Page 1 of 53 Diabetes

Id1 promotes obesity by suppressing brown adipose thermogenesis and white adipose browning

Mallikarjun Patil, Bal Krishan Sharma, Sawsan Elattar, Judith Chang, Shweta Kapil, Jinling Yuan & Ande Satyanarayana*

Department of Biochemistry and Molecular Biology, Molecular Oncology & Biomarkers Program, Georgia Cancer Center, Augusta University, RoomCN3150, 1410 Laney Walker Blvd., Augusta, GA 30912

*To whom correspondence should be addressed (Ande Satyanarayana, 1410 Laney Walker Blvd., Rm CN3150, Augusta, GA 30912, Phone: 7067230029; Fax: 7067210101; email: ([email protected])

Running title: Id1 suppresses BAT thermogenesis and WAT browning

Number of figures & tables: Figures 8; Suppl. Figures 4; Suppl. Tables 2

1

Diabetes Publish Ahead of Print, published online March 7, 2017 Diabetes Page 2 of 53

Abstract

Obesity results from increased energy intake or defects in energy expenditure. Brown adipose tissue

(BAT) is specialized for energy expenditure, a process called adaptive thermogenesis. Peroxisome proliferator activated receptor γ coactivator 1α (PGC1α) controls BATmediated thermogenesis by

regulating the expression of Ucp1. Inhibitor of differentiation 1 (Id1) is a helixloophelix transcription

factor that plays important roles in cell proliferation and differentiation. Here, we demonstrate a novel

function of Id1 in BAT thermogenesis and programming of beige adipocytes in white adipose tissue

(WAT). We found that adipose tissuespecific overexpression of Id1 causes ageassociated and highfat

dietinduced obesity in mice. Id1 suppresses BAT thermogenesis by binding to and suppressing PGC1α

transcriptional activity. In WAT, Id1 is mainly localized in the stromal vascular fraction (SVF) where

the adipose progenitor/precursors reside. Lack of Id1 increased beige gene and Ucp1 expression in the

WAT in response to cold exposure. Furthermore, brownlike differentiation is increased in Id1deficient

mouse embryonic fibroblasts (MEFs). At the molecular level, Id1 directly interacted with and

suppressed Ebf2 transcriptional activity leading to reduced expression of Prdm16, which determines beige/brown adipocyte cell fate. Overall, our study highlights the existence of novel regulatory

mechanisms between Id1/PGC1α and Id1/Ebf2 in controlling brown fat metabolism which has

significant implications in the treatment of obesity and its associated diseases such as diabetes.

2 Page 3 of 53 Diabetes

Introduction

Obesity is a significant risk factor for a vast number of diseases, such as cardiovascular disease, type 2

diabetes, hypertension, fatty liver disease, and numerous cancers (1; 2). Obesity results from increased

energy intake or defects in energy expenditure. The excessive energy is stored as triglycerides in the

white adipose tissue (WAT). A second type of adipose tissue known as brown adipose tissue (BAT) has

been well known for a long time in rodents and newborn humans (3). BAT is specialized for energy

expenditure, a process called adaptive thermogenesis, a physiological mechanism during which energy

is dissipated to generate heat in response to cold temperatures and possibly diet (3; 4). The very densely

packed mitochondria in BAT execute heat production through a unique protein called uncoupling

protein1 (UCP1). UCP1 uncouples mitochondrial oxidative phosphorylation from ATP production and

dissipates chemical energy as heat, thereby strongly increasing energy expenditure (5). Peroxisome

proliferator activated receptor γ coactivator 1α (PGC1α) promotes BATmediated thermogenesis by

directly regulating the expression of Ucp1. Due to their critical role in thermogenesis, PGC1α and UCP1

expression and activity are tightly controlled by several other factors that either positively or negatively

regulate PGC1α and UCP1. Some of the important factors that positively regulate PGC1α or UCP1 are

FOXC2, SRC1, CREB, IRF4, SIRT3 and p38 MAPK, whereas RIP140, LXRα, Cidea, pRB, SRC2,

Twist1 and TRPV4 negatively regulate PGC1α and/or UCP1 (4; 69).

In addition to WAT and BAT, a third type of adipocytes arises in the body in response to certain

physiological stimuli. For example, in response to cold and βAR or PPARγ agonists, pools of UCP1

expressing brownlike adipocytes were detected in mouse WAT (10; 11). These adipocytes are called

‘beige’ or ‘brite’ cells. Beige and brown adipocytes share a number of thermogenic genes such as

Pgc1a, Ucp1, Cidea, and Prdm16; however, beige cells also express several unique genes such as

CD40, CD137, Tmem26 and Tbx1 that apparently reflect their distinct developmental origin (12).

3 Diabetes Page 4 of 53

Lineage tracing studies revealed that the beige adipocytes are derived from the adipose progenitor/precursor cells (13). In these cells, early B cell factor2 (Ebf2) and PPARγ cooperatively

induce the expression of Prdm16, which determines the beige adipocyte cell fate (1416). PRDM16 can bind to and promote the transcriptional activity of numerous factors such as PPARα, PPARγ and

PGC1α, thus functioning as a critical activator of brown/beige adipocyte cell fate (15; 17). Induced

expression of Prdm16 or Ebf2 is sufficient to convert white adipose precursors into brownlike UCP1

expressing cells in vitro (14; 17). Similarly, forced expression of Ebf2 or Prdm16 stimulates beige cell formation in subcutaneous WAT, and transgenic mice display increased energy expenditure and reduced weight gain in response to highfat diet (18; 19).

Inhibitor of differentiation 1 (Id1) is a helixloophelix (HLH) transcription factor that lacks a DNA binding domain. Id1 directly binds to other transcription factors and suppresses their transcriptional activity, thereby playing an important role in a number of cellular processes such as cell proliferation, differentiation, hematopoiesis and cancer cell metabolic adaptation (2022). However, the adipose tissuespecific function of Id1 and its relative contribution to body energy expenditure remains unclear.

Moreover, although Id1 is expressed in both BAT and WAT, the specific involvement of Id1in BAT mediated thermogenesis or browning of WAT has not been fully established. To understand the adipose tissuespecific function of Id1, we have generated adipose tissuespecific Id1 transgenic mice (aP2

Id1Tg+). We discovered that adiposespecific overexpression of Id1 causes age and HFDassociated

obesity in male mice, whereas females are resistant to Id1induced obesity. At the molecular level, Id1

directly binds to the central regulators of BAT thermogenesis and WAT browning, PGC1α and Ebf2 (4;

16; 19), and suppresses their transcriptional activity. Our results highlight a novel function of Id1 in

BATthermogenesis and WAT browning, and thus, in body energy homeostasis.

4 Page 5 of 53 Diabetes

Research Design and Methods

Mice and diet

The aP2Id1Tg+ transgenic expression vector and mice were generated using services from Cyagen, CA,

USA. Out of 4 transgenic lines generated, one line showed strong expression of Id1 in WAT and BAT

and was utilized for this study. The following primers were used to distinguish the aP2Id1Tg+ transgenic

from control aP2Id1Tg mice: Tg F: ATCTTTAAAAGCGAGTTCCCT; Tg R:

CTCCGACAGACCAAGTACCA; Internal control F: ACTCCAAGGCCACTTATCACC; Internal

control R: ATTGTTACCAACTGGGACGACA. Endogenous mouse βactin served as internal control.

C57BL/6J background aP2Id1Tg+, aP2Id1Tg, Id1+/+ and Id1/ mice were used for this study. Mice

were housed in a barrier facility under standard conditions with a 12h lightdark cycle. Mice were

handled in compliance with the U.S. National Institutes of Health (NIH) guidelines for animal care and

use. All animal protocols were reviewed and approved by the Institutional Animal Care and Use

Committee (IACUC) of Augusta University. Mice were fed a standard chow normal diet (ND)

containing 6% crude fat (Harlan Teklad Rodent Diet 2918). For the highfatdiet (HFD) experiments, 1

monthold mice were switched from ND to HFD (60 kcal% fat diet; D12492; Research Diets, New

Brunswick) and fed for 12 weeks. For cold exposure studies, mice were housed in standard cages

without bedding, and the cages were placed in the cold room (4°C) for 412 h. Mice were then

euthanized and tissues harvested.

Quantitative realtime PCR (qPCR)

Total RNA from cells and tissues were prepared using TRIzol Reagent (Lifetechnologies # 15596026)

according to the manufacturer’s instructions. Total RNA (2 µg) from each sample was reverse

transcribed into cDNA using RevertAid RT Kit (Thermo Scientific # K1691) according to the

manufacturer’s instructions. QuantitativePCR (qPCR) analysis was performed using Power SYBR

5 Diabetes Page 6 of 53

Green PCR Master Mix (Life Technologies # 4367659) according to the manufacturer’s instructions, in a 20 l final reaction volume, in 96well plates (Applied Biosciences # 4346907). The qPCR primers used are provided in the Suppl. Table 1.

Statistical analysis

The quantitative data for the experiments were presented as means ±SD. Statistical analyses were performed by the unpaired Student’s t test, and a p value of <0.05 was considered statistically

significant. Oxymax metabolic data were analyzed by oneway repeated measures ANOVA and a p value of <0.01 was considered statistically significant.

Detailed methods were provided in the online supplement additional information

Results

Id1 is expressed in adipose tissues, and its expression is strongly induced during brown adipocyte differentiation

To investigate if Id1 plays any role in adipose tissue metabolism, we analyzed the expression pattern of

Id1protein in mouse BAT, inguinal (iWAT), epididymal (eWAT) and retroperitoneal (rWAT) WAT. Id1 is expressed in all the adipose tissues, and its expression is relatively stronger in BAT compared to iWAT, eWAT and rWAT (Fig 1AB). In addition, analysis of Id1 mRNA revealed a relatively higher

Id1 mRNA in BAT and eWAT compared to iWAT and rWAT (Fig 1C). To investigate if Id1 is required for brown adipocyte differentiation, we induced differentiation in the HIB1B brown preadipocyte cell line and detected a 3fold induction of Id1 mRNA in day 4 differentiated cells compared to undifferentiated cells (Fig 1DE). The adipocyte differentiation marker aP2 and the brown adipocyte specific thermogenic genes Pgc1α and Ucp1 were also induced in day 4 differentiated cells as expected

(Fig 1E). Further analysis at the protein level revealed low levels of Id1 in undifferentiated HIB1B cells,

6 Page 7 of 53 Diabetes

and its expression is steadily elevated during the differentiation process (Fig 1F). To further confirm that

Id1 expression is induced in differentiated cells compared to nondifferentiated HIB1B cells, we

performed Id1/aP2 costaining, which revealed increased Id1 staining in aP2positive cells (Fig 1G).

These observations suggest that Id1 might play a role in brown adipocyte differentiation or in brown

adipocyteassociated thermogenesis.

Id1 is not required for brown adipocyte differentiation

Strong induction of Id1 during brown adipocyte differentiation prompted us to ask whether Id1 promotes

brown adipocyte differentiation. Id1 expression is strongly induced in differentiated HIB1B cells,

whereas its expression is low in undifferentiated cells (Fig 1F), therefore, we asked if forced expression

of Id1 in undifferentiated HIB1B cells accelerates differentiation. To this end, we generated an Id1

overexpressed (Id1OE) HIB1B cell line (Suppl. Fig 1A). Analysis of the expression levels of various

mitochondrial and thermogenesis genes in nondifferentiated control and Id1OE HIB1B cells revealed

downregulation of Pgc1α, Ucp1, Cox4, Cpt1b, Dio2 and ERRα in Id1OE compared to control HIB1B

cells (Fig 1H). In addition, mitochondrial content is also reduced in Id1OE cells compared to control

cells (Fig 1IJ). Consequently, Id1OE cells displayed reduced respiration (O2 consumption) compared

to control nondifferentiated HIB1B cells (Fig 1K). These results indicate that Id1 might suppress

mitochondrial and thermogenesis genes, and thus, Id1 could play a role in brown adipocyte

differentiation. However, induction of brown adipogenesis in the control and Id1OE HIB1B cells

revealed no significant difference as revealed by OilRedO staining, and the expression levels of

adipocyte differentiation marker aP2, and brown adipocyte markers PGC1α and Ucp1 in fully

differentiated adipocytes (Suppl. Fig 1BC). To investigate if loss of Id1 has any effect on brown

adipocyte differentiation, we knocked down Id1 in HIB1B cells by shRNA (Suppl. Fig 1D) and induced

brown adipocyte differentiation. We did not observe any significant difference in the differentiation

between control and Id1shRNA HIB1B cells as revealed by OilRedO staining, and the expression

7 Diabetes Page 8 of 53

levels of aP2, PGC1α and Ucp1 in fully differentiated adipocytes (Suppl. Fig 1EF). These results together suggest that Id1 is not required for brown adipocyte differentiation and Id1 might have a different function in brown adipocytes.

