The role of secreted factors in white and brown adipogenesis

Ulm University Medical Center

Department of Pediatrics and Adolescent Medicine

Director: Prof. Dr. Klaus-Michael Debatin

Experimental Endocrinology and Metabolism Research Head: Prof. Dr. Pamela Fischer-Posovszky

The role of secreted factors in white and brown adipogenesis

Dissertation submitted in partial fulfillment of the requirements for the degree of Doctor rerum naturalium (Dr. rer. nat.) of the International Graduate School in Molecular Medicine Ulm

Daniel Halbgebauer

Waiblingen

2021

ii

Current Dean of the Graduate School: Prof. Dr. Knöll Thesis Advisory Committee First Supervisor: Prof. Dr. Fischer-Posovszky Second Supervisor: Prof. Dr. Jan Tuckermann Third Supervisor: PD Dr. Marcel Scheideler External Reviewer: Prof. Dr. Engeli Day Doctorate Awarded: 17.05.2021

iii

Parts of the results gained in my thesis have previously been published in the following

publication:

Browning capabilities of human primary adipose-derived stromal cells compared to SGBS cells.

Halbgebauer D, Dahlhaus M, Wabitsch M, Fischer-Posovszky P, Tews D.

Sci Rep. 2020;10(1):1-8.

doi:10.1038/s41598-020-64369-7

PMID: 32541826

This publication is an open access article distributed under the terms of the Creative

Commons CC BY license, which permits unrestricted use, distribution, and reproduction in

any medium, provided the original work is properly cited.

iv

Contents

1. Introduction ............................................................................................. 1

1.1 Obesity ..................................................................................................... 1

1.2 Adipose tissue .......................................................................................... 1

1.3 Brown adipose tissue thermogenesis ............................................................ 3

1.4 Relevance for brown adipose tissue in humans ......................................... 5

1.5 Browning of the white adipose tissue ........................................................ 6

1.6 Activation of brown adipose tissue and white adipocyte browning as a new

therapeutic approach for obesity .............................................................. 8

1.7 Transforming growth factor beta ............................................................ 10

1.8 Latent transforming growth factor beta binding proteins ......................... 13

1.9 Aim of the study ..................................................................................... 16

2. Material and Methods ........................................................................... 19

2.1 Materials ................................................................................................ 19

2.2 Methods................................................................................................. 37

3. Results ................................................................................................... 52

3.1 The role of LTBPs in adipogenesis and their impact on UCP1 expression in

vitro ....................................................................................................... 52

3.2 The role of TGFβ in SGBS adipogenesis .................................................... 65

3.3 TGFβ2 is reduced in LTBP2- and LTBP3-deficient cells .............................. 73

3.4 TGFβ2 deficiency leads to alterations in UCP1 and metabolic function ..... 73

3.5 TGFβ2 correlates significantly with LTBP3 and UCP1 expression in human

subcutaneous adipose tissue .................................................................. 76

4. Discussion .............................................................................................. 77

4.1 LTBP3 deficiency leads to a decrease of UCP1 expression in SGBS cells .... 77

v

4.2 Inhibition of the TGFβ pathway leads to a decrease of UCP1 in SGBS

adipocytes .............................................................................................. 81

4.3 Repression of brown marker gene UCP1 by LTBP3-deficiency could be

mediated by alterations in TGFβ2 bioavailability ........................................ 85

4.4 Indications for LTBP3-TGFβ2 mediated browning in human WAT ............. 86

4.5 Conclusion, limitations and outlook ........................................................ 88

5. Summary ............................................................................................... 90

6. References ............................................................................................. 92

7. Register of illustrations and figure permissions .................................... 115

Appendix I .......................................................................................................... 118

Appendix II ......................................................................................................... 123

Acknowledgements ............................................................................................ 126

Statutory declaration .......................................................................................... 126

Curriculum vitae ................................................................................................. 128

vi

List of Abbreviations

AC adenylyl cyclase

adip. adipocytes

ADIPOQ Adiponectin

Ant/Rot Antimycin A/Rotenone

ATP adenosine-triphosphate

BAT brown adipose tissue

Batokine brown adipose tissue adipokine

BMI Body mass index

BMP bone morphogenic protein

BSA bovine serum albumin

cAMP cyclic adenosine-monophosphate

Cas9 CRSIPR associated protein 9

CRISPR clustered regularly interspaced short palindromic repeats

CS citrate synthase

DMSO dimethyl sulfoxide

dn deep neck

DNA deoxyribonucleic acid

dNTP deoxyribonucleotide triphosphate

DPBS Dulbecco’s phosphate buffered saline

DTT dithiothreitol

ECL enhanced chemiluminescence

ECM extracellular matrix

EDTA ethylenediaminetetraacetic acid

ELISA enzyme-linked immunosorbent assay

EV empty vector

eWAT epididymal white adipose tissue

FCCP carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone

FCS fetal calf serum

FDA federal drug enforcement

vii

FGF21 fibroblast growth factor 21

GAPDH glycerinaldehyd-3-phosphat-dehydrogenase

GLUT4 glucose transporter type 4

hASC human adipose stromal cell

HCL hydrochloric acid

HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

HPRT hypoxanthine-guanine phosphoribosyltransferase

HRP horse radish peroxidase

HSL hormone sensitive lipase

iBAT interscapular brown adipose tissue

IBMX 3-isobutyl-1-methylxanthine

ingWAT inguinal white adipose tissue

ITR inverted repeat

kb kilo base pair

KD knockdown

kDa kilo Dalton

KO knockout

LAP latency associated protein

LLC large latent complex

LTBP latent transforming growth factor beta binding protein

mRNA messenger ribonucleic acid

MuLE Multiple lentiviral expression system

MuSE Multiple sleeping beauty expression system

Myf5 myogenic factor 5

NTC non-targeting control

Oligo Oligomycin

PCR polymerase chain reaction

PGC1α PPARγ coactivator 1-alpha

pgWAT perigonadal white adipose tissue

PKA protein kinase A

viii

PPARγ peroxisome proliferator-activated receptor gamma

pre. preadipocytes

Puro Puromycin

qPCR quantitative polymerase chain reaction

rpm revolutions per minute

RT-PCR reverse-transcription polymerase chain reaction

SB sleeping beauty

sc subcutaneous

SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis

SGBS Simpson-Golabi-Behmel syndrome

sgRNA single guide RNA

siRNA small interfering ribonucleic acid

SLC small latent complex

SMAD “small” mothers against decapentaplegic (Drosophila)

TBS tris buffered saline

TGFβ transforming growth factor beta

TGFβR transforming growth factor beta receptor

TZD thiazolidinediones

UCP1 uncoupling protein 1

vs versus

WAT white adipose tissue

WHO world health organization

WT wild type

1

1. Introduction

1.1 Obesity

Obesity is defined by the world health organization (WHO) as a disease with “abnormal or

excessive fat accumulation that may impair health” [1]. Overweight or obesity can be

measured with the body mass index (BMI), which is calculated by the weight for height in

kg/m². Obese people have a BMI greater or equal 30 kg/m2, whereas overweight people

have a BMI equal or above 25 kg/m2 [1].

Obesity and its associated metabolic disorders are rapidly increasing worldwide [1]. In the

last 40 years the prevalence for obesity has nearly been tripled. In 2016, 650 million adult

people were obese and an additional 1.25 billion people were overweight. Obesity brings

along metabolic disorders like type 2 diabetes mellitus or cardiovascular diseases, which

are the number one cause for death worldwide [2]. Furthermore, childhood obesity is

associated with a higher chance of premature death, obesity and disability in adulthood. In

2018, 40 million of children under the age of 5 years were already overweight or obese [1].

From a physical perspective, obesity is based on an imbalance of energy intake and

expenditure, whereby high amounts of visceral adipose tissue are associated with

metabolic disorders such as cardiovascular diseases and diabetes type 2 [3,4]. The current

methods for the prevention of obesity and its treatment fails in the long term [3].

Approaches aiming at reducing energy intake or increasing energy expenditure through

exercise often fail due to low discipline and methods like bariatric surgery are not available

for the majority of people due to insufficient healthcare or are just not affordable [3]. All

these facts reveal obesity as a challenging pandemic disease which takes great influence

on our society and healthcare system.

1.2 Adipose tissue

The adipose tissue is a connective tissue which inter alia consists of pre- and adipocytes,

although macrophages, endothelial cells, fibroblasts and leukocytes can also be found [5].

There are two distinct types of adipose tissues in mammals [5]. The white adipose tissue

(WAT) and the brown adipose tissue (BAT). White adipocytes are mainly distributed in the

2

subcutaneous (sc) and visceral adipose tissue and can be recognized by an unilocular lipid

droplet (Figure 1) [5]. Brown adipocytes are histologically distinguishable from white

adipocytes by their large number of lipid vacuoles and a high density of mitochondria.

Furthermore, brown adipocytes express the specific BAT marker uncoupling protein 1

(UCP1) [5].

BAT can be found in the deep neck, supraclavicular, axillary, suprarenal, paravertebral, and

peri-aortic region of humans [6–9]. Of note, there is a third type of adipose tissue, the beige

adipose tissue which arises from WAT browning through cold exposure or β3 adrenergic

stimulation [10,11]. However, without these stimuli, browning is also reversable, which is

often referred as whitening [10]. Comparable to classical brown fat cells, beige adipocytes

are also rich in mitochondria and express UCP1, but they have a distinct gene expression

profile [10].

Figure 1: Morphology of different types of adipocytes. Left: White adipocyte with a unilocular lipid droplet. Due to certain stimuli like cold exposure or β3 adrenergic stimulation white adipocytes can undergo browning into beige adipocytes (middle), which have higher amounts of mitochondria and multilocular lipid droplets. However, browning is also reversable and often referred as whitening. Right: The morphology of brown adipocytes is characterized by small multilocular lipid droplets and high amounts of mitochondria.

One important role of the WAT is the energy storage and supply of the body with free fatty

acids and glycerol in case of need [12]. White adipocytes store surplus energy of the body

in form of triglycerides and play therefore a crucial role in energy homeostasis and energy

storage [4]. Of note, WAT is also an important endocrine organ which secretes adipokines

acting on other organs like liver and muscle [4,5]. Adiponectin and leptin are just two

important adipokines to mention [5]. Both are known to modulate metabolic pathways in

peripheral tissues. BAT accounts only for about 4% of the total body fat mass in adults,

whereby WAT is making up the other 96% [4]. In contrast to WAT, the primary function of

BAT is not to store energy in form of triglycerides, but to produce heat by oxidizing fatty

3

acids and glucose from intracellular energy storages and the circulation [12,13]. This

enables survival during nightly cold periods in winter as well as during cold stress after

childbirth and possibly offers advantages with a reduced supply of macronutrients [13].

Furthermore, BAT plays an important role in thermoregulation of human newborns [14].

In mice, BAT is the key organ regulating body temperature and energy homeostasis [13].

These properties are attributed to the expression and activation of UCP1, the key factor of

thermogenic capacity [13,15]. Importantly, recent findings suggest that BAT is relevant for

adult humans as well [8]. For a long time, BAT was thought to be only present in small

mammals and human infants, until 2007, when four independent studies revealed

evidence for BAT in human adults [6–9]. Compared to WAT there is just a small amount of

BAT in humans, but it has the potential to have a high impact on thermoregulation and

increasing energy expenditure [16–18].

1.3 Brown adipose tissue thermogenesis

Non-shivering thermogenesis occurs in BAT depots and strongly depends on the activation

of BAT [13,19]. It has been shown that non-shivering thermogenesis is dependent on UCP1,

which is an intermembrane protein residing in the inner mitochondrial membrane of brown

and beige adipocytes [13,20]. UCP1-deficient mice are not able to keep up their body

temperature during cold exposure, revealing the high impact of UCP1 on thermoregulation

[21]. A cold environment stimulates the activation of thermoreceptors on human skin,

which send afferent signals to the hypothalamus which leads to a release of norepinephrine

in sympathetic nerve fibers within the BAT [20,22]. Subsequently, norepinephrine binds to

β3-adrenoreceptors on the adipocytes surface (Figure 2), activating adenylyl cyclase (AC).

This activation triggers the production of cyclic adenosine monophosphate (cAMP), leading

to the activation of protein kinase A (PKA) which further phosphorylates the hormone

sensitive lipase (HSL) and perilipin A, followed by the hydrolysis of triglycerides into glycerol

and free fatty acids which are transported into mitochondria. These free fatty acids change

the conformation of UCP1, leading to its activation. UCP1 then uncouples the proton

motive force by transporting protons through the inner mitochondrial membrane into the

mitochondrial matrix. The energy of the cellular oxidation is then released as heat. The

activity of UCP1 is generally inhibited by purine nucleotides [22].

4

Under UCP1-inactivated conditions, the respiratory chain in the mitochondrial membrane

generates a proton gradient to produce adenosine-triphosphate (ATP). The ATP synthase

uses this proton motive force to recycle adenosine diphosphate (ADP) to ATP. If UCP1 is

activated by free fatty acids, the proton motive force is uncoupled from ATP generation. In

consequence, the cell compensates for the energy loss by increasing its respiratory chain

activity. As a result, the oxygen consumption is increased and more energy supply from the

citric acid cycle and β-oxidation is needed. Upon chronic activation, PKA leads to induction

and stabilization of UCP1 mRNA expression in brown and beige adipose tissue via

phosphorylation of cAMP-response-element-binding-protein (CREB) and p38 [23,24].

Figure 2: Molecular mechanism of BAT activation: Norepinephrine (NE) or other β3-agonists bind to the β3 adrenoreceptor on the adipocytes surface [22]. This activates the Gsα subunit of the G-protein Gs (Gs) which leads to cAMP synthesis by the adenylate cyclase (AC). Subsequent the protein kinase A (PKA) is activated, accompanied by the phosphorylation of hormone sensitive lipase (HSL), followed by the cleavage of triglycerides into free fatty acids and glycerin. These free fatty acids are transported into the mitochondria and activate the uncoupling protein 1 (UCP1). UCP1 sits in the inner mitochondrial membrane uncoupling the proton motive force by transporting protons through the membrane. The energy of the cellular oxidation is then released as heat. At the same time the cell tries to maintain adenosine triphosphate (ATP) production, thereby increasing oxygen consumption and demanding more energy from the citric acid cycle and β-oxidation. Figure 2 is taken from P.G. Crichton, Y. Lee, E.R.S. Kunji, The molecular features of uncoupling protein 1 support a conventional mitochondrial carrier-like mechanism, Biochimie. 134 (2017) 35–50. https://doi.org/10.1016/j.biochi.2016.12.016 with permission of Elsevier.

5

1.4 Relevance for brown adipose tissue in humans

It has been shown in several rodent studies that activation of BAT has significant positive

effects on metabolic health [10,25,26]. Rodents with chronically activated BAT were

resistant to high fat diet, had lower blood glucose levels and higher insulin sensitivity

[10,27]. Activation of BAT in rodents was achieved by exposing the animals either to chronic

cold or treatment with β3 adrenoreceptor agonists.

In humans, BAT was first described in the interscapular region of fetuses and newborns

[27,28]. Until recently, it had been assumed that human adults are mostly devoid of

functional BAT. However, the functionality was firstly described between 2007-2009 in four

independent studies, where it was shown that biopsies taken from the deep neck and

supraclavicular adipose tissue of human subjects are UCP1-positive [6–9]. Interestingly, the

BAT mass correlated negatively with the BMI as well as percent-body fat mass in humans,

suggesting that BAT is a regulator of body weight [6]. Further studies revealed that these

BAT depots could be activated by cold as it was shown before in rodents [8,9,29]. Subjects

exposed to acute cold showed a higher body temperature in the supraclavicular region and

increased glucose and fatty acid uptake [6,9,30,31]. In addition, subjects with higher BAT

mass had a significant overall higher glucose uptake rate, higher insulin sensitivity, and

higher basal metabolic rate [17,29,32]. Active BAT is associated with a higher overall energy

expenditure in humans and can therefore influence the energy homeostasis [8,9].

Interestingly, cold or β3-adrenergic stimulation led also to a significant increase in BAT mass

and UCP1 expression in humans and rodents [10,26].

However, it has also been shown that people lose BAT mass during ageing and that obese

people have little to no BAT [30,31,33,34]. On the other hand, obese people without active

BAT could restore its activity after weight loss due to bariatric surgery [33]. Moreover, cold

acclimation in obese subjects led to an reappearance of BAT [35]. This indicates that BAT

can still be recruited in adults, suggesting a relevant function.

6

1.5 Browning of the white adipose tissue

Browning is defined by a significant increase in UCP1 expression in adipocytes within the

WAT depots and was first described in 1984 in rodents [11,26]. Through chronic cold

exposure or β-adrenergic stimulation, adipocytes within the WAT of rodents phenocopy

brown adipocytes and are therefore referred to as beige adipocytes [12,13,26]. Many

factors have been discovered so far, known to stimulate white adipocyte browning in

rodents or in vitro. Activation of transcription factors like peroxisome proliferator activator

gamma coactivator 1 alpha (PGC1α) and PR domain containing protein 16 (PRDM16) are

known to trigger browning in white adipocytes. Furthermore, stimulation with extracellular

factors like bone morphogenic proteins (BMP), thyroid hormones, transforming growth

factor beta (TGFβ), and fibroblast growth factor 21 (FGF21) as well as natriuretic peptides

have been shown amongst others to regulate browning [36–41].

It is still controversially discussed if beige adipocytes occur in WAT due to

transdifferentiation of white adipocytes or de novo differentiation [42–44]. Currently there

are two different models discussed which are based on Cre-labeling studies in mice [42].

Myogenic factor 5 (Myf5) positive cells are known to be the precursor cells of the myogenic

lineages. In the transdifferentiation model, myocytes, brown adipocytes and white

adipocytes share the same origin, a multipotent Myf5 and Pax3-positive precursor cell,

whereas PRDM16 controls the switch from myogenic to adipogenic differentiation [42,45].

Browning occurs then through transdifferentiation of mature white adipocytes. But it is still

assumable that different lineages express Myf5 and Pax3 independently of each other [42].

The other model suggests that white and beige adipocytes have a common origin. Cells

labeled for Prx1, a transcription factor selective for precursor cells of the scWAT and not

for BAT or visceral WAT, revealed that beige adipocytes could emerge from the same

precursors of sc WAT by de novo differentiation [42]. In rodents one can distinguish

between the classical WAT, BAT, and beige adipose tissue [42]. Several marker genes have

been identified, which were differently expressed between these types of adipose tissue

[42]. In addition, the interscapular brown (iBAT) and the inguinal WAT (iWAT) are highly

temperature sensitive and under housing and cold conditions one will observe white

adipocyte browning in iWAT and distinct differences in morphology between the iBAT,

iWAT, and perigonadal WAT (pgWAT) (Figure 3). In humans, the composition of adipose

tissue found in the neck shows high inter-individual heterogeneity, which makes it much

7

more complicated to distinguish between the adipose tissue types [46,47]. UCP1-positive

cells referred as brown adipocytes can be found most likely in the deep neck and

supraclavicular region, but one will also find UCP1-negative cells within this tissue area. The

deeper the adipose tissue region in the neck of humans, the more classical BAT was found

[48]. In some studies, human brown adipocytes expressed markers, which fit to beige

markers of rodents [49–51]. On contrary to that, other studies found similar expression

patterns of human BAT to classical BAT of rodents [47,52,53].

These controversial results often lead to misunderstanding in the nomenclature in the

context of browning, beige adipose tissue, BAT, and WAT [54]. As previously mentioned,

browning is originally defined by an increase of UCP1 expression in the adipose tissue

depots [26]. Therefore, the term browning includes both, the de novo differentiation and

the transdifferentiation model. In the following, the term browning is used, when speaking

of a shift from a white towards a beige or brown phenotype within the murine or human

WAT. In this thesis, only the de novo differentiation model was considered for the in vitro

experiments. As mentioned above, it is still not clear whether UCP1-expressing adipocytes

from human WAT are beige or brown cells. Since this thesis focused on human adipose

stromal cells (hASCs) from the sc WAT, terms such as “an adipogenesis towards a brown

adipocyte phenotype” are used when a significant alteration in UCP1 expression was

observed in these cells.

Even if one cannot clearly distinguish between a beige and white adipose tissue depot in

humans as it is seen in rodents under cold conditions, there is evidence for browning of

white adipocytes in human sc adipose tissue [49,55,56]. The treatment with thyroid

hormones in a subject suffering from papillary thyroid carcinoma led to the appearance of

BAT marker genes in the sc tissue [57]. Furthermore, in patients suffering from

pheochromocytoma, UCP1-positive cells were found in the omental adipose tissue [44]. It

is assumed that high catecholamine levels secreted by the tumors induce browning in the

WAT [44]. This was also seen in patients with severe adrenergic stress [58]. Increased

mitochondrial density and the expression of UCP1 was observed in the sc WAT of patients

suffering from burn injuries [58]. In another study sc adipose tissue biopsies of individual

subjects were taken during summer and winter [55]. Interestingly, lean subjects had a

significantly higher UCP1 expression during winter compared to summer, which suggests a

potential for the induction of browning in human WAT [55].

8

Figure 3: Distribution of adipose tissue and its temperature sensitivity in mice. Under standard housing condition one can clearly distinguish between the different adipose tissue types in mice, shown by hematoxylin and eosin staining. It becomes even clearer after cold acclimation at 6°C. Multilocular lipid droplets are observed in the interscapular brown adipose tissue (iBAT), some in the inguinal white adipose tissue (ingWAT) and none in the perigonadal WAT (pgWAT). Figure 3 is taken from J. Sanchez-Gurmaches, C.M. Hung, D.A. Guertin, Emerging Complexities in Adipocyte Origins and Identity, Trends Cell Biol. 26 (2016) 313–326. https://doi.org/10.1016/j.tcb.2016.01.004 in agreement with Elsevier.

1.6 Activation of brown adipose tissue and white adipocyte browning as a new

therapeutic approach for obesity

As current obesity treatment methods often fail in the long term, new approaches for

weight loss should be found [59]. Since BAT activation in humans and browning of WAT in

rodents showed positive effects on metabolic health and body weight, the topic came into

focus as a new treatment approach for obesity [16,25,60,61]. Increasing the amount of

energy expenditure in obese patients by activating BAT or browning would thus lead to a

negative energy balance, weight loss, and improvements of obesity-related metabolic

disorders like diabetes or arteriosclerosis.

Several studies investigated the impact of BAT on overall metabolic health in humans over

time [18,55,62–65]. Human subjects were either exposed to chronic mild cold or

investigated during summers and winters. Cold exposition did not only lead to a higher BAT

activation but also to an increase in total BAT mass, a higher loss of fat mass, and a higher

glucose uptake rate compared to control groups [18,62–64]. However, a major problem is

9

that obese patients have almost none to no brown adipose tissue. It was shown that the

positive effects of BAT activation on the metabolism by cold stimulation were blunted in

obese individuals [30]. Several pharmacological, physiological, and nutritional compounds

can activate BAT or induce browning in rodents or UCP1 in vitro in cultured human

adipocytes but they were not tested or failed to activate BAT in humans so far

[26,27,61,66]. Different β3 agonists failed because of their cardiovascular side effects by

binding also to β1 and β2 receptors in humans [39,67–70]. Moreover, ephedrine failed to

activate BAT in another study with human subjects [39]. Nevertheless, it was shown

recently that human BAT activation can be induced by a selective β3 receptor stimulation

with the FDA-approved drug Mirabegron [71–73]. Subjects which were treated with

200 mg Mirabegron showed a significant increase in BAT activity and an increase of

200 Kcal of resting energy expenditure per day [71]. Furthermore, chronical Mirabegron

treatment increased human BAT as well as insulin sensitivity, and led to an increase of

brown marker genes in the sc adipose tissue of individuals [72,73]. However, besides these

positive effects, subjects had also a higher blood pressure and a higher heart rate.

Pharmacological induction of browning in human WAT in order to increase energy

expenditure represents another approach to target obesity.

One approach is the treatment with thyroid hormones since they have already been shown

to induce browning in a human patient [57]. Furthermore, triiodothyronine induced UCP1

expression in the ingWAT of rodents and activation of the thyroid receptor led to browning

and increased energy expenditure in mice [74,75]. However, thyroid hormones have similar

side effects as β3-receptor agonists, namely a higher risk for heart failure and hyperthermia

[76]. Another group of pharmacological compounds with high browning capacity are the

thiazolidinediones (TZD), which are strong peroxisome proliferator activator gamma

(PPARγ) agonists. PPARγ is the key regulator of adipogenesis [77]. Stimulation with TZDs

induced UCP1 in vitro and in vivo [78,79]. Approved TZDs have been already used for

treatment of type 2 diabetes but some of them have been banned by the FDA due to severe

side effects like heart failure, weight gain and bone loss [80,81]. A promising browning

therapeutic is FGF21. Mice treated with FGF21 showed a significant increase in UCP1

expression in the ingWAT [82]. Moreover, FGF21 null mice showed a significant lower

thermoregulation and an impaired response to cold. FGF21 is currently under investigation

in clinical trials and FGF21 analogs (LY2405319 and PF-05231023) have already been shown

10

to decrease body weight and improve lipid profile in human obese subjects or subjects with

type 2 diabetes after 25-28 days of treatment [83,84]. However, there are no studies on

long term treatment in humans yet. Furthermore, mice treated with FGF21 or

overexpressed FGF21 showed severe bone loss [82]. A significant reduction in markers for

bone formation was also obtained in humans treated with PF-05231023, which doubts

FGF21 as a suitable drug for overweight or its related metabolic disorders [84].

To date, no pharmaceuticals have been described in clinical studies that specifically induce

browning in human WAT without severe side effects. New factors should be found for

future pharmacological targeting to trigger WAT browning or an increase of BAT mass in

obese humans and humans with related metabolic disorders.

New approaches try to target the origins of the white or brown/beige adipocytes. Among

other theories it is assumed that white and beige and even brown adipocytes have the

same precursors, as all these cell types have a mesenchymal stem cell background and

different factors can control the direction of differentiation [42]. Therefore, several groups

isolated stromal cells from the deep neck adipose tissue and/or sc adipose tissue of humans

and differentiated these cells in vitro [47,48,52,85,86]. Interestingly, differentiated cells

derived from the deep neck (dn) or supraclavicular region showed a beige or brown

phenotype, whereas cells from the sc tissue a white phenotype. Furthermore, to find

potential targets, the expression patterns of the isolated hASCs were analyzed and

compared [47,85,86]. This new approach revealed several targets which may be involved

or important for the stimulation of an adipogenesis towards a brown adipocyte phenotype.

Especially adipokines are of interest since it was already shown that secreted factors

expressed in ASCs and adipocytes of the BAT (“batokines”) stimulate the thermogenic

activity in an autocrine or paracrine manner [87]. For example, the TGFβ isoforms are

secreted factors and have already been shown to either stimulate or inhibit UCP1

expression in vivo and in vitro [36,88].

1.7 Transforming growth factor beta

All TGFβ isoforms, i.e. TGFβ1, 2 and 3 are expressed in mammals and can be found in all

tissues and cell types [89,90]. The active isoforms are homodimeric peptides activating the

SMAD2/3 pathway [91]. The active TGFβ1 and TGFβ3 dimer bind to phosphorylated TGFβ

11

receptor2 (TGFβR2) on the cell surface, which then leads to recruitment and

transphosphorylation of the TGFβR1 (Figure 4) [91]. The activated TGFβR1 subsequently

phosphorylates SMAD2/3 intracellularly. This leads to binding of the Co-SMAD (SMAD4) to

phosphorylated SMAD2/SMAD3 which localize as a complex into the nucleus, regulating

gene expression. However, the TGFβ2 isoform has low affinity for TGFβR2, therefore using

a coreceptor for SMAD activation [91]. The peptide binds at first to the TGFβR3, which then

activates TGFβR2 and subsequently TGFβR1, which leads as well to SMAD2/3 pathway

activation.

Figure 4: TGFβ-SMAD signaling pathway The dimeric transforming growth factor β (TGFβ) peptide binds to the autophosphorylated TGFβ receptor2 (TGFβRII). This leads to the recruitment of TGFβ receptor 1 (TβRI), forming a heterodimeric complex. TβRI is then transphosphorylated by TβRII, activating it. Thereupon, the TβRI kinase phosphorylates SMAD2 and SMAD3 which activates intracellular signaling. Phosphorylated SMAD2/3 binds then to SMAD4, enabling the internalization into the nucleus, where they interact with other transcription factors, repressing or activating certain gene expression. Figure 4 is taken from E.H. Budi, D. Duan, R. Derynck, Transforming Growth Factor-β Receptors and Smads: Regulatory Complexity and Functional Versatility, Trends Cell Biol. 27 (2017) 658–672. https://doi.org/10.1016/j.tcb.2017.04.005 in agreement with AAA

The TGFβ isoforms control several vital processes, such as embryonic development,

immune regulation, and inflammation [89]. Furthermore, these cytokines exert their

regulatory role in a bifunctional manner, either inhibiting or stimulating cell proliferation

and differentiation, dependent on the cell type [89]. Knockout mouse models revealed vital

functions of all three isoforms in vivo. TGFβ1-deficient mice died within 3 weeks after birth

due to severe inflammation [92,93]. On the other hand, TGFβ2 null mice had

underdeveloped organs and died perinatally [94]. Moreover, TGFβ3-deficient mice died

directly after birth due to lung dysfunction [95]. These results indicate a crucial role for

TGFβ2 and TGFβ3 in early development and an important role for TGFβ1 in early immune

response.

12

1.7.1 Transforming growth factor beta in the context of adipose tissue

High circulating TGFβ levels are associated with obesity in mice and humans [96,97]. In

human, there is a positive correlation between TGFβ1 and BMI [97]. In addition, overweight

and obese individuals had significantly higher TGFβ1 plasma levels compared to lean [97].

In leptin deficient ob/ob mice, TGFβ is expressed 2.3 fold higher in the epididymal WAT

(eWAT) compared to WT mice [98]. Furthermore, the levels of phosphorylated SMAD3

(pSMAD3) was even 8-fold higher in ob/ob eWAT compared to WT [98].

Early on it was described that TGFβ1 plays a role in proliferation and subsequent adipogenic

differentiation. It inhibits differentiation in murine 3T3-L1 cells, murine brown

preadipocytes and in hASCs [36,96,98,99]. Of note, several studies suggested different

mechanisms of the TGFβ/SMAD pathway influencing the adipogenesis. In NIH3T3

adipocytes activated SMAD3 was bound to CCAAT/enhancer-binding protein beta (C/EBPβ)

and C/EBPδ, thereby inhibiting their activity as a transcription factor for PPARγ expression

and adipogenesis [100]. A study by Yadav et al. revealed new insights on the mechanism of

TGFβ in adipogenesis [97]. They showed that phosphorylated SMAD3 binds to the PGC1α

promoter in the nucleus, inhibiting its expression [97]. This suggests also a role for TGFβ in

browning since PGC1α is known to be a regulator of PPARγ and an inducer of brown marker

genes including UCP1 [101]. Indeed, browning and an increased mitochondrial biogenesis

was obtained in the WAT of SMAD3-deficient mice and in mice treated with an anti TGFβ

1,2,3 antibody [97,98]. Moreover, heterozygous TGFβR1-deficient mice had significant

browning of the eWAT compared to WT mice [102]. Interestingly, SMAD3 null mice also

had improved insulin sensitivity on high fat diet (HFD) and smaller adipocyte size compared

to WT. In addition, mice lacking SMAD3 were protected from insulin resistance and obesity

on HFD [103]. These mice had improved glucose tolerance and insulin sensitivity. The

blockage of the TGFβ pathway with a TGFβ 1,2,3 antibody in ob/ob and diet-induced

obesity mice (DIO) resulted in the protection from obesity and type 2 diabetes [97].

A special property in connection with adipose tissue was shown for TGFβ2 [88]. Mice

infused with recombinant TGFβ2 by an osmotic pump were resistant to high fat diet and

showed a higher glucose tolerance and insulin sensitivity [88]. Furthermore, TGFβ2

increased the UCP1 expression in murine brown adipose tissue and in in vitro differentiated

human brown adipocytes [88]. This was not true for TGFβ1 and TGFβ3. A specific pathway

for TGFβ2 on adipocytes was not discussed and is still unknown.

13

TGFβ bioavailability and signaling has an impact on adipogenic differentiation and

browning of WAT. However, the bioavailability is strongly dependent on the presence of

the latent TGFβ binding protein (LTBP) [104]. TGFβ1-3 are expressed as inactive pro-

peptides, which are secreted in a complex with LTBP, which is known as the large latent

complex (LLC). The importance of LTBP for TGFβ bioavailability was shown in a study with

mice, where a cysteine of the TGFβ1 propeptide, which is crucial for LTBP binding, was

substituted by a serine [105]. These mice showed comparable phenotypes to TGFβ1

knockout mice, inter alia a shorter lifespan. The strong dependency of TGFβ activity on

LTBP suggests that LTBPs have an impact on the generation of a white or brown adipocyte

phenotype.

1.8 Latent transforming growth factor beta binding proteins

LTBPs are crucial for TGFβ bioavailability and play an important role in extracellular matrix

(ECM) modulation and stabilization [104,106,107]. In humans and rodents there are four

members of the LTBP family: LTBP1, LTBP2, LTBP3, and LTBP4. All of them are known to

interact with fibrillin in microfibrils [104]. LTBPs are expressed in a variety of cell types but

mostly in cells with a mesenchymal origin, like the adipogenic, osteogenic, or myogenic

lineages [108]. Structurally, all LTBPs contain calcium binding epidermal growth factor

domains and 8-Cys type domains, some of which are TGFβ-binding domains [104,106]. It

was demonstrated that LTBPs can bind intracellularly to TGFβ, by forming disulfide bridges

with the inactive propeptide of TGFβ, also called latency-associated peptide (LAP) (Figure 5)

[107,109]. When bound to LTBP, the mature TGFβ is cleaved off from its pro-form by furin

in the Golgi vesicle, but still bound by disulfide bridges to it (small latent complex, SLC). The

LTBP-SLC complex, called LLC is then secreted into the ECM where it incorporates by

binding to fibrillin and/or fibronectin [104]. LTBP keeps the mature TGFβ in a latent state

until proteases, integrins, or pH change leads to a release of the mature active TGFβ from

the LLC. Although structurally similar, the LTBP isoforms have also different biochemical

and functional properties [104,110]. Binding studies revealed that the affinity of LTBPs to

TGFβ is different between the isoforms. Only LTBP1 and LTBP3 are able to bind TGFβ1,

TGFβ2 and TGFβ3. LTBP4 can only bind the TGFβ1 isoform and LTBP2 is not able to bind

14

any of them [110] (Table 1). Until now there is no data published on any connection or

correlation of LTBPs with obesity.

Figure 5: Secretion of the large latent complex (LLC). The transforming growth factor beta (TGFβ) propeptide binds intracellular to the latent TGFβ binding protein (LTBP). This leads to structural changes and the mature TGFβ dimer is cleaved from its propeptide by furin in the Golgi-vesicle. However, the mature TGFβ is still bound by disulfide bridges to its propeptide, called small latent complex (SLC). The whole LTBP-SLC, also known as large latent complex (LLC), is then secreted into the extracellular matrix (ECM). The Mature TGFβ is then released by proteases, integrins, or pH change and can bind to its receptor. Figure 5 is taken from V. Todorovic, D.B. Rifkin, LTBPs, more than just an escort service, J. Cell. Biochem. 113 (2012) 410–418. https://doi.org/10.1002/jcb.23385 in agreement with John Wiley and Sons.

