Revisiting adipose thermogenesis for delaying aging and age-related diseases: Opportunities and challenges

Abstract

Adipose tissue undergoes significant changes in structure, composition, and function with age including altered adipokine secretion, decreased adipogenesis, altered immune cell profile and increased inflammation. Considering the role of adipose tissue in whole-body energy homeostasis, age-related dysfunction in adipose metabolism could potentially contribute to an increased risk for metabolic diseases and accelerate the onset of other age-related diseases. Increasing cellular energy expenditure in adipose tissue, also referred to as thermogenesis, has emerged as a promising strategy to improve adipose metabolism and treat obesity-related metabolic disorders. However, translating this strategy to the aged population comes with several challenges such as decreased thermogenic response and the paucity of safe pharmacological agents to activate thermogenesis. This mini-review aims to discuss the current body of knowledge on aging and thermogenesis and highlight the unexplored opportunities (cellular mechanisms and secreted factors) to target thermogenic mechanisms for delaying aging and age-related diseases. Finally, we also discuss the emerging role of thermogenic adipocytes in healthspan and lifespan extension.

Introduction

Adipose thermogenesis has been an attractive therapeutic target for obesity and metabolic disorders because it increases energy expenditure by catabolizing fuel reserves in the body. Thermogenic stimulation via cold or β3-AR agonists stimulates lipolysis, activates brown adipose tissue (BAT) and induces the formation of brown adipocytes in white adipose tissue (WAT), referred to as beige adipocytes through a process called ‘beiging’ (Kaisanlahti and Glumoff, 2019). Activated beige and brown adipocytes act as a metabolic sink for glucose, free fatty acids and branched-chain amino acids and result in an overall increase in whole-body energy expenditure (thermogenesis) (Wolfrum and Gerhart-Hines, 2022). In addition to energy expenditure, beiging is also linked to improved glucose and lipid homeostasis through increased glucose tolerance and insulin sensitivity both in rodents and humans (Cohen et al., 2014; Finlin et al., 2020). Hence, there is a growing interest in identifying mechanisms to recruit and activate brown and beige adipocytes to counteract metabolic dysfunction in humans. However, realizing the therapeutic benefits of thermogenesis in the target population of obese and older adults is challenging because the response to thermogenic stimulation is reportedly diminished with aging. Indeed, decreased thermogenesis with aging is suggested to be one of the drivers of increased metabolic dysfunction in the aging population. The lack of clinically safe pharmacotherapies to activate thermogenesis without adverse cardiovascular side effects in the aging population further adds to this challenge. Despite these challenges, there is emerging evidence linking increased adipose thermogenesis with extended longevity. A better understanding of the role of thermogenic adipocytes in healthspan extension will pave way for the identification of new targets to prevent or delay age-related diseases.

Thermogenic mechanisms in adipose tissue

Adaptive thermogenesis is a heat-generating cellular process aimed at restoring homeothermy during exposure to cold conditions. There are two kinds of thermogenic mechanisms. Shivering thermogenesis is an acute response and is primarily mediated by skeletal muscle contraction leading to increased fuel utilization and heat generation (Haman and Blondin, 2017). In contrast, non-shivering thermogenesis (NST) is a chronic adaptational response to maintain a stable body temperature in cold environments. NST is mediated through increased fuel mobilization and utilization in activated thermogenic adipocytes which results in an overall increase in energy expenditure and improvements in lipid and glucose metabolism (Roesler and Kazak, 2020). In adipocytes, NST can occur through both classical (UCP1-dependent mechanisms) and recently discovered UCP1-independent mechanisms.

UCP1-dependent thermogenesis:

BAT is a major site of non-shivering thermogenesis and is critical for thermoregulation in newborns and also in small mammals during chronic cold stress conditions. The electron transport chain activity is tightly coupled with ATP production in most cell types. Brown adipocytes (BAs) in BAT are, however, an exception, in that they express uncoupling protein 1 (UCP1) in their inner mitochondrial membrane. UCP1 uncouples cellular respiration from ATP synthesis and thereby generates heat. This thermoregulatory function of BAT is anatomically supported by dense sympathetic innervation to support heat generation and vascularization to distribute the generated metabolic heat throughout the whole body (Gaspar et al., 2021). Cold stimulation results in increased activation of the sympathetic nervous system and release of norepinephrine (NE) at the sympathetic nerve terminals both in BAT and WAT. NE acts through β3-adrenergic receptors, which are predominantly expressed on the surface of adipocytes, culminating in triglyceride lipolysis to release free fatty acids and glycerol. Fatty acids released during lipolysis bind to the nuclear receptor PPARα, which then translocates to the nucleus and increases the transcription of various genes involved in the thermogenic machinery including UCP1, the mitochondrial electron transport chain, and fatty acid oxidation.