Reduced energy expenditure and increased adiposity in aP2-Id1Tg+ mice

Previously, aP2promoter driven transgenic mice have been extensively utilized to study BAT thermogenesis and WAT browning (15; 19; 23). Therefore, to investigate the specific function of Id1 in

BATmediated thermogenesis and adipose tissue metabolism in a more physiologically relevant in vivo system, we generated aP2Id1Tg+ transgenic mice where Id1 is specifically overexpressed in adipose

tissues (Fig 2AC). We found that Id1 expression is 34 fold higher in the adipose tissues of aP2Id1Tg+ compared to aP2Id1Tg control mice (Fig 2C). To investigate if adiposespecific overexpression of Id1 causes metabolic alterations leading to changes in body adiposity, we measured body weight and fat mass in aP2Id1Tg+ and control mice that were fed a normal chow diet (ND). aP2Id1Tg+ males steadily

gained more body weight, and the body weight and fat mass were significantly increased in 6monthold

adult aP2Id1Tg+ males compared to aP2Id1Tg males (Fig 2DE), whereas lean mass, bone mass and interscapular BAT weight are unchanged between the two genotypes (Suppl. Fig 2AE). In contrast, females did not show any difference in the body weight and fat mass between the two genotypes in 2 monthold young and 6monthold adult mice (Fig 2FG).These observations suggest that females are resistant to Id1induced weight gain, whereas aP2Id1Tg+ male mice accumulate more fat mass in the course of time. To investigate if increased fat mass and weight gain in aP2Id1Tg+ mice are due to

changes in metabolic activity, we measured different metabolic parameters in 2 and 6 month old mice by indirect opencircuit calorimeter (Oxymax/CLAMS), and detected reduced O2 consumption (VO2), CO2

Tg+ Tg release (VCO2) and heat production (thermogenesis) in aP2Id1 compared to aP2Id1 mice (Fig

2H, Suppl. Fig 3A), indicating that body energy expenditure is reduced in aP2Id1Tg+ male mice.

Physical activity and food consumption were unchanged between the two genotypes (Fig 2IJ, Suppl.

8 Page 9 of 53 Diabetes

Fig 3BC), further confirming that reduced energy expenditure is mainly responsible for increased body

weight and fat mass gain in aP2Id1Tg+ male mice. Consequently, we observed increased lipid

accumulation in the BAT and white adipocyte hypertrophy in 6monthold aP2Id1Tg+ compared to aP2

Id1Tg mice (Fig 2KM). These observations suggest that adipose tissuespecific overexpression of Id1

suppresses body energy expenditure and promotes weight gain in male mice.

aP2-Id1Tg+ male mice are prone to HFDinduced obesity

On a normal chow diet (ND, 6 kcal %), aP2Id1Tg+ male mice gained more body weight in the course of

time and displayed reduced energy expenditure (Fig 2DE, H). Therefore, we asked if aP2Id1Tg+ mice

are prone to HFDinduced obesity. In response to 12 weeks of HFD (60 kcal %), aP2Id1Tg+ mice

gained more body weight compared to aP2Id1Tg mice (Fig 3AB). aP2Id1Tg+ mice displayed an

increased amount of visceral fat (Fig 3C), larger eWAT pads (Fig 3DE) and increased accumulation of

body fat (Fig 3F), and higher blood glucose, insulin and leptin levels (Suppl. Table 2) compared to aP2

Id1Tg mice. Consequently, aP2Id1Tg+ male mice displayed increased accumulation of lipids in and

around the interscapular BAT (Fig 3G) and increased liver size and weight compared to aP2Id1Tg mice

(Fig 3HI). Histological analysis of WAT from aP2Id1Tg+ mice showed white adipocyte hypertrophy

(Fig 3JL) and increased lipid accumulation in the BAT and liver compared to aP2Id1Tg mice (Fig 3M

N). To investigate if increased weight gain in aP2Id1Tg+ mice in response to HFD is due to changes in

metabolic activity, we measured different metabolic parameters. Food consumption and physical

activity were unchanged between the two genotypes (Fig 3OP). However, we detected reduced O2

Tg+ consumption (VO2), CO2 release (VCO2) and heat production (thermogenesis) in aP2Id1 compared

to aP2Id1Tg mice (Fig 3QS). Moreover, thermogenic and mitochondrial gene expression was reduced

in the BAT of HFDfed aP2Id1Tg+ compared to aP2Id1Tg mice (Suppl. Fig 3DE). In contrast to

males, aP2Id1Tg+ female mice did not gain any significant body weight compared to aP2Id1Tg controls

9 Diabetes Page 10 of 53

in response to HFD (not shown). These results together indicate that aP2Id1Tg+ male mice are prone to

HFDinduced obesity due to reduced energy expenditure.

Id1 suppresses BAT thermogenesis by inhibiting PGC1α transcriptional activity

To investigate how Id1 is expressed in vivo in response to thermogenic stimuli such as cold exposure or

HFD, we exposed wildtype mice to 4°C for 4 h which led to strong induction of Id1 in BAT (Fig 4A).

Similarly, 1 week of HFD feeding led to strong induction of Id1 in the BAT of wildtype mice (Fig 4B), indicating that Id1 could be involved in BAT thermogenesis. Therefore, we asked if reduced energy expenditure in aP2Id1Tg+ mice is due to downregulation of thermogenesis genes in BAT. At room temperature (RT; 23°C), we detected some reduction in the expression of Ucp1 in the BAT of aP2

Id1Tg+ mice but it has not reached statistical significance, whereas Elovl3 is significantly downregulated compared to aP2Id1Tg BAT (Fig 4CD). However, further expression analysis at the protein level

revealed reduced levels of thermogenic proteins Ucp1, Dio2 and Cytc in the aP2Id1Tg+ compared to aP2Id1Tg BAT (Fig 4E). Moreover, when mice were exposed to 4°C for 4 h, induction of thermogenic

and mitochondrial genes was reduced in aP2Id1Tg+ compared to aP2Id1Tg BAT (Fig 4FG).

Consistent with reduced expression of thermogenic proteins, interscapular BAT harvested from aP2

Id1Tg+ mice displayed reduced basal and uncoupled respiration compared to aP2Id1Tg BAT (Fig 4H).

Since increased expression of Id1 suppressed the expression of thermogenic genes and reduced O2

consumption, we asked if loss of Id1 results in increased expression of thermogenic genes. To this end,

we utilized Id1+/+ and Id1/ mice and detected a slight upregulation of Ucp1 and a strong induction of

Elovl3 at RT in Id1/ compared to Id1+/+ BAT (Fig 4IJ). However, at protein level, Ucp1 and Cytc are

increased in Id1/ compared to Id1+/+ BAT (Fig 4K). In addition, exposure of mice to 4°C for 4 h led to

a significant upregulation of thermogenic genes such as Ucp1, PPARγ, Cidea, and Dio2 in Id1/

compared to Id1+/+ BAT (Fig 4LM). Moreover, interscapular BAT isolated from Id1/ mice displayed

increased basal and uncoupled respiration compared to Id1+/+ BAT (Fig 4N). These results together

10 Page 11 of 53 Diabetes

indicate that Id1 suppresses thermogenic gene expression and lack of Id1 leads to increased expression

of thermogenic genes.

A number of factors activate or inhibit thermogenesis by either positively or negatively regulating

PGC1α/UCP1 thermogenesis pathway (4; 68; 24). Since Id1 lacks a DNAbinding domain and

regulates other transcription factors by direct interaction (20; 25), we reasoned that Id1 directly binds to

a transcriptional activator of thermogenesis and suppresses its transcriptional activity. Alternatively, Id1

directly binds to a negative regulator of thermogenesis and promotes its inhibitory action, leading to

reduced energy expenditure. To investigate these possibilities, we coexpressed Id1 and some of the

known positive and negative regulators of thermogenesis in HEK293 cells and performed a series of co

IPs. These analyses revealed a direct interaction between Id1 and PGC1α (Fig 5A), whereas Id1 did not

interact with the other factors that we have tested. Further analysis revealed a direct interaction between

Id1 and PGC1α in day 6 differentiated HIB1B cells indicating that Id1 indeed interacts with PGC1α

endogenously (Fig 5B). In response to cold exposure Id1 expression is induced (Fig 4A), and

thermogenic gene expression is suppressed in aP2Id1Tg+ and increased in Id1/ mice BAT compared to

controls (Fig 4F, L), Therefore, we asked if thermogenic stimulus promotes Id1 interaction with PGC1α

to prevent excessive thermogenesis. To test this possibility, we exposed wildtype mice to 4°C for 4h

and detected increased interaction between Id1 and PGC1α in the BAT of mice exposed to 4°C

compared to mice at RT (Fig 5C). To determine if the Id1/PGC1α interaction affects PGC1α

transcriptional activity, we performed a luciferase reporter assay in Cos7 cells utilizing a luciferase

reporter driven by the GalUAS sequence and fulllength PGC1α fused with the DNAbinding domain

(DBD) of yeast Gal4 as described previously (26). We found that the activity of Gal4PGC1α was

strongly inhibited when Id1 was coexpressed (Fig 5D). Moreover, coexpression of Ucp1 promoter

driven luciferase reporter vector along with PPARγ/RXRα/PGC1α or PPARγ/RXRα/PGC1α/Id1

expression vectors revealed significant suppression of Ucp1 promoterdriven luciferase activity when

11 Diabetes Page 12 of 53

Id1 is coexpressed (Fig 5E), suggesting that binding of Id1 indeed suppresses PGC1α transcriptional activity. These results together suggest that Id1 suppresses BAT thermogenesis by binding to and inhibiting the transcriptional activity of PGC1α.

Id1deficiency results in increased beige/thermogenic gene expression in iWAT

In addition to BAT, Id1 is also expressed in WAT (Fig 1A). WAT consists of adipocytes and the stromal vascular fraction (SVF) where the preadipocytes/progenitors reside (27). To investigate if Id1 is expressed in the white adipocytes or in the SVF, we performed Id1 and CD45 [a general marker of SVF,

(28)] costaining on iWAT and detected a strong colocalization of Id1 and CD45 (Fig 6A), indicating that Id1 is mainly localized in the SVF of WAT, and is thus possibly involved in adipose progenitor cell differentiation. To investigate if Id1 enhances or inhibits coldinduced brownlike (beige) differentiation of white preadipocytes in iWAT, we exposed Id1+/+ and Id1/ mice to 4°C for 12 h and analyzed the expression levels of the beige markers Tbx1, Tmem26, CD40 and CD137 (12) and thermogenic genes in iWAT. We detected significant upregulation of Tmem26, PPARγ, Ucp1 and Prdm16 in the iWAT of

Id1/ compared to Id1+/+ mice (Fig 6BC). Moreover, UCP1 staining on the iWAT of Id1+/+ and Id1/ mice after 12 h of cold exposure revealed strong UCP1 staining on Id1/ iWAT and is not detectable on

Id1+/+ iWAT (Fig 6D). These observations suggest that loss of Id1 enhances browning of iWAT.

Differentiation of MEFs into brownlike cells is increased in the absence of Id1

To further investigate whether loss of Id1 increases differentiation of white preadipocytes into brown like cells, we harvested the SVF from Id1+/+ and Id1/ WAT and attempted to culture white adipose progenitors as described previously (16; 29). However, Id1/ progenitor cells failed to grow and showed a senescencelike phenotype (data not shown). This is consistent with previous observations that Id1 promotes cell proliferation, and lack of Id1 results in premature senescence in certain cell types (30).

Therefore, we were unable to perform in vitro differentiation experiments in Id1/ progenitor cells. To

12 Page 13 of 53 Diabetes

overcome this obstacle, we used an alternative in vitro system; early passage (P2) MEFs. MEFs are un

programmed cells, share several characteristics with mesenchymal stem cells, and can differentiate into

various mesenchymal lineages (31). MEFs can be induced to differentiate into brownlike cells that

express mitochondrial and thermogenesis genes (24; 32). To test whether Id1 is involved in the

differentiation of MEFs into brownlike cells, we treated Id1+/+ and Id1/ E13.5 MEFs with brown

adipocyte differentiation media that consists of rosiglitazone, as described previously (24). We detected

increased adipocyte differentiation in Id1/ compared to Id1+/+ cultures as revealed by enhanced

expression of the adipocyte differentiation marker aP2 and OilRedO staining (Fig 7AB). Expression

of Id1 is induced during differentiation of MEFs into brown adipocytes in Id1+/+ cultures, and is

completely absence in Id1/ cells as expected (Fig 7A). Consistent with increased brown adipogenesis,

Id1/ differentiated cells displayed increased mitochondrial content and increased cellular respiration

compared to Id1+/+ cells (Fig 7CE). At the molecular level, Id1/ cells displayed increased expression

of beige and mitochondrial genes in day 10 differentiated cells, whereas their expression is similar in

undifferentiated MEFs (Fig 7FI). We also detected increased expression of thermogenic genes such as

Pparγ, Pgc1α, Ucp1, Ebf2 and Prdm16 in differentiated Id1/ cells compared to Id1+/+ cells (Fig 7JK).

Interestingly, the expression of Ebf2 and Prdm16, the major determinants of brown adipocyte identity

were increased even in the nondifferentiated Id1/ MEFs (Fig 7J). Further analysis at the protein level

revealed elevated levels of aP2, PPARγ, PGC1α, Ebf2 and PRDM16 in Id1/ differentiating cells

compared to Id1+/+ cells (Fig 7L). We also analyzed RB and RIP140 which are known to suppress WAT

browning (33). Although RB and pRB levels are unchanged, we found a detectable reduction in RIP140

levels in Id1/ cells (Fig 7L). These results suggest that lack of Id1 promotes differentiation of MEFs

into brownlike cells due to induced expression of PPARγ, Ebf2, PRDM16 and PGC1α.

Id1 directly binds to and suppresses Ebf2 transcriptional activity

13 Diabetes Page 14 of 53

Previously, it was demonstrated that by recruiting PPARγ to the Prdm16 promoter, Ebf2 induces the expression of Prdm16, which determines brown adipocyte cell fate (1416; 34). Ebf2 is a helixloop

helix transcription factor, and Id proteins are known to interact with helixloophelix proteins with very

high affinity (20; 25), indicating the possibility of a direct interaction between Id1 and Ebf2. Therefore,

we asked if Id1 downregulates Prdm16 expression by directly interacting with and suppressing Ebf2

transcriptional activity. To this end, we first asked whether Id1 and Ebf2 colocalize in WAT. We

stained for Id1 and Ebf2 on iWAT and detected colocalization of Id1 and Ebf2 (Fig 8AC), indicating

the possibility of a direct interaction between Id1 and Ebf2. Next, we coexpressed Id1 and Ebf2 in

HEK293 cells and performed coIP which revealed a direct interaction between Id1 and Ebf2 (Fig 8D).