1.8.1 LTBP1

Among all LTBP isoforms, LTBP1 is the best studied member of this family. It binds to all

three TGFβ isoforms and incorporates into the matrix by binding with its N-terminal end to

fibrillin [104]. In contrast to the other LTBP members, LTBP1 is also able to bind to

fibronectin with its C-terminal end, which suggests an important role of fibrillin-fibronectin

linkage [111,112]. So far, integrin-mediated TGFβ activation was only shown for LTBP1

[113]. In the ECM, integrin can bind to the SLC and exerts force on LTBP1, which then

releases the mature TGFβ from the LLC [104,114]. However, for integrin-mediated TGFβ

release, the incorporation of LTBP1 into the matrix and binding to fibronectin is necessary,

15

since tension between the integrin and LTBP1 is needed [114]. Furthermore, in vitro studies

with antibodies or siRNA against LTBP1 revealed that LTBP1 is important for the

differentiation of endothelial cells [115]. The knockout of the long LTBP1 isoform in mice

resulted in lethality after birth due to persistent truncus arteriosus, which can be attributed

to a low TGFβ bioavailability [116,117]. On the other hand, the knockout of the short LTBP1

form had no effect on the phenotype in mice.

1.8.2 LTBP2

LTBP2 is the only LTBP member which cannot bind to any TGFβ isoform [104,110].

However, LTBP2-deficient mice revealed important functions for LTBP2 in microfibril

formation in the ciliary zones [118]. In addition, humans with LTBP2 null-mutations suffer

from primary congenital glaucoma [119,120]. It is suggested that LTBP2 plays a crucial role

for ECM modulation. LTBP2 binds not only to fibrillin in the matrix, but also to fibulin-5

which supports elastin fiber assembly [121,122]. Furthermore, it was shown that LTBP2 can

inhibit the maturation of BMP11 and secretion of myostatin intracellularly, thereby

influencing the bioavailability of these TGFβ superfamily proteins [123,124].

1.8.3 LTBP3

In contrast to the other LTBP isoforms, LTBP3 is only secreted if bound to the LAP of the

TGFβ isoforms, suggesting that LTBP3 does not have any TGFβ-independent functions

[125]. Mice lacking LTBP3 had premature ossification of the skull base or increased bone

density indicating a TGFβ deficiency which was verified by a lower SMAD3 phosphorylation

in the tissue [126,127]. Moreover, it was also shown that LTBP3 inhibits secretion of pro-

myostatin and maturation of pro-BMP11 by binding to the proprotein convertase 5/6A in

the endoplasmic reticulum [123].

1.8.4 LTBP4

LTBP4 has only low binding affinity for TGFβ1 [104,110]. However, LTBP4-deficient mice

developed cardiomyopathies and rectal-anal tumors which are associated with low levels

of TGFβ1 [128]. In addition, LTBP4 is assigned a role independent of TGFβ1. Studies

16

revealed that LTBP4 plays a crucial role in elastogenesis [129]. LTBP4 enhances fibulin 5

deposition onto microfibrils, thereby potentially building a bridge between fibrillin and

fibulin [104]. This integrates elastin into microfibrils which forms elastic fibers. The

knockout of LTBP4 showed an impaired elastogenesis due to insufficient binding of fibulin

to microfibrils [129,130].

Table 1: In vivo knockout phenotype and transforming growth factor β (TGFβ) binding properties of the latent TGFβ binding protein (LTBP) isoforms.

LTBP1 LTBP2 LTBP3 LTBP4

Knockout phenotype in vivo

▪ Lethality

▪ Persistent truncus arteriosus

▪ Lens luxation

▪ Increased bone density ▪ Premature ossification

of the skull base

▪ Cardiomyopathies

▪ Rectal-anal tumors

TGFβ binding

▪ TGFβ1

▪ TGFβ2

▪ TGFβ3

▪ None ▪ TGFβ1

▪ TGFβ2

▪ TGFβ3

▪ TGFβ1

1.9 Aim of the study

Obesity is a severe problem of the western civilization, since it leads to several

cardiovascular diseases, which are the number one cause of death worldwide [1,2]. Until

now obesity treatment methods are very limited and often fail in the long term [3].

There is evidence for the presence of functional BAT in human adults [6–9]. Browning and

the activation of BAT have been shown to significant increase overall metabolic health

among others by higher glucose uptake and fatty acid uptake [17,29,32]. Recent clinical

studies suggest that prolonged cold exposure leads to recruitment of BAT accompanied by

body weight reduction and improvement of obesity-associated metabolic diseases like type

2 diabetes [18,64,65,71]. Therefore, activation of BAT and stimulation of WAT browning

has become a new approach as an anti-obesity treatment [16,25,60,61].

Until now the exact mechanism how ASCs differentiate either into brown/beige or white

adipocytes is not completely understood. Thus, identification of factors driving these

distinct types of adipogenesis is mandatory and could lead to the development of

therapeutic targets.

In a previous study by our group (Tews et. al) hASCs of paired adipose tissue samples from

the sc and the dn region were isolated from patients undergoing neck surgery for

17

malignancies or nodular goiter [47]. These cells were ex vivo differentiated using identical

differentiation protocols and the expression of white and brown marker genes was

analyzed. Of note, adipocytes differentiated from cells derived from the dn region

expressed higher levels of brown marker genes as compared to adipocytes differentiated

from sc cells [47]. This was supported by other studies, which could also demonstrate that

stromal cells isolated from different adipose tissue regions reveal differences in function

and gene expression after in vitro differentiation [48,52,85]. To identify novel targets that

are potentially involved in the generation of a brown adipocyte phenotype, the gene

expression of the isolated hASCs was analyzed via microarray [47].

It was shown that adipokines can act in an autocrine or paracrine manner to stimulate or

inhibit the expression of brown marker gene UCP1 in vitro and in vivo [87]. Based on these

findings and the introduced results by Tews et al. we hypothesize that factors differently

secreted in the respective adipose tissues may play a role in stimulating or inhibiting an

adipogenesis towards a brown adipocyte phenotype [47]. Therefore, the focus in this work

was on secreted factors as potential targets to stimulate the adipogenesis towards a brown

phenotype. The microarray analysis by Tews et al. revealed 422 genes which were

significantly different expressed between dn and sc samples [47]. These were then ranked

for fold-change expression and screened for secreted factors (see figure 6 for study design).

Figure 6: Study design to identify new factors involved in the adipogenesis towards a brown adipocyte phenotype. Human adipose stromal cells (hASCs) were isolated from the deep neck (dn) and subcutaneous (sc) region of human subjects (1) [47]. These cells were then in vitro differentiated (2) and used for microarray analysis (3). 422 genes were significantly different expressed with a p-value below 0.05. This dataset was then screened for secreted factors which were ranked for the fold change expression between dn and sc. LTBP1, a secreted factor was 2-fold higher expressed in the hASCs derived from the dn compared to sc. Screening the literature revealed also other LTBPs differently expressed between brown and white hASCs [85]. Therefore, we decided to investigate the role of the LTBP family in the adipogenesis and their impact on the UCP1 expression in vitro.

18

Among the genes with a significant differential expression between dn and sc hASCs, LTBP1

was found to be enriched in dn hASCs. Of note, members of this family have also been

found to be higher expressed in brown compared to white hASCs in another study [85]. As

mentioned previously, the LTBP isoforms are secreted factors and known to regulate TGFβ

bioavailability. Several studies have been published on TGFβ and its role in adipogenesis

and WAT browning (see 1.7.1). However, to date, the role of LTBP family members in

adipogenesis and in the context of adipose tissue browning has not been addressed.

Therefore, the aim of this thesis was to investigate the role of the LTBP family members

in adipogenesis and their impact on UCP1 expression and metabolic function in vitro.

19

2. Material and Methods

2.1 Material

2.1.1 Chemicals, kits and reagents

Name Company Order number

Acetyl-CoA Sigma-Aldrich #A2056

Agarose Sigma -Aldrich #A9539

Ampicillin AppiChem #A0839

Antimycin A Sigma-Aldrich #A8674

BactoAgar BD #214010

BactoTryptone BD #211705

BactoYeast Extract BD #212750

BamHI NE Biolabs #R0136S

BfuAI NE Biolabs #R0701S

Biotin Sigma-Aldrich #B-4639

Bolt 4-12% Bis-Tris Plus gels ThermoFisher

Scientific

#NW04125BOX

#NW04122BOX

#NW04120BOX

Bolt LDS sample buffer, 4x ThermoFisher

Scientific

#B0007

Bolt MES SDS running buffer, 20X ThermoFisher

Scientific

#B0002

20

Bolt sample reducing agent, 10X ThermoFisher

Scientific

#B0009

Bradford protein assay concentrate, 5x Bio-Rad #5000006

BSA Sigma-Aldrich #A7906

BSA protein assay standard ThermoFisher

Scientific

#23209

cOmplete protease inhibitor cocktail

tablets

Roche #04693116001

Cortisol Sigma-Aldrich #H-0888

Dexamethasone Sigma-Aldrich #D-1756

D-Glucose Invitrogen #Q100-37

Dibutyryl-cAMP Sigma-Aldrich #D0627

Direct-zol RNA Miniprep Kit Zymo Research #R2052

DMEM ThermoFisher

Scientific

#41966029

DMEM base Sigma-Aldrich #D5030

DMEM/F12 ThermoFisher

Scientific

#31330-038

DMSO Honeywell #41641

dNTP mix, 20 mM Amersham #27-2094-01

DPBS ThermoFisher

Scientific

#14190094

DTNB Sigma-Aldrich #D8130

21

DTT solution, 100 mM ThermoFisher

Scientific

#Y00147

DuoSET ELISA Ancillary Reagent kit 1 R&D #DY007

DuoSET ELISA Ancillary Reagent kit 2 R&D #DY008

ECORI-HF NE-Biolabs #R3101S

EDTA Carl Roth #8043

Ethanol Honeywell #32205

FCCP Sigma-Aldrich #C2920

FCS ThermoFisher

Scientific

#10270106

Formaldehyde solution, 37-38 % (w/v) AppliChem #131328

Free glycerol reagent Sigma-Aldrich #F6428

Gateway LR Clonase II plus enzyme mix ThermoFisher

Scientific

#12538200

GeneArt Genomic Cleavage Detection Kit ThermoFisher

Scientific

#A24372

GlutaMAX, 100x ThermoFisher

Scientific

#35050061

Glycerol Carl Roth #7530

Glycerol standard solution, ~2.8 mM Sigma-Aldrich #G7793

HCl solution, 1 M Honeywell #35328

HEPES, 1 M BioChrome #L1613

Hexan Sigma-Aldrich #296090

Human TGF-beta 1 DuoSET ELISA R&D #DY240

22

Human TGF-beta 2 DuoSet ELISA R&D #DY302

Human TGF-beta 3 DuoSet ELISA R&D #DY243

IBMX Sigma-Aldrich #I-5879

Insulin Sigma-Aldrich #I-1507

Isopropanol VWR #20842

Janus Green B Carl Roth #7697

Kanamycin ThermoFisher

Scientific

#11815024

KCl Merck #104935

L-glutamine solution, 200 mM, 100X ThermoFisher

Scientific

#25030024

Lipofectamine2000 ThermoFisher

Scientific

#11668019

MEM non-essential amino acids solution,

100x

ThermoFisher

Scientific

#11140035

MgCl2 Merck #105833

MgSO4 ∙ 7 H2O Sigma-Aldrich #230391

Monarch DNA Gel Extraction Kit NE Biolabs #T1020S

Monarch PCR & DNA Cleanup Kit NE Biolabs #T1030S

Na2HPO4 AppliChem #A3567

NaCl Merck #106406

NAH2PO4 ∙ H2O Merck Millipore #106349

NaN3 Sigma-Aldrich #S2002

23

Neon Transfection system 100 µl kit ThermoFisher

Scientific

#MPK10025

Nile Red, 1 mg/ml Sigma-Aldrich #N3013

Novex NuPAGE 3-8% Tris-Acetate-Protein-

Gels

ThermoFisher

Scientific

#EA03752BOX

Novex NuPAGE LDS sample buffer (4x) ThermoFisher

Scientific

#NP0007

Novex NuPAGE Tris-Acetate-Running

buffer

ThermoFisher

Scientific

#LA0041

Novex Tris-Glycin native running buffer ThermoFisher

Scientific

#LC2672

Novex Tris-Glycin native sample buffer ThermoFisher

Scientific

#LC2673

Oligomycin Sigma-Aldrich #04876

Opti-MEM I ThermoFisher

Scientific

#31985047

Oxalacetate Sigma-Aldrich #171263

Panthotenat Sigma-Aldrich #P-5155

Penicillin/streptomycin solution, 100x ThermoFisher

Scientific

#15140122

peqGOLD DNA ladder mix VWR #25-2040

peqGOLD Plasmid MiniPrep kit I VWR #12-6943-02

phosSTOP phosphatase inhibitor cocktail

tablets

Roche #04906845001

24

PureLink HiPure plasmid MidiPrep kit ThermoFisher

Scientific

#K210005

Puromycin Sigma-Aldrich #P8833

Quick ligation kit NE Biolabs #M2200S

Random Primers ThermoFisher

Scientific

#48190011

Recombinant TGFβ1 Abcam #50036

Rosiglitazone Cayman #71740

Rotenone Sigma-Aldrich #R8875

SB 431542 R&D #1614

Seahorse XF base medium Agilent #103334-100

Seahorse XF Calibrant solution Agilent #100840-000

Seahorse XF glucose solution 1M Agilent #103577-100

SeeBlue Plus2 pre-stained protein standard ThermoFisher

Scientific

#LC5925

Skim milk powder Sigma-Aldrich #70166

Sodium pyruvate solution, 100 mM ThermoFisher

Scientific

#11360039

Sso Advanced Universal Probes Supermix Biorad #1725281

SuperScript II Reverse Transcriptase Invitrogen #18064014

T4 PNK NE Biolabs #M0201S

TE Buffer ThermoFisher

Scientific

#12090015

25

Trans-Blot Turbo RTA 0.2µM PVDF Transfer

kit

Biorad #1704272

#1704273

Transferrin Sigma-Aldrich #T-2252

Triglyceride reagent Sigma-Aldrich #T2449

Triiodothyronine Sigma-Aldrich #T-6397

Tris base Sigma-Aldrich #T1503

Tris HCl Sigma-Aldrich #10812846001

Triton

Trypsin/EDTA solution, 0.05 %/0.02% Merck #L2143

Tween 20 AppliChem #A1389

Westar Nova 2.0 Western blot substrate

solutions

Cyanagen #XLS071

XbaI NE Biolabs #R0145S

2.1.2 Equipment

Name Company Order number

cell culture dishes, 10 cm Sarstedt #83.3902

cell culture dishes, 6 cm Corning #353002

cell culture flasks, 175 cm² Sarstedt #83.3912.002

cell culture flasks, 175 cm² Corning #353112

cell culture plates, 12 well Corning #353043

cell scrapers Sarstedt #83.1830

cell strainer, 70 µm Falcon #352350

26

cryotubes Sarstedt #72.379.992

Finnpipette F2 Multichannel Thermo Scientific #4662020 #4662040

Finnpipette F2 single Thermo Scientific #4642090 #4642010 #4642020 #4642060 #4642080

glass pipettes VWR #612-1701

glass slides VWR #631-1551

Mr. Frosty freezing container Thermo Scientific #10110051

Multipipette Eppendorf #EP4982000012

multipipette tips, 0.5 ml, non-sterile Eppendorf #0030089421

multipipette tips, 10 ml, non-sterile Eppendorf #0030089820

multipipette tips, 5 ml Eppendorf #0030089669

multipipette tips, 5 ml, non-sterile Eppendorf #0030089456

PCR plates, 96-well, white shell/clear well BioRad #HSP9601

pipette tips, 1250 µl Sarstedt #70.1186

pipette tips, 20 µl Sarstedt #70.1116

pipette tips, 200 µl Sarstedt #70.760.002

plastic tubes with plug cap 13 ml Sarstedt #62.515.006

plastic tubes with screwed cap, 15 ml Sarstedt #62.554.502

plastic tubes with screwed cap, 50 ml Sarstedt #62.547.254

plate sealing films Bio-Rad #MSB1001

reaction tubes, 0.5 ml Sarstedt #72.704.400

27

reaction tubes, 1.5 ml Sarstedt #72.706.400

reaction tubes, 2 ml Sarstedt #72.695.400

seahorse XFe96 cell culture multiplate Agilent #101085-004

serological pipettes, 1 ml Corning #357521




sterile filter 0.2 µM GE Healthcare #10462200

sterile filter units, 0.2 µM, 250 ml ThermoFisher

Scientific

#568-0020

sterile filter units, 0.2 µM, 500 ml ThermoFisher

Scientific

#569-0020

syringes, 20 ml BD #300629

syringes, 50 ml BD #300863

2.1.3 Devices

Name Company

Bolt Mini Gel Tank electrophoresis chamber ThermoFisher Scientific

BZ-9000 microscope Keyence

Chemidoc MP Biorad

Infinite 200Pro multimode plate reader Tecan

Min-Sub Cell GT Cell electrophoresis chamber Biorad

NanoDrop 2000 spectrophotometer ThermoFisher Scientific

28

Neon Transfection system ThermoFisher Scientific

peqStar 2X Universal Gradient thermocycler VWR

Seahorse XFe96 analyzer Agilent

Trans Blot Turbo Biorad

2.1.4 Buffers and solutions

Protein lysis buffer, pH 7.5 + 1% Triton X-100

+ 10 mM Tris

+ 10% glycerol

+ 150 mM NaCl

+ 2 mM EDTA

in dH2O

and ad

+ 1 mM DTT

+ 1X protease inhibitor

TBS 10X, pH 7.6 + 200 mM Tris base

+ 1.5 M NaCl

in dH2O

TBST + 1X TBS

+ 0.1% Tween 20

in dH2O

Antibody buffer + 2.0% (w/v) BSA

+ 0.02% (v/v) NaN3

in TBST

Western blot blocking buffer + 5.0 % (w/v) milk powder in TBST

29

2.1.5 Bacteria and mammalian cells

Name Company/Origin Order number

One Shot TOP10 Chemically Competent E. coli Thermo Fisher #C404010

One shot Stbl3 Chemically Competent E. coli Thermo Fisher #C737303

Human SGBS cells [131] -

2.1.6 Cell culture medium

SGBS basal medium (0F) + 100 U/ml Penicillin/streptomycin + 17 μM Pantothenate + 33 μM Biotin in DMEM:F12 (1:1)

SGBS growth medium + 10 % (v/v) FCS In 0F Medium SGBS Differentiation medium I (Quick medium) + 0.01 mg/ml Transferrin (day 0 - day 4) + 0.2 nM Triiodthyronine (T3)

+ 100 nM Cortisol + 2 μM Rosiglitazone + 20 nM Insulin + 25 nM Dexamethasone + 250 μM IBMX in 0F Medium

SGBS Differentiation medium II (3FC) + 0.01 mg/ml Transferrin (day 4-day 14) + 20 nM Insulin

+ 100 nM Cortisol + 0.2 nM Triiodthyronin (T3) in 0F medium

Seahorse XF assay medium + 1X glutamaxx + 1 mM sodium pyruvate + 10 mM glucose solution + 1%(w/v) BSA in Seahorse XF Base medium

30

2.1.7 Bacterial cell culture medium

LB agar, pH 7.00 + 10 g/l BactoTryptone + 5 g/l BactoYeast Extract + 15 g/l Bacto Agar + 10 g/l NaCl in dH2O Add either: 100 µg/ml Ampicillin or 100 µg/ml Kanamycin

LB medium 5X, pH 7.00 + 50 g/l BactoTryptone + 25 g/l BactoYeast Extract + 50 g/l NaCl in dH2O Add either: 100 µg/ml Ampicillin or 100 µg/ml Kanamycin

SOB medium + 10 mM MgCl2 + 10 mM MgSO4

+ 10 mM NaCl + 2.5 mM KCl + 5 g/l BactoYeast Extract

+ 20 g/l BactoTryptone in dH2O

SOC medium + 20 mM glucose in SOB medium

2.1.8 qPCR and sequencing Primer

Table 2: Primer used for qPCR or sequencing

Gene Primer Sequence 5’→3’

Adiponectin forward reverse

GGCCGTGATGGCAGAGAT CCTTCAGCCCCGGGTACT

COX8 forward reverse

TTTGTGGTGTACTCCGTGCCATCATGTC GTAAGCCCAACGGCCAATTCCATGATCC

CPT1B forward reverse

CATGGTGGGCAAGTAACTATGTGAGTGACTG GCACGTCTGTATTCTTGATGAGCACAAGG

31

DIO2 forward reverse

CGATGCCTACAAACAGGTGAAATTGGGTGAG AGCCTCATCAATGTAGACCAGCAGGAAGTCAG

GLUT4 forward reverse

TTCCAACAGATAGGCTCCGAAG AAGCACCGCAGAGAACACAG

HPRT forward reverse

GAGATGGGAGGCCATCACATTGTAGCCCTC CTCCACCAATTACTTTTATGTCCCCTGTTGACTGGTC

LTBP1 forward reverse

CGAGCATCTGTAAAGTGACCT AAGATGGCAAATTACCACTCGG


AGCACCAACCACTGTATCAAAC CTCATCGGGAATGACCTCCTC


TCCCCAGGGCTACAAGAGG AGACACAGCGATAGGAGCCA


CTGCCCATTCTGCGGAACAT GCCGAGTAGTGGTAACCAGG

NDUF8B forward reverse

CCGCCAAGAAGTATAATATGCGT TATCCACACGGTTCCTGTTGT

PGC1α forward reverse

CTCAAATATCTGACCACAAACGATGACCCTC GTTGTTGGTTTGGCTTGTAAGTGTTGTGAC

PPARG forward reverse

GATCCAGTGGTTGCAGATTACAA GAGGGAGTTGGAAGGCTCTTC

PRDM16 forward reverse

TCTACATTCCTGAAGACATTCCGATCCCA CCGTCAGTATTTGCTCCCATCCGAAGTC

SMAD4 forward reverse

TACACCAACAAGTAATGATGCC AATCCTTTCCGACCAGCC

TGFβ1 forward reverse

CTAATGGTGGAAACCCACAACG TATCGCCAGGAATTGTTGCTG


CCCCGGAGGTGATTTCCATC GGGCGGCATGTCTATTTTGTAAA


GGAAAACACCGAGTCGGAATA GCGGAAAACCTTGGAGGTAAT

TGFβR1 forward reverse

ACGGCGTTACAGTGTTTCTG GCACATACAAACGGCCTATCTC


GTAGCTCTGATGAGTGCAATGAC CAGATATGGCAACTCCCAGTG


TGGGGTCTCCAGACTGTTTTT CTGCTCCATACTCTTTTCGGG

UCP1 forward reverse

GGAAAGAAACAGCACCTAGTTTAGGAAGCA CGTCAAGCCTTCGGTTGTTGCTATTATTCTG

32

UQCRC2 forward reverse

TTCAGCAATTTAGGAACCACCC GGTCACACTTAATTTGCCACCAA

M13+ (sequencing

primer)

forward ACGTTGTAAAACGACGGCCAGT

2.1.9 siRNAs

Table 3: Small interfering RNAs (siRNA) used for in vitro gene knockdown

siRNA pool Sequence 5’→3’

Non-targeting control 1. UAAGGCUAUGAAGAGAUAC 2. AUGUAUUGGCCUGUAUUAG 3. AUGAACGUGAAUUGCUCAA 4. UGGUUUACAUGUCGACUAA

SMAD4 1. GUGUGCAGUUGGAAUGUAA 2. GUACAGAGUUACUACUUAG 3. GAGUAUUGGUGUUCCAUUG 4. GUAAUGCUCCAUCAAGUAU

2.1.10 Western blot antibodies

Table 4: Antibodies used for Western blot

Antigen clone host supplier Order number

Anti Goat Rabbit Santa Cruz sc-2768

Anti Mouse

520 nm

Goat Biorad #12005866

Anti Mouse

HRP


Anti Rabbit

520 nm


Anti Rabbit

HRP


33

GAPDH

Recombinant

rhodamine labeled

Biorad #12004168

LTBP1 Monoclonal Mouse R&D MAB388

LTBP2 Polyclonal Goat R&D AF3850

OXPHOS Monoclonal Mouse abcam ab110411

PPARG Monoclonal Rabbit Cell signaling #2443

TGFβ1 Monoclonal Rabbit abcam ab179695

TGFβ2 Monoclonal Mouse abcam ab36495

TGFβ3 Polyclonal Rabbit abcam ab227711

Tubulin

Recombinant

rhodamine labeled

Biorad #12004166

UCP1 Monoclonal Mouse R&D MAB6158

2.1.11 Single guide RNAs used for CRISPR/Cas9 mediated gene knockout

Table 5: sgRNA used for CRISPR/Cas9 mediated gene knockout

Target sgRNA Sequence 5’→3’

Empty vector (EV) Top Bottom

ACCGGTCACCGATCGAGAGCTAG AAACCTAGCTCTCGATCGGTGAC

LTBP1 Top Bottom

ACCGGTGGAACTGTGGGTACCTCC AAACGGAGGTACCCACAGTTCCAC

LTBP2 Top Bottom

ACCGCAGGTGGCAGAACTTCCCGG AAACCCGGGAAGTTCTGCCACCTG

LTBP3 Top Bottom

ACCGCGGCGTTCGAGCTCTCAATG AAACCATTGAGAGCTCGAACGCCG

LTBP4 Top Bottom

ACCGCTGCGATCCTGGGTACCACG AAACCGTGGTACCCAGGATCGCAG

TGFβ1 Top Bottom

ACCGTTGATGTCACCGGAGTTGTG AAACCACAACTCCGGTGACATCAA

TGFβ2 Top Bottom

ACCGCGACGAAGAGTACTACGCCA AAACTGGCGTAGTACTCTTCGTCG

34

TGFβ3 Top Bottom

ACCGATAAATTCGACATGATCCAG AAACCTGGATCATGTCGAATTTAT

2.1.12 Plasmids

Table 6: Plasmids used for cloning

Name Supplier Order

number

Comment Properties

pMuLE ENTR U6

stuffer sgRNA

scaffold L1-R5

Adgene #1000000060 Kindly provided

by Ian Frew

For sgRNA

insertion

pMuLE ENTR U6

stuffer sgRNA

scaffold L4-L5


by Ian Frew

For sgRNA

insertion

pMuLE ENTR U6

stuffer sgRNA

scaffold L1-R5


by Ian Frew

For sgRNA

insertion

pMuLE ENTR CMV-

hCas9 L3-L2


by Ian Frew

Cas9 enzyme

pMuLE ENTR CMV-

hCas9 L5-L2


by Ian Frew

Cas 9 enzyme

pSBtet-GP Adgene #60495 Kindly provided

by Eric Kowarz

Sleeping Beauty

recognition sites,

eGFP, Puromycin

resistance

pLenti6.3/V5- DEST Thermo-

Fisher

Scientific

#V53306 Destination

vector for LR

Gateway cloning

35

2.1.13 Cleavage assay primer

Table 7: Primer used for Gene Art Genomic Cleavage Assay

Gene Cleavage primer Sequence 5’→3’ LTBP1 forward

reverse GTGAGGGACAGCAGTGCTC CTCTTCCCCTGAGACAAAAAGTCCA


CAGGAAGGCAGGGGC GGCCCAGCTGTGCCG


AGCAAGCACGCCATCTACG AAGGAAAGGCAGATCCCGACT


GGGGGTTTAATCCAGAGGCC ATCTTGCCTCCCTGCTTC

2.1.14 Software

Name Version Company

BZII Viewer and Analyzer 2.1 Keyence

CFX Manager 3.1 Biorad

Image Lab 6.0.1 Biorad

Inkscape vector graphic

editor

0.92 Inkscape

Office 365 Microsoft

Prism 7.0 GraphPad

pMuSE SB eGFP-

P2A-PuroR-DEST

- - Cloned by Jan-

Bernd Funcke in

our department

Combination of

pLenti6.3/V5-

DEST and pSBtet-

GP [132,133]

pCMV(CAT)T7-

SB100


by Zsuzsanna

Izsvak

Sleeping Beauty

transposase

36

SnapGene 4 GSL Biotech

Wave 2.6 Agilent

37

2.2 Methods

2.2.1 Cell culture

Simpson-Golabi-Behmel Syndrome (SGBS) cells are hASCs isolated from the scWAT of a 3-

month old infant who suffered from SGBS [131]. In the following, the term SGBS

preadipocytes will be used for this cell strain. These SGBS preadipocytes can be easily

subjected to adipogenesis and reach a differentiation rate up to 95% [131]. In addition,

they can be kept until the 30th generation without affecting differentiation [131].

SGBS preadipocytes were grown in T175 cell culture flasks with growth medium and split

at latest before full confluence was reached. To prepare SGBS preadipocytes for seeding,

cells were washed once with Dulbecco’s phosphate buffered saline (DPBS), subjected to

2 ml trypsin and incubated for 5 minutes at 37°C, 5% (v/v) CO2. Subsequently,

preadipocytes were resuspended in growth medium and centrifuged for 10 min at

1100 rpm. Thereupon, supernatant was discarded, and cells were resuspended once more

in 10 ml growth medium. Afterwards cell number was determined by counting the

preadipocytes with a light microscope and a Neubauer cell counting chamber.

For differentiation, SGBS cells were seeded into cell culture vessels and grown (37°C,

5%(v/v) CO2) for 72 h in growth medium until full confluence was obtained (d0). Cells were

then washed once with DPBS and growth medium was substituted with Quick medium.

After another four days the medium was substituted with 3FC medium without a washing

step. Cells were then kept for further ten days in the incubator until mature lipid laden

adipocytes were obtained (d14). If not stated differently, cells were cultured based on this

method in all experiments. Seeding densities are depicted in table 8.

For prolonged storage of SGBS preadipocytes, up to 500.000 cells were resuspended in 1 ml

growth medium containing 10 % DMSO and added to kryo tubes. These tubes were frozen

in freezing containers (Mr. Frosty, Thermo scientific) at -80°C and on the next day

transferred to liquid nitrogen for long term storage.

38

Table 8: Seeding densities for SGBS cell differentiation approach

2.2.2 Differentiation rate and microscopic pictures

Differentiation rate of mature adipocytes was determined on day 14 by counting the

number of preadipocytes and adipocytes in three independent segments using a light

microscope equipped with a net micrometer. If the cell contained at least 3 lipid droplets,

it was considered to be an adipocyte. On the same day, pictures of mature adipocytes were

taken at 10-fold magnification using a BZ-900 microscope (Keyence).

2.2.3 Determination of triglyceride content

Triglyceride accumulation in SGBS adipocytes was determined enzymatically on d14 by

measuring the intracellular glycerol content. Therefore, triglycerides were isolated by

washing the wells 2-times with 500 µl Hexan:Isopropanol (60:40 ratio). The solved

triglycerides were then transferred into a microcentrifuge tube. This tube was incubated

with an open lid overnight until solvent was completely evaporated. Samples were then

stored at -20°C or resuspended in 100 µl isopropanol. Per sample and well 0.33 µl were

Cell culture

vessel

Cell density/well

followed by 72 h

incubation

Medium

volume (ml)

Usage

12-well plate 20,000 1

RNA isolation, generation of cell

supernatant, determination of

differentiation rate,

determination of triglyceride

content

6 cm dish 150,000 5 Protein isolation

96-well

Seahorse XF96

cell culture

microplates

2,500 0.1 Measurement of cellular oxygen

consumption and extracellular

acidification rate

39

then transferred into a 96-well plate. Ten µl of a glycerol standard (1250 µg/ml, 625 µg/ml

312 µg/ml, 156.3 µg/ml, 78 µg/ml, 39 µg/ml diluted in isopropanol) was transferred into

the 96 well plate as well. Subsequently, 160 µl/well triglyceride reagent to cleave

triglycerides into glycerol and 40 µl/well glycerol reagent for detection of free glycerol were

added to each well. After 15 minutes of incubation at room temperature (RT), the

absorption was measured at 550 nm and triglyceride content was calculated using a linear

standard curve.

2.2.4 RNA isolation and qRT-PCR

Total RNA of SGBS cells was harvested with TRIzol reagent and then either stored at -80°C

or isolated subsequently with the ZYMO Direct-zol RNA Miniprep kit, according to the

manufacturer’s protocol. RNA concentration was determined with a NanoDrop 2000

spectrophotometer (Thermo fisher) measuring the wavelength at 260 nm and using the

following equation:

c=(A*ε)/b

c= concentration (ng/µl); A= absorbance (AU); ε=40 ng*cm/µl; b= pathlength (cm)

For cDNA synthesis 500-1000 ng of RNA was mixed with 1 µl Random Primer (3 µg/ml) and

subjected to denaturation (85°C for 3 min). Afterwards samples were immediately stored

on ice and each complemented with 10 µl of RT-PCR Master-mix, containing 1 µl

Superscript II RT (200 units), 1 µl dNTP (10 µM), 1 µl H2O, 2 µl DTT (0.1 M) and 5 µl first

strand buffer(5X). Samples were then subjected to a peqStar 2X Universal Gradient

thermocycler using the following conditions:

25°C for 15 min

42°C for 50 min

70°C for 15 min

4°C until termination

Afterwards samples were diluted with RNAse free water to a final concentration of 5 ng/µl

cDNA.

40

For quantitative real time PCR (qRT-PCR) 1 µL Primer (5 µM), 5 µl Sso Advanced Universal

SYBR Green Supermix and 3 µl H2O per well was pipetted in advance into a hard shell 96-

well PCR plate (Biorad). Thereupon 1 µl cDNA (5 ng/µl) was added per well. Finally, the PCR

assay was performed in a CFX96 Real-time PCR detection System (Biorad) with the

following conditions:

30 s at 95 °C

40 Cycles of

5s at 95°C

30s at 60°C

Verification of the correct amplicon size was determined by analyzing the melting

temperature of the PCR products. If not stated differently Hypoxanthine-Guanin-

Phosphoribosyltransferase (HPRT) was used as a house keeping gene and the expression

was calculated either using the 2-ΔCt or the 2-ΔΔCT method for normalization.

2.2.5 Protein isolation, separation and immunoblotting

For protein isolation, cells in 6 cm dishes were washed once with ice cold DPBS and

subjected subsequently to 100 µl of protein lysis buffer, followed by harvesting with a cell

scraper and transferring into a microcentrifuge tube. Samples were then incubated for

30 min on ice and centrifuged for 30 min at 14,000 g. Subsequently, supernatant was

transferred into a new microcentrifuge tube and stored at -20°C for further experiments.

To determine the protein concentration of the cell lysates a Bradford assay was applied

according to the manufacture’s protocol (Biorad). Protein samples were pre-diluted 1:10

and measured in duplicates. For concentration determination a bovine serum albumin

(BSA) standard was used and previously diluted in dH2O to 500 µg/ml, 300 µg/ml, 200

µg/ml, 100 µg/ml, 50 µg/ml and 0 µg/ml. Absorbance was measured at 550 nm using a

multimoded plate reader (Tecan).

Isolated protein samples (12-15 µg) were separated by sodium dodecyl sulfate

polyacrylamide gel electrophoresis (SDS-PAGE) using a Bolt Mini Gel Tank electrophoresis

chamber. Samples were mixed with 4x sample buffer and adjusted with H2O to a final

volume of 25-40 µl. For denaturation conditions, 10x reducing reagent was included in the

41

mix. Samples were then heated at 70°C for 10 min and transferred onto the gel. SeeBlue

Plus2 pre-stained protein standard served as a reference. Different SDS-PAGE running

conditions are depicted in table 9.

