The evolutionary impact and influence of oestrogens on adipose tissue structure and function

Abstract

Oestrogens are sex steroid hormones that have gained prominence over the years owing to their crucial roles in human health and reproduction functions which have been preserved throughout evolution. One of oestrogens actions, and the focus of this review, is their ability to determine adipose tissue distribution, function and adipose tissue ‘health’. Body fat distribution is sexually dimorphic, affecting males and females differently. These differences are also apparent in the development of the metabolic syndrome and other chronic conditions where oestrogens are critical. In this review, we summarize the different molecular mechanisms, pathways and resulting pathophysiology which are a result of oestrogens actions in and on adipose tissues.

This article is part of a discussion meeting issue ‘Causes of obesity: theories, conjectures and evidence (Part I)’.

1. Introduction

Clinical and research communities have long appreciated the differences between men and women with respect to disease prevalence [1,2]. The focus of this review is on oestrogens, in all their forms and through their receptors, as key drivers of sex differences in disease prevalence and how these differences relate to their impact on adipose tissues [3]. Throughout evolution, sex differences in body fat distribution have persisted. Women store fat in a healthy way because they are programmed to gain weight ‘healthfully’ during pregnancy. Additionally, women need to mobilize stored fat and calories during breastfeeding to provide nutrition to their young. In a seminal manuscript focusing on the topic of evolution and body adiposity, Speakman argues the increased capacity to store body fat is an evolutionary advantage. Speakman suggests accrual of body adiposity provides a selective advantage for sustaining metabolic processes, fighting illness and supporting physical activity [4]. It appears oestrogens are critical drivers of these functions, which is why women can store fat and extra calories more healthily than men, a phenomenon that leads to reductions in disease prevalence in pre-menopausal women when compared to men [4–6].

Historically, oestrogens were thought to mainly control reproduction. The role of oestrogens in regulating metabolic function has expanded and here we discuss the evolving literature of how oestrogens influence adipose tissue function. Oestrogens act directly on the adipocyte, or fat cell, as well as influence different regions of the brain providing a mechanism for neural communication to the fat tissue and different cells within adipose tissue [4,7,8]. Discoveries focusing on oestrogens' relation to fat storage and function is not just relevant to women. Men have oestrogens too—in fact, oestrogens are made from testosterone in adipose tissues [9]. Overall, the more we learn about what sex hormones and sex chromosomes do in tissues such as adipose tissues, the more we will understand how they can be augmented to potentially reduce disease risk.

In this review, we initially discuss oestrogens role in body fat distribution. We then discuss the role of circulating oestrogens and how they influence adipose tissue function. We follow this discussion by talking about the role of the oestrogen receptors (ERs) in adipose tissues. We provide a brief discussion on what was learned through genome-wide association studies (GWAS) regarding how oestrogen-regulated genes influence sex differences in body fat distribution [10]. We expand our discussion on the role sex plays in determining adipose tissue function by discussing the role sex chromosomes have in influencing adiposity [10,11]. Oestrogens also influence adipose tissue lineage and epigenetics as additional methods they impact metabolism. There are new and evolving data regarding oestrogens' influence on the adipose tissue microenvironment as well as oestrogens' influence on lipogenesis and lipolysis within adipose tissues. Lastly, we briefly discuss the emerging field of oestrogens' role in brown and beige adipocytes. Our goal is to expand the readers knowledge of the breath, depth and scope of how oestrogens and their receptors directly influence all facets of adipose tissue biology and function. Numerous articles outline oestrogens’ role in adipose/fat tissue in one way or another; however, this review brings together a host of published articles that have covered a particular facet of how oestrogens influence adipose tissue and synthesizes the data into a more complete picture of the biological and evolutionary role of oestrogens in adipose tissue function. Additionally, this review does not limit its scope to the role of one oestrogen, but rather seeks to understand oestrogens as a family and their functions through a family of receptors.

2. The evolutionary biology in sex differences of adipose tissue distribution

The aetiology of sex differences in adipose tissue distribution can begin to be explained by evolutionary biology. Biologically, men and women differ in their proportion of body adiposity and where their adipose tissues are deposited, even after correcting for body mass index (BMI). Sex differences in body adiposity are influenced by sex hormones and sex chromosomes. Sex differences in body adiposity begin early in life and further evolve during puberty [8].