To determine if Id1 binding to Ebf2 affects the transcriptional activity of Ebf2, we coexpressed a

Prdm16 promoterdriven luciferase reporter vector along with PPARγ/RXRα/Ebf2 or

PPARγ/RXRα/Ebf2/Id1 expression vectors. We detected a significant suppression of Prdm16 promoter

driven luciferase activity when Id1 is coexpressed with PPARγ/RXRα/Ebf2 (Fig 8E), whereas in the

absence of Ebf2, Id1 did not suppress Prdm16 promoterdriven luciferase activity (Fig 8F), suggesting

that Id1 downregulates Prdm16 expression by binding to and suppressing the transcriptional activity of

Ebf2. Since Id1 does not have a DNA binding domain and directly binds to Ebf2 and PGC1α, we

wondered if Id1 colocalizes with Ebf2 and PGC1α on their regulatory regions such as Prdm16 and

Ucp1 promoters. ChIPqPCR assays in day 4 differentiated HIB1B cells revealed that Id1 is localized to

the Prdm16 promoter region but not to the Ucp1 promoter region (Fig 8GH). These observations

indicate that Id1 can associate with Ebf2 at the Ebf2regulatory regions, whereas Id1/PGC1α interaction precludes PGC1α from associating with its regulatory regions. As a result, Id1 is not associated with the

Ucp1 promoter.

14 Page 15 of 53 Diabetes

Discussion

Id1 has been known to play critical roles in cell proliferation, cellular differentiation and tumorigenesis

(35; 36). Previously, we and others have demonstrated that Id1 whole body knockout mice (Id1/) are

lean, and are protected from highfatdiet (HFD)induced insulin resistance and hepatosteatosis (37; 38).

However, the adipose tissuespecific function of Id1 and its relative contribution to body energy

expenditure remains unclear since it is a whole body Id1deficient mouse. In the current work, we

discovered a novel function of Id1 in BATmediated thermogenesis and its potential involvement in the

programing of preadipocytes into brownlike adipocytes. During HIB1B brown adipocyte

differentiation, Id1 expression pattern closely mimicked the expression pattern of PPARγ, PGC1α and

UCP1. However, unlike PPARγ, which is absolutely required for both white and brown adipocyte

differentiation (39), Id1 is dispensable for brown adipocyte differentiation. Previously, it was

demonstrated that Id1 is also not required for white adipocyte differentiation (37), indicating that Id1 is

dispensable for both white and brown adipocyte differentiation. aP2Id1Tg+ mice displayed reduced

Tg energy expenditure, indicating that Id1 suppresses BAT thermogenesis. In aP2Id1 BAT, thermogenic

gene expression is mildly reduced compared to controls at RT. However, in the course time, this can

have a significant effect on body weight as it slowly but steadily reduces body energy expenditure. This

Tg could be the reason that the aP2Id1 mice gained weight slowly during aging. Id1 lacks a DNA binding

domain, therefore, the only way Id1 could inhibit thermogenesis is by directly binding to and

suppressing the transcriptional activity of a factor that promotes thermogenesis. Our attempt to identify a

possible interaction between Id1 and some of the known activators/suppressors of thermogenesis very

surprisingly revealed a direct interaction between Id1 and PGC1α with a concomitant suppression of

PGC1α transcriptional activity. If Id1 suppresses PGC1α transcriptional activity, then, why PGC1α

downstream target UCP1 is only slightly reduced when Id1 levels are strongly increased during HIB1B

differentiation? A possible explanation could be that PPARγ, PGC1α and UCP1 are highly expressed in

15 Diabetes Page 16 of 53

fully differentiated HIB1B cells and in the BAT where majority of the cells are mature fully differentiated brown adipocytes. The transcription factors Twist 1 and Cidea which inhibit PGC1α and

UCP1 are also highly expressed in BAT (26; 40). Therefore, it appears that both the activators and suppressors of thermogenesis coexist in the BAT and in fully differentiated brown adipocytes. When thermogenesis is strongly activated it triggers a negative feedback loop which further induces the expression of the thermogenic suppressors such as Id1 which in turn prevents excessive thermogenesis by inhibiting PGC1α and/or UCP1. Consistent with this explanation, we found that cold exposure

enhanced the expression of Id1 which in turn resulted in increased interaction between Id1 and PGC1α.

Conversely, cold exposure results in increased expression of UCP1 in the BAT of Id1/ mice (37).

In the adipose progenitor cells, Ebf2 and PPARγ cooperatively induce the expression of Prdm16 which

controls beige adipocyte determination. We found that both Id1 and Ebf2 are mainly localized in the

SVF of iWAT where the adipose progenitors reside. Furthermore, Id1 directly interacted with Ebf2 and

suppressed its transcriptional activity, leading to impaired expression of Prdm16. Consequently, in Id1/ iWAT, the expression of beige and thermogenic genes, including Prdm16, is increased in response to prolonged cold exposure. In addition, the differentiation of MEFs into brownlike cells is also

significantly increased in the absence of Id1. This is in contrast to our observation that Id1 is dispensable

for HIB1B brown adipocyte differentiation. A possible reason that the knockdown or overexpression of

Id1 did not impact differentiation of HIB1B cells could be due to the fact that HIB1B cells are a preprogrammed brown adipocyte cell line and upon brown adipogenic stimulation they certainly

differentiate into brown adipocytes. Therefore, it is possible that the HIB1B cell differentiation is no longer influenced by the factors that are involved in the earlier cell fate determination. It appears that Id1 participates in the earlier programming of preadipocytes into beige adipocytes by regulating Ebf2

transcriptional activity. Ebf2 not only determines beige but also brown adipocyte cell fate, and loss of

Ebf2 results in significant reduction of BAT and an almost total loss of brown fatspecific gene

16 Page 17 of 53 Diabetes

expression in Ebf2/ embryos (16). If Id1 functions as an upstream inhibitor of Ebf2, then why does

elevated expression of Id1 (aP2Id1Tg+) or loss of Id1 (Id1/) have no detectable effect on BAT

development? A possible explanation could be that in aP2Id1Tg+ mice, Id1 expression is directed by the

aP2 promoter/enhancer which may not be active in the adipose progenitors during embryonic

development. Therefore, Id1 may not have been induced during brown adipocyte cell fate determination

in aP2Id1Tg+ mice, and therefore did not interfere with Ebf2/Prdm16mediated brown adipocyte

programming and BAT development. In Id1/ mice, obviously, Ebf2 is free from Id1mediated

suppression and induces Prdm16 expression leading to normal BAT development. In contrast to brown

adipocytes, beige adipocytes are generated in response to a substantial physiological stimulus, such as

prolonged cold exposure or β3ARadrenergic agonists (12; 41; 42). Interestingly, Id1 expression is also

strongly induced in response to cold exposure. Once induced, Id1 functions as an upstream

transcriptional suppressor of Ebf2, and in the absence of Id1, Ebf2mediated browning of iWAT is

enhanced. Together, our results suggest that Id1 regulates two pathways PGC1α/Ucp1 thermogenic and

Ebf2/Prdm16 adipose progenitor cell programing pathways by inhibiting PGC1α in the BAT and Ebf2

in the progenitor cells. Nevertheless, we cannot exclude the possibility that Id1 also interacts with

PGC1α in the progenitor cells. However, Id1/PGC1α interaction may not prevent the progenitors to

differentiate into brownlike cells since neither PGC1α nor Id1 are required for brown adipocyte

differentiation.

In contrast to males, female aP2Id1Tg+ mice are protected from age and HFDassociated obesity. Sex

dependent differences in energy intake, storage and expenditure are well known, and these physiological

mechanisms are also under the control of sex hormones (43). For example, ovarian estrogens control

energy homeostasis (44), and estradiol activates thermogenesis in BAT through the sympathetic nervous

system (45). In our study, Id1mediated suppression of energy expenditure caused obesity in males but

not in females, suggesting that female sex hormones can counteract the effects of some of the negative

17 Diabetes Page 18 of 53

regulators of energy expenditure. Therefore, future studies should be focused at exploring which of these factors are under the control of female sex hormones. In the current study, we utilized aP2Id1Tg mice and recent studies demonstrate that aP2 promoterdriven expression is not highly specific to adipose tissues and it can have certain low nonspecific expression in other tissues such as lung, heart, muscle, liver and testes (4648). In our aP2Id1Tg mouse model, we did not detect any difference in the lean mass and bone density between control and aP2Id1Tg+ mice. Moreover, female aP2Id1Tg+ mice did not

show any detectable phenotypes in adipose or other tissues, suggesting that aP2 promoterdriven Id1

expression did not cause any significant side effects. In conclusion, our study demonstrates that Id1 by suppressing PGC1αmediated BAT thermogenesis and Ebf2mediated beige adipocyte programing promotes energy storage and obesity, which is a significant risk factor for insulin resistance and diabetes. Therefore, Id1 could potentially function as a molecular target to reverse obesity. Id1/ mice are viable, fertile and live normally suggesting that most cell types do not require Id1 for their normal function. Therefore, targeting Id1 in vivo could be a relatively safe strategy to increase energy expenditure, reduce adiposity and treat obesity and its associated diseases such as diabetes.

Author contributions

M.P., B.K.S, S.E., J.C., and S.K performed experiments and collected and analyzed data. J.Y maintained and provided animals for experiments. A.S. supervised the study, and A.S., and M.P. analyzed and interpreted the data and wrote the manuscript.

Acknowledgments

The authors thank Dr. Jonathan Keller (Mouse Cancer Genetics Program, NCIFrederick) for providing

Id1/ mice, Dr. Bruce Spiegelman (Department of Cell Biology, Harvard Medical School) for providing

Gal4PGC1α vectors and HIB1B cells, Dr. Patrick Seale (Department of Cell and Developmental

18 Page 19 of 53 Diabetes

Biology, Perelman School of Medicine, University of Pennsylvania) for providing Prdm16 luciferase

reporter vectors and Dr. Mark Christian (Division of Metabolic and Vascular Health, The University of

Warwick) for providing Ucp1 luciferase vector. We also thank Dr. Ali Arbab, Roxan Ara and Chris

Middleton (Tumor angiogenesis section, Georgia Cancer Center, Augusta University) for technical help

with CT scans, Dr. Jenfeng Pang (Molecular Oncology Program, Georgia Cancer Center, Augusta

University) for help with Oxymax readings, and Dr. RheaBeth Markowitz (Georgia Cancer Center,

Augusta University) for reviewing and editing the manuscript. This research is supported by the

National Cancer Institute (NCI) K22CA168828 (AS), and National Institute of Diabetes and Digestive

and Kidney Diseases (NIDDK) DP2DK105565 (AS) of the National Institutes of Health. The authors

declare no conflicts of interest. M.P and A.S are the guarantors of this work and has full access to all the

data in the study and take responsibility for the integrity of the data and the accuracy of the data

analysis.

References

1. Vucenik I, Stains JP: Obesity and cancer risk: evidence, mechanisms, and recommendations. Annals of the New York Academy of Sciences 2012;1271:3743 2. Poirier P, Eckel RH: Obesity and cardiovascular disease. Current atherosclerosis reports 2002;4:448 453 3. Cannon B, Nedergaard J: Brown adipose tissue: function and physiological significance. Physiological reviews 2004;84:277359 4. Seale P, Kajimura S, Spiegelman BM: Transcriptional control of brown adipocyte development and physiological functionof mice and men. Genes & development 2009;23:788797 5. Kozak LP, AnunciadoKoza R: UCP1: its involvement and utility in obesity. International journal of obesity 2008;32 Suppl 7:S3238 6. Finck BN, Kelly DP: PGC1 coactivators: inducible regulators of energy metabolism in health and disease. The Journal of clinical investigation 2006;116:615622 7. Kong X, Banks A, Liu T, Kazak L, Rao RR, Cohen P, Wang X, Yu S, Lo JC, Tseng YH, Cypess AM, Xue R, Kleiner S, Kang S, Spiegelman BM, Rosen ED: IRF4 is a key thermogenic transcriptional partner of PGC1alpha. Cell 2014;158:6983 8. Picard F, Gehin M, Annicotte J, Rocchi S, Champy MF, O'Malley BW, Chambon P, Auwerx J: SRC 1 and TIF2 control energy balance between white and brown adipose tissues. Cell 2002;111:931941 9. Sharma BK, Patil M, Satyanarayana A: Negative regulators of brown adipose tissue (BAT)mediated thermogenesis. Journal of cellular physiology 2014;229:19011907

