1. health and fitness
2. disease

A ROLE OF VISCERAL ADIPOSE AND GASTRIC TISSUE IN INFLAMMATORY
CONDITIONS ASSOCIATED WITH METABOLIC SYNDROME
by
Rohini Mehta
A Dissertation
Submitted to the
Graduate Faculty
of
George Mason University
in Partial Fulfillment of
The Requirements for the Degree
of
Doctor of Philosophy
Biosciences
Committee:
_________________________________________
Dr. Ancha Baranova,
Dissertation Director
_________________________________________
Dr. Alan Christensen,
Committee Member
_________________________________________
Dr. Patrick Gillevet,
Committee Member
_________________________________________
Dr. Soren Mogelsvang,
Committee Member
_________________________________________
Dr. Jim Willet, Department
Chairperson
_________________________________________
Dr. Timothy L. Born,
Associate Dean for Student
and Academic Affairs,
College of Science
_________________________________________
Vikas Chandhoke, Dean,
College of Science
Date: __________________________________
Spring Semester 2013
George Mason University
Fairfax, VA
A Role of Visceral Adipose and Gastric Tissue in Inflammatory Conditions Associated
With Metabolic Syndrome
A Dissertation submitted in partial fulfillment of the requirements for the degree of
Doctor of Philosophy at George Mason University
By
Rohini Mehta
Master of Science
Maharaja Sayajirao University, 2005
Director: Dr. Ancha Baranova, Associate Professor
School of Systems Biology
Spring Semester 2013
George Mason University
Fairfax, VA
Copyright: 2013 Rohini Mehta
All Rights Reserved
ii
DEDICATION
I dedicate the dissertation to my husband Deapesh, who was always there for me with his
unconditional support and patience. I also dedicate this dissertation to my parents and
sister.
iii
ACKNOWLEDGEMENTS
I wish to thank my committee members who were more than generous with their
expertise and precious time. A special thanks to Dr. Ancha Baranova, my committee
chair for her countless hours of reflecting, reading, encouraging, and most of all patience
throughout the entire process. Thank you Dr. Alan Christensen, Dr. Patrick Gillevet, Dr.
Soren Mogelsvang for your helpful insights.
I would like to acknowledge and thank Dr. Zobair Younossi, for allowing me to
conduct my research at Inova Fairfax Hospital and his confidence in my abilities. Special
thanks goes to Dr. Aybike Birerdinc and Dr. Michael Estep for always being there and
motivating me. Finally I would like to thank Ms. Diane St Germain whose patience and
promptness made the completion of this research an enjoyable experience.
iv
TABLE OF CONTENTS
Page
List of Tables…………………………………………………………………………….vii
List of Figures………………………………………………………………….................ix
List of Abbreviations/Symbols……………………………………………......................xii
Abstract................................................................................................................……….xiii
Chapter 1: Introduction ............................................................................................. …1
A: Obesity..................................................................................................... …2
B: Adipose Tissue .................................................................................... …...16
C: Gastric Tissue ..................................................................................... ……61
D: Liver…………………………………………………………………………76
Chapter 2: Specific Aims………………………………………………………………..81
Chapter 3: Reference Gene Validation………… ......................................................... 84
A: Reference Gene Validation in VAT………………………………………....84
v
B: Reference Gene Validation in Stomach…………………………………….115
C: Reference Gene Validation in HepG2……………………………………...136
Chapter 4: Gene Expression of Brown and White Adipose Tissue…………………….176
Chapter 5: Gene Expression in Stomach……………………………………………….239
A : Inflammatory Panel ………………………………………………………..239
B : Obesity Panel ……………………………………………………………....281
Appendices …………….………………………………………………………………322
Appendix Table 1 ……………………………………………………………....322
Appendix Table 2 ……………………………………………………………....337
List of References.……………………………………………………………………..354
vi
LIST OF TABLES
Table
Page
Table 1: Comparison of highly ranked gene lists identified by all three algorithms (n=9)
........................................................................................................................... ……98
Table 2: Comparison of highly ranked gene lists identified by all three algorithms (n=19)
................................................................................................................................... 99
Table 3: BestKeeper correlation analysis (n=9) ......................................................... 105
Table 4: BestKeeper correlation analysis (n=19) ....................................................... 106
Table 5: Comparison of highly ranked genes by all three software (n=23) ................ 125
Table 6: BestKeeper correlation analysis (n=23) ....................................................... 129
Table 7: Comparison of reference gene stability ranked based on intra- and inter-group
variation by NormFinder .......................................................................................... 145
Table 8: Primer sequences of reference genes ........................................................... 161
Table 9: A list of gene profiled in VAT of morbidly obese and non-obese subjects and
their roles in energy homeostasis .............................................................................. 161
Table 10: Demographic and clinical data .................................................................. 208
Table 11: Clinical and Gene Expression Data ........................................................... 218
vii