Table 9: SDS-PAGE running conditions for different protein properties (SDS: sodium dodecyl sulfate; MES: 2-(N-Morpholino)ethansulfonic acid; LDS: lithium dodecyl sulfate)

Protein properties Type of Gel Sample Buffer Running Buffer Conditions

Denatured/Native

Mixed molecular

weight

Bolt 4-12% Bis-

Tris Plus gels

(10-17 wells)

Bolt LDS

sample buffer

Bolt MES SDS

running buffer

200 V

25- 35 min

Native

High molecular

weight

NuPAGE 3-8%

Tris-Acetate-

Protein gels

(12-well)

Novex Tris-

Glycine Native

Sample Buffer

Novex Tris-

Glycine Native

Running Buffer

150 V

60 min

Denatured

High molecular

weight

NuPAGE 3-8%

Tris-Acetate-

Protein gels

(12-well)

NuPAGE LDS

Sample Buffer

NuPAGE Tris-

Acetate SDS

Running Buffer

150 V

90 min

For Western blot, filter paper and nitrocellulose membrane were previously incubated for

10 min in transfer buffer (Biorad). Proteins were then transferred on membrane using the

Trans-Blot Turbo transfer system (Biorad) according to the manufactures protocol (2.5 A,

19 V, 7 min) Hereinafter, the membrane was blocked for 1h in 1x tris-buffered saline with

Tween 20 (TBST) with 5% (w/v) fat free milk powder and subsequently washed 3 times for

10 min in 1xTBST. Afterwards membrane was incubated with the primary antibody

overnight at 4°C. Thereupon the membrane was washed again 3 times for 10 min in 1x TBST

and then incubated for another hour with the secondary antibody (in 5% milk TBST)

followed by another three washing steps á 10 min in TBST. Depending on the sensitivity of

the primary antibody, secondary antibodies with either fluorescent or horse radish

42

peroxidase (HRP) labeling were used and detected by fluorescence or enhanced

chemiluminescence (ECL), respectively. Membranes were analyzed using a ChemiDoc MP

Imager (Biorad).

2.2.6 Citrate synthase assay

To measure the citrate synthase activity as a surrogate for mitochondrial content, a kinetic

assay was performed. Citrate synthase catalyzes citrate from oxaloacetate and acetyl CoA

thereby reducing CoA. This reduced form transforms 5,5’ dithiobis 2 nitrobenzoic acid

(DTNB) into 2-nitro-5benzoic acid (TNB). The activity of citrate synthase can then be

assessed by the absorbance of TNB at 412 nm.

The following reaction media was prepared per 96-well and sample:

4 µl of 5 mM DTNB

6 µl of 10 mM acetyl CoA

19 µl of 1M Tris HCl pH 8.1

151 µl of H2O

This solution was added to 5 µg of protein sample in 10 µl H2O. Each sample was then

transferred to a 96-well plate. The 96-well plate containing the samples and reaction media

was then subjected to a multimoded plate reader (Tecan) with the following measurement:

-Baseline: 7 cycles measurement with an absorbance at 412 nm

-Injection: Injection of 10 µl/well oxaloacetate solution (10 mM)

-Reaction: 7 cycles measurement with an absorbance at 412 nm

Citrate synthase activity was calculated with the following equation

Units (µmol/min/mg) = ΔA412/min ∗ V(ml)

ε ∗ L(cm) ∗ mg protein

ΔA412 /min= The absorption at 413 nm/min at baseline activity subtracted from the

absorption at 413 nm/min at the endogenous activity.

V(ml) = sample volume (0.2 ml)

43

ε (mM-1cm-1) = extinction coefficient (13.6 for TNB)

L(cm) = pathlength (0.625 cm for 96-well plate)

2.2.7 Determination of the cellular oxygen consumption rate

To measure the cellular oxygen consumption rate (OCR) of SGBS adipocytes, an

extracellular Flux Analyzer (Seahorse XF96, Agilent) was used. Therefore, mature

adipocytes in 96-well Seahorse XF96 cell culture microplates were washed twice with 0F

medium on the day before measurement (for seeding and differentiation see 2.2.1) and

200 µL/well of Seahorse XF Calibrant solution was added to a Flux plate and subsequently

incubated at 37°C, no CO2. On the day of measurement, cells in the 96-well plate were

washed once with 100 µl/well Seahorse assay medium and finally 155 µl assay medium was

added. The plate was then incubated for 1h at 37°C, no CO2. The Seahorse Flux plate was

prepared for injection with 0.5 mM/well dibutyryl cyclic adenosine monophosphate

(cAMP), 2 µM/well Oligomycin, 4 µM/well carbonylcyanide-p-trifluoro-

methoxyphenylhydrazone (FCCP) and 1.5 µM/well Antimycin A and Rotenone (Ant/Rot)

according to the manufacture’s protocol. Cellular oxygen consumption rate was then

measured in the 96-well plate with following parameters: 3-5 cycles basal; 10-12 cycles

after cAMP injection, 5 cycles after Oligomycin injection, 3 cycles after FCCP injection, 3

cycles after Ant/Rot injection.

2.2.8 Normalization for lipid content and cell density

To normalize for lipid content, cells were stained with Nile Red, a lipophilic dye which

incorporates into neutral lipids and can be detected by its fluorescence. After the

measurement of the OCR, cells in the 96 well plate were fixed in 4% PFA/PBS (100 µl/well)

for a minimum of 20 min and then washed with DPBS. Subsequently 100 µl of 10%

DMSO/0.1% Nile Red in DPBS was added per well and incubated for 20 min, protected from

light. Afterwards Nile Red staining was analyzed using a Tecan plate reader at 550 nm

excitation and 640 nm emission. Afterwards the plate was washed again once with DPBS

and subsequently 25 µl of a 0.3% Janus green solution was added per well, followed by 30

min incubation time. Thereupon, the plates were washed with water and then tapped on

a paper towel until the plate was dry. 100 µl of a 0.5 M HCl solution was added per well

44

and incubated for further 10 min. Subsequently, the Janus green staining was measured

using a Tecan plate reader at 630 nm absorption. The data was used to normalize the

measured OCR data to lipid content (NileRed) or cell content (JanusGreen) of the individual

well and as a determination for differentiation rate or cell amount.

2.2.9 siRNA Transfection

For siRNA transfection SGBS preadipocytes were seeded as described before (2.2.1) and

transfected with the siRNA after 24 h of cell growth and incubated for 48 h until d0 was

reached (see 2.2.1). For transfection of a single 12 well, 1 µl siRNA (20 µM) was added to

49 µl of OptiMEM and mixed with 2.5 µl lipofectamine 2000, previously added to 47.5 µl

OptiMEM. This stock solution was incubated for 20 min and subsequent pipetted drop by

drop on SGBS preadipocytes to a final siRNA concentration of 20 nM.

2.2.10 Multiple Gateway cloning for CRISPR/Cas gene knockout in vitro and transposase

mediated stable transfection

CRISPR/Cas9 system has become the gold standard for stable genetic modification in cells.

The Cas9 protein, discovered in the bacterial immune system is a RNA guided DNA

endonuclease allowing specific gene manipulation [134,135]. A convenient way to clone

CRISPR/Cas9 knockout constructs is the multiple lentiviral expression system (MuLE) [136].

This has been frequently used as a flexible tool to recombine up to four different ENTRY

vectors containing e.g fluorescent proteins, resistance genes, the Cas9 gene or the

designed single guide RNAs (sgRNA) into a destination vector (Figure 7). The recombination

is achieved by LR Gateway cloning. In this thesis the system was used to recombine up to

four entry vectors with different sgRNAs and the Cas9 into a destination vector [136,137].

Designing the intended sgRNA is achieved by public online tools e.g. the broad institute. By

typing in the target gene, the program suggests several options for sgRNA sequences

ranked for lowest off target effects and highest specificity in gene targeting calculated by

algorithms [138,139].

The final constructs for CRISPR/Cas9 mediated knockout are named multiple sleeping

beauty expression vector (MuSE). These vectors contain an enhanced green fluorescent

protein (eGFP) under the control of a RPL13A promoter which is used for visual transfection

45

control. Under the control of the same promotor a puromycin resistance (PuroR) gene for

selection of transfected cells is encoded. Furthermore, a cytomegalovirus (CMV) promoter

controls the Cas9 and a U6 promoter the intended sgRNA. The complete cassette is flanked

by inverted repeats (ITR), which are Sleeping Beauty (SB) transposase recognition sites.

The usage of SB transposase system offers a way for stable gene integration into

mammalian cells [140]. Originally the sleeping beauty transposase was discovered as an

inactive gene in fish [141]. Genetic modification of this sequence created a new synthetic

transposon system. Through further modification of the gene a high activity was achieved,

enabling the generation of stable transgenic cell lines [140]. The SB transposase works

according to the cut and paste principle. It recognizes the ITRs which flank the transposon

(genes of interest). This transposon ends are then brought together by the SB transposase

and subsequent excised. Afterwards, the transposon is integrated by the SB transposase

into the host genome at TA dinucleotide sites. Compared to a common lentiviral

transfection, the SB transposase reveals a faster and safer way for a stable host gene

integration with high molecular weight plasmids [140].

Figure 7: LR Gateway cloning

with multiple fragments.

Multisite gateway cloning is a

powerful tool for rapid and easy

recombination of up to 4

different entry vectors. By using

the multiple lentiviral expressing

system (MuLe) two (A) and up to

four fragments (B) can be

recombined and cloned into a

destination vector within 16 h.

Specific attL and attR sites

mediate the recombination step.

CMR=Chloramphenicol-

resistance

ccdB=ccdB toxin gene.

46

For cloning of CRISPR/Cas9 constructs, partially modified plasmids of the multiple lentiviral

expression (MuLE) system kit (Adgene) were used [132,136]. The respective plasmids used

for cloning are listed in 2.1.12 table 6 and the related sequences and vector maps of

products are attached in the appendix (Figure S1 and S2)

1) Digestion of pMuLE ENTR plasmids

For preparation the pMuLE ENTR U6 stuffer sgRNA scaffold plasmids (see 2.1.12) were

digested for 2 h at 50°C using the restriction enzyme BfuAI with NEBuffer 3.1 according to

the manufacture’s protocol. Digested plasmids were then subjected to an agarose gel

electrophoresis (1% (w/v) agarose, 120 V, 30 min) to separate the insert from the

backbone. The plasmids backbone was then cut from the gel and extracted using the

Monarch DNA Extraction kit (NEB) according to its protocol. The product contained the

purified linearized plasmid, ready for insertion of the intended sgRNA.

2) sgRNA design

For Crispr/Cas 9 knockout constructs the respective sgRNA for the target gene was

designed using the broads institutes analysis tool [142], which assures maximal on-target

and minimal off-target effects using approved scoring models for sgRNA prediction [138].

The top and the complementary bottom sgRNA flanked by restriction sites for further

ligation, was then synthesized at Biomers GmbH (see 2.1.11 for sequences).

3) Phosphorylation and annealing

The first step of cloning was the annealing and phosphorylation of top and bottom sgRNA

with the following formula:

1 µl sgRNA top (100 µM)

1 µl sgRNA bottom (100 µM)

1 µl T4 DNA Ligase reaction buffer (10X)

6.5 µl H2O

0.5 µl T4 PNK (10 U/µl)

47

All reagents were mixed in a microcentrifuge tube and incubated in a peqStar 2X Universal

Gradient thermocycler at:

37°C 30 min

95°C 5 min

95°C to 8°C with 5°C/min cooling

4) Ligation of sgRNA duplexes into the opened entry vector

sgRNA duplexes were then diluted 1/200 in H2O to a total concentration of 50 nM and

ligated into the corresponding entry vector. For each ligation, the following approach was

pipetted into a microcentrifuge tube, followed by 10 min incubation at room temperature:

Digested pMuLE ENTR U6 stuffer sgRNA scaffold 50 ng

sgRNA duplex (50 nM) 1 µl

Quick ligase reaction buffer (2X) 5 µl

Quick ligase 1 µl

H2O fill up to a total of 11 µl

5) Transformation of ligated plasmids into TOP10 E. coli

For the amplification of the plasmid, the ligated plasmids were transformed into TOP10

E. coli bacteria. For this purpose, the bacteria were thawed 5-10 min on ice and then 7 µl

of the plasmid solution was added to the bacteria and incubated for 30 min on ice. The

bacteria were then heat shocked for 1.30 min at 42°C in a water bath. Subsequently, the

tubes with bacteria were incubated on ice for 2 min and 250 µl SOC medium was added to

each sample and shaken for 1h at 37°C, 220 rpm. Thereupon, each batch was plated on

agar plates with kanamycin (50 µg/ml) and incubated for 12-18h at 37°C.

6) Control digest and sequencing

Bacterial cultures were inoculated into liquid LB medium containing 100 µg/ml kanamycin

and were shaken overnight at 37°C. As a backup, 500 µl of each batch was mixed with 250 µl

glycerol and frozen at -80°C. The plasmids were isolated according to the instructions of

48

the peqGOLD miniprep kit. DNA concentrations were measured with a NanoDrop 2000

spectrophotometer (Thermo fisher) and afterwards a control digest was prepared.

Therefore 250 ng plasmid was mixed with 2.5 µl Cut smart buffer (10x) and 0.5 µl EcoRI-HF

enzyme and digested for 1-2 h at 37°C. Samples were subsequently mixed with loading dye

(6X) and applied to a 1% agarose gel (120 V, 400 mA, 100 W, 30 min) and analyzed with the

ChemiDoc MP imager.

To exclude mutations and to verify correct insertion, the entry vectors containing the

respective sgRNA were sent to Sanger sequencing (Eurofins genomics) using a M13+

forward primer.

7) Gateway LR reaction

The entry vector with the sgRNA was then recombined with the entry vector containing the

Cas9 and cloned into the destination vector using LR gateway cloning [137]. For

recombination of single sgRNA constructs see table 10. Table 11 depicts the recombination

for a TGFβ1-3 KO plasmid containing three sgRNAs.

The destination vector contains a puromycin resistance, GFP and the recognition sites for

the sleeping beauty transposon. For the LR reaction 1 µl entry sgRNA vector (10 fmol) was

added with 1 µl entry Cas9 vector (10 fmol) and 1 µl destination vector (20 fmol) to a PCR

tube and filled up to 5 µl with TE buffer. 2 µl LR Clonase II Plus Enzyme Mix (5x) was then

added to each sample and incubated for 24 h at 25°C. After that, the reaction was stopped

by adding 1 µl Proteinase K (10X), incubating for 10 min at 37°C.

8) Transformation in Stbl3 E. coli, plasmid isolation and control digestion

7 µl of the LR reaction tube was used for transformation of Stbl3 E. coli bacteria as described

above (5). Agar plates and liquid LB medium for inoculation contained 100 µg/ml Ampicillin

instead of Kanamycin. Plasmids were isolated as described before (6) and 500 ng of DNA

was digested for control, using 2.5 µl NEBuffer 3.1 and 0.5 µl BamHI at 37°C for 2 h. Samples

were mixed with loading dye (6X) and applied to a 1% Agarose gel (120 V, 400 mA, 100 W,

30 min) and analyzed with the ChemiDoc MP imager.

49

9) Midiprep DNA isolation

Bacteria containing the correct plasmids were inoculated from kryopreserved glycerol

stocks in an Erlenmeyer flask containing 100 ml liquid LB medium with 100 µg/ml

Ampicillin. Flasks were shaken at 37°C overnight with 220 rpm. Plasmids were then isolated

with the PureLink HiPure Plasmid Midiprep Kit adhering to the manufacture’s protocol.

DNA pellet was eluted with 100 µl TE buffer.

Table 10: Plasmids used for LR Gateway cloning for single CRIPSR/Cas9 knockout constructs

sgRNA ENTRY vector Recombined with…

(LR Gateway cloning)

Product

LTBP1

LTBP2

LTBP3

LTBP4

TGFβ2

TGFβ1

pMuLE ENTR U6 stuffer

sgRNA scaffold L1-R5

-pMuLE ENTR CMV-

hCas9 L5-L2

-pMuSE SB eGFP-P2A-

PuroR-DEST R1-R2

pMuSE SB CRISPR/Cas9

sgRNA

(LTBP1/LTBP2/LTBP3/

LTBP4/TGFβ2/ TGFβ1)

Table 11: Plasmids used for LR Gateway cloning for triple CRIPSR/Cas9 knockout construct

sgRNA ENTRY vector Recombined with…

(LR Gateway cloning)

Product

TGFβ1 pMuLE ENTR U6 stuffer

sgRNA scaffold L1-R5

-pMuSE SB eGFP-P2A-

PuroR-DEST R1-R2

-pMuLE ENTR CMV-

hCas9 L3-L2

-pMuSE SB

CRISPR/Cas9 TGFβ1-3


sgRNA scaffold L4-L5


sgRNA scaffold R3-R4

50

2.2.11 Electroporation

For transfection of SGBS cells with CRISPR/Cas9 knockout constructs a Neon

electroporation transfection system (Invitrogen) was used. 10.5 µg of Plasmid DNA per

knockout construct was mixed with 500 ng pCMV(CAT)T7-SB100 and adjusted to a final

volume of 110 µl with R buffer (Neon kit). Neon transfection system was prepared

according to the manufacture’s protocol. SGBS preadipocytes were trypsinized as

described before (see 2.2.1) and 1.1x106 cells were used for one transfection approach,

added into a 1.5 ml microcentrifuge tube and centrifuged for 5 min at 300g. Hereinafter,

the supernatant was discarded and cells resuspended in 110 µl plasmid-buffer solution.

Subsequently, 100 µl of this cell-plasmid-buffer suspension was taken up with a Neon

pipette (Neon kit) and the electroporation was performed according to the manufacture’s

protocol. Cells were electroporated at 1400 V with a pulse of 3x 10 ms. A mock control

containing no plasmid DNA was also subjected to electroporation and later used as a

selection control. After electroporation, cells were placed in a 10 cm dish with 10 ml

antibiotic free growth medium and incubated at 37°C, 5% CO2 (v/v). On the next day

medium was changed to normal growth medium, additionally containing 5 µg/ml

puromycin to start the selection process. When the mock control cells were completely

dead (after 7-10 days), medium was changed to normal growth medium without

puromycin. Cells which underwent selection were then transferred into T175 flasks,

cultivated and prepared for seeding or freezing (see 2.2.1).

2.2.12 Genomic cleavage assay

To control for DNA deletion in target gene regions of CRISPR/Cas9 mediated LTBP knockout

SGBS cells, a genomic cleavage assay was performed using the GeneArtGenomic cleavage

detection kit (Thermo). To address this, the genomic DNA regions which were assumed to

be mutated were amplified, denatured and re-annealed. A specific enzyme detects

mismatches in the re-annealed DNA strands and breaks it. A positive control was done to

verify successful amplification and cleavage. A cleavage exhibits a reduction of parental

DNA and cleaved bands compared to the control (NC), indicating a mutation in this DNA

region. The designed cleavage assay primers are listed in table 7.

51

2.2.13 Generation of SGBS supernatant

To generate FCS free supernatant of SGBS preadipocytes the LTBP knockout cells and

control cells were seeded in 12 well plates as described previously (2.2.1). After 72 h in

growth medium, the medium was aspirated, and cells immediately washed once with

DPBS. Afterwards cells were subjected to 0F medium (1 ml /well) and incubated for further

4 days at 37°C 5% CO2 (v/v) before the supernatant was transferred into microcentrifuge

tubes and stored at –20°C.

2.2.14 TGFβ ELISA

Enzyme linked immunosorbent assays (ELISAs) were used to determine TGFβ1, TGFβ2, and

TGFβ3 content of SGBS supernatants. To measure also bound TGFβ the supernatant

samples were previously heated for 5 min at 80°C to activate TGFβ. ELISAs were then

performed adhering to the manufacture’s protocol.

2.2.15 Statistics

GraphPad Prism 7 software was used for statistical analysis. Students t-test was applied to

compare two groups with a two-tailed calculation and assumption of Gaussian distribution.

When groups were normalized to control an unpaired t-test was used, otherwise a paired

t-test was applied. For more than two time points a two-way ANOVA was applied with

multiple comparison and tukey correction. For correlation analysis linear regression was

applied. A p-value of <0.05 was considered as statistically significant.

52

3. Results

Based on the previously introduced study by Tews et al., we hypothesized that hASCs

respective of the adipose tissue region, secrete different factors which act in an auto- or

paracrine manner, stimulating or inhibiting a differentiation towards a brown adipocyte

phenotype [47]. We identified LTBP1 as a factor higher expressed in hASCs derived from

the dn vs sc, suggesting a role in the adipogenesis towards a brown adipocyte phenotype.

The LTBP protein family has 4 members [104]. Screening the literature revealed that LTBP3

and LTBP1 were also higher expressed in isolated brown hASCs vs white hASCs in another

study [85]. Furthermore, LTBP2 and LTBP3 were differently expressed as well, when

comparing murine ASCs from BAT vs WAT [86]. Therefore, the expression of all four LTBP

members (LTBP1, LTBP2, LTBP3, LTBP4) were characterized during adipogenesis of in vitro

differentiated hASCs from mamma tissue and in SGBS preadipocytes. The SGBS cell strain

is commonly used as an in vitro model for human white preadipocytes [131,143]. These

cells can be easily subjected to adipogenesis and represent an excellent model to study any

effect on the UCP1 expression and metabolic function [79,144,145]. In a next step, we

aimed to establish CRISPR/Cas9 mediated LTBP knockout SGBS preadipocytes to study the

impact of LTBPs on adipogenic differentiation, mitochondrial content, UCP1 expression and

metabolic function.

3.1 The role of LTBPs in adipogenesis and their impact on UCP1 expression in vitro

3.1.1 LTBP1, LTBP3 and LTBP4 are enriched in hASCs from the deep neck adipose tissue

In order to validate the array findings by Tews et al., the cDNA samples of the isolated

hASCs derived from the dn and sc region of human adipose tissue were used to determine

the mRNA expression of LTBPs [47]. Indeed, LTBP1 was 2.5-fold higher expressed in dn vs

sc (Figure 8). LTBP2 was equally expressed between dn- and sc-derived samples. However,

LTBP3 (1.8-fold) and LTBP4 (10-fold) were also significantly higher expressed in hASCs

derived from the dn region compared to sc.

53

Figure 8: Expression of LTBP family members in isolated human adipose stromal cells (hASCs) from subcutaneous (sc) and deep neck (dn) adipose tissue. mRNA expression of all 4 members of the LTBP family was determined by qPCR in isolated hASCs derived from the dn and sc adipose tissue. HPRT served as a reference gene. Mean values +SEM of n=6 from paired patient samples are shown. *p<0.05, paired student’s t-test

3.1.2 LTBP expression is decreased during adipogenic differentiation

To characterize the expression of LTBP isoforms in human WAT samples and in isolated and

in vitro differentiated hASCs of WAT, tissue samples and hASCs were derived and isolated

from mammary adipose tissue of patients who underwent plastic surgery. All LTBPs, except

LTBP4 showed comparable regulation during adipogenesis in hASCs (Figure 9), i.e. high

expression levels in hASCs and decreased levels in mature adipocytes (hASC adip).

Figure 9: Expression of LTBP isoforms in human adipose tissue and in isolated and in in vitro differentiated human adipose stromal cells (hASCs). Samples of human adipose tissue are derived from mamma adipose tissue of n=28 patients which underwent plastic surgery. Mean age±SD = 45±15 years, Mean BMI±SD = 28.2±4.9 kg/m². These samples belong to our in-house biobank. hASCs have been isolated in a previous study from mamma adipose tissue [79]. LTBP mRNA expression in human mamma adipose tissue and in isolated hASCs and in vitro differentiated adipocytes (adip.). Mean values +SEM n=28 (tissue) and n=9 (hASCs), ***p<0.001, ****p<0.0001, two-way ANOVA, multiple comparison.

Interestingly, tissue expression varied between LTBP isoforms compared to hASCs. LTBP1

expression was equally in tissue compared to isolated stromal cells. In contrast, LTBP2

expression was lower compared to hASCs and LTBP3 was significantly higher in adipose

tissue compared to isolated stromal cells. Furthermore, LTBP4 was 100-fold higher

expressed in tissue samples. Of note, the expression levels of isolated mamma ASCs were

54

in line with the levels seen in isolated hASCs derived from supraclavicular sc adipose tissue

(Figure 8).

Primary cells are not suitable for prolonged experiments due to senescence after a few

passages in vitro. Therefore, the human SGBS preadipocyte cell strain was used to

investigate the function of LTBP family members in adipocytes. SGBS cells were subjected

to adipogenic differentiation (Figure 10). Before (d0) and at early stages of differentiation

(d2), the cells have a fibroblast like morphology (Figure 10A). Commencing from d4, the

incorporation of intracellular lipid droplets can be observed, which is steadily increasing

during the time course (d7, d10, d14). After d14, more than 70% of the total cell population

represents mature, lipid laden adipocytes. The expression of adipogenic marker genes was

significantly upregulated, e.g. PPARγ, adiponectin and glucose transporter 4 (GLUT4)

(Figure 10B). Furthermore, UCP1 was abundantly expressed in SGBS cells with increasing

amount during adipogenesis. All LTBP family members were expressed in SGBS cells, with

LTBP4 showing the lowest abundance (Figure 10C). All LTBPs showed similar expression

patterns during adipogenesis, with a rapid increase during early adipogenesis (d0-d2)

followed by a decrease until mature adipocytes were observed (d14). Of note, the LTBP

expression pattern and levels during adipogenesis are similar to that of in vitro

differentiated hASCs (Figure 9), which underlines the SGBS cells as a suitable cell model for

in vitro studies.

55

Figure 10: Expression of differentiation marker and LTBP members in SGBS cells during adipogenesis. SGBS preadipocytes (d0) were subjected to adipogenic differentiation for 14 days. (A) Microscopic picture of SGBS cells on different timepoints during adipogenesis. (B) mRNA expression of the differentiation marker PPARγ, adiponectin, GLUT4 and UCP1 during adipogenic differentiation were measured by qRT-PCR. (C) Expression of LTBP1-4 mRNA on different days during adipogenesis. HPRT served as a reference gene. Mean +SEM n=3; *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001; two-way ANOVA vs. d0.

56

3.1.3 CRISPR/Cas9 mediated knockout of LTBPs in SGBS cells

To elucidate the role of LTBPs in adipocytes, SGBS preadipocytes deficient for LTBP1, LTBP2,

LTBP3 and LTBP4 were generated using the CRISPR/Cas9 system. SGBS cells were

transfected in the preadipocyte state with the final vectors containing the Cas9 and the

sgRNA targeting either LTBP1, LTPB2, LTBP3, or LTBP4, and were then subjected to

puromycin selection for 7 days.

SGBS cells can be kept in culture until the 30th generation without affecting the

differentiation rate [131]. However, these cells are not immortalized and have a prolonged

doubling time of 38.4 h [131]. Thus, the generation of single cell clones was considered to

be not possible and bulk cultures were studied instead. Successful knockout of the target

gene was proven by cleavage assay, mRNA expression and Western blot. Each LTBP

knockout strain showed a reduction in DNA product of the target gene region compared to

the NC and a cleaved band smaller in size (Figure 11A). Furthermore, a significant decrease

in mRNA expression of LTBPs compared to empty vector control (EV) was observed in each

knockout bulk culture of SGBS preadipocytes (Figure 11B) and in mature adipocytes after

14 days of adipogenesis (Figure 11C). A stable downregulation of at least 80% was

measured in pre- and adipocytes. Deficiency of LTBP1 and LTBP2 was also confirmed by

Western blot, where protein in knockout cells was almost diminished compared to EV

control (Figure 11D and E). LTBP3 and LTBP4 antibodies were tested for Western blot but

both LTBPs could not be detected properly.

57

Figure 11: Transfection with CRISPR/Cas9 knockout constructs leads to stable downregulation of LTBPs in SGBS cells After stable transfection of SGBS preadipocytes with LTBP CRIPSR/Cas9 constructs, the cells were controlled for specific LTBP knockout. (A) Agarose gel of DNA products after GeneArt genomic cleavage assay to detect DNA mutations. PS = positive control; NC = negative control, Cleav. = Cleaved DNA. (B) LTBP mRNA expression of the respective knockout and empty vector control (EV) SGBS preadipocyte strain and (C) in adipocytes. Mean +SEM, n=4-5, ****p<0.0001, student’s t-test. (D) Western blot analysis of EV and LTBP1 knockout cells with α-Tubulin as loading control. (E) Western blot in EV and LTBP2 knockout cells with α-Tubulin as loading control.

58

3.1.4 LTBP deficiency does not alter the adipogenesis and mitochondrial content in

SGBS cells.

LTBPs are crucial for TGFβ secretion into the extracellular matrix [109]. TGFβ is known to

play a role in the regulation of adipogenic differentiation [36,96,98,99]. Furthermore, LTBPs

themselves can modulate the extracellular matrix composition which is known to have

influence on the adipogenesis as well [104,146].

Deficiency of LTBPs in SGBS cells had no effect on the morphology of mature adipocytes,

shown in Figure 12A by microscopic pictures. Moreover, there were no significant changes

in the differentiation rate between knockout strains and EV control (Figure 12B). In line

with this, the triglyceride content (Figure 12C) as well as the fluorescence after staining

with Nile Red, a lipophilic dye incorporating into neutral lipids (Figure 12D), was

comparable to the EV control as well. This was also confirmed by measuring the mRNA

expression of key adipogenic marker genes PPARγ, adiponectin and GLUT4 (Figure 12E),

which were comparable in the different cell strains. The only exception was adiponectin,

which was slightly, but statistically significant increased among LTBP1 knockout compared

to EV control. Browning of WAT is commonly accompanied by an increase in mitochondrial

mass in vivo [13]. mRNA expression of genes involved in mitochondrial metabolism were

not significantly altered between LTBP knockout adipocytes and empty vector (EV) control

(Figure 12F).

To further investigate mitochondrial content, protein expression of genes involved in the

respiratory chain were assessed by Western blot. Again, there were no significant

differences seen in LTBP-deficient SGBS cells, assured by densitometric analysis (Figure

12G, one representative is shown). Determination of citrate synthase (CS) activity serves as

a marker of intact mitochondria and mitochondrial content. CS activity was significantly

reduced by 15% in LTBP2-deficient adipocytes compared to the EV, whereas the other

LTBPs showed no significant alterations (Figure 12H).

59

Figure 12: Reduced LTBP expression has no effect on adipogenic differentiation and mitochondrial content. LTBP-deficient cells were subjected to differentiation until lipid laden adipocytes were observed. Different adipogenic and mitochondrial markers were measured to draw conclusions on the differentiation rate and mitochondrial content. (d14). (A) Microscopic pictures of LTBP knockout adipocytes and empty vector (EV) control with 10-times magnification. (B) Differentiation rate was counted on d14. Fold change compared to EV is shown. (C) Triglyceride content of mature adipocytes, n=3 (D) Nile Red staining of mature adipocytes, fold change of intensity compared to EV, n=4-5. (E) mRNA expression of key adipogenic marker was assessed by qRT-PCR. Fold change compared to EV control is shown. (F) Mitochondrial marker genes on d14 assessed by qRT-PCR. HPRT served as a reference gene and fold change to EV control is shown n=4 (LTBP2, LTBP3, LTBP4) n=5(LTBP1) (G) OXPHOS genes of the respiratory chain were assessed by Western blot. One representative is shown. (H) Citrate synthase activity was measured spectrophotometric by substrate conversion. Fold change to control is shown. Mean +SEM of, *p<0.05, students t-test vs. EV control.

60

3.1.5 Deficiency of LTBP2 and LTBP3 leads to a decrease in browning marker UCP1

The overall aim of this study was to identify potential targets which are involved in the

generation of UCP1-expressing cells. To address this, the mRNA expression of marker genes

known to be higher expressed in brown or beige adipocytes compared to white adipocytes

were determined.

Figure 13: LTBP2 and LTBP3 deficiency leads to a decreased UCP1 expression After 14 day of adipogenesis different adipocyte marker genes were analyzed. (A) mRNA expression of markers known to be upregulated in brow adipocytes versus white adipocytes. HPRT serves as a reference gene and fold change to EV control is shown n=4 (LTBP2, LTBP3, LTBP4) n=5(LTBP1). (B) Representative Western blot is shown and densitometric analysis was performed in n=3-5 independent experiments, GAPDH serves as a loading control. Fold change to EV is shown n=3 (LTBP3) n=4 (LTBP2, LTBP4) n=5(LTBP1). Mean +SEM, *p<0.05, ***p<0.001, ****p<0.001 students t-test vs. EV control.

The UCP1 expression was checked on mRNA and protein level, since browning is defined

by a significant upregulation of UCP1 [26]. A significant reduction of UCP1 was observed in

LTBP2-deficient cells (35%) and a 37% reduction in LTBP3-deficient SGBS adipocytes

compared to EV (Figure 13A). Brown marker genes type II iodothyronine deiodinase (DIO2)

and the transcription coregulator PR domain containing 16 (PRDM16) were not altered

compared to control cells (EV) (Figure 13A). However, PGC1α expression was 1.3-fold

61

upregulated in LTBP3-deficient SGBS adipocytes compared to EV cells. In line with the

mRNA expression, LTBP3 deficiency led also to downregulation of UCP1 protein expression

assessed by Western blot and densitometric analysis (Figure 13B). Of note, UCP1 protein

was reduced by 45% compared to EV control. PPARγ protein expression was equal

indicating no changes in differentiation. On the contrary to LTBP3 deficiency, the UCP1

protein expression was not altered in LTBP2-deficient cells compared to the EV cells.

3.1.6 LTBP deficiency leads to metabolic alterations in SGBS adipocytes

To address if the changes in UCP1 expression influence the metabolic function in LTBP-

deficient cells, the oxygen consumption rate (OCR) was measured with an extracellular flux

analyzer (seahorse). A schematic graph of a seahorse plot and how different values are

calculated is depicted in figure 14.

To mimic beta-adrenergic stimulation cells were treated with cAMP during the

measurement. cAMP induces lipolysis in the cells and the resulting free fatty acids activate

UCP1 which then should result in a higher oligomycin-insensitive OCR. Through this, one

can draw conclusions on the activity of UCP1. Indeed, there was a decrease in OCR after

stimulation with cAMP in LTBP2- and LTBP3-deficient cells compared to the control (Figure

15B and C). Basal respiration after cAMP injection was 22% lower in LTBP2- and 18% lower

in LTBP3-deficient cells. In addition, the response of the cellular respiration to cAMP

stimulation was reduced by 43% in LTBP2- and 44% reduced in LTBP3-deficient adipocytes.

This is in line with the UCP1 mRNA expression of LTBP2- and LTBP3-deficient adipocytes.

Interestingly, LTBP1 deficiency resulted in a significant increase in OCR after cAMP

stimulation (Figure 15A). Basal respiration before (1.2-fold) and after cAMP injection (1.2-

fold) was significantly increased. However, there were no significant changes in UCP1

mRNA and protein expression or on the mitochondrial content in LTBP1-deficient cells.

Diminishing LTBP4 had no effect on the OCR in SGBS adipocytes (Figure 15D).

To address changes in ATP production and proton leak of LTBP-deficient cells compared to

EV, cells were treated with oligomycin (Oligo) which inhibits the ATP synthase. The

corresponding OCR values of LTBP3-deficient cells after oligomycin treatment were

significantly lower, indicating a significantly lower ATP production (11%) (Figure 15C). In

contrast, ATP production was 1.4-fold higher in LTBP1-deficient cells compared to EV cells

62

(Figure 15A). Moreover, FCCP treatment led to complete uncoupling of the proton motive

force and resulted in maximum respiration (Max. resp.). Only LTBP1-deficient cells showed

alterations upon FCCP treatment. The OCR was 1.1-fold higher compared to control

adipocytes, suggesting an increased mitochondrial content in these cells (Figure 15A),

which could not be confirmed by the gene expression data (Figure 13).