UCP1 expression is not only observed in brown adipocytes, but also in beige or brite adipocytes, which are induced in the WAT by several stimuli including cold, β3-adrenergic agonists, PPARγ agonists, cancer cachexia, tissue injury and exercise (Lee et al., 2012). The mechanism of induction of beige adipocytes is dependent on the adipose depot. In the epididymal WAT (eWAT), genetic tracing and single-cell sequencing studies show that adipose stem cells positive for PDGFRα, Sma, Myh11 or PDGFRβ undergo de novo proliferation to form BAs in response to cold or β3-AR activation (Berry et al., 2016; Burl et al., 2018; Lee et al., 2012; Long et al., 2014; Vishvanath et al., 2016). On the other hand, beige adipocytes are induced in the inguinal WAT (iWAT) with little or no cell proliferation suggesting conversion of pre-existing WA to BA (Lee et al., 2012). Although activated beige adipocytes are morphologically similar to brown adipocytes (UCP1 expression, multilocular appearance and dense mitochondria), their developmental timing, lineage and maintenance are distinct from those of BAT (reviewed in (Ikeda et al., 2018)). While BAs are constitutively maintained, the existence of beige adipocytes is transient. Beige adipocytes lose their brown phenotype and revert to WA with a unilocular appearance in as little as 2 weeks after the withdrawal of thermogenic stimulation (Altshuler-Keylin et al., 2016; Rosenwald et al., 2013). Upon cold or β3-AR stimulation, activated UCP1+ brown and beige adipocytes drive fuel mobilization and oxidation (glucose, free fatty acids, branched-chain amino acids), thereby leading to improvements in whole-body energy homeostasis and glucose metabolism.

The presence of active BAT in adult humans has been unequivocally proven based on the landmark ¹⁸FDG-PET/CT studies back in 2009 (Cypess et al., 2009; Saito et al., 2009; van Marken Lichtenbelt et al., 2009). In addition, studies show that cold exposure and treatment with the β3-AR agonist mirabegron promote the recruitment of new thermogenic beige adipose tissue even in adults who did not have pre-existing brown fat in basal conditions. Further, BAT or beige adipose tissue activation was associated with increased energy expenditure, reduced body fat mass and/or improvements in insulin sensitivity (Finlin et al., 2018; Finlin et al., 2020; Lee et al., 2014; van der Lans et al., 2013; Yoneshiro et al., 2013), indicating that UCP1-dependent thermogenesis is physiologically relevant in humans. Moreover, in obese insulin-resistant patients who underwent treatment with mirabegron and displayed substantial subcutaneous WAT beiging, there were also improvements in pancreatic β-cell function and favorable muscle fiber type switching (Finlin et al., 2020). Since β3-ARs are not expressed in the pancreas or in muscle, these non-adipose metabolic effects are presumably mediated through secreted factors from thermogenic beige adipocytes in WAT. It is also important to note that there is an ongoing controversy about which β-adrenergic receptors (β1, 2 or 3) mediate human BAT thermogenesis. Unlike rodents, human BAT has reportedly low levels of β3-AR expression and recent studies have demonstrated the possibility that β1 or β2-AR might also contribute to increased UCP1 expression and BAT thermogenesis in humans (Blondin et al., 2020; Cero et al., 2021; Riis-Vestergaard et al., 2020). Human BAT composition is also debated. Some studies have shown that human BAT shares both brown and beige adipocyte gene expression signature (Jespersen et al., 2013) while another study indicated that it may be entirely composed of beige adipocytes (Sharp et al., 2012). It seems like the location of the human BAT determines its phenotype. Depending on whether the human neck BAT biopsy was obtained in deep versus intermediate region, they resemble either mouse classical brown or beige adipose tissue, respectively (Cypess et al., 2013).