Table 12: Primer sequences of BAT specific genes ................................................... 234
Table 13: Demographic and clinical characteristics ................................................... 244
Table 14: List of inflammatory genes significantly upregulated in gastric tissues of
patients with the following pathological conditions ................................................... 249
Table 15: Correlations between inflammatory gene expression with the spectrum of
NAFLD .................................................................................................................... 254
Table 16: Multiple regression analysis ...................................................................... 264
Table 17: Clinical and Demographic Data................................................................. 289
Table 18: List of obesity related genes significantly upregulated in gastric tissues of
patients with the following pathological conditions ................................................... 264
Table 19: Correlations between obesity gene expression and spectrum of NAFLD ... 298
viii
LIST OF FIGURES
Figure
Page
Figure 1:Obesity and Inflammation........................................................................... …9
Figure 2:NAFLD Spectrum ........................................................................................ 12
Figure 3:Energy Homeostasis ..................................................................................... 14
Figure 4:Adipose Tissue Inflammation ....................................................................... 20
Figure 5:Lipogenesis .................................................................................................. 26
Figure 6:Fat Homeostasis ........................................................................................... 29
Figure 7:Glycolysis-Lipogenesis ................................................................................. 34
Figure 8:Major AT depots. ......................................................................................... 38
Figure 9:Electron Transport Chain (ETC) ................................................................... 51
Figure 10:UCP1 mediated thermogenesis ................................................................... 53
Figure 11:Location of BAT in Adults and Infants ....................................................... 57
Figure 12:Location of BAT in Adults and Infants ....................................................... 57
ix
Figure 13:Regions within the Stomach and Gastric Tissue .......................................... 66
Figure 14:Gastric Glands Structure ............................................................................. 68
Figure 15:Energy Homeostasis ................................................................................... 75
Figure 16:Gene expression stability M of candidate reference genes in visceral adipose
tissue as calculated by GeNorm software .................................................................... 94
Figure 17:Determination of optimal number of reference genes for normalization by
pairwise variation analysis using GeNorm software .................................................... 96
Figure 18:Determination of the most stable reference genes using NormFinder ........ 102
Figure 19:Gene expression stability M of candidate reference genes in gastric tissue
calculated by GeNorm software ................................................................................ 122
Figure 20:Determination of optimal number of reference genes for normalization by
pairwise variation analysis using GeNorm software .................................................. 123
Figure 21:Determination of the most stable reference genes using NormFinder ........ 127
Figure 22:Schematic representation of Fas mediated cellular apoptosis and role of antiapoptotic proteins ..................................................................................................... 141
Figure 23:Determination of the most stable reference genes using NormFinder ........ 149
Figure 24:Key regulators of the BAT functioning and thermogenesis ....................... 181
Figure 25:Regulation of UCP1 gene expression and activity by nor-epinephrine and free
fatty acids ................................................................................................................. 184
Figure 26:Tissue specific role of PPARD in thermogenesis and energy homeostasis . 186
x
Figure 27:PGC1A and regulation of UCP1 expression .............................................. 189
Figure 28:Role of PGC1A in BAT and energy homeostasis ...................................... 192
Figure 29:Role of SIRT2 in regulation of thermogenic gene expression .................... 196
Figure 30:Gene expression profile in visceral AT of obese subjects compared to lean
subjects..................................................................................................................... 219
Figure 31:Potential role of nuclear factors in thermogenesis and lipolysis in omental
adipose tissue as it relates to morbid obesity ............................................................. 222
Figure 32:Venn diagram depicting overlap in results of group comparison and correlation
analysis in different cohorts ..................................................................................... 269
Figure 33:Venn diagram depicting results of the statistical analysis .......................... 271
Figure 34:Inflammation-related genes in stomach and obesity associated Non-Alcoholic
Fatty Liver Disease (NAFLD) .................................................................................. 274
Figure 34:Role of CPE in processing of propeptides involved in energy homeostasis 305
Figure 35:Differentially expressed genes in obese patients with NAFLD .................. 312
