Role of developmental transcription factors in white, brown and beige adipose tissues

    Role of Developmental Transcription Factors in White, Brown and Beige Adipose Tissue Catriona Hilton, Fredrik Karpe, Katherine E. Pinnick PII: DOI: Reference:

S1388-1981(15)00051-7 doi: 10.1016/j.bbalip.2015.02.003 BBAMCB 57764

To appear in:

BBA - Molecular and Cell Biology of Lipids

Received date: Revised date: Accepted date:

16 October 2014 8 January 2015 3 February 2015

Please cite this article as: Catriona Hilton, Fredrik Karpe, Katherine E. Pinnick, Role of Developmental Transcription Factors in White, Brown and Beige Adipose Tissue, BBA Molecular and Cell Biology of Lipids (2015), doi: 10.1016/j.bbalip.2015.02.003

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

SC R

Catriona Hilton1, Fredrik Karpe1, 2 and Katherine E. Pinnick1

IP

T

Role of Developmental Transcription Factors in White, Brown and Beige Adipose Tissue

1

Oxford Centre for Diabetes, Endocrinology and Metabolism, Radcliffe Department of Medicine,

University of Oxford, Oxford, UK 2

NU

NIHR Oxford Biomedical Research Centre, OUH Trust, Churchill Hospital, Oxford, UK

MA

Correspondence:

Dr C Hilton ([email protected]) or Dr KE Pinnick ([email protected])

D

Oxford Centre for Diabetes, Endocrinology and Metabolism,

TE

Churchill Hospital, Oxford,

CE P

OX3 7LE.

AC

UK.

Running title: Developmental Transcription Factors in Adipose Tissue

Word count: 8501

Figures: 1

1

ACCEPTED MANUSCRIPT Abstract In this review we discuss the role of developmental transcription factors in adipose tissue biology with a focus on how these developmental genes may contribute to regional variation in adipose

T

tissue distribution and function. Regional, depot-specific, differences in lipid handling and signalling

IP

(lipolysis, lipid storage and adipokine/lipokine signalling) are important determinants of metabolic

SC R

health. At a cellular level, preadipocytes removed from their original depot and cultured in vitro retain depot-specific functional properties, implying that these are intrinsic to the cells and not a function of their environment in situ. High throughput screening has identified a number of developmental transcription factors involved in embryological development, including members of

NU

the Homeobox and T-Box gene families, that are strongly differentially expressed between regional white fat depots and also between brown and white fat. However, the significance of depot-specific

MA

developmental signatures remains unclear. Developmental transcription factors determine body patterning during embryogenesis. The divergent developmental origins of regional adipose tissue depots may explain their differing functional characteristics. There is evidence from human genetics

D

that developmental genes determine adipose tissue distribution: in GWAS studies a number of

TE

developmental genes have been identified as being correlated with anthropometric measures of adiposity and fat distribution. Additionally, compelling functional studies have recently implicated

CE P

developmental genes in both white adipogenesis and the so-called ‘browning’ of white fat. Understanding the genetic and developmental pathways in adipose tissue may help uncover novel

AC

ways to intervene with the function of adipose tissue in order to promote health.

2

ACCEPTED MANUSCRIPT Introduction Adipose tissue (AT) can be viewed as a heterogeneous multi-depot organ comprising the energy-storing white AT (WAT) and the thermogenic brown AT (BAT) [1]. In addition, a third class of

T

adipocyte can be found within certain WAT depots, the so called “brite” or “beige” adipocytes,

IP

which appear functionally similar to classical brown adipocytes [2, 3]. Clear anatomical divisions can

SC R

be drawn between WAT and BAT; the major WAT depots in humans are located intra-abdominally (omental, mesenteric and perirenal; jointly referred to as visceral) and subcutaneously (abdominal, gluteal and femoral). Smaller WAT depots are found throughout the human body including in the pericardial region, the retro-orbital space, within the bone marrow and on the face [4, 5]. By

NU

comparison, BAT is mostly located in the neck and supraclavicular regions of adult humans [6]. The emergence of clusters of beige adipocytes within certain WAT depots is inducible by various stimuli

MA

including prolonged cold exposure and beta-adrenergic agonists through a process now referred to as ‘browning’ [7]. However, there remains much uncertainty relating to the development of different AT classes and sub-depots.

D

At the morphological level some examples of AT depot differences are very evident; for

TE

instance, adipocytes found in the well demarcated BAT depots are characterised by multilocular lipid droplets, express UCP1 and contain numerous mitochondria, consistent with the thermogenic

CE P

properties of these tissues [8, 9]. Beige adipocytes, in common with brown adipocytes, are UCP1 positive thermogenic cells with multilocular lipid droplets [10]. White adipocytes, in contrast, are defined by the presence of a large unilocular lipid droplet reflecting the important functional role the white adipocyte plays in lipid storage [5]. Other depot differences are less apparent, but equally

AC

important. Although white adipocytes from different anatomical regions may appear superficially similar, upon closer inspection, depot-specific differences can be observed relating to their histological arrangement [5, 11]. There are differences in adipocyte size between AT depots, for example visceral adipocytes are often reported to be smaller than subcutaneous [12-14] although not all findings are in agreement which likely reflects differences in techniques and participants [15, 16]. Moreover, differences in cell size, collagen fiber arrangement and microvasculature have been noted between subcutaneous WAT depots isolated from various regions of the body [17]. The subcutaneous abdominal AT depot is further divided by Scarpa’s fascia into the deep subcutaneous and superficial subcutaneous AT layers. Histologically, the deep subcutaneous AT layer contains a larger proportion of small adipocytes [18], higher expression of inflammatory genes and increased macrophage infiltration [19]. The WAT depots also display functional differences in terms of fatty acid handling, adipokine production and capacity for preadipocyte proliferation and adipogenesis [20, 21]. Furthermore, the cellular mechanisms by which AT expands during obesity appears to differ 3

ACCEPTED MANUSCRIPT between depots; an over-feeding study by Tchoukalova et al showed that upper-body WAT depots expand through increased adipocyte size (hypertrophy) whereas lower-body depots expand as a result of increased adipocyte number (hyperplasia) [22].

T

The distinct functional characteristics of individual AT depots are likely, in part, to contribute

IP

to the paradoxical relationships observed between body fat distribution and metabolic health [2325]. In humans, lower-body WAT depots (e.g. gluteal and femoral) appear to confer a level of

SC R

metabolic protection and are associated with improved metabolic profiles including lower fasting plasma glucose and triglycerides, and higher HDL-cholesterol [26]. By comparison, the upper-body WAT depots (e.g. subcutaneous abdominal and visceral) are more closely linked with an increased

NU

risk of cardiovascular disease and type 2 diabetes [23-25]. There are also differences between abdominal deep and superficial subcutaneous WAT with the depth of the deep layer correlating with insulin resistance and cardiovascular risk [18]. A further division can be drawn between visceral and

MA

subcutaneous obesity [27, 28], with visceral WAT displaying stronger associations with metabolic risk factors than subcutaneous WAT, independently of the overall level of obesity [27]. Ectopic deposition of fat (pericardial, hepatic and inter-muscular) is also associated with increased

D

cardiometabolic risk, however because of the close relationship between visceral WAT and ectopic

TE

fat deposition it becomes difficult to dissect the individual contribution of these fat depots [29]. Furthermore, the prevalence and activity of BAT also relates to metabolic risk factors. BAT is a

CE P

thermogenic organ important for non-shivering temperature maintenance in infants and also present and metabolically active in adult humans [8]. In adults BAT is inversely associated with BMI [30, 31], fasting plasma glucose [32] and type 2 diabetes status [31]. Recent findings have shown

AC

that the induction of BAT by cold exposure is associated with increased energy expenditure and a reduction in overall body fat, suggesting that BAT recruitment may pose an important anti-obesity target [33].

Growing evidence suggests the functional characteristics of various fat depots are determined by intrinsic mechanisms rather than simply being a function of their anatomical locations. For instance, the transplantation of subcutaneous WAT into the visceral compartment of mice leads to improved insulin sensitivity, glucose tolerance and adiposity. By comparison the transplantation of visceral WAT is without effect suggesting that the metabolic improvements exerted by subcutaneous WAT are the result of intrinsic properties of this tissue [34, 35]. At the cellular level, preadipocytes isolated from their native WAT depots retain many of the functional features of that depot in vitro including depot-specific lipolytic activity [36], adipokine secretion [36, 37] and fatty acid handling [38, 39]. Regional differences are also apparent relating to preadipocyte proliferative and adipogenic capacity, as well as sensitivity to apoptotic stimuli (recently reviewed

4

ACCEPTED MANUSCRIPT [20]). The depot-of-origin memory exhibited by preadipocytes is consistent with the view that different AT depots comprise of distinct preadipocyte subpopulations. In the quest to understand these intrinsic differences investigators have performed high-throughput transcriptional profiling of

T

regional AT depots to identify variations in genetic transcripts. This has led to the identification of

IP

unique transcriptional ‘signatures’ that are AT-depot specific [40-43]. Of particular note are the striking differences in expression of transcription factors involved in embryonic development and

SC R

body pattern specification, and which are apparent in isolated preadipocytes even after repeated population doublings [44]. Whether these genes are residual developmental markers and whether they play active roles in AT function remains unclear.

NU

In this review we will provide an overview of the proposed embryological origins of individual AT depots and summarise the developmental transcription factors which have been identified as differentially expressed between depots. We will explore the role of these genes in AT

MA

biology, considering their contribution to regional variation in AT distribution and function and the subsequent impact this may have on metabolic health.

TE

D

Embryological Origins of Adipose Tissue

Each class of AT (WAT, BAT and beige) exhibits its own unique transcriptional profile characterised

CE P

by striking differences in the expression of developmental transcription factors [45], yet our understanding of the embryological origins of individual fat depots is still far from complete. WAT depots are heterogeneous in terms of their embryological origin [46]. In the human embryo primitive fat lobules become visible between gestational weeks 14 - 16. Fat depots in the head and

AC

neck are the first to develop, followed by the trunk and then the limbs [47]. In mice, using doxycycline-inducible labelling of mature adipocytes it has been possible to accurately pin-point the development of different AT depots. Specifically, subcutaneous WAT develops between embryonic days 14-18 whereas epididymal (visceral) fat develops postnatally [48]. With the exception of facial AT, which derives from the ectoderm [49], AT is generally considered to derive from the mesoderm [50]. Lineage tracing studies in mice have recently shown that visceral WAT develops from Wilms’ tumour gene (Wt1)-expressing cells found specifically in the lateral plate mesoderm. By comparison, Wt1-expressing cells make no contribution to subcutaneous WAT leaving the exact embryological origins of this depot unknown [46]. These findings suggest that individual AT depots arise from different anatomical regions of the mesoderm. There is further heterogeneity amongst individual visceral depots with regards to the contribution of Wt1 cells, demonstrating the complexity of AT as a multi-depot organ.