UCP1 independent thermogenesis:

Genetic ablation of BAT or beige fat leads to the development of obesity and insulin resistance in rodents at usual laboratory temperatures (18–22°C) (Cohen et al., 2014; Lowell et al., 1993). On the other hand, UCP1 knock-out mice (UCP1^−/−) do not develop obesity or diabetes at these temperatures or at thermoneutrality (29°C) (Dieckmann et al., 2022; Enerback et al., 1997; Liu et al., 2003), although few studies have demonstrated otherwise (Feldmann et al., 2009; Luijten et al., 2019; Pahlavani et al., 2019). Several variables including inducible vs. germline deletion, C57BL/6N vs. 6J background strain, presence or absence of appropriate littermate controls, microbiota and diet composition, all could have contributed to this discrepancy and it is still uncertain on whether UCP1 ablation promotes obesity. However, while UCP1^−/−mice are sensitive and develop hypothermia to acute cold exposure, they adapt well to gradual cold exposure. These distinct metabolic phenotypes between BAT or beige fat-deficient mice and UCP1^−/−mice highlight the existence of UCP1-independent thermogenesis. In addition, cold and β3-AR treatment increased the metabolic rate of subcutaneous (inguinal) WAT and visceral (epididymal) WAT even in UCP1^−/−mice (Granneman et al., 2003; Ukropec et al., 2006) suggesting that UCP1 is not required for WAT thermogenesis.

Several UCP1-independent thermogenic mechanisms based on ATP-consuming futile metabolic substrate cycling have been documented in recent years (Brownstein et al., 2022; Roesler and Kazak, 2020). One of the major futile cycles involves the breakdown of triglycerides (TGs) to release FA and glycerol followed by re-esterification to form TGs. Free glycerol released during lipolysis is converted to glycerol-3-phosphate by glycerol kinase (GyK) in an ATP-dependent manner which can be used for re-esterification. As an alternative source, the glycerol-3-phosphate needed for re-esterification can also be generated de novo by glyceroneogenesis, a process mediated by phosphoenolpyruvate carboxykinase (PEPCK-C). While adipocytes under basal conditions lack Gyk expression, thermogenic stimulation, antidiabetic agents and UCP1 ablation induce the expression of Gyk and PEPCK-C in WAT and contributes to non-canonical thermogenesis in beige and brown fat (Guan et al., 2002; Mottillo et al., 2014; Oeckl et al., 2022). Another UCP1-independent mechanism depends on the ATP-consuming phosphorylation and dephosphorylation of creatine in the mitochondria of thermogenic adipocytes (Kazak et al., 2015). Creatine kinase B (CKB) dictates the phosphorylation reaction of creatine, and the ATP demand for this reaction drives mitochondrial oxidation of substrates, thus leading to an overall increase in energy expenditure (EE). Reducing creatine levels by adipocyte-specific genetic deletion of the enzyme involved in creatine biosynthesis (glycine amidinotransferase, GATM) and cell-surface creatine transporter significantly impaired EE and pre-disposed the animals to obesity on a high-fat diet (Kazak et al., 2017; Kazak et al., 2019). Similarly, calcium cycled between the sarcoplasmic-endoplasmic (SER) stores and cytosol by the SERCA2b pump also consumes ATP and drives electron transport chain (ETC) activity thus contributing to thermogenesis in a UCP1-independent manner (Ikeda et al., 2017). Although all of these non-canonical thermogenic mechanisms have been documented in brown and beige adipocytes, we still do not know whether they occur within the same adipocyte or whether there are heterogeneous sub-populations that exclusively engage in specific substrate cycling. In addition, we also do not know their collective or singular relative contribution to the overall EE in comparison to the UCP1-dependent pathway and if the effects are transient or enduring after the cessation of thermogenic stimulation. Nonetheless, the UCP1 independent mechanism holds great promise for improving metabolic health, especially for the aged population where beiging or UCP1 induction is impaired.

Age-related changes in thermogenic adipose tissue: Knowns and unknowns

Thermogenic adipose tissue mass:

In humans, BAT identified by imaging and histological analysis was reported to be present not only in the classical interscapular region but also in other areas like supraclavicular, cervical, axillary, peri-renal and mediastinal (chest) depots at birth. These substantial depots, comprising up to 5% of body weight, play a significant role in non-shivering thermogenesis in the immediate post-natal period. While the interscapular depot slowly involutes beginning shortly after birth, thermogenic adipose tissue is maintained in other regions during childhood. Studies show a transient increase in the mass and activity of BAT around puberty which correlated with skeletal muscle volume (Gilsanz et al., 2012). After that brief phase, BAT mass slowly declines with aging and reaches a plateau in the sixth decade of life. Some studies show a negative correlation between BAT mass with BMI: levels are proportionally higher in lean than in obese adults (Leitner et al., 2017). It is not clear why some adults maintain BAT mass with aging and whether loss of BAT in others increases their risk for obesity or metabolic diseases. Imaging studies of this kind have limitations, and it is highly challenging to get an accurate measurement of BAT in humans due to its heterogenous presence within WAT. These technical challenges contributed substantially to the variability observed with BAT volume in healthy adults between studies, which produced estimates varying from 14 ml to 665 ml (van der Lans et al., 2014). Nonetheless, understanding the mechanisms behind BAT retention in adulthood and its potential association with healthy aging can open up novel strategies to target thermogenic adipose tissue for age-related metabolic dysfunction.

