Interplay between diet, exercise and the molecular circadian clock in orchestrating metabolic adaptations of adipose tissue

Abstract

Disruption of circadian rhythmicity induced by prolonged light exposure, altered sleep patterns and shift work is associated with the development of obesity and related metabolic disorders, including type 2 diabetes and cardiovascular diseases. White and brown adipose tissue activity shows circadian rhythmicity, with daily variations in the regulation of metabolic processes such as lipolysis, glucose and lipid uptake, and adipokine secretion. The role of the circadian clock in the regulation of energy homeostasis has raised interest in clock‐related strategies to mitigate metabolic disturbances associated with type 2 diabetes, including ‘resynchronizing’ metabolism through diet or targeting a particular time of a day to potentiate the effect of a pharmacological or physiological treatment. Exercise is an effective intervention to prevent insulin resistance and type 2 diabetes. Beyond its effect on skeletal muscle, exercise training also has a profound effect on adipose tissue. Adipose tissue partly mediates the beneficial effect of exercise on glucose and energy homeostasis, via its metabolic and endocrine function. The interaction between zeitgeber time and diet or exercise is likely to influence the metabolic response of adipose tissue and therefore impact the whole‐body phenotype. Understanding the impact of circadian clock systems on human physiology and how this is regulated by exercise in a tissue‐specific manner will yield new insights for the management of metabolic disorders.

Introduction

The circadian clock is a key homeostatic regulator that is synchronized by photic and non‐photic stimuli and controls many genomic and physiological responses (Eckel‐Mahan & Sassone‐Corsi, 2013; Gerhart‐Hines & Lazar, 2015). The circadian programme is regulated at both a central and a peripheral level, with the suprachiasmatic nucleus (SCN) in the hypothalamus acting as a master clock that synchronizes peripheral oscillators in most cells of the body. On a molecular level, the circadian clock is composed of an auto‐regulatory transcription–translation feedback loop with the transcriptional activators CLOCK and BMAL1 and their target genes that encode Period (PER) and cryptochrome (CRY), which form a repressor complex, and reverse‐erb α (REV‐ERBα) that together inhibit CLOCK and BMAL1 transcription activity (Fig. 1). A variety of stimuli including the timing of food intake, dietary composition, temperature and the level of physical activity can alter the activity of circadian clocks in peripheral tissues, highlighting the interplay between circadian rhythm and metabolism (Gabriel & Zierath, 2019). Disrupted circadian rhythms lead to insulin resistance and increased risk for the development of type 2 diabetes (Eckel‐Mahan & Sassone‐Corsi, 2013; Gerhart‐Hines & Lazar, 2015; Bescos et al. 2018; Vetter et al. 2018). Thus, synchronizing exercise or nutrient strategies to the molecular circadian clock may maximize the health‐promoting benefits of these interventions to enhance insulin sensitivity and prevent type 2 diabetes.

Figure 1.

Circadian regulation of metabolic and endocrine function of adipose tissue and the contribution to glucose and lipid oscillations in plasma

White adipocyte tissue lipolysis and lipid storage undergo diurnal oscillations. Adipokines such as adiponectin and leptin are released into circulation in a circadian pattern and peak at different times of the day. Brown adipose tissue displays a circadian rhythm of glucose and fatty acid uptake, with higher thermogenesis activity in the active phase. On the molecular level, the intrinsic clock influences adipocyte function via direct regulation of the expression of several genes, including the lipolytic enzymes ATGL and HSL, the fat storage factor PPARγ and the thermogenic protein UCP1. This schematic representation is mainly derived from experimental evidence from mouse models and therefore it requires additional corroboration in humans. The peak in leptin level in humans is identified by the red triangle.