White adipose tissue (WAT) is divided by its anatomical location into either subcutaneous adipose tissue (SAT) or visceral adipose tissue (VAT). Adipose tissue can be broadly categorized as either SAT which is anatomically located below the skin, whereas VAT is located inside the abdominal wall and is associated with abdominal organs [8]. Importantly, venous outflow enters the portal vein, which is thought to be critical for VAT's metabolic effects [12]. In general, females, have more SAT, typically creating a ‘pear-shape’ adipose tissue distribution, referred to as a gynoid body fat distribution. Interestingly, not all SAT is created equally. For instance, abdominal and gluteo-femoral SAT have differentially expressed genes and distinct metabolic profiles. For instance, higher levels of gluteo-femoral SAT appear to be linked to reduced cardiovascular and metabolic risks [13,14].

Males and post-menopausal women have a higher ratio of VAT to SAT creating an ‘apple-shape’ body habitus, also called android body fat distribution. Several studies have shown correlations between VAT deposition and metabolic disease, whereas lower-body (gluteal and femoral) adipose tissue deposition appears to be protective and/or is associated with less negative metabolic consequences [9,15,16].

Oestrogens influence on body fat distribution is further supported by findings from rodent studies, which demonstrate removal of the ovaries (ovariectomy) increases overall adiposity and enhances deposition of VAT, whereas of 17 beta-oestradiol replacement reduces overall adiposity and favours redistribution of the adipose tissue into SAT. In humans, post-menopausal women experience a shift in their adipose tissue deposition towards accrual of adipose tissue in VAT [17,18]. Additionally, menopause is associated with a specific increase in total VAT relative to SAT and a slight decrease in fat-free body mass in healthy-weight and mildly obese women. Research using dual-energy X-ray absorptiometry has further supported the connection between increased deposition of adiposity in VAT in post-menopausal women. It is worth noting that even non-obese women undergo significant increases in VAT and per cent body adiposity during the transition into menopause. Moreover, when comparing pre- and post-menopausal women with matched BMI controls, the pre-menopausal group exhibited a significantly lower percentage of VAT [17,19,20]. Additionally, there are data indicating ethnic differences in female body adiposity. Individuals with low levels of body fat from different races have no notable disparities in VAT; however, as adiposity levels increased, both Asians and whites exhibited higher quantities of VAT compared to African Americans. This pattern of Asians having higher amounts of VAT was consistent across various total adipose tissue levels [21–23]. One prevailing hypothesis is as pre-menopausal women gain weight, they preferentially deposit adipose tissue in SAT, but as SAT expansion reaches capacity, these women begin to deposit adipose tissue in VAT. Understanding how menopausal status impacts adipose tissue distribution in obese women is still an important area of research. Lastly, oestrogens also influence the capacity of the adipocyte to be filled with lipid. As will be discussed further below, oestrogens also influence the nature of fat cells themselves, making them more ‘expandable’—which refers to the ability of the fat cell to store more calories [9,16,24]. Once a cell has reached its capacity to store lipid, lipid is deposited ectopically which is associated with increased inflammation and the metabolic syndrome.

Sex differences in adipose tissue distribution may have evolved owing to differences in the metabolic requirements of reproduction. The nutritional and metabolic demands for reproduction are significantly higher in women when compared to men. SAT deposition in females provides fuel during peak times of energy requirements such as during breastfeeding, which may account for the asymmetry seen in adipose tissue deposition [24].

Androgens, such as testosterone, also influence VAT deposition. In men, there might have been a selective advantage to storing small amounts of adipose tissue in VAT because adipose tissue turnover is higher in VAT in men when compared to women [25,26]. VAT deposition in men is associated with elevated free fatty acids, triglyceride levels, postprandial insulin and higher rates of lipolysis, all of which lead to increased fatty acid deposition and the development of metabolic syndrome [25,26].

In evolutionary terms, our ancestors had little opportunity to accrue excess body adiposity in VAT owing to a limited food supply and the high-energy requirements of resource procurement. There are disadvantages associated with storing excess VAT in both men and women owing to greater risks of morbidity and mortality [26,27].

3. Circulating oestrogens

Oestrogens are a group of compounds that include oestrone, oestradiol and oestriol. Oestradiol is further separated into 17 alpha-oestradiol (17α-E2) and 17 β-oestradiol (17β-E2). The most predominant form is 17β-E2. Typically, the common term ‘oestrogen’ refers to 17β-oestradiol [28,29]. Oestrogens are synthesized from cholesterol and metabolized in the liver by the cytochrome P450 (CYP) superfamily of enzymes. The most common enzyme CYP19A1, also called aromatase, is notable for converting androgens to oestrogens [30]. The aromatase enzyme is found in a variety of tissues including the brain, adipose tissues, vessels, bone and the gonads. The highest levels of aromatase expression are found in the gonads, which is why the main site of oestrogen synthesis in pre-menopausal females occurs in the granulosa cells of the ovaries and reproductive tract. In the years following menopause, adipose tissue (specifically VAT) becomes a major source of oestrogens, yet despite this localization, post-menopausal circulating oestrogens are lower than at any time during the menstrual cycle in pre-menopausal women. Adipose tissue levels of oestrogens in post-menopausal women are also lower than in pre-menopausal women [28,31,32].