Browning potential of WAT:

Aging is known to impair the ability of WAT browning in response to thermogenic stimulation. Several studies used cold stimulation to investigate BAT function and the WAT browning response in aging mice (Berry et al., 2017; Goldberg et al., 2021). The sympathetic tone to adipose tissue declines with age (Bahler et al., 2016), and therefore cold stimulation, which is mediated through the sympathetic nervous system (SNS), may not be appropriate to investigate thermogenesis in aged animals. Direct stimulation of β3-AR using pharmacological agents will bypass the SNS and thus may provide a better method for thermogenic stimulation in aged animals. In studies involving β3-AR treatment, there are conflicting reports on the browning potential of aged WAT. Some studies reported no change in β3-AR treatment-induced UCP1 expression (Shin et al., 2017; Wang et al., 2019) while others reported an increase in UCP1 expression in aged rodents (Rogers et al., 2012; Tournissac et al., 2021) and aged humans (Finlin et al., 2020) when compared to untreated controls. Even though the thermogenic response to adrenergic agonists is diminished with aging when compared to young counterparts, it might still be sufficient to improve metabolic outcomes in aged animals. While the majority of the studies did not investigate this, a recent study by Tournissac et al. showed reduced body weight, fasting glucose and insulin and improved glucose tolerance in 15-month-old mice treated with CL316,243 (a β3-AR agonist) for 1 month (Tournissac et al., 2021). They also reported increased BAT thermogenesis marked by higher temperature and UCP1 expression, but the effect on WAT browning was not assessed. One of the critical differences between this study and the others is the duration of the thermogenic stimulation in aged animals. While most studies investigated thermogenic response in aged animals with acute stimulation, Tournissac et al. (Tournissac et al., 2021) used a chronic approach, and their work suggests that chronic β3-AR treatment or cold stimulation can improve thermogenesis in aging. In fact, unpublished data from our lab with 6 weeks of CL316,243 treatment in 18-month-old C57BL/6J mice showed similar metabolic benefits with improved glucose tolerance, reduced fat mass and improved adipokine profile. More importantly, our studies showed the induction of UCP1-independent thermogenic mechanisms, specifically futile TG-FA cycling, in WAT as a synergistic mechanism responsible for the metabolic benefits of β3-AR agonist treatment in aged mice. We believe that in lieu of UCP1, thermogenic stimulation forces aged adipose tissue to respond via alternative mechanisms, such as futile substrate cycling, to increase cellular energy expenditure. These promising preliminary studies highlight the potential of UCP1-independent thermogenesis for the aged population and should be explored more in future studies. It is not clear, however, to what extent metabolic benefits can be mediated through these alternate mechanisms.

Progenitor cell differentiation vs transdifferentiation:

De novo beige adipocyte formation within WAT occurs by proliferation and differentiation of adipocyte progenitor cells (PDGFRα+) residing in the stromal vascular fraction (SVF). Aging has been shown to decrease both the number and the differentiation potential of these progenitor cells in WAT (Berry et al., 2017; Shin et al., 2017). Elegant studies by Berry et al. (Berry et al., 2017) point to senescence as the potential mechanism underlying impaired differentiation of beige progenitor cells in WAT. Senescence is the cellular stress response that culminates in irreversible proliferative arrest in proliferating cells. SVF cells of 6 months old mice and old humans expressed high levels of senescence markers compared to their young counterparts. Genetic overexpression of p21 in smooth muscle actin+ (SMA) cells in adipose tissue disrupted the beiging potential of SVF cells. On the other hand, blocking senescence in aged progenitor cells by genetic deletion of the Ink4a/ARF locus improved beige adipogenesis. In another study, genetic clearance of senescent cells using INK-ATTAC mice also improved adipogenesis in middle-aged mice (Xu et al., 2015). Also, overexpression of Sirt1 in adipose-derived Myf5^−ve stem cells isolated from elderly donors blocked p53/p21 mediated senescence and improved their differentiation potential (Khanh et al., 2018). The circulating pro-aging factors may affect the adipose tissue microenvironment to promote senescence and thermogenic failure. Heterochronic parabiosis studies show that circulating factors in young blood can reduce the senescence burden in aged WAT (Ghosh et al., 2019). While these studies explain the thermogenic failure in visceral or epididymal WAT where de novo proliferation contributes to beiging, it is still not clear what mechanisms contribute to impaired beiging in subcutaneous or inguinal WAT, where pre-existing WAs transdifferentiated to become BAs. Aging-induced intrinsic defects in WAT, such as impaired mitochondrial function/biogenesis, immune cell dysregulation or altered extracellular matrix composition could all potentially affect transdifferentiation in iWAT. However, the exact mechanisms are yet to be confirmed.