White and brown adipose tissue play an important role in the regulation of lipid metabolism, whole‐body energy homeostasis and thermogenesis. Adipose tissue integrates several external stimuli, including nutrients and hormones, which modulate its function in a diurnal pattern. Accordingly, adipose tissue dysfunction is central to the development of obesity and type 2 diabetes, and other related metabolic disorders. Regular physical exercise is a potent modulator of glucose and energy homeostasis by affecting cellular processes in a variety of peripheral tissues including skeletal muscle, liver and adipose tissue (Hawley et al. 2014). While skeletal muscle is one of the main tissues directly affected by physical activity, exercising skeletal muscle ‘communicates’ with adipose tissue to improve whole‐body glucose and energy homeostasis (Townsend & Wright, 2019). The role of adipose tissue in mediating the health‐promoting effects of exercise is a topic of emerging interest. Intimate links between gene regulation and the circadian clock are likely to contribute to the plasticity of a variety of insulin‐sensitive organs in response to exercise and nutritional interventions. This review focuses on the circadian regulation of adipose tissue metabolism, and the interplay between diet, exercise and the circadian clock in orchestrating the adipose tissue adaptation and, in a more integrative way, the consequences of these factors for the whole‐body metabolism.

White and brown adipose tissues contribute to the circadian regulation of energy homeostasis

White adipose tissue activity shows circadian rhythmicity, with daily variations in the expression of genes involved in metabolic processes such as lipid metabolism and energy expenditure (Zvonic et al. 2006; Shostak et al. 2013a). The integration of several stimuli, including nutrients and hormones, regulate the balance between substrate storage and mobilization in adipose tissue throughout the day, consistent with the oscillation of the circulating levels of free fatty acids (FFAs) and triglycerides (TGs). However, growing evidence suggests that the endogenous adipose clock plays an important role in lipid metabolism. Indeed, white and brown adipose tissue gene expression of the components of the core clock machinery is cyclic, with a significant percentage of genes displaying associated coordinated circadian expression profiles (Zvonic et al. 2006). The key enzymes of lipolysis, including adipose triglyceride lipase (ATGL) and hormone‐sensitive lipase (HSL), are under the transcriptional control of BMAL1–CLOCK and the modulation of these transcriptional activators ultimately affects the plasma FFA level in transgenic mouse models (Shostak et al. 2013b). Of note, the increase in mRNA level of ATGL and HSL precedes the elevation of the plasma FFA concentration by several hours, suggesting that in vivo, the clock is likely responsible for the fine‐tuning of the regulation of lipolysis, while acute induction of lipolysis is activated by phosphorylation events under the control of hormonal cues (Duncan et al. 2007). PER2 directly inhibits the activity of peroxisome proliferator‐activated receptor γ (PPARγ), a master regulator of adipogenesis and promotor of lipid storage (Grimaldi et al. 2010). Metabolomic studies in humans reveal that many circulating metabolites, with the majority being lipid products, show circadian oscillations, pointing to a specific role of the endogenous circadian clocks in the regulation of the plasma lipid profile (Dallmann et al. 2012; Skene et al. 2018). In addition to a metabolic role, white adipose tissue contributes to energy homeostasis through the production of adipokines. For example, leptin, an adipokine that regulates food intake and energy expenditure, is released in a pulsatile pattern with a peak in the early non‐active phase in humans (Heptulla et al. 2001). Accordingly, decreased leptin pulsatility has been suggested to contribute to leptin resistance and the development of obesity (Arble et al. 2011). Conversely, adiponectin, an adipokine implicated in the regulation of insulin sensitivity and fatty acid (FA) oxidation, peaks in the early active phase. The diurnal pattern of adipokine release may contribute to the withdrawal of circulating FAs in the active phase, and the control of glucose and energy homeostasis (Gomez‐Abellan et al. 2010; Froy & Garaulet, 2018). Adipokine levels also vary in response to environmental factors such as light–dark and feeding–fasting cycles that typically occur in time‐of‐day‐dependent patterns (Shea et al. 2005). Thus, adipokine regulation can provide a feedback loop between the systemic environment and adipose tissue to regulate whole‐body metabolism.