5

ACCEPTED MANUSCRIPT Mesenchymal stem cells (MSC) give rise to cells of the adipocyte lineage through a process involving an early commitment step followed by terminal adipocyte differentiation. In humans the number of mature white adipocytes remains relatively constant throughout adult life, reflecting the

T

fine balance between adipocyte death and adipogenesis. It has been estimated that approximately

IP

10% of white adipocytes turnover each year [51] and interestingly this rate was reported not to be influenced by age or changes in energy balance. Terminal adipogenic differentiation sees the

SC R

committed adipocyte precursors or “preadipocytes” acquire the functional characteristics of the mature adipocyte. The adipogenic differentiation of preadipocytes is tightly regulated by a welldefined transcription factor cascade [52] involving the transient expression of early transcriptional

NU

regulators (CEBPβ, CEBPδ) followed by the induction of critical adipogenic regulators (CEBPα, PPARγ) which co-ordinate the expression of many adipocyte genes. Although the stages of terminal adipogenic differentiation have been well characterised [52] early events in the commitment of MSC

MA

to the adipocyte lineage are less clear. Resident progenitors have been identified within the adipose stromovascular fraction [53] and there is evidence to support the view that MSC originate from the perivascular region. By genetically marking PPARG-expressing cells Tang et al were able to

D

investigate adipose lineage specification in mice in vivo. They demonstrated that adipocyte

TE

progenitors are already committed during postnatal development and traced the source of these progenitors to the AT vasculature [54]. Further studies have also pointed to a perivascular or

CE P

endothelial contribution to the adipose lineage [55, 56]. However, there are also other non-adipose sources of progenitor cells to be considered; bone marrow-derived haematopoietic stem cells, for instance, appear capable of migrating to AT depots and undergoing adipogenic differentiation [57,

AC

58].

A great deal of interest has been directed towards uncovering the origins of the brown and beige adipocyte lineages. The beige adipocytes found within WAT depots and possess a unique gene expression profile compared to both BAT and WAT [10]. An early transcriptional profiling study by Timmons et al. identified a striking enrichment of myogenic genes (including myogenin, Myf5+, MyoD) in rodent brown preadipocytes when compared to white preadipocytes, suggesting that the developmental origins of brown adipocytes were more similar to skeletal muscle than to WAT [59] and challenging the view that brown and white adipocytes shared a common early adipogenic precursor [4]. This was followed by lineage tracing studies in mice where it was reported that both brown adipocytes and skeletal muscle cells were derived from Myf5+ expressing precursors, while by comparison the Myf5+ lineage did not give rise to white and beige adipocytes [60]. It has since been suggested that beige adipocytes may share more similarities with smooth, rather than skeletal, muscle [61]. The close relationship between BAT and skeletal muscle was highlighted further when

6

ACCEPTED MANUSCRIPT BAT was shown to be capable of inter-conversion to skeletal muscle (and vice versa) through the modulation of the BAT transcriptional regulator PRDM16 [60]. Thus the view was adopted that white and brown adipocyte lineages could be distinguished using Myf5+ as a marker [62]. However, in

T

reality the divisions and level of plasticity between white, brown and beige adipocytes seem to be

IP

far more complex. A subset of white preadipocytes have in fact been found to derive from Myf5+ progenitors, traditionally thought to only give rise to brown fat [63, 64], and preadipocytes from

SC R

both the Myf5- and Myf5+ lineages can give rise to beige adipocytes following beta-adrenergic stimulation [63]. Work conducted by Shan et al. has shown there to be varying proportions of Myf5+ lineage cells between different WAT depots which may account for the discrepancies between

NU

lineage tracing studies which have examined the contribution of Myf5 [60, 64]. Furthermore, these Myf5+ progenitors can give rise to cells of both the myogenic and adipogenic lineages [64]. Another important question is to what extent trans-differentiation of white adipocytes

MA

contributes to the recruitment of beige adipocytes. In rodents, quantitative analysis has shown that although the number of beige adipocytes increase in WAT during cold exposure overall there is no difference in the total number of adipocytes or any sign of increased mitosis [65], which is consistent

D

with the view that trans-differentiation of white adipocytes contributes to beige adipocyte

TE

recruitment [65]. Using an in vivo lineage tracing approach it has recently been shown that mature white adipocytes have the potential to inter-convert to beige adipocytes [66]. Cold-induced beige

CE P

adipocytes were found to undergo “whitening” when exposed to a warm phase however approximately 50% of these cells later returned to the beige phenotype when exposed to a second cold phase. These findings indicate that bi-directional inter-conversion of white and beige adipocytes

AC

can occur in WAT [66]. Also of interest is the notion that the mechanisms for beige adipocyte recruitment may differ between WAT depots. Recent lineage tracing studies have indicated that in response to beta3-adrenergic stimulation beige adipocytes differentiate from precursors within some WAT depots [48, 67], and that these precursors also have the potential to differentiate into white adipocytes in the absence of a ‘browning’ stimulus, particularly under conditions of high fat feeding. However, in other WAT depots the primary method of recruitment of beige adipocytes was by trans-differentiation of existing white adipocytes [67]. This direct transformation of white into beige adipocytes is accompanied by an increase in noradrenergic innervation [65]. A recent murine study has suggested that regional differences in the capacity for browning are determined by intrinsic cues rather than variation in the abundance of beige cell progenitors [68]. The majority of studies investigating the origins of the various adipocyte lineages are conducted in murine models and the full implications for human AT biology remain to be elucidated.

7

ACCEPTED MANUSCRIPT WAT and Metabolic Health Obesity is an important risk factor for a myriad of health problems including type 2 diabetes, cardiovascular disease and certain cancers [69]. However, approximately 10-30% of obese

T

individuals do not exhibit the typical features of obesity-associated metabolic disease such as insulin

IP

resistance, hypertension and dyslipidaemia [70]. This subgroup has been labelled the “metabolically

SC R

healthy obese” and would seem to imply that it is not the absolute amount of AT that determines metabolic health. Furthermore, not all fat depots confer equal metabolic risk. Indeed, waist to hip ratio (WHR), as a proxy measure of body fat distribution, is a better indicator of myocardial infarction risk than body mass index (BMI) [23]. The visceral WAT compartment in particular has

NU

been shown to correlate with metabolic risk more strongly than subcutaneous WAT [71], and in fact, the relative proportion of visceral to subcutaneous AT is a predictor of metabolic risk independent of

MA

both overall adiposity and visceral AT [71]. Although a causative relationship between visceral WAT and metabolic health has not been proven [72], it has been suggested that visceral fat may promote hepatic insulin resistance by directly releasing bioactive factors such as non-esterified fatty acids and

D

cytokines into the portal vein [73]. Subsequently, hepatic steatosis and inflammation may mediate

TE

systemic insulin resistance through the generation of inflammatory cytokines [74]. There is also evidence that functional differences between VAT and SAT could contribute to the stronger

CE P

relationship between VAT and metabolic risk. These include increased lipolytic activity and a reduced capacity for preadipocyte proliferation and adipogenesis in the visceral compartment (reviewed in [20]). The expression and secretion of a number of adipokines also differs between VAT and SAT. For example, adiponectin, which exerts insulin-sensitizing actions [75], is expressed at significantly lower

AC

levels in VAT compared to SAT in humans [76], whereas the pro-inflammatory cytokine, interleukin6, is secreted at higher levels from adipocytes from the visceral depot compared to those from the subcutaneous depot [77]. Equally, the lack of adequate amounts of subcutaneous AT may lead to a failure of the subcutaneous AT to appropriately store lipids resulting in increased visceral and ectopic fat deposition [78]. Indeed, the necessity to maintain functional AT expansion, the so-called “adipose tissue expandability” hypothesis [79] is illustrated by the morbidly obese AdTG-ob/ob mouse which is genetically modified to possess an unlimited capacity for AT expansion and is markedly more insulin sensitive than ob/ob littermates [80]. A distinction in relation to metabolic risk can also be made according to upper- and lowerbody obesity. Individuals with preferential lower-body adiposity, who display a ‘pear’ body shape, are in general metabolically healthier than BMI-matched individuals with upper-body adiposity or an ‘apple’ body shape [81]. Specifically, the accumulation of upper-body subcutaneous and visceral fat is associated with increased risk of myocardial infarction [23] and type 2 diabetes [24] whereas fat 8

ACCEPTED MANUSCRIPT accumulation on the gluteofemoral regions is paradoxically associated with reduced risk. In addition, when adjusted for total fat mass, DXA-quantified gynoid fat is inversely associated with insulin resistance (HOMA-IR), fasting plasma triglycerides and C-reactive peptide in an apparently healthy

T

adult population. By comparison, the opposite is true for android fat mass which displays a positive

IP

association with these risk factors [43].

It is interesting to note that the associations between metabolic risk and upper- or lower-

SC R

body adiposity differ not just in terms of the strength of association but also in directionality. For this reason it is proposed that the gluteofemoral AT depots exert a protective effect which may reflect the distinct underlying traits of these depots (as extensively reviewed [21, 81]). Overall, it would

NU

seem that the upper-body depots play a more active role in the storage and mobilisation of dietderived lipids whilst lower-body depots turnover more slowly and provide stable long-term fat storage [82, 83]. The adipokine secretory profiles of gluteofemoral and abdominal subcutaneous AT

MA

also differ. Using in vivo arterio-venous sampling to measure adipokine release from gluteofemoral AT we have shown that lower-body AT displays a more “protective” adipokine secretory profile releasing higher amounts of the insulin-sensitising lipokine palmitoleate and lower amounts of the

D

pro-inflammatory cytokine interleukin-6 [39, 43]. Further research is warranted to better understand

CE P

BAT and Metabolic Health

TE

the mechanisms underlying the protective effects of gluteofemoral AT.

BAT is inversely associated with age, overall adiposity as assessed by BMI, and type 2 diabetes in humans [6, 31] and, because of its potential involvement in whole-body energy expenditure, poses

AC

an attractive anti-obesity target. The relevance of BAT to metabolic health has been most extensively studied in rodents. Indeed, the study of Ucp1-deficient mice housed under thermoneutral conditions has highlighted the potential involvement of BAT in controlling adiposity; Ucp1-deficient mice displayed a 50% increase in fat mass compared to wild type controls [84]. However, whereas in rodents BAT is a well-defined entity that is clearly distinct from WAT and beige fat, there is evidence that this is not the case in humans. Rather it has been challenged that BAT depots in adult humans are comprised primarily of beige adipocytes [10, 85]. Despite sharing common properties it remains unclear whether beige adipocytes perform the same role as brown adipocytes [86] and whether there are sufficient numbers of beige adipocytes to influence metabolic health. There is some preliminary evidence to suggest that this may be the case. Deletion of PRDM16, a transcriptional regulatory protein involved in beige adipocyte induction, has been associated with a metabolic phenotype [87]. Adipocyte-specific deletion of PRDM16 inhibited the induction of thermogenic activity in beige adipocytes in subcutaneous AT depots following cold 9

ACCEPTED MANUSCRIPT exposure or β3 agonist treatment. On a high-fat diet PRDM16-deficient mice went on to develop obesity, insulin resistance and hepatic steatosis [87]. Furthermore, mice overexpressing PRDM16

T

exhibit increased beige adipocytes and are metabolically healthier on a high-fat diet [88]

IP

What dictates body fat distribution?