Summarizing these data, the deficiency of LTBP2 and LTBP3 exhibits a significant decrease

of UCP1 expression. Furthermore, metabolic function was changed inter alia by a lower

response to cAMP, supporting the gene expression data.

Figure 14: Schematic view of a seahorse plot. An example of the oxygen consumption rate (OCR) over time is shown in red. Through induction of lipolysis with dibutyryl-cyclic adenosine monophosphate (cAMP), inhibition of ATP synthase with oligomycin, full uncoupling of the proton motive force by carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP), and inhibition of the respiratory chain by antimycin A and rotenone, the non-mitochondrial oxygen consumption, basal respiration (resp.), cAMP response, ATP production, proton leak, and maximum respiration (Max. resp.) can be calculated. Colored boxes represent these values and demonstrate how these are calculated.

63

Figure 15: LTBP deficiency alters the metabolic function of SGBS adipocytes CRISPR/Cas9 mediated LTBP-deficient SGBS cells were subjected to adipogenesis. On day 14 of differentiation the oxygen consumption rate (OCR) was measured with an extracellular flux analyzer and normalized with Nile Red to the amount of lipid content. (A) LTBP1 KO, (B) LTBP2 KO (C), LTBP3 KO, (D) LTBP4. To calculate basal and maximum respiration, as well as ATP production, proton leak and cAMP response, cells were treated with dibutyryl-cyclic adenosine monophosphate (cAMP), oligomycin (Oligo), carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP) and antimycin/rotenone (Ant/Rot). Mean +SEM of n=4 (LTBP2, LTBP3, LTBP4) and n=5 (LTBP1) is shown. **p<0.01, ***p<0.001, students t-test vs. EV control.

64

3.1.7 LTBP3 correlates positive with UCP1 expression in human subcutaneous adipose

tissue

The in vitro data suggests a role for LTBP2 and LTBP3 in the regulation of brown marker

gene UCP1. To elucidate any connection of LTBPs to adipocyte browning in human WAT,

the UCP1 and LTBP expression was examined in human sc mamma adipose tissue of n=28

patients (age=45±15 years, BMI=28.2±4.9 kg/m²), which underwent plastic surgery. There

was no correlation of LTBP1, LTBP2 and LTBP4 mRNA expression with UCP1 (Figure 16).

However, a significant positive correlation for LTBP3 and UCP1 mRNA expression was

observed in human WAT (p=0.0382), using a linear regression.

Figure 16: LTBP3 correlates significant positive with UCP1 mRNA expression in human subcutaneous

adipose tissue. mRNA expression of LTBPs and UCP1 was determined in human sc adipose tissue samples

derived from mamma tissue of n=28 patients which underwent plastic surgery. LTBP mRNA expression was plotted against UCP1 mRNA expression and a linear regression was applied. Adipose tissue samples belong to our in-house biobank. Mean age±SD = 45±15 years, Mean BMI±SD = 28.2±4.9 kg/m². P<0.05 is considered as statistically significant.

65

3.2 The role of TGFβ in SGBS adipogenesis

The absence of LTBP2 and LTBP3 in SGBS adipocytes resulted in alterations in UCP1

expression and metabolic function, but the mechanism remained to be investigated. To

examine whether the observed effects of LTBP deficiency are dependent on TGFβ signaling,

it was investigated if the TGFβ ligands TGFβ1, TGFβ2 and TGFβ3 are co-expressed with

LTBPs and whether the TGFβ receptors are present in SGBS pre- and adipocytes.

3.2.1 TGFβ2 and TGFβR3 are co-expressed with LTBPs during adipogenesis of SGBS cells

The TGFβ isoforms are differentially regulated upon adipogenesis in SGBS cells

(Figure 16A). TGFβ1 is the most abundant ligand and downregulated during adipogenesis.

Interestingly, comparison of the expression data to that of the LTBPs during adipogenesis,

revealed an overlapping expression pattern only for TGFβ2 with a peak on day 4 of

adipogenesis (Figure 17A and 10C). In contrast, TGFβ3 expression was low in preadipocytes

with a steady increase until day 14. Tracing the receptor expression during the time course

of adipogenic differentiation revealed different regulation patterns in between the

receptor isoforms (Figure 17B). TGFβR1 expression was stable during the whole

adipogenesis, whereas the expression of TGFβR2 reached a peak on day two of

differentiation and decreased during adipogenesis until d14. Of note, TGFβR3, which is

exclusively involved in TGFβ2 signaling, was highly expressed on the first days of

adipogenesis, with similar expression to TGFβ2 and the LTBPs (Figure 17 and 10C).

Additionally, the expression of TGFβ ligands and their receptors was examined in hASCs

derived from the dn and sc adipose tissue (Figure 18). Of note, expression of ligands as well

as of receptors was not significantly altered comparing dn vs sc samples as it was seen

before with LTBP1, LTBP3, and LTBP4 (Figure 18 and Figure 8). In SGBS cells as well as in

primary cells, TGFβ1 is the highest expressed ligand (Figure 18). In contrast, TGFβ2 and

TGFβ3 were less expressed. TGFβ receptors were abundantly expressed in isolated hASCs.

In line with the data from SGBS cells, all three receptors and ligands are expressed in hASCs

to a similar extent as compared to SGBS preadipocytes (Figure 17 and 18).

66

Figure 17: TGFβ ligands and TGFβ receptors are expressed during adipogenesis of SGBS cells. SGBS preadipocytes (d0) were subjected to adipogenesis for 14 days until lipid laden adipocytes were observed (d14). (A) mRNA expression of TGFβ ligands and (B) TGFβ receptors (TGFβR) during adipogenic differentiation measured by qRT-PCR. Mean +SEM n=3; *p<0.05; **p<0.01;****p<0.0001; ANOVA vs. d0

Figure 18: TGFβ ligand and receptor expression of isolated hASCs from dn and sc adipose tissue Expression of TGFβ ligands and receptors was investigated in the samples which were previously isolated and used in the study by Tews et al.[47] (A) mRNA expression of TGFβ ligands and (B) TGFβ receptors (TGFβR) in isolated hASCs derived from dn and sc adipose tissue. Mean values +SEM of n=7 paired patient samples are shown.

It was reported that LTBP3 is able to bind all three TGFβ ligands and it was shown in several

studies that these ligands and their mediated pathway influences browning [88,96,102].

Since LTBP3 deficiency led to a decrease in UCP1 expression and all three ligands are

expressed in SGBS preadipocytes, we were interested whether the inhibition of the TGFβ

pathway results in a similar UCP1 expression compared to the LTBP3-deficient adipocytes.

Therefore, in a first step a knockout of TGFβ1, TGFβ2, and TGFβ3 at the same time was

generated in SGBS preadipocytes.

67

3.2.2 TGFβ1-3 deficiency leads to a decrease of UCP1 in SGBS adipocytes

After transfection of SGBS preadipocytes with the triple knockout construct, stable bulk

cultures were generated by puromycin selection. TGFβ mRNA was reduced by more than

75% in pre- and more than 60% in adipocytes (Figure 19A and B). Protein expression of

TGFβ1-3 was almost diminished in SGBS bulk cultures (Figure 19C). Inhibition of TGFβ

ligands by blocking antibodies or the knockout of a TGFβ pathway in vivo resulted in

browning of WAT [96,97,102]. Therefore, the expression of brown marker genes and the

metabolic function in EV control and TGFβ1-3-deficient SGBS adipocytes was examined in

this approach.

Figure 19: Transfection with CRISPR/Cas9 triple knockout construct leads to stable downregulation of TGFβ ligands in SGBS cells during adipogenesis. After stable transfection of SGBS preadipocytes with CRIPSR/Cas9 TGFβ1-3 knockout construct, the cells were controlled for TGFβ ligand specific knockout. (A) mRNA expression of TGFβ ligands in triple knockout and EV SGBS preadipocytes (B) and in adipocytes. Mean +SEM, fold change with n=4, ***p<0.001, ****p<0.0001, student’s t-test. (C) Representative Western blot of n=3 independent experiments with GAPDH as loading control. Mean +SEM, fold change, **p<0.01, ****p<0.0001, student’s t-test vs EV.

Microscopic pictures revealed no morphological changes of TGFβ1-3-deficient adipocytes

compared to EV (Figure 20A). Furthermore, counted differentiation rate (Figure 20B) and

lipid staining with Nile Red (Figure 20C) exhibited no differences as well. In line, the mRNA

expression of differentiation marker also showed no alterations between EV and TGFβ1-3-

deficient adipocytes (Figure 20D). In addition, mRNA expression of markers associated with

mitochondrial metabolism were not altered between EV control and TGFβ-deficient

adipocytes (Figure 20E). In line, there was no alteration of OXPHOS proteins upon TGFβ

deficiency as shown by Western blot analysis (Figure 20F). The activity of CS was

comparable between EV cells and TGFβ1-3 KO adipocytes (Figure 20G).

68

Figure 20: No alterations on adipogenic differentiation and mitochondrial content after CRISPR/Cas9 mediated TGFβ1-3 knockout: TGFβ1-3 deficiency was achieved by CRISPR/Cas9 in SGBS preadipocytes, which were subjected to adipogenic differentiation for 14 days. (A) Microscopic pictures of EV and TGFβ1-3-deficient SGBS adipocytes, 10-fold magnification. (B) Counted differentiation rate of mature adipocytes. (C) Lipids of mature adipocytes were stained with Nile Red and fluorescent intensity was measured at 550 nm excitation; fold change is shown. (D) mRNA expression of differentiation marker PPARG, Glut4 and Adiponectin. (E) Mitochondrial marker genes on d14 assessed by qRT-PCR. HPRT served as a reference gene. (F) Oxphos genes of the respiratory chain were assessed by Western blot. One representative is shown of 4 independent experiments. GAPDH served as loading control. (G) Citrate synthase activity was enzymatically determined. Mean +SEM, n=4.

Of note, investigation of UCP1 mRNA and protein levels revealed a decrease in expression

of 50% in SGBS adipocytes with TGFβ1-3 deficiency (Figure 21A and C). PGC1α is known to

be a marker for brown adipogenesis and proposed to be influenced by TGFβ signaling

[41,97] but its mRNA expression was not significantly altered between transfected SGBS

strains (Figure 21B). Furthermore, the oxygen consumption rate was determined with

injection of cAMP to induce lipolysis in EV and TGFβ1-3-deficient adipocytes (Figure 21D).

In contrast to my expectation, no significant alterations were observed, even after cAMP

69

injection. Calculation of cAMP response, proton leak or ATP production revealed

comparable results between EV and TGFβ1-3-deficient cells. Why differences in UCP1

expression did not lead to alterations in metabolic function is not clear at this point and

remains to be resolved.

Figure 21: Decreased UCP1 expression in TGFβ1-3 deficient SGBS adipocytes. After 14 day of adipogenesis PGC1α and UCP1 expression were determined in EV control and TGFβ1-3 KO SGBS cells and the oxygen consumption rate (OCR) was measured with an extracellular flux analyzer and normalized with Nile Red to the amount of lipid content. (A) mRNA expression of uncoupling protein 1 (UCP1) and (B) PGC1α. HPRT serves as a reference gene. (C) Representative Western blot is shown and densitometric analysis was performed in n=4 independent experiments, GAPDH serves as a loading control. (D) OCR plot over time. To calculate basal and maximum respiration, as well as ATP production, proton leak and cAMP response, cells were treated with dibutyryl-cAMP (cAMP), oligomycin (Oligo), carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP) and antimycin/rotenone (Ant/Rot). Mean +SEM of n=4, *p<0.05, **p<0.01, student’s t-test

70

3.2.3 Inhibition of TGFβ receptor 1 kinase leads to a decrease in UCP1 expression

Active TGFβ ligands act on the cells through receptor activation. Binding to phosphorylated

TGFβR2 on the cell surface enables the recruitment and transphosphorylation of TGFβR1

[91]. Thereupon activated TGFβR1 phosphorylates SMAD2/3 intracellularly, leading to

changes in gene expression. In this approach, TGFβ signaling is blocked during SGBS

adipogenesis with the compound SB 431542 (SB). SB inhibits TGFβR1 kinase activity

thereby preventing SMAD2/3 phosphorylation. At first, the optimal inhibitor concentration

was titrated by adding 0.5 ng/ml recombinant TGFβ1 (rTGFβ1) to SB treated SGBS

preadipocytes followed by measurement of SMAD2/3 phosphorylation (Figure 22A). A

concentration of 10 µM SB inhibited SMAD3 phosphorylation to a basal level and was

chosen for SGBS treatment on day 0 and day 4 of adipogenesis. Microscopic pictures of

mature adipocytes on day 14 revealed differences in the morphological appearance upon

SB treatment. Cells treated with TGFβR1 inhibitor were smaller in size and appeared more

rounded compared to vehicle control (Figure 22B), but this did not influence the counted

differentiation rate which was equal (Figure 22C). However, analyzing expression of

differentiation marker genes in SGBS adipocytes revealed significant changes in mRNA

expression after TGFβR1 inhibition (Figure 22D and E). PPARγ was 1.55-fold upregulated

and adiponectin was 2.2-fold higher. Interestingly, the mRNA expression of browning

marker UCP1 was decreased by 77% in SB treated adipocytes (Figure 22F) which could be

observed on the protein level as well. UCP1 was 78% lower in adipocytes treated with the

inhibitor (Figure 22G). In line with the mRNA data, PPARγ protein level was also higher

compared to vehicle control. These findings are in line with the results seen in TGFβ-

deficient SGBS cells, showing that the inhibition of the TGFβ-pathway in SGBS cells leads to

a repression of UCP1 expression.

71

Figure 22: Inhibition of TGFβ receptor1 kinase leads to significant alterations in differentiation marker and UCP1 expression. (A) Phosphorylation of SMAD2/3 in SGBS preadipocytes after treatment with different concentrations of SB followed by stimulation with or without recombinant TGFβ1 (rTGFβ1). α-Tubulin and unphosphorylated SMAD3 served as loading control. (B) SGBS wild type preadipocytes were subjected to differentiation and treated either with DMSO as a vehicle control or with 10 µM of the inhibitor SB 431542 (SB) on day 0 and day 4 of adipogenesis. Microscopic picture of SGBS adipocytes are shown with 4-fold magnification. (C) Counted differentiation rate. (D) mRNA expression of differentiation marker adiponectin, (E) PPARγ, and (F) browning marker UCP1 in mature adipocytes. HPRT served as a reference gene. (G) Representative blot of UCP1 and PPARγ protein expression. Densitometric analysis for UCP1 was done in three independent experiments. α-Tubulin served a loading control. Mean+SEM of n=3, **p<0.01, ****p<0.0001, student’s t-test

3.2.4 SMAD4 downregulation decreases UCP1 expression in SGBS adipocytes

To provide prove that TGFβ-signaling regulates UCP1 expression, the pathway at the level

of intracellular signaling was inhibited and a siRNA-mediated knockdown of SMAD4 was

performed. SGBS preadipocytes were transfected with a siRNA targeting SMAD4 (siSMAD4)

or a non-targeting control (NTC). A knockdown of SMAD4 by 62% was achieved after 48h

(Figure 23A) and cells were subjected to adipogenic differentiation.

72

Figure 23: Decrease of UCP1 expression and oxygen consumption rate upon SMAD4 knockdown. SGBS preadipocytes were transfected with siRNA targeting SMAD4 (siSMAD4) and a non-targeting siRNA control (NTC). After 48h cells were subjected to adipogenesis for 14 days and mature adipocytes were analyzed for UCP1 expression and the oxygen consumption was measured with an extracellular flux analyzer. (A) SMAD4 mRNA expression in siRNA transfected SGBS preadipocytes after 48h of transfection. HPRT served as a reference gene. (B) Microscopic pictures of mature adipocytes with 10-fold magnification. (C) Counted differentiation rate of siRNA transfected adipocytes. (D) mRNA of adipogenic differentiation marker PPARγ, GLUT4, adiponectin, and (E) browning marker UCP1. HPRT served as a reference gene. Mean+SEM for n=4 (F) Representative Western blot and densitometric analysis of UCP1 of n=3 independent experiments. GAPDH served as a loading control and PPARγ as differentiation marker. Mean+SEM for n=3. (G) Oxygen consumption rate (OCR) over time, which was normalized with Nile Red to the amount of lipid content. To calculate basal and maximum respiration, as well as ATP production, proton leak and cAMP response, cells were treated with dibutyryl-cAMP (cAMP), oligomycin (oligo), carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP) and antimycin/rotenone (AA/Rot). Dotted lines indicate treatment intervals. Mean +SEM of n=3. *p<0.05, **p<0.01, student’s t-test

There was no difference in cellular morphology (Figure 23B) and the rate of adipogenic

differentiation (Figure 23C), as well as mRNA expression of differentiation marker PPARγ,

GLUT4 and Adiponectin (Figure 23D) was comparable between SMAD4 knockdown and

controls. Interestingly, UCP1 mRNA expression and protein expression were significantly

reduced upon SMAD4 knockdown. A reduction of 93% UCP1 mRNA expression was

73

observed and 72% less protein after densitometric analysis of Western blot

(Figure 23E and F). Moreover, knockdown of SMAD4 resulted in a decrease of cellular

oxygen consumption measured with an extracellular flux analyzer (Figure 23G). ATP

production (25%) and response to cAMP injection (46%) was significantly decreased in

SMAD4 downregulated SGBS adipocytes.

3.3 TGFβ2 is reduced in LTBP2- and LTBP3-deficient cells

To implement whether LTBP deficiency leads to a lower TGFβ secretion, the supernatant of

the LTBP-deficient preadipocytes was analyzed for TGFβ content with ELISAs (Figure 24).

No significant alterations of TGFβ1 content in SGBS supernatant could be detected and

TGFβ3 was below the reliable detection limit. However, TGFβ2 was significantly decreased

(28%) in supernatant of LTBP3-deficient SGBS preadipocytes (Figure 24 B). Unexpectedly

this was also true for LTBP2-deficient cells (29%). As already mentioned previously, it was

reported that LTBP2 is not able to bind any TGFβ [104].

Figure 24: LTBP2 and LTBP3 deficiency alters the TGFβ2 content in SGBS preadipocyte supernatant TGFβ content in supernatant of LTBP deficient SGBS preadipocytes and empty vector control (EV) was measured with an enzyme-linked immunosorbent assay. (A) TGFβ1 content(pg/ml) and (B) TGFβ2 content (pg/ml). TGFβ3 was below the reliable detection limit. Mean+SEM for n=3, *p<0.05 vs EV, student’s t-test

3.4 TGFβ2 deficiency leads to alterations in UCP1 and metabolic function

Both, LTBP2- and LTBP3-deficient cells showed a similar reduction of TGFβ2 content in the

supernatant. Furthermore, both had a decreased UCP1 expression and a significant

reduced response to cAMP. It has been reported earlier that in contrast to TGFβ1 and

TGFβ3, administration of recombinant TGFβ2 increased UCP1 expression in murine brown

adipose tissue in vivo and in immortalized brown adipocytes in vitro [88].

74

Figure 25: TGFβ2 deficiency leads to a higher differentiation rate TGFβ2 deficiency was achieved by CRISPR/Cas9 in SGBS preadipocytes, which were subjected to adipogenic differentiation for 14 days. (A) mRNA expression of TGFβ2 in knockout and empty vector (EV) SGBS preadipocytes (B) and in adipocytes. (C) Representative Western blot of n=4 independent experiments with GAPDH as loading control. (D) Microscopic pictures of EV and TGFβ2-deficient SGBS adipocytes, 10-fold magnification. (E) Counted differentiation rate and (F) triglyceride content of mature adipocytes. (G) mRNA expression of differentiation marker PPARG, Glut4 and Adiponectin. Mean +SEM with n=4, **p<0.01, student’s t-test.

To study the role of TGFβ2 in this context in more detail, CRISPR/Cas9 mediated TGFβ2-

deficient SGBS cells were generated. After stable transfection and puromycin selection,

TGFβ2 mRNA expression was significantly decreased in SGBS pre- (67%) and adipocytes

(70%) compared to EV control (Figure25A and B). In addition, TGFβ2 protein was

significantly reduced in SGBS preadipocytes (Figure 25C). After 14 days of differentiation

microscopic pictures and counted differentiation rate of mature adipocytes revealed

(Figure 25D) a significant higher differentiation for TGFβ2-deficient adipocytes (Figure 24E).

However, no significant differences were observed measuring triglyceride content and

mRNA expression of differentiation marker (Figure 25F and G).

75

Figure 26: TGFβ2 deficiency leads to a reduction of UCP1 and a decreased response to cAMP. After 14 days of adipogenesis UCP1 expression was determined in EV control and TGFβ2-deficient SGBS cells and the oxygen consumption rate (OCR) was measured with an extracellular flux analyzer and normalized with Janus Green. (A) mRNA expression of UCP1 was determined by q-RT-PCR. HPRT serves as a reference gene (B) Representative Western blot is shown and densitometric analysis was performed in n=4 independent experiments, GAPDH serves as a loading control. (C) OCR over time of n=3 individual experiments. To calculate basal and maximum respiration, as well as ATP production, proton leak and cAMP response, cells were treated with dibutyryl-cAMP (cAMP), oligomycin (Oligo), carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP) and antimycin/rotenone (Ant/Rot). Mean +SEM of n=4, *p<0.05, student’s t-test.

Despite a higher differentiation rate of TGFβ2-deficient cells, the UCP1 expression was

decreased compared to EV control. mRNA expression was altered but did not reach

statistical significance (p=0.053) (Figure 26A). However, protein expression was

significantly reduced by 35% in TGFβ2-deficient SGBS adipocytes (Figure 26B). In line with

that, the response to cAMP and the basal respiration after cAMP induction was significantly

reduced (Figure 26D). Moreover, the maximum respiration obtained after FCCP

administration was reduced as well in TGFβ2-deficient cells compared to EV control. TGFβ1

deficiency did not alter UCP1 expression or metabolic function in SGBS cells (Appendix

Figure S4).

76

3.5 TGFβ2 correlates significantly positive with LTBP3 and UCP1 expression in human

subcutaneous adipose tissue

To elucidate if the impact of LTBP3 on the regulation of UCP1 mediated by TGFβ2

bioavailability and signaling may also play a role in human WAT, the correlation of LTBP3

expression with the TGFβ ligands and receptors in human sc WAT was examined. LTBP3

correlated positively with TGFβ2 (R²=0.2291, p=0.01) but not with TGFβ1 or TGFβ3 mRNA

expression (Figure 27A). In addition, none of the other LTBPs which bind TGFβ correlated

with any ligand (Appendix, Table S2). Moreover, there was a statistically significant positive

correlation of LTBP3 with all three receptor isoforms with R²=0.4641 for TGFβR1

(p<0.0001), R²=0.326 for TGFβR2 (p=0.0015) and R²=0.5053 for TGFβR3 (p<0.0001) (Figure

27B). Of note, TGFβ2 mRNA expression correlated significantly positive with UCP1 with

R2 = 0.8735 (p<0.0001) (Figure 27C).

Figure 27: TGFβ2 mRNA expression correlates positive with UCP1 in human subcutaneous adipose tissue samples. mRNA expression of LTBP3, UCP1, TGFβ ligands and receptors was determined in human sc adipose tissue samples derived from mamma tissue of n=28 patients which underwent plastic surgery. (A) Correlation of LTBP3 mRNA expression to TGFβ ligand and (B) to TGFβ receptor (TGFβR) expression. (C) Correlation of TGFβ mRNA expression and UCP1 mRNA expression. Adipose tissue samples belong to our in-house biobank. Mean age±SD = 45±15 years, Mean BMI±SD = 28.2±4.9 kg/m². p<0.05 is considered as statistically significant.

77

4. Discussion

Browning of white adipose tissue in vivo and activating brown adipose tissue in humans is

associated with weight loss and an overall better metabolic health [17,29,32]. The switch

from a white to a brown adipocyte phenotype by inhibiting or stimulating targets involved

in this process, is a new approach for the treatment of obesity and its related metabolic

disorders [16,25,60,61]. How and why ASCs within the adipose tissue differentiate either

into brown/beige or white adipocytes has not yet been resolved completely. Therefore, it

is important to identify new factors which are involved in this process. Furthermore, these

factors could potentially represent future targets for the treatment of obesity.

Our group has shown that in vitro differentiated adipocytes derived from isolated ASCs

from human dn highly express brown marker genes compared to the adipocytes derived

from ASCs isolated from the sc adipose tissue [47]. Furthermore, Tews et al. have

demonstrated that the gene expression of these ASCs is differently regulated dependent

on the adipose tissue’s origin [47]. Through ranking of these differentially regulated genes

and literature comparison, the LTBP family was identified as potential factors involved in

the regulation of brown marker gene UCP1 [85]. Since there is nothing known to date about

the role of LTBPs in the context of adipose tissue browning, the aim of this thesis was to

study the role of LTBP members in adipogenesis and investigate their impact on the UCP1

expression and on the metabolic function in vitro.

4.1 LTBP2 and LTBP3 deficiency leads to a decrease of UCP1 expression in SGBS cells

Nothing has been described so far about the role of LTBPs in adipose tissue. It was reported

that LTBPs are highly expressed in mesenchymal stem cells from adipose tissue but their

role in the adipogenesis towards a white or brown adipocyte phenotype has not been

addressed so far [108]. LTBP1 was 2-fold higher in dn hASCs compared to sc in the

microarray analysis by Tews et al. [47]. This was verified in this study, with a statistically

significant difference of LTBP1 in dn hASCs compared to sc. In addition, investigation of the

LTBP3 and LTBP4 expression exhibited a significant higher level in dn hASCs.

The human SGBS cell strain served as a model to study the impact of LTBPs on UCP1

regulation and on the metabolic function in vitro. Therefore, it was a prerequisite, that all

78

four LTBP member are expressed during adipogenesis. Importantly, expression levels of

LTBPs in human SGBS pre- and adipocytes were comparable to isolated and in vitro

differentiated hASCs. Based on this, the SGBS cells were considered as a suitable model for

the in vitro studies. The high expression of LTBPs in adipose tissue samples and the

regulation during adipogenesis in in vitro differentiated human primary cells and SGBS cells

suggests that LTBPs may play an import role in the regulation of adipogenesis, especially in

the first days of differentiation. Furthermore, the higher expression of LTBP1, LTBP3, and

LTBP4 in hASCs of the dn compared to sc may indicate a role for these isoforms in the

adipogenesis towards a brown adipocyte phenotype.

In this study LTBP-deficient SGBS preadipocytes were introduced to examine possible

effects on adipogenic differentiation, brown marker genes or metabolic function. Indeed,

changes in UCP1 expression and metabolic function were observed in LTBP2- and LTBP3-

deficient SGBS adipocytes, whereas the adipogenic differentiation was not affected.

Deficiency of LTBP1 in SGBS cells showed no consistent results on expression of UCP1 and

on a functional level. If LTBP1 was involved in generating a brown adipocyte phenotype,

one might have expected a decrease in UCP1 after LTBP1 deficiency in SGBS cells, based on

its higher expression in dn hASCs compared to sc. However, no significant changes were

observed in UCP1 expression in LTBP1-deficient SGBS adipocytes. Furthermore, functional

analysis revealed even a higher OCR after cAMP induction and a higher maximum

respiration. Small changes in a higher differentiation rate in LTBP1-deficient cells may be

the reason, indicated by a significant higher adiponectin expression. However, this was not

supported by the differentiation rate, or triglyceride content. Interestingly, the knockout

of LTBP1 in vivo is lethal, due to reduced TGFβ bioavailability and the phenotype was

overlapping with the TGFβ1 knockout mouse [104]. LTBP1 is known to bind all three TGFβ

isoforms, but measurement of secreted TGFβ in SGBS preadipocytes revealed no significant

differences between control and LTBP1-deficient cells. This either indicates that LTBP1 is

not important for TGFβ secretion in SGBS preadipocytes or LTBP3 and LTBP4 are

compensating for the loss of LTBP1. It seems that LTBP1 may have other roles in hASCs than

promoting a brown adipocyte phenotype. LTBP1 was shown to be a regulator of TGFβ in

late development, in the mineralization and maturation phase of osteogenic differentiation

in vitro, whereas LTBP3 played a role in early commitment [104]. This may indicate that

LTBP1 also does not play an important role in early adipogenesis.

79

In summary, LTBP1 deficiency did not alter UCP1 expression and the results did not support

a role for LTBP1 in stimulating an adipogenesis towards a brown adipocyte phenotype.

LTBP4 can only bind TGFβ1 but with low affinity [104,110]. LTBP4 deficiency did not

influence the TGFβ1 secretion in vitro, indicating a nonessential role in this context. LTBP4

seems to play a subordinary role in SGBS cells, as it was little expressed in preadipocytes as

well as in adipocytes and the deficiency had no effect on differentiation rate or UCP1

expression. Interestingly, LTBP4 expression was 100-fold higher in human WAT samples

compared to isolated preadipocytes, which indicates that LTBP4 is highly expressed by

other cell types like macrophages, fibroblasts or endothelial cells in the adipose tissue.

LTBP2 was equally expressed in hASCs of dn and sc adipose tissue. However, deficiency of

LTBP2 in SGBS cells led to an inhibition of UCP1 expression and function. In contrast, protein

expression showed no differences. LTBP2 is the only LTBP isoform which does not bind any

TGFβ ligand and was described to be highly expressed in elastic tissues [104,110].

Unexpectedly, LTBP2 deficiency revealed a reduction of secreted TGFβ2. As well as other

LTBPs, LTBP2 is secreted and associates with the extracellular matrix. It interacts with

fibrillin-1 in the ECM and shares the same binding site of fibrillin-1 as LTBP1 [121,122]. Both

are competing with comparable affinity for fibrillin-1 binding. Thus, it is suggested that

LTBP2 can indirectly regulate the activation of TGFβ by preventing LTBP1 matrix

incorporation [104]. If these properties led to the lower TGFβ2 levels in LTBP2-deficient

SGBS preadipocytes remains unclear.

This study for the first time demonstrates that the LTBP2-deficiency has a significant impact

on the UCP1 expression and metabolic function in SGBS cells. A possible mechanism is via

TGFβ2 bioavailability, which is discussed later (see 4.3). Another mechanism could be the

inhibition of BMP11 and myostatin [123,124]. LTBP2 was shown to inhibit proprotein

convertase 5/6A (PC5/6) by binding to it. PC5/6 cleaves pro-BMP11 into the mature BMP11.

Furthermore, LTBP2 can also bind to pro-myostatin, thereby inhibiting its secretion. Both,

myostatin and BMP11 have been shown to inhibit UCP1 expression in vitro and in vivo

[36,147].

Our hypothesis is that secreted factors act in an auto- or paracrine manner on hASCs,

stimulating or repressing an adipogenesis towards a brown adipocyte phenotype. The

significantly higher expression of LTBP3 in the dn hASCs compared to sc supports this thesis

80

and suggests a role for LTBP3 in generating a brown adipocyte phenotype. Furthermore, in

another work to study the characterizations of hASCs of different adipose tissue origins,

LTBP3 was significantly higher expressed in hASCs from supraclavicular compared to sc

adipose tissue [85]. Interestingly, Dabovic et al. published a study where they determined

which organs might be affected by LTBP3 in early murine embryonic development [148]. In

situ hybridization revealed that LTBP3 was only detected in four organs of 16.5 days old

murine embryos, inter alia in the BAT. This again supports the notion that LTBP3 plays a

role in an adipogenesis towards a brown adipocyte phenotype.

The published LTBP3 knockout mouse had premature ossification of the skull base or

increased bone density indicating a TGFβ deficiency, but data on the effects on adipose

tissue had not been reported [126,127]. In this thesis, it was demonstrated that LTBP3 has

a significant impact on the expression of key brown marker gene UCP1. In contrast to all

other LTBP isoforms, LTBP3 deficiency in SGBS cells showed a decrease in UCP1 mRNA as

well as in protein expression by over 35%. Furthermore, the metabolic function was altered

after cAMP injection, supporting the expression data. It was published that LTBP3 is an

important regulator of TGFβ bioavailability and can bind all three TGFβ isoforms [104]. Until

now there is no extracellular TGFβ-independent function known for LTBP3, as it was shown

for LTBP1, LTBP2 and LTBP4 [104]. In addition, LTBP3 was reported to be only secreted

when bound to a TGFβ isoform [149]. Indeed, significantly lower TGFβ2 levels were

observed in the supernatant of LTBP3-deficient preadipocytes, which may lead to a

repressed UCP1 expression in adipocytes. In addition, it was shown that LTBP3 can also

bind to pro-myostatin and inhibit the proprotein convertase 5/6A (PC5/6) in the

endoplasmic reticulum [123,124]. If a LTBP3 deficiency in SGBS cells could lead to a higher

BMP11 and myostatin secretion and if this has an impact on the UCP1 expression remains

to be investigated.

In summary these results revealed that LTBP deficiency has no effect on the adipogenic

differentiation in SGBS cells. In addition, UCP1 expression was not altered in LTBP1 and

LTBP4-deficient SGBS cells. Moreover, TGFβ levels in SGBS preadipocyte supernatant of

LTBP1- and LTBP4-deficient SGBS preadipocytes was not reduced, indicating a dispensable

role for TGFβ bioavailability in SGBS cells. However, deficiency of LTBP2 and LTBP3 in SGBS

cells led to a significant decrease of UCP1 and a regulation of metabolic function.

81

Supernatant of both, LTBP2- and LTBP3-deficient SGBS preadipocytes revealed a significant

reduction of TGFβ2 levels.

These data indicate that LTBP2 and LTBP3 are involved in promoting a brown adipocyte

phenotype.

4.2 Inhibition of the TGFβ pathway leads to a decrease of UCP1 in SGBS adipocytes

LTBPs belong to the extracellular matrix proteins and are therefore secreted by cells. Except

for LTBP2, they are known to bind TGFβ ligands, which are then secreted by the cell as a

LTBP-TGFβ complex. Therefore, LTBPs are crucial for the regulation of TGFβ bioavailability.

Animal models of TGFβ1, TGFβ2 or TGFβ3 deficiency were reported to be lethal, which

hampered a detailed characterization of their function in vivo [89]. While information on

the role of LTBPs in adipocyte biology is scarce, the TGFβ pathway has been shown to play

a fundamental role in adipogenesis and adipocyte metabolism. Clinical studies revealed

that high TGFβ1 serum levels are associated with obesity and a higher BMI [96,97].

Furthermore, it has been shown in several studies that the TGFβ pathway has an influence

on white adipocyte browning in vivo and in vitro [97,98,102]. SMAD3 KO in mice resulted

in WAT browning and SMAD3 KD in murine 3T3-L1 adipocytes increased UCP1 expression.

On the other hand, TGFβ2 has been shown to induce UCP1 expression in murine BAT in

vivo and in brown adipocytes in vitro [88].

To study the effect of TGFβ on the expression of brown adipocyte marker in SGBS cells, the

TGFβ pathway was inhibited in three independent approaches. First, TGFβ1-3-deficient

preadipocytes were generated. Second, the pathway was inhibited by blocking SMAD2/3

phosphorylation with a TGFβR1 kinase inhibitor. Third, SMAD4 was silenced in SGBS

preadipocytes. Of note, TGFβ1-3-deficient SGBS adipocytes revealed a significantly lower

UCP1 expression compared to the EV control. In line, the inhibition of TGFβR1 kinase in

SGBS cells resulted in a significant reduction of UCP1 mRNA and protein as well. Moreover,

to inhibit the TGFβ SMAD2/3 axis downstream, a siRNA-mediated SMAD4 knockdown was

performed in SGBS preadipocytes. This led to a significant decrease in UCP1 expression of

over 90% on mRNA and over 70% on protein level and also inhibited UCP1 on a functional

level. These results demonstrate that inhibition of the TGFβ signaling pathway represses

UCP1 expression and function in SGBS cells.