4. Oestrogen receptors

Oestrogens regulate gene transcription via binding and activating ligand-dependent transcription factors called nuclear hormone receptors (NHRs) [30]. Oestrogens bind predominately to two ERs, ER alpha and beta (ERα; ERβ), also known as ESR1 and ESR2. The main difference between the two-receptor proteins is that ERβ has a shorter amino-terminal domain than ERα [33]. As members of the NHR superfamily of transcription regulators, the ERs are both composed of various functional domains termed A/B, C, D and E/F. The amino-terminal domain is composed of the A/B domain and controls gene transcription via its zinc-finger mediated sequence binding. The C region corresponds to the DNA binding domain and contributes to ER dimerization upon binding. The D domain is the hinge region that connects the C and E domains as well as binds chaperone proteins. Importantly, this region also contains the nuclear localization signal that is exposed when oestrogen binds the receptor, allowing for the complex to translocate to the nucleus and start the downstream process. Lastly, two activation factors (AF1 and AF2), regulate transcriptional activities [33,34].

Upon ER activation and nuclear localization, the ER complex binds DNA, interacts with transcription factors, and becomes tethered to the transcriptional co-regulators to communicate with the DNA polymerase complex to initiate transcription. Interestingly, ligand-activated ERs function in a cyclical fashion—the ER transcription complex cycles on-and-off target promoters if oestrogens are present. This process could indicate an evolutionary mechanism by which cells adapt to the external environment [35,36].

Although ERα and ERβ are the predominant ERs, oestrogens bind to and activate another cell-signalling pathway via the G protein-coupled oestrogen receptor (GPER), a receptor that is part of a larger class of classic G protein-coupled receptors. Binding of oestrogens to GPER stimulates the Gα subunit, which initiates a cascade of events by activating adenylate cyclase to increase cyclic adenosine monophosphate production. This cascade in turn leads to the activation of protein kinase A, calcium mobilization and ultimately to the transactivation of epidermal growth factor receptor [37,38].

5. Impact of oestrogen receptors on adipose tissue function

(a) . Oestrogen receptor alpha/ESR1

Adipose tissue function is influenced by ERα and this has been demonstrated in human studies and mouse models. Impairments in oestrogens ability to interact with ERα, regardless of sex, cause dysregulation in glucose homeostasis, increased adiposity and recapitulate various aspects of the metabolic syndrome [37,38]. Initial observations of the role of ERα and adipose tissues came from mouse-knockout models devoid of the ERα gene, Esr1 (ERαKO). ERαKO mice have increased adiposity, with a near doubling of VAT relative to their age-matched wild-type counterparts. Further exploration into the role of adipose tissue ERα suggested ERα selective deletion from adipocytes resulted in male and female mice with larger adipocytes and a higher expression of markers of macrophage infiltration and markers of fibrosis when compared to wild-type mice [39]. These important findings suggest that adipocyte ERα protect against excess adiposity, inflammation and fibrosis in both sexes. By contrast to these findings, as will be discussed in further detail below, selective activation of ERα in adipose tissues induces lipolysis, thus providing free fatty acids as fuel which facilitates adipose beiging [40]. These data collectively suggest an important role for adipocyte ERα in regulating adipose tissue function.

(b) . Oestrogen receptor beta/ESR2

The role of ERβ in adipose tissues is less clear. Deletion of the Erβ (Esr2) gene results in a subtle phenotype and causes mild disruption in adipose tissue function [41,42]. In fact, initial metabolic characterization of mice lacking ERβ indicated that there was no appreciable impact on adipose tissue function [41]. However, analysis of ERαKO mice suggests an important role for ERβ in adipose tissues. ERαKO mice have a 10-fold increase in circulating 17β-oestradiol and the excess oestradiol binds to ERβ contributing to the metabolic phenotype in ERαKO mice [43]. Additionally, the removal of the ovaries and reductions in circulating oestrogens in ERβ null mice protects these mice from obesity and a metabolic phenotype despite the fact these mice have increased adiposity without insulin resistance [44]. Lastly, in rodent models, selective activation of ERβ increases energy expenditure through enhanced adipose tissue mitochondrial activity [45,46] and activation of ERβ leads to an increase of mitochondrial biogenesis, which ultimately enhances lipolysis of adipose tissues [46–49].