xi
LIST OF ABBREVIATIONS
WHO
BMI
AT
NAFLD
TNF-α
IL-6
iNOS
TGF-β
CRP
ICAM
MCP-1
IR
HDL
MHO
MS
HS
TGA
T2D
FFA
MMPs
FAS
NEFA
VLDL
LDL
ATP
ACC1
SAT
VAT
WAT
BAT
ApoE
CETP
18
F-FDG
NAD
NADH
World Health Organization
Body Mass Index
Adipose Tissue
Non-alcoholic fatty liver disease
Tumor necrosis factor-alpha
Interleukin-6
Inducible nitric oxide synthase
Transforming growth factor-beta
C-reactive protein
Intercellular adhesion molecule
Monocyte chemotactic protein-1
Insulin resistance
High-density lipoprotein
Metabolically healthy obese
Metabolic Syndrome
Fatty liver or hepatic steatosis
Triacylglycerols
Type 2 diabetes
Free fatty acids
Matrix metalloproteases
Fatty acid synthases
Non-esterified fatty acids
Very Low density lipoproteins
Low density lipoproteins
Adenosine triphosphate
Acetyl-CoA carboxylase 1
Subcutaneous Adipose Tissue
Visceral Adipose Tissue
White Adipose Tissue
Brown Adipose Tissue
Apolipoprotein E
Cholesteryl ester transfer protein
Fludeoxyglucose18F
Nicotinamide adenine dinucleotide:
Nicotinamide adenine dinucleotide hydride
xii
ETC
UCP1
CIDEA
Rb
PRDM16
PGC-1α
CtBP-1
CtBP-2
PGC-1β
GHS-R1a
UAG
NPY
VTA
ACTB
GAPDH
B2M
RPII
YWHAZ
UBC
HPRT1
β-ME
Electron Transport Chain
Uncoupling protein-1
Cell-death inducing DFF45-like effector A
Retinoblastoma protein
PRD1-BF1-RIZ1 homologous domain-containing 16
PPARG coactivator 1alpha
C-terminal-binding protein-1
C-terminal-binding protein-1
PPARA-coactivator-1B
Growth hormone secretagogue receptor type 1a
Unacylated ghrelin
Neuropeptide Y
Ventral tegmental area
β-Actin
Glyceraldehyde-3-phosphate dehydrogenase
Beta-2-microglobulin
RNA polymerase II
Tyrosine-3 monooxygenase/Tryptophan-5 monooxygenase activation
protein, zeta polypeptide:
Ubiquitin C
Hypoxanthine phosphoribosyl transferase 1
Beta-mercaptoethanol
xiii
ABSTRACT
A ROLE OF THE VISCERAL ADIPOSE AND THE GASTRIC TISSUE IN
INFLAMMATORY CONDITIONS ASSOCIATED WITH METABOLIC SYNDROME
Rohini Mehta, PhD
George Mason University, 2013
Dissertation Director: Dr. Ancha Baranova
Obesity has reached epidemic proportions globally. Obesity is accompanied by comorbidities affecting multiple peripheral tissues. Obesity associated non-alcoholic fatty
liver disease (NAFLD) is the leading cause of chronic liver disease in the US. The
molecular mechanisms underlying the development and the progression of NAFLD,
however, remain poorly understood. Stomach is an important endocrine organ that
produces a number of bioactive peptides with important roles in the metabolism of
energy. Stomach is located in the vicinity of the liver, but so far it has been largely
neglected as an organ with potentially important role in obesity and associated NAFLD.
This study provides the evidence for the contribution of the stomach-specific expression
changes in a number of molecules previously implicated in inflammation and energy
homeostasis
to
the
pathogenesis
of
obesity-associated
NAFLD.
White adipose tissue (WAT) represents the majority of adipose tissue in humans, is the
main fat storage organ. In contrast, brown adipose tissue (BAT) has a unique ability to
spend energy through producing heat in mitochondrial “uncoupling”. The facultative
nature of BAT activity and its distribution within WAT depots complicate the detection
of BAT in adult humans. In this study, the expression of genes involved in transcriptional
regulation of brown adipocyte-specific UCP1 gene and BAT differentiation was assessed
as quantitative indicator of BAT activity in adipose of adult humans.
CHAPTER 1: Introduction
Obesity:
Obesity is the abnormal or excessive increase in adiposity. It is
currently defined by World Health Organization (WHO) as Body Mass
Index (BMI) of > 30kg/m2. In the 21st century, it has reached epidemic
proportions in various parts of the Western World (Alkhouri et al., 2009)
and it continues to increase globally (Hossain et al., 2007), (MacLaren et al.,
2010). In developing countries, the prevalence of obesity is increasing in the
younger age groups (Hossain et al., 2007).
Obesity can be explained from a biochemical as well as a
thermodynamic standpoint. From a biochemical perspective, the disruption
in the homeostasis between food intake and energy expenditure results in the
subsequent deposition of excess fatty acids in the adipocytes in the form of
triglycerides (TGA) (Kypreos et al., 2009). From a thermodynamic point of
view, obesity is the result of a positive imbalance between food intake and
1
energy expenditure (Sethi and Vidal-Puig, 2007). Obesity, alters metabolic
and endocrine function of adipose tissue (AT) and leads to an increased
release of free fatty acids (FFA), hormones, and pro-inflammatory molecules
that contribute to obesity associated complications (Arner P, 2001),
(Weisberg et al., 2003). Importantly, obesity is accompanied by changes in
not only the AT, but also in non-adipose tissues, for example, in the liver
and the muscles (Coppack S W, 2005).
Adiposity, the fraction of total body mass comprised of lipids stored
in the AT, is closely correlated with important physiological parameters such
as blood pressure, systemic insulin sensitivity, serum TGA and leptin
concentrations (Weisberg et al., 2003). A strong positive correlation between
the degree of adiposity and several obesity-associated disorders such as
hypertension, dyslipidemia, and glucose intolerance is well known
(Weisberg et al., 2003). In particular, among the obesity associated
conditions is a spectrum of liver abnormalities known as non-alcoholic fatty
liver disease (NAFLD), which is characterized by increased intrahepatic
triglyceride content (steatosis) often accompanied by inflammation,
2
ballooning degeneration and/or Mallory-Denk bodies and fibrosis (nonalcoholic steatohepatitis) (Fabbrini et al., 2010).