82

Interestingly, in vitro studies showed that treatment with recombinant TGFβ resulted in an

inhibition of differentiation and treatment with TGFβR1 kinase inhibitor SB431542 in an

increase in differentiation [36,102]. Unexpectedly, no effects on the differentiation rate

were observed in all three approaches. However, the treatment with TGFβR kinase

inhibitor resulted only in an increase in differentiation marker on mRNA and protein level,

not on adipocyte number. On the other hand, treatment of SGBS cells with recombinant

TGFβ1 resulted in an inhibition of differentiation (Appendix Figure S3). It has been reported

that phosphorylation of SMAD2 and SMAD3 by TGFβR activation leads to a repression of

adipogenic transcription factors. Yadav et al. showed that phosphorylated SMAD3 interacts

with C/EBPβ and C/EBPδ thereby inhibiting the induction of transcription factor PPARγ [97].

Activation of pSMAD2 resulted in a decreased expression of C/EBPα and PPARγ, whereby

the mechanism is unclear. It is conceivable that there was no effect on the differentiation

rate, as SGBS cells are already maximally differentiated under basal conditions [143]. This

is inter alia due to the strong PPARγ agonist rosiglitazone which is contained in the

differentiation medium.

In this study a significant reduction of UCP1 expression was demonstrated upon TGFβ1-3

deficiency. Unexpectedly, the metabolic function was not altered after cAMP injection. The

reason for this remains unclear. The UCP1 protein expression was reduced in all 4 individual

experiments on mRNA and protein level. Moreover, the data of TGFβR inhibition and

SMAD4 KD supported the observed decrease in UCP1 expression.

A repression of UCP1 due to TGFβ inhibition contrasts with the published in vivo and in vitro

results. Yadav et al. inhibited all three TGFβ isoforms in vivo and investigated the effects on

the adipose tissue and overall metabolism [97]. For this purpose, murine models for obesity

(Lepob/ob) and type 2 diabetes (DIO) were treated intraperitonally with an antibody (AB)

capturing TGFβ1-3, and therefore preventing receptor activation. Interestingly, TGFβ1-3 AB

treated mice had lower fat mass, lower level of triglycerides and improved glucose and

insulin tolerance compared to control-AB treated mice. Moreover, these mice showed

increased mitochondrial DNA content and increased levels of BAT markers including UCP1

in WAT. There was nothing reported on any side effects. Furthermore, in vitro studies using

TGFβR1 kinase inhibitor SB431542 during adipogenic differentiation exhibited an increased

UCP1 expression in murine stroma vascular cells from WAT and in human induced

83

pluripotent stem cells [102,150]. It was also shown that the usage of SB431542 leads to a

higher number of differentiated cells, which would explain a higher UCP1 expression [150].

However, significantly lower UCP1 levels were observed upon SB431542 treatment in SGBS

cells compared to the vehicle control and at the same time a significant higher expression

of differentiation markers. This indicates that the inhibition of UCP1 seems not to be an

effect of differentiation on the SGBS cells, rather a specific effect. Interestingly, Takeda et

al. described a similar effect on UCP1 expression [151]. They stimulated dermal fibroblasts

with an “optimized chemical cocktail” containing inter alia SB431542 to differentiate these

cells into chemical induced brown adipocytes (ciBAs). These mature ciBAS expressed basal

levels of UCP1. However, when SB was omitted in the differentiation cocktail, the UCP1

expression was significantly higher (⁓5-fold). They speculated that this is due to a higher

number of “adipocyte like” cells. A potential mechanism which leads to these observations

was not discussed. In this work, there were no differences observed in adipocyte number

of SB and control treated SGBS cells.

The TGFβR activation leads to phosphorylation of SMAD2/3, which then bind to SMAD4 for

the internalization into the nucleus [91]. In vivo, the SMAD4 knockout was reported to be

lethal in early embryonal stage, which makes it difficult to study the exact role of SMAD4

in adipocyte browning [152]. However, SMAD3 knockout in vivo is not lethal and its impact

on the metabolism and on the expression of brown marker genes has been studied as well

by Yadav et al. [97]. Mice with a homozygous knockout for SMAD3 had a reduced fat mass

and higher glucose uptake in WAT. Furthermore, the WAT of SMAD3-deficient mice

showed a morphology similar to that of BAT, with smaller and unilocular adipocytes. These

observations were supported by a higher UCP1 mRNA expression and protein level in WAT

compared to WT mice. In addition, 3T3-L1 cells expressing a short hairpin RNA against

SMAD3 showed elevated levels of brown marker genes in vitro compared to the control

[97].

Summarizing the results to study the impact of the TGFβ-SMAD pathway on brown marker

genes and on the metabolic function in SGBS cells revealed a significant decrease in UCP1

expression in all three approaches. Interestingly, these are the opposite results compared

to the published literature of in vitro and in vivo experiments. One can speculate if this is

due to the differences between human and mouse or in vivo and in vitro. It has been

demonstrated that TGFβ is also highly expressed and secreted by other cell types residing

84

in the adipose tissue, such as macrophages and leukocytes, thus influencing the

adipogenesis as well [89]. This exogenous TGFβ is absent in our in vitro model.

Furthermore, a SMAD4 knockdown could not only affect TGFβ action, but also diminish the

effect of BMP4 and BMP7 on promoting a brown adipocyte phenotype. The intracellular

signaling of BMPs, which belong to the TGF superfamily, is mediated via phosphorylation

of SMAD1, SMAD5 and SMAD8 which also require SMAD4 for the internalization into the

nucleus [91]. In vivo and in vitro experiments have demonstrated that BMP4 and BMP7 can

induce adipogenesis towards a brown adipocyte phenotype [36]. On the other hand, this

was not supported by the publication of Modica et al. [153]. They performed a SMAD4

knockdown in vitro in immortalized murine brown adipocytes which resulted in an increase

in UCP1.

Yadav et al. have also shown that TGFβ signaling can repress the expression of PGC1α via

promotor regulation, speculating that this inhibits both, differentiation and browning [97].

No differences in the PGC1α expression of TGFβ-deficient cells versus EV control were

observed in this work, proving at least that there was no repression of transcription factor

PGC1α. The reasons why the inhibition of the TGFβ-SMAD axis led to an inhibition of UCP1

in SGBS cells remains unclear at this point. If this is an SGBS cell specific effect needs to be

investigated by repeating the same approaches in other hASCs derived from the sc WAT.

However, in this work, it was clearly demonstrated in three independent approaches that

inhibition of the TGFβ pathway represses UCP1 expression in human SGBS cells in vitro.

This data demonstrates new insights in the context of TGFβ and the regulation of key

browning marker UCP1.

Comparing the observations of TGFβ pathway inhibition in SGBS cells with the LTBP-

deficient adipocytes reveals overlapping results in UCP1 expression on mRNA and protein

level with LTBP3-deficient cells. Furthermore, the supernatant of LTBP3-deficient

preadipocytes exhibited a significant reduced level of TGFβ2. These results support a role

for LTBP3 in generating a brown adipocyte phenotype and suggests that the effects on

UCP1 and metabolic function seen in LTBP3-deficient SGBS cells are mediated via the TGFβ-

pathway.

85

4.3 Repression of brown marker gene UCP1 by LTBP3-deficiency could be mediated by

alterations in TGFβ2 bioavailability

TGFβ2 levels were significantly reduced in LTBP3-deficient preadipocytes. Thus, this work

focused on this ligand as a potential factor mediating an adipogenesis towards a brown

adipocyte phenotype.

Of note, comparing the TGFβ ligand and receptor expression during SGBS adipogenesis to

the expression pattern of LTBPs, revealed a co-regulation only for TGFβ2 and TGFβR3 with

the LTBPs. This suggests a special role for TGFβ2 in the adipogenesis of SGBS cells. Indeed,

TGFβ2 deficiency in SGBS cells revealed a significant reduction of UCP1 on protein level by

35%. In addition, induction of lipolysis by cAMP exhibited a significant lower OCR in TGFβ2-

deficient adipocytes compared to control cells. This observation supports a reduction of

UCP1 in TGFβ2-deficient adipocytes. Of note, TGFβ1 knockout did not alter UCP1

expression or metabolic function in SGBS cells. Recently, Takahashi et al. published a study

on TGFβ2 and its impact on the metabolism in vitro and in vivo [88]. They found that the

secretion of TGFβ2 in adipocytes from the sc WAT is induced by exercise and that TGFβ2

can regulate the metabolism in mice. In vitro treatment with recombinant TGFβ2 resulted

in a higher fatty acid uptake and glucose uptake inter alia in 3T3-L1 adipocytes and in

human immortalized brown adipocytes (WT-1) [88]. Furthermore, UCP1 expression was

significantly increased in WT-1 adipocytes. This was not observed for TGFβ1 and TGFβ3.

Moreover, mice treated with a β-3 receptor agonist showed a significant reduction of

TGFβ1 and TGFβ3 expression in sc WAT, whereby TGFβ2 was not altered. Mice challenged

with HFD and infused with recombinant TGFβ2 by an osmotic pump showed an overall

improved metabolism including a better glucose tolerance and insulin sensitivity compared

to control mice on HFD [88]. It was shown that mainly muscle tissue was responsible for

these improvements. However, TGFβ2 treatment also significantly increased UCP1

expression in BAT. In addition, they have shown that TGFβ2 inhibits macrophage infiltration

into the scWAT and hypothesized that TGFβ2 acts as in immune suppressor [88]. They

suggest that this reduction of inflammation in scWAT led to an improved overall

metabolism in TGFβ2 treated mice. Unfortunately, they did not discuss the observed in

vitro results or a potential mechanism of TGFβ2 on the UCP1 expression [88]. In addition,

they did observe different results for each TGFβ ligand, but did not comment on the special

role of TGFβ2, since the same pathways are assumed for all three ligands. However, there

86

are some different properties for TGFβ2 compared to TGFβ1 and TGFβ3. In contrast to the

other TGFβ isoforms, TGFβ2 has a low affinity for TGFβR2, thus it binds first to the co-

receptor TGFβR3 which then activates TGFβR2 [91]. Moreover, it is not completely

understood how TGFβ2 is activated in the ECM. It was demonstrated that LTBP1 and LTBP3

can bind this isoform and that cells secrete this complex into the ECM [104]. TGFβ1 and

TGFβ3 can be released from the LLC by integrin binding but the pro-TGFβ2 is missing this

binding sequence making an integrin-mediated activation in the ECM unlikely [104].

Therefore, other factors or circumstances like pH change or proteases must lead to TGFβ2

release in the ECM. Interestingly, Petrus et al. exhibited another difference to the other

isoforms. TGFβ1 and TGFβ3 expression was significantly higher in human sc WAT of an

obese cohort compared to control [154]. However, TGFβ2 expression was not altered due

to obesity or after weight loss of this cohort. These data may explain differences in function

of TGFβ2 compared to TGFβ1 and TGFβ3, but the mechanism remains unknown. Another

possibility for the lower UCP1 expression of TGFβ2-deficient SGBS cells is the influence of

the p38 MAPK pathway. It was shown that p38 MAPK can be phosphorylated by TGFβR

activation [91]. In different studies, it has been shown that the activation of p38 e.g. by

natriuretic peptides, stimulates UCP1 expression [24,40]. Whether this has an impact in our

cell system remains to be investigated. A specific pathway for TGFβ2 distinct of TGFβ1 or

TGFβ3 is not known yet.

My observations and the published effects of TGFβ2 on the adipocyte metabolism indicate

that LTBP3 regulates UCP1 expression in an auto- or paracrine manner via control of TGFβ2

bioavailability [88]. However, the mechanism by which UCP1 is regulated is unknown and

remains to be determined.

4.4 Indications for LTBP3-TGFβ2 mediated browning in human WAT

The overall idea of this study was to identify potential new target structure for the

treatment of obesity and its associated cardiovascular diseases. Therefore, it was

important to show if the observed LTBP3-TGFβ2 axis may also play a role in browning of

human sc WAT. Correlation of mRNA expression analysis in WAT of patients who

underwent plastic surgery revealed interesting results. First, the LTBP3 expression

correlated positive with TGFβ2 expression. Of note, LTBP3 did not correlate with any other

87

TGFβ ligand in those patient samples. This indicates an important relation of LTBP3 with

TGFβ2 in human adipose tissue, since it was shown that LTBP3 secretion is dependent on

the expression of TGFβ [149]. Even more important, the LTBP3 expression correlated

significantly and positively with the UCP1 expression in these WAT samples. No publication

was found on a negative correlation of LTBP3 to obesity, insulin resistance or any disease

which could be associated with low UCP1 levels. McInerney-Leo et al. reported on patients

with heterozygous mutations in the LTBP3 gene [155]. These subjects suffered from

acromicric dysplasia or geleophysica dysplasia, including a shorter statue or brachydactyly.

It was reported that these patients had normal proportions to their height. Nothing was

described in detail on BMI, body fat mass or insulin resistance. It seems that LTBP3 plays

an important role in the early development of humans since some of these patients also

died in early childhood. In line to the previously described LTBP3 knockout mouse, LTBP3

takes a crucial role in the early bone formation [126,127]. However, the role of LTBP3 in

WAT may be different and has not been elucidated to date. As previously described, the

TGFβ2 knockout mouse is lethal due to insufficient development of vital organs in the

embryonal stage [94]. Humans with heterozygous TGFβ2 mutations suffered inter alia from

Loeys-Dietz syndrome, which is often characterized by a higher TGFβ signaling, higher

SMAD2/3 phosphorylation and aortic aneurysm [156]. These observations are often seen

within mutations of the TGFβ isoforms and are also known from the Marfan syndrome. The

complete mechanism is not understood, but it is assumed that the other TGFβ isoforms are

overcompensating the loss of functional TGFβ2 protein. As previously described, TGFβ2

treatment in vivo led to an increase in UCP1 in BAT and to an improved overall metabolism

[88]. The authors of this study did not observe any side effects after 13 days of treatment.

In the same study they also performed an adipocyte specific TGFβ2 knockout in mice. There

were no changes in the glucose tolerance test compared to control mice. Unfortunately,

they did not show the UCP1 expression in the WAT of these mice or report anything on the

context of white adipocyte browning. However, in this thesis, a significant correlation of

TGFβ2 expression to UCP1 was found in human WAT samples with R²=0.8735. This supports

the in vitro observations and the hypothesis that LTBP3 alters UCP1 expression via TGFβ2

and suggests a role for LTBP3 in WAT browning in vivo.

88

4.5 Conclusion, limitations and outlook

This study has demonstrated that the LTBP3-TGFβ2 pathway is involved in the adipogenesis

towards a brown adipocyte phenotype in vitro and may also be an important axis in human

WAT browning. However, these findings also reveal some limitations and open questions

which should be addressed in future experiments. Not much is known to date about LTBP3

in the context of adipose tissue and the release of TGFβ2 by LTBPs, but insight in its

pathway could reveal new targets to stimulate the adipogenesis towards a brown

adipocyte phenotype. The same pathways are assumed for all three TGFβs, but the ligands

have already been shown to have different functions [87]. Especially in the context of

adipose tissue browning, it would be interesting to reveal any new pathways. RNAseq or

proteomics analysis in the LTBP3- and TGFβ2-deficient SGBS cells may detect any

differences in pathway activation. Another approach would be the in vitro treatment with

each TGFβ ligand followed by RNAseq or proteomics.

Despite the new findings in this study it should be considered that all approaches were

done in vitro in a 2-dimensional cell culture model only using SGBS cells. The adipose tissue

is a highly complex organ with several different cell types [5]. Especially immune cells with

the secretion of cytokines are known to have an impact on the development and expression

profile of pre- and adipocytes [157].

In this thesis, the focus was on LTBPs, the TGFβ-pathway, and the cytokine TGFβ2. LTBPs

are known to be matrix proteins, keeping TGFβs there in a latent state [104]. The SGBS cells

express and secrete extracellular matrix proteins in vitro [158], but whether this is enough

for adequate LTBP binding and e.g integrin-activated TGFβ secretion remains to be

investigated. Furthermore, LTBPs and TGFβ isoforms are also known to be expressed by

other cells residing in the adipose tissue [89,108]. My measurements revealed also higher

expression for LTBP3 and LTBP4 in adipose tissue samples compared to isolated hASCs. This

should be considered thinking of a potential role for LTBP3 in stimulating UCP1 expression

in vivo. The LTBP3 or TGFβ2, which is secreted by other cell types than ASCs may have an

impact as well. Therefore, an interesting approach would be an adipocyte specific knockout

for LTBP3 in mice. This could be achieved with a Cre-loxP system and under the control of

an adiponectin promoter. Challenging these mice with HFD, housing under cold conditions

or treatment with a β3-receptor agonist could give insights on the role of LTBP3 in WAT

and its impact on WAT browning and on the overall metabolism. The same approach would

89

be interesting for TGFβ2. As previously mentioned, an adipocyte specific knockout for

TGFβ2 has already been generated, but the authors focused more on the role of TGFβ2 in

exercise and on the overall metabolism of these mice [88]. Approaches with cold or β3-

adrenergic stimulation of mice with adipocytes specific TGFβ2 knockout were not done yet.

The overall goal of my study was to identify a new factor which is involved in the

adipogenesis towards a brown adipocyte phenotype and may be a target for the treatment

of obesity and its associated cardiovascular diseases. Although, LTBP3 and TGFβ2 was

identified to regulate UCP1 expression, it remains to be elucidated if these are targets for

a treatment of obesity or cardiovascular diseases. High levels of LTBP3 have been

associated with metastatic tumors [104]. Furthermore, it was reported that TGFβ2 has an

impact on several cellular processes such as differentiation, proliferation and immune

suppression, suggesting that TGFβ2 treatment in humans could have severe side effects

[89]. Stimulating a higher LTBP3 expression or TGFβ2 secretion locally in the adipose tissue

could overcome this. Several approaches to implement a local treatment of the adipose

tissue have been published so far. It was shown that a special patch, locally applied on the

sc adipose tissue, stimulated browning in vivo by injecting rosiglitazone with microneedles

[159]. Another interesting approach is the transplantation of preadipocytes which are

preprogrammed to undergo adipogenesis into UCP1-expressing cells. Very recently Wang

et al. activated the UCP1 gene expression with CRISPR/Cas in human preadipocytes derived

from the sc WAT and transplanted them into the sc WAT of obese mice [160]. These mice

showed an improved glucose tolerance, insulin sensitivity, and higher energy expenditure

compared to control mice. However, it is not clear when this transplantation method will

become clinically relevant. Injection with nanoparticles containing “batokines” or its mRNA

into the sc adipose tissue of obese patients to induce browning could be a realistic scenario.

Nanoparticles specifically targeting the adipose tissue have already been developed and

pharma companies have mRNA-based drugs in phase 1 and phase 2 of clinical trials [161–

164].

In conclusion my findings indicate LTBP3 as a new factor regulating UCP1 expression via

TGFβ2 bioavailability.

90

5. Summary

Obesity is a global epidemic and the main risk factor for cardiovascular diseases which is

the number one cause of death worldwide. Current treatment methods for obesity are very

limited and often fail in the long term.

In the last decade it has been demonstrated that not only human newborns but also adults

have functional brown adipose tissue mass. Furthermore, there is evidence for

uncoupling protein 1 (UCP1)-expressing adipocytes within the white adipose tissue (WAT)

of humans. White adipocyte browning is defined by a significant increase in UCP1

expression in the adipose tissue depot and has been shown to significantly increase overall

metabolic health inter alia by higher glucose uptake and fatty acid uptake, thus providing

a strategy for obesity treatment. An approach for implementation, is to stimulate human

adipose stromal cells (hASCs) within the WAT to undergo adipogenesis into UCP1-

expressing adipocytes. However, the mechanism how these cells differentiate either into

adipocytes with a white or with a brown phenotype is not fully understood. Our group has

shown that hASCs isolated from deep neck (dn) adipose tissue differentiated in vitro into

adipocytes with a brown phenotype and hASCs derived from the subcutaneous (sc) adipose

tissue differentiated into adipocytes with a white phenotype. Microarray analysis revealed

different expression patterns of these hASCs dependent on their adipose tissue origin.

Therefore, we hypothesized, that hASCs secrete factors which stimulate or inhibit the

adipogenesis into a brown adipocyte phenotype in an auto- or paracrine manner.

After target ranking of these differentially regulated genes and literature screening the

latent transforming growth factor beta binding proteins (LTBP1-4) were identified as

potential factors involved in generating a brown adipocyte phenotype. Except for LTBP2,

they are known to play a crucial role in the bioavailability of transforming growth factors

(TGFβ). These TGFβ cytokines in turn are known to be involved in regulating UCP1

expression in vitro and WAT browning in vivo. The SMAD2/SMAD3 mediated TGFβ pathway

has already been shown to inhibit white adipocyte browning and TGFβ2 was reported to

increase UCP1 expression in murine BAT. The aim of this study was to investigate the

impact of the LTBP isoforms on brown marker genes and the metabolic function in vitro.

91

LTBP1, LTBP3 and LTBP4 were significantly higher expressed in the dn hASCs and all four

LTBPs were expressed during the adipogenesis of a human sc preadipocyte cell strain (SGBS

cells). LTBP1 and LTBP4 were not involved in the generation of a brown phenotype in SGBS

cells. However, LTBP2 deficiency showed a reduced UCP1 mRNA expression and altered

metabolic function. Moreover, deficiency of LTBP3 in SGBS adipocytes led to a repression

of UCP1 expression on mRNA and protein level. Furthermore, the metabolic function of

these cells was altered as well.

Inhibition of the TGFβ-pathway in human SGBS cells led in three different approaches to an

adipogenesis towards a white phenotype. TGFβ deficiency, TGFβ receptor1 kinase

inhibition and knockdown of SMAD4 led to a significant decrease in UCP1 expression,

whereby the differentiation rate was not affected in these cells. These results are in

accordance with the observed UCP1 expression in LTBP2- and LTBP3-deficient cells.

Measuring the TGFβ content in the supernatant of LTBP2- and LTBP3-deficient

preadipocytes revealed a significant lower level for TGFβ2 compared to the control. The

TGFβ2 knockout in turn resulted in a decreased UCP1 expression and metabolic function as

well. Moreover, the LTBP3 and TGFβ2 axis seems to play also a relevant role in human WAT

browning. LTBP3 expression correlated positively with the TGFβ2 expression in human sc

WAT samples. Furthermore, both LTBP3 and TGFβ2 correlated significantly positive with

the UCP1 expression in the same patient samples. As a result, one can hypothesize that

LTBP3 regulates the adipogenesis towards a brown phenotype via TGFβ2 bioavailability.

The findings of this thesis revealed LTBP3 as a new factor involved in generating a brown

adipocyte phenotype. Furthermore, the inhibition of TGFβ pathway demonstrated new

insights of TGFβ and the regulation of UCP1 expression. The specific pathway how LTBP3

and TGFβ2 mediate a shift towards a brown adipocyte phenotype should be addressed in

the future, thereby identifying a new mechanism to induce UCP1 expression in human

WAT. As a consequence, LTBP3 or TGFβ2 may become new promising targets for the

treatment of obesity.

92

6. References

[1] WHO, Obesity and overweight, World Heal. Organ. (2017).

https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight

(accessed June 3, 2020).

[2] WHO, Cardiovascular diseases (CVDs), World Heal. Organ. (2017).

https://www.who.int/news-room/fact-sheets/detail/cardiovascular-diseases-(cvds)

(accessed June 3, 2020).

[3] M. Blüher, Obesity: global epidemiology and pathogenesis, Nat. Rev. Endocrinol. 15

(2019) 288–298. https://doi.org/10.1038/s41574-019-0176-8.

[4] E.K. Oikonomou, C. Antoniades, The role of adipose tissue in cardiovascular health

and disease, Nat. Rev. Cardiol. 16 (2019) 83–99. https://doi.org/10.1038/s41569-

018-0097-6.

[5] L. Luo, M. Liu, Adipose tissue in control of metabolism, J. Endocrinol. 231 (2016) R77–

R99. https://doi.org/10.1530/JOE-16-0211.

[6] A.M. Cypess, S. Lehman, G. Williams, I. Tal, D. Rodman, A.B. Goldfine, F.C. Kuo, E.L.

Palmer, Y.-H. Tseng, A. Doria, G.M. Kolodny, C.R. Kahn, Identification and importance

of brown adipose tissue in adult humans., N. Engl. J. Med. 360 (2009) 1509–1517.

https://doi.org/10.1097/OGX.0b013e3181ac8aa2.

[7] J. Nedergaard, T. Bengtsson, B. Cannon, Unexpected evidence for active brown

adipose tissue in adult humans., Am. J. Physiol. Endocrinol. Metab. 293 (2007) E444–

E452. https://doi.org/10.1152/ajpendo.00691.2006.

[8] K.A. Virtanen, M.E. Lidell, J. Orava, M. Heglind, R. Westergren, T. Niemi, M.

Taittonen, J. Laine, N.J. Savisto, S. Enerback, P. Nuutila, Functional brown adipose

tissue in healthy adults, N Engl J Med. 360 (2009) 1518–1525.

https://doi.org/360/15/1518 [pii] 10.1056/NEJMoa0808949.

93

[9] W.D. Van Marken Lichtenbelt, J.W. Vanhommerig, N.M. Smulders, J.M.A.F.L.

Drossaerts, G.J. Kemerink, N.D. Bouvy, P. Schrauwen, G.J.J. Teule, Cold-activated

brown adipose tissue in healthy men, N. Engl. J. Med. 360 (2009) 1500–1508.

https://doi.org/10.1056/NEJMoa0808718.

[10] A. Bartelt, J. Heeren, Adipose tissue browning and metabolic health, Nat. Rev.

Endocrinol. 10 (2014) 24–36. https://doi.org/10.1038/nrendo.2013.204.

[11] P. Young, J.R.S. Arch, M. Ashwell, Brown adipose tissue in the parametrial fat pad of

the mouse, 1984. https://doi.org/10.1016/0014-5793(84)80822-4.

[12] E.D. Rosen, B.M. Spiegelman, What we talk about when we talk about fat, Cell. 156

(2014) 20–44. https://doi.org/10.1016/j.cell.2013.12.012.

[13] B. Cannon, J. Nedergaard, Brown Adipose Tissue: Function and Physiological

Significance, Physiol. Rev. 84 (2004) 277–359.

https://doi.org/10.1152/physrev.00015.2003.

[14] V. Gilsanz, H.H. Hu, S. Kajimura, Relevance of brown adipose tissue in infancy and

adolescence, Pediatr. Res. 73 (2013) 3–9. https://doi.org/10.1038/pr.2012.141.

[15] M. Klingenberg, Uncoupling protein - A useful energy dissipator, J. Bioenerg.

Biomembr. 31 (1999) 419–430. https://doi.org/10.1023/A:1005440221914.

[16] A.L. Poher, J. Altirriba, C. Veyrat-Durebex, F. Rohner-Jeanrenaud, Brown adipose

tissue activity as a target for the treatment of obesity/insulin resistance, Front.

Physiol. 6 (2015) 4. https://doi.org/10.3389/fphys.2015.00004.

[17] M. Chondronikola, E. Volpi, E. Børsheim, C. Porter, P. Annamalai, S. Enerbäck, M.E.

Lidell, M.K. Saraf, S.M. Labbe, N.M. Hurren, C. Yfanti, T. Chao, C.R. Andersen, F.

Cesani, H. Hawkins, L.S. Sidossis, Brown adipose tissue improves whole-body glucose

homeostasis and insulin sensitivity in humans, Diabetes. 63 (2014) 4089–4099.

https://doi.org/10.2337/db14-0746.

94

[18] M.J.W. Hanssen, J. Hoeks, B. Brans, A.A.J.J. Van Der Lans, G. Schaart, J.J. Van Den

Driessche, J.A. Jörgensen, M. V. Boekschoten, M.K.C. Hesselink, B. Havekes, S.

Kersten, F.M. Mottaghy, W.D. Van Marken Lichtenbelt, P. Schrauwen, Short-term

cold acclimation improves insulin sensitivity in patients with type 2 diabetes mellitus,

Nat. Med. 21 (2015) 863–865. https://doi.org/10.1038/nm.3891.

[19] B. Cannon, J. Nedergaard, Nonshivering thermogenesis and its adequate

measurement in metabolic studies, J. Exp. Biol. 214 (2011) 242–253.

https://doi.org/10.1242/jeb.050989.

[20] E.T. Chouchani, L. Kazak, B.M. Spiegelman, New Advances in Adaptive

Thermogenesis: UCP1 and Beyond, Cell Metab. 29 (2019) 27–37.

https://doi.org/10.1016/j.cmet.2018.11.002.

[21] S. Enerbäck, A. Jacobsson, E.M. Simpson, C. Guerra, H. Yamashita, M.E. Harper, L.P.

Kozak, Mice lacking mitochondrial uncoupling protein are cold-sensitive but not

obese, 1997. https://doi.org/10.1038/387090a0.

[22] P.G. Crichton, Y. Lee, E.R.S. Kunji, The molecular features of uncoupling protein 1

support a conventional mitochondrial carrier-like mechanism, Biochimie. 134 (2017)

35–50. https://doi.org/10.1016/j.biochi.2016.12.016.

[23] B. Wicksteed, L.M. Dickson, PKA differentially regulates adipose depots to control

energy expenditure, Endocrinology. 158 (2017) 464–466.

https://doi.org/10.1210/en.2017-00038.

[24] W. Cao, K.W. Daniel, J. Robidoux, P. Puigserver, A. V. Medvedev, X. Bai, L.M. Floering,

B.M. Spiegelman, S. Collins, p38 Mitogen-Activated Protein Kinase Is the Central

Regulator of Cyclic AMP-Dependent Transcription of the Brown Fat Uncoupling

Protein 1 Gene, Mol. Cell. Biol. 24 (2004) 3057–3067.

https://doi.org/10.1128/mcb.24.7.3057-3067.2004.

[25] A. Vargas-Castillo, R. Fuentes-Romero, L.A. Rodriguez-Lopez, N. Torres, A.R. Tovar,

Understanding the Biology of Thermogenic Fat: Is Browning A New Approach to the

Treatment of Obesity?, Arch. Med. Res. 48 (2017) 401–413.

https://doi.org/10.1016/j.arcmed.2017.10.002.

95

[26] C.T. Herz, F.W. Kiefer, Adipose tissue browning in mice and humans, J. Endocrinol.

241 (2019) R97–R109. https://doi.org/10.1530/joe-18-0598.

[27] M. Harms, P. Seale, Brown and beige fat: development, function and therapeutic

potential, Nat. Med. 19 (2013) 1252–1263. https://doi.org/10.1038/nm.3361.

[28] W. Aherne, D. Hull, Brown adipose tissue and heat production in the newborn

infant., J. Pathol. Bacteriol. 91 (1966) 223–234.

https://doi.org/10.1002/path.1700910126.

[29] K.A. Iwen, J. Backhaus, M. Cassens, M. Waltl, O.C. Hedesan, M. Merkel, J. Heeren, C.

Sina, L. Rademacher, A. Windjäger, A.R. Haug, F.W. Kiefer, H. Lehnert, S.M. Schmid,

Cold-Induced Brown Adipose Tissue Activity Alters Plasma Fatty Acids and Improves

Glucose Metabolism in Men, J. Clin. Endocrinol. Metab. 102 (2017) 4226–4234.

https://doi.org/10.1210/jc.2017-01250.

[30] J. Orava, P. Nuutila, T. Noponen, R. Parkkola, T. Viljanen, S. Enerbäck, A. Rissanen,

K.H. Pietiläinen, K.A. Virtanen, Blunted metabolic responses to cold and insulin

stimulation in brown adipose tissue of obese humans, Obesity. 21 (2013) 2279–

2287. https://doi.org/10.1002/oby.20456.

[31] C. Pfannenberg, M.K. Werner, S. Ripkens, I. Stef, A. Deckert, M. Schmadl, M. Reimold,

H.U. Häring, C.D. Claussen, N. Stefan, Impact of age on the relationships of brown

adipose tissue with sex and adiposity in humans, Diabetes. 59 (2010) 1789–1793.

https://doi.org/10.2337/db10-0004.

[32] V. Ouellet, S.M. Labbé, D.P. Blondin, S. Phoenix, B. Guérin, F. Haman, E.E. Turcotte,

D. Richard, A.C. Carpentier, Brown adipose tissue oxidative metabolism contributes

to energy expenditure during acute cold exposure in humans, J. Clin. Invest. 122

(2012) 545–552. https://doi.org/10.1172/JCI60433.

[33] G.H.E.J. Vijgen, N.D. Bouvy, G.J.J. Teule, B. Brans, J. Hoeks, P. Schrauwen, W.D. Van

Marken Lichtenbelt, Increase in brown adipose tissue activity after weight loss in

morbidly obese subjects, J. Clin. Endocrinol. Metab. 97 (2012).

https://doi.org/10.1210/jc.2012-1289.

96

[34] T. Yoneshiro, S. Aita, M. Matsushita, Y. Okamatsu-Ogura, T. Kameya, Y. Kawai, M.

Miyagawa, M. Tsujisaki, M. Saito, Age-related decrease in cold-activated brown

adipose tissue and accumulation of body fat in healthy humans, Obesity. 19 (2011)

1755–1760. https://doi.org/10.1038/oby.2011.125.

[35] M.J.W. Hanssen, A.A.J.J. Van Der Lans, B. Brans, J. Hoeks, K.M.C. Jardon, G. Schaart,

F.M. Mottaghy, P. Schrauwen, W.D. Van Marken Lichtenbelt, Short-term cold

acclimation recruits brown adipose tissue in obese humans, Diabetes. 65 (2016)

1179–1189. https://doi.org/10.2337/db15-1372.

[36] M.J. Lee, Transforming growth factor beta superfamily regulation of adipose tissue

biology in obesity, Biochim. Biophys. Acta - Mol. Basis Dis. 1864 (2018) 1160–1171.

https://doi.org/10.1016/j.bbadis.2018.01.025.

[37] Y.H. Tseng, E. Kokkotou, T.J. Schulz, T.L. Huang, J.N. Winnay, C.M. Taniguchi, T.T.

Tran, R. Suzuki, D.O. Espinoza, Y. Yamamoto, M.J. Ahrens, A.T. Dudley, A.W. Norris,

R.N. Kulkarni, C.R. Kahn, New role of bone morphogenetic protein 7 in brown

adipogenesis and energy expenditure, Nature. 454 (2008) 1000–1004.

https://doi.org/10.1038/nature07221.

[38] F.F. Fisher, S. Kleiner, N. Douris, E.C. Fox, R.J. Mepani, F. Verdeguer, J. Wu, A.

Kharitonenkov, J.S. Flier, E. Maratos-Flier, B.M. Spiegelman, FGF21 regulates PGC-1α

and browning of white adipose tissues in adaptive thermogenesis, Genes Dev. 26

(2012) 271–281. https://doi.org/10.1101/gad.177857.111.

[39] A.M. Cypess, Y.C. Chen, C. Sze, K. Wang, J. English, O. Chan, A.R. Holman, I. Tal, M.R.

Palmer, G.M. Kolodny, C.R. Kahn, Cold but not sympathomimetics activates human

brown adipose tissue in vivo, Proc. Natl. Acad. Sci. U. S. A. 109 (2012) 10001–10005.

https://doi.org/10.1073/pnas.1207911109.

[40] M. Bordicchia, D. Liu, E.Z. Amri, G. Ailhaud, P. Dessì-Fulgheri, C. Zhang, N. Takahashi,

R. Sarzani, S. Collins, Cardiac natriuretic peptides act via p38 MAPK to induce the

brown fat thermogenic program in mouse and human adipocytes, J. Clin. Invest. 122

(2012) 1022–1036. https://doi.org/10.1172/JCI59701.