(c) . Ratio of oestrogen receptor alpha to oestrogen receptor beta in adipose tissues

Recent data suggest the relative ratio of ERα to ERβ in adipose tissues is critical in determining the function of these receptors. A study by Park et al. [50], found the ratio of ERα to ERβ in humans in different adipose tissue depots influences adipose tissue function and insulin sensitivity. Specifically, in pre-menopausal women, there is an increased ratio of ERα to ERβ when compared to post-menopausal women. Moreover, the administration of 17-β-oestradiol to pre-and post-menopausal women changes the ratio of ERα to ERβ within WAT, suggesting that oestrogens influence the adipose tissue ER ratio in humans [49]. Other studies have found that the proportion of ERα and ERβ messenger RNA (mRNA) levels is different in VAT than in SAT depots, with VAT depots showing overall higher ERα mRNA levels than SAT [50,51]. Finally, there are data to suggest both ERs are required for proper adipose tissue energetics and mediate metabolic flexibility. Whether a certain ratio of ERs are required for optimizing these effects is not clear [52]. Additional exploration into the impact of ER isotype activation, ER expression profiling, ER receptor–DNA interaction, and non-genomic activities of oestrogens/ERs and their relative ratio within adipose tissues is needed to enhance our understanding of the role of both ERs in adipose tissue function.

6. Influence of splice variants and polymorphisms in the oestrogen receptor alpha gene (ESR1) on metabolism and adipose tissue function

Polymorphisms represent natural sequence variants (alleles) which occur in the general population. The ERα gene (ESR1) in humans is a large genomic sequence spanning about 300 kb on chromosome 6p25.1 and it has eight exons spanning greater than 140 kb. ERα comprises 595 amino acids and has a molecular weight of 66 kDa. As described above, it has six domains, A–F from the N- to the C-terminal regions. Regulation of ERα is largely done through its nine promoter (A, B, C, D, T2, T1, E1, F and ENH1) regions upstream from the transcriptional start site of the human ESR1. Several studies have indicated that the A and ENH1 promoters are the most used regions in ESR1 transcription. There are many (more than 30) transcription factors and complexes that bind to and regulate this receptor which suggests the complexity of its expression [52–54].

Several ERα polymorphisms exist in nature. In humans, ERα single-nucleotide polymorphisms (SNPs) have been linked to breast cancer susceptibility, osteoporosis, hypertension and adipose tissue distribution [53,54]. Specifically, there are two SNPs in the first intron of the ERα gene: T → C polymorphism and an A → G polymorphism. In humans, SNPs in the ERα gene impact adipose tissue distribution [55,56]. For instance, in a study by Yamada et al., SNPs in middle-aged Japanese pre-menopausal women were associated with increased total and VAT deposition when compared to women who lacked this particular polymorphism [57,58]. Moreover, a study of Egyptian women with ERα SNPs was associated with the development of metabolic syndrome [59,60]. One study by Speer et al., found an association between ERα gene SNPs and obesity phenotype. In particular, the A → G SNP or a combination of the T → C and A → G SNPs are associated with increased VAT in middle-aged women. However, the T → C polymorphism alone was not associated with a shift in body adipose tissue distribution [55].

Further studies have shown that the A → G SNP is associated with the android body fat distribution in pre-menopausal women, whereas the same SNP in post-menopausal women is associated with a reduction in total body and abdominal adipose tissue deposition. Interestingly, there appears to be an inverse relationship between the ESR1 rs2234693 polymorphism and the prevalence of obesity in women, where carriers of the rs2234693-C minor allele were 38% less likely to be obese [60]. The mechanisms by which these processes occur is still under investigation. Currently, human and rodent data indicate there are five SNPs in ERβ in both males and females which are associated with obesity [51,61,62].

7. Genome-wide association studies and sex differences in adipose tissue distribution

The advent of genetic sequencing through GWAS suggests body fat distribution is influenced by a set of genes and is heritable [63]. Analysis of genetic variants across the genome has identified more than 100 genetic loci which influence body adiposity. Additionally, a study in 1990 by Selby et al. [64], in male twins showed that there was a 31% heritability of the waist-to-hip ratio. Another population-based study estimated the heritability of body fat distribution to be about 39% [65]. Interestingly, the heritability of adipose tissue distribution appears to be greater in women when compared to men, suggesting an evolutionary propensity for sexually dimorphic adipose tissue distribution. Additionally, the genetic loci associated with body fat distribution are physically located near genes that are expressed in adipose tissue [66,67].