Obesity has received attention as the state of low grade chronic
inflammation characterized by both macrophage infiltration of the AT and
increased production of pro-inflammatory cytokines (Arner P, 2001),
(Weisberg et al., 2003), (Alkhouri et al., 2009), (Hajer et al., 2008).
Compared with the AT of lean individuals, increased amounts of proinflammatory proteins such as Tumor necrosis factor-alpha (TNF-α),
Interleukin-6 (IL-6), Inducible nitric oxide synthase (iNOS), Transforming
growth factor-beta (TGF-β), C-reactive protein (CRP), soluble intercellular
adhesion molecule (ICAM), and monocyte chemotactic protein-1 (MCP-1),
and pro-coagulant proteins such as plasminogen activator inhibitor type-1
(PAI-1), tissue factor, and factor VII have been reported in the AT of obese
individuals (Weisberg et al., 2003). The increased pro-inflammatory
cytokine levels in circulation play a role in insulin resistance (IR); this
notion is supported by the observation that the prevention of macrophage
accumulation in the AT protects against inflammation and the development
of IR (Alkhouri et al., 2009).
3
However, there is still a debate concerning the pathogenic
mechanisms that result in macrophage recruitment (Alkhouri et al., 2009).
Increase in hypoxia in the expanded AT is believed to result in the
production and the release of certain chemokines (Alkhouri et al., 2009).
Furthermore, adipocyte death by necrosis results in a crown-like structure
formation, which also attracts macrophages. However, whether these events
are a consequence of inflammation or they precede it remains unclear
(Alkhouri et al., 2009).
Obese subjects also differ in their degree of IR, levels of adipokine
production and presence of associated co-morbities (MacLaren et al., 2010).
It is obvious that obesity is an important risk factor for IR, but many obese
individuals remain insulin sensitive (Despres and Lemieux, 2006), (Koska et
al., 2008), (Younossi et al., 2008). Interestingly, a subgroup of obese
individuals have been found with normal triglyceride, glucose, and highsensitivity CRP levels, decreased high-density lipoprotein (HDL) cholesterol
levels, insulin resistance, and hypertension and are classified as
metabolically healthy obese (MHO) (Sims, 2001). Adding to this
complexity, not all forms of obesity are genetically or etiologically similar.
4
For instance, visceral obesity presents a greater risk for obesity-related
disorders than subcutaneous obesity (Weisberg et al., 2003), (Koska et al.,
2008).
i.
Obesity And Inflammation:
Obesity is a state of chronic low grade inflammation (Trayhurn,
2005), (Stienstra et al., 2007). This concept may be illustrated by the
increased plasma levels of several pro-inflammatory markers including
cytokines and acute phase proteins like CRP in obese individuals. Many of
the inflammatory markers found in plasma of obese individuals appear to
originate from the AT (Baranova et al., 2007), (Estep et al., 2009).
For many years, human AT was considered as a passive fat reservoir
(Mora and Pessin, 2002). However, less than 10 years ago it was shown to
actively alter homeostasis of energy and other aspects of body functioning
(Kershaw and Flier, 2004). AT releases over 50 bioactive peptides
collectively called adipokines (Ahima and Flier, 2000). These molecules are
shown to be involved in the modulation of the fat mass (Greenberg and
5
Obin, 2006), long term appetite regulation (Stanley et al., 2005), immune
response, the blood pressure and bone mass control, as well as thyroid and
reproductive function (Trayhurn, 2005). Inflammation is a phenomenon
originating largely from the fat compartment (Baranova et al., 2007), (Estep
et al., 2009). AT derived adipokines have a systemic influence (Faintuch et
al., 2007) as mediators of inflammation and nutrient status signaling
components (Lee et al., 2009) such as glucose homeostasis and lipogenesis.
Stomach is a central organ involved in the process of digestion. Thus,
its role in the nutrient status signaling would not be surprising (Weigt and
Malfertheiner, 2009). However, studies of obesity focusing on the role of
stomach are limited. The stomach is shown to be an important source of
paracrine peptides such as leptin (Bado et al., 1998), (Baratta, 2002), ghrelin
(Badman and Flier, 2005) and obestatin (Sakata and Sakai, 2010). The latter
two gastric peptides along with AT derived leptin have been shown to play a
critical role in regulating appetite (Murphy and Bloom, 2006), (Sakata and
Sakai, 2010).
6
One of the tissues that might be directly affected by the proinflammatory secretion profile of AT is the liver. Liver is also located in the
proximity of the stomach via portal circulation. The molecular mechanisms
underlying the development of steatosis and progression to steatohepatitis in
NAFLD remains poorly understood. Lipid peroxidation, cytokines, and other
pro-inflammatory compounds are all believed to play a vital role in the
transition (Stienstra et al., 2007). The close interplay of AT and stomach in
regulating appetite and thus energy homeostasis suggests the likelihood of
the influence of AT derived inflammatory molecules on stomach (Figure1).
