cold-induced thermogenesis, resting metabolic rate is also increased in association with a meal, known as the specific dynamic action or thermal effect of food. Although this effect is partially due to the obligatory costs of digestion and resorption, a facultative component mediated by brown fat has been discussed for a long time (Glick, Teague, & Bray, 1981;Rothwell & Stock, 1979).
A recent study demonstrated that a single mixed meal activates human brown fat to the same degree as cold exposure (U Din et al., 2018). While obligatory costs for food digestion and resorption are inevitable and comprehensible, the existence of a facultative component is still debated, including the source of thermogenesis (brown fat-dependent or not), the functional significance, and the potential mediators (Cannon & Nedergaard, 2004;Kozak, 2010). In this context, the hypothesis of thermoregulatory feeding proposed that during a meal brown fat thermogenesis may serve as a feedback signal to the brain to control meal initiation and termination (Brobeck, 1948;Himms-Hagen, 1995). Thus, aside from the mere capacity to burn excess calories, brown fat would play a role in the control of energy intake (Glick, 1982;Himms-Hagen, 1995). In rats, brown fat temperature increases about 15 min prior to meal initiation, and drops upon meal termination (Blessing, Mohammed, & Ootsuka, 2013). An increased periprandial tone of the sympathetic nervous system was suggested to activate thermogenesis by the release of noradrenaline from the sympathetic nerves in brown fat, which may lead to meal termination. Accordingly, acute pharmacological activation of brown fat in fasted mice using a 3 -AR agonist reduced cumulative food intake during a refeeding trial (Grujic et al., 1997;Susulic et al., 1995).

New Findings
• What is the topic of this review?
Brown fat's role in meal-associated thermogenesis and the related consequences for energy balance regulation with a focus on the gut hormone secretin, which has been identified as the endocrine molecular mediator of mealassociated brown fat thermogenesis.
• What advances does it highlight?
The finding of the secretin-induced gut-brown fat-brain axis creates new opportunities to manipulate brown fat and thereby energy balance in a natural way while living in a thermoneutral environment. The role of brown fat as a mere catabolic heater organ needs to be revised and more attention should be directed towards the regulatory role of brown fat beyond energy expenditure. Therefore, so far unknown triggers other than the SNS are likely to contribute.
As the release of gastrointestinal peptides is one of the first physiological responses to eating, periprandially secreted gut hormones are potential evokers of meal-associated brown fat thermogenesis. All along the gastrointestinal tract, specialized endocrine cells produce peptide hormones, making the gut one of the largest endocrine organ in the body (Coate, Kliewer, & Mangelsdorf, 2014). During a meal, secretion of gut hormones initiates complex neuroendocrine responses encoding information on the nutritional status (Cummings & Overduin, 2007). Gut hormones not only act locally to orchestrate gastrointestinal motility, secretion, digestion and nutrient absorption, but also promote the central perception of satiation via central neuronal circuits in the brain controlling food intake and limiting meal size (Chaudhri, Small, & Bloom, 2006). Most prominently, glucagon-like peptide (GLP-1), cholecystokinin (CCK), oxyntomodulin, peptide YY (PYY) and secretin inhibit food intake as demonstrated in various animal experiments, whereas ghrelin promotes hunger (Batterham et al., 2002;Chaudhri et al., 2006;Cheng, Chu, & Chow, 2011;Dakin et al., 2004;Gibbs, Young, & Smith, 1973;Turton et al., 1996;Weller, Smith, & Gibbs, 1990). Indeed, some of these gut hormones activate brown fat through their effects on the efferent SNS tone, such as CCK and GLP-1 (Beiroa et al., 2014;Blouet & Schwartz, 2012). Some gut hormones also modulate fat metabolism in white adipocytes, which is of particular interest as the mobilization of free fatty acids is an essential prerequisite for the activation of UCP1. For example, PYY inhibits lipolysis via Y2-receptor (Y2R) coupling to signalling by G i (Valet et al., 1990), which is a specific isoform of the G-proteinsubunit initiating a signalling cascade that inhibits adenylyl cyclase and thereby decreases intracellular cAMP levels. Secretin, on the other hand, stimulates lipolysis through the secretin receptor (SCTR) coupling to G S signalling in white adipocytes by activation of adenylyl cyclase and rising cAMP levels (Butcher & Carlson, 1970 & Chow, 2014). None of these gut hormones had been demonstrated to directly activate brown fat thermogenesis during a meal.