97

[41] W. Wang, P. Seale, Control of brown and beige fat development, Nat. Rev. Mol. Cell

Biol. 17 (2016) 691–702. https://doi.org/10.1038/nrm.2016.96.

[42] J. Sanchez-Gurmaches, C.M. Hung, D.A. Guertin, Emerging Complexities in Adipocyte

Origins and Identity, Trends Cell Biol. 26 (2016) 313–326.

https://doi.org/10.1016/j.tcb.2016.01.004.

[43] G. Barbatelli, I. Murano, L. Madsen, Q. Hao, M. Jimenez, K. Kristiansen, J.P.

Giacobino, R. De Matteis, S. Cinti, The emergence of cold-induced brown adipocytes

in mouse white fat depots is determined predominantly by white to brown adipocyte

transdifferentiation, Am. J. Physiol. - Endocrinol. Metab. 298 (2010) E1244-53.

https://doi.org/10.1152/ajpendo.00600.2009.

[44] A. Frontini, A. Vitali, J. Perugini, I. Murano, C. Romiti, D. Ricquier, M. Guerrieri, S.

Cinti, White-to-brown transdifferentiation of omental adipocytes in patients

affected by pheochromocytoma, Biochim. Biophys. Acta - Mol. Cell Biol. Lipids. 1831

(2013) 950–959. https://doi.org/10.1016/j.bbalip.2013.02.005.

[45] P. Seale, B. Bjork, W. Yang, S. Kajimura, S. Chin, S. Kuang, A. Scimè, S. Devarakonda,

H.M. Conroe, H. Erdjument-Bromage, P. Tempst, M.A. Rudnicki, D.R. Beier, B.M.

Spiegelman, PRDM16 controls a brown fat/skeletal muscle switch, Nature. 454

(2008) 961–967. https://doi.org/10.1038/nature07182.

[46] M. Esteve Ràfols, Adipose tissue: Cell heterogeneity and functional diversity,

Endocrinol. y Nutr. (English Ed. 61 (2014) 100–112.

https://doi.org/10.1016/j.endoen.2014.02.001.

[47] D. Tews, V. Schwar, M. Scheithauer, T. Weber, T. Fromme, M. Klingenspor, T.F. Barth,

P. Möller, K. Holzmann, K.M. Debatin, P. Fischer-Posovszky, M. Wabitsch,

Comparative gene array analysis of progenitor cells from human paired deep neck

and subcutaneous adipose tissue., Mol. Cell. Endocrinol. 395 (2014) 41–50.

https://doi.org/10.1016/j.mce.2014.07.011.

98

[48] A.M. Cypess, A.P. White, C. Vernochet, T.J. Schulz, R. Xue, C.A. Sass, T.L. Huang, C.

Roberts-Toler, L.S. Weiner, C. Sze, A.T. Chacko, L.N. Deschamps, L.M. Herder, N.

Truchan, A.L. Glasgow, A.R. Holman, A. Gavrila, P.O. Hasselgren, M.A. Mori, M.

Molla, Y.H. Tseng, Anatomical localization, gene expression profiling and functional

characterization of adult human neck brown fat, Nat. Med. 19 (2013) 635–639.

https://doi.org/10.1038/nm.3112.

[49] M.E. Lidell, M.J. Betz, O. Dahlqvist Leinhard, M. Heglind, L. Elander, M. Slawik, T.

Mussack, D. Nilsson, T. Romu, P. Nuutila, K.A. Virtanen, F. Beuschlein, A. Persson, M.

Borga, S. Enerbäck, Evidence for two types of brown adipose tissue in humans, Nat.

Med. 19 (2013) 631–634. https://doi.org/10.1038/nm.3017.

[50] J. Wu, P. Boström, L.M. Sparks, L. Ye, J.H. Choi, A.H. Giang, M. Khandekar, K.A.

Virtanen, P. Nuutila, G. Schaart, K. Huang, H. Tu, W.D. Van Marken Lichtenbelt, J.

Hoeks, S. Enerbäck, P. Schrauwen, B.M. Spiegelman, Beige adipocytes are a distinct

type of thermogenic fat cell in mouse and human, Cell. 150 (2012) 366–376.

https://doi.org/10.1016/j.cell.2012.05.016.

[51] L.Z. Sharp, K. Shinoda, H. Ohno, D.W. Scheel, E. Tomoda, L. Ruiz, H. Hu, L. Wang, Z.

Pavlova, V. Gilsanz, S. Kajimura, Human BAT Possesses Molecular Signatures That

Resemble Beige/Brite Cells, PLoS One. 7 (2012) e49452.

https://doi.org/10.1371/journal.pone.0049452.

[52] N.Z. Jespersen, T.J. Larsen, L. Peijs, S. Daugaard, P. Homøe, A. Loft, J. De Jong, N.

Mathur, B. Cannon, J. Nedergaard, B.K. Pedersen, K. Møller, C. Scheele, A classical

brown adipose tissue mrna signature partly overlaps with brite in the supraclavicular

region of adult humans, Cell Metab. 17 (2013) 798–805.


[53] J.M.A. de Jong, W. Sun, N.D. Pires, A. Frontini, M. Balaz, N.Z. Jespersen, A. Feizi, K.

Petrovic, A.W. Fischer, M.H. Bokhari, T. Niemi, P. Nuutila, S. Cinti, S. Nielsen, C.

Scheele, K. Virtanen, B. Cannon, J. Nedergaard, C. Wolfrum, N. Petrovic, Human

brown adipose tissue is phenocopied by classical brown adipose tissue in

physiologically humanized mice, Nat. Metab. 1 (2019) 830–843.

https://doi.org/10.1038/s42255-019-0101-4.

99

[54] M. Rosenwald, C. Wolfrum, The origin and definition of brite versus white and

classical brown adipocytes, Adipocyte. 3 (2014) 4–9.

https://doi.org/10.4161/adip.26232.

[55] P.A. Kern, B.S. Finlin, B. Zhu, N. Rasouli, R.E. McGehee, P.M. Westgate, E.E. Dupont-

Versteegden, The effects of temperature and seasons on subcutaneous white

adipose tissue in humans: Evidence for thermogenic gene induction, J. Clin.

Endocrinol. Metab. 99 (2014) E2772–E2779. https://doi.org/10.1210/jc.2014-2440.

[56] N. Martínez-Sánchez, J.M. Moreno-Navarrete, C. Contreras, E. Rial-Pensado, J.

Fernø, R. Nogueiras, C. Diéguez, J.M. Fernández-Real, M. López, Thyroid hormones

induce browning of white fat, J. Endocrinol. 232 (2017) 351–362.

https://doi.org/10.1530/JOE-16-0425.

[57] M.C. Skarulis, F.S. Celi, E. Mueller, M. Zemskova, R. Malek, L. Hugendubler, C.

Cochran, J. Solomon, C. Chen, P. Gorden, Thyroid hormone induced brown adipose

tissue and amelioration of diabetes in a patient with extreme insulin resistance, J.

Clin. Endocrinol. Metab. 95 (2010) 256–262. https://doi.org/10.1210/jc.2009-0543.

[58] L.S. Sidossis, C. Porter, M.K. Saraf, E. Børsheim, R.S. Radhakrishnan, T. Chao, A. Ali,

M. Chondronikola, R. Mlcak, C.C. Finnerty, H.K. Hawkins, T. Toliver-Kinsky, D.N.

Herndon, Browning of Subcutaneous White Adipose Tissue in Humans after Severe

Adrenergic Stress, Cell Metab. 22 (2015) 219–227.


[59] A. Ruban, K. Stoenchev, H. Ashrafian, J. Teare, Current treatments for obesity, Clin.

Med. J. R. Coll. Physicians London. 19 (2019) 205–212.

https://doi.org/10.7861/clinmedicine.19-3-205.

[60] R.K.C. Loh, B.A. Kingwell, A.L. Carey, Human brown adipose tissue as a target for

obesity management; beyond cold-induced thermogenesis, Obes. Rev. 18 (2017)

1227–1242. https://doi.org/10.1111/obr.12584.

[61] C.M. Kusminski, P.E. Bickel, P.E. Scherer, Targeting adipose tissue in the treatment

of obesity-associated diabetes, Nat. Rev. Drug Discov. 15 (2016) 639–660.

https://doi.org/10.1038/nrd.2016.75.

100

[62] A.A.J.J. Van Der Lans, J. Hoeks, B. Brans, G.H.E.J. Vijgen, M.G.W. Visser, M.J.

Vosselman, J. Hansen, J.A. Jörgensen, J. Wu, F.M. Mottaghy, P. Schrauwen, W.D. Van

Marken Lichtenbelt, Cold acclimation recruits human brown fat and increases

nonshivering thermogenesis, J. Clin. Invest. 123 (2013) 3395–3403.

https://doi.org/10.1172/JCI68993.

[63] D.P. Blondin, S.M. Labbé, H.C. Tingelstad, C. Noll, M. Kunach, S. Phoenix, B. Guérin,

É.E. Turcotte, A.C. Carpentier, D. Richard, F. Haman, Increased brown adipose tissue

oxidative capacity in cold-acclimated humans, J. Clin. Endocrinol. Metab. 99 (2014)

E438-46. https://doi.org/10.1210/jc.2013-3901.

[64] T. Yoneshiro, S. Aita, M. Matsushita, T. Kayahara, T. Kameya, Y. Kawai, T. Iwanaga,

M. Saito, Recruited brown adipose tissue as an antiobesity agent in humans, J. Clin.

Invest. 123 (2013) 3404–3408. https://doi.org/10.1172/JCI67803.

[65] K.Y. Chen, R.J. Brychta, J.D. Linderman, S. Smith, A. Courville, W. Dieckmann, P.

Herscovitch, C.M. Millo, A. Remaley, P. Lee, F.S. Celi, Brown fat activation mediates

cold-induced thermogenesis in adult humans in response to a mild decrease in

ambient temperature, J. Clin. Endocrinol. Metab. 98 (2013) E1218-23.

https://doi.org/10.1210/jc.2012-4213.

[66] S. Srivastava, R.L. Veech, Brown and brite: The fat soldiers in the anti-obesity fight,

Front. Physiol. 10 (2019). https://doi.org/10.3389/fphys.2019.00038.

[67] C. Weyer, P.A. Tataranni, S. Snitker, E. Danforth, E. Ravussin, Increase in insulin

action and fat oxidation after treatment with CL 316,243, a highly selective β3-

adrenoceptor agonist in humans, Diabetes. 47 (1998) 1555–1561.

https://doi.org/10.2337/diabetes.47.10.1555.

[68] L.M. Redman, L. De Jonge, X. Fang, B. Gamlin, D. Recker, F.L. Greenway, S.R. Smith,

E. Ravussin, Lack of an effect of a novel β3-adrenoceptor agonist, TAK-677, on energy

metabolism in obese individuals: A double-blind, placebo-controlled randomized

study, J. Clin. Endocrinol. Metab. 92 (2007) 527–531.

https://doi.org/10.1210/jc.2006-1740.

101

[69] J.R.S. Arch, The discovery of drugs for obesity, the metabolic effects of leptin and

variable receptor pharmacology: Perspectives from β3- adrenoceptor agonists,

Naunyn. Schmiedebergs. Arch. Pharmacol. 378 (2008) 225–240.

https://doi.org/10.1007/s00210-008-0271-1.

[70] T.M. Larsen, S. Toubro, M.A. Van Baak, K.M. Gottesdiener, P. Larson, W.H.M. Saris,

A. Astrup, Effect of a 28-d treatment with L-796568, a novel β 3-adrenergic receptor

agonist, on energy expenditure and body composition in obese men, Am. J. Clin.

Nutr. 76 (2002) 780–788. https://doi.org/10.1093/ajcn/76.4.780.

[71] A.M. Cypess, L.S. Weiner, C. Roberts-Toler, E.F. Elía, S.H. Kessler, P.A. Kahn, J. English,

K. Chatman, S.A. Trauger, A. Doria, G.M. Kolodny, Activation of Human Brown

Adipose Tissue by a β3-Adrenergic Receptor Agonist, Cell Metab. 21 (2015) 33–38.

https://doi.org/10.1016/J.CMET.2014.12.009.

[72] A.E. O’Mara, J.W. Johnson, J.D. Linderman, R.J. Brychta, S. McGehee, L.A. Fletcher,

Y.A. Fink, D. Kapuria, T.M. Cassimatis, N. Kelsey, C. Cero, Z.A. Sater, F. Piccinini, A.S.

Baskin, B.P. Leitner, H. Cai, C.M. Millo, W. Dieckmann, M. Walter, N.B. Javitt, Y.

Rotman, P.J. Walter, M. Ader, R.N. Bergman, P. Herscovitch, K.Y. Chen, A.M. Cypess,

Chronic mirabegron treatment increases human brown fat, HDL cholesterol, and

insulin sensitivity, J. Clin. Invest. 130 (2020) 2209–2219.

https://doi.org/10.1172/JCI131126.

[73] B.S. Finlin, H. Memetimin, A.L. Confides, I. Kasza, B. Zhu, H.J. Vekaria, B. Harfmann,

K.A. Jones, Z.R. Johnson, P.M. Westgate, C.M. Alexander, P.G. Sullivan, E.E. Dupont-

Versteegden, P.A. Kern, Human adipose beiging in response to cold and mirabegron,

JCI Insight. 3 (2018). https://doi.org/10.1172/jci.insight.121510.

[74] G. Medina-Gomez, R.M. Calvo, M.-J. Obregon, Thermogenic effect of

triiodothyroacetic acid at low doses in rat adipose tissue without adverse side effects

in the thyroid axis, Am. J. Physiol. Metab. 294 (2008) E688–E697.

https://doi.org/10.1152/ajpendo.00417.2007.

102

[75] J.Z. Lin, A.J. Martagón, S.L. Cimini, D.D. Gonzalez, D.W. Tinkey, A. Biter, J.D. Baxter,

P. Webb, J.Å. Gustafsson, S.M. Hartig, K.J. Phillips, Pharmacological Activation of

Thyroid Hormone Receptors Elicits a Functional Conversion of White to Brown Fat,

Cell Rep. 13 (2015) 1528–1537. https://doi.org/10.1016/j.celrep.2015.10.022.

[76] I. Klein, K. Ojamaa, Thyroid Hormone and the Cardiovascular System, N. Engl. J. Med.

344 (2001) 501–509. https://doi.org/10.1056/NEJM200102153440707.

[77] M. Aprile, M.R. Ambrosio, V. D’Esposito, F. Beguinot, P. Formisano, V. Costa, A.

Ciccodicola, PPARG in Human Adipogenesis: Differential Contribution of Canonical

Transcripts and Dominant Negative Isoforms, PPAR Res. 2014 (2014).

https://doi.org/10.1155/2014/537865.

[78] J.X. Rong, Y. Qiu, M.K. Hansen, L. Zhu, V. Zhang, M. Xie, Y. Okamoto, M.D. Mattie, H.

Higashiyama, S. Asano, J.C. Strum, T.E. Ryan, Adipose mitochondrial biogenesis is

suppressed in db/db and high-fat diet-fed mice and improved by rosiglitazone,

Diabetes. 56 (2007) 1751–1760. https://doi.org/10.2337/db06-1135.

[79] D. Halbgebauer, M. Dahlhaus, M. Wabitsch, P. Fischer-Posovszky, D. Tews, Browning

capabilities of human primary adipose-derived stromal cells compared to SGBS cells,

Sci. Rep. 10 (2020) 1–8. https://doi.org/10.1038/s41598-020-64369-7.

[80] R.E. Soccio, E.R. Chen, M.A. Lazar, Thiazolidinediones and the promise of insulin

sensitization in type 2 diabetes, Cell Metab. 20 (2014) 573–591.


[81] A. Abbas, J. Blandon, J. Rude, A. Elfar, D. Mukherjee, PPAR- γ Agonist in

Treatment of Diabetes: Cardiovascular Safety Considerations, Cardiovasc. Hematol.

Agents Med. Chem. 10 (2012) 124–134.

https://doi.org/10.2174/187152512800388948.

[82] D. Cuevas-Ramos, R. Mehta, C.A. Aguilar-Salinas, Fibroblast growth factor 21 and

browning of white adipose tissue, Front. Physiol. 10 (2019).

https://doi.org/10.3389/fphys.2019.00037.

103

[83] G. Gaich, J.Y. Chien, H. Fu, L.C. Glass, M.A. Deeg, W.L. Holland, A. Kharitonenkov, T.

Bumol, H.K. Schilske, D.E. Moller, The effects of LY2405319, an FGF21 Analog, in

obese human subjects with type 2 diabetes, Cell Metab. 18 (2013) 333–340.


[84] S. Talukdar, Y. Zhou, D. Li, M. Rossulek, J. Dong, V. Somayaji, Y. Weng, R. Clark, A.

Lanba, B.M. Owen, M.B. Brenner, J.K. Trimmer, K.E. Gropp, J.R. Chabot, D.M. Erion,

T.P. Rolph, B. Goodwin, R.A. Calle, A long-acting FGF21 molecule, PF-05231023,

decreases body weight and improves lipid profile in non-human primates and type

2 diabetic subjects, Cell Metab. 23 (2016) 427–440.


[85] K. Shinoda, I.H.N. Luijten, Y. Hasegawa, H. Hong, S.B. Sonne, M. Kim, R. Xue, M.

Chondronikola, A.M. Cypess, Y.H. Tseng, J. Nedergaard, L.S. Sidossis, S. Kajimura,

Genetic and functional characterization of clonally derived adult human brown

adipocytes, Nat. Med. 21 (2015) 389–394. https://doi.org/10.1038/nm.3819.

[86] A. Perdikari, G.G. Leparc, M. Balaz, N.D. Pires, M.E. Lidell, W. Sun, F. Fernandez-

Albert, S. Müller, N. Akchiche, H. Dong, L. Balazova, L. Opitz, E. Röder, H. Klein, P.

Stefanicka, L. Varga, P. Nuutila, K.A. Virtanen, T. Niemi, M. Taittonen, G. Rudofsky, J.

Ukropec, S. Enerbäck, E. Stupka, H. Neubauer, C. Wolfrum, BATLAS: Deconvoluting

Brown Adipose Tissue, Cell Rep. 25 (2018) 784-797.e4.

https://doi.org/10.1016/j.celrep.2018.09.044.

[87] F. Villarroya, R. Cereijo, J. Villarroya, M. Giralt, Brown adipose tissue as a secretory

organ, Nat. Rev. Endocrinol. 13 (2017) 26–35.

https://doi.org/10.1038/nrendo.2016.136.

[88] H. Takahashi, C.R.R. Alves, K.I. Stanford, R.J.W. Middelbeek, P. Nigro, R.E. Ryan, R.

Xue, M. Sakaguchi, M.D. Lynes, K. So, J.D. Mul, M.-Y. Lee, E. Balan, H. Pan, J.M.

Dreyfuss, M.F. Hirshman, M. Azhar, J.C. Hannukainen, P. Nuutila, K.K. Kalliokoski, S.

Nielsen, B.K. Pedersen, C.R. Kahn, Y.-H. Tseng, L.J. Goodyear, TGF-β2 is an exercise-

induced adipokine that regulates glucose and fatty acid metabolism, Nat. Metab.

2019 12. 1 (2019) 291. https://doi.org/10.1038/s42255-018-0030-7.

104

[89] M. Morikawa, R. Derynck, K. Miyazono, TGF- β and the TGF-β family: Context-

dependent roles in cell and tissue physiology, Cold Spring Harb. Perspect. Biol. 8

(2016) a021873. https://doi.org/10.1101/cshperspect.a021873.

[90] H. Chang, C.W. Brown, M.M. Matzuk, Genetic analysis of the mammalian

transforming growth factor-β superfamily, Endocr. Rev. 23 (2002) 787–823.

https://doi.org/10.1210/er.2002-0003.

[91] R. Derynck, E.H. Budi, Specificity, versatility, and control of TGF-b family signaling,

Sci. Signal. 12 (2019) eaav5183. https://doi.org/10.1126/scisignal.aav5183.

[92] M.M. Shull, I. Ormsby, A.B. Kier, S. Pawlowski, R.J. Diebold, M. Yin, R. Allen, C.

Sidman, G. Proetzel, D. Calvin, N. Annunziata, T. Doetschman, Targeted disruption of

the mouse transforming growth factor-β1 gene results in multifocal inflammatory

disease [14], Nature. 359 (1992) 693–699. https://doi.org/10.1038/359693a0.

[93] A.B. Kulkarni, C.G. Huh, D. Becker, A. Geiser, M. Lyght, K.C. Flanders, A.B. Roberts,

M.B. Sporn, J.M. Ward, S. Karlsson, Transforming growth factor β1 null mutation in

mice causes excessive inflammatory response and early death, Proc. Natl. Acad. Sci.

U. S. A. 90 (1993) 770–774. https://doi.org/10.1073/pnas.90.2.770.

[94] L.P. Sanford, I. Ormsby, A.C. Gittenberger-de Groot, H. Sariola, R. Friedman, G.P.

Boivin, E. Lou Cardell, T. Doetschman, TGFβ2 knockout mice have multiple

developmental defects that are non-overlapping with other TGFβ knockout

phenotypes, Development. 124 (1997) 2659–2670.

[95] V. Kaartinen, J.W. Voncken, C. Shuler, D. Warburton, D. Bu, N. Heisterkamp, J.

Groffen, Abnormal lung development and cleft palate in mice lacking TGF–β3

indicates defects of epithelial–mesenchymal interaction, Nat. Genet. 11 (1995) 415–

421. https://doi.org/10.1038/ng1295-415.

[96] C.K. Tan, H.C. Chong, E.H.P. Tan, N.S. Tan, Getting “Smad” about obesity and

diabetes, Nutr. Diabetes. 2 (2012) e29–e29. https://doi.org/10.1038/nutd.2012.1.

105

[97] H. Yadav, C. Quijano, A.K. Kamaraju, O. Gavrilova, R. Malek, W. Chen, P. Zerfas, D.

Zhigang, E.C. Wright, C. Stuelten, P. Sun, S. Lonning, M. Skarulis, A.E. Sumner, T.

Finkel, S.G. Rane, Protection from obesity and diabetes by blockade of TGF-β/Smad3

signaling, Cell Metab. 14 (2011) 67–79. https://doi.org/10.1016/j.cmet.2011.04.013.

[98] Y. Tsurutani, M. Fujimoto, M. Takemoto, H. Irisuna, M. Koshizaka, S. Onishi, T.

Ishikawa, M. Mezawa, P. He, S. Honjo, Y. Maezawa, Y. Saito, K. Yokote, The roles of

transforming growth factor-β and Smad3 signaling in adipocyte differentiation and

obesity, Biochem. Biophys. Res. Commun. 407 (2011) 68–73.

https://doi.org/10.1016/j.bbrc.2011.02.106.

[99] R.L. Sparks, B.J. Allen, E.E. Strauss, TGF‐β blocks early but not late differentiation‐

specific gene expression and morphologic differentiation of 3T3 T proadipocytes, J.

Cell. Physiol. 150 (1992) 568–577. https://doi.org/10.1002/jcp.1041500318.

[100] L. Choy, R. Derynck, Transforming growth factor-β inhibits adipocyte differentiation

by Smad3 interacting with CCAAT/enhancer-binding protein (C/EBP) and repressing

C/EBP transactivation function, J. Biol. Chem. 278 (2003) 9609–9619.

https://doi.org/10.1074/jbc.M212259200.

[101] H. Liang, W.F. Ward, PGC-1α: A key regulator of energy metabolism, Am. J. Physiol.

- Adv. Physiol. Educ. 30 (2006) 145–151.

https://doi.org/10.1152/advan.00052.2006.

[102] U.D. Wankhade, J.H. Lee, P.K. Dagur, H. Yadav, M. Shen, W. Chen, A.B. Kulkarni, J.P.

McCoy, T. Finkel, A.M. Cypess, S.G. Rane, TGF-β receptor 1 regulates progenitors that

promote browning of white fat, Mol. Metab. 16 (2018) 160–171.

https://doi.org/10.1016/j.molmet.2018.07.008.

[103] C.K. Tan, N. Leuenberger, M.J. Tan, Y.W. Yan, Y. Chen, R. Kambadur, W. Wahli, N.S.

Tan, Smad3 deficiency in mice protects against insulin resistance and obesity

induced by a high-fat diet, Diabetes. 60 (2011) 464–476.

https://doi.org/10.2337/db10-0801.

106

[104] I.B. Robertson, M. Horiguchi, L. Zilberberg, B. Dabovic, K. Hadjiolova, D.B. Rifkin,

Latent TGF-β-binding proteins, Elsevier, 2015.

https://doi.org/10.1016/j.matbio.2015.05.005.

[105] K. Yoshinaga, H. Obata, V. Jurukovski, R. Mazzieri, Y. Chen, L. Zilberberg, D. Huso, J.

Melamed, P. Prijatelj, V. Todorovic, B. Dabovic, D.B. Rifkin, Perturbation of

transforming growth factor (TGF)-β1 association with latent TGF-β binding protein

yields inflammation and tumors, Proc. Natl. Acad. Sci. U. S. A. 105 (2008) 18758–

18763. https://doi.org/10.1073/pnas.0805411105.

[106] D.B. Rifkin, W.J. Rifkin, L. Zilberberg, LTBPs in biology and medicine: LTBP diseases,

Matrix Biol. 71–72 (2018) 90–99. https://doi.org/10.1016/j.matbio.2017.11.014.

[107] V. Todorovic, D.B. Rifkin, LTBPs, more than just an escort service, J. Cell. Biochem.

113 (2012) 410–418. https://doi.org/10.1002/jcb.23385.

[108] M.R. Davis, R. Andersson, J. Severin, M. de Hoon, N. Bertin, J.K. Baillie, H. Kawaji, A.

Sandelin, A.R.R. Forrest, K.M. Summers, Transcriptional profiling of the human

fibrillin/LTBP gene family, key regulators of mesenchymal cell functions, Mol. Genet.

Metab. 112 (2014) 73–83. https://doi.org/10.1016/J.YMGME.2013.12.006.

[109] I.B. Robertson, D.B. Rifkin, Regulation of the bioavailability of TGF-β and TGF-β-

related proteins, Cold Spring Harb. Perspect. Biol. 8 (2016).

https://doi.org/10.1101/cshperspect.a021907.

[110] J. Saharinen, J. Keski-Oja, Specific sequence motif of 8-Cys repeats of TGF-β binding

proteins, LTBPs, creates a hydrophobic interaction surface for binding of small latent

TGF-β, Mol. Biol. Cell. 11 (2000) 2691–2704.

https://doi.org/10.1091/mbc.11.8.2691.

[111] C. Unsöld, M. Hyytiäinen, L. Bruckner-Tuderman, J. Keski-Oja, Latent TGF-β binding

protein LTBP-1 contains three potential extracellular matrix interacting domains, J.

Cell Sci. 114 (2001) 187–197.

107

[112] S.L. Dallas, P. Sivakumar, C.J.P. Jones, Q. Chen, D.M. Peters, D.F. Mosher, M.J.

Humphries, C.M. Kielty, Fibronectin regulates latent transforming growth factor-β

(TGFβ) by controlling matrix assembly of latent TGFβ-binding protein-1, J. Biol.

Chem. 280 (2005) 18871–18880. https://doi.org/10.1074/jbc.M410762200.

[113] J.P. Annes, Y. Chen, J.S. Munger, D.B. Rifkin, Integrin αvβ6-mediated activation of

latent TGF-β requires the latent TGF-β binding protein-1, J. Cell Biol. 165 (2004) 723–

734. https://doi.org/10.1083/jcb.200312172.

[114] L. Fontana, Y. Chen, P. Prijatelj, T. Sakai, R. Fässler, L.Y. Sakai, D.B. Rifkin, Fibronectin

is required for integrin o:vβ6‐mediated activation of latent TGF‐β complexes

containing LTBP‐1, FASEB J. 19 (2005) 1798–1808. https://doi.org/10.1096/fj.05-

4134com.

[115] A. Gualandris, J.P. Annes, M. Arese, I. Noguera, V. Jurukovski, D.B. Rifkin, The Latent

Transforming Growth Factor-beta -binding Protein-1 Promotes In Vitro

Differentiation of Embryonic Stem Cells into Endothelium, Mol. Biol. Cell. 11 (2000)

4295–4308. https://doi.org/10.1091/mbc.11.12.4295.

[116] V. Todorovic, D. Frendewey, D.E. Gutstein, Y. Chen, L. Freyer, E. Finnegan, F. Liu, A.

Murphy, D. Valenzuela, G. Yancopoulos, D.B. Rifkin, Long form of latent TGF-β

binding protein 1 (Ltbp1L) is essential for cardiac outflow tract septation and

remodeling, Development. 134 (2007) 3723–3732.

https://doi.org/10.1242/dev.008599.

[117] M. Horiguchi, V. Todorovic, K. Hadjiolova, R. Weiskirchen, D.B. Rifkin, Abrogation of

both short and long forms of latent transforming growth factor-β binding protein-1

causes defective cardiovascular development and is perinatally lethal, Matrix Biol.

43 (2015) 61–70. https://doi.org/10.1016/j.matbio.2015.03.006.

[118] T. Inoue, T. Ohbayashi, Y. Fujikawa, H. Yoshida, T.O. Akama, K. Noda, M. Horiguchi,

K. Kameyama, Y. Hata, K. Takahashi, K. Kusumoto, T. Nakamura, Latent TGF-β

binding protein-2 is essential for the development of ciliary zonule microfibrils, Hum.

Mol. Genet. 23 (2014) 5672–5682. https://doi.org/10.1093/hmg/ddu283.

108

[119] M. Narooie-Nejad, S.H. Paylakhi, S. Shojaee, Z. Fazlali, M. Rezaei Kanavi, N.

Nilforushan, S. Yazdani, F. Babrzadeh, F. Suri, M. Ronaghi, E. Elahi, C. Paisán-Ruiz,

Loss of function mutations in the gene encoding latent transforming growth factor

beta binding protein 2, LTBP2, cause primary congenital glaucoma, Hum. Mol. Genet.

18 (2009) 3969–3977. https://doi.org/10.1093/hmg/ddp338.

[120] M. Ali, M. McKibbin, A. Booth, D.A. Parry, P. Jain, S.A. Riazuddin, J.F. Hejtmancik, S.N.

Khan, S. Firasat, M. Shires, D.F. Gilmour, K. Towns, A.L. Murphy, D. Azmanov, I.

Tournev, S. Cherninkova, H. Jafri, Y. Raashid, C. Toomes, J. Craig, D.A. Mackey, L.

Kalaydjieva, S. Riazuddin, C.F. Inglehearn, Null Mutations in LTBP2 Cause Primary

Congenital Glaucoma, Am. J. Hum. Genet. 84 (2009) 664–671.

https://doi.org/10.1016/j.ajhg.2009.03.017.

[121] M. Hirai, M. Horiguchi, T. Ohbayashi, T. Kita, K.R. Chien, T. Nakamura, Latent TGF-β-

binding protein 2 binds to DANCE/fibulin-5 and regulates elastic fiber assembly,

EMBO J. 26 (2007) 3283–3295. https://doi.org/10.1038/sj.emboj.7601768.

[122] P. Vehviläinen, M. Hyytiäinen, J. Keski-Oja, Matrix association of latent TGF-beta

binding protein-2 (LTBP-2) is dependent on fibrillin-1, J. Cell. Physiol. 221 (2009) 586–

593. https://doi.org/10.1002/jcp.21888.

[123] X. Sun, R. Essalmani, D. Susan-Resiga, A. Prat, N.G. Seidah, Latent transforming

growth factor β-binding proteins-2 and -3 inhibit the proprotein convertase 5/6A, J.

Biol. Chem. 286 (2011) 29063–29073. https://doi.org/10.1074/jbc.M111.242479.

[124] S.B. Anderson, A.L. Goldberg, M. Whitman, Identification of a novel pool of

extracellular pro-myostatin in skeletal muscle, J. Biol. Chem. 283 (2008) 7027–7035.

https://doi.org/10.1074/jbc.M706678200.

[125] Y. Chen, B. Dabovic, J.P. Annes, D.B. Rifkin, Latent TGF-β binding protein-3 (LTBP-3)

requires binding to TGF-β for secretion, FEBS Lett. 517 (2002) 277–280.

https://doi.org/10.1016/S0014-5793(02)02648-0.

109

[126] S. Morkmued, J. Hemmerle, E. Mathieu, V. Laugel-Haushalter, B. Dabovic, D.B. Rifkin,

P. Dollé, K. Niederreither, A. Bloch-Zupan, Enamel and dental anomalies in latent-

transforming growth factor beta-binding protein 3 mutant mice, Eur. J. Oral Sci. 125

(2017) 8–17. https://doi.org/10.1111/eos.12328.

[127] B. Dabovic, Y. Chem, C. Colarossi, L. Zambuto, H. Obata, D.B. Rifkin, Bone defects in

latent TGF-β binding protein (Ltbp)-3 null mice; a role for Ltbp in TGF-β presentation,

J. Endocrinol. 175 (2002) 129–141. https://doi.org/10.1677/joe.0.1750129.

[128] A. Sterner-Kock, I.S. Thorey, K. Koli, F. Wempe, J. Otte, T. Bangsow, K. Kuhlmeier, T.

Kirchner, S. Jin, J. Keski-Oja, H. Von Melchner, Disruption of the gene encoding the

latent transforming growth factor-β binding protein 4 (LTBP-4) causes abnormal lung

development, cardiomyopathy, and colorectal cancer, Genes Dev. 16 (2002) 2264–

2273. https://doi.org/10.1101/gad.229102.

[129] K. Noda, B. Dabovic, K. Takagi, T. Inoue, M. Horiguchi, M. Hirai, Y. Fujikawa, T.O.

Akama, K. Kusumoto, L. Zilberberg, L.Y. Sakai, K. Koli, M. Naitoh, H. Von Melchner, S.

Suzuki, D.B. Rifkin, T. Nakamura, Latent TGF-β binding protein 4 promotes elastic

fiber assembly by interacting with fibulin-5, Proc. Natl. Acad. Sci. U. S. A. 110 (2013)

2852–2857. https://doi.org/10.1073/pnas.1215779110.

[130] B. Dabovic, I.B. Robertson, L. Zilberberg, M. Vassallo, E.C. Davis, D.B. Rifkin, Function

of latent TGFβ binding protein 4 and fibulin 5 in elastogenesis and lung development,

J. Cell. Physiol. 230 (2015) 226–236. https://doi.org/10.1002/jcp.24704.

[131] M. Wabitsch, R.E. Brenner, I. Melzner, M. Braun, P. Möller, E. Heinze, K.M. Debatin,

H. Hauner, Characterization of a human preadipocyte cell strain with high capacity

for adipose differentiation, Int. J. Obes. 25 (2001) 8–15.

https://doi.org/10.1038/sj.ijo.0801520.

[132] D. Tews, T. Pula, J.B.B. Funcke, M. Jastroch, M. Keuper, K.M.M. Debatin, M.

Wabitsch, P. Fischer-Posovszky, Elevated UCP1 levels are sufficient to improve

glucose uptake in human white adipocytes, Redox Biol. 26 (2019) 101286.

https://doi.org/10.1016/j.redox.2019.101286.

110

[133] T. Reisser, D. Halbgebauer, J. Scheurer, L. Wolf, F. Leithäuser, N. Beyersdorf, P.

Fischer-Posovszky, K.-M. Debatin, G. Strauss, In vitro-generated alloantigen-specific

Th9 cells mediate antileukemia cytotoxicity in the absence of graft-versus-host

disease, Leukemia. (2020) 4–9. https://doi.org/10.1038/s41375-020-0731-2.

[134] M. Jinek, K. Chylinski, I. Fonfara, M. Hauer, J.A. Doudna, E. Charpentier, A

programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity,

Science (80-. ). 337 (2012) 816–821. https://doi.org/10.1126/science.1225829.