In the last decade, brown adipose tissue has re‐emerged as an important organ for the regulation of glucose and energy homeostasis. Brown adipose tissue activity is rapidly modulated throughout the course of a day, thereby affecting thermogenic plasticity and influencing whole‐body metabolism. Fluorodeoxyglucose positron emission tomography (PET) imaging analysis of anaesthetized C57Bl/6 male and female mice over the course of the light–dark cycle reveals a diurnal pattern of glucose uptake in brown adipose tissue (van der Veen et al. 2012), possibly driven by glucocorticoids (Ramage et al. 2016). In humans, brown adipose tissue explants display circadian rhythmicity of the gene expression of insulin‐sensitive glucose transporter protein 4 (GLUT4) and the mitochondrial uncoupling protein 1 (UCP1), with increased glucose uptake in brown adipose tissue in the early morning as assessed in vivo using PET–computed tomography scanning (Lee et al. 2016). Although the diurnal variation in brown adipose tissue activity did not influence blood glucose oscillations, a modest correlation was detected with the supraclavicular temperature response and glucose excursions among individuals with high brown adipose tissue abundance (Lee et al. 2016). In male C57Bl/6J mice, brown adipose tissue also displays a diurnal rhythm in FA uptake, with increased TG‐derived FA uptake at the onset of the active period (van den Berg et al. 2018). These results indicate that photic cues and modulation of the light exposure duration influence brown adipose tissue activity in part via the central circadian clock machinery. Indeed, TG‐derived FA uptake in brown adipose tissue and thermogenic activity has been previously shown to be under the direct control of the SCN via the nuclear factor REV‐ERBα (Gerhart‐Hines et al. 2013; Moran‐Ramos et al. 2017). While the circadian brown adipose tissue activity predicts the variation in plasma TG and FA level, and determines the postprandial handling of TGs in mice, this finding has not been corroborated in humans (van den Berg et al. 2018).

Disruption of the clock alters adipose tissue metabolism

Disruption of circadian rhythmicity induced by prolonged light exposure, altered sleep patterns and shift working is associated with the development of obesity and related disorders, including type 2 diabetes and cardiovascular diseases (Karlsson et al. 2001). Considering the important role of adipose tissue in the regulation of glucose and lipid metabolism, misalignment or perturbation of the internal adipose clock is a plausible mechanism for the development of metabolic disorders. In rodents, advance phase shift or time‐restricted sleeping increases the expression of lipogenic genes in white adipose tissue (Husse et al. 2012; Herrero et al. 2015). Moreover, prolonged light exposure in mice leads to obesity by decreasing brown adipose tissue activity via impaired adrenergic signalling (Kooijman et al. 2015). In humans, forced dissynchrony of the clock by altered eating and sleeping patterns to mimic conditions of acute jet lag or chronic shift work decreases plasma leptin levels, increases glucose and insulin levels, and disrupts the cortisol rhythm (Scheer et al. 2009). These changes in endocrine function indicate that circadian misalignment can affect adipose tissue metabolism, which if maintained, could lead to the development of obesity.

A role for the clock machinery in adipose tissue function is directly supported from studies of animal models in which components of the core clock machinery have been mutated or ablated. Clock‐mutant Clock^Δ19 mice exhibit a loss of rhythmicity in mRNA expression of adipose tissue lipogenic and lipolytic genes, leading to decreased lipolysis, increased adipose tissue mass and adipocyte hypertrophy (Shostak et al. 2013b). Studies of Rev‐Erbα knockout mice link circadian and thermogenic networks together by regulating brown adipose tissue function. Indeed, UCP1 rhythmicity was abolished and its protein abundance was persistently upregulated in Rev‐Erbα ^−/− mice, concomitant with increased glucose uptake and thermogenic activity (Gerhart‐Hines et al. 2013). Further compelling evidence supporting a role of the clock machinery in adipose function has emerged from the generation of tissue‐specific mouse models. Adipose‐specific deletion of BMAL1 (Ad‐Bmal1 ^−/− mice) decreased the concentration of adipose tissue polyunsaturated FA‐containing TGs, which increased saturated versus unsaturated FAs both in circulation and in hypothalamic neurons, and in turn altered feeding behaviour to drive obesity in mice challenged with high‐fat diet (Paschos et al. 2012). Thus, a well‐functioning circadian clock in adipose tissue is critical for whole‐body energy homeostasis and body weight balance.