Sympathetic nerve output to WAT:

Centrally mediated sympathetic nerve output drives thermogenesis in WAT and BAT in response to even mild decreases in ambient temperature (Δ 5°C)(Chen et al., 2013). NE secreted from the sympathetic nerve terminals activate β3-ARs on brown and white adipocytes to stimulate lipolysis, beiging and thermogenesis. In fact, the sympathetic nerve density correlated with the number of cold-induced BAs in WAT (Murano et al., 2009), highlighting the significance of neural control in thermogenesis. Also, iWAT sympathetic denervation blocked β3-AR stimulation-induced UCP1 expression (Contreras et al., 2014), suggesting that the basal sympathetic tone is necessary for maintaining adipose phenotype plasticity. Aging is associated with both reduced innervation (Song et al., 2020) and decreased responsiveness to sympathetic stimulation in WAT. The mechanisms responsible for reduced WAT innervation with aging are not known. Recently, nerve growth factor secreted from activated eosinophils in WAT has been shown to drive cold-induced increased arborization of sympathetic nerves (Meng et al., 2022). Based on this study, age-induced loss and dysregulation of type 2 innate lymphoid cells (ILC2) in WAT (Goldberg et al., 2021) could potentially contribute to denervation with aging. In addition, trophic factors from PDGFRα+ cells have been reported to influence nerve density in WAT (Song et al., 2020), and in principle, age-induced senescence in PDGFRα+ adipose resident progenitor cells could also diminish innervation with aging. Understanding the intercellular interactions between the senescent adipose progenitor cells, immune cells and sympathetic nerves in aging WAT might lead to the identification of strategies to increase innervation and improve thermogenesis with aging.

Thermogenic adipose secretome:

In addition to energy expenditure and its associated metabolic benefits, thermogenic brown and beige adipocytes secrete a variety of bioactive lipid mediators (also termed as lipokines), as well as peptides, metabolites and adipokines that exert both paracrine actions locally and regulate systemic metabolism through endocrine actions. Cold exposure and exercise have been shown to increase circulating levels of the lipokine, 12,13-dihydroxy-9Z-octadecenoic acid (12,13-diHOME), which is an oxidized derivative of linoleic acid secreted from BAT (Lynes et al., 2017). 12,13-diHOME has been shown to activate fatty acid uptake in the skeletal muscle, indicating that thermogenic adipose tissue regulates metabolism in distant organs through lipokines (Stanford et al., 2018). Complementing the actions of lipokines, brown and beige adipocytes also secrete metabolites like 3-methyl-2-oxovaleric acid, 5-oxoproline and β-hydroxyisobutyric acid, which are catabolites of branched-chain amino acids (BCAA) degradation (Whitehead et al., 2021). These metabolites induce the expression of mitochondrial and oxidative genes in the muscle and increase whole-body EE (Whitehead et al., 2021). Proteomics-based mouse studies have also shed some light on the secreted proteins from thermogenic brown and beige adipocytes. Some of the common proteins that arose in several proteomics studies with thermogenic adipose remodeling included; extracellular matrix (COL3A1 and PCOLCE), growth factors (insulin-like growth factor-binding protein-4 (IGFBP4)), follistatin-like protein-1 (FSTL1), and chemerin, as reviewed by Villarroya J et al. (Villarroya et al., 2019). A recent proteomics study using primary human brown adipocytes shows ependymin-related protein 1 (EPDR1) as a novel batokine that regulates whole-body energy metabolism (Deshmukh et al., 2019). BAT also secretes proteins, including bone morphogenic protein-8b (BMP-8b), neuregulin-4 (NRG4), nerve growth factor and S100B protein that act locally on vascular cells and neurons (Villarroya et al., 2019). Especially, the autocrine actions of BMP8b mediates the secretion of NRG4 which promotes sympathetic axon growth and also upregulates the secretion of pro-angiogenic factors to increase vasculrization (Pellegrinelli et al., 2018). NRG4 has also been reported to have metabolic benefits in the liver independent of BMP8b where it activates receptor tyrosine kinases ErbB3 and ErbB4 leading to STAT5 phosphorylation and inhibition of hepatic de novo lipogenesis (Wang et al., 2014). In addition, there are also humoral factors like adiponectin and FGF-21 which are part of the thermogenic adipose secretome and have been shown to have a beneficial systemic impact (Abu-Odeh et al., 2021; Villarroya et al., 2019). Adiponectin activates the AMPK axis to mediate pleotropic effects on several tissues including reducing hepatic gluconeogenesis, increasing fatty acid oxidation and glucose uptake in muscle and acts centrally to regulate whole-body energy expenditure (Kubota et al., 2007; Yamauchi et al., 2002). FGF21 also exerts complementary actions to adiponectin by increasing insulin sensitivity and improving lipid metabolism (Villarroya et al., 2019). More specifically, studies in BAT-specific FGF21 knock out mice show that BAT-secreted FGF21 exerts cardioprotective effects by preventing hypertension-induced cardiac remodeling (Ruan et al., 2018). There is no information available on how aging affects the secretory function of thermogenic adipocytes (both UCP1-positive and UCP-negative cells). Considering the physiological significance of these secreted factors, it is also important to investigate if these secreted factors can be pharmacologically targeted to mimic the beneficial effects of thermogenesis in the aged population.