SECRETIN -THE ENDOCRINE MOLECULAR MEDIATOR OF MEAL-ASSOCIATED BROWN FAT THERMOGENESIS
We recently revealed a novel endocrine gut-brown fat-brain axis triggered by secretin release from the intestine during a meal . Our transcriptome analysis of murine brown fat demonstrated that SCTR is abundantly expressed, while receptors for other gastrointestinal peptides are absent. Therefore, the gut hormone secretin, which is classically known as a stimulant of pancreatic water and bicarbonate secretion upon food intake, represented the top candidate in our search for novel endocrine mediators of meal-associated brown fat thermogenesis. Indeed, in primary brown adipocytes, secretin increased UCP1-dependent respiration.
Further detailed in vitro analysis established that secretin initiates the canonical SCTR-cAMP-PKA-lipolysis-UCP1 pathway in brown adipocytes. In vivo, the thermogenic effect of secretin was consolidated by indirect calorimetry and multispectral optoacoustic tomography in three different mouse models. The thermogenic effect was (i) UCP1dependent, (ii) comparable to noradrenaline, (iii) significantly present at room temperature and thermoneutrality, and (iv) accompanied by a rise in the temperature of interscapular brown fat (T iBAT ), the largest brown fat depot in mice (Cinti, 2005). In addition, plasma secretin levels were decreased by fasting and increased significantly within 1 h after refeeding, which was congruent with changes in T iBAT . Furthermore, a single secretin injection prior to refeeding of fasted mice reduced food intake in a UCP1-dependent manner, which was reflected by the regulation of anorexigenic and orexigenic hypothalamic peptides in wild-type, but not UCP1-knockout (KO) mice. This effect on food intake was induced by a direct activation of brown fat, independent of the SNS, as assessed by pretreatment with propranolol, which blocks the sympathetic neuronal input to brown adipose tissue. We reasoned that any signal working through the brain from activation of secretin receptors either in vagal afferents or hypothalamic nuclei will be blocked. As pretreatment with propranolol did not alter secretin's effect on brown fat activity and food intake, but completely abolished the effect of the 3 -adrenergic receptor agonist CL-316,243, we exclude a gut-brain-SNS-brown fat route. Our findings are in contrast to a previous study reporting that the anorexigenic action of secretin depends on the activation of secretin receptors in vagal sensory nerves and melanocortin signalling in the brain (Cheng et al., 2011).
In further experiments we substantiated the importance of mealinduced endogenous secretin release for the function of the endocrine gut-brown fat-brain axis by antibody-based neutralization of secretin activity. In mice refed after an overnight fasting, neutralization caused attenuation of the meal-associated rise in T iBAT as well as an increase in cumulative food intake. The link between secretin-induced brown fat thermogenesis and satiation was underlined by a negative correlation of food intake and meal-associated rise in T iBAT . Meal-pattern analyses revealed that the interplay of secretin and brown fat regulates food intake in mice. Although total food intake was not altered in the absence of either secretin or intact brown fat, the number of meals per night and inter-meal bout lengths were decreased, while meal size and meal duration were increased. Taken together, our findings demonstrate that secretin mediates a gut-brown fat-brain axis in the control of satiation. Conclusively, targeting this endocrine axis might hold promise for developing novel obesity therapies as it promotes negative energy balance through both increasing energy expenditure and decreasing energy intake ( Figure 1). In addition to elucidating a novel mechanism of satiation, these findings have a number of interesting implications. First, they may explain the presence of functional brown fat in humans and many even larger mammalian species, for which allometric modelling of thermogenic brown fat capacity (Heldmaier, 1971) and comparative genomics of Ucp1 (Gaudry, Campbell, & Jastroch, 2019) exclude nonshivering thermogenesis in brown fat. Thus, in humans, heat dissipated by brown fat may serve a regulatory role, rather than a homeostatic role in thermoregulation. Second, these findings qualify brown fat as an even more attractive therapeutic target that not only increases energy expenditure but also reduces food intake at the same time. Based on these findings one may hypothesize that any brown fat activating stimulus could potentially induce satiation. By manipulating both sides of the energy balance at the same time, one of the most challenging difficulties during weight loss interventions, i.e. compensation (Hall et al., 2012), could be overcome.