SC R

The genetic predisposition to body shape is relatively strong, with heritability estimates for waist-tohip ratio falling in the range of 40% -70% [89, 90]. However, despite this the precise genetic mechanisms determining body fat distribution remain poorly understood. The study of rare monogenic forms of partial lipodystrophy have identified a limited set of genes which, when

NU

perturbed, result in the selective loss of certain WAT depots (CIDEC, PLIN1, AKT2, PPARG, LMNA) [91]. These genes all appear to play important roles in adipocyte function [92], although it is

MA

sometimes unclear how they exert a regional phenotype. In addition, over 20 genetic loci have now been associated with body fat distribution through genome-wide association studies (GWAS) using anthropometric measures or computed tomography [93]. Perhaps rather disappointingly these loci

D

explain only a small proportion of the variation for body fat distribution (less than 5%) and, since

TE

most are located outside exonic regions, it remains an ongoing challenge to identify the causative genes and to understand their biological relevance. Common polymorphisms associated with body

CE P

fat distribution have also been identified using a candidate approach to screen genes with known mechanistic roles in adipocyte biology. Polymorphisms have been identified in genes relating to catecholamine action (the adrenergic receptors) [94] and adipogenesis (PPARG) [95] and also in genes involved in steroid hormone metabolism and action such as NR3C1 (the glucocorticoid

AC

receptor) [96] and ESR1 (the oestrogen receptor a) [97]. But it would seem likely that there are further genetic determinants of body fat distribution which remain to be uncovered. There are also other mediators which can influence body fat distribution. The powerful effects steroid hormones exert on fat distribution have been widely reported [98]. Body shape exhibits striking sexual dimorphism. In general, men have less total body fat than women of the same BMI, whilst being more prone to upper-body fat deposition [99]. Furthermore, sex hormones have been shown to directly alter body composition and fat distribution in prospective measurements of transsexuals receiving hormone treatment [100]. The contribution that glucocorticoids make to body fat distribution is evident in individuals exhibiting the clinical condition of Cushing’s syndrome in which elevated cortisol levels are accompanied by increased central adiposity and a diminution of peripheral AT depots [101]. Changes in body fat distribution are also associated with the ageing process. Typically, an increase in total adiposity is apparent with a specific increase in visceral fat but a relative loss of subcutaneous fat, particularly in the gluteofemoral depot 10

ACCEPTED MANUSCRIPT [102]. Notably in females an increase in visceral adiposity is associated with post-menopausal oestrogen deficiency [103]. Body fat distribution can also be influenced by environmental factors, for

T

example, by drugs such as antiretroviral therapies [104].

IP

The Role of Depot-Specific Developmental Transcription Factors in Body Fat Distribution

SC R

Depot-specific differences in WAT adipocyte function are often maintained when cells are removed from their native tissues and cultured in vitro [44], suggesting that depot-specific features are determined by intrinsic mechanisms rather than the influence of the local milieu. Preadipocytes isolated from separate WAT depots also exhibit different propensities for apoptosis, proliferation

NU

and adipogenesis both in humans [105, 106] and in rodents [107]. Indeed, distinct preadipocyte subpopulations have been identified which display differing distribution between SAT and VAT [106].

MA

This implies that some of the depot-specific properties reflect differences in the developmental origins of WAT depots.

Several groups have investigated the mRNA expression profiles of different AT depots using

D

genome-wide transcriptomic approaches. These studies have consistently noted distinct

TE

transcriptional signatures between BAT and WAT [41, 108, 109] and also between separate WAT depots (visceral, abdominal SAT, gluteal) [40, 42, 43, 105, 107, 110, 111]. Interestingly, many of the

CE P

genes identified are known to play a role in embryological development and body patterning. In one study, developmental genes accounted for 18.4% of the gene transcripts differing between depots, and 24.7% in undifferentiated preadipocytes [111]. Notably, depot-specific developmental gene signatures are retained in vitro, both in primary preadipocytes [40, 42, 43] and in immortalised

AC

preadipocyte cell lines after multiple rounds of cell division [43, 111]. Furthermore, a large number of these differentially expressed developmental genes are up-regulated during adipogenesis in mouse embryonic stem cells indicating that they play important roles in adipocyte development [112] . This begs the question of whether differential expression of developmental genes could be responsible for some of the functional differences between depots, rather than just being markers of their embryological origin. In the following sections we will outline how expression studies, GWAS and in vivo and in vitro functional experiments have provided evidence for a mechanistic role for developmental transcription factors in determining the differences between fat depots.

The Homeobox (HOX) Family

11

ACCEPTED MANUSCRIPT The HOX gene family are a group of related transcription factors that are important in determining body patterning along the anterior-posterior (A-P) axis during embryogenesis. They display colinearity, that is, their chromosomal arrangement in general mirrors their expression along the A-P

T

axis. Interestingly, whilst HOX expression is turned on or off to dictate body pattern in simple

IP

organisms, it is thought that quantitative differences in HOX expression are important for development of the complex tissues found in higher order species [113, 114]. HOX genes have been

SC R

identified in several studies as amongst the most differentially expressed genes between AT depots (Figure 1) [40, 41, 105, 108, 109, 115].

NU

Depot-specific expression of HOX genes: As outlined in Figure 1, HOX genes display striking expression patterns between VAT and SAT [40, 107], as well as between individual visceral (omental, mesenteric, perirenal) [111] and subcutaneous (abdominal, gluteal) depots [42, 43, 105]. Over the

MA

last decade GWAS have revolutionised the hunt for common variants that predispose to disease phenotypes. Several GWAS have focused on finding causative genes which influence obesity and body fat distribution. These studies identified a locus near HOXC13 as being associated with WHR

D

[116], and both HOXC13 and HOXB5 as risk alleles for childhood obesity [117]. The AT expression of

TE

several HOX genes has also been linked with obesity, body fat distribution and changes in nutritional status. HOXA5, for example, appears to be a marker of upper-body WAT, being more highly

CE P

expressed in visceral [40] and abdominal SAT [42, 43]. Of particular interest is the observation that HOXA5 expression is positively associated with BMI and WHR in humans [40], suggesting this gene could be involved in the expansion of regional WAT depots in obesity. However, in obese ob/ob mice

AC

expression of Hoxa5 in the perirenal and epididymal visceral depots was actually decreased compared to lean controls, as was the expression of Hoxc8 and Hoxc9 [41]. In response to fasting, expression of Hoxc9 specifically in the mesenteric AT of obese mice is markedly increased while Hoxc8 expression decreases in both visceral and subcutaneous depots, and Hoxa5 expression is unchanged [41]. Although it is important to note the differences observed between humans and rodents, which are likely due to the comparison of dissimilar depots, these observations make it tempting to speculate that the HOX network may have a functional role in determining WAT expansion. This hypothesis has been given further weight by Dankel et al who demonstrated upregulation of HOX genes (including HOXA5, HOXA9, HOXB5, HOXC6) in human AT following extreme weight loss after bariatric surgery [118]. The authors of this study also carried out promoter analysis of the genes that were differently expressed before and after weight loss. They found that 48% of the differentially expressed genes contained at least one potential Homeobox binding site in their promoter regions [118]. However, not all studies have observed associations between the

12

ACCEPTED MANUSCRIPT expression of HOX network genes and obesity [42]. In addition, depot-specific expression of homeobox-domain containing genes (e.g. SHOX2) [41] and HOX gene cofactors (e.g. PBX1, PBX3, MEIS1, MEIS2) has also been reported [42, 119]. There are also differences in HOX gene expression

T

between WAT and BAT: Hoxc8 and Hoxc9 appear to be WAT-specific being more highly expressed in

SC R

IP

WAT and beige adipocytes compared to BAT [45, 59].

HOX functional studies: The transcriptional profiling studies suggest that HOX genes and their associated cofactors may play important regulatory roles in controlling AT development and remodelling throughout adult life, rather than being simple AT depot markers. Using in vitro models,

NU

temporal changes in the expression of specific HOX genes have been observed during adipogenesis of 3T3-L1 cells (Hoxa4, Hoxa7, Hoxd4) [120] and primary human preadipocytes (HOXA10, HOXB8)

MA

[42]. However, evidence supporting a functional role for these genes in adipocyte biology remains limited. Mori et al have recently proposed that HOXC8 may function as an important regulator of the WAT lineage by inhibiting brown adipogenesis [121]. They demonstrated that the “browning” of

D

human preadipocytes derived from subcutaneous WAT is associated with a strong down-regulation

TE

of HOXC8 and that HOXC8 suppresses the expression of brown fat genes by targeting the crucial regulator of brown adipogenesis CEBPβ [121]. By comparison, treatment of WAT adipocytes with

CE P

rosiglitazone induces expression of Hoxc9 [2]. The down-regulation of HOXC8 in human adiposederived stem cells (hASCs) is associated with increased proliferation and impaired differentiation along the osteogenic lineage, however no change in adipogenic capacity was reported [122]. HOXC8 appears to be an important depot-specific determinant of cell fate.

AC

The homeobox-domain containing gene Shox2 displays dramatically higher expression in abdominal SAT than in the visceral depot in both rodents and humans [123] and is more highly expressed in gluteal preadipocytes compared to abdominal subcutaneous [43]. The AT-specific ablation of Shox2 in mice is reported to protect against diet-induced obesity (particularly of the subcutaneous depot) however, this was not accompanied by protection from insulin resistance [123]. The reduction in adipocyte cell size in Shox2-deficient mice was accompanied by markedly higher rates of lipolytic activity and increased expression of the β3 adrenergic receptor (Adrb3) [123]. Lee et al. went on to demonstrate that Shox2 directly interacts with CEBPa to repress its activation of Adrb3. These findings demonstrate the potential for Homeobox genes to regulate regional AT function and thus contribute to body fat distribution. Given these findings, genes which interact with the HOX genes pose another interesting class of genes to investigate in relation to body fat distribution. HOX cofactors include the PBX and MEIS gene classes which appear to enhance HOX gene DNA-binding specificity [124]. Indeed, Pbx has been shown to augment the targeting of specific 13

ACCEPTED MANUSCRIPT DNA sites by HOX proteins to influence skeletal patterning and chondrocyte proliferation and differentiation [125]. Pbx1-deficient MSC derived from mouse embryos are unable to generate adipocytes [126]. Conversely, the opposite results are seen in human post-natal hASCs, where

T

knock-down reduced proliferation and enhanced adipogenesis. The authors propose that Pbx1 plays

IP

a complex role in the generation and maintenance of preadipocytes [126]. Interestingly, Pbx1 has also been shown to attenuate osteogenesis in MSC by binding to Pbx sites in the promoters of

SC R

osteoblast-related genes and impairing their transcriptional activation by Hoxa10 [127]. Thus the PBX genes would appear to play important roles in cell fate determination. Transcriptional network analysis has identified PBX1 and the related HOX cofactor, MEIS1, as two transcription factors which

NU

are up-regulated during mESC adipogenesis [112]. Furthermore, in silico prediction of transcription factor binding sites amongst genes co-expressed during mESC adipogenesis identified an overrepresentation of the MEIS1 binding motif [112]. The MEIS family appear to regulate Hox protein-

MA

DNA binding specificity by forming heterotrimeric complexes with the Hox and Pbx genes [124]. A mechanism of action for Meis1 and Meis2 has not yet been determined in AT however knock-down of another family member, Prep1, has been shown to have an inhibitory effect on adipogenesis

D

[128]. Further studies are required to elucidate the complex interactions between HOX genes and

CE P

TE

their cofactors in the development of AT.