Cycling of the intracellular clock is not intrinsically altered in white adipose tissue of obese healthy or type 2 diabetic subjects, at least in the early stages of the disease, indicating that clock disruption in white adipose tissue may not be a primary cause linked to the development of insulin resistance (Otway et al. 2011). Consistently, acute sleep loss alters the circadian clock in skeletal muscle, but not in the subcutaneous adipose tissue (Cedernaes et al. 2018). Nevertheless, in isolated adipocytes from obese patients transduced in vitro with a circadian reporter construct, the increased age of the donor correlates with a shortened circadian period in BMAL1 expression. This suggests that ageing‐related development of metabolic disorders could be associated with adipose tissue clock perturbation (Kolbe et al. 2019).

Synchronizing nutrient interventions to the molecular circadian clock to alter adipose tissue metabolism

Adipose tissue is characterized by its exquisite ability to adapt to an energy surplus through processes of adipocyte hypertrophy and/or hyperplasia, concomitant with remodelling of the vascular and immune environment. In adipocytes, activation of nutrient‐sensing nuclear receptors such as PPAR or G protein‐coupled receptor (GPCR) family members enable the adipose tissue to adapt to metabolic stimuli and induce the appropriate pathways (Goto et al. 2011; Quesada‐Lopez et al. 2016). Under a physiological setting, altered dietary patterns have also been shown to interfere with the molecular clock. For example, oscillations of clock gene expression are diminished in adipose tissue of mice fed a high‐fat diet, with decreased amplitude of the mRNA expression of Pparγ and its target lipogenic genes (Kohsaka et al. 2007). Thus, activation of the nuclear receptor PPARγ by fat‐soluble hormones or lipid compounds may also serve as a mechanism to integrate metabolic output and feedback to the molecular clock in adipose tissue (Yang et al. 2006; Chen & Yang, 2014). While high‐fat diet‐induced obesity may alter the molecular clock and impair metabolism, weight loss may have beneficial effects. For example, hypocaloric diet‐induced weight loss increases clock gene expression in adipose tissue in overweight humans (Pivovarova et al. 2016). Strikingly, the weight loss‐induced changes of clock gene expression were tightly correlated with the mRNA level of genes involved in fat metabolism, energy metabolism, autophagy and inflammatory responses (Pivovarova et al. 2016). However, whether these changes can impact the adipocyte proteome and improve its metabolic function remains to be determined.