19 Diabetes Page 20 of 53

10. Vitali A, Murano I, Zingaretti MC, Frontini A, Ricquier D, Cinti S: The adipose organ of obesity prone C57BL/6J mice is composed of mixed white and brown adipocytes. Journal of lipid research 2012;53:619629 11. Barbatelli G, Murano I, Madsen L, Hao Q, Jimenez M, Kristiansen K, Giacobino JP, De Matteis R, Cinti S: The emergence of coldinduced brown adipocytes in mouse white fat depots is determined predominantly by white to brown adipocyte transdifferentiation. American journal of physiology Endocrinology and metabolism 2010;298:E12441253 12. Wu J, Bostrom P, Sparks LM, Ye L, Choi JH, Giang AH, Khandekar M, Virtanen KA, Nuutila P, Schaart G, Huang K, Tu H, van Marken Lichtenbelt WD, Hoeks J, Enerback S, Schrauwen P, Spiegelman BM: Beige adipocytes are a distinct type of thermogenic fat cell in mouse and human. Cell 2012;150:366376 13. Wang QA, Tao C, Gupta RK, Scherer PE: Tracking adipogenesis during white adipose tissue development, expansion and regeneration. Nature medicine 2013;19:13381344 14. Seale P, Bjork B, Yang W, Kajimura S, Chin S, Kuang S, Scime A, Devarakonda S, Conroe HM, ErdjumentBromage H, Tempst P, Rudnicki MA, Beier DR, Spiegelman BM: PRDM16 controls a brown fat/skeletal muscle switch. Nature 2008;454:961967 15. Seale P, Kajimura S, Yang W, Chin S, Rohas LM, Uldry M, Tavernier G, Langin D, Spiegelman BM: Transcriptional control of brown fat determination by PRDM16. Cell metabolism 2007;6:3854 16. Rajakumari S, Wu J, Ishibashi J, Lim HW, Giang AH, Won KJ, Reed RR, Seale P: EBF2 determines and maintains brown adipocyte identity. Cell metabolism 2013;17:562574 17. Kajimura S, Seale P, Kubota K, Lunsford E, Frangioni JV, Gygi SP, Spiegelman BM: Initiation of myoblast to brown fat switch by a PRDM16C/EBPbeta transcriptional complex. Nature 2009;460:11541158 18. Seale P, Conroe HM, Estall J, Kajimura S, Frontini A, Ishibashi J, Cohen P, Cinti S, Spiegelman BM: Prdm16 determines the thermogenic program of subcutaneous white adipose tissue in mice. The Journal of clinical investigation 2011;121:96105 19. Stine RR, Shapira SN, Lim HW, Ishibashi J, Harms M, Won KJ, Seale P: EBF2 promotes the recruitment of beige adipocytes in white adipose tissue. Molecular metabolism 2016;5:5765 20. Norton JD: ID helixloophelix proteins in cell growth, differentiation and tumorigenesis. Journal of cell science 2000;113 ( Pt 22):38973905 21. Perry SS, Zhao Y, Nie L, Cochrane SW, Huang Z, Sun XH: Id1, but not Id3, directs longterm repopulating hematopoietic stemcell maintenance. Blood 2007;110:23512360 22. Sharma BK, Kolhe R, Black SM, Keller JR, Mivechi NF, Satyanarayana A: Inhibitor of differentiation 1 transcription factor promotes metabolic reprogramming in hepatocellular carcinoma cells. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 2016;30:262275 23. Vernochet C, Mourier A, Bezy O, Macotela Y, Boucher J, Rardin MJ, An D, Lee KY, Ilkayeva OR, Zingaretti CM, Emanuelli B, Smyth G, Cinti S, Newgard CB, Gibson BW, Larsson NG, Kahn CR: Adiposespecific deletion of TFAM increases mitochondrial oxidation and protects mice against obesity and insulin resistance. Cell metabolism 2012;16:765776 24. Hansen JB, Jorgensen C, Petersen RK, Hallenborg P, De Matteis R, Boye HA, Petrovic N, Enerback S, Nedergaard J, Cinti S, te Riele H, Kristiansen K: Retinoblastoma protein functions as a molecular switch determining white versus brown adipocyte differentiation. Proceedings of the National Academy of Sciences of the United States of America 2004;101:41124117 25. O'Toole PJ, Inoue T, Emerson L, Morrison IE, Mackie AR, Cherry RJ, Norton JD: Id proteins negatively regulate basic helixloophelix transcription factor function by disrupting subnuclear compartmentalization. The Journal of biological chemistry 2003;278:4577045776 26. Pan D, Fujimoto M, Lopes A, Wang YX: Twist1 is a PPARdeltainducible, negativefeedback regulator of PGC1alpha in brown fat metabolism. Cell 2009;137:7386

20 Page 21 of 53 Diabetes

27. Rodeheffer MS, Birsoy K, Friedman JM: Identification of white adipocyte progenitor cells in vivo. Cell 2008;135:240249 28. Bourin P, Bunnell BA, Casteilla L, Dominici M, Katz AJ, March KL, Redl H, Rubin JP, Yoshimura K, Gimble JM: Stromal cells from the adipose tissuederived stromal vascular fraction and culture expanded adipose tissuederived stromal/stem cells: a joint statement of the International Federation for Adipose Therapeutics and Science (IFATS) and the International Society for Cellular Therapy (ISCT). Cytotherapy 2013;15:641648 29. Aune UL, Ruiz L, Kajimura S: Isolation and differentiation of stromal vascular cells to beige/brite cells. Journal of visualized experiments : JoVE 2013; 30. Alani RM, Young AZ, Shifflett CB: Id1 regulation of cellular senescence through transcriptional repression of p16/Ink4a. Proceedings of the National Academy of Sciences of the United States of America 2001;98:78127816 31. Yusuf B, Gopurappilly R, Dadheech N, Gupta S, Bhonde R, Pal R: Embryonic fibroblasts represent a connecting link between mesenchymal and embryonic stem cells. Development, growth & differentiation 2013;55:330340 32. Sellayah D, Bharaj P, Sikder D: Orexin is required for brown adipose tissue development, differentiation, and function. Cell metabolism 2011;14:478490 33. Scime A, Grenier G, Huh MS, Gillespie MA, Bevilacqua L, Harper ME, Rudnicki MA: Rb and p107 regulate preadipocyte differentiation into white versus brown fat through repression of PGC1alpha. Cell metabolism 2005;2:283295 34. Timmons JA, Wennmalm K, Larsson O, Walden TB, Lassmann T, Petrovic N, Hamilton DL, Gimeno RE, Wahlestedt C, Baar K, Nedergaard J, Cannon B: Myogenic gene expression signature establishes that brown and white adipocytes originate from distinct cell lineages. Proceedings of the National Academy of Sciences of the United States of America 2007;104:44014406 35. Zebedee Z, Hara E: Id proteins in cell cycle control and cellular senescence. Oncogene 2001;20:83178325 36. Sikder HA, Devlin MK, Dunlap S, Ryu B, Alani RM: Id proteins in cell growth and tumorigenesis. Cancer cell 2003;3:525530 37. Satyanarayana A, Klarmann KD, Gavrilova O, Keller JR: Ablation of the transcriptional regulator Id1 enhances energy expenditure, increases insulin sensitivity, and protects against age and diet induced insulin resistance, and hepatosteatosis. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 2012;26:309323 38. Zhao Y, Ling F, Griffin TM, He T, Towner R, Ruan H, Sun XH: Upregulation of the Sirtuin 1 (Sirt1) and peroxisome proliferatoractivated receptor gamma coactivator1alpha (PGC1alpha) genes in white adipose tissue of Id1 proteindeficient mice: implications in the protection against diet and age induced glucose intolerance. The Journal of biological chemistry 2014;289:2911229122 39. Rosen ED, Walkey CJ, Puigserver P, Spiegelman BM: Transcriptional regulation of adipogenesis. Genes & development 2000;14:12931307 40. Zhou Z, Yon Toh S, Chen Z, Guo K, Ng CP, Ponniah S, Lin SC, Hong W, Li P: Cideadeficient mice have lean phenotype and are resistant to obesity. Nature genetics 2003;35:4956 41. Ohno H, Shinoda K, Spiegelman BM, Kajimura S: PPARgamma agonists induce a whitetobrown fat conversion through stabilization of PRDM16 protein. Cell metabolism 2012;15:395404 42. Harms M, Seale P: Brown and beige fat: development, function and therapeutic potential. Nature medicine 2013;19:12521263 43. Shi H, Seeley RJ, Clegg DJ: Sexual differences in the control of energy homeostasis. Frontiers in neuroendocrinology 2009;30:396404 44. MauvaisJarvis F, Clegg DJ, Hevener AL: The role of estrogens in control of energy balance and glucose homeostasis. Endocrine reviews 2013;34:309338

21 Diabetes Page 22 of 53

45. Martinez de Morentin PB, GonzalezGarcia I, Martins L, Lage R, FernandezMallo D, Martinez Sanchez N, RuizPino F, Liu J, Morgan DA, Pinilla L, Gallego R, Saha AK, Kalsbeek A, Fliers E, Bisschop PH, Dieguez C, Nogueiras R, Rahmouni K, TenaSempere M, Lopez M: Estradiol regulates brown adipose tissue thermogenesis via hypothalamic AMPK. Cell metabolism 2014;20:4153 46. Jeffery E, Berry R, Church CD, Yu S, Shook BA, Horsley V, Rosen ED, Rodeheffer MS: Characterization of Cre recombinase models for the study of adipose tissue. Adipocyte 2014;3:206211 47. Mullican SE, Tomaru T, Gaddis CA, Peed LC, Sundaram A, Lazar MA: A novel adiposespecific gene deletion model demonstrates potential pitfalls of existing methods. Molecular endocrinology 2013;27:127134 48. Lee KY, Russell SJ, Ussar S, Boucher J, Vernochet C, Mori MA, Smyth G, Rourk M, Cederquist C, Rosen ED, Kahn BB, Kahn CR: Lessons on conditional gene targeting in mouse adipose tissue. Diabetes 2013;62:864874

Figure legends

Figure 1: Id1 expression is induced during brown adipocyte differentiation in HIB1B cells. (AB)

Expression levels of Id1 and βactin in different adipose tissues of 2monthold C57BL/6 mice (A).

After acquiring images, band intensities were measured and normalized to βactin by the LiCor

Odyssey system (B). (C) Id1 mRNA transcript levels in different adipose tissues of 2monthold

C57BL/6 mice. Compared to BAT or eWAT Id1 expression level is lower in iWAT and rWAT.

Expression data are mean ± SD; n=3; *p<0.05. (D) OilRedO staining of nondifferentiated (day 0), day

6 and day 8 differentiated HIB1B brown preadipocyte cells. (E) Id1, aP2, PGC1α and Ucp1 mRNA transcript levels in day 4 differentiated HIB1B cells compared to nondifferentiated cells. Expression data are mean ± SD; n=3; **p<0.005, ***p<0.0005. (F) The expression patterns of Id1, PPARγ, PGC1α,

UCP1, aP2 and βactin during HIB1B brown adipocyte differentiation at the indicated time points. (G)

Immunofluorescence costaining of aP2 and Id1 on day 4 differentiated HIB1B cells showing strong expression of Id1 in aP2positive cells. Scale bar 100µm. (H) Relative mRNA transcript levels of indicated genes in control and Id1OE nondifferentiated HIB1B cells. Expression data are mean ± SD; n=3; *p<0.05, **p<0.005. (I) FACS analysis of unstained and Nonyl Acridine Orange (NAO)stained control and Id1OE nondifferentiated HIB1B cells. (J) The relative fluorescence levels of NAO in

22 Page 23 of 53 Diabetes

control and Id1OE cells (n=3); *p<0.05. (K) O2 consumption rate of control and Id1OE non

differentiated HIB1B cells, (n=4); **p<0.005.

Figure 2: Adipose tissuespecific overexpression of Id1 reduces energy expenditure. (A) aP2Id1

expression vector used to generate adipose tissuespecific aP2Id1Tg+ transgenic mice. (B) Gel

photograph showing the PCR bands of internal control (βactin) and Id1 transgene. (C) Expression

levels of Id1 protein and βactin in the eWAT and BAT of aP2Id1Tg control and aP2Id1Tg+ transgenic

mice. (D) Body weight of male aP2Id1Tg and aP2Id1Tg+ mice (n=810) at the indicated time points.

(E) Fat mass density in 2 and 6month old aP2Id1Tg and aP2Id1Tg+ male mice measured by multi

slice CT scanner (n=56). Data are mean ± SD; *p<0.05. (F) Body weight of female aP2Id1Tg and aP2

Id1Tg+ mice (n=810) at the indicated time points. (G) Fat mass density in 2 and 6month old aP2Id1Tg

Tg+ and aP2Id1 female mice measured by multislice CT scanner (n=5). Data are mean ± SD. (HJ) O2

Tg consumption rate, CO2 release, heat production, food consumption and physical activity in aP2Id1

and aP2Id1Tg+ male mice measured for 4 days using Oxymax/CLAMS animal monitoring system (n=5

6). Data are mean ± SD; **p<0.001. (K) Representative H&E stained BAT and eWAT sections. (LM)

Adipocyte size distribution (L), and the average size of adipocytes (M) in the eWAT of 6month old

aP2Id1Tg and aP2Id1Tg+ male mice.

Figure 3: aP2-Id1Tg+ mice are prone to HFDinduced obesity. (A) Body weights of HFDfed male

aP2Id1Tg and aP2Id1Tg+ mice (n=9) at the indicated time points. (BD) Representative pictures of

aP2Id1Tg and aP2Id1Tg+ male mice (B), visceral fat (C), and eWAT pads (D) after 12 weeks of HFD.

(E) Average weight of eWAT pads in aP2Id1Tg and aP2Id1Tg+ mice (n=6). Data are mean ± SD;

*p<0.05. (F) Fat mass density in aP2Id1Tg and aP2Id1Tg+ male mice after 12 weeks of HFD measured

by multislice CT scanner (n=6). Data are mean ± SD; *p<0.05. (GH) Representative photos of

interscapular BAT (G) and liver (H) of aP2Id1Tg and aP2Id1Tg+ male mice. BAT weight is

23 Diabetes Page 24 of 53

significantly higher in HFD fed aP2Id1Tg+ mice compared to aP2Id1Tg mice due to higher lipid accumulation (n=6). Data are mean ± SD; *p<0.05. (I) Average liver weight in aP2Id1Tg and aP2

Id1Tg+ mice after 12 weeks of HFD (n=6). Data are mean ± SD; *p<0.05. (JL) Representative H&E stained eWAT sections (J), adipocyte size distribution (K), and the average size of adipocytes (L) in the eWAT of HFDfed aP2Id1Tg and aP2Id1Tg+ male mice (n=6). Data are mean ± SD; *p<0.05. (MN)

Representative H&E stained BAT and liver sections of HFDfed aP2Id1Tg and aP2Id1Tg+ male mice.