The T-Box (TBX) family

The T-Box family is a second group of transcription factors which play a vital role in determining

AC

body plan in early embryogenesis and in dictating organogenesis and cell fate later in development. Of particular relevance to AT biology, the T-Box family have been shown to be involved in mesoderm formation and patterning [129].

Depot-specific expression of TBX genes: The TBX family includes several members which display depot-specific expression patterns in AT. In rodents TBX15 is highly expressed in BAT and “beige” competent AT depots but is down-regulated by cold acclimation [45, 130]. In WAT TBX15 is expressed at markedly higher levels in visceral compared to subcutaneous AT in lean humans [40] but displays an opposite pattern of expression in mice [40, 41]. Expression of TBX15 is strongly down-regulated with BMI and WHR in visceral AT, but, interestingly, displays a positive association with BMI and WHR in subcutaneous AT. Indeed, in obese individuals it appears that the direction of differential expression of TBX15 is actually reversed with higher expression observed in subcutaneous AT [131]. These observations suggest TBX15 may play different roles in different

14

ACCEPTED MANUSCRIPT depots thus contributing to body fat distribution. In support of this, a common polymorphism near the TBX15 locus has been associated with WHR, and differential expression was also reported between gluteal and abdominal subcutaneous AT [116]. Other members of the T-Box family which

IP

T

also display depot-specific expression patterns in AT include TBX5 [43] and TBX1 [10].

TBX functional studies: Experimental findings support the view that TBX15 has AT depot-specific

SC R

roles. Using the 3T3-L1 murine cell line it has been demonstrated that Tbx15 acts as a negative regulator of adipogenesis, and that it also influences mitochondrial mass and function [132]. However, another group found that in cultured murine cells knock-down of Tbx15 only affected

NU

adipogenesis or expression of BAT markers in depots that were considered brown or ‘beige’ competent (i.e. subcutaneous WAT), and not in classical white adipocytes derived from the visceral region [130]. The authors therefore concluded that Tbx15 may be involved in determining the

MA

thermogenic programming of beige adipocytes [130]. We have recently identified another member of the T-box family which appears to be involved in the regulation of adipogenesis [43]. TBX5 was found to display a marked difference in expression between the gluteal and abdominal

D

subcutaneous depots, being almost exclusively expressed in abdominal preadipocytes. Silencing

TE

TBX5 in human abdominal subcutaneous preadipocytes reduced proliferative capacity and led to a failure to differentiate along the adipogenic lineage. These findings suggest TBX5 plays an important

CE P

functional role in the development of the abdominal subcutaneous AT depot. Tbx1 has also been identified as a marker of murine beige fat and human BAT [10, 85]. Its functional role in the determination of cell fate or in the function of beige fat remains to be elucidated. There is evidence

AC

to suggest that TBX2 may also play a role in adipogenesis. A study performed on multi-potent stem cells derived from a haemangioma demonstrated that silencing TBX2 inhibits adipogenesis [133] but it is unclear whether TBX2 also plays a role in healthy AT. Finally, the down-regulation of TBX3 is associated with reduced proliferation of hASC [134]. However, although osteogenic differentiation was impaired in these cells no effect on adipogenesis was reported [134]. Together these findings suggest that the TBX transcription factors have important roles in both the development and function of adipocytes, and that regional expression of this gene family may significantly influence body fat distribution.

Other Differentially Expressed Developmental Transcription Factors The FOX genes: The forkhead box (FOX) proteins are a large family of 50 transcription factors organised into 19 subgroups which share a conserved forkhead DNA-binding domain and play an important role in development during embryogenesis and in adult tissues [135]. Multiple members 15

ACCEPTED MANUSCRIPT of the FOX family have been shown to be up-regulated during the early stages of adipogenesis in mESC [112] and some also display AT depot-specific expression patterns. In obese humans, FOXA3 exhibits higher expression in visceral than abdominal subcutaneous AT [136]. In line with this

T

observation mice with ablation of FOXA3 have reduced epididymal (i.e. visceral) fat mass [136],

IP

suggesting a depot-specific role in AT expansion. FOXA3 has been shown to enhance adipogenesis through cooperation with CEBPβ and CEBPδ [136], but the precise mechanism by which it might

SC R

preferentially expand different fat depots is uncertain. FOXC1 is more highly expressed in gluteal AT compared to abdominal subcutaneous [43]. Recently Foxc1 has been identified as a critical regulator of haematopoietic stem cells [137]. The authors demonstrated that Foxc1 is highly expressed in bone

NU

marrow MSC which can give rise to the adipogenic and osteogenic lineages. Deletion of Foxc1 in the MSC resulted in increased adipocytes within the bone marrow. In vitro studies showed that overexpression of Foxc1 in the OP9 murine preadipocyte cell line significantly decreased their adipogenic

MA

capacity [137]. Several other members of the FOX family have also been shown to be regulated during adipogenesis in cell culture models [50, 138]. Some of these act as positive regulators of white adipogenesis, such as FOXF2 [138]. Conversely, other members of the family, such as FOXA2

D

[139], FOXO1 [140] FOXC2 [141], act to inhibit white adipogenesis. Further in vivo characterisation of

TE

these genes has demonstrated that Foxa2+/- mice display increased weight gain on a high-fat diet compared with littermate controls [139] whilst Foxo1 ablation in mice protects from diet-induced

CE P

insulin resistance and leads to reduced adipocyte hypertrophy in epididymal fat pads [140]. Transgenic mice overexpressing FOXC2 in both BAT and WAT display resistance to diet-induced obesity and are insulin sensitive [142]. Notably, the intra-abdominal WAT depot in FOXC2 transgenic

AC

mice acquired characteristics of BAT in terms of mRNA profile (e.g. induction of Ucp1), histological features and increased oxygen consumption. These findings implicate FOXC2 as an important regulator of beige adipocyte development. By performing mechanistic studies the authors were also able to demonstrate that FOXC2 mediates these effects by increasing sensitivity of the β-adrenergic cAMP-protein kinase A pathway [142]. In an obese human population FOXC2 expression within visceral AT inversely correlates with fasting serum insulin levels [143]. The 5’ untranslated region of the FOXC2 gene contains a common polymorphism (C-512T), the TT genotype of which is associated with higher expression of FOXC2 within visceral than subcutaneous WAT. Furthermore, the T allele was associated with enhanced insulin sensitivity [143].

The IRX genes: The Iroquois (Iro) gene family share a TALE homeodomain regulatory sequence and are organised genomically into two clusters each containing three genes [144]. The Iro genes IRX1

16

ACCEPTED MANUSCRIPT and IRX2 are expressed at differing levels in gluteal and abdominal subcutaneous AT, both being more highly expressed in the abdominal depot [42, 43]. The functional role of IRX1 and IRX2 is unclear however the related family member IRX3 lies adjacent to the FTO gene, which has been well

T

described in GWAS studies to associate with obesity [145]. Ragvin et al demonstrated that the GWAS

IP

variants in FTO overlap with regulatory regions for IRX3 [146] and it has since been shown that Irx3deficient mice exhibit a marked reduction in body weight, largely due to reduced fat mass. Irx3-

SC R

deficient mice also exhibit signs of browning of WAT depots with the up-regulation of brown markers such as Ucp1, Cidea and Prdm16 [147]. Furthermore IRX3, as well as an additional family member IRX6, were amongst genes which were found to be up-regulated in AT following significant

NU

weight loss in bariatric patients [118]. The findings suggest that members of the IRX family play

MA

important roles in AT function and further studies are required to investigate this.

The TCF family: The TCF transcription factors act within the canonical (β-catenin dependent) Wnt signalling cascade to modulate the expression of numerous transcripts involved in embryonic

D

development, cell proliferation, adipogenesis and apoptosis [148]. The Wnt family are well

TE

established as regulators of adipogenesis [149] and polymorphisms within or near genes encoding Wnt signalling proteins (RSPO3, KREMEN1) have been identified through GWAS as associated with

CE P

body fat distribution [116]. Of particular note, members of the TCF family, TCF21 and TCF7L2, have been observed to be AT depot-specific. In humans, TCF7L2 is enriched within abdominal subcutaneous AT when compared to gluteal AT [43]. Variants in the TCF7L2 gene have been

AC

associated with increased type 2 diabetes risk and a mechanism of action has been proposed through an effect on Wnt signalling in enteroendocrine cells [150]. However, there is also evidence to suggest that TCF7L2 may contribute to increased diabetes risk by exerting effects on adiposity and adipocyte biology. Indeed, the TCF7L2 risk allele is associated with a lower BMI in patients with type 2 diabetes [151]. TCF7L2 can be alternatively spliced to give rise to transcripts of differing length. A short (Ex12−13−13a−mRNA) TCF7L2 variant has been shown to be up-regulated during adipogenic differentiation of the human Simpson-Golabi-Behmel syndrome (SGBS) preadipocyte cell strain [152]. Furthermore, expression of the short TCF7L2 variant was decreased in human subcutaneous AT following weight loss after bariatric surgery [152]. With the well-known role that Wnt signalling plays in adipocyte development it seems likely that regional expression of genes in this pathway may contribute to body fat distribution. A second member of the TCF family, Tcf21, was observed to be specific to white adipocytes in vitro [59] and was the most highly differentially expressed gene in visceral preadipocytes compared to subcutaneous [107]. Tcf21 also displays enhanced expression in

17

ACCEPTED MANUSCRIPT WAT in mice, as opposed to ‘brite’ or BAT [45]. However the exact functional role Tcf21 plays in AT remains uncertain.

T

The TWIST genes

IP

The TWIST genes encode a family of conserved basic helix-loop-helix (bHLH) transcription factors that are essential to mesoderm development during embryogenesis [153]. TWIST1 expression in

SC R

subcutaneous WAT is higher than in visceral WAT and is reduced in obesity and increased by weight loss [154]. Within WAT, TWIST1 regulates fatty acid oxidation and the secretion of several inflammatory cytokines [155] whereas in BAT TWIST1 interacts with PGC-1α to reduce BAT

NU

metabolism [156].

Other developmental markers of BAT: By performing detailed transcriptomic analysis of murine

MA

primary brown and white preadipocytes Timmons et al identified a set of distinct markers for the two cell lineages [59]. In addition to members of the HOX (Hoxa7, Hoxc8, Hoxc9) and TBX (Tbx15) families, which we have previously discussed, they identified other developmental transcription

D

factors including Meox2, Lhx8, Zic1 and Tcf21. Meox2, Lhx8 and Zic1 were all shown to be markedly

TE

expressed in brown preadipocytes whilst being essentially absent in white preadipocytes. Interestingly, the same group later demonstrated that this novel set of brown markers was not

CE P

expressed in white preadipocytes which were continuously exposed to rosiglitazone to induce “browning” [2], although as discussed above beige adipocytes may arise by different mechanisms depending on their location within the WAT depots.