The role of the circadian clock in the regulation of the energy homeostasis has raised interest in the notion of ‘chrono‐nutrition’ whereby dietary challenges such as altering the macronutrient content or timing of meals can reprogramme the clock and affect whole‐body physiology and metabolism (Asher & Sassone‐Corsi, 2015). Strategies to mitigate metabolic disturbances associated with type 2 diabetes include ‘resynchronizing’ metabolism through diet or targeting the time of day to potentiate the effect of a physiological or pharmacological treatment. As an example, time‐restricted feeding, without reducing calorie intake, improves circadian rhythms, increases brown adipose tissue thermogenesis, and prevents obesity and associated metabolic disorders in mice fed a high‐fat diet (Hatori et al. 2012). Interestingly, time‐restricted feeding in circadian mutant mice also improves the adipose tissue profile, by decreasing adipocyte hypertrophy and inflammation and increasing adiponectin secretion (Chaix et al. 2019). While the effects of time‐restricted feeding in humans are not fully characterized, some preliminary studies suggest that restoring feeding–fasting cycles with diurnal rhythms may have beneficial effects on metabolism. Indeed, time‐restricted feeding, with feeding time restricted to less than 8 h a day, sustains robust diurnal rhythms and has beneficial effects on insulin sensitivity in people with increased risk for type 2 diabetes (Panda, 2016; Sutton et al. 2018). Moreover, varying the calorie consumption throughout the day may influence body weight balance and adiposity. Greater caloric intake at breakfast versus dinner promotes weight loss, and improves fasting glucose levels, insulin sensitivity, and the blood lipid profiles in obese women fed an isocaloric diet (Jakubowicz et al. 2013). Thus, time‐targeted interventions as ‘chrono‐nutrition’ may represent a therapeutic strategy to improve adipose tissue function and manage metabolic diseases such as obesity and diabetes. However, in addition to nutrition, exercise is also a potent lifestyle intervention to combat metabolic dysfunction. The subsequent sections will focus on the role of exercise in adipocyte physiology and whether training at a specific time of day favours the adaptive response to exercise to increase insulin sensitivity and prevent metabolic diseases.

Exercise training in the fight against metabolic disorders: muscling in on adipose tissue

Regular physical activity is a major public health recommendation for the prevention and the management of type 2 diabetes and associated cardiovascular risks (Egan & Zierath, 2013). Myriad data demonstrate that exercise training improves glycaemic control, blood pressure and TG level in people with type 2 diabetes (Chudyk & Petrella, 2011). The beneficial effects of exercise on glucose and lipid metabolism are commonly explained by the improvement of skeletal muscle function (Gabriel & Zierath, 2017), but exercise also has a profound effect on insulin sensitivity in adipose tissue (Wallberg‐Henriksson & Zierath, 2015).

Adipose tissue has long been appreciated as a source of substrates for skeletal muscle, by releasing FAs in response to increased energetic demands associated with each exercise bout (Martin, 1996). Adipose tissue undergoes metabolic and genomic changes in response to acute exercise and training (Fig. 2). For example, exercise training decreases size and increases insulin sensitivity of isolated rat adipocytes (Craig et al. 1981), which may partly contribute to reductions in body fat and improvements in glucose metabolism (Weiss & Holloszy, 2007; Mavros et al. 2013; Drenowatz et al. 2015). The effects of exercise training on the metabolic and morphological properties of adipocytes are rapidly diminished within days after the last training session (Craig et al. 1983), indicating that genomic processes are involved. In mice, the exercise training‐induced remodelling of white adipose tissue, including reduced adipocyte size and increased FA metabolism, is associated with profound modifications of the transcriptional profile (Gollisch et al. 2009; Stanford et al. 2015b). Remarkably, exercise training induces a ‘browning’ of white adipose tissue, to generate a ‘brown adipose tissue‐like’ phenotype characterized by the presence of small multilocular lipid droplets and increased expression of thermogenic genes including Ucp1, Prdm16, Pgc1α and Cox8β (Stanford et al. 2015a). Several stimuli have been proposed to account for the browning of white adipose tissue including increased sympathetic innervation, or tissue cross‐talk between the working skeletal muscle and adipose tissue via secreted ‘myokines’ including irisin (Bostrom et al. 2012) and meteorin‐like 1 (Rao et al. 2014), or metabolites such as β‐aminoisobutyric acid (Roberts et al. 2014) and lactate (Carriere et al. 2014). Collectively, this has highlighted the importance of skeletal muscle–adipose tissue crosstalk to improve whole‐body glucose and energy homeostasis in response to exercise (Stanford & Goodyear, 2018). While the physiological relevance of browning as an adaptive response to exercise seems tenuous, this phenomenon could contribute to increased whole‐body energy expenditure and partly mediate the beneficial effect of exercise on glucose and lipid homeostasis (Seale et al. 2011).