(OS) Food consumption, physical activity, O2 consumption rate, CO2 release and heat production, in

HFDfed (12 weeks) aP2Id1Tg and aP2Id1Tg+ male mice measured for 4 days using Oxymax/CLAMS animal monitoring system (n=5). Data are mean ± SD; **p<0.0001.

Figure 4: Thermogenic protein levels and O2 consumption rate are reduced in the BAT of aP2-

Id1Tg+ mice. (A) Expression levels of Id1 and βactin in the BAT of 2monthold wildtype mice at RT and after exposing mice to 4°C for 4h. (B) Expression levels of Id1 and βactin in the BAT of 2month old NDfed or HFDfed (1 week) wildtype mice. (CD) Relative mRNA transcript levels of thermogenic and mitochondrial genes in the BAT of aP2Id1Tg and aP2Id1Tg+ male mice at RT. (E)

Expression levels of indicated proteins in the BAT of 2monthold aP2Id1Tg and aP2Id1Tg+ male mice at RT. (FG) Relative mRNA transcript levels of thermogenic and mitochondrial genes in the BAT of aP2Id1Tg and aP2Id1Tg+ male mice after exposing mice to 4°C for 4h. (H) Basal respiration and uncoupled respiration (after blocking ATP synthase with oligomycin) were determined in the BAT explants of aP2Id1Tg and aP2Id1Tg+ male mice using OxygraphPlus System (n=4). Data are mean ±

SD; **p<0.05. (IJ) Relative mRNA transcript levels of thermogenic and mitochondrial genes in the

BAT of Id1+/+ and Id1/ male mice at RT. (K) Expression levels of indicated proteins in the BAT of 2 monthold Id1+/+ and Id1/ mice at RT. After acquiring images, band intensities were measured and normalized to βactin by the LiCor Odyssey system. (LM) Relative mRNA transcript levels of thermogenic and mitochondrial genes in the BAT of Id1+/+ and Id1/ male mice after exposing mice to

24 Page 25 of 53 Diabetes

4°C for 4h. All the qPCR data are mean ± SD, n=68; *p<0.05, **p<0.005. (N) Basal respiration and

uncoupled respiration were determined in the BAT explants of Id1+/+ and Id1/ mice (n=4). Data are

mean ± SD; **p<0.05.

Figure 5: Id1 suppresses PGC1α transcriptional activity. (A) CoIP followed by Western blot

showing a direct interaction between Id1 and PGC1α (HAtag), whereas Id1 did not directly interact

with other proteins that were tested. Input: 2% of IP reaction. (B) CoIP followed by Western blot

showing a direct interaction between endogenous Id1 and PGC1α in day 6 differentiated HIB1B cells.

Input: 5% of IP reaction. (C) CoIP followed by Western blot showing a direct interaction between

endogenous Id1 and PGC1α in the BAT of mice at RT and after exposing mice to 4°C for 4h. Input: 5%

of IP reaction. Higher exposure image was also included to show PGC1α input bands which were not

visible under lower exposure. Band intensities were measured and normalized to PGC1α input bands by

the LiCor Odyssey system. (D) Transcriptional activity of PGC1α was assayed by cotransfecting Cos

7 cells with luciferase reporter driven by GalUAS sequence and Gal4PGC1α plasmid (fulllength

PGC1α fused with the DNAbinding domain (DBD) of yeast Gal4) with different concentrations of Id1

expression vector. Data are mean ± SD, n=3; *p<0.05, **p<0.005, ***p<0.0005. (E) Transcriptional

activity of PGC1α was assayed by cotransfecting Cos7 cells with luciferase reporter driven by Ucp1

promoter, PPARγ, RXRα (PPARγ binding partner) and different concentrations of Id1 plasmid. Data are

mean ± SD, n=2; *p<0.05, **p<0.005.

Figure 6: Increased beige/thermogenic gene expression in the iWAT Id1-/- mice. (A)

Immunofluorescence staining showing the localization pattern of CD45 and Id1 in the iWAT of 2

monthold wildtype mice. iWAT sections incubated only with fluorescent secondary antibodies served

as ve controls. Scale bar 100µm. (BC) Relative mRNA transcript levels of beige and thermogenic

genes in the iWAT of Id1+/+ and Id1/ male mice after exposing them to 4°C for 12 h. Data are mean ±

25 Diabetes Page 26 of 53

SD, n=7; *p<0.05. (D) Representative IHC staining showing the expression pattern of UCP1 in the iWAT of Id1+/+ and Id1/ mice after exposing them to 4°C for 12 h, n=5; 3 out of 5 Id1/ mice showed strong UCP1 staining in the iWAT, and no staining was detected in the Id1+/+ mice. Scale bar 10µm. iWAT sections incubated only with HRPconjugated secondary antibody served as –ve control.

Figure 7: Increased beige/thermogenic gene expression in Id1-/- MEFs.

(A) Representative OilRedO staining pictures of day 10 differentiated Id1+/+ and Id1/ MEFs. Scale bar 20µm. (B) Expression patterns of Id1, aP2 and βactin in differentiating MEFs at the indicated time points after inducing brown adipogenesis. (CD) FACS analysis of unstained and Nonyl Acridine

Orangestained day 10 differentiated Id1+/+ and Id1/ MEFs (C), and the quantification of fluorescence level (D). Data are mean ± SD, n=4; **p<0.005. (E) Basal and uncoupled respiration (after blocking

ATP synthase with oligomycin) rate of day 10 differentiated Id1+/+ and Id1/ MEFs. Data are mean ±

SD, n=3; *p<0.05. (FK) Relative mRNA transcript levels of beige (FG), mitochondrial (HI) and thermogenic (JK) genes in nondifferentiated (F, H, J) and day 10 differentiated (G, I, K) Id1+/+ and

Id1/ MEFs. Expression data are mean ± SD; n=3; *p<0.05, **p<0.005, ***p<0.0005. (L) Expression levels of indicated proteins in Id1+/+ and Id1/ differentiating MEFs at the indicated time points after inducing brown adipogenesis.

Figure 8: Id1 suppresses Ebf2 transcriptional activity. (AC) Immunofluorescence staining showing the localization patterns of CD45/Ebf2 and Id1/Ebf2 in the iWAT of 2monthold wildtype mice. Scale bar 100µm (AB), and 20µm (C). (D) CoIP followed by Western blot showing a direct interaction between Id1 and Ebf2, Input: 2% of IP reaction. (E) Relative reporter activity of Prdm16 promoter driven luciferase activity in the presence of PPARγ, RXRα (PPARγ binding partner), Ebf2 and different concentrations of Id1 plasmid. Data are mean ± SD, n=2; **p<0.005. (F) Relative reporter activity of

Prdm16 promoterdriven luciferase activity in the presence of PPARγ, RXRα, and in the presence and

26 Page 27 of 53 Diabetes

absence of Ebf2 and Id1. Data are mean ± SD, n=2; **p<0.005. Id1 suppressed Prdm16 promoterdriven

luciferase activity only when Ebf2 is present. (GH) ChIPqPCR analysis of Ebf2 and Id1 binding to the

Prdm16 promoter (G), and PGC1α and Id1 binding to the Ucp1 promoter (H) in day 4 differentiated

HIB1B cells after normalizing to 18s binding. Data are mean ± SD; n=2; *p<0.05, ***p<0.0005.

27 Diabetes Page 28 of 53 “Patil_Fig1”

A 1.5 B 1.5 C -actin) -actin) 1 1 * Id1 * 0.5 0.5 β-actin

Id1 levels ( β 0 BAT iWAT eWAT rWAT BAT iWAT eWAT rWAT Relative mRNA levels 0 BAT ifat efat rfat BAT iWATIfat eWATEfat rWATrfat

D Day 0 Day 6 Day 8 F

Id1

PPARγ

PGC1α 8 E ND *** D (day 4) 6 UCP1 *** 4 ** ** aP2 2 β-actin Relative mRNA levels 0 Id1 PGC1 α Ucp1 aP2 0 2 4 6 8 ID1 PGC UCP1 AP2 days alpha G Hib1B 1.5 Hib1B + Id1-OE Day 4 aP2 H * * 1.2 * ** * * * 0.9 0.6 0.3

Relative levels mRNA Relative 0 ap2 Nrf1 Nrf2 Dio2 Ucp1 ACCa Evol3 Evol6 Cox 4 Cidea Cpt1b PPARg Erralpha Id1 I Unstained PGC1 alpha Hib1B Hib1B + Id1-OE

Merge 250000 600 7 Hib1B + Id1 Hib1B + Hib1B + Id1 Hib1B + Hib1B J * K * Hib1B 200000500 6

5 150000400 - 4 - OE OE

100000300 3 /µg protein) 2 consumption

50000200 2 O pmol 1 ( fluorescence Relative 1000 0 1 1 Page 29 of 53 Diabetes “Patil_Fig2”

TB A fl ori B C aP2-Id1Tg+ aP2-Id1Tg- m IC-413bp Id1 aP2 β-actin pUC ori Tg-228bp eWAT BAT H SV40pA Id1 aP2-Id1Tg- 14000 O Consumption aP2-Id1Tg- aP2-Id1Tg- 2 aP2-Id1Tg+ ) hr Tg+ 11000 aP2-Id1 E aP2-Id1Tg+ D 8000 40 25000 * ) (ml/Kg/ 3 35 * * * 20000 2 5000 30 * p<0.001 15000 VO 2000 25 01 96h 53

20 10000 105 157 209 261 313 365 417 469 521 14000 CO Release 15 2

Males ) hr Fat mass (mm 5000 11000

Body weight (gm) weight Body 10 1 2 3 4 5 6 0 1 2 3 4 5 6 2M 6M 8000

Months 1 2 (ml/Kg/

2 5000

Tg- Tg- aP2-Id1 p<0.001 aP2-Id1 VCO 2000 Tg+ aP2-Id1Tg+ G aP2-Id1 1

0 53 96h

F 10000 105 157 209 261 313 365 417 469 521 ) 80 3 30 Heat production 8000

) hr 60 25 6000 40 20 4000

Kcal/Kg/ 20 15 Females Fat mass (mm 2000 p<0.001 0 Body weight (gm) weight Body 10 0 1 1 2 3 4 5 6 0 53 96h 1m 2m 3m 4m 5m 6m 2M 1 6M 105 157 209 261 313 365 417 469 521 Months I J n.s 80000 5 n.s K 2 Months 6 Months 4 aP2-Id1Tg- aP2-Id1Tg+ aP2-Id1Tg- aP2-Id1Tg+ 60000 3

/day) 40000 BAT gm 2 ( 20000 Total activity activity Total

1 Food consumptionFood 0 (counts/animal/day) 0 1 1 eWAT M aP2-Id1Tg- Tg+ 25000 aP2-Id1

* 20000 L 6 Months 15000 7% 20% 10000

32% (Arbitrary units) <5000 Arbitrary 5000 5000-10000 units 45% 48% >10000 Average size of adipocytes of size Average 0 50% 1 aP2-Id1Tg- aP2-Id1Tg+ Diabetes Page 30 of 53 “Patil_Fig3”

aP2-Id1Tg- aP2-Id1Tg+ aP2-Id1Tg- aP2-Id1Tg+ aP2-Id1Tg-

D 50 HFD * B C A E Tg+ * aP2-Id1 40 1.5 * ) 30 gm 1 20 aP2-Id1Tg- pad ( pad Tg+ Body weight (gm) weight Body aP2-Id1 10 0.5

0 2 4 6 8 12 eFAT Weeks 0 1 Tg- Tg+ Tg- Tg+ aP2-Id1 aP2-Id1 aP2-Id1 G aP2-Id1 aP2-Id1Tg- aP2-Id1Tg- 400±70mg 495±55mg J F I aP2-Id1Tg+ eWAT aP2-Id1Tg+ 2.5 25000 ) *

* )

3 2 20000 gm Interscapular BAT 15000 1.5 >20000 K aP2-Id1Tg- aP2-Id1Tg+ 10000 <10000 H 1 10000 2% - 20000

Fat mass (mm 5000 Liver weight weight ( Liver 0.5 32% 25% 27% 0 0 66% units Arbitrary Liver 1 1 48% aP2-Id1Tg- Tg- Tg+ L Tg+ aP2-Id1 aP2-Id1 aP2-Id1 M B

aP2 aP2

AT * aP2 25000 aP2

- - 80000 Id1

- Id1 4 - Id1 P Id1 O n.s

n.s Tg+ 20000 Tg 70000 Tg+ /day) Tg -

- 60000 gm 3 15000 N 50000

10000 Liver 2 40000 30000 (Arbitrary units) (Arbitrary 5000 1 20000 10000

Average size size of adipocytes Average 0 Total activity (counts/animal/day) activity Total Food consumption consumption Food ( 0 0 1 1 1 Q R aP2-Id1Tg- S Tg+ aP2-Id1 10000 O Consumption 10000 CO Release 50 Heat production 2 2 ) hr ) hr 8000 8000 40 ) hr 6000 6000 30 (ml/Kg/

4000 4000 20 (ml/Kg/

2 2

2000 2000 Kcal/Kg/ 10 VO

p<0.0001 VCO p<0.0001 p<0.0001 0 0 0 1 1 0 1 96h 0 96h 73 73 73 0 96h 145 217 289 361 433 505 145 217 289 361 433 505 145 217 289 361 433 505 Page 31 of 53 Diabetes “Patil_Fig4”