It is likely that the gastric mucosa is one of the target organs of AT proinflammatory cytokines. Conversely, cytokines of stomach may have both
systemic and, probably, liver-specific impacts, which may increase the
susceptibility of obese individuals to NAFLD and metabolic syndrome
(MS). Thus, gastric mucosa of obese individuals needs to be investigated.
7
Figure 1: Obesity and Inflammation. The close interplay of the adipose
and gastric tissue in energy homeostasis suggests the likelihood of the
influence of AT derived inflammatory molecules on gastric tissue.
Conversely, cytokines of gastric tissue may have both systemic and,
probably, liver-specific impact, which may increase the susceptibility of
obese individuals to NAFLD.
8
ii.
Obesity and NAFLD:
NAFLD is a common disease of this modern time. It comprises a
morphological spectrum of liver pathologies ranging from simple TGA
accumulation in hepatocytes (fatty liver or hepatic steatosis; HS) to
inflammatory and hepatocellular ballooning injury (NASH), eventually
leading to fibrosis and cirrhosis (Figure 2). NAFLD has burst out onto the
clinical landscape over the past 25 years and since then has got recognition
as the most common cause of liver enzyme levels abnormalities in adults
(Marchesini et al., 2003), (Qureshi and Abrams, 2007). The prevalence of
NAFLD in the general US population is approximately 20% (Vernon et al.,
2011). Similar data has been obtained in the Japanese and Italian populations
(Qureshi and Abrams, 2007). Although a number of medical conditions have
9
been identified risk factors for development of NAFLD, obesity is by far the
most dominant factor (Haynes et al., 2004).
The pathophysiology of NAFLD is complex. The available data
suggests that environmental factors such as diet, exercise, and/or toxins are
likely to be important in its causation and progression. Although in most
cases fatty liver does not progress to a more severe liver disease,
approximately 20% to 30% of patients have histological signs of fibrosis and
necroinflammation, indicating the presence of NASH (Vernon et al., 2011).
These patients are at higher risk of developing cirrhosis, terminal liver
failure, and hepatocellular carcinoma (Marchesini et al., 2003). Lipid
peroxidation, cytokines, and other pro-inflammatory compounds produced
by the AT are believed to play a vital role in the transition to NASH
(Stienstra et al., 2007).
The great majority of NAFLD occurs in the setting of MS, in which
IR plays a pivotal role (Marchesini et al., 2003).
10
Figure 2: NAFLD spectrum. Liver lesions ranging from simple triglyceride
accumulation in hepatocytes (fatty liver or hepatic steatosis; HS) to
inflammatory
and
hepatocellular
ballooning
injury
(non-alcoholic
steatohepatitis; NASH), eventually leading to fibrosis and cirrhosis.
11
Clearly, obesity is a complex disease. It involves both genetic and
environmental perturbations to a complex array of networks connecting
peripheral tissues such as adipose, gut, muscle, intestine, liver, and pancreas
tissues with the hypothalamus (Figure 3). Consequently, the resulting energy
imbalance has a systemic effect (Dobrin et al., 2009). Importantly, after
several years of research on AT, the molecular mechanisms leading to or
underlying obesity and its associated disorders are still ambiguous. Thus, it
is essential to explore the role of additional peripheral tissues such as
stomach in the pathogenesis of obesity and associated co-morbidities.
12
Figure 3: Energy Homeostasis. A complex interplay of multiple tissues and
hormones ensures homeostasis between energy intake, expenditure and
storage (Image Source: Ahima, 2008).
13
iii.
Obesity And Metabolic Syndrome:
Obesity is a dominant underlying factor for MS. MS is recognized as
a constellation of metabolic abnormalities that serves as a major indicator for
cardiovascular diseases (CVDs) (Despres and Lemieux, 2006). The five
screening variables used to identify those with MS are high waist
circumference, high circulating levels of TGA and low circulating levels of
HDL-cholesterol, high fasting glycemia and blood pressure (Despres and
Lemieux, 2006). The presence of any three of these factors amounts to
definite MS diagnosis, which paves the path to IR and Type 2 diabetes
(T2D).
There is ample evidence indicating that impaired FFA metabolism in
the visceral AT could contribute to the insulin-resistant state observed
among individuals with visceral obesity (Despres and Lemieux, 2006).
However, another mechanistic possibility is failure of subcutaneous AT to
14
act as an ‘energy sink’. This would lead to resistance to insulin when an
individual has to handle a calorie surplus due to excess energy intake and/or
reduced energy expenditure (Despres and Lemieux, 2006). According to this
compelling hypothesis, the relative deficits in the capacity of subcutaneous
fat to store excess energy would result in increased accumulation of fat at
undesired sites such as the liver, the skeletal muscle, the heart and even
pancreatic β-cells; a phenomenon that has been described as ectopic fat
deposition (Despres and Lemieux, 2006). This ectopic accumulation of lipid
in liver, muscle and pancreatic β -cells is apparently associated with IR, MS
and T2D (Ferré and Foufelle, 2007). The support for this theory comes from
the fact that essentially fatless transgenic mice show liver and muscle IR and
eventually develop diabetes (Gavrilova et al., 2000), (Kim et al., 2000).