AC

MEOX2 encodes the protein MOX2 which regulates mesodermal patterning [157] and is essential for normal vertebrate limb myogenesis [158]. In perivascular adipocytes MEOX2 has been shown to counteract the stimulatory effects of IGF1 on adipocyte proliferation and differentiation [159]. MEOX2 has also been reported to exert anti-angiogenic effects in endothelial cells [160] and overexpression of MEOX2 induces premature fibroblast senescence [161]. The depot-specific expression of MEOX2 in AT suggests this transcription factor may play an important role in the development of BAT. Cypess et al [6] described the nature of AT in adult human subjects undergoing neck surgery. They observed a gradient in the histological properties of samples taken from different horizontal levels, rather than discrete compartments of BAT and WAT. This was accompanied by a gradient in mRNA expression profiles with the AT closest to the skin expressing almost no UCP1 whilst the deepest AT expressed the highest level of UCP1, with a corresponding gradient in the expression of ZIC1 and LHX8. A second group also confirmed ZIC1 and LHX8 as markers of human BAT collected

18

ACCEPTED MANUSCRIPT from the supraclavicular region [162]. ZIC1 encodes a transcriptional activator from the zinc finger protein of the cerebellum family and its function would seem to be crucial for a variety of developmental processes including development of the neural crest and left-right axis patterning. It

T

remains unclear whether these developmental transcription factors are necessary for determining

IP

BAT function or if they are a markers of anatomical position and embryological origin. Another developmental transcription factor implicated in BAT development is Ebf2 which

SC R

belongs to the COE (collier/olfactory-1/early B cell factor) family of transcription factors which are critical for neurogenesis and neuronal migration. Ebf2 is preferentially expressed in brown [163, 164] and Myf5+ beige [164] adipocytes and is required for maintaining the identity of brown precursors

NU

[164] and differentiated brown adipocytes [163]. It has been suggested that Ebf2 determines adipocyte lineage by regulating PPARy binding activity [163].

MA

How might developmental gene signatures be maintained?

As we have seen, numerous developmental transcription factors display AT depot-specific expression patterns which are maintained into adulthood. Some genes, such as the HOX family, are

D

known to be modulated by endocrine hormones including oestrogen, progesterone and

TE

testosterone [165]. These hormones can influence HOX gene expression both during embryonic development and also in adult tissues, and this may provide one mechanism for the differences in

CE P

fat distribution observed between genders and in some endocrine conditions. However, the observation that isolated preadipocytes retain depot-specific expression patterns when cultured in vitro also points to the involvement of epigenetic mechanisms [166]. Epigenetic regulation involves

AC

heritable differences in gene expression that are not the result of differences in the DNA sequence. This can be achieved by the action of depot-specific non-coding RNAs (e.g. microRNAs, long noncoding RNAs), through the altered modification of histone proteins or through differences in DNA methylation. MicroRNAs are endogenous small non-coding RNAs that form an elegant layer of regulation by post-transcriptionally modulating gene expression. MicroRNAs are important orchestrators of AT biology [167]. For example, microRNA(miR)-155 [168] and miR-27 [169] regulate brown and beige adipocyte differentiation by targeting CEBPβ [168] and PPARγ, respectively [169]. Several microRNAs have been shown to act on developmental patterning genes [170], such as miR-196a which is a positive regulator of “browning” of WAT via its targeting of HOXC8 [121]. Of particular interest, some of these microRNAs also have depot-specific expression patterns in human AT [171, 172], although their precise contribution to depot-specific properties remains to be investigated. A related family which also acts at the level of transcriptional fine tuning are the long non-coding RNAs

19

ACCEPTED MANUSCRIPT (lncRNAs). Research into this important class of transcriptional regulators in AT is still in its infancy however multiple lncRNAs have been shown to display regulated expression patterns during adipogenesis in both white and brown preadipocytes [173]. Furthermore, RNA interference studies

T

identified a set of lncRNAs which were functionally required for adipogenesis [173]. Some lncRNAs

IP

are known to modulate HOX gene expression (e.g. Hotair, Hog and Tog) [174] and it is therefore plausible that regional differences in lncRNA expression may be involved in the maintenance of AT

SC R

depot-specific developmental gene expression. Intriguingly, one of these lncRNAs, HOX antisense intergenic RNA (HOTAIR), is one of the most differentially expressed transcripts to be identified between gluteal and abdominal subcutaneous AT [43, 175]. HOTAIR is expressed from within the

NU

HOXC gene cluster on chromosome 12 and has been shown to silence the HOXD locus in trans [176]. Furthermore, the role of HOTAIR as a depot-specific regulator of adipogenesis has recently been investigated by Divoux et al who ectopically expressed HOTAIR in preadipocytes derived from

MA

abdominal subcutaneous AT (which display very low expression of HOTAIR) [175]. The overexpression of HOTAIR resulted in increased adipogenic differentiation and expression of proadipogenic genes [175] and implicates HOTAIR as an important depot-specific regulator of

D

adipogenesis. These findings clearly highlight the need for further investigation into the role lncRNAs

TE

play in AT development and function.

Changes in histone protein modification and DNA methylation are also likely to be important

CE P

epigenetic determinants of AT development and function. Histone acetyltransferases (HATs) and deacetylases (HDACs) are two important classes of histone modifying enzymes which have been implicated in the regulation of key adipogenic and myogenic transcription factors e.g. PPARG, PGC1a

AC

and MYOD (the role of HATs and HDACs in adipogenesis has recently been extensively reviewed [177]). Notably, the histone deacetylase sirtuin 1 (Sirt1) has been shown to promote the browning of WAT by deacetylating Pparγ at lysine residues required to recruit Prdm16, thus favouring the induction of BAT genes [178]. Furthermore, the overexpression of Sirt2 inhibits white adipogenesis in 3T3-L1 cells whereas a reduction in Sirt2 expression is pro-adipogenic. Sirt2 targets a number of transcription factors including the developmental gene Foxo1 [179]. Thus depot-specific differences in the expression of histone modifying enzymes, such as the sirtuins, may play an important role in determining depot differences in AT development and function. The methylation of DNA at cytosine residues located within CpG dinucleotides produces a stable modification of the DNA which regulates gene expression by altering protein-DNA interactions and chromatin structure. Discrete differences in the DNA methylation profiles of white and brown AT have been reported in rodents [180] and between individual WAT depots in humans [119]. It is possible that these heritable DNA methylation patterns are involved in maintaining the depot-

20

ACCEPTED MANUSCRIPT specific expression of developmental genes. Indeed, it has recently been shown that genes displaying differential methylation between abdominal and gluteal subcutaneous AT in humans are enriched for members of the HOX gene family [119]. Furthermore, by employing bisulphite

T

sequencing we have recently demonstrated that immortalized preadipocyte cell lines derived from

IP

abdominal and gluteal subcutaneous AT exhibit distinct depot-specific differences in promoter DNA methylation for the developmental genes HOTAIR and TBX5 [43]. These differences in DNA

SC R

methylation were consistent with the depot-specific expression patterns of HOTAIR and TBX5 which were maintained in the abdominal and gluteal preadipocyte cell lines across multiple cell generations [43]. The observation that differentially methylated regions in AT are often related to

NU

developmental genes is in keeping with previous comparisons of healthy human tissues (including blood, brain and muscle) which have reported tissue-specific differentially methylated regions to be over-represented at genetic loci critical for developmental processes [181]. It is possible that depot-

MA

specific DNA methylation patterns are required for the programming and maintenance of specific

D

adipocyte cell lineages.

TE

CONCLUSIONS

Human AT distribution is an important independent risk factor for metabolic disease and so

CE P

much research has sought to uncover the determinants of this relationship. Transcription factors involved in embryogenesis and body patterning are overrepresented in transcripts that are expressed in an AT depot-specific manner and have also been implicated functionally in determining

AC

the characteristics of the different AT depots. As such, developmental transcription factors including the HOX and TBX gene families look promising as candidates for modulating metabolic risk profiles. Given the clinical importance of understanding AT distribution and our limited knowledge to date of the mechanisms involved, this emerging area of research promises to be exciting.

21

ACCEPTED MANUSCRIPT ACKNOWLEDGEMENTS Funding to support our contribution to this research field has been received from the Medical Research Council, Novo Nordisk UK Research Foundation, Heart Research UK, EU FP6 MolPAGE

T

(LSHG/512066) and EU FP7 LipidomicNet (#202272). We also wish to thank the NIHR Oxford

IP

Biomedical Research Centre for supporting the Oxford Biobank without which our human AT work

AC

CE P

TE

D

MA

NU

SC R

could not have been done.

22

ACCEPTED MANUSCRIPT FIGURE LEGENDS

T

Figure 1: Depot-specific expression of developmental transcription factors in regional AT depots.

IP

Arrows indicate expression levels relative to the abdominal subcutaneous AT depot since this is the most consistent comparison in the literature. Data are compiled from Karastergiou et al. [42],

SC R

Gehrke et al. [119], Pinnick et al. [43], Gesta et al. [40], Yamamoto et al. [41], Tchkonia et al. [111], Xu et al. [136], Macotela et al. [107], Vohl et al. [110], Timmons et al. [59], Wu et al. [10] and Sharp

AC

CE P

TE

D

MA

NU

et al. [85].

23

ACCEPTED MANUSCRIPT

8. 9. 10. 11. 12. 13. 14.

15. 16. 17. 18.

19. 20. 21. 22. 23.

T

IP

SC R

NU

7.

MA

6.

D

5.

TE

4.

CE P

3.