E A RT 4oC B ND HFD aP2-Id1Tg- aP2-Id1Tg+

Id1 Id1 PGC1α

β-actin β-actin Ucp1

Tg- 1.0 0.82 3 aP2-Id1 C RT 1.5 D aP2-Id1Tg+ Dio2 2 1 p=0.08 * 1.0 0.74 1 0.5 Cyt-c

Relative levels mRNA Relative 0 0 1.0 0.8 Relative levels mRNA Relative Cyt-c Cox5b Cox7a cytc cox5 cox7 aP2-Id1Tg- β-actin 4oC * 1.5 F * 1.5 aP2-Id1Tg+ * * * G * 1 1 H Basal Uncoupled 5 respiration 2.5 respiration 0.5 0.5 4 2 0 0 * Relative levels mRNA Relative Relative levels mRNA Relative * cytcCyt-c cox5Cox5b cox7Cox7a 3 1.5

/µg protein) 2 1 consumption

2 1 0.5 O pmol ( Id1+/+ 0 0 * 8 I RT 1.5 J Id1-/- 6 1 4 p=0.1 2 0.5 K Id1+/+ Id1-/- 0 0 Relative levels mRNA Relative Relative levels mRNA Relative cytcCyt-c cox5b Cox5b Cox7acox7 PGC1α o 4 C +/+ 2.5 Id1 L * 1.5 -/- 2 * M Id1 * ** Ucp1 1.5 1 1.0 1.66 1 0.5 Dio2 0.5

Relative levels mRNA Relative 0 0 Relative levels mRNA Relative Cyt-c Cox5b Cox7a cytc cox5 cox7 Cyt-c 1.0 1.23 N Basal Uncoupled

6 respiration 2 respiration β-actin * ** 4 1.5 1

/µg protein) 2 consumption

0.5 2 O pmol ( 0 0 +/+ -/- +/+ -/- Id1wt Id1ko Id1wt Id1KO Diabetes Page 32 of 53 A “Patil_Fig5” C RT 4°C IB-Id1 IB IB-HA- (PGC1α) PGC1α (Low) 3.8 1.0 IB-Id1 PGC1α (High) IB-PPARγ 1.0 1.41

Id1 IB-Id1

IB-Rb D PGC1α

DBD 120 Gal4-UAS 100 IB-Id1 * 80 IB-Foxc2 60 ** 40 *** IB-Id1 20

Relative reporter activity 0 IB-CREB Gal4-UAS 40 40 40 40 40 40 (ng) Gal4-PGC1α 40 40 40 40 40 40 (ng) Id1 0 10 20 40 80 120 (ng) IB-Id1

IB-IRF4 E

120 100 IB-Id1 80 60 * IB-Sirt3 **

activity ** 40 ** 20 Relative reporter Relative B 0 Ucp1 pro 100 100 100 100 100 (ng) IB-Id1 PGC1α 100 100 100 100 100 (ng) PPARγ 100 100 100 100 100 (ng) IB-PGC1α RXRα 100 100 100 100 100 (ng) Id1 0 100 200 300 400 (ng) Page 33 of 53 Diabetes “Patil_Fig6”

A CD45 Id1 Merge

CD45 –ve control Id1 –ve control

Id1+/+ +/+ C B Id1 Id1-/- 4 Id1-/- 4 * 3 * 3 P=0.1 * * * 2 2 1 1 0 Relative levels mRNA Relative

Relative levels mRNA Relative 0 Tbx1 Tmem26 Cd40 Cd137

D Id1+/+ Id1-/- -ve control UCP1 Diabetes Page 34 of 53 “Patil_Fig7”

B +/+ A Id1+/+ Id1-/- Id1 Id1-/- Id1

aP2

β-actin 0 2 4 6 8 10 12 0 2 4 6 8 10 12 Days C D E +/+ 100 Unstained Id1+/+ 2.5 * Id1

250 ** +/+ 10 * -/- 80 Id1 -/- Id1 Id1-/- Id1 200 8 2 60 150 6 1.5 40 Counts 100 /µg of protein) 4 1 consumption

20 50 2 2 0.5 0 O 0 1 2 3 fluorescence Relative 10 10 10 10 0 pmoles 0 0 ( FL1-Height 1 Basal1 Uncoupled1

Id1+/+ MEFs (ND) MEFs (D10) 6 -/- 8 ** F Id1 G ** L Id1+/+ Id1-/- 4 6 4 Id1 2 2 * 0 0 aP2 Relative mRNA levels Tbx1 Tmem26 CD40 CD137 Relative mRNA levels Tbx1 Tmem26 Cd40 Cd137 RB

1.5 3 pRB H I * 1 2 * * RIP140 0.5 1 PPARγ 0 0 Relative mRNA levels Relative mRNA levels Cyt-c Cox5b Cox7a Cytcytc-c Cox5bcox5 cox7a1Cox7a cytc cox5 cox7 PGC1α

2 8 K J Ebf2 * * ** 1.5 * 6 *** ** ** PRDM16 1 4 ** **

0.5 2 β-actin Relative mRNA levels Relative mRNA levels 0 4 8 12 0 4 8 12 Days 0 0 Page 35 of 53 Diabetes “Patil_Fig8”

D Merge A CD45 Ebf2 IB-Id1

IB-Ebf2

E

Id1 Ebf2 Merge ** ** B 300

200 activity 100

Relative reporter Relative 0 -25kb Prdm16 100 100 100 100 (ng) PPARγ 1000 20ng 100 40ng 100 100ng 100 (ng) C Id1 Ebf2 Merge RXRα 100 100 100 100 (ng) Ebf2 100 100 100 100 (ng) Id1 0 40 80 120 (ng)

F ** 400

300 n.s 200 G H

*** activity 15 15 *** 100

10 10 reporter Relative 0

* -25kb Prdm16 100 100 100 100 (ng) 5 5 n.s PPARγ 100 100 100 100 (ng) RXRα 100 100 100 100 (ng) (Ucp1 Promoter) (Ucp1

Fold EnrichmentFold Ebf2 0 0 100 100 (ng)

Fold EnrichmentFold 0 0 Id1 0 100 0 100 (ng) (PRDM16 Promoter)(PRDM16 ebf2Ebf2 id1Id1 neg IgG trl PGC1α Id1 IgG Diabetes Page 36 of 53

Id1 promotes obesity by suppressing brown adipose thermogenesis and white adipose browning

Mallikarjun Patil, Bal Krishan Sharma, Sawsan Elattar, Judith Chang, Shweta Kapil, Jinling Yuan & Ande Satyanarayana

Suppl. Information

Id1 partially suppresses Ebf2-mediated re-programing of C2C12 myoblasts into brown adipocytes

Forced expression of Ebf2 alone is sufficient to reprogram C2C12 myoblasts into brown adipocytes (1).

Since Id1 directly interacted with Ebf2 and suppressed its transcriptional activity (Fig 8D-F), we asked if

Id1 inhibits Ebf2-mediated reprograming of C2C12 myoblasts into brown adipocytes. We generated

Ebf2-expressing C2C12 cells (Suppl. Fig 4A), and as demonstrated previously (1), Ebf2- expressing

cells differentiated into adipocytes as determined by aP2 expression and Oil-Red-O staining (Suppl. Fig

4B-C). We then expressed Id1 in Ebf2-C2C12 cells (Suppl. Fig 4D) and observed reduced differentiation in Ebf2/Id1-C2C12 cells compared to Ebf2-C2C12 cells on day 4 (Suppl. Fig 4E).

Consequently, we detected reduced induction of Prdm16, PGC1α and Ucp1 mRNA in Id1 overexpressing day 4 differentiated Ebf2-C2C12 cells, whereas in non-differentiated cells, except for

PGC1α, expression of the other genes was not affected (Suppl. Fig 4F-G). However, by day 8 the overall differentiation was not different between Ebf2-C2C12 and Ebf2/Id1-C2C12 cells as determined by Oil-Red-O staining (Suppl. Fig 4H). Similarly, the expression of Prdm16, PGC1α and Ucp1 were also unchanged in day 8 differentiated Ebf2-C2C12 and Ebf2/Id1-C2C12 cells (Suppl. Fig 4I). This is mainly due to the fact that Ebf2 expression is very strongly induced during C2C12 differentiation and

Id1 expression is down-regulated (Suppl. Fig 4J); therefore, Ebf2 is able to escape from Id1-mediated inhibition. Consistent with this explanation, we detected a direct interaction between Id1 and Ebf2 in day 0 cells (Suppl. Fig 4K). However, we were unable to immunoprecipitate Id1from latter time points

1 Page 37 of 53 Diabetes

(day 2-6) due to the fact that Id1 levels are very low. These results together suggest that Id1 was unable

to sustain the inhibition on Ebf2-mediated programing of C2C12 cells into brown adipocytes due to the

fact that Id1 is strongly downregulated during C2C12 differentiation.

Additional Information

Immunoblotting and co-immunoprecipitation

Whole-cell protein lysates were prepared from cells and tissues using RIPA lysis buffer

(ThermoScientific # 89901) supplemented with complete mini-protease inhibitor tablet (Roche #

11836153001). For Western blot analysis, 50µg of protein was separated on NuPAGE precast gels

(Invitrogen), transferred using a XCell II Blot module (Invitrogen # 090707-098) onto Immobilon-FL

membranes (Millipore # IPFL00010), and probed with specific primary antibodies: mouse Id1

(Biocheck # BCH-1, 37-2), mouse/human Id1 (Biocheck # BCH-1/195-14), Rb (Santa Cruz # (sc-50),

pRb (Cell Signaling, # 8516), RIP140 (Santa Cruz # sc-8997), Foxc2 (abcam # ab5060), SRC1 (Santa

Cruz # sc-136077), PPARγ (Thermo Scientific # MA5-14889), PPARγ (Cell Signaling # 2443), Ebf2

(R&D Systems # AF7006), CREB (Cell Signaling # 9197), PGC1α (Santa Cruz # sc-13067), aP2 (Santa

Cruz # sc-18661), Ucp1 (abcam # ab10983), PRDM16 (abcam # ab106410), Sirt3 (Cell Signaling #

5490), IRF4 (Santa Cruz # sc-6059), HA-Tag (Cell Signaling # 3724), Cyt-c (Cell Signaling # 4280),

Dio2 (Proteintech # 26513-1-AP) and β-actin (Sigma # A5441). All the antibodies were used at a

dilution of 1:1000 except Id1 (1:500), Ebf2 (1:500) and β-actin (1:10,000). The following IRDye-

conjugated secondary antibodies were used: donkey anti-mouse IRDye800CW (Licor # 926-32212) and

donkey anti-rabbit IRDye800CW (Licor # 926-32213), donkey anti-mouse IRDye680RD (Licor # 925-

68072) and donkey anti-rabbit IRDye680RD (Licor # 925-68073). Licor Odyssey Classic Imager was

used to detect Western blots. For co-immunoprecipitation assays, 1mg of protein from whole-cell lysates

and 20 μl of Protein A/G PLUS-Agarose immunoprecipitation reagent (sc-2003), Rabbit IgG (Santa

2 Diabetes Page 38 of 53

Cruz # sc-2027) and Mouse IgG (Santa Cruz # sc-2025) were used and co-IPs were performed as

described previously (2).

Oxymax/CLAMS metabolic analysis

Mice (aP2-Id1Tg+ and aP2-Id1Tg-) were acclimated for 12 h in the metabolic cages, and their metabolic

rates were measured for 96 h in an indirect open-circuit calorimeter (Oxymax Comprehensive Lab

Animal Monitoring System; Columbus Instruments). O2 consumption, CO2 release, heat production,

food consumption and physical activity were measured at room temperature (RT) and normalized to

body weight to account for the disparity in body weight between the two groups.

Multi-slice CT scans

The fat volume in aP2-Id1Tg+ and control aP2-Id1Tg- mice was determined by multi-slice microCT

imaging (Mediso Imaging System # Nanoscan SpectCT) as described previously (3). Lean mass and

fluid content was measured by LF90II BCA-Analyzer (Bruker). Bone mass was measured by Lunar

PIXImus2.

H&E and Oil-Red-O staining

Oil-Red-O staining and the staining intensity measurements on differentiated adipocytes, and H&E staining on different tissues were performed as described previously (4).

Adipocyte size measurement

H&E stained WAT sections were photographed under 40X magnification using AmScope Inverted microscope equipped with 9MP Camera and ToupView software. The adipocyte area was measured using imageJ software and presented as arbitrary units. A total of 200 adipocytes were measured from each genotype.

3 Page 39 of 53 Diabetes

Immunocytochemical staining

Day 4 differentiated HIB1B cells cultured on coverslips in 6-well plates were washed twice with cold

PBS, fixed with ice-cold acetone: methanol (1:1) for 5 min, and washed 3 times with ice-cold PBS and

incubated with Id1 (#BCH-1, 37-2; BioCheck) and aP2 (Santa Cruz # sc-18661) antibodies at 4°C

overnight. The cells were then washed 3 times with PBS/0.2% Tween 20 and incubated with AlexaFluor

555 goat anti-rabbit (Life Technologies # A21428) and AlexaFluor 488 goat anti-mouse (abcam #

ab150105) secondary antibodies for 1 h at room temperature. The coverslips were washed 3 times with

PBS/0.2% Tween 20, air dried for 10 min, and mounted with fluorescence mounting medium (Vector

Laboratories Vectashield # H-1200) containing DAPI. The optical single-slice images were captured

with the Zeiss Axio fluorescence laser-scanning microscope.