THE MISERY OF WEIGHT LOSS
In the course of evolution, humans and their hominin ancestors developed sophisticated physiological mechanisms to expand and defend their body fat stores to survive periods of famine. In our modern times with many people significantly exceeding a healthy expansion of body fat, these physiological mechanisms rather prove to be a health risk than an evolutionary advantage. Lowering energy intake, whether voluntarily by lifestyle changes or assisted by pharmacological interventions, is inevitably counteracted by a reduction in metabolic rate to defend the current body mass through energy homeostatic systems (Mole, 1990

Meal composition
Cold living habits F I G U R E 1 Targeting brown fat for energy balance regulation. Classically, brown fat is activated in response to cold exposure by noradrenaline from the sympathetic nervous system. Additionally, brown fat is activated in association with a meal by the gut hormone secretin. Upon activation, brown fat increases energy expenditure and decreases energy intake. Thus, it has an overall negative impact on energy balance. The augmentation of prandial released secretin, e.g. by secretagogues or cleavage inhibitors, represents besides sympathomimetics an alternative to pharmacologically activating brown adipose tissue the same time counteracts hyperphagia would be the first-line therapy for the treatment of metabolic diseases. The adipocyte-derived protein hormone leptin, a key factor in energy balance regulation (Allison & Myers, 2014;Gautron & Elmquist, 2011), may be regarded as a showcase for such a factor. Leptin promotes satiety and prevents the lowering of energy expenditure in response to caloric restriction (Bolze et al., 2016;Doring, Schwarzer, Nuesslein-Hildesheim, & Schmidt, 1998). Despite these promising endocrine actions, leptin treatment of people with obesity was not successful, most likely due to leptin resistance (Rosenbaum & Leibel, 2014). In this respect, we regard secretin, perhaps in combination with other gastrointestinal hormones and leptin, as a promising candidate.

ACHIEVING BROWN FAT ACTIVATION WHILE LIVING IN A COMFORTABLE THERMONEUTRAL ENVIRONMENT
Several novel molecular mediators for the recruitment of brown fat and/or the browning of white adipose tissue have been identified (Bartelt & Heeren, 2014). The mere augmentation of thermogenic capacity, however, is only one step in targeting brown fat for therapeutic purposes in obesity prevention and treatment because UCP1 is constitutively inactive in brown adipocytes (Li, Fromme, Schweizer, Schottl, & Klingenspor, 2014). Classically, increased sympathetic tone in response to cold exposure will trigger intracellular signalling events that mobilize fatty acids and activate uncoupled respiration. However, humans spend most of their time in thermoneutral environments equipped with heating systems and put on warm clothing which normally precludes cold-induced brown fat thermogenesis. Even though cold acclimation can recruit cold-inducible heating capacity in human brown fat, this capacity will not be utilized while remaining in thermoneutrality. Cold mimetics or other pharmacological molecules to stimulate brown fat would be desirable, but so far, only a few activators of UCP1-mediated thermogenesis have been identified (Braun, Oeckl, Westermeier, Li, & Klingenspor, 2018). Thus, gut-derived secretin as a novel direct brown fat activator is of prime interest. With three to four meals ingested in a regular day, it is likely that periprandial secretion of secretin triggers repeated activation of brown fat thermogenesis according to daily meal patterns. This implies more frequent bouts of brown fat activation in a thermoneutral environment than previously anticipated. Moreover, it has been recently shown that brown adipose tissue volume and activity, measured by 18 F-FDG, are not associated with energy intake and meal-induced appetite-related sensations in young, healthy adults (Sanchez-Delgado et al., 2020). We hypothesize, that meal-associated thermogenesis determines energy intake and appetite, rather than cold-induced activation. From an evolutionary perspective, this would prevent a limitation of energy intake during cold exposure when brown adipose tissue is activated to survive in the cold. The intensity of meal-associated brown fat activity may depend on caloric intake and meal type. Future studies will need to address the effect size and duration of thermogenic action of endogenous secretin release on brown fat.