Cinti, S., The adipose organ at a glance. Dis Model Mech, 2012. 5(5): p. 588-94. Petrovic, N., et al., Chronic peroxisome proliferator-activated receptor gamma (PPARgamma) activation of epididymally derived white adipocyte cultures reveals a population of thermogenically competent, UCP1-containing adipocytes molecularly distinct from classic brown adipocytes. J Biol Chem, 2010. 285(10): p. 7153-64. Schulz, T.J., et al., Identification of inducible brown adipocyte progenitors residing in skeletal muscle and white fat. Proc Natl Acad Sci U S A, 2011. 108(1): p. 143-8. Gesta, S., Y.H. Tseng, and C.R. Kahn, Developmental origin of fat: tracking obesity to its source. Cell, 2007. 131(2): p. 242-56. Wronska, A. and Z. Kmiec, Structural and biochemical characteristics of various white adipose tissue depots. Acta Physiol (Oxf), 2012. 205(2): p. 194-208. Cypess, A.M., et al., Identification and importance of brown adipose tissue in adult humans. N Engl J Med, 2009. 360(15): p. 1509-17. Cousin, B., et al., Occurrence of brown adipocytes in rat white adipose tissue: molecular and morphological characterization. J Cell Sci, 1992. 103 ( Pt 4): p. 931-42. Nedergaard, J., T. Bengtsson, and B. Cannon, Unexpected evidence for active brown adipose tissue in adult humans. Am J Physiol Endocrinol Metab, 2007. 293(2): p. E444-52. Cannon, B. and J. Nedergaard, Brown adipose tissue: function and physiological significance. Physiol Rev, 2004. 84(1): p. 277-359. Wu, J., et al., Beige adipocytes are a distinct type of thermogenic fat cell in mouse and human. Cell, 2012. 150(2): p. 366-76. Frontini, A. and S. Cinti, Distribution and development of brown adipocytes in the murine and human adipose organ. Cell Metab, 2010. 11(4): p. 253-6. Van Harmelen, V., et al., Leptin secretion from subcutaneous and visceral adipose tissue in women. Diabetes, 1998. 47(6): p. 913-7. O'Connell, J., et al., The relationship of omental and subcutaneous adipocyte size to metabolic disease in severe obesity. PLoS One, 2010. 5(4): p. e9997. Zhang, Y., et al., Fat cell size and adipokine expression in relation to gender, depot, and metabolic risk factors in morbidly obese adolescents. Obesity (Silver Spring), 2014. 22(3): p. 691-7. Boivin, A., et al., Regional differences in adipose tissue metabolism in obese men. Metabolism, 2007. 56(4): p. 533-40. Maslowska, M.H., et al., Regional differences in triacylglycerol synthesis in adipose tissue and in cultured preadipocytes. J Lipid Res, 1993. 34(2): p. 219-28. Sbarbati, A., et al., Subcutaneous adipose tissue classification. Eur J Histochem, 2010. 54(4): p. e48. Marinou, K., et al., Structural and functional properties of deep abdominal subcutaneous adipose tissue explain its association with insulin resistance and cardiovascular risk in men. Diabetes Care, 2014. 37(3): p. 821-9. Tordjman, J., et al., Structural and inflammatory heterogeneity in subcutaneous adipose tissue: relation with liver histopathology in morbid obesity. J Hepatol, 2012. 56(5): p. 1152-8. Tchkonia, T., et al., Mechanisms and metabolic implications of regional differences among fat depots. Cell Metab, 2013. 17(5): p. 644-56. White, U.A. and Y.D. Tchoukalova, Sex dimorphism and depot differences in adipose tissue function. Biochim Biophys Acta, 2014. 1842(3): p. 377-92. Tchoukalova, Y.D., et al., Regional differences in cellular mechanisms of adipose tissue gain with overfeeding. Proc Natl Acad Sci U S A, 2010. 107(42): p. 18226-31. Yusuf, S., et al., Obesity and the risk of myocardial infarction in 27,000 participants from 52 countries: a case-control study. Lancet, 2005. 366(9497): p. 1640-9.

AC

1. 2.

24

ACCEPTED MANUSCRIPT

30. 31.

32. 33. 34. 35.

36. 37.

38. 39. 40. 41. 42. 43. 44. 45.

T

IP

SC R

NU

29.

MA

28.

D

27.

TE

26.

CE P

25.

Snijder, M.B., et al., Associations of hip and thigh circumferences independent of waist circumference with the incidence of type 2 diabetes: the Hoorn Study. Am J Clin Nutr, 2003. 77(5): p. 1192-7. Karpe, F. and K.E. Pinnick, Biology of upper-body and lower-body adipose tissue-link to whole-body phenotypes. Nat Rev Endocrinol, 2014. Snijder, M.B., et al., Low subcutaneous thigh fat is a risk factor for unfavourable glucose and lipid levels, independently of high abdominal fat. The Health ABC Study. Diabetologia, 2005. 48(2): p. 301-8. Fox, C.S., et al., Abdominal visceral and subcutaneous adipose tissue compartments: association with metabolic risk factors in the Framingham Heart Study. Circulation, 2007. 116(1): p. 39-48. Liu, J., et al., Impact of abdominal visceral and subcutaneous adipose tissue on cardiometabolic risk factors: the Jackson Heart Study. J Clin Endocrinol Metab, 2010. 95(12): p. 5419-26. Despres, J.P., Body fat distribution and risk of cardiovascular disease: an update. Circulation, 2012. 126(10): p. 1301-13. van Marken Lichtenbelt, W.D., et al., Cold-activated brown adipose tissue in healthy men. N Engl J Med, 2009. 360(15): p. 1500-8. Ouellet, V., et al., Outdoor temperature, age, sex, body mass index, and diabetic status determine the prevalence, mass, and glucose-uptake activity of 18F-FDG-detected BAT in humans. J Clin Endocrinol Metab, 2011. 96(1): p. 192-9. Lee, P., et al., A critical appraisal of the prevalence and metabolic significance of brown adipose tissue in adult humans. Am J Physiol Endocrinol Metab, 2010. 299(4): p. E601-6. Yoneshiro, T., et al., Recruited brown adipose tissue as an antiobesity agent in humans. J Clin Invest, 2013. 123(8): p. 3404-8. Tran, T.T., et al., Beneficial effects of subcutaneous fat transplantation on metabolism. Cell Metab, 2008. 7(5): p. 410-20. Hocking, S.L., D.J. Chisholm, and D.E. James, Studies of regional adipose transplantation reveal a unique and beneficial interaction between subcutaneous adipose tissue and the intra-abdominal compartment. Diabetologia, 2008. 51(5): p. 900-2. van Harmelen, V., et al., Increased lipolysis and decreased leptin production by human omental as compared with subcutaneous preadipocytes. Diabetes, 2002. 51(7): p. 2029-36. Hocking, S.L., et al., Intrinsic depot-specific differences in the secretome of adipose tissue, preadipocytes, and adipose tissue-derived microvascular endothelial cells. Diabetes, 2010. 59(12): p. 3008-16. Caserta, F., et al., Fat depot origin affects fatty acid handling in cultured rat and human preadipocytes. Am J Physiol Endocrinol Metab, 2001. 280(2): p. E238-47. Pinnick, K.E., et al., Gluteofemoral adipose tissue plays a major role in production of the lipokine palmitoleate in humans. Diabetes, 2012. 61(6): p. 1399-403. Gesta, S., et al., Evidence for a role of developmental genes in the origin of obesity and body fat distribution. Proc Natl Acad Sci U S A, 2006. 103(17): p. 6676-81. Yamamoto, Y., et al., Adipose depots possess unique developmental gene signatures. Obesity (Silver Spring), 2010. 18(5): p. 872-8. Karastergiou, K., et al., Distinct developmental signatures of human abdominal and gluteal subcutaneous adipose tissue depots. J Clin Endocrinol Metab, 2013. 98(1): p. 362-71. Pinnick, K.E., et al., Distinct Developmental Profile of Lower-Body Adipose Tissue Defines Resistance Against Obesity-Associated Metabolic Complications. Diabetes, 2014. Tchkonia, T., et al., Fat depot-specific characteristics are retained in strains derived from single human preadipocytes. Diabetes, 2006. 55(9): p. 2571-8. Walden, T.B., et al., Recruited vs. nonrecruited molecular signatures of brown, "brite," and white adipose tissues. Am J Physiol Endocrinol Metab, 2012. 302(1): p. E19-31.

AC

24.

25

ACCEPTED MANUSCRIPT

53. 54. 55. 56. 57.

58. 59.

60. 61. 62.

63.

64. 65. 66. 67. 68. 69.

T

IP

SC R

52.

NU

51.

MA

50.

D

49.

TE

48.

CE P

47.

Chau, Y.Y., et al., Visceral and subcutaneous fat have different origins and evidence supports a mesothelial source. Nat Cell Biol, 2014. 16(4): p. 367-75. Poissonnet, C.M., A.R. Burdi, and S.M. Garn, The chronology of adipose tissue appearance and distribution in the human fetus. Early Hum Dev, 1984. 10(1-2): p. 1-11. Wang, Q.A., et al., Tracking adipogenesis during white adipose tissue development, expansion and regeneration. Nat Med, 2013. 19(10): p. 1338-44. Billon, N., et al., The generation of adipocytes by the neural crest. Development, 2007. 134(12): p. 2283-92. Billon, N. and C. Dani, Developmental origins of the adipocyte lineage: new insights from genetics and genomics studies. Stem Cell Rev, 2012. 8(1): p. 55-66. Spalding, K.L., et al., Dynamics of fat cell turnover in humans. Nature, 2008. 453(7196): p. 783-7. Rosen, E.D. and O.A. MacDougald, Adipocyte differentiation from the inside out. Nat Rev Mol Cell Biol, 2006. 7(12): p. 885-96. Rodeheffer, M.S., K. Birsoy, and J.M. Friedman, Identification of white adipocyte progenitor cells in vivo. Cell, 2008. 135(2): p. 240-9. Tang, W., et al., White fat progenitor cells reside in the adipose vasculature. Science, 2008. 322(5901): p. 583-6. Gupta, R.K., et al., Zfp423 expression identifies committed preadipocytes and localizes to adipose endothelial and perivascular cells. Cell Metab, 2012. 15(2): p. 230-9. Tran, K.V., et al., The vascular endothelium of the adipose tissue gives rise to both white and brown fat cells. Cell Metab, 2012. 15(2): p. 222-9. Crossno, J.T., Jr., et al., Rosiglitazone promotes development of a novel adipocyte population from bone marrow-derived circulating progenitor cells. J Clin Invest, 2006. 116(12): p. 32208. Tomiyama, K., et al., Characterization of transplanted green fluorescent protein+ bone marrow cells into adipose tissue. Stem Cells, 2008. 26(2): p. 330-8. Timmons, J.A., et al., Myogenic gene expression signature establishes that brown and white adipocytes originate from distinct cell lineages. Proc Natl Acad Sci U S A, 2007. 104(11): p. 4401-6. Seale, P., et al., PRDM16 controls a brown fat/skeletal muscle switch. Nature, 2008. 454(7207): p. 961-7. Long, J.Z., et al., A smooth muscle-like origin for beige adipocytes. Cell Metab, 2014. 19(5): p. 810-20. Seale, P., S. Kajimura, and B.M. Spiegelman, Transcriptional control of brown adipocyte development and physiological function--of mice and men. Genes Dev, 2009. 23(7): p. 78897. Sanchez-Gurmaches, J., et al., PTEN loss in the Myf5 lineage redistributes body fat and reveals subsets of white adipocytes that arise from Myf5 precursors. Cell Metab, 2012. 16(3): p. 348-62. Shan, T., et al., Distinct populations of adipogenic and myogenic Myf5-lineage progenitors in white adipose tissues. J Lipid Res, 2013. 54(8): p. 2214-24. Vitali, A., et al., The adipose organ of obesity-prone C57BL/6J mice is composed of mixed white and brown adipocytes. J Lipid Res, 2012. 53(4): p. 619-29. Rosenwald, M., et al., Bi-directional interconversion of brite and white adipocytes. Nat Cell Biol, 2013. 15(6): p. 659-67. Lee, Y.H., et al., In vivo identification of bipotential adipocyte progenitors recruited by beta3adrenoceptor activation and high-fat feeding. Cell Metab, 2012. 15(4): p. 480-91. Li, Y., et al., Intrinsic differences in BRITE adipogenesis of primary adipocytes from two different mouse strains. Biochim Biophys Acta, 2014. 1841(9): p. 1345-52. Haslam, D.W. and W.P. James, Obesity. Lancet, 2005. 366(9492): p. 1197-209.