Immunohistochemistry (IHC)

For IHC on iWAT, slides were deparaffinized in xylol, and antigen retrieval was performed by boiling

the slides for 30 min in H2O. After cooling for 20 min and washing in PBS, endogenous peroxidase was

blocked with 0.3% hydrogen peroxide for 30 min. The slides were then incubated with UCP1 antibody

(1:300 dilution, abcam # ab10983) overnight at 4°C. UCP1 staining was detected using UltraVision

Quanto Detection System HRP DAB (Thermofisher # TL-060-QHL). The staining pattern was analyzed

and photographed using AmScope inverted microscope equipped with 9MP camera and ToupView

software. For Id1, CD45 and Ebf2 stainings, slides were incubated after antigen retrieval with primary

antibodies Id1, 1:100 dilution (Biocheck # BCH-1, 37-2), Ebf2, 1:50 dilution (R&D Systems # AF7006)

and CD45, 1:200 dilution (abcam # ab10558) overnight at 4°C. The slides were then washed 3 times

with PBS/0.2% Tween 20 and incubated with AlexaFluor 555 goat anti-rabbit (Life Technologies #

A21428) and AlexaFluor 488 anti-mouse (abcam # ab150105) secondary antibodies for 1 h at RT. The

slides were washed 3 times with PBS/0.2% Tween 20, air dried for 10 min, and mounted with

fluorescence mounting medium (Vector Laboratories Vectashield # H-1200) containing DAPI. For

4 Diabetes Page 40 of 53

CD45/Id1, CD45/Ebf2 and Id1/Ebf2 double stainings, primary antibody incubations were performed

sequentially on consecutive days. The optical images were captured with the Zeiss Axio fluorescence

laser-scanning microscope.

Brown adipocyte differentiation in HIB1B cells, MEFs and C2C12 myoblasts

HIB1B brown pre-adipocyte cells were a kind gift from Dr. Bruce Spiegelman (Harvard Medical

School). HIB1B cells were cultured in 10cm dishes in DMEM (HyClone Laboratories # SH30022.01)

supplemented with 10% heat-inactivated fetal bovine serum (FBS; Gemini Bio-Products #100106) and

1% penicillin/streptomycin (Life Technologies # 15140) in a humidified cell culture incubator (Thermo

Fisher Scientific # Heracell 150i) maintained at 37°C, 5% CO2, and 20% O2. After the cells reached

100% confluence, the media was replaced with adipogenesis differentiation media [DMEM, 10% FBS,

20nM insulin (I2643; Sigma) and 1nM triiodothyronine (T3) (Sigma # T6397) for 48 h. Adipocyte

differentiation was induced by treating cells for an additional 48h in differentiation medium further

supplemented with 0.5µM dexamethasone (Sigma # D1756), 0.5mM 3-isobutyl-1-methylxanthine

(Sigma # I7018), and 125 µM indomethacin (Sigma # I7378) (induction media). After induction, cells

were returned to differentiation medium, and the differentiated cells were collected at different time

points for further analysis. Mouse embryonic fibroblasts (MEFs) were cultured in 100mm dishes and induced to differentiate into brown adipocytes as described previously (5). Briefly, E13.5 Id1+/+ and Id1-

/- MEFs (passage 2) were cultured in DMEM supplemented with 10% FBS and 1%

penicillin/streptomycin. After the cells became fully confluent, the cells were treated for 48h with media

containing 10% FBS, 0.5 mM 3-isobutyl-1-methylxanthine, 125 µM indomethacin, 1 µM

dexamethasone, 850 nM insulin, 1 nM T3, and 1µM rosiglitazone (Sigma # R2408). Two days after

induction, cells were switched to maintenance medium containing 10% FBS, 850 nM insulin, 1 nM T3,

and 1 µM rosiglitazone, and the differentiated cells were collected at different time points for further

analysis. C2C12 cells were purchased from ATCC (CRL-1772, lot # 62042837). The full-length mouse

5 Page 41 of 53 Diabetes

Ebf2 gene was PCR amplified and sub-cloned into the retroviral vector PLNCX2. PLNCX2-Ebf2 was

used for viral packaging in ampho-phoenix cells. C2C12 cells were transduced with retrovirus

expressing Ebf2 using polybrene (Millipore # TR-1003-G) at a concentration of 5µg/ml at 35°C. Viral

transduction was conducted twice with an interval of 24h and the cells were then selected with G418

(750µg/ml, Sigma # G8168) to generate stable Ebf2-expressing C2C12 cells. The Ebf2-C2C12 cells

were transiently transfected with pcDNA3 mId1 (gift from Robert Benezra, Addgene plasmid # 16060)

to overexpress Id1, and brown adipogenesis was induced in C2C12, Ebf2-C2C12 and Ebf2/Id1-C2C12

cells as described previously (1).

Retrovirus-mediated Id1 overexpression in HIB1B cells

The Id1 gene was excised from pCDNA3 mId1 (Addgene plasmid # 16060) by restriction digestion and

sub-cloned into the retroviral vector PLNCX2. PLNCX2-Id1 was then used for viral packaging in

ampho-phoenix cells. HIB1B cells were transduced with retrovirus expressing Id1 using polybrene at a

concentration of 5µg/ml at 35°C. Viral transduction was conducted twice with an interval of 24h, and

the cells were then selected with G418 (200µg/ml) to generate stable Id1-expressing HIB1B cells.

Mitochondrial staining and FACS analysis

For measuring mitochondrial content in MEFs and HIB1B cells, cells growing in complete medium

were incubated with 10 nM NAO (Nonyl Acridine Orange, Molecular Probes # A1372) dissolved in

PBS/5% FBS for 15 min at 37°C. Medium containing NAO was removed from the cells, then the cells

were washed twice with PBS, trypsinized, resuspended in PBS, and analyzed by the FACSCalibur flow

cytometry system equipped with CellQuest Pro Software (BD Biosciences).

O2 consumption measurement

Cellular respiration analysis was done as described previously (6). Cells were trypsinized, a single-cell

suspension was placed in respiration buffer, and the O2 consumption rate was measured using

6 Diabetes Page 42 of 53

OxygraphPlus System (Hansatech Instruments). For tissues, BAT was weighed and 25mg of tissue was

minced using a razor blade and equal amount of wet tissue was used to measure O2 consumption rate.

After measuring basal respiration, 1µM (final conc.) oligomycin was added to the respiration buffer to

measure uncoupled respiration. After the measurements, cells and tissues were lysed using RIPA lysis buffer, and whole cell protein concentration was measured. O2 consumption was normalized to protein

concentration.

Glucose, insulin, and adipokine measurements

Blood glucose levels were measured from the tail blood by One Touch Ultra II Glucometer (LifeScan)

and One Touch Ultra glucose strips (LifeScan). For insulin and leptin measurements, mice were starved

4 h, and blood serum was prepared, aliquoted and stored at -80°C until use. Insulin concentration was

determined by insulin (mouse) ultrasensitive EIA kit (Alpco # 80-INSMSU-E01). Leptin and

adiponectin levels were measured by leptin (murine/rat) EIA kit (Alpco # 22-LEPMS-E01) and adiponectin EIA Kit (Alpco # 17-ADPMS-E01) according to the manufacturer’s protocols. The ELISA plates were read using Synergy HTX Multi-Mode Plate Reader (BioTek).

Transient transfection, shRNA and overexpression vectors

ContinuumTM Transfection Reagent (Gemini Bio-products # 400-700) was used to transfect the cells

with control, shRNA, or expression vectors according to the manufacturer’s protocol. pCF-CREB was a gift from Marc Montminy (addgene plasmid # 22968) (7), RcCMV/Rb was a gift from Bob Weinberg

(Addgene plasmid # 1763), pBabe human FOXC2 was a gift from Bob Weinberg (Addgene plasmid #

15535) (8), pAd-Track HA PGC1α was a gift from Pere Puigserver (Addgene plasmid # 14427) (9), pcDNA3 hId1 was a gift from Robert Benezra (Addgene plasmid # 16061), pBABE puro PPARγ2 was a gift from Bruce Spiegelman (Addgene plasmid # 8859) (10), pMIG-IRF4 was a gift from David

Baltimore (Addgene plasmid # 58987) (11), SIRT3 Flag was a gift from Eric Verdin (Addgene Plasmid

7 Page 43 of 53 Diabetes

# 13814) (12), and pBabe hygro human RXRα was a gift from Ronald Kahn (Addgene plasmid #

11440). Control shRNA (shGL) and Id1-shRNA vectors were a kind gift from Jonathan Keller (NCI-

Frederick) and were described previously (13).

Luciferase reporter assays

Gal4-UAS and full-length PGC1α fused with DNA-binding domain (DBD) of yeast Gal4 vectors were

kind gift from Dr. Bruce Spiegelman (Harvard Medical School). For the luciferase reporter assays, Cos7

cells (3x104) were seeded onto 24-well plates. The next day, cells were co-transfected with Gal4-UAS

(40ng/well), DBD-Gal4-PGC1α (40ng/well), Id1 plasmid (different concentrations, 0-120ng) and

galactosidase expression vector (control reporter) (100ng/well). The mass of transfected plasmids was

equalized with an empty vector (pCDNA3.1). After 48 hr, cells were harvested and the luciferase

activity was measured using the Luciferase assay system (Promega # E1500) and Synergy HTX Multi-

Mode Plate Reader (BioTek). Luciferase activity was divided by β-gal (control reporter) to normalize

for transfection efficiency. For the Prdm16 promoter driven luciferase assay, the -25Kb Prdm16

promoter-driven luciferase vector was obtained from Dr. Patrick Seale (UPenn). HEK293 cells were

seeded at ~65% confluency in 12-well plates. The next day, cells were co-transfected with -25Kb

Prdm16 promoter-driven luciferase vector (100ng/well), PPARγ2 (100ng/well), RXRα (100ng/well),

Id1 plasmid (different concentrations, 0-120ng) and galactosidase expression vector (control reporter)

(100ng/well). The mass of transfected plasmids was equalized with an empty vector (pCDNA3.1). After

48 hr, cells were harvested and the luciferase activity was measured as described above. Ucp1 promoter-

driven luciferase assays were carried out in a similar fashion in Cos7 cells.

ChIP-qPCR

ChIP-qPCR was performed as described previously (1) using the ChIP assay kit (Millipore cat# 17-295).

Briefly, HIB1B preadipocytes were differentiated for 4 days and cross-linked with 1% formaldehyde

8 Diabetes Page 44 of 53

(final conc.) in the growth media at 370C for 10 min. Cells were then harvested and re-suspended in SDS lysis buffer at a final concentration of 107 cells per ml. Then the cells were sonicated for 18 cycles with

10 second intervals and centrifuged at 13000 RPM for 10 min. The supernatant was diluted 10 fold in

ChIP dilution buffer/cocktail inhibitor and pre-cleared using 75µl of salmon sperm DNA/ProteinG

agarose (Cell Signaling # 9007). Immunoprecipitations were carried out using IgG, Ebf2, PGC1α and

Id1 antibodies overnight and washed twice sequentially with low salt, high salt, lithium chloride and TE

buffers. DNA/protein complexes were eluted using SDS/NaHCO3 buffer and reverse cross linked at

650C for 4 h in the presence of 0.2M NaCl. The elutes were then treated with proteinase K for 1 h and

DNA was extracted using phenol chloroform and used for qPCR. Primer sequences were provided in

Suppl. Table 1.

Suppl. Bibliography

1. Rajakumari S, Wu J, Ishibashi J, Lim HW, Giang AH, Won KJ, Reed RR, Seale P: EBF2 determines and maintains brown adipocyte identity. Cell metabolism 2013;17:562-574 2. Pan D, Fujimoto M, Lopes A, Wang YX: Twist-1 is a PPARdelta-inducible, negative-feedback regulator of PGC-1alpha in brown fat metabolism. Cell 2009;137:73-86 3. Luu YK, Lublinsky S, Ozcivici E, Capilla E, Pessin JE, Rubin CT, Judex S: In vivo quantification of subcutaneous and visceral adiposity by micro-computed tomography in a small animal model. Medical engineering & physics 2009;31:34-41 4. Satyanarayana A, Klarmann KD, Gavrilova O, Keller JR: Ablation of the transcriptional regulator Id1 enhances energy expenditure, increases insulin sensitivity, and protects against age and diet induced insulin resistance, and hepatosteatosis. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 2012;26:309-323 5. Seale P, Kajimura S, Yang W, Chin S, Rohas LM, Uldry M, Tavernier G, Langin D, Spiegelman BM: Transcriptional control of brown fat determination by PRDM16. Cell metabolism 2007;6:38-54 6. Zhang H, Bosch-Marce M, Shimoda LA, Tan YS, Baek JH, Wesley JB, Gonzalez FJ, Semenza GL: Mitochondrial autophagy is an HIF-1-dependent adaptive metabolic response to hypoxia. The Journal of biological chemistry 2008;283:10892-10903 7. Du K, Asahara H, Jhala US, Wagner BL, Montminy M: Characterization of a CREB gain-of-function mutant with constitutive transcriptional activity in vivo. Molecular and cellular biology 2000;20:4320- 4327 8. Mani SA, Yang J, Brooks M, Schwaninger G, Zhou A, Miura N, Kutok JL, Hartwell K, Richardson AL, Weinberg RA: Mesenchyme Forkhead 1 (FOXC2) plays a key role in metastasis and is associated with aggressive basal-like breast cancers. Proceedings of the National Academy of Sciences of the United States of America 2007;104:10069-10074