Overall, a complete picture of the effects of aging on thermogenic adipose tissue is still lacking. Future studies investigating the unknown and/or unexplored mechanisms that contribute to impaired thermogenesis in aging may offer potential therapeutic targets for improving age-related metabolic dysfunction (Fig. 1).

Figure 1:

Age-related changes in the thermogenic adipose tissue. The figure summarizes the effects of aging on the cellular mechanisms mediating thermogenesis, browning potential of WAT, BAT mass, thermogenic adipose tissue secretome, and sympathetic innervation to WAT. The unknown aspects or mechanisms are indicated by a question mark at the end. The figure was created using Biorender.com.

A potential role for thermogenic adipocytes in healthspan and lifespan extension

Adipose tissue plays a pivotal role in the interplay between metabolism and aging (Finkel, 2015; Palmer and Kirkland, 2016). Age-related changes in adipose tissue metabolism contribute to increased disease vulnerability through systemic influences including chronic ‘sterile’ inflammation, altered adipokine secretion, ectopic lipid accumulation and insulin resistance (Mau et al., 2020; Miller et al., 2017; Muzumdar et al., 2008; Palmer and Kirkland, 2016). On the other hand, the anti-aging effects of several interventions including calorie restriction, 17α-estradiol and acarbose may be mediated, at least partially, through modulation of adipose tissue metabolism (Kulkarni et al., 2018; Mau et al., 2020; Miller et al., 2017; Stout et al., 2017). Based on these experimental observations, it seems likely that adipose tissue metabolism may contribute in important ways to systemic aging (Muzumdar et al., 2008). Emerging evidence points to a critical role for increased thermogenesis in the adipose tissue in longevity and potentially delaying the onset of age-related diseases (Darcy and Bartke, 2017; Darcy et al., 2016; Ortega-Molina et al., 2012; Vatner et al., 2018). Genetic models of longevity including Ames dwarf mice, Snell dwarf mice, and growth hormone receptor KO (GHR-KO) mice have enhanced BAT function, increased UCP1 expression (or beiging) in WAT depots, and increased expression of genes regulating thermogenic function in their adipose tissue (Darcy and Bartke, 2017; Darcy et al., 2016; Li et al., 2020). A recent paper by Li et al. also showed UCP1 induction in both BAT and WAT depots of long-lived mice with genetic deletion of pregnancy-associated plasma protein-A (PAPP-A) (a protease that cleaves some IGF1 binding proteins and regulates IGF1 concentration in tissues) (Li et al., 2022). In addition, these long-lived mice display altered macrophage polarization (increase in M2/M1 ratio) and decreased adipose inflammation (Li et al., 2020; Li et al., 2022; Liu et al., 2004; Menon et al., 2014). At the whole-body level, Ames dwarf mice and GHR-KO mice demonstrate increased oxygen consumption, energy expenditure, and decreased respiratory quotient (RQ) (Westbrook et al., 2009), all of which are in accord with the increased adipose thermogenesis observed in these mice. The core body temperature of the Ames dwarf mice and GHR-KO mice is lower than their normal littermates. Some of the potential explanations as discussed by Westbrook et al. (Westbrook et al., 2009) include the reduced thyroid hormones (both T3 and T4) and increased heat loss due to decreased mass-surface area. Both of these mechanisms potentially contribute to the reduced core body temperature, which may in turn stimulate the activation of thermogenic mechanisms in an attempt to restore homeothermy.