AC

46.

26

ACCEPTED MANUSCRIPT

77.

78. 79. 80. 81. 82. 83. 84.

85. 86. 87. 88. 89. 90. 91. 92. 93.

T

IP

SC R

76.

NU

75.

MA

74.

D

73.

TE

72.

CE P

71.

van Vliet-Ostaptchouk, J.V., et al., The prevalence of metabolic syndrome and metabolically healthy obesity in Europe: a collaborative analysis of ten large cohort studies. BMC Endocr Disord, 2014. 14(1): p. 9. Kaess, B.M., et al., The ratio of visceral to subcutaneous fat, a metric of body fat distribution, is a unique correlate of cardiometabolic risk. Diabetologia, 2012. 55(10): p. 2622-30. Frayn, K.N., Visceral fat and insulin resistance--causative or correlative? Br J Nutr, 2000. 83 Suppl 1: p. S71-7. Item, F. and D. Konrad, Visceral fat and metabolic inflammation: the portal theory revisited. Obes Rev, 2012. 13 Suppl 2: p. 30-9. Cai, D., et al., Local and systemic insulin resistance resulting from hepatic activation of IKKbeta and NF-kappaB. Nat Med, 2005. 11(2): p. 183-90. Kadowaki, T., et al., Adiponectin and adiponectin receptors in insulin resistance, diabetes, and the metabolic syndrome. J Clin Invest, 2006. 116(7): p. 1784-92. Lihn, A.S., et al., Lower expression of adiponectin mRNA in visceral adipose tissue in lean and obese subjects. Mol Cell Endocrinol, 2004. 219(1-2): p. 9-15. Fried, S.K., D.A. Bunkin, and A.S. Greenberg, Omental and subcutaneous adipose tissues of obese subjects release interleukin-6: depot difference and regulation by glucocorticoid. J Clin Endocrinol Metab, 1998. 83(3): p. 847-50. Despres, J.P. and I. Lemieux, Abdominal obesity and metabolic syndrome. Nature, 2006. 444(7121): p. 881-7. Danforth, E., Jr., Failure of adipocyte differentiation causes type II diabetes mellitus? Nat Genet, 2000. 26(1): p. 13. Kim, J.Y., et al., Obesity-associated improvements in metabolic profile through expansion of adipose tissue. J Clin Invest, 2007. 117(9): p. 2621-37. Manolopoulos, K.N., F. Karpe, and K.N. Frayn, Gluteofemoral body fat as a determinant of metabolic health. International Journal of Obesity, 2010. 34(6): p. 949-959. Guo, Z., et al., Regional postprandial fatty acid metabolism in different obesity phenotypes. Diabetes, 1999. 48(8): p. 1586-92. McQuaid, S.E., et al., Femoral adipose tissue may accumulate the fat that has been recycled as VLDL and nonesterified fatty acids. Diabetes, 2010. 59(10): p. 2465-73. Feldmann, H.M., et al., UCP1 ablation induces obesity and abolishes diet-induced thermogenesis in mice exempt from thermal stress by living at thermoneutrality. Cell Metab, 2009. 9(2): p. 203-9. Sharp, L.Z., et al., Human BAT possesses molecular signatures that resemble beige/brite cells. PLoS One, 2012. 7(11): p. e49452. Owens, B., Cell physiology: The changing colour of fat. Nature, 2014. 508(7496): p. S52-3. Cohen, P., et al., Ablation of PRDM16 and beige adipose causes metabolic dysfunction and a subcutaneous to visceral fat switch. Cell, 2014. 156(1-2): p. 304-16. Seale, P., et al., Prdm16 determines the thermogenic program of subcutaneous white adipose tissue in mice. J Clin Invest, 2011. 121(1): p. 96-105. Pausova, Z., et al., Heritability estimates of obesity measures in siblings with and without hypertension. Hypertension, 2001. 38(1): p. 41-7. Souren, N.Y., et al., Anthropometry, carbohydrate and lipid metabolism in the East Flanders Prospective Twin Survey: heritabilities. Diabetologia, 2007. 50(10): p. 2107-16. Garg, A., Clinical review#: Lipodystrophies: genetic and acquired body fat disorders. J Clin Endocrinol Metab, 2011. 96(11): p. 3313-25. Vigouroux, C., et al., Molecular mechanisms of human lipodystrophies: from adipocyte lipid droplet to oxidative stress and lipotoxicity. Int J Biochem Cell Biol, 2011. 43(6): p. 862-76. Wagner, R., et al., The genetic influence on body fat distribution. Drug Discovery Today: Disease Mechanisms, 2013. 10(1–2): p. e5-e13.

AC

70.

27

ACCEPTED MANUSCRIPT

100. 101.

102. 103. 104.

105. 106.

107. 108. 109. 110. 111.

112.

113. 114.

T

IP

SC R

NU

99.

MA

98.

D

97.

TE

96.

CE P

95.

Ukkola, O., et al., Interactions among the alpha2-, beta2-, and beta3-adrenergic receptor genes and obesity-related phenotypes in the Quebec Family Study. Metabolism, 2000. 49(8): p. 1063-70. Kim, K.S., et al., Effects of peroxisome proliferator-activated receptor-gamma 2 Pro12Ala polymorphism on body fat distribution in female Korean subjects. Metabolism, 2004. 53(12): p. 1538-43. Syed, A.A., et al., Association of glucocorticoid receptor polymorphism A3669G in exon 9beta with reduced central adiposity in women. Obesity (Silver Spring), 2006. 14(5): p. 759-64. Gallagher, C.J., et al., Association of the estrogen receptor-alpha gene with the metabolic syndrome and its component traits in African-American families: the Insulin Resistance Atherosclerosis Family Study. Diabetes, 2007. 56(8): p. 2135-41. Bjorntorp, P., Hormonal control of regional fat distribution. Hum Reprod, 1997. 12 Suppl 1: p. 21-5. Goodman-Gruen, D. and E. Barrett-Connor, Sex differences in measures of body fat and body fat distribution in the elderly. Am J Epidemiol, 1996. 143(9): p. 898-906. Elbers, J.M., et al., Effects of sex steroid hormones on regional fat depots as assessed by magnetic resonance imaging in transsexuals. Am J Physiol, 1999. 276(2 Pt 1): p. E317-25. Wajchenberg, B.L., et al., Estimation of body fat and lean tissue distribution by dual energy Xray absorptiometry and abdominal body fat evaluation by computed tomography in Cushing's disease. J Clin Endocrinol Metab, 1995. 80(9): p. 2791-4. Kuk, J.L., et al., Age-related changes in total and regional fat distribution. Ageing Res Rev, 2009. 8(4): p. 339-48. Lovejoy, J.C., et al., Increased visceral fat and decreased energy expenditure during the menopausal transition. Int J Obes (Lond), 2008. 32(6): p. 949-58. Villarroya, F., P. Domingo, and M. Giralt, Drug-induced lipotoxicity: lipodystrophy associated with HIV-1 infection and antiretroviral treatment. Biochim Biophys Acta, 2010. 1801(3): p. 392-9. Levi, B., et al., Depot-specific variation in the osteogenic and adipogenic potential of human adipose-derived stromal cells. Plast Reconstr Surg, 2010. 126(3): p. 822-34. Tchkonia, T., et al., Abundance of two human preadipocyte subtypes with distinct capacities for replication, adipogenesis, and apoptosis varies among fat depots. Am J Physiol Endocrinol Metab, 2005. 288(1): p. E267-77. Macotela, Y., et al., Intrinsic differences in adipocyte precursor cells from different white fat depots. Diabetes, 2012. 61(7): p. 1691-9. Cantile, M., et al., HOX gene network is involved in the transcriptional regulation of in vivo human adipogenesis. J Cell Physiol, 2003. 194(2): p. 225-36. Cheung, L., et al., Human mediastinal adipose tissue displays certain characteristics of brown fat. Nutr Diabetes, 2013. 3: p. e66. Vohl, M.C., et al., A survey of genes differentially expressed in subcutaneous and visceral adipose tissue in men. Obes Res, 2004. 12(8): p. 1217-22. Tchkonia, T., et al., Identification of depot-specific human fat cell progenitors through distinct expression profiles and developmental gene patterns. Am J Physiol Endocrinol Metab, 2007. 292(1): p. E298-307. Billon, N., et al., Comprehensive transcriptome analysis of mouse embryonic stem cell adipogenesis unravels new processes of adipocyte development. Genome Biol, 2010. 11(8): p. R80. Lempradl, A. and L. Ringrose, How does noncoding transcription regulate Hox genes? Bioessays, 2008. 30(2): p. 110-21. Pollock, R.A., et al., Gain of function mutations for paralogous Hox genes: implications for the evolution of Hox gene function. Proc Natl Acad Sci U S A, 1995. 92(10): p. 4492-6.

AC

94.

28

ACCEPTED MANUSCRIPT

122.

123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133.

134. 135. 136. 137.

T

IP

SC R

121.

NU

120.

MA

119.

D

118.

TE

117.

CE P

116.