MANIPULATING MEAL-ASSOCIATED BROWN FAT THERMOGENESIS TO ACHIEVE WEIGHT LOSS
A recently developed novel optoacoustic imaging tool for the noninvasive assessment of metabolic processes without using contrast agents could be applied to determine meal-associated brown fat thermogenesis in humans. This technique employs haemoglobin as an intrinsic tissue biosensor and resolved oxygen utilization (rate of tissue oxygen saturation) as a metabolic indicator, enabling the labelfree measurement of brown fat activation in mice and human subjects . Characterizing the physiological underpinnings Alternatively, the prandial surge of endogenous secretin could be boosted in a timely manner to promote satiation. Secretagogues or cleavage-inhibitors of secretin may efficiently accelerate meal termination and thereby reduce caloric intake by increasing the yield of prandial secretin (Klingenspor, 2019). Since duodenal acidification is the primary stimulus for secretin release from S-cells in the epithelium of the duodenum into the circulation, meal composition could also be a crucial determinant for secretin release and subsequent brown fat activation. Therefore, nutritional interventions with meals tailored to boost prandial secretin release would be the most natural and least invasive approach to promote satiation ( Figure 1). Conversely, one would assume that a reduction of duodenal acidification results in lower secretin levels and thus attenuated meal-associated activation of brown fat. Proton pump inhibitors (PPI) and histamine H2 receptor antagonists are groups of medications whose main action is a pronounced and long-lasting reduction of gastric acid secretion. Interestingly, PPI use has been reported to be associated with a significant weight gain in men, while energy intake, physical activity and sedentary behaviour were unchanged (Czwornog & Austin, 2015;Yoshikawa, Nagato, Yamasaki, Kume, & Otsuki, 2009). Furthermore, it has been demonstrated that children prescribed with PPIs and H2 receptor antagonists were slightly more likely to develop obesity, with increasing manifestation depending on medication duration (Stark, Susi, Emerick, & Nylund, 2019). Although most of these effects are attributed to changes in gut microbiota, the impact on secretin levels, meal-associated brown fat thermogenesis and food intake needs to be further investigated.

DIMENSIONS OF BROWN FAT'S CONTRIBUTION TO ENERGY EXPENDITURE AND WEIGHT LOSS IN THE LONG RUN
When targeting brown fat for obesity treatment it has to be taken into account that the incidence of active brown fat is altered by age, sex, body mass index, plasma glucose, season, outdoor temperature and medication Hanssen et al., 2015;Lee, Greenfield, Ho, & Fulham, 2010;Nedergaard & Cannon, 2010;Ouellet et al., 2011;Persichetti et al., 2013;Skillen, Currie, & Wheat, 2012).
Although there is some evidence, that the prevalence of brown fat is not affected by body mass index, there seems to be no doubt that active brown fat declines with age (Gerngross, Schretter, Klingenspor, Schwaiger, & Fromme, 2017). Thus, not only must the thermogenic process be activated, but the total mass and oxidative capacity of brown fat in the human body must be increased. To date it is unclear whether sufficient brown fat capacity for energy expenditure can be recruited in people with obesity to reduce their body fat mass effectively (Marlatt & Ravussin, 2017 Oxygen consumption (U Din et al., 2016); BAT mass (Leitner et al., 2017); 20 kJ (l O 2 ) −1 (Leonard, 2010); AT density 0.925 g.ml −1 (Martin et al., 1994) Norman et al., 2003). However, there is no question that much more attention should be directed to the regulatory role of brown fat beyond energy expenditure.
In conclusion, the identification of the secretin-driven gut-brown fat-brain axis gives an impetus to rethink the role of brown fat as a mere catabolic heating organ and qualifies it as an even more attractive target for the treatment of obesity. Any stimulus of brown fat may potentially induce satiation, but may also increase resting energy expenditure without causing unwanted side effects. Further studies on the function and sensitivity of the endocrine gut-brown fat-brainaxis in people with obesity are urgently needed to exploit the potential of brown fat to impact the complex and redundant system of energy balance controlling hunger and satiety, energy partitioning and energy expenditure and promote metabolic health.

COMPETING INTERESTS
The Technical University of Munich has applied for a patent (PCT/EP2017/062420).