Surgical implantation of AT in these mice improves the insulin sensitivity of
their liver and muscles, consistent with the idea that subcutaneous fat is a
metabolic sink to buffer an energy surplus. In humans, the severe insulinresistant states found in patients with lipodystrophic conditions are also
consistent with the role of subcutaneous AT as a depot buffering an excess
of energy (Gavrilova et al., 2000), (Kim et al., 2000).
15
Hence, the relationship between amounts of subcutaneous and visceral
AT with IR, a proxy for MS, in obesity is not straightforward.
A. Adipose Tissue
Human AT is a heterogeneous, metabolically active organ (Arner P,
2001). In fact, with the current knowledge on its ability to alter metabolic
signaling, it can most appropriately be considered to be a peripheral organ.
i.
Structure:
AT is described by histologists as a loose connective tissue with
collagen as a major contributor of non-cellular mass (Mariman and Wang,
2010). In adult mammals, the major bulk of AT is a loosely held association
of lipid-filled parenchymal adipocytes which are embedded in a framework
of collagen fibers containing stromal-vascular cells namely fibroblastic
connective tissue cells, leukocytes, macrophages, pre-adipocytes and nerve
cells (Kershaw and Flier, 2004), (Qureshi and Abrams, 2007).
16
In humans, AT is the only organ with unlimited growth potential at
any stage of life (Qureshi and Abrams, 2007). Its size fluctuates as a
function of both adipocyte number and volume (Qureshi and Abrams, 2007).
In the first phase of adipogenesis, pre-adipocytes are believed to differentiate
into adipocytes and are hence committed to TGA storage. This involves
considerable alteration in cell shape (Mariman and Wang, 2010). In the
second phase of adipogenesis, the mature adipocytes apparently continue to
accumulate vast amounts of TGA. The increased uptake of fat by mature
adipocytes results in increased cell size (hypertrophy) and eventually
increased tissue mass (hyperplasia) (Mariman and Wang, 2010). Thus, AT
has the potential to expand by increase in adipocyte number (hyperplasy)
and/or by increase in the size of adipocytes (hypertrophy) (Qureshi and
Abrams, 2007).
However, there is a limit to the amount of fat that can be taken up by
mature adipocytes. This is believed to be determined by a number of factors
like: the ability of the extracellular matrix (ECM) to expand and the limited
oxygen supply to the cell contents as a result of altered cell surface-volume
ratio (Mariman and Wang, 2010). The later creates a condition of hypoxia
17
which induces apoptosis in some of the adipocytes. This may give rise to the
vicious cycle of increased infiltrating macrophages and inflammation.
Alternatively, the expanding adipocytes may release cytokines to attract
macrophages.
These
macrophages
through
the
release
of
matrix
metalloproteases (MMPs) may stimulate adipose tissue remodeling. Under
conditions of positive energy balance, the persistent macrophage infiltration
may trigger a systemic inflammatory state that is typical for obese
individuals (Figure 4).
18
Figure 4: Adipose Tissue Inflammation. Under conditions of positive
energy, expansion of adipose tissue may result in hypoxia. The resulting
adipocyte apoptosis gives rise to vicious cycle of increased and persistant
infiltration of macrophages and subsequent inflammation. Alternatively,
expanding adipocytes may release cytokines to attract macrophages.
Macrophages may trigger an inflammatory state characteristic of obesity.
19
ii.
Function:
Traditionally, AT was considered to be a passive tissue that served as
a fat storage organ mainly in the form of TGAs and FFA (Mora and Pessin,
2002). This view is however, no longer valid (Kershaw and Flier, 2004).
Adipose Tissue as an Endocrine Organ:
The first report suggesting an active metabolic role for AT and a role
beyond repository for lipids was published by Von Gierke (Von Gierke E,
1906), who recognized the role of AT in glycogen storage. After the
discovery of leptin in 1994, an endocrine role of AT became obvious. Now,
AT is commonly viewed as a complex, dynamic endocrine metabolic organ
(Kershaw and Flier, 2004) with key roles in mammalian physiology
20
(Trayhurn and Beattie, 2001), (Arner P, 2001), (Klein et al., 2006),
(Mariman and Wang, 2010). The ability to expand in size during periods of
positive energy balance and shrink under conditions of limited food intake
and excessive energy spending, affords this organ the function of long term
energy reservoir. This unique capability is conferred by insulin, which, in
adipocytes, markedly stimulates both glucose uptake and lipogenesis (Mora
and Pessin, 2002). In addition to fat storage, the AT is also capable of
mobilizing stored fat for oxidation in other organs (Trayhurn and Beattie,
2001). Further, AT functions as an endocrine/paracrine regulator of energy
metabolism, provides thermal insulation to the body in adverse weather
conditions and protects internal organs from mechanical damage by
providing shock insulation (Mariman and Wang, 2010).