Sevastianova, K., et al., Comparison of dorsocervical with abdominal subcutaneous adipose tissue in patients with and without antiretroviral therapy-associated lipodystrophy. Diabetes, 2011. 60(7): p. 1894-900. Heid, I.M., et al., Meta-analysis identifies 13 new loci associated with waist-hip ratio and reveals sexual dimorphism in the genetic basis of fat distribution. Nat Genet, 2010. 42(11): p. 949-60. Bradfield, J.P., et al., A genome-wide association meta-analysis identifies new childhood obesity loci. Nat Genet, 2012. 44(5): p. 526-31. Dankel, S.N., et al., Switch from stress response to homeobox transcription factors in adipose tissue after profound fat loss. PLoS One, 2010. 5(6): p. e11033. Gehrke, S., et al., Epigenetic regulation of depot-specific gene expression in adipose tissue. PLoS One, 2013. 8(12): p. e82516. Cowherd, R.M., et al., Developmental profile of homeobox gene expression during 3T3-L1 adipogenesis. Biochem Biophys Res Commun, 1997. 237(2): p. 470-5. Mori, M., et al., Essential role for miR-196a in brown adipogenesis of white fat progenitor cells. PLoS Biol, 2012. 10(4): p. e1001314. Kim, Y.J., et al., miR-196a regulates proliferation and osteogenic differentiation in mesenchymal stem cells derived from human adipose tissue. J Bone Miner Res, 2009. 24(5): p. 816-25. Lee, K.Y., et al., Shox2 is a molecular determinant of depot-specific adipocyte function. Proc Natl Acad Sci U S A, 2013. 110(28): p. 11409-14. Moens, C.B. and L. Selleri, Hox cofactors in vertebrate development. Dev Biol, 2006. 291(2): p. 193-206. Selleri, L., et al., Requirement for Pbx1 in skeletal patterning and programming chondrocyte proliferation and differentiation. Development, 2001. 128(18): p. 3543-57. Monteiro, M.C., et al., PBX1: a novel stage-specific regulator of adipocyte development. Stem Cells, 2011. 29(11): p. 1837-48. Gordon, J.A., et al., Pbx1 represses osteoblastogenesis by blocking Hoxa10-mediated recruitment of chromatin remodeling factors. Mol Cell Biol, 2010. 30(14): p. 3531-41. Liotti, A., et al., The transcription factor Prep1 impairs adipose tissue differentiation. Diabetologia, 2013. 56: p. S21-S21. Naiche, L.A., et al., T-box genes in vertebrate development. Annu Rev Genet, 2005. 39: p. 219-39. Gburcik, V., et al., An essential role for Tbx15 in the differentiation of brown and "brite" but not white adipocytes. Am J Physiol Endocrinol Metab, 2012. 303(8): p. E1053-60. Schleinitz, D., et al., Fat depot-specific mRNA expression of novel loci associated with waisthip ratio. Int J Obes (Lond), 2014. 38(1): p. 120-5. Gesta, S., et al., Mesodermal developmental gene Tbx15 impairs adipocyte differentiation and mitochondrial respiration. Proc Natl Acad Sci U S A, 2011. 108(7): p. 2771-6. Todorovich, S.M. and Z.A. Khan, Elevated T-box 2 in infantile hemangioma stem cells maintains an adipogenic differentiation-competent state. Dermatoendocrinol, 2013. 5(3): p. 352-7. Lee, H.S., et al., Tbx3, a transcriptional factor, involves in proliferation and osteogenic differentiation of human adipose stromal cells. Mol Cell Biochem, 2007. 296(1-2): p. 129-36. Lam, E.W., et al., Forkhead box proteins: tuning forks for transcriptional harmony. Nat Rev Cancer, 2013. 13(7): p. 482-95. Xu, L., et al., The winged helix transcription factor Foxa3 regulates adipocyte differentiation and depot-selective fat tissue expansion. Mol Cell Biol, 2013. 33(17): p. 3392-9. Omatsu, Y., et al., Foxc1 is a critical regulator of haematopoietic stem/progenitor cell niche formation. Nature, 2014. 508(7497): p. 536-40.

AC

115.

29

ACCEPTED MANUSCRIPT

145. 146. 147. 148. 149. 150. 151. 152. 153.

154. 155. 156. 157.

158.

159. 160.

T

IP

SC R

144.

NU

143.

MA

142.

D

141.

TE

140.

CE P

139.

Gerin, I., et al., On the role of FOX transcription factors in adipocyte differentiation and insulin-stimulated glucose uptake. J Biol Chem, 2009. 284(16): p. 10755-63. Wolfrum, C., et al., Role of Foxa-2 in adipocyte metabolism and differentiation. J Clin Invest, 2003. 112(3): p. 345-56. Nakae, J., et al., The forkhead transcription factor Foxo1 regulates adipocyte differentiation. Dev Cell, 2003. 4(1): p. 119-29. Davis, K.E., M. Moldes, and S.R. Farmer, The forkhead transcription factor FoxC2 inhibits white adipocyte differentiation. J Biol Chem, 2004. 279(41): p. 42453-61. Cederberg, A., et al., FOXC2 is a winged helix gene that counteracts obesity, hypertriglyceridemia, and diet-induced insulin resistance. Cell, 2001. 106(5): p. 563-73. Ridderstrale, M., et al., FOXC2 mRNA Expression and a 5' untranslated region polymorphism of the gene are associated with insulin resistance. Diabetes, 2002. 51(12): p. 3554-60. Cavodeassi, F., J. Modolell, and J.L. Gomez-Skarmeta, The Iroquois family of genes: from body building to neural patterning. Development, 2001. 128(15): p. 2847-55. Frayling, T.M., et al., A common variant in the FTO gene is associated with body mass index and predisposes to childhood and adult obesity. Science, 2007. 316(5826): p. 889-94. Ragvin, A., et al., Long-range gene regulation links genomic type 2 diabetes and obesity risk regions to HHEX, SOX4, and IRX3. Proc Natl Acad Sci U S A, 2010. 107(2): p. 775-80. Smemo, S., et al., Obesity-associated variants within FTO form long-range functional connections with IRX3. Nature, 2014. 507(7492): p. 371-5. Clevers, H. and R. Nusse, Wnt/beta-catenin signaling and disease. Cell, 2012. 149(6): p. 1192-205. Christodoulides, C., et al., Adipogenesis and WNT signalling. Trends Endocrinol Metab, 2009. 20(1): p. 16-24. Grant, S.F., et al., Variant of transcription factor 7-like 2 (TCF7L2) gene confers risk of type 2 diabetes. Nat Genet, 2006. 38(3): p. 320-3. Helgason, A., et al., Refining the impact of TCF7L2 gene variants on type 2 diabetes and adaptive evolution. Nat Genet, 2007. 39(2): p. 218-25. Kaminska, D., et al., Adipose tissue TCF7L2 splicing is regulated by weight loss and associates with glucose and fatty acid metabolism. Diabetes, 2012. 61(11): p. 2807-13. Thisse, B., M. el Messal, and F. Perrin-Schmitt, The twist gene: isolation of a Drosophila zygotic gene necessary for the establishment of dorsoventral pattern. Nucleic Acids Res, 1987. 15(8): p. 3439-53. Pettersson, A.T., et al., Twist1 in human white adipose tissue and obesity. J Clin Endocrinol Metab, 2011. 96(1): p. 133-41. Pettersson, A.T., et al., A possible inflammatory role of twist1 in human white adipocytes. Diabetes, 2010. 59(3): p. 564-71. Pan, D., et al., Twist-1 is a PPARdelta-inducible, negative-feedback regulator of PGC-1alpha in brown fat metabolism. Cell, 2009. 137(1): p. 73-86. Candia, A.F., et al., Mox-1 and Mox-2 define a novel homeobox gene subfamily and are differentially expressed during early mesodermal patterning in mouse embryos. Development, 1992. 116(4): p. 1123-36. Mankoo, B.S., et al., The concerted action of Meox homeobox genes is required upstream of genetic pathways essential for the formation, patterning and differentiation of somites. Development, 2003. 130(19): p. 4655-64. Liu, P., et al., Involvement of IGF-1 and MEOX2 in PI3K/Akt1/2 and ERK1/2 pathways mediated proliferation and differentiation of perivascular adipocytes. Exp Cell Res, 2014. Chen, Y. and D.H. Gorski, Regulation of angiogenesis through a microRNA (miR-130a) that down-regulates antiangiogenic homeobox genes GAX and HOXA5. Blood, 2008. 111(3): p. 1217-26.

AC

138.

30

ACCEPTED MANUSCRIPT

168. 169. 170. 171.

172. 173. 174. 175. 176. 177. 178. 179. 180.

181.

T

IP

SC R

167.

NU

166.

MA

165.

D

164.

TE

163.

CE P

162.

Irelan, J.T., et al., A functional screen for regulators of CKDN2A reveals MEOX2 as a transcriptional activator of INK4a. PLoS One, 2009. 4(4): p. e5067. Jespersen, N.Z., et al., A classical brown adipose tissue mRNA signature partly overlaps with brite in the supraclavicular region of adult humans. Cell Metab, 2013. 17(5): p. 798-805. Rajakumari, S., et al., EBF2 determines and maintains brown adipocyte identity. Cell Metab, 2013. 17(4): p. 562-74. Wang, W., et al., Ebf2 is a selective marker of brown and beige adipogenic precursor cells. Proc Natl Acad Sci U S A, 2014. 111(40): p. 14466-71. Daftary, G.S. and H.S. Taylor, Endocrine regulation of HOX genes. Endocr Rev, 2006. 27(4): p. 331-55. Pinnick, K.E. and F. Karpe, DNA methylation of genes in adipose tissue. Proc Nutr Soc, 2011. 70(1): p. 57-63. Hilton, C., M.J. Neville, and F. Karpe, MicroRNAs in adipose tissue: their role in adipogenesis and obesity. Int J Obes, 2012. Chen, Y., et al., miR-155 regulates differentiation of brown and beige adipocytes via a bistable circuit. Nat Commun, 2013. 4: p. 1769. Sun, L. and M. Trajkovski, MiR-27 orchestrates the transcriptional regulation of brown adipogenesis. Metabolism, 2014. 63(2): p. 272-82. Yekta, S., C.J. Tabin, and D.P. Bartel, MicroRNAs in the Hox network: an apparent link to posterior prevalence. Nat Rev Genet, 2008. 9(10): p. 789-96. Rantalainen, M., et al., MicroRNA expression in abdominal and gluteal adipose tissue is associated with mRNA expression levels and partly genetically driven. PLoS One, 2011. 6(11): p. e27338. Kloting, N., et al., MicroRNA expression in human omental and subcutaneous adipose tissue. PLoS One, 2009. 4(3): p. e4699. Sun, L., et al., Long noncoding RNAs regulate adipogenesis. Proc Natl Acad Sci U S A, 2013. 110(9): p. 3387-92. Dasen, J.S., Long noncoding RNAs in development: solidifying the Lncs to Hox gene regulation. Cell Rep, 2013. 5(1): p. 1-2. Divoux, A., et al., Identification of a novel lncRNA in gluteal adipose tissue and evidence for its positive effect on preadipocyte differentiation. Obesity (Silver Spring), 2014. Rinn, J.L., et al., Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell, 2007. 129(7): p. 1311-23. Zhou, Y., J. Peng, and S. Jiang, Role of histone acetyltransferases and histone deacetylases in adipocyte differentiation and adipogenesis. Eur J Cell Biol, 2014. 93(4): p. 170-7. Qiang, L., et al., Brown remodeling of white adipose tissue by SirT1-dependent deacetylation of Ppargamma. Cell, 2012. 150(3): p. 620-32. Jing, E., S. Gesta, and C.R. Kahn, SIRT2 regulates adipocyte differentiation through FoxO1 acetylation/deacetylation. Cell Metab, 2007. 6(2): p. 105-14. Sakamoto, H., et al., Cell type-specific methylation profiles occurring disproportionately in CpG-less regions that delineate developmental similarity. Genes Cells, 2007. 12(10): p. 112332. Illingworth, R., et al., A novel CpG island set identifies tissue-specific methylation at developmental gene loci. PLoS Biol, 2008. 6(1): p. e22.

AC

161.

31

ACCEPTED MANUSCRIPT Highlights Brown, beige and regional white adipose tissue depots have different structural and functional properties.

T

Developmental transcription factors are strongly differently expressed between depots.

AC

CE P

TE

D

MA

NU

SC R

IP

There is evidence that developmental genes have a causal role in regional adipose tissue distribution and function.

32

ACCEPTED MANUSCRIPT Conflict of interest

AC

CE P

TE

D

MA

NU

SC R

IP

T

We currently have a review article in press which we have referred to in this article and we therefore attach a copy. None of the authors have any conflict of interest to declare. Funding to support our contribution to this research field has been received from the Medical Research Council, Novo Nordisk UK Research Foundation, Heart Research UK, EU FP6 MolPAGE (LSHG/512066) and EU FP7 LipidomicNet (#202272).

33

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

AC

CE P

Figure 1

34