In addition to mutations that alter GH-signaling, dietary interventions have also been associated with increased UCP1 induction in adipose tissue. Calorie restriction (CR) has been demonstrated to promote the development of UCP1+ beige adipocytes in both subcutaneous and visceral WAT depots (Fabbiano et al., 2016; Huang et al., 2022; Xinna Li, 2022) and prevent mitochondrial dysfunction in BAT with aging (Valle et al., 2008). More importantly, beige WAT depots, rather than BAT, have been reported to be the major glucose disposal sites in CR mice, highlighting the metabolic significance of beige adipocytes in CR (Fabbiano et al., 2016). Time-restricted feeding (TRF), a dietary paradigm in which the food intake is restricted to an 8 hr window has also been shown to induce the browning of subcutaneous WAT and prevent obesity in response to a cafeteria diet in rats (Aouichat et al., 2020). In addition to fasting and calories, diet composition also affects adipose beiging. Low protein diets have also been demonstrated to increase energy expenditure in a UCP1-dependent manner (Hill et al., 2017; Rothwell and Stock, 1987). Recently, the beneficial metabolic effects of low-protein diets have been attributed to reduced consumption of specific dietary amino acids. Among different kinds of amino acids, an isoleucine (a branched-chain amino acid)-restricted diet mimicked the metabolic effects of a low-protein diet and activated the FGF21-UCP1 axis to increase energy expenditure (Yu et al., 2021). Also, life span-extending drug interventions like acarbose, 17α-estradiol, and canagliflozin increased UCP1 expression in both BAT and WAT depots and increased M2/M1 ratio in adipose tissue (Xinna Li, 2022). UCP1 induction in response to 17α-estradiol and canagliflozin occurred in a sex-specific manner that correlates with lifespan extension (higher UCP1 expression and longer lifespan only in males)(Xinna Li, 2022). Overall, the adipose tissue changes related to UCP1 induction and favorable macrophage polarization appear to be shared phenotypes of delayed aging irrespective of the intervention (mutation, dietary regimen, or drug treatment) (Fig. 2).

Figure 2:

Adipose tissue-specific shared mechanisms of delayed aging. UCP1 induction and favorable macrophage polarization leading to decreased inflammation in both BAT and WAT are common adipose phenotypes observed in several animal models of delayed aging and extended longevity. Dwarf mice-Snell and Ames dwarf mice, GKO-Growth hormone receptor KO mice, PAPP-A-pregnancy-associated plasma protein-A KO mice, and CR-calorie restriction.

The upstream mechanisms that activate UCP1 in adipose tissue may vary depending on the type of intervention. Reduced core body temperature has been postulated as a driver for thermogenesis in mice with extended longevity. The data, however, are inconsistent. Housing at thermoneutrality abolished the thermogenic adipose phenotype in Ames dwarf mice (Darcy et al., 2018) but not in BCAA-restricted mice (Yu et al., 2021). The lifespan benefits observed in GHRKO mice were also not affected by housing at thermoneutral conditions from weaning (Fang et al., 2020). One of the alternative stimuli for UCP1 induction could be lipid mobilization triggered by negative energy balance either at the central level (sympathetic stimulation) or at the level of adipose tissue, in that free fatty acids released by TG hydrolysis induce UCP1 expression in the adipose tissue.