9 Page 45 of 53 Diabetes

9. Lerin C, Rodgers JT, Kalume DE, Kim SH, Pandey A, Puigserver P: GCN5 acetyltransferase complex controls glucose metabolism through transcriptional repression of PGC-1alpha. Cell metabolism 2006;3:429-438 10. Tontonoz P, Hu E, Spiegelman BM: Stimulation of adipogenesis in fibroblasts by PPAR gamma 2, a lipid-activated transcription factor. Cell 1994;79:1147-1156 11. So AY, Sookram R, Chaudhuri AA, Minisandram A, Cheng D, Xie C, Lim EL, Flores YG, Jiang S, Kim JT, Keown C, Ramakrishnan P, Baltimore D: Dual mechanisms by which miR-125b represses IRF4 to induce myeloid and B-cell leukemias. Blood 2014;124:1502-1512 12. North BJ, Marshall BL, Borra MT, Denu JM, Verdin E: The human Sir2 ortholog, SIRT2, is an NAD+-dependent tubulin deacetylase. Molecular cell 2003;11:437-444 13. Wagner K, Zhang P, Rosenbauer F, Drescher B, Kobayashi S, Radomska HS, Kutok JL, Gilliland DG, Krauter J, Tenen DG: Absence of the transcription factor CCAAT enhancer binding protein alpha results in loss of myeloid identity in bcr/abl-induced malignancy. Proceedings of the National Academy of Sciences of the United States of America 2006;103:6338-6343

Suppl. Table 1: Mouse primers used for real-time qPCR

Gene Forward primer (5’-3’) Reverse primer (5’-3’)

Id1 GGTCCGAGGCAGAGTATTACA CCTGAAAAGTAAGGAAGGGGGA

aP2 ACACCG AGATTTCCTTCAAACTG CCATCTAGGGTTATGATGCTCTTC A

Cidea TGCTCTTCTGTATCGCCCAGT GCCGTGTTAAGGAATCTGCTG

Dio2 CAGTGTGGTGCACGTCTCCAATC TGAACCAAAGTTGACCACCAG

Pgc1α CCCTGCCATTGTTAAGACC TGCTGCTGTTCCTGTTTTC

Pparγ GTGCCAGTTTCGATCCGTAGA GGCCAGCATCGTGTAGATGA

Prdm16 CAGCACGGTGAAGCCATTC GCGTGCATCCGCTTGTG

Ebf2 GCTGCGGGAACCGGAACGAGA ACACGACCTGGAACCGCCTCA

Ucp1 ACTGCCACACCTCCAGTCATT CTTTGCCTCACTCAGGATTGG

Elovl3 TCCGCGTTCTCATGTAGGTCT GGACCTGATGCAACCCTATGA

Elovl6 TGCTGCATCCAGTTGAAGAC TGCCATGTTCATCACCTTGT

Nrf1 GAACTGCCAACCACAGTCAC TTTGTTCCACCTCTCCATCA

Nrf2 GCTTTTGGCAGAGACATTCC ATCAGCCAGCTGCTTGTTTT

10 Diabetes Page 46 of 53

Cytc GCAAGCATAAGACTGGACCAAA TTGTTGGCATCTGTGTAAGAGAATC

Cox4 ACCAAGCGAATGCTGGACAT GGCGGAGAAGCCCTGAA

Cox5b GCTGCATCTGTGAAGAGGACAAC CAGCTTGTAATGGGTTCCACAGT

Cox7a CAGCGTCATGGTCAGTCTGT AGAAAACCGTGTGGCAGAGA

Cpt1b CGAGGATTCTCTGGAACTGC GGTCGCTTCTTCAAGGTCTG

Errα TTCTGCACAGCTTCCACATC GGAAGAATTCGTCACCCTCA

Cd40 TTGTTGACAGCGGTCCATCTA CCATCGTGGAGGTACTGTTTG

Cd137 CGTGCAGAACTCCTGTGATAAC GTCCACCTATGCTGGAGAAGG

Tbx1 GGCAGGCAGACGAATGTTC TTGTCATCTACGGGCACAAAG

Tmem26 ACCCTGTCATCCCACAGAG TGTTTGGTGGAGTCCTAAGGTC

18S GTAACCCGTTGAACCCCATT CCATCCAATCGGTAGTAGCG

Prdm16 GACCTCCTGCCTTCCCTGAGG GCTGCCTGAGCTGGGCCAGCC ChIP UCP1- AGTGAAGCTTGCTGTCACTC GTCTGAGGAAAGGGTTGACC ChIP 18s ChIP AGTCCCTGCCCTTTGTACACA CGATCCGAGGGCCTCACT

Suppl. Table 2: Metabolic parameters in aP2-Id1Tg- and aP2-Id1Tg+ mice (n=7). *p<0.05

Blood Glucose(mg/dL) aP2-Id1Tg- aP2-Id1Tg+

2-month-old 138.7±8.8 158.7±19.9

6-month-old 166.6±5 169.3±22

High-fat diet* 148.6±46.1 191±51.3 *

Insulin (ng/ml)

2-month-old 0.6±0.13 1.12±0.42

11 Page 47 of 53 Diabetes

6-month-old 0.91±0.55 1.41±0.77

High-fat diet* 4.2±2.32 15.32±4.02 *

Leptin (ng)

2-month-old 1.2±0.2 1.3±0.3

6-month-old 6.7±2.1 8.3±3.5

High-fat diet* 22.4±4.7 34.6±2.2 *

Adiponectin (µg/ml)

2-month-old 8.5±2 8.4±1.4

6-month-old 6.5±1.61 9±1.09

High-fat diet* 5.6±2.19 8.84±0.8

*12 Weeks of high-fat diet.

Suppl. Figure legends

Suppl. Figure 1: Id1 is dispensable for brown adipocyte differentiation. (A) Id1 and β-actin levels

in vector control and mouse-Id1 overexpressed (Id1-OE) HIB1B cells. (B) Oil-Red-O staining of day 8

differentiated control and Id1-OE HIB1B brown adipocytes. Scale bar 20µm. (C) Relative mRNA

transcript levels of indicated genes in control and Id1-OE day 8 differentiated HIB1B cells. Expression

data are mean ± SD; n=3. (D) Id1 and β-actin levels in scramble-shRNA or Id1-shRNA expressing

HIB1B cells. (E) Oil-Red-O staining of day 8 differentiated control or Id1-shRNA HIB1B adipocytes.

Scale bar 10µm. (F) Relative mRNA transcript levels of indicated genes in control and Id1-shRNA day

8 differentiated HIB1B cells. Expression data are mean ± SD; n=3. Oil Red O-stained cells were de-

stained with isopropanol and the amount of staining was quantified by reading absorbance at 510nm.

The staining intensity was not significantly different between control and Id1-shRNA cells, and between

control and Id1-OE cells.

12 Diabetes Page 48 of 53

Suppl. Figure 2: Lean mass, bone mass and BAT content are unchanged in aP2-Id1Tg+ mice. (A-B)

Percentage of lean mass (A) and fluid content (B) in 6-month old aP2-Id1Tg- and aP2-Id1Tg+ male mice

measured by LF90II BCA-Analyzer (n=5). (C) Bone mass in 6-month old aP2-Id1Tg- and aP2-Id1Tg+

male mice measured by Lunar PIXImus2 (n=5). (D-E) Interscapular BAT weight in 6 month old aP2-

Id1Tg- and aP2-Id1Tg+, and Id1+/+ and Id1-/- male mice (n=5). All the data are mean ± SD.

Suppl. Figure 3: Reduced energy expenditure in adult aP2-Id1Tg+ mice, and reduced expression of

Tg+ thermogenic genes in HFD-fed aP2-Id1 mice. (A-C) O2 consumption rate, CO2 release, heat

production, food consumption and physical activity in 6 month old aP2-Id1Tg- and aP2-Id1Tg+ male mice measured for 4 days using Oxymax/CLAMS animal monitoring system (n=3). Data are mean ± SD,

*p<0.01, ***p<0.0001. (D-E) Relative mRNA transcript levels of thermogenic and mitochondrial genes

in the BAT of aP2-Id1Tg- and aP2-Id1Tg+ male mice after 12 weeks of HFD. Data are mean ± SD, n=5;

*p<0.05, **p<0.005.

Suppl. Figure 4: Id1 partially suppresses Ebf2-mediated differentiation of C2C12 cells into brown

adipocytes. (A) Ebf2 protein levels in C2C12 and C2C12-Ebf2-expressing cells. (B) aP2 expression

levels in C2C12 and Ebf2-C2C12 cells at the indicated time points after inducing brown adipogenesis.

(C) Oil-Red-O staining pictures of day 4 differentiated control and Ebf2-C2C12 cells. Scale bar 20µm.

(D) Expression levels of Ebf2 and Id1 after co-expressing Id1 in Ebf2-expressing C2C12 cells. (E) Oil-

Red-O staining pictures of day 4 differentiated Ebf2-C2C12 and Ebf2/Id1-C2C12 cells. Scale bar 20µm.

(F-G) Relative mRNA transcript levels of indicated genes in day 0 and 4 differentiated Ebf2-C2C12 and

Ebf2/Id1-C2C12 cells. Data are mean ± SD, n=3; *p<0.05, **p<0.005. (H) Oil-Red-O staining pictures

of day 8 differentiated Ebf2-C2C12 and Ebf2/Id1-C2C12 cells. Scale bar 20µm. (I) Relative mRNA

transcript levels of indicated genes in day 8 differentiated Ebf2-C2C12 and Ebf2/Id1-C2C12 cells. Data

13 Page 49 of 53 Diabetes

are mean ± SD, n=3. (J) Western blot showing the expression patterns of Ebf2 and Id1 at the indicated

time points after inducing brown adipocyte differentiation in Ebf2-C2C12 and Ebf2/Id1-C2C12 cells.

(K) Co-IP followed by Western blot showing a direct interaction between Id1 and Ebf2 in C2C12 cells.

Input: 2% of IP reaction.

14 Diabetes Page 50 of 53 “Patil_Suppl._Fig1”

B A Day 8 Control Id1-OE

Id1

β-actin 2.77±0.3 3.5±0.4

2 C Day 8 Hib1B Hib1B + Id1-OE 1.5

1

0.5 Relative mRNA levels 0 PGC1aPGC1α Ucp1ucp1 PPARpparγ aP2ap2

E D Day 8 Control Id1-shRNA

Id1

β-actin 3.78±0.4 3.53±0.3

2 F Day 8 Hib1B 1.5 Hib1B + Id1-shRNA

1

0.5

Relative mRNA levels 0 PGC1α Ucp1 PPARγ aP2 Page 51 of 53 Diabetes “Patil_Suppl._Fig2”

aP2-Id1Tg- aP2-Id1Tg- aP2-Id1Tg- A B C aP2-Id1Tg+ aP2-Id1Tg+ aP2-Id1Tg+ 80 9 0.07 ) 8 2 0.06

7 /cm 60 0.05

6 gm 0.04 5 40

Fluid (%) Fluid 4 0.03 3 0.02 Lean mass (%)

20 ( mass Bone 2 0.01 1 0 0 0 1 1 1 E D aP2-Id1Tg- 600 Id1+/+ aP2-Id1Tg+ Id1-/- 600 500 500 400 400 300 BAT weight (mg) weight BAT

BAT weight (mg) weight BAT 300 200 200 100 100 Interscapular

Interscapular 0 0 1 1

- - - - 7 * * Tg

Tg Tg+ Tg+

Page 52 of 53 of 52 Page Id1 Id1 Id1 Id1 - - - - 3 * * Cox7a 6 aP2 aP2 aP2 * * aP2 5 **

2 * * 4 * * Cox5b “Patil_Suppl._Fig3” 3 -c -c 1 2 * * ** High fat diet (12 diet (12 weeks) High fat Cyt * * 1 E E 2 1 0 D 1.5 0.5

3 2 1 0 levels mRNA Relative

levels mRNA Relative Diabetes

Tg-

aP2-Id1

Tg+

- -

96h aP2-Id1

96h

96h Tg+ Tg 550 550 551

Id1 Id1

- - 496 489 489 1 n.s

aP2 aP2 441

428 428 C

386

367 367 0

331

306 306

80000 60000 40000 20000 (counts/animal/day)

276

245 245

activity Total

221

184 184

166

123 123 111

1

Release n.s 62 62

56 2 6 Month old mice old 6 Month

Consumption

B B 1 2 1 1 CO p<0.0001 p<0.01 Heat production Heat p<0.01 0 0 O 0 0 0 0

A 5 4 3 2 1 0

0

/day) gm ( 20 40 60

2000 5000 8000

2000 5000 8000

Kcal/Kg/ )

11000 hr

consumption Food

2 2

(ml/Kg/ VCO ) 11000 hr 2 2

(ml/Kg/ VO ) hr Page 53 of 53 Diabetes “Patil_Suppl._Fig4”

C2C12+Ebf2 A B C2C12 C2C12+Ebf2 C C2C12

0 2 4 6 0 2 4 6 Ebf2 aP2 Day 4 β-actin β-actin

D C2C12 Ebf2 Ebf2+Id1 Ebf2 H Id1 C2C12+Ebf2 J β-actin EBF2+ V EBF2+ Id1 0 2 4 6 8 0 2 4 6 8 days Ebf2 C2C12+Ebf2 C2C12+Ebf2+Id1 E (low)

Ebf2 (high)

Day 8 C2C12+Ebf2+Id1

Day 4 Id1

aP2

1.5 Day 0 β-actin F * 1

0.5 Ebf2/Id1 Ebf2 0 Ebf2-C2C12 Relative levels mRNA Relative K aP2 Prdm16 PGC1α UCP1 - ap2 prdm pgc ucp C2C12 I Ebf2/Id1-C2C12 - 2.5 Day 4 C2C12 2.5 Day 8 G 2 * 2

IB-Id1 1.5 ** ** * 1.5 1 1 0.5 0.5 IB-Ebf2 0 0 Relative levels mRNA Relative Relative levels mRNA Relative aP2 Prdm16 PGC1α UCP1 aP2 Prdm16 PGC1α UCP1 ap2 prdm pgc ucp