Figure 2.

Exercise adaptation of adipose tissue and regulation by circadian factors

Evidence in mouse models and humans suggests that adipose tissue responds to exercise training by metabolic and genomic remodelling. Acute exercise induces lipolysis and inflammation. Exercise training leads to a profound remodelling of the metabolic and endocrine profile of adipose tissue. Skeletal muscle and adipose tissue communicate during exercise and play a role in adaptation to exercise training. Several physiological and metabolic processes are regulated in a circadian pattern, in response to hormonal, neuronal or feeding patterns or the molecular clock machinery. Changes in the molecular clock function can influence adipose tissue physiology, which in turn could potentiate the adaptive response to exercise. ^*Evidence from mouse models. PGC‐1α, peroxisome proliferator‐activated receptor γ coactivator 1α.

Adipose tissue can also play an endocrine role to mediate exercise‐induced improvements in glucose homeostasis. Transplantation of subcutaneous adipose tissue from mice trained for 11 days decreased fasting blood glucose and insulin level in a host sedentary mouse, likely via increased glucose uptake into oxidative skeletal muscle and brown adipose tissue (Stanford et al. 2015b). Although adipokines were proposed as the causative mechanism, the precise molecular mechanism remains elusive (Stanford et al. 2015b). However, recent studies suggested the existence of an adipokine secreted from brown adipose tissue, namely the lipokine 12,13‐dihydroxy‐9Z‐octadecenoic acid (12,13‐diHOME), which is released in response to cold exposure or acute exercise to potentially activate skeletal muscle and brown adipose tissue FA uptake and oxidation (Lynes et al. 2017; Stanford et al. 2018).

Much of the earlier work to understand the role of exercise on adipocyte biology was performed in rodents. However, exercise also affects adipocyte metabolism in humans. A recent study of healthy young men indicates the transcriptomic and epigenetic profile of abdominal subcutaneous adipose biopsies is altered after acute exercise, further highlighting the plasticity of adipose tissue (Fabre et al. 2018). While evidence for adipose tissue browning in humans in response to exercise training is lacking (Tsiloulis et al. 2018), profound adipocyte remodelling occurs, with increased angiogenesis and modification of the endocrine profile (Walton et al. 2015). Moreover, an exercise intervention (6‐months’ supervised training) altered the subcutaneous adipose tissue transcriptome, with an enrichment of oxidative phosphorylation pathways (Ronn et al. 2014). These training‐induced genomic changes might provide a mechanism for the beneficial effect of exercise in human adipocytes. Additionally, systemic changes in secreted factors may play a role. For example, in healthy individuals, the plasma level of the lipokine 12,13‐diHOME is acutely increased after 40 min of aerobic exercise, indicating a possible endocrine response of adipose tissue (Stanford et al. 2018). Nevertheless, emerging data suggests that the adipose tissue‐specific adaptation to exercise may be partly defective in obese insulin‐resistant individuals. For example, the angiogenic response to aerobic exercise is blunted, concomitant with reduced adrenergically induced lipolysis in subcutaneous adipose tissue (Walton et al. 2015; Verboven et al. 2018), while the beneficial effect on inflammation is maintained (Khadir et al. 2018).

While great strides have been made in identifying tissue‐specific secreted factors that communicate information regarding the metabolic or energetic stressors induced by exercise, the identification of specific adipokines and metabolites that signal from the adipocyte to the peripheral tissues during exercise is not a trivial undertaking, particularly in humans. Model systems to identify exercise‐induced adipokines are challenging given that exercise may not be fully recapitulated in isolated cellular systems, and whole‐body approaches may be impractical or imprecise for adequate sampling directly from adipocytes. An area of fertile research going forward is the identification of systemic factors that integrate the individual organ response to exercise with the entire body.