8. Sex chromosomes influence on adipose tissue distribution

Most studies to date have focused on the role of sex hormones and their contribution to adipose tissue distribution and sex chromosomes have typically not been factored into the overall analysis [68,69]. However, sexual dimorphism in body adiposity arises prior to onset of puberty, suggesting factors such as sex chromosomes account for body fat distribution. Females have two XX chromosomes, whereas males have one X and one Y sex chromosome [69]. Additionally, the sex-determining-region Y (SRY) located on the Y sex chromosome encodes a transcription factor that is responsible for initiating testes development in males. In the absence of the SRY, the formation of ovaries and ovarian-related hormones result, and testes do not develop [70].

The influence of sex chromosomes on sex differences naturally begins at the earliest stages of development. Sex chromosomes influence embryonic development and fetal size with males being larger and heavier than females even prior to gonadal differentiation [71,72]. An example of sex chromosomal influence is evident in the event of the formation of two X sex chromosomes in the presence of one Y sex chromosome (XXY). In XXY sex chromosome mouse models, the presence of the Y chromosome is associated with increased adiposity and dyslipidemia. Further evidence is found in humans where the XXY genotype confers Klinefelter syndrome (KS). KS occurs in approximately 1 in 600 births and is considered one of the most common sex chromosome disorders. Males with KS have hypogonadism, are infertile, and have increased measures of abdominal obesity, elevated fasting glucose levels, elevated triglyceride levels, reduced high-density lipoprotein levels and hypertension [72–75].

Furthermore, in rodent models, evidence for the effects of sex chromosomes on body adiposity and body fat distribution have been found from studies investigating the manipulation of gonadal development [71–73]. For instance, manipulation of the SRY region, which is essential in the development of testes, allows for the generation of different combinations of testes and sex chromosomes. In the four-core genotype mouse model, the SRY region is deleted from the Y sex chromosome and expressed on a non-sex chromosome (or autosome) [76]. This halts the development of testes, and its transgenic expression in females allows for the generation of four sexes: XX mice with either male or female gonads, and XY mice with either male or female gonads. These mouse models have been instrumental in understanding the influence of sex chromosomes on adiposity [77,78].

Investigations using the four-core mouse models have also led to important findings in distinguishing the impact of hormones versus sex chromosomes in determining body weight and adipose tissue deposition. Specifically, mice underwent gonadectomy around the age of 75 days to remove any gonadal hormone production and were then followed to the age of 10 months. Despite the removal of the hormonal influence, XX mice were heavier and had a greater measured body adiposity than XY mice. This difference was also greater in the XX gonadal female mice, despite the absence of gonadal secretion [77,78]. Additionally, XX mice gained more SAT versus XY mice, further supporting the influence of sex chromosomes in adiposity. Interestingly, when XO, XY and XXY were compared to XX mice after gonadectomy, the XX animals have higher body weight and VAT compared to mice with the Y sex chromosome. This indicates that the presence of the Y sex chromosome appears to have no effect on the adipose distribution and instead suggests that the presence of two X chromosomes determines body weight and fat distribution [11,79,80]. These and other findings indicate that the male-to-female differences in the number of X chromosomes influence body weight and adipose deposition in the opposite direction to the male–female differences in gonadal hormones.

9. Oestrogens influence on adipose tissue lineage and commitment

Adipose tissue is believed to have mesodermal origin. This makes up the axial, intermediate, lateral plate and paraxial mesoderm, which generates local adipose tissue and gives rise to the differential expression of genes involved in the development of different adipose tissue depots. Adipose tissue contains adipose progenitor cells that undergo adipogenesis and become lipid-laden adipocytes. These embryonic stem cells are influenced by differing signalling systems such as Nodal, Wingless and Int-1 (WNT), and Hedgehog among others, which are instrumental in influencing differentiation into mesodermal stem cells [80,81]. Adipocyte development is an emerging area of research, and suggests the mesodermal stem cells differentiate to a common white pre-adipocyte and then progresses into SAT or VAT [82].

In addition to mesenteric origin, other adipocyte precursors include myeloid progenitors, neural crest cells and pericytes. Interestingly, SAT and VAT differ with respect to the expression of genes depending on their precursors. For instance, WAT from VAT has higher expression of Nc2fl, HoxA4, HoxA5, Gpc4 and HoxC8 while those adipocytes from SAT have higher expression of Sfpr2, En1, Shox2, Tbx15, Twist1, HoxA10 and HoxC9. This variance is associated with differences in transcriptional regulation and gene expression and has been associated with a different propensity in the development of metabolic disorders associated with body fat distribution patterns [83,84].