Although there is a clear link between thermogenic mechanisms in adipose tissue and longevity, several questions remain unanswered. First, it is not known to what extent the UCP1+ thermogenic adipocytes contribute to the healthspan and lifespan benefits of anti-aging interventions. Genetic manipulation of UCP1 in mice (deletion or overexpression) has provided some insights into the potential anti-aging role of thermogenesis. Targeted overexpression of UCP1 to the skeletal muscle increased median lifespan and reduced the incidence of lymphoma and atherosclerosis (Gates et al., 2007). Metabolically, these animals had higher energy expenditure, reduced adiposity, and increased circulating adiponectin levels (Gates et al., 2007), suggesting a potential link between UCP1-mediated uncoupling and healthspan extension. On the other hand, Kontani et al. reported that global deletion of UCP1 did not affect lifespan at standard housing conditions, although these mice did develop diet-induced obesity with age (Kontani et al., 2005). UCP1 KO mice have inherent developmental defects in thermoregulation and hence studies using inducible UCP1 KO mice should be performed to avoid this confounding variable. Further, adipose tissue-specific UCP1 gene targeted studies are lacking, and hence the direct role of adipose thermogenesis in lifespan and healthspan regulation is not well understood. Secondly, the role of UCP1-independent mechanisms in lifespan and healthspan regulation is even less studied. Third, the influence of age at the time of initiation of a dietary or pharmacological anti-aging intervention on adipose thermogenesis is also worthy of investigation. Most of the studies discussed in this section utilized mice that were less than one year old, and studies of age effects on UCP1-dependent and UCP1-independent thermogenesis would be useful guides to the possible translational potential of these systems in aging humans. Lastly, we still do not know if activating thermogenesis via β3-AR agonists can mimic some of the metabolic benefits observed in the long-lived mice, although the dose and duration have to be fine-tuned to avoid the cardiovascular side effects observed with the existing non-specific β3-AR agonists.

Thermogenesis and age-related diseases

Several lines of evidence support a protective role for thermogenesis in age-related diseases. β3-AR agonist (CL 316,243) treatment for 1 month decreased body weight, improved glucose metabolism, and increased BAT thermogenesis in 15-month-old WT and 3xTg AD mice (Tournissac et al., 2021). Further, CL treatment also improved the cognitive outcomes like recognition index and reduced insoluble Aβ aggregates in the hippocampus of 3xTg AD mice (Tournissac et al., 2021). Mirabegron treatment improved learning and memory and also displayed anti-depressant and anxiolytic properties in young mice (Tanyeri et al., 2021), although its effect on age-related cognitive impairment needs further investigation. Mirabegron treatment for 12 weeks in obese insulin-resistant middle-aged humans (mean age 54 years) not only improved glucose and adipose metabolism (reduced fibrosis, increased lipolysis, and altered macrophage polarization) but also improved pancreatic β-cell function and reduced lipotoxicity and increased type I fibers in the skeletal muscle (Finlin et al., 2020). This suggests that mirabegron might also have therapeutic implications for sarcopenia. More studies are needed to assess the impact of thermogenic stimulation on other age-related diseases.

Conclusions and future perspectives

There is no doubt that the manipulation of cells and circuits involved in thermogenesis holds promise for improving age-related metabolic dysfunction and healthspan extension. However, several questions have to be answered to bring this idea to fruition. First, the mechanistic role of thermogenic adipocytes in the healthspan benefits exerted by anti-aging interventions should be explored. Secondly, more studies are needed to determine if UCP1-independent mechanisms and the thermogenic adipose secretomes are preserved in aging, because data on this point would open up novel avenues to improve systemic metabolism in aging. In addition, one of the major translational challenges is the lack of safe and effective pharmacological agents to activate thermogenesis in humans. The existing drugs, specifically β3-AR agonists like Mirabegron, have cardiovascular side effects in clinical studies. Identification of novel agents that target non-β3-AR mechanisms (such as ABHD5 ligands that promote lipolysis (Sanders et al., 2015)) might be a safe alternative to reduce the cardiovascular risk associated with β3-AR agonists in the elderly population. More work on sex differences in these pathways is also needed. There is some preliminary evidence showing sex-specific differences in thermogenic mechanisms, and future studies should include both sexes to explore this topic more carefully. Finally, it is also important to consider that thermogenesis is a double-edged sword. Contrary to the beneficial metabolic effects, WAT browning is one of the early mechanisms that contribute to cancer cachexia and tumor progression. Considering the high prevalence of cancer in the aging population, careful consideration should be given to selecting ideal candidates for thermogenesis activators in future studies.

Highlights.

• Age-related changes in adipose tissue contribute to systemic aging

• Activation of thermogenesis is an attractive therapeutic target for improving adipose and systemic metabolism

• Thermogenic stimulation as a strategy for improving metabolic dysfunction in aged population is challenging

• Emerging evidence from animal models of longevity suggests a potential role for thermogenic adipocytes in lifespan and healthspan extension

• Understanding how thermogenic adipocytes play a role in healthspan extension might pave way for developing novel targets to delay aging and the onset of age-related diseases.