Interaction between circadian rhythms and the metabolic response of adipose tissue to exercise

Many of the pathways and genes identified in the control of whole‐body and cellular glucose and energy homeostasis are under circadian control (Eckel‐Mahan & Sassone‐Corsi, 2013). Furthermore, many of these circadian controlled genes have been described as being altered in response to exercise (Schroder & Esser, 2013) or nutrient stress (high‐fat diet) (Dyar et al. 2018). In mice, gene expression and metabolic adaptations to exercise vary according to zeitgeber time (Peek et al. 2017). Mitochondrial function also displays a circadian rhythm with oxidative metabolism peaking in the late afternoon (Deschenes et al. 1998; van Moorsel et al. 2016). Thus, optimizing the timing of a training bout to coincide with the greatest genomic or physiological response throughout the day may augment the potency of exercise as a therapeutic tool in regards to health outcomes.

The interaction between zeitgeber time and the metabolic response of the adipocyte is likely to be influenced by exercise, given the recent paradigm underscoring how intra‐ and inter‐tissue relationships coordinate the organismal response to nutrient stressors (Dyar et al. 2018). Therefore, the field is ripe to address several questions related to exercise and nutrient timing. For example, can exercise reset the adipose tissue intrinsic clock? While this question has not yet been experimentally addressed, acute exercise does affect the SCN circadian pattern, which may signal to the adipose tissue clock (Maywood et al. 1999). Conversely, the deleterious effect of high‐fat diet on the circadian clock and the concomitant increase in FA synthase gene expression in white adipose tissue is prevented by treatment of mice with the insulin‐sensitizer drug rosiglitazone (Ribas‐Latre et al. 2019). Therefore, it is plausible that as a physiological ‘insulin sensitizer’, exercise may reset the clock in adipose tissue, thereby improving whole‐body glucose metabolism and reducing adiposity.

Another question of interest is whether exercise timing can augment the metabolic effects on adipose tissue? The adaptive response of adipose tissue to exercise is likely dependent on nutritional status (fed or fasted) and time of day (day or night), with nutritional or hormonal status putatively potentiating or inhibiting the induction of specific molecular pathways that will contribute to adipose remodelling. Indeed, in overweight men, acute exercise performed after a meal attenuates the increase in mRNA expression of several genes involved in lipolysis and glucose metabolism such as ATGL, HSL, PDK4 and IRS2, compared to the same exercise performed in the fasting state (Chen et al. 2017). Although only the change in IRS2 was recapitulated at the protein level, it was found that the acute nutritional status affects the magnitude of the exercise response, and could therefore alter the adaptive response of adipose tissue to exercise training. As another example, in mice, browning of white adipose tissue is linked to vascular endothelial growth factor (VEGF) abundance and adrenergic signalling (During et al. 2015; Aldiss et al. 2018). Several studies indicate that the molecular clock interacts with hypoxia‐inducible factor (HIF) signalling and regulates the skeletal muscle glycolytic response and Vegfα expression during exercise in mice (Peek et al. 2017; Wu et al. 2017). Exercise‐induced catecholamine release also depends on the time of the day, while receptor expression displays a circadian pattern in adipose tissue (Yang et al. 2006; Scheer et al. 2010). Therefore, ‘browning’ of white adipose tissue and induction of metabolic programmes may be more effective if the exercise bout is performed when the adrenergic or hypoxia signalling is amplified. Moreover, in human myotubes, the secretion of myokines such as interleukin 6 and VEGF is under the control of the circadian clock (Perrin et al. 2015), suggesting that the exercise‐induced myokines profile can vary over the time of the day and therefore contribute to a different response in adipose tissue (Fig. 2).