[135] P. Mali, L. Yang, K.M. Esvelt, J. Aach, M. Guell, J.E. DiCarlo, J.E. Norville, G.M. Church,

RNA-guided human genome engineering via Cas9, Science (80-. ). 339 (2013) 823–

826. https://doi.org/10.1126/science.1232033.

[136] J. Albers, C. Danzer, M. Rechsteiner, H. Lehmann, L.P. Brandt, T. Hejhal, A. Catalano,

P. Busenhart, A.F. Gonçalves, S. Brandt, P.K. Bode, B. Bode-Lesniewska, P.J. Wild, I.J.

Frew, A versatile modular vector system for rapid combinatorial mammalian

genetics, J. Clin. Invest. 125 (2015) 1603–1619. https://doi.org/10.1172/JCI79743.

[137] F. Katzen, Gateway® recombinational cloning: A biological operating system, Expert

Opin. Drug Discov. 2 (2007) 571–589. https://doi.org/10.1517/17460441.2.4.571.

[138] J.G. Doench, N. Fusi, M. Sullender, M. Hegde, E.W. Vaimberg, K.F. Donovan, I. Smith,

Z. Tothova, C. Wilen, R. Orchard, H.W. Virgin, J. Listgarten, D.E. Root, Optimized

sgRNA design to maximize activity and minimize off-target effects of CRISPR-Cas9.,

Nat. Biotechnol. 34 (2016) 184–91. https://doi.org/10.1038/nbt.3437.

[139] K.R. Sanson, R.E. Hanna, M. Hegde, K.F. Donovan, C. Strand, M.E. Sullender, E.W.

Vaimberg, A. Goodale, D.E. Root, F. Piccioni, J.G. Doench, Optimized libraries for

CRISPR-Cas9 genetic screens with multiple modalities, Nat. Commun. 9 (2018).

https://doi.org/10.1038/s41467-018-07901-8.

[140] E. Kowarz, D. Löscher, R. Marschalek, Optimized Sleeping Beauty transposons rapidly

generate stable transgenic cell lines, Biotechnol. J. 10 (2015) 647–653.

https://doi.org/10.1002/biot.201400821.

111

[141] L. Mátés, M.K.L. Chuah, E. Belay, B. Jerchow, N. Manoj, A. Acosta-Sanchez, D.P.

Grzela, A. Schmitt, K. Becker, J. Matrai, L. Ma, E. Samara-Kuko, C. Gysemans, D.

Pryputniewicz, C. Miskey, B. Fletcher, T. Vandendriessche, Z. Ivics, Z. Izsvák,

Molecular evolution of a novel hyperactive Sleeping Beauty transposase enables

robust stable gene transfer in vertebrates, Nat. Genet. 41 (2009) 753–761.

https://doi.org/10.1038/ng.343.

[142] Broad Institute, sgRNA Designer: CRISPRko, (2018).

https://portals.broadinstitute.org/gpp/public/analysis-tools/sgrna-design (accessed

April 2, 2020).

[143] P. Fischer-Posovszky, F.S. Newell, M. Wabitsch, H.E. Tornqvist, Human SGBS cells - A

unique tool for studies of human fat cell biology, Obes. Facts. 1 (2008) 184–189.

https://doi.org/10.1159/000145784.

[144] D. Tews, T. Fromme, M. Keuper, S.M. Hofmann, K.M. Debatin, M. Klingenspor, M.

Wabitsch, P. Fischer-Posovszky, Teneurin-2 (TENM2) deficiency induces UCP1

expression in differentiating human fat cells, Mol. Cell. Endocrinol. 443 (2017) 106–

113. https://doi.org/10.1016/j.mce.2017.01.015.

[145] D. Tews, P. Fischer-Posovszky, T. Fromme, M. Klingenspor, J. Fischer, U. Rüther, R.

Marienfeld, T.F. Barth, P. Möller, K.M. Debatin, M. Wabitsch, FTO deficiency induces

UCP-1 expression and mitochondrial uncoupling in adipocytes, Endocrinology. 154

(2013) 3141–3151. https://doi.org/10.1210/en.2012-1873.

[146] E.C.M. Mariman, P. Wang, Adipocyte extracellular matrix composition, dynamics and

role in obesity, Cell. Mol. Life Sci. . 67 (2010) 1277. https://doi.org/10.1007/S00018-

010-0263-4.

[147] A. Koncarevic, S. Kajimura, M. Cornwall-Brady, A. Andreucci, A. Pullen, D. Sako, R.

Kumar, A. V. Grinberg, K. Liharska, J.A. Ucran, E. Howard, B.M. Spiegelman, J. Seehra,

J. Lachey, A novel therapeutic approach to treating obesity through modulation of

TGFβ signaling, Endocrinology. 153 (2012) 3133–3146.

https://doi.org/10.1210/en.2012-1016.

112

[148] B. Dabovic, Y. Chen, J. Choi, E.C. Davis, L.Y. Sakai, V. Todorovic, M. Vassallo, L.

Zilberberg, A. Singh, D.B. Rifkin, Control of Lung Development by Latent TGF-β

Binding Proteins, (n.d.). https://doi.org/10.1002/jcp.22479.

[149] C. Penttinen, J. Saharinen, K. Weikkolainen, M. Hyytiäinen, J. Keski-Oja, Secretion of

human latent TGF-beta-binding protein-3 (LTBP-3) is dependent on co-expression of

TGF-beta., J. Cell Sci. 115 (2002) 3457–68.

http://www.ncbi.nlm.nih.gov/pubmed/12154076 (accessed March 26, 2020).

[150] A.L. Hafner, J. Contet, C. Ravaud, X. Yao, P. Villageois, K. Suknuntha, K. Annab, P.

Peraldi, B. Binetruy, I.I. Slukvin, A. Ladoux, C. Dani, Brown-like adipose progenitors

derived from human induced pluripotent stem cells: Identification of critical

pathways governing their adipogenic capacity, Sci. Rep. 6 (2016).

https://doi.org/10.1038/srep32490.

[151] Y. Takeda, P. Dai, A developed serum-free medium and an optimized chemical

cocktail for direct conversion of human dermal fibroblasts into brown adipocytes,

Sci. Rep. 10 (2020) 1–13. https://doi.org/10.1038/s41598-020-60769-x.

[152] C. Sirard, J.L. De La Pompa, A. Elia, A. Itie, C. Mirtsos, A. Cheung, S. Hahn, A.

Wakeham, L. Schwartz, S.E. Kern, J. Rossant, T.W. Mak, The tumor suppressor gene

Smad4/Dpc4 is required for gastrulation and later for anterior development of the

mouse embryo, Genes Dev. 12 (1998) 107–119.

https://doi.org/10.1101/gad.12.1.107.

[153] S. Modica, L.G. Straub, M. Balaz, W. Sun, C. Wolfrum, L. Varga, B. Ukropcova, J.

Ukropec, L. Varga, P. Stefanicka, M. Profant, B. Ukropcova, H. Neubauer, E. Simon,

Bmp4 Promotes a Brown to White-like Adipocyte Shift, Cell Rep. 16 (2016) 2243–

2258. https://doi.org/10.1016/j.celrep.2016.07.048.

113

[154] P. Petrus, N. Mejhert, P. Corrales, S. Lecoutre, Q. Li, E. Maldonado, A. Kulyté, Y.

Lopez, M. Campbell, J.R. Acosta, J. Laurencikiene, I. Douagi, H. Gao, C. Martínez-

Álvarez, P. Hedén, K.L. Spalding, A. Vidal-Puig, G. Medina-Gomez, P. Arner, M. Rydén,

P. Petrus, A. Vidal-Puig, Y. Lopez, H. Gao, S. Lecoutre, J. Laurencikiene, E. Maldonado,

M. Rydén, N. Mejhert, A. Kulyté, P. Hedén, M. Campbell, Q. Li, G. Medina-Gomez, P.

Arner, K.L. Spalding, P. Corrales, J.R. Acosta, Transforming Growth Factor-β3

Regulates Adipocyte Number in Subcutaneous White Adipose Tissue, Cell Rep. 25

(2018) 551-560.e5. https://doi.org/10.1016/j.celrep.2018.09.069.

[155] A.M. McInerney-Leo, C. Le Goff, P.J. Leo, T.J. Kenna, P. Keith, J.E. Harris, R. Steer, C.

Bole-Feysot, P. Nitschke, C. Kielty, M.A. Brown, A. Zankl, E.L. Duncan, V. Cormier-

Daire, Mutations in LTBP3 cause acromicric dysplasia and geleophysic dysplasia, J.

Med. Genet. 53 (2016) 457–464. https://doi.org/10.1136/jmedgenet-2015-103647.

[156] M.E. Lindsay, D. Schepers, N.A. Bolar, J.J. Doyle, E. Gallo, J. Fert-Bober, M.J.E.

Kempers, E.K. Fishman, Y. Chen, L. Myers, D. Bjeda, G. Oswald, A.F. Elias, H.P. Levy,

B.M. Anderlid, M.H. Yang, E.M.H.F. Bongers, J. Timmermans, A.C. Braverman, N.

Canham, G.R. Mortier, H.G. Brunner, P.H. Byers, J. Van Eyk, L. Van Laer, H.C. Dietz,

B.L. Loeys, Loss-of-function mutations in TGFB2 cause a syndromic presentation of

thoracic aortic aneurysm, Nat. Genet. 44 (2012) 922–927.

https://doi.org/10.1038/ng.2349.

[157] C.C. Chan, M.S.M.A. Damen, P.C. Alarcon, J. Sanchez-Gurmaches, S. Divanovic,

Inflammation and Immunity: From an Adipocyte’s Perspective, J. Interf. Cytokine

Res. 39 (2019) 459–471. https://doi.org/10.1089/jir.2019.0014.

[158] A. Rosenow, T.N. Arrey, F.G. Bouwman, J.P. Noben, M. Wabitsch, E.C.M. Mariman,

M. Karas, J. Renes, Identification of novel human adipocyte secreted proteins by

using SGBS cells, J. Proteome Res. 9 (2010) 5389–5401.

https://doi.org/10.1021/pr100621g.

[159] Y. Zhang, Q. Liu, J. Yu, S. Yu, J. Wang, L. Qiang, Z. Gu, Locally Induced Adipose Tissue

Browning by Microneedle Patch for Obesity Treatment, ACS Nano. 11 (2017) 9223–

9230. https://doi.org/10.1021/acsnano.7b04348.

114

[160] C.-H. Wang, M. Lundh, A. Fu, R. Kriszt, T.L. Huang, M.D. Lynes, L.O. Leiria, F. Shamsi,

J. Darcy, B.P. Greenwood, N.R. Narain, V. Tolstikov, K.L. Smith, B. Emanuelli, Y.-T.

Chang, S. Hagen, N.N. Danial, M.A. Kiebish, Y.-H. Tseng, CRISPR-engineered human

brown-like adipocytes prevent diet-induced obesity and ameliorate metabolic

syndrome in mice., Sci. Transl. Med. 12 (2020).

https://doi.org/10.1126/scitranslmed.aaz8664.

[161] Pipeline | Biontech, (n.d.). https://biontech.de/de/science/pipeline (accessed

September 7, 2020).

[162] Pipeline – CureVac, (n.d.). https://www.curevac.com/en/pipeline/ (accessed

September 7, 2020).

[163] C. Jiang, M.A. Cano-Vega, F. Yue, L. Kuang, N. Narayanan, G. Uzunalli, M.P. Merkel,

S. Kuang, M. Deng, Dibenzazepine-Loaded Nanoparticles Induce Local Browning of

White Adipose Tissue to Counteract Obesity, Mol. Ther. 25 (2017) 1718–1729.

https://doi.org/10.1016/j.ymthe.2017.05.020.

[164] Y. Xue, X. Xu, X.Q. Zhang, O.C. Farokhzad, R. Langer, Preventing diet-induced obesity

in mice by adipose tissue transformation and angiogenesis using targeted

nanoparticles, Proc. Natl. Acad. Sci. U. S. A. 113 (2016) 5552–5557.

https://doi.org/10.1073/pnas.1603840113.

115

7. Register of illustrations and figure permissions

Table 1: In vivo knockout phenotype and TGFβ binding properties of the LTBP isoforms. ............................ 16

Table 2: Primer used for qPCR or sequencing .............................................................................................. 30

Table 3: Small interfering RNAs used for in vitro gene knockdown .............................................................. 32

Table 4: Antibodies used for Western blot ................................................................................................... 32

Table 5: sgRNA used for CRISPR/Cas9 mediated gene knockout .................................................................. 33

Table 6: Plasmids used for cloning ............................................................................................................... 34

Table 7: Primer used for Gen Art Genomic Cleavage Assay .......................................................................... 35

Table 8: Seeding densities for SGBS cell differentiation approach ................................................................ 38

Table 9: SDS-PAGE running conditions for different protein properties ....................................................... 41

Table 10: Plasmids used for LR Gateway cloning for single CRIPSR/Cas9 knockout constructs ..................... 49

Table 11: Plasmids used for LR Gateway cloning for triple CRIPSR/Cas9 knockout construct ........................ 49

Figure 1: Morphology of different types of adipocytes. ................................................................................. 2

Figure 2: Molecular mechanism of BAT activation. ........................................................................................ 4

Figure 3: Distribution of adipose tissue and its temperature sensitivity in mice ............................................. 8

Figure 4: TGFβ-SMAD signaling pathway...................................................................................................... 11

Figure 5: Secretion of the large latent complex (LLC). .................................................................................. 14

Figure 6: Study design to identify new factors involved in the adipogenesis towards a brown adipocyte

phenotype. .................................................................................................................................................. 17

Figure 7: LR Gateway cloning with multiple fragments ................................................................................ 45

Figure 8: Expression of LTBP family members in isolated hASCs from subcutaneous and deep neck adipose

tissue ........................................................................................................................................................... 53

Figure 9: Expression of LTBP isoforms in human adipose tissue and in isolated and in in vitro differentiated

human adipose derived stem cells ............................................................................................................... 53

Figure 10: Expression of differentiation marker and LTBP members in SGBS cells during adipogenesis........ 55

Figure 11: Transfection with CRISPR/Cas9 knockout constructs leads to stable downregulation of LTBPs in

SGBS cells .................................................................................................................................................... 57

Figure 12: Reduced LTBP expression has no effect on adipogenic differentiation and mitochondrial content..

.................................................................................................................................................................... 59

Figure 13: LTBP2 and LTBP3 deficiency leads to a decreased UCP1 expression ............................................ 60

Figure 14: Schematic view of a seahorse plot .............................................................................................. 62

Figure 15: LTBP deficiency alters the metabolic function of SGBS adipocytes .............................................. 63

Figure 16: LTBP3 correlates significant positive with UCP1 mRNA expression in human subcutaneous adipose

tissue ........................................................................................................................................................... 64

Figure 17: TGFβ ligands and TGFβ receptors are expressed during adipogenesis of SGBS cells. ................... 66

Figure 18: TGFβ ligand and receptor expression of isolated hASCs from dn and sc adipose tissue ............... 66

116

Figure 19: Transfection with CRISPR/Cas9 triple knockout construct leads to stable downregulation of TGFβ

ligands in SGBS cells during adipogenesis .................................................................................................... 67

Figure 20: No changes on adipogenic differentiation and mitochondrial content after CRISPR/Cas9 mediated

TGFβ1-3 knockout:. ..................................................................................................................................... 68

Figure 21: Decreased UCP1 expression in TGFβ1-3 deficient SGBS adipocytes. ............................................ 69

Figure 22: Inhibition of TGFβ receptor1 kinase leads to significant alterations in differentiation marker and

UCP1 expression. ......................................................................................................................................... 71

Figure 23: Decrease of UCP1 expression and oxygen consumption rate upon SMAD4 knockdown. ............. 72

Figure 24: LTBP2 and LTBP3 deficiency alters the TGFβ2 content in SGBS preadipocyte supernatant .......... 73

Figure 25: TGFβ2 deficiency leads to a higher differentiation rate ............................................................... 74

Figure 26: TGFβ2 deficiency leads to a reduction of UCP1 and a decreased response to cAMP. ................... 75

Figure 27: TGFβ2 mRNA expression correlates positive with UCP1 in human subcutaneous adipose tissue

samples. ...................................................................................................................................................... 76

117

Figure permission

Figure 2 was taken from P.G. Crichton, Y. Lee, E.R.S. Kunji, The molecular features of

uncoupling protein 1 support a conventional mitochondrial carrier-like mechanism,

Biochimie. 134 (2017) 35–50. https://doi.org/10.1016/j.biochi.2016.12.016.

I have read and understood the user license: https://creativecommons.org/licenses/by-nc-

nd/4.0/; accessed on 14.09.2020

Figure 3 was taken from J. Sanchez-Gurmaches, C.M. Hung, D.A. Guertin, Emerging

Complexities in Adipocyte Origins and Identity, Trends Cell Biol. 26 (2016) 313–326.

https://doi.org/10.1016/j.tcb.2016.01.004

in agreement with Elsevier.

Figure 4 was taken from E.H. Budi, D. Duan, R. Derynck, Transforming Growth Factor-β

Receptors and Smads: Regulatory Complexity and Functional Versatility, Trends Cell Biol.

27 (2017) 658–672. https://doi.org/10.1016/j.tcb.2017.04.005.

I have read an understood the terms and conditions of using AAA material in a thesis or

dissertation: https://www.sciencemag.org/help/reprints-and-permissions; accessed on

14.09.2020.

Figure 5 was taken from V. Todorovic, D.B. Rifkin, LTBPs, more than just an escort service,

J. Cell. Biochem. 113 (2012) 410–418. https://doi.org/10.1002/jcb.23385

in agreement with John Wiley and Sons.

118

Appendix I

pMuSE SB CRISPR/Cas9 sgRNA:

Figure S1: Vector map of the final CRISPR/Cas9 knockout construct for single gene knockout. The vector carries eGFP (bright green) and puromycin resistance (mint green) under the control of a RPL13A promoter. Both genes are separated by a P2A sequence which enables cleavage of the gene product into two separate protein. Cas9 (red) expression is controlled by a CMV promotor and the expression of the introduced sgRNA cassette (orange) is under the control of a U6 promoter. These genes are flanked by inverted repeats (ITRs, light blue), the SB transposase recognition sites.

Sequence pMuSE SB CRISPR/Cas9 sgRNA:

atgaccgagtacaagcccacGGTGCGCCTCGCCACCCGCGACGACGTCCCCAGGGCCGTACGCACCCTCGCCGCCGCGTTCGCCGACTACCCCGCCACGCGCCACACCGTCGATCCGGACCGCCACATCGAGCG

GGTCACCGAGCTGCAAGAACTCTTCCTCACGCGCGTCGGGCTCGACATCGGCAAGGTGTGGGTCGCGGACGACGGCGCCGCGGTGGCGGTCTGGACCACGCCGGAGAGCGTCGAAGCGGGGGCGGTG

TTCGCCGAGATCGGCCCGCGCATGGCCGAGTTGAGCGGTTCCCGGCTGGCCGCGCAGCAACAGATGGAAGGCCTCCTGGCGCCGCACCGGCCCAAGGAGCCCGCGTGGTTCCTGGCCACCGTCGGCGTC

TCGCCCGACCACCAGGGCAAGGGTCTGGGCAGCGCCGTCGTGCTCCCCGGAGTGGAGGCGGCCGAGCGCGCCGGGGTGCCCGCCTTCCTGGAGACCTCCGCGCCCCGCAACCTCCCCTTCTACGAGCGG

CTCGGCTTCACCGTCACCGCCGACGTCGAGGTGCCCGAAGGACCGCGCACCTGGTGCATGACCCGCAAGCCCGGTGCCTGATTCGAAGGCCTGTCGTGAAGCTTGGGGATCAATTCTCTAGAGCTCGCTG

ATCAGCCTCGACTGTGCCTTCTAGTTGCCAGCCATCTGTTGTTTGCCCCTCCCCCGTGCCTTCCTTGACCCTGGAAGGTGCCACTCCCACTGTCCTTTCCTAATAAAATGAGGAAATTGCATCGCATTGTCTG

AGTAGGTGTCATTCTATTCTGGGGGGTGGGGTGGGGCAGGACAGCAAGGGGGAGGATTGGGAAGACAATAGCAGGCATGCTGGGGATGCGGTGGGCTCTATGGCTTCTGAGGCGGAAAGAACCAGCT

GGGCCTACTAGCTACTCGGGACCCCTTACCGAAACATCGCCGCATTCTGCAGAGGAGTCGAGTGTATGTAAACTTCTGACCCACTGGGAATGTGATGAAAGAAATAAAAGCTGAAATGAATCATTCTCTCT

ACTATTATTCTGATATTTCACATTCTTAAAATAAAGTGGTGATCCTAACTGACCTAAGACAGGGAATTTTTACTAGGATTAAATGTCAGGAATTGTGAAAAAGTGAGTTTAAATGTATTTGGCTAAGGTGTA

TGTAAACTTCCGACTTCAACTGTATAGGGATCCGCTTCCTCGCTCACTGACTCGCTGCGCTCGGTCGTTCGGCTGCGGCGAGCGGTATCAGCTCACTCAAAGGCGGTAATACGGTTATCCACAGAATCAGG

GGATAACGCAGGAAAGAACATGTGAGCAAAAGGCCAGCAAAAGGCCAGGAACCGTAAAAAGGCCGCGTTGCTGGCGTTTTTCCATAGGCTCCGCCCCCCTGACGAGCATCACAAAAATCGACGCTCAAG

TCAGAGGTGGCGAAACCCGACAGGACTATAAAGATACCAGGCGTTTCCCCCTGGAAGCTCCCTCGTGCGCTCTCCTGTTCCGACCCTGCCGCTTACCGGATACCTGTCCGCCTTTCTCCCTTCGGGAAGCGT

GGCGCTTTCTCATAGCTCACGCTGTAGGTATCTCAGTTCGGTGTAGGTCGTTCGCTCCAAGCTGGGCTGTGTGCACGAACCCCCCGTTCAGCCCGACCGCTGCGCCTTATCCGGTAACTATCGTCTTGAGTC

CAACCCGGTAAGACACGACTTATCGCCACTGGCAGCAGCCACTGGTAACAGGATTAGCAGAGCGAGGTATGTAGGCGGTGCTACAGAGTTCTTGAAGTGGTGGCCTAACTACGGCTACACTAGAAGAAC

AGTATTTGGTATCTGCGCTCTGCTGAAGCCAGTTACCTTCGGAAAAAGAGTTGGTAGCTCTTGATCCGGCAAACAAACCACCGCTGGTAGCGGTGGTTTTTTTGTTTGCAAGCAGCAGATTACGCGCAGAA

AAAAAGGATCTCAAGAAGATCCTTTGATCTTTTCTACGGGGTCTGACGCTCAGTGGAACGAAAACTCACGTTAAGGGATTTTGGTCATGAGATTATCAAAAAGGATCTTCACCTAGATCCTTTTAAATTAAA

AATGAAGTTTTAAATCAATCTAAAGTATATATGAGTAAACTTGGTCTGACAGTTACCAATGCTTAATCAGTGAGGCACCTATCTCAGCGATCTGTCTATTTCGTTCATCCATAGTTGCCTGACTCCCCGTCGT

GTAGATAACTACGATACGGGAGGGCTTACCATCTGGCCCCAGTGCTGCAATGATACCGCGAGACCCACGCTCACCGGCTCCAGATTTATCAGCAATAAACCAGCCAGCCGGAAGGGCCGAGCGCAGAAG

119

TGGTCCTGCAACTTTATCCGCCTCCATCCAGTCTATTAATTGTTGCCGGGAAGCTAGAGTAAGTAGTTCGCCAGTTAATAGTTTGCGCAACGTTGTTGCCATTGCTACAGGCATCGTGGTGTCACGCTCGTC

GTTTGGTATGGCTTCATTCAGCTCCGGTTCCCAACGATCAAGGCGAGTTACATGATCCCCCATGTTGTGCAAAAAAGCGGTTAGCTCCTTCGGTCCTCCGATCGTTGTCAGAAGTAAGTTGGCCGCAGTGTT

ATCACTCATGGTTATGGCAGCACTGCATAATTCTCTTACTGTCATGCCATCCGTAAGATGCTTTTCTGTGACTGGTGAGTACTCAACCAAGTCATTCTGAGAATAGTGTATGCGGCGACCGAGTTGCTCTTG

CCCGGCGTCAATACGGGATAATACCGCGCCACATAGCAGAACTTTAAAAGTGCTCATCATTGGAAAACGTTCTTCGGGGCGAAAACTCTCAAGGATCTTACCGCTGTTGAGATCCAGTTCGATGTAACCCA

CTCGTGCACCCAACTGATCTTCAGCATCTTTTACTTTCACCAGCGTTTCTGGGTGAGCAAAAACAGGAAGGCAAAATGCCGCAAAAAAGGGAATAAGGGCGACACGGAAATGTTGAATACTCATACTCTTC

CTTTTTCAATATTATTGAAGCATTTATCAGGGTTATTGTCTCATGAGCGGATACATATTTGAATGTATTTAGAAAAATAAACAAATAGGGGTTCCGCGCACATTTCCCCGAAAAGTGCCACCTGATGCGGTG

TGAAATACCGCACAGATGCGTAAGGAGAAAATACCGCATCAGGAAATTGTAAGCGTTAATATTTTGTTAAAATTGGATCCCTATACAGTTGAAGTCGGAAGTTTACATACACTTAAGTTGGAGTCATTAAA

ACTCGTTTTTCAACTACTCCACAAATTTCTTGTTAACAAACAATAGTTTTGGCAAGTCAGTTAGGACATCTACTTTGTGCATGACACAAGTCATTTTTCCAACAATTGTTTACAGACAGATTATTTCACTTATA

ATTCACTGTATCACAATTCCAGTGGGTCAGAAGTTTACATACACTAAGTTCGACTCCTCTGCAGAATGCGGCGATGTTTCGGTAAGGGGTCCGCTATCTAGACACAAGTTTGTACAAAAAAGCAGGCTCTA

TCGATCACGAGACTAGCCTCGAGCGGCCGCCCCCTTCACCGAGGGCCTATTTCCCATGATTCCTTCATATTTGCATATACGATACAAGGCTGTTAGAGAGATAATTGGAATTAATTTGACTGTAAACACAAA

GATATTAGTACAAAATACGTGACGTAGAAAGTAATAATTTCTTGGGTAGTTTGCAGTTTTAAAATTATGTTTTAAAATGGACTATCATATGCTTACCGTAACTTGAAAGTATTTCGATTTCTTGGCTTTATAT

ATCTTGTGGAAAGGACGAAACACCGxxxxxxxxxxxxxxxxxxxxGTTTTAGAGCTAGAAATAGCAAGTTAAAATAAGGCTAGTCCGTTATCAACTTGAAAAAGTGGCACCGAGTCGGTGCTTTTTTTGAATTCG

ACAACTTTGTATACAAAAGTTGGCGATGTACGGGCCAGATATACGCGTTGACATTGATTATTGACTAGTTATTAATAGTAATCAATTACGGGGTCATTAGTTCATAGCCCATATATGGAGTTCCGCGTTACA

TAACTTACGGTAAATGGCCCGCCTGGCTGACCGCCCAACGACCCCCGCCCATTGACGTCAATAATGACGTATGTTCCCATAGTAACGCCAATAGGGACTTTCCATTGACGTCAATGGGTGGAGTATTTACG

GTAAACTGCCCACTTGGCAGTACATCAAGTGTATCATATGCCAAGTACGCCCCCTATTGACGTCAATGACGGTAAATGGCCCGCCTGGCATTATGCCCAGTACATGACCTTATGGGACTTTCCTACTTGGCA

GTACATCTACGTATTAGTCATCGCTATTACCATGGTGATGCGGTTTTGGCAGTACATCAATGGGCGTGGATAGCGGTTTGACTCACGGGGATTTCCAAGTCTCCACCCCATTGACGTCAATGGGAGTTTGTT

TTGGCACCAAAATCAACGGGACTTTCCAAAATGTCGTAACAACTCCGCCCCATTGACGCAAATGGGCGGTAGGCGTGTACGGTGGGAGGTCTATATAAGCAGAGCTCTCTGGCTAACTAGAGAACCCACT

GCTTACTGGCTTATCGAAATTAATACGACTCACTATAGGGAGACCCAAGCTTGGTACCGCCACCATGGACAAGAAGTACTCCATTGGGCTCGATATCGGCACAAACAGCGTCGGCTGGGCCGTCATTACG

GACGAGTACAAGGTGCCGAGCAAAAAATTCAAAGTTCTGGGCAATACCGATCGCCACAGCATAAAGAAGAACCTCATTGGCGCCCTCCTGTTCGACTCCGGGGAGACGGCCGAAGCCACGCGGCTCAAA

AGAACAGCACGGCGCAGATATACCCGCAGAAAGAATCGGATCTGCTACCTGCAGGAGATCTTTAGTAATGAGATGGCTAAGGTGGATGACTCTTTCTTCCATAGGCTGGAGGAGTCCTTTTTGGTGGAGG

AGGATAAAAAGCACGAGCGCCACCCAATCTTTGGCAATATCGTGGACGAGGTGGCGTACCATGAAAAGTACCCAACCATATATCATCTGAGGAAGAAGCTTGTAGACAGTACTGATAAGGCTGACTTGC

GGTTGATCTATCTCGCGCTGGCGCATATGATCAAATTTCGGGGACACTTCCTCATCGAGGGGGACCTGAACCCAGACAACAGCGATGTCGACAAACTCTTTATCCAACTGGTTCAGACTTACAATCAGCTTT

TCGAAGAGAACCCGATCAACGCATCCGGAGTTGACGCCAAAGCAATCCTGAGCGCTAGGCTGTCCAAATCCCGGCGGCTCGAAAACCTCATCGCACAGCTCCCTGGGGAGAAGAAGAACGGCCTGTTTG

GTAATCTTATCGCCCTGTCACTCGGGCTGACCCCCAACTTTAAATCTAACTTCGACCTGGCCGAAGATGCCAAGCTTCAACTGAGCAAAGACACCTACGATGATGATCTCGACAATCTGCTGGCCCAGATCG

GCGACCAGTACGCAGACCTTTTTTTGGCGGCAAAGAACCTGTCAGACGCCATTCTGCTGAGTGATATTCTGCGAGTGAACACGGAGATCACCAAAGCTCCGCTGAGCGCTAGTATGATCAAGCGCTATGA

TGAGCACCACCAAGACTTGACTTTGCTGAAGGCCCTTGTCAGACAGCAACTGCCTGAGAAGTACAAGGAAATTTTCTTCGATCAGTCTAAAAATGGCTACGCCGGATACATTGACGGCGGAGCAAGCCAG

GAGGAATTTTACAAATTTATTAAGCCCATCTTGGAAAAAATGGACGGCACCGAGGAGCTGCTGGTAAAGCTTAACAGAGAAGATCTGTTGCGCAAACAGCGCACTTTCGACAATGGAAGCATCCCCCACC

AGATTCACCTGGGCGAACTGCACGCTATCCTCAGGCGGCAAGAGGATTTCTACCCCTTTTTGAAAGATAACAGGGAAAAGATTGAGAAAATCCTCACATTTCGGATACCCTACTATGTAGGCCCCCTCGCC

CGGGGAAATTCCAGATTCGCGTGGATGACTCGCAAATCAGAAGAGACCATCACTCCCTGGAACTTCGAGGAAGTCGTGGATAAGGGGGCCTCTGCCCAGTCCTTCATCGAAAGGATGACTAACTTTGATA

AAAATCTGCCTAACGAAAAGGTGCTTCCTAAACACTCTCTGCTGTACGAGTACTTCACAGTTTATAACGAGCTCACCAAGGTCAAATACGTCACAGAAGGGATGAGAAAGCCAGCATTCCTGTCTGGAGAG

CAGAAGAAAGCTATCGTGGACCTCCTCTTCAAGACGAACCGGAAAGTTACCGTGAAACAGCTCAAAGAAGACTATTTCAAAAAGATTGAATGTTTCGACTCTGTTGAAATCAGCGGAGTGGAGGATCGCT

TCAACGCATCCCTGGGAACGTATCACGATCTCCTGAAAATCATTAAAGACAAGGACTTCCTGGACAATGAGGAGAACGAGGACATTCTTGAGGACATTGTCCTCACCCTTACGTTGTTTGAAGATAGGGA

GATGATTGAAGAACGCTTGAAAACTTACGCTCATCTCTTCGACGACAAAGTCATGAAACAGCTCAAGAGGCGCCGATATACAGGATGGGGGCGGCTGTCAAGAAAACTGATCAATGGGATCCGAGACAA

GCAGAGTGGAAAGACAATCCTGGATTTTCTTAAGTCCGATGGATTTGCCAACCGGAACTTCATGCAGTTGATCCATGATGACTCTCTCACCTTTAAGGAGGACATCCAGAAAGCACAAGTTTCTGGCCAGG

GGGACAGTCTTCACGAGCACATCGCTAATCTTGCAGGTAGCCCAGCTATCAAAAAGGGAATACTGCAGACCGTTAAGGTCGTGGATGAACTCGTCAAAGTAATGGGAAGGCATAAGCCCGAGAATATCG

TTATCGAGATGGCCCGAGAGAACCAAACTACCCAGAAGGGACAGAAGAACAGTAGGGAAAGGATGAAGAGGATTGAAGAGGGTATAAAAGAACTGGGGTCCCAAATCCTTAAGGAACACCCAGTTGA

AAACACCCAGCTTCAGAATGAGAAGCTCTACCTGTACTACCTGCAGAACGGCAGGGACATGTACGTGGATCAGGAACTGGACATCAATCGGCTCTCCGACTACGACGTGGATCATATCGTGCCCCAGTCT

TTTCTCAAAGATGATTCTATTGATAATAAAGTGTTGACAAGATCCGATAAAAATAGAGGGAAGAGTGATAACGTCCCCTCAGAAGAAGTTGTCAAGAAAATGAAAAATTATTGGCGGCAGCTGCTGAACG

CCAAACTGATCACACAACGGAAGTTCGATAATCTGACTAAGGCTGAACGAGGTGGCCTGTCTGAGTTGGATAAAGCCGGCTTCATCAAAAGGCAGCTTGTTGAGACACGCCAGATCACCAAGCACGTGG

CCCAAATTCTCGATTCACGCATGAACACCAAGTACGATGAAAATGACAAACTGATTCGAGAGGTGAAAGTTATTACTCTGAAGTCTAAGCTGGTCTCAGATTTCAGAAAGGACTTTCAGTTTTATAAGGTG

AGAGAGATCAACAATTACCACCATGCGCATGATGCCTACCTGAATGCAGTGGTAGGCACTGCACTTATCAAAAAATATCCCAAGCTTGAATCTGAATTTGTTTACGGAGACTATAAAGTGTACGATGTTAG

GAAAATGATCGCAAAGTCTGAGCAGGAAATAGGCAAGGCCACCGCTAAGTACTTCTTTTACAGCAATATTATGAATTTTTTCAAGACCGAGATTACACTGGCCAATGGAGAGATTCGGAAGCGACCACTT

ATCGAAACAAACGGAGAAACAGGAGAAATCGTGTGGGACAAGGGTAGGGATTTCGCGACAGTCCGGAAGGTCCTGTCCATGCCGCAGGTGAACATCGTTAAAAAGACCGAAGTACAGACCGGAGGCTT

CTCCAAGGAAAGTATCCTCCCGAAAAGGAACAGCGACAAGCTGATCGCACGCAAAAAAGATTGGGACCCCAAGAAATACGGCGGATTCGATTCTCCTACAGTCGCTTACAGTGTACTGGTTGTGGCCAAA

GTGGAGAAAGGGAAGTCTAAAAAACTCAAAAGCGTCAAGGAACTGCTGGGCATCACAATCATGGAGCGATCAAGCTTCGAAAAAAACCCCATCGACTTTCTCGAGGCGAAAGGATATAAAGAGGTCAA

AAAAGACCTCATCATTAAGCTTCCCAAGTACTCTCTCTTTGAGCTTGAAAACGGCCGGAAACGAATGCTCGCTAGTGCGGGCGAGCTGCAGAAAGGTAACGAGCTGGCACTGCCCTCTAAATACGTTAATT

TCTTGTATCTGGCCAGCCACTATGAAAAGCTCAAAGGGTCTCCCGAAGATAATGAGCAGAAGCAGCTGTTCGTGGAACAACACAAACACTACCTTGATGAGATCATCGAGCAAATAAGCGAATTCTCCAA

AAGAGTGATCCTCGCCGACGCTAACCTCGATAAGGTGCTTTCTGCTTACAATAAGCACAGGGATAAGCCCATCAGGGAGCAGGCAGAAAACATTATCCACTTGTTTACTCTGACCAACTTGGGCGCGCCTG

CAGCCTTCAAGTACTTCGACACCACCATAGACAGAAAGCGGTACACCTCTACAAAGGAGGTCCTGGACGCCACACTGATTCATCAGTCAATTACGGGGCTCTATGAAACAAGAATCGACCTCTCTCAGCTC

GGTGGAGACAGCAGGGCTGACCCCAAGAAGAAGAGGAAGGTGTGAGCGGCCGCGACTCTAGAGGTACCCAGCTTTCTTGTACAAAGTGGTGGCCTGTCAGGCCAAGCTTCCATCGATAGACATGATAA

GATACATTGATGAGTTTGGACAAACCACAACAAGAATGCAGTGAAAAAAATGCTTTATTTGTGAAATTTGTGATGCTATTGCTTTATTTGTAACCATTATAAGCTGCAATAAACAAGTTAACAACAACAATT

GCATTCATTTTATGTTTCAGGTTCAGGGGGAGATGTGGGAGGTTTTTTAAAGCAAGTAAAACCTCTACAAATGTGGTACCTAGGCGAGACCCTGTCTCACAAAATAAAGTAAGCCCGGACTGAGTGCGGA

AAGGCGGGCCTGGCGGGTCTGGTCTCCCCATGCGGGCCACCAGAGGCCCTGCAGCCTTCAGTCGCTTGAAGGGGTAATGGCGCTTCCACTCACAAACATGGCGGACAGAGCGTGTGAACGAGATGAAC

AGCCCCTCAAAAATATGGCCGCCGAGGCTGGACGGCCGTGCCCCAGCAGCACCGCCTCCGCGCCCCACGTGATCTCTCGCCGGGCACAGCGCTGACCGCGGAGGTCCAACCGGAAGAATGTCCGGATTG

GACATTCGGAAGAGGGCCCGCCTTCCCTGGGGAATCTCTGCGCACGCGCAGAACGCTTCGACCAATGAAAACACAGGAAGCCGTCCGCGCAACCGCGTTGCGTCACTTCTGCCGCCCCTGTTTCAAGGTA

TATAGCCGTAGACGGAACTTCGCCTTTCTCTCGGCCTTAGCGCCATTTTTTTGGGTGAGTGTTTTTTGGTTCCTGCGTTGGGATTCCGTGTACAATCCATAGACATCTGACCTCGGCACTTAGCATCATCACA

GCAAACTAACTGTAGCCTTTCTCTCTTTCCCTGTAGAAACCTCTGCACCTGAGGCCACCATGTTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGAGCTGGACGGCGACGTAAA

CGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATGCCACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCCTGGCCCACCCTCGTGACCACCCTGACCTAC

GGCGTGCAGTGCTTCAGCCGCTACCCCGACCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGCGCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCG

CCGAGGTGAAGTTCGAGGGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAACATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATC

ATGGCCGACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGCAGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCTGCTG

CCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGAAGCGCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATGGACGAGCTGTACAAG

GGCAGTGGAGCTACTAACTTCAGCCTGCTGAAGCAGGCTGGTGACGTCgaggagaatcctggcccc

120

xxxxxxxxxxxxxxxxxxxx=

GGTCACCGATCGAGAGCTAG: Empty vector

GTGGAACTGTGGGTACCTCC: LTBP1 KO

CAGGTGGCAGAACTTCCCGG: LTBP2 KO

CGGCGTTCGAGCTCTCAATG: LTBP3 KO

CTGCGATCCTGGGTACCACG: LTBP4 KO

AGAAAACTATAAAGTCCACT: TGFβ2 KO

TTGATGTCACCGGAGTTGTG: TGFβ1 KO

pMuSE SB CRISPR/Cas9 TGFβ1-3KO:

Figure S2: Vector map of the final CRISPR/Cas9 knockout construct for triple gene knockout of TGFβ1-3. The vector carries eGFP (bright green) and puromycin resistance (mint green) under the control of a RPL13A promoter. Both genes are separated by P2A which leads to two separate proteins. A CMV promoter controls the Cas9 (red) and a U6 promoter each sgRNA (orange and purple). These genes are flanked by inverted repeats (ITRs, light blue), the Sleeping Beauty transposase recognition sites.