Moreover, sex hormones also regulate various aspects of the adipocyte lineage progression and mature adipocyte function, and this appears to vary based on circulating oestrogen levels [85]. ERα expression (not ERβ) is highest in mature adipocytes when compared to pre-adipocytes. Adipose progenitor cells are a minority component of the adipose tissue stromal vascular fraction, which also contains immune cells, fibroblasts, neuronal cells, endothelial and smooth muscle cells. Adipose progenitor cells are essential to the development and maintenance of adipose tissue and may underlie aspects of adipose tissue plasticity. Importantly, progenitor cells of adipocytes reside in the perivascular niche and oestrogens regulate their function in a manner that potentially influences adipose lineage [85,86]. In studies where mice had their ovaries removed, low levels of circulating oestrogens alter adipose progenitor cell number and proliferation [79,84,86]. Moreover, oestrogen activation of ERα signalling promotes white adipose progenitor cell identity and lineage formation. In the absence of ERα, cellular reprogramming of white adipose progenitor cells yields smooth muscle or brown adipose tissue (BAT) differentiation [86]. These findings are important and can have significant physiological consequences as adipose progenitor cells influence in vivo plasticity, allowing the formation of a variety of cell types depending on the relevant morphogenic signalling context. Sexual dimorphisms also influence oestrogens regulation of pre-adipocytes, adipocyte growth and function. Studies by Anderson et al. [32] have shown that oestrogens stimulate the proliferation of pre-adipocytes from both sexes. However, SAT and VAT pre-adipocytes from females were more responsive to oestrogens and proliferated faster when compared to cells from males [32]. Overall, the above findings indicate that oestrogens and ERα are important components of pre-adipocyte differentiation, adipose tissue distribution, inflammation and fibrosis.

10. Ostrogens influence on inflammation and epigenetic regulation of adipose tissues

Excess fat accumulation, regardless of depot, is linked to the development of metabolic disorders such as increased insulin resistance, type 2 diabetes mellitus and dyslipidemia [1,2]. Oxidative stress, endoplasmic reticulum stress and lipo-toxicity as a result from excess lipid accumulation are associated with ‘unhealthy’ adipose tissues. These cellular responses are associated with chronic low-grade inflammation in adipose tissues which in turn has been linked to the development of metabolic syndrome [87]. Underlying the inflammatory response is a myriad of released cytokines and mobilization of T cells, B cells, macrophages, monocytes and granulocytes [88,89].

Epigenetic modifications of cells such as adipocytes have been implicated in the development of excess fat accumulation and obesity. Epigenetic modifications influence DNA methylation, histone modifications, and the association of transcription factors and non-coding RNA species which lead to chromatic structure changes and gene expression changes. Adipose tissue development is influenced by DNA methylation and demethylation, histone tail modification and chromatin remodelling [90]. Importantly, in the regulation of adipocytes, ERs, under the influence of oestrogens, have been shown to be involved in the epigenetic process through the recruitment of regulatory and epigenetic remodelling enzymes [91,92].

ERα has been shown to promote de novo DNA methylation by indirect recruitment and activation of DNA methyltransferase-3 (DNMT3) to oestrogen response elements (EREs). Mapping of the binding patterns of ERα in different adipose tissues has shown a strong preference for binding to intergenic regions where CpG islands are also found. During adipocyte differentiation, CpG island methylation is influenced by oestrogens binding to and activating ERα gene transcription. The underlying mechanism involves ERα-induced recruitment of protein co-repressors and other protein complexes involved in the methylation process of CpG islands in the ERE [93,94].

Interestingly, both ERα and ERβ are also involved in demethylation. It is believed that activation of ERs leads to transcriptional inhibition of DNMT1, which leads to reductions in adipogenesis. Moreover, investigations by Al-Qahtani et al. [95] suggest oestrogens influence VAT lipogenic gene expression epigenetically through demethylation thereby improving fatty acid utilization in mice in VAT.

11. Oestrogens influence on adipose tissue microenvironment

Oestrogens are also generated within tissues by the activation of aromatase which is encoded by the Cyp19 gene. The process under which oestrogens are produced is a complex multistep process and studies have shown the aromatase enzyme is extensively expressed in many cells. Aromatase expression in pre-menopausal women is found primarily in the granulosa cells and the corpus luteum of the ovaries. Other sites of aromatase expression include male gonads, epididymis, germ cells, syncytotrophoblasts, fetal tissue, adipose mesenchymal tissue, bone osteoblasts, osteoclasts, skeletal muscle, smooth muscle and vascular endothelium.