The metabolic properties of adipose tissue are influenced in a distinct depot‐ and sex‐dependent manner (Karastergiou & Fried, 2017). Thus, adipose tissues from distinct depots may respond differently to exercise partly due to the sensitivity or rhythmicity of the response to various hormones, especially glucocorticoids, but also due to their intrinsic circadian gene expression profiles (Garaulet et al. 2011; Lee et al. 2014). In particular, the circadian rhythm of adipokine expression is depot‐specific, with substantial differences noted between subcutaneous and visceral human adipose tissue explants (Garaulet et al. 2011). Therefore, exercise‐induced adipokine secretion may be regulated differentially in men versus women based on sex‐dependent hormone and adipose tissue depot distribution profiles. Accordingly, subcutaneous adipose tissue seems to drive the metabolic response to exercise in mouse models (Stanford et al. 2015b). Other factors, such as the circadian oscillation of circulating immune cells, may influence the adipose tissue status and therefore the metabolic response (Zhao et al. 2017).

Concluding remarks

Type 2 diabetes is one of the fastest‐growing diseases world‐wide. For many patients, pharmacological intervention is required to manage this disease, yet effective insulin sensitizers are lacking from the current diabetes pharmacopeia. Exercise is an effective intervention to prevent insulin resistance and type 2 diabetes, and should be considered a first line treatment. While exercise training mainly targets skeletal muscle, it is a physiological ‘insulin sensitizer’ that also has a profound effect on adipose tissue. Several lines of evidence show that adipose tissue partly mediates the beneficial effect of exercise on glucose tolerance, via its metabolic and endocrine function. Diet and exercise have an additive effect on insulin sensitivity, involving metabolic and genomic adaptations at the level of skeletal muscle and adipose tissue, which are influenced by the time of day. Given the circadian role of white and brown adipose tissues in the regulation of metabolism, we speculate that time‐targeted diet and exercise strategies may potentiate adipose tissue metabolism and therefore contribute to the management of metabolic disorders. Understanding the relationship between nutrient state and circadian homeostasis, as well as the impact of circadian clock systems on human physiology and how this is regulated by exercise and nutrition in a tissue‐specific manner, will be a challenge for the field. Ultimately, a greater understanding of these relationships may yield insight into the connection between genes, behaviour and the development of metabolic diseases.

Additional information

Competing interests

We confirm that the authors do not have any conflicts of interests regarding the material in this review.

Author contributions

Both authors have read and approved the final version of this manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

The authors are supported by grants from the Novo Nordisk Foundation (NNF14OC0011493, NNF14OC0009941, NNF18CC0034900), Swedish Research Council (2015‐00165) to J.R.Z., and L.D. is supported by a Novo Nordisk postdoctoral fellowship run in partnership with Karolinska Institutet.

Glossary.

Chrono nutrition: the concept of adapting the macronutrient content or timing of a meal to the circadian rhythm to affect whole‐body physiology and metabolism.

Genomic response: a response depending on gene transcription and protein synthesis events.

Time restricted feeding: daily eating pattern in which all nutrient intake occurs within a few hours.

Zeitgeber time: time as defined after the synchronization of the circadian system by a zeitgeber, an external cue that can reset the time‐keeping system of organisms.

Biographies

Lucile Dollet is currently a Post‐Doctoral fellow in the Department of Physiology and Pharmacology at Karolinska Institutet where she is studying mechanisms by which exercise training affects adipocyte metabolism and adipose–skeletal muscle cross talk in diabetes.

Juleen R. Zierath is Professor of Physiology at Karolinska Institutet and University of Copenhagen. She delivered The Physiological Society's Annual Review Prize Lecture in 2018. As an exercise physiologist, she has a long‐standing interest in the health‐promoting benefits of physical exercise. The ultimate goal of her work is to identify and validate molecular candidates to prevent or treat insulin resistance in type 2 diabetes. She is President of the European Association for the Study of Diabetes. She is a member of the Royal Swedish Academy of Sciences and the Nobel Assembly. She was Chair of the Nobel Committee at Karolinska Institutet between 2013 and 2015 and is currently a member of this Committee.