121

Sequence pMuSE SB CRISPR/Cas9 TGFβ1-3 KO:

atgaccgagtacaagcccacGGTGCGCCTCGCCACCCGCGACGACGTCCCCAGGGCCGTACGCACCCTCGCCGCCGCGTTCGCCGACTACCCCGCCACGCGCCACACCGTCGATCCGGACCGCCACATCGAGCG

GGTCACCGAGCTGCAAGAACTCTTCCTCACGCGCGTCGGGCTCGACATCGGCAAGGTGTGGGTCGCGGACGACGGCGCCGCGGTGGCGGTCTGGACCACGCCGGAGAGCGTCGAAGCGGGGGCGGTG

TTCGCCGAGATCGGCCCGCGCATGGCCGAGTTGAGCGGTTCCCGGCTGGCCGCGCAGCAACAGATGGAAGGCCTCCTGGCGCCGCACCGGCCCAAGGAGCCCGCGTGGTTCCTGGCCACCGTCGGCGTC

TCGCCCGACCACCAGGGCAAGGGTCTGGGCAGCGCCGTCGTGCTCCCCGGAGTGGAGGCGGCCGAGCGCGCCGGGGTGCCCGCCTTCCTGGAGACCTCCGCGCCCCGCAACCTCCCCTTCTACGAGCGG

CTCGGCTTCACCGTCACCGCCGACGTCGAGGTGCCCGAAGGACCGCGCACCTGGTGCATGACCCGCAAGCCCGGTGCCTGATTCGAAGGCCTGTCGTGAAGCTTGGGGATCAATTCTCTAGAGCTCGCTG

ATCAGCCTCGACTGTGCCTTCTAGTTGCCAGCCATCTGTTGTTTGCCCCTCCCCCGTGCCTTCCTTGACCCTGGAAGGTGCCACTCCCACTGTCCTTTCCTAATAAAATGAGGAAATTGCATCGCATTGTCTG

AGTAGGTGTCATTCTATTCTGGGGGGTGGGGTGGGGCAGGACAGCAAGGGGGAGGATTGGGAAGACAATAGCAGGCATGCTGGGGATGCGGTGGGCTCTATGGCTTCTGAGGCGGAAAGAACCAGCT

GGGCCTACTAGCTACTCGGGACCCCTTACCGAAACATCGCCGCATTCTGCAGAGGAGTCGAGTGTATGTAAACTTCTGACCCACTGGGAATGTGATGAAAGAAATAAAAGCTGAAATGAATCATTCTCTCT

ACTATTATTCTGATATTTCACATTCTTAAAATAAAGTGGTGATCCTAACTGACCTAAGACAGGGAATTTTTACTAGGATTAAATGTCAGGAATTGTGAAAAAGTGAGTTTAAATGTATTTGGCTAAGGTGTA

TGTAAACTTCCGACTTCAACTGTATAGGGATCCGCTTCCTCGCTCACTGACTCGCTGCGCTCGGTCGTTCGGCTGCGGCGAGCGGTATCAGCTCACTCAAAGGCGGTAATACGGTTATCCACAGAATCAGG

GGATAACGCAGGAAAGAACATGTGAGCAAAAGGCCAGCAAAAGGCCAGGAACCGTAAAAAGGCCGCGTTGCTGGCGTTTTTCCATAGGCTCCGCCCCCCTGACGAGCATCACAAAAATCGACGCTCAAG

TCAGAGGTGGCGAAACCCGACAGGACTATAAAGATACCAGGCGTTTCCCCCTGGAAGCTCCCTCGTGCGCTCTCCTGTTCCGACCCTGCCGCTTACCGGATACCTGTCCGCCTTTCTCCCTTCGGGAAGCGT

GGCGCTTTCTCATAGCTCACGCTGTAGGTATCTCAGTTCGGTGTAGGTCGTTCGCTCCAAGCTGGGCTGTGTGCACGAACCCCCCGTTCAGCCCGACCGCTGCGCCTTATCCGGTAACTATCGTCTTGAGTC

CAACCCGGTAAGACACGACTTATCGCCACTGGCAGCAGCCACTGGTAACAGGATTAGCAGAGCGAGGTATGTAGGCGGTGCTACAGAGTTCTTGAAGTGGTGGCCTAACTACGGCTACACTAGAAGAAC

AGTATTTGGTATCTGCGCTCTGCTGAAGCCAGTTACCTTCGGAAAAAGAGTTGGTAGCTCTTGATCCGGCAAACAAACCACCGCTGGTAGCGGTGGTTTTTTTGTTTGCAAGCAGCAGATTACGCGCAGAA

AAAAAGGATCTCAAGAAGATCCTTTGATCTTTTCTACGGGGTCTGACGCTCAGTGGAACGAAAACTCACGTTAAGGGATTTTGGTCATGAGATTATCAAAAAGGATCTTCACCTAGATCCTTTTAAATTAAA

AATGAAGTTTTAAATCAATCTAAAGTATATATGAGTAAACTTGGTCTGACAGTTACCAATGCTTAATCAGTGAGGCACCTATCTCAGCGATCTGTCTATTTCGTTCATCCATAGTTGCCTGACTCCCCGTCGT

GTAGATAACTACGATACGGGAGGGCTTACCATCTGGCCCCAGTGCTGCAATGATACCGCGAGACCCACGCTCACCGGCTCCAGATTTATCAGCAATAAACCAGCCAGCCGGAAGGGCCGAGCGCAGAAG

TGGTCCTGCAACTTTATCCGCCTCCATCCAGTCTATTAATTGTTGCCGGGAAGCTAGAGTAAGTAGTTCGCCAGTTAATAGTTTGCGCAACGTTGTTGCCATTGCTACAGGCATCGTGGTGTCACGCTCGTC

GTTTGGTATGGCTTCATTCAGCTCCGGTTCCCAACGATCAAGGCGAGTTACATGATCCCCCATGTTGTGCAAAAAAGCGGTTAGCTCCTTCGGTCCTCCGATCGTTGTCAGAAGTAAGTTGGCCGCAGTGTT

ATCACTCATGGTTATGGCAGCACTGCATAATTCTCTTACTGTCATGCCATCCGTAAGATGCTTTTCTGTGACTGGTGAGTACTCAACCAAGTCATTCTGAGAATAGTGTATGCGGCGACCGAGTTGCTCTTG

CCCGGCGTCAATACGGGATAATACCGCGCCACATAGCAGAACTTTAAAAGTGCTCATCATTGGAAAACGTTCTTCGGGGCGAAAACTCTCAAGGATCTTACCGCTGTTGAGATCCAGTTCGATGTAACCCA

CTCGTGCACCCAACTGATCTTCAGCATCTTTTACTTTCACCAGCGTTTCTGGGTGAGCAAAAACAGGAAGGCAAAATGCCGCAAAAAAGGGAATAAGGGCGACACGGAAATGTTGAATACTCATACTCTTC

CTTTTTCAATATTATTGAAGCATTTATCAGGGTTATTGTCTCATGAGCGGATACATATTTGAATGTATTTAGAAAAATAAACAAATAGGGGTTCCGCGCACATTTCCCCGAAAAGTGCCACCTGATGCGGTG

TGAAATACCGCACAGATGCGTAAGGAGAAAATACCGCATCAGGAAATTGTAAGCGTTAATATTTTGTTAAAATTGGATCCCTATACAGTTGAAGTCGGAAGTTTACATACACTTAAGTTGGAGTCATTAAA

ACTCGTTTTTCAACTACTCCACAAATTTCTTGTTAACAAACAATAGTTTTGGCAAGTCAGTTAGGACATCTACTTTGTGCATGACACAAGTCATTTTTCCAACAATTGTTTACAGACAGATTATTTCACTTATA

ATTCACTGTATCACAATTCCAGTGGGTCAGAAGTTTACATACACTAAGTTCGACTCCTCTGCAGAATGCGGCGATGTTTCGGTAAGGGGTCCGCTATCTAGACACAAGTTTGTACAAAAAAGCAGGCTCTA

TCGATCACGAGACTAGCCTCGAGCGGCCGCCCCCTTCACCGAGGGCCTATTTCCCATGATTCCTTCATATTTGCATATACGATACAAGGCTGTTAGAGAGATAATTGGAATTAATTTGACTGTAAACACAAA

GATATTAGTACAAAATACGTGACGTAGAAAGTAATAATTTCTTGGGTAGTTTGCAGTTTTAAAATTATGTTTTAAAATGGACTATCATATGCTTACCGTAACTTGAAAGTATTTCGATTTCTTGGCTTTATAT

ATCTTGTGGAAAGGACGAAACACCGTTGATGTCACCGGAGTTGTGGTTTTAGAGCTAGAAATAGCAAGTTAAAATAAGGCTAGTCCGTTATCAACTTGAAAAAGTGGCACCGAGTCGGTGCTTTTTTTGA

ATTCGACAACTTTGTATACAAAAGTTGCTATCGATCACGAGACTAGCCTCGAGCGGCCGCCCCCTTCACCGAGGGCCTATTTCCCATGATTCCTTCATATTTGCATATACGATACAAGGCTGTTAGAGAGAT

AATTGGAATTAATTTGACTGTAAACACAAAGATATTAGTACAAAATACGTGACGTAGAAAGTAATAATTTCTTGGGTAGTTTGCAGTTTTAAAATTATGTTTTAAAATGGACTATCATATGCTTACCGTAACT

TGAAAGTATTTCGATTTCTTGGCTTTATATATCTTGTGGAAAGGACGAAACACCGCGACGAAGAGTACTACGCCAGTTTTAGAGCTAGAAATAGCAAGTTAAAATAAGGCTAGTCCGTTATCAACTTGAAA

AAGTGGCACCGAGTCGGTGCTTTTTTTGAATTCGCACCCAACTTTTCTATACAAAGTTGCTATCGATCACGAGACTAGCCTCGAGCGGCCGCCCCCTTCACCGAGGGCCTATTTCCCATGATTCCTTCATATT

TGCATATACGATACAAGGCTGTTAGAGAGATAATTGGAATTAATTTGACTGTAAACACAAAGATATTAGTACAAAATACGTGACGTAGAAAGTAATAATTTCTTGGGTAGTTTGCAGTTTTAAAATTATGTT

TTAAAATGGACTATCATATGCTTACCGTAACTTGAAAGTATTTCGATTTCTTGGCTTTATATATCTTGTGGAAAGGACGAAACACCGATAAATTCGACATGATCCAGGTTTTAGAGCTAGAAATAGCAAGTT

AAAATAAGGCTAGTCCGTTATCAACTTGAAAAAGTGGCACCGAGTCGGTGCTTTTTTTGAATTCGACAACTTTGTATAATAAAGTTGGCGATGTACGGGCCAGATATACGCGTTGACATTGATTATTGACT

AGTTATTAATAGTAATCAATTACGGGGTCATTAGTTCATAGCCCATATATGGAGTTCCGCGTTACATAACTTACGGTAAATGGCCCGCCTGGCTGACCGCCCAACGACCCCCGCCCATTGACGTCAATAATG

ACGTATGTTCCCATAGTAACGCCAATAGGGACTTTCCATTGACGTCAATGGGTGGAGTATTTACGGTAAACTGCCCACTTGGCAGTACATCAAGTGTATCATATGCCAAGTACGCCCCCTATTGACGTCAAT

GACGGTAAATGGCCCGCCTGGCATTATGCCCAGTACATGACCTTATGGGACTTTCCTACTTGGCAGTACATCTACGTATTAGTCATCGCTATTACCATGGTGATGCGGTTTTGGCAGTACATCAATGGGCGT

GGATAGCGGTTTGACTCACGGGGATTTCCAAGTCTCCACCCCATTGACGTCAATGGGAGTTTGTTTTGGCACCAAAATCAACGGGACTTTCCAAAATGTCGTAACAACTCCGCCCCATTGACGCAAATGGG

CGGTAGGCGTGTACGGTGGGAGGTCTATATAAGCAGAGCTCTCTGGCTAACTAGAGAACCCACTGCTTACTGGCTTATCGAAATTAATACGACTCACTATAGGGAGACCCAAGCTTGGTACCGCCACCAT

GGACAAGAAGTACTCCATTGGGCTCGATATCGGCACAAACAGCGTCGGCTGGGCCGTCATTACGGACGAGTACAAGGTGCCGAGCAAAAAATTCAAAGTTCTGGGCAATACCGATCGCCACAGCATAAA

GAAGAACCTCATTGGCGCCCTCCTGTTCGACTCCGGGGAGACGGCCGAAGCCACGCGGCTCAAAAGAACAGCACGGCGCAGATATACCCGCAGAAAGAATCGGATCTGCTACCTGCAGGAGATCTTTAG

TAATGAGATGGCTAAGGTGGATGACTCTTTCTTCCATAGGCTGGAGGAGTCCTTTTTGGTGGAGGAGGATAAAAAGCACGAGCGCCACCCAATCTTTGGCAATATCGTGGACGAGGTGGCGTACCATGA

AAAGTACCCAACCATATATCATCTGAGGAAGAAGCTTGTAGACAGTACTGATAAGGCTGACTTGCGGTTGATCTATCTCGCGCTGGCGCATATGATCAAATTTCGGGGACACTTCCTCATCGAGGGGGAC

CTGAACCCAGACAACAGCGATGTCGACAAACTCTTTATCCAACTGGTTCAGACTTACAATCAGCTTTTCGAAGAGAACCCGATCAACGCATCCGGAGTTGACGCCAAAGCAATCCTGAGCGCTAGGCTGTC

CAAATCCCGGCGGCTCGAAAACCTCATCGCACAGCTCCCTGGGGAGAAGAAGAACGGCCTGTTTGGTAATCTTATCGCCCTGTCACTCGGGCTGACCCCCAACTTTAAATCTAACTTCGACCTGGCCGAAG

ATGCCAAGCTTCAACTGAGCAAAGACACCTACGATGATGATCTCGACAATCTGCTGGCCCAGATCGGCGACCAGTACGCAGACCTTTTTTTGGCGGCAAAGAACCTGTCAGACGCCATTCTGCTGAGTGAT

ATTCTGCGAGTGAACACGGAGATCACCAAAGCTCCGCTGAGCGCTAGTATGATCAAGCGCTATGATGAGCACCACCAAGACTTGACTTTGCTGAAGGCCCTTGTCAGACAGCAACTGCCTGAGAAGTACA

AGGAAATTTTCTTCGATCAGTCTAAAAATGGCTACGCCGGATACATTGACGGCGGAGCAAGCCAGGAGGAATTTTACAAATTTATTAAGCCCATCTTGGAAAAAATGGACGGCACCGAGGAGCTGCTGGT

AAAGCTTAACAGAGAAGATCTGTTGCGCAAACAGCGCACTTTCGACAATGGAAGCATCCCCCACCAGATTCACCTGGGCGAACTGCACGCTATCCTCAGGCGGCAAGAGGATTTCTACCCCTTTTTGAAAG

ATAACAGGGAAAAGATTGAGAAAATCCTCACATTTCGGATACCCTACTATGTAGGCCCCCTCGCCCGGGGAAATTCCAGATTCGCGTGGATGACTCGCAAATCAGAAGAGACCATCACTCCCTGGAACTTC

GAGGAAGTCGTGGATAAGGGGGCCTCTGCCCAGTCCTTCATCGAAAGGATGACTAACTTTGATAAAAATCTGCCTAACGAAAAGGTGCTTCCTAAACACTCTCTGCTGTACGAGTACTTCACAGTTTATAA

CGAGCTCACCAAGGTCAAATACGTCACAGAAGGGATGAGAAAGCCAGCATTCCTGTCTGGAGAGCAGAAGAAAGCTATCGTGGACCTCCTCTTCAAGACGAACCGGAAAGTTACCGTGAAACAGCTCAA

AGAAGACTATTTCAAAAAGATTGAATGTTTCGACTCTGTTGAAATCAGCGGAGTGGAGGATCGCTTCAACGCATCCCTGGGAACGTATCACGATCTCCTGAAAATCATTAAAGACAAGGACTTCCTGGACA

ATGAGGAGAACGAGGACATTCTTGAGGACATTGTCCTCACCCTTACGTTGTTTGAAGATAGGGAGATGATTGAAGAACGCTTGAAAACTTACGCTCATCTCTTCGACGACAAAGTCATGAAACAGCTCAA

GAGGCGCCGATATACAGGATGGGGGCGGCTGTCAAGAAAACTGATCAATGGGATCCGAGACAAGCAGAGTGGAAAGACAATCCTGGATTTTCTTAAGTCCGATGGATTTGCCAACCGGAACTTCATGCA

GTTGATCCATGATGACTCTCTCACCTTTAAGGAGGACATCCAGAAAGCACAAGTTTCTGGCCAGGGGGACAGTCTTCACGAGCACATCGCTAATCTTGCAGGTAGCCCAGCTATCAAAAAGGGAATACTG

CAGACCGTTAAGGTCGTGGATGAACTCGTCAAAGTAATGGGAAGGCATAAGCCCGAGAATATCGTTATCGAGATGGCCCGAGAGAACCAAACTACCCAGAAGGGACAGAAGAACAGTAGGGAAAGGAT

GAAGAGGATTGAAGAGGGTATAAAAGAACTGGGGTCCCAAATCCTTAAGGAACACCCAGTTGAAAACACCCAGCTTCAGAATGAGAAGCTCTACCTGTACTACCTGCAGAACGGCAGGGACATGTACGT

GGATCAGGAACTGGACATCAATCGGCTCTCCGACTACGACGTGGATCATATCGTGCCCCAGTCTTTTCTCAAAGATGATTCTATTGATAATAAAGTGTTGACAAGATCCGATAAAAATAGAGGGAAGAGT

122

GATAACGTCCCCTCAGAAGAAGTTGTCAAGAAAATGAAAAATTATTGGCGGCAGCTGCTGAACGCCAAACTGATCACACAACGGAAGTTCGATAATCTGACTAAGGCTGAACGAGGTGGCCTGTCTGAG

TTGGATAAAGCCGGCTTCATCAAAAGGCAGCTTGTTGAGACACGCCAGATCACCAAGCACGTGGCCCAAATTCTCGATTCACGCATGAACACCAAGTACGATGAAAATGACAAACTGATTCGAGAGGTGA

AAGTTATTACTCTGAAGTCTAAGCTGGTCTCAGATTTCAGAAAGGACTTTCAGTTTTATAAGGTGAGAGAGATCAACAATTACCACCATGCGCATGATGCCTACCTGAATGCAGTGGTAGGCACTGCACTT

ATCAAAAAATATCCCAAGCTTGAATCTGAATTTGTTTACGGAGACTATAAAGTGTACGATGTTAGGAAAATGATCGCAAAGTCTGAGCAGGAAATAGGCAAGGCCACCGCTAAGTACTTCTTTTACAGCAA

TATTATGAATTTTTTCAAGACCGAGATTACACTGGCCAATGGAGAGATTCGGAAGCGACCACTTATCGAAACAAACGGAGAAACAGGAGAAATCGTGTGGGACAAGGGTAGGGATTTCGCGACAGTCCG

GAAGGTCCTGTCCATGCCGCAGGTGAACATCGTTAAAAAGACCGAAGTACAGACCGGAGGCTTCTCCAAGGAAAGTATCCTCCCGAAAAGGAACAGCGACAAGCTGATCGCACGCAAAAAAGATTGGG

ACCCCAAGAAATACGGCGGATTCGATTCTCCTACAGTCGCTTACAGTGTACTGGTTGTGGCCAAAGTGGAGAAAGGGAAGTCTAAAAAACTCAAAAGCGTCAAGGAACTGCTGGGCATCACAATCATGG

AGCGATCAAGCTTCGAAAAAAACCCCATCGACTTTCTCGAGGCGAAAGGATATAAAGAGGTCAAAAAAGACCTCATCATTAAGCTTCCCAAGTACTCTCTCTTTGAGCTTGAAAACGGCCGGAAACGAAT

GCTCGCTAGTGCGGGCGAGCTGCAGAAAGGTAACGAGCTGGCACTGCCCTCTAAATACGTTAATTTCTTGTATCTGGCCAGCCACTATGAAAAGCTCAAAGGGTCTCCCGAAGATAATGAGCAGAAGCAG

CTGTTCGTGGAACAACACAAACACTACCTTGATGAGATCATCGAGCAAATAAGCGAATTCTCCAAAAGAGTGATCCTCGCCGACGCTAACCTCGATAAGGTGCTTTCTGCTTACAATAAGCACAGGGATAA

GCCCATCAGGGAGCAGGCAGAAAACATTATCCACTTGTTTACTCTGACCAACTTGGGCGCGCCTGCAGCCTTCAAGTACTTCGACACCACCATAGACAGAAAGCGGTACACCTCTACAAAGGAGGTCCTG

GACGCCACACTGATTCATCAGTCAATTACGGGGCTCTATGAAACAAGAATCGACCTCTCTCAGCTCGGTGGAGACAGCAGGGCTGACCCCAAGAAGAAGAGGAAGGTGTGAGCGGCCGCGACTCTAGAG

GTACCCAGCTTTCTTGTACAAAGTGGTGGCCTGTCAGGCCAAGCTTCCATCGATAGACATGATAAGATACATTGATGAGTTTGGACAAACCACAACAAGAATGCAGTGAAAAAAATGCTTTATTTGTGAAA

TTTGTGATGCTATTGCTTTATTTGTAACCATTATAAGCTGCAATAAACAAGTTAACAACAACAATTGCATTCATTTTATGTTTCAGGTTCAGGGGGAGATGTGGGAGGTTTTTTAAAGCAAGTAAAACCTCTA

CAAATGTGGTACCTAGGCGAGACCCTGTCTCACAAAATAAAGTAAGCCCGGACTGAGTGCGGAAAGGCGGGCCTGGCGGGTCTGGTCTCCCCATGCGGGCCACCAGAGGCCCTGCAGCCTTCAGTCGCT

TGAAGGGGTAATGGCGCTTCCACTCACAAACATGGCGGACAGAGCGTGTGAACGAGATGAACAGCCCCTCAAAAATATGGCCGCCGAGGCTGGACGGCCGTGCCCCAGCAGCACCGCCTCCGCGCCCCA

CGTGATCTCTCGCCGGGCACAGCGCTGACCGCGGAGGTCCAACCGGAAGAATGTCCGGATTGGACATTCGGAAGAGGGCCCGCCTTCCCTGGGGAATCTCTGCGCACGCGCAGAACGCTTCGACCAATG

AAAACACAGGAAGCCGTCCGCGCAACCGCGTTGCGTCACTTCTGCCGCCCCTGTTTCAAGGTATATAGCCGTAGACGGAACTTCGCCTTTCTCTCGGCCTTAGCGCCATTTTTTTGGGTGAGTGTTTTTTGG

TTCCTGCGTTGGGATTCCGTGTACAATCCATAGACATCTGACCTCGGCACTTAGCATCATCACAGCAAACTAACTGTAGCCTTTCTCTCTTTCCCTGTAGAAACCTCTGCACCTGAGGCCACCATGTTGAGCA

AGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGAGCTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATGCCACCTACGGCAAGCTGACCCTGAA

GTTCATCTGCACCACCGGCAAGCTGCCCGTGCCCTGGCCCACCCTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGACCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGC

CCGAAGGCTACGTCCAGGAGCGCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAGGGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAG

GAGGACGGCAACATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCGACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGG

CAGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCTGCTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGAAGCGCGATCAC

ATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATGGACGAGCTGTACAAGGGCAGTGGAGCTACTAACTTCAGCCTGCTGAAGCAGGCTGGTGACGTCgaggagaatcctggcccc

123

Appendix II

Figure S3: TGFβ1 treatment inhibits the adipogenic differentiation of SGBS cells. SGBS preadipocytes were treated with different concentrations of recombinant TGFβ1 (r TGFβ1) during adipogenesis. (A) Pictures were taken on day 14 and (B) protein expression of Adiponectin was assessed by Western Blot. Tubulin served as a reference, n=1.

124

Figure S4: TGFβ1 deficiency does not alter UCP1 expression or metabolic function. TGFβ1 deficiency was achieved by CRISPR/Cas9 in SGBS preadipocytes, which were subjected to adipogenic differentiation for 14 days. (A) Representative Western blot of EV and TGFβ1-deficient SGBS preadipocytes of n=3 independent experiments with GAPDH as loading control. (B) Microscopic pictures of EV and TGFβ1-deficient SGBS adipocytes, 10-fold magnification. (C) mRNA expression of differentiation marker PPARG, GLUT4, Adiponectin and (D) UCP1 was determined by q-RT-PCR. HPRT serves as a reference gene. Mean +SEM of n=4 (E) Representative Western blot is shown of EV and TGFβ1-deficient SGBS adipocytes and densitometric analysis was performed in n=3 independent experiments, GAPDH serves as a loading control. (F) OCR over time of n=4 individual experiments. To calculate basal and maximum respiration, as well as ATP production, proton leak and cAMP response, cells were treated with dibutyryl-cAMP (cAMP), oligomycin (Oligo), carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP) and antimycin/rotenone (Ant/Rot). Mean +SEM of n=4

125

Table S1: p-values of correlated mRNA expression in human WAT. UCP1 and LTBP mRNA expression was correlated with the TGFβ ligands and receptors mRNA expression in human mamma subcutaneous adipose tissue of n=28 patients which underwent plastic surgery. Mean age±SD = 45±15 years, Mean BMI±SD = 28.2±4.9 kg/m². A linear regression was applied and p<0.05 is considered as statistically significant (marked in bolt).

LTBP1 LTBP2 LTBP3 LTBP4 UCP1

UCP1 p=0.7582 p=0.2710 p=0.0382 p=0.1573 -

TGFβ1 p=0.2355 p=0.0258 p=0.3372 p=0.8546 p =0.3361

TGFβ2 p=0.2585 p=0.1180 p=0.0100 p=0.1618 p<0.0001

TGFβ3 p=0.1968 p=0.1585 p=0.2336 p=0.4400 p=0.3786

TGFβR1 p=0.654 p=0.0197 p<0.0001 p=0.2762 p<0.0001

TGFβR2 p=0.0884 p=0.7880 p=0.0015 p=0.0871 p=0.3467

TGFβR3 p=0.3178 p=0.0266 p<0.0001 p=0.0149 p=0.3668

126

Acknowledgements

The acknowledgements were removed for data privacy protection reasons.

127

Statutory declaration

I hereby declare that I wrote the present dissertation with the topic:

“Role of secreted factors in white and brown adipogenesis”

independently and used no other aids that those cited. In each individual case, I have clearly

identified the source of the passages that are taken word for word or paraphrased from

other works.

I also hereby declare that I have carried out my scientific work according to the principles

of good scientific practice in accordance with the current “ Satzung der Universität Ulm zur

Sicherung guter wissenschaftlicher Praxis” [Rules of the University of Ulm for Assuring

Good Scientific Practice]

Ulm, 25.11.2020

Daniel Halbgebauer

128

Curriculum vitae

Personal information

Name: Daniel Halbgebauer

Year of Birth: 1990

Dissertation

06/2017-12/2020 PhD student at the University Medical Center Ulm

➢ Department of Pediatrics and Adolescent Medicine

➢ Experimental endocrinology and metabolism research

➢ Title: Role of secreted factors for white and brown adipocyte differentiation

Education

10/2014-04/2017 Master of science Biochemistry University of Ulm

➢ Thesis at the department of peptide pharmaceuticals ➢ Title: Development of a biomedical bilayer hydrogel for

catching and killing Pseudomonas aeruginosa

09/2015-05/2016 2 Semesters at the University of Massachusetts-Amherst-USA,

➢ Fellowship of the Baden Württemberg exchange program

10/2011-09/2014 Bachelor of science Biochemistry University of Ulm

➢ Bachelor thesis at the internal medicine II, section for sports and rehabilitation medicine

➢ Title: The effect of estradiol and testosterone on the irisin signaling pathway.

09/2010 - 05/2011 Civil service Paulinenpflege Winnenden

09/2001 - 07/2010 Abitur Lessing-Gymnasium Winnenden

129

Conferences

Annual meeting of the German Obesity Association 2018, Wiesbaden, Germany Poster presentation: Investigation of the browning capacity in human cell models during adipogenesis Daniel Halbgebauer, Martin Wabitsch, Meike Dahlhaus, Pamela Fischer-Posovszky, Daniel Tews Poster presentation: Investigation of the functional role of Matrix Gla Protein in adipocytes Daniel Halbgebauer, Jan-Bernd Funcke, Heike Neubauer, Bradford Hamilton, Martin Wabitsch, Pamela Fischer-Posovszky, Daniel Tews

Publications:

Halbgebauer D, Dahlhaus M, Wabitsch M, Fischer-Posovszky P, Tews D. Browning capabilities of human primary adipose-derived stromal cells compared to SGBS cells. Sci Rep. 2020;10(1):1-8. doi:10.1038/s41598-020-64369-7

Reisser T, Halbgebauer D, Scheurer J, Wolf L, Leithäuser F, Beyersdorf N, Fischer-Posovszky P, Debatin KM, Strauss G. In vitro-generated alloantigen-specific Th9 cells mediate antileukemia cytotoxicity in the absence of graft-versus-host disease. Leukemia. February 2020:1-6. doi:10.1038/s41375-020-0731-2

Dahlhaus M, Roos J, Engel D, Tews D, Halbgebauer D, Funcke JB, Kiener S, Schuler PJ, Döscher J, Hoffmann TK, Zinngrebe J, Rojewski M, Schrezenmeier H, Debatin KM, Wabitsch M, Fischer-Posovszky P. CD90 is dispensable for white and beige/brown adipocyte differentiation. Int J Mol Sci. 2020;21(21):1-19. doi:10.3390/ijms21217907

Roos J, Dahlhaus M, Funcke JB, Kustermann M, Strauss G, Halbgebauer D, Boldrin E, Holzmann K, Möller P, Trojanowski BM, Baumann B, Michael K, Martin D, Posovszky PF. miR-146a regulates insulin sensitivity via NPR3. Cell Mol Life Sci. 2020;(0123456789). doi:10.1007/s00018-020-03699-1

Bodenberger N, Kubiczek D, Halbgebauer D, Rimola V, Wiese S, Mayer D, Rodriguez Alfonso AA, Ständker L, Stenger S, Rosenau F. Lectin-Functionalized Composite Hydrogels for “capture-and-Killing” of Carbapenem-Resistant Pseudomonas aeruginosa. Biomacromolecules. 2018;19(7):2472-2482. doi:10.1021/acs.biomac.8b00089

Documents

The role of secreted factors in white and brown adipogenesis