Gene expression of Cyp19 is under tissue-specific regulation [96]. The gene is located at chromosome 15q21.1 and contains nine coding exons and a 5′-untranslated region. Many studies have characterized the existence of 11 tissue-specific promoter/first exons, which regulate CYP19A1 expression. For instance, promoter I.4 is used to direct expression of aromatase in adipose tissues, while 1f is specific to the brain. Regulation of aromatase within different tissues is quite extensive and beyond the scope of this review, but it is important to discuss that the regulators of peripheral aromatase expression include hormones such as oestrogens, androgens, cytokines and other growth factors. Interestingly, no ERE has been reported in the promoter regions, though oestrogens play an important role in aromatase transcription by recruitment of ERα leading to histone modifications. Testosterone and oestradiol have also been implicated in inducing differential expression of aromatase [97–100].

Initial studies on the role of aromatase in regulating levels of oestrogens were uncovered by using models of aromatase deficiency in mouse and human studies [101,102]. Aromatase deficiency in humans is extremely rare and results in low levels of oestrogens, reproductive abnormalities, delayed bone maturation, truncal obesity, hyperinsulinemia and dyslipidemia [103–106]. These phenotypes are recapitulated in aromatase-null mice [105]. To further elucidate the importance of aromatase in adipose tissues, studies in mouse models overexpressing the aromatase enzyme in adipose tissues resulted in increased levels of oestrogens, reductions in adiposity and improvements in insulin sensitivity [107].

12. Oestrogens influence adrenergic receptor expression in adipose tissues

Androgens and oestrogens influence the activation of adrenergic receptors (ARs) alpha (αAR) and beta (βAR) receptors. Activation of βAR stimulates lipolysis of adipose tissues while αAR increases adipose tissue mass. Given that women and men have different body types and distinctive adipose tissue distribution (gynoid and android), AR expression studies suggest there are sex differences in AR expression pattern within adipose tissues. For example, women tend to have greater expression of αAR and a lower βAR response in gluteal–femoral SAT (thus higher adiposity) when compared to abdominal SAT and VAT [108–110].

Moreover, in SAT, oestrogens upregulate the αAR receptor in adipose tissues which results in decreased lipolysis. By contrast, oestrogens do not appear to affect the concentration of αAR receptors in adipocytes from VAT [111]. SAT from pre-menopausal women have higher αAR density and lower lipolytic activity in response to adrenaline than do SAT from men [112]. Importantly, the pro-lipolytic effects of oestrogens have been found to be blunted specifically in female SAT via oestrogen-mediated increase in anti-lipolytic αAR [113]. Interestingly, this was not observed in VAT which may help to explain why only SAT, and not VAT in females, is affected by changes in serum levels of oestrogens and how oestrogens overall have anti-obesity effects but at the same time promotes fat storage subcutaneously [112,113].

13. Oestrogens influence on lipolysis/lipogenesis within adipose tissues

Lipid storage in adipose tissue is dependent on the rate of fatty acid uptake and its conversion to triglyceride molecules. When metabolic demand is increased, triglycerides are hydrolyzed and released into circulation to be used as energy for other tissues. At the core of this process is lipoprotein lipase (LPL), an enzyme that catalyses the rate-limiting step in the hydrolysis of circulating lipoproteins. Studies investigating the role of oestrogens in regulating adipose LPL synthesis have been mixed, with some studies supporting findings that oestrogens decrease LPL function, while others suggest that oestrogens do not alter LPL activity [114,115]. For instance, a study in females found that LPL production decreases in gonadal adipose tissue when treated with oestrogens [116,117]. Other rodent and in vitro studies have also supported the role of oestrogens in LPL gene expression. By contrast, previous research supports a different role of oestrogens—one where oestrogens play little if any role in LPL synthesis [117]. Studies investigating LPL activity in SAT in humans have found no difference in pre- and post-menopausal women [116]. Nevertheless, despite the incongruence of the above studies in animal models and humans, there is enough evidence to demonstrate that oestrogens play an important role in adipose tissue lipid metabolism.

14. Oestrogens influence on adipogenesis via peroxisome proliferator-activated receptor gamma

Oestrogens play a direct role in suppressing lipid storage genes and preferentially activating lipolytic pathways. For instance, in adipocytes and cancer cells, both ERα and ERβ receptor activation has been shown to directly interact with peroxisome proliferator-activated receptor gamma (PPARy) [118]. PPARy is an important protein that regulates gene expression via direct binding of DNA. In particular, PPARy is an isoform of the peroxisome proliferator-activated receptor family that is an important regulator of glucose and lipid metabolism in adipose tissue. Briefly, PPARy increases adipocyte size via activating triacylglycerol synthesis and storage and promotes the maturation of pre-adipocytes into mature adipose cells. Oestrogens have been found to affect PPARy expression via activation of ERα and ultimately affecting lipid deposition [119,120]. Specifically, stimulation of ERα has been shown to directly inhibit PPARy activity in several animal and human studies. One investigation found that in rodents which underwent ovariectomy, ERα agonism reduced the expression of PPARy. In agreement, cell models that are deficient in ERα have greater PPARy expression than untreated cells, further reinforcing the inhibitory effect of ERα on PPARy [121,122].