Collectively, over 50 hormones and bioactive peptides are released by
the AT. The list continues to increase to include adiponectin, resistin,
visfatin, apelin, vaspin, hepcidine, TNF-α, chemerin, omentin, MCP-1, PAI
and others. These hormones modulate AT function (Greenberg and Obin,
2006), act as systemic inflammatory mediators and nutrient status sensors
for glucose homeostasis and lipogenesis (Lee et al., 2009).
21
Adipose Tissue as a Participant in Lipid Homeostasis:
AT, especially white AT, is the main reservoir of energy. Adipocytes
store energy in the form of cytoplasmic TGA fat droplet. The TGAs come
either from dietary FA or from de novo synthesis of TGA from non-lipid
substrates such as glucose (Ferré and Foufelle, 2007) by the process of
lipogenesis. Lipogenesis encompasses fatty acid synthesis from acetyl-CoA
and malonyl-CoA precursors through the action of enzymes called fatty acid
synthases (FAS). This is followed by esterification of fatty acids with
glycerol to produce TGA (Figure 5) (Kersten, 2001).
The AT is considered as the body's largest storage organ for energy in
the form of TGAs, which are mobilized through lipolysis process, to provide
fuel to other organs and to deliver substrates to liver for gluconeogenesis
(glycerol) and lipoprotein synthesis (free fatty acids;FFA). The release of
glycerol and FFAs from the AT is mainly dependent on hormone-sensitive
lipase which is intensively regulated by hormones and agents, such as
insulin (inhibition of lipolysis) and catecholamines (stimulation of lipolysis)
(Large et al., 2004).
22
Adipocytes not only store fat, but also release FFA into the
circulation, which are used by most other organs for fuel when glucose is
limiting (Figure 6). These FFAs are generated by breaking down TGAs
(lipolysis). TGAs contain more energy per unit mass than glycogen which is
another source of energy. The ability to be stored anhydrously, further adds
to the importance of TGA as an energy source. By contrast, glycogen has
only half the energy content per unit of pure mass, and must be stored in
association with water, thus decreasing its efficiency as an energy source
(Rosen and Spiegelman, 2006). Upon specific stimuli such as food scarcity
and/or increase in energy expenditure, TGA reserves are released to provide
fuel for energy generation (Sethi and Vidal-Puig, 2007). This ability of AT
is conferred by the action of lipases. Lipases enzymatically break down
TGA into glycerol and FFA (Figure 6) that can then be transported by the
blood to mitochondrial dense tissues such as the liver and muscle where they
are used up in fatty acid oxidation (Trayhurn and Beattie, 2001). There is
also evidence that both glycerol and FFA can be re-esterified in adipocytes,
thereby allowing precision for FFA flux regulation (Figure 6) (Sethi and
Vidal-Puig, 2007).
23
Figure 5: Lipogenesis. The origin of triglycerides (TGA) can be either from
dietary source or de novo synthesis from non-lipid substrates such as
24
glucose.The process of lipogenesis encompasses fatty acid synthesis from
acetyl-CoA and malonyl-CoA precursors through the action of enzymes
called fatty acid synthases. This is followed by esterification of fatty acids
with glycerol to give rise to TGA.
Dietary TGAs:
Most of the plasma TGA is provided by dietary lipids. Dietary TGAs
are digested by pancreatic lipase in the small intestine, forming
monoacylglycerol and fatty acids (Large et al., 2004). These end products
are then taken up by the intestinal enterocytes and reconverted back to
TGAs. The TGAs are then packaged along with cholesterol and
aplolipoproteins into chylomicron particles within the enterocytes. The core
of each chylomicron consists of TGAs along with cholesterol esters whilst
25
apolipoproteins, cholesterol and phospholipids form the external layer
(Large et al., 2004), (Pattrick et al., 2009). Chylomicrons are then
transported from the intestine to the liver (Qureshi and Abrams, 2007) and
via mesenteric lymphatic channels also to the AT and muscle. Here the
chylomicrons are taken up by the adipocytes, hepatocytes or muscles and are
either stored or oxidized for energy. The liver is the major organ for
processing chylomicrons into various lipoprotein forms such as VLDL and
LDL which are then transported to peripheral tissues such as the AT. In the
AT, under stimulation by hormones, such as insulin, the cellular lipoprotein
lipases hydrolyze these lipoproteins to release FFA. These FFAs are either
released into blood stream to serve as energy source or are repackaged with
glycerol to form TGAs. TGAs are then added to the central lipid droplet in
the AT (Large et al., 2004) (Figure 6).
26
Figure 6: Fat Homeostasis. Role of liver and WAT in fat storage and
mobilization.