15. Oestrogens influence on brown and beige adipose tissue function and mass

As previously described, adipose tissue is typically distinguished by its structure and features and divided into WAT, BAT and beige adipose tissue (BET). The colour of the adipose tissue differs by the amount of lipid content within a lipid droplet. WAT has a large amount of stored lipid and few mitochondria. Contrarily, BAT has an increased number of mitochondria and a smaller amount of lipid stores when compared to WAT [123,124].

BAT can use lipids as energy in the form of heat in a process called non-shivering thermogenesis [124]. BAT is principally found in newborns which require a high amount of BAT to maintain temperature. In the postnatal period, BAT is rapidly depleted, and it is only found in small amounts in adults, with women having a higher amount when compared to men. Recent studies have shown that BAT may be involved in glucose metabolism and in the regulation of body weight and energy balance [125,126]. Thus, thermogenic fat is clinically desirable and may possess anti-obesity and anti-diabetic properties. For this reason, BAT has been the focus of intense research in the past decade. Evidence exists that adrenergic and oestrogenic signalling may be involved in BAT remodelling. Because ARs are also under the control of sex hormones, it is important to investigate the effect of oestrogens on the expression of AR in BAT [127,128].

A more recent focus within adipose tissue biology is BET. Graff and colleagues demonstrated that smooth muscle cells could generate BET cells but not BAT [127]. Further, they showed that the deletion of ERα within the adipose lineage results in a fate switch favouring smooth muscle cells that were accompanied by BET formation [127,128]. Yet, in opposition, Clegg and co-workers [40] demonstrated that the activation of ERα promotes BET cell formation within WATs. The effects of oestrogens on thermogenic fat and its activity may also be indirect. An additional observation showed that melanocortin receptors are expressed along neuronal projections that can stimulate Ucp1 expression and may also be activated by oestrogens and facilitate BET formation [129,130]. However, it appears that oestrogen's role in thermogenesis may be more complex than initially hypothesized, and further studies aimed at specific tissues and cell types will be informative. Because of the therapeutic potential of BET, it will be critical to continue to evaluate the role of oestrogens and ERs in controlling thermogenesis and BET.

16. Summary

Oestrogens have an overt function in adipose tissue distribution in males and females, and this is associated with the pleiotropic roles oestrogens have on human health and disease which have arisen from evolution. Decades of clinical and basic science research have elucidated the myriad of mechanisms and signalling pathways by which oestrogens drive sexual dimorphisms. Indeed, it is now known that oestrogens play an essential role not only in reproduction but is also important in adipose tissue expansion and function [5]. In this review, we discuss some of the ways oestrogens influence our health by focusing on their impact on adipose tissues. Beginning with pre-adipocytes, oestrogens promote the differentiation of precursors into mature adipocytes and thereby control the accumulation of fat in adipose tissues and contribute to the female and male, gynoid and android bodies, respectively [7]. This means that women tend to have more subcutaneous fat—an adaptation thought to be evolutionarily advantageous in supporting childbirth and breastfeeding [5–7]. Moreover, oestrogens have metabolic effects on adipose tissues, including increasing insulin sensitivity, promoting glucose uptake and decreasing inflammation [1–4]. These effects may help protect against metabolic disorders such as obesity, diabetes mellitus and cardiovascular disease, differences apparent in epidemiological data demonstrating a lower incidence of these conditions in pre-menopausal women [6,7]. Additionally, oestrogens regulate the expression of genes involved in BAT and BET function and thermogenesis [40]. These findings are important in the context of the current obesity epidemic where new therapies and drug development can potentially find innovative ways to increase energy expenditure and facilitate weight loss. Nonetheless, more research is still required to discover novel oestrogen-binding targets and determine exactly how oestrogens and their receptors regulate adipose tissue function.

Data accessibility

This article has no additional data.

Authors' contributions

O.B.: writing—original draft, writing—review and editing; B.F.P.: writing—original draft, writing—review and editing; D.J.C.: conceptualization, writing—original draft, writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

We received no funding for this study.