De novo TGAs:
27
In addition to dietary TGAs, the liver synthesizes TGAs de novo from
circulating FFA and glycerol by the process of de novo lipogenesis taking
place in the postprandial state. De novo lipogenesis facilitates glucose
utilization under the influence of insulin. Newly synthesized TGAs are
secreted into blood as very low density lipoproteins (VLDL) and low density
lipoproteins (LDL) (chylomicrons) that are then either stored in AT as reesterified TGA or metabolized into FFA during conditions of energy
expenditure or starvation (Figure 6). Thus, the sources of FFA for hepatic
TGA formation are either the diet derived plasma pool of FFA or the lipids
synthesized de novo.
The energy rich Western diet provides surplus of dietary TGA and an
FFA in plasma, further augmented by high levels of the blood glucose
available for the conversion into FA (Qureshi and Abrams, 2007). Together
with a sedentary lifestyle, Western diet amounts to a positive energy balance
and an accumulation of surplus energy in the form of fat droplets in the
storage organ – AT. While adipocytes have a unique capacity to store excess
FFAs as TGA droplets, cells of non-adipose tissue have a limited capacity
for storage of lipids. Hence, when the FFAs levels exceed the storage
28
capacity of adipocytes, it is expected that fat gets deposited at ectopic sites
such as heart, kidney, blood vessels, liver, muscle and pancreas (Schaffer,
2003), (Montani et al., 2004), (Qureshi and Abrams, 2007). Importantly, the
hepatic uptake of FFA from the plasma pool is not regulated as evidenced by
the influx being directly proportional to the plasma FFA concentration. This
increases the susceptibility of liver to ectopic accumulation of TGA in the
presence of high levels of circulating liposomes and FFAs.
This ectopic accumulation of fat could result in mechanical stress on the
organ by physical compression or altered metabolism of the organ because
of altered TGA metabolism and secretion of bioactive peptides from
metabolically stressed tissue (Unger and Orci, 2001).The lipid accumulation
may eventually lead to cell dysfunction or cell death, a phenomenon known
as lipotoxicity (Schaffer, 2003).
Adipose Tissue as a Participant in Glucose Homeostasis:
Relatively steady serum glucose levels are maintained despite a habit
of intermittent food consumption. The glucose homeostasis is achieved
29
through the concerted action of several different organs such as AT, liver,
pancreas, muscle and hypothalamus (Herman and Kahn, 2006), (Rosen and
Spiegelman, 2006) mediated by both endocrine and non-endocrine
mechanisms. The importance of AT for glucose homeostasis is supported by
the development of hyperglycemia and IR under conditions of obesity as
well as lipodystrophy (Rosen and Spiegelman, 2006).
In postprandial state, insulin secretion by pancreas promotes glucose
uptake in AT and muscle. Insulin also prevents the liver from overproducing
glucose by suppressing glycogenolysis and gluconeogenesis. Many ATderived molecules such as leptin, TNF-α and IL-6 are also involved in
insulin sensitivity. Leptin, for instance, increases insulin sensitivity in
muscle and liver (Rosen and Spiegelman, 2006), while the reports on role of
IL-6 and TNF-α on insulin sensitivity are conflicting. Fasting AT releases
FFA that reduces glucose output by adipocytes and muscle and promote
hepatic glucose production. Another mechanism by which AT regulates
glucose homeostasis is by acting as a glucose sponge through its abundance
of insulin–dependent glucose transporters such as GLUT4.
30
The process of glucose homeostasis and lipogenesis are closely related.
Plasma glucose can stimulate lipogenesis through several mechanisms.
Firstly, glucose is the substrate of lipogenesis. Glucose is converted to
pyruvate via glycolysis with the release of adenosine triphosphate (ATP) in
the cell cytoplasm. In the presence of oxygen and pyruvate dehydrogenase,
pyruvate undergoes conversion to acetyl CoA in the mitochondria which
then enters the Kreb’s cycle. This begins a series of reactions culminating in
the generation of citrate. Mitochondrial citrate is then shuttled to the cytosol
where it is converted to acetyl-CoA by ATP citrate lyase. The Acetyl-CoA
carboxylase 1 (ACC1) enzyme then converts acetyl-CoA to malonyl-CoA,
which is used by fatty acid synthase to form different long chain FA in the
cytosol thus, concluding de novo lipogenesis (Figure 7) (Ferré and Foufelle,
2007), (Qureshi and Abrams, 2007).
31
Figure 7: Glycolysis-Lipogenesis. The process of glycolysis and de novo
lipogenesis are closely related.
32
AT is targeted by a number of signaling molecules. A variety of
membrane receptors (Kershaw and Flier, 2004) including receptors for
several well-known pro-inflammatory molecules (e.g., TNF-α, IL-6) are
expressed on adipocytes (Weisberg et al., 2003). This notion supports the
models in which adipocytes are both the source and target of proinflammatory signals (Weisberg et al., 2003). The milieus of adipocyte
hormones and their respective receptors trigger regulatory downstream
responses in AT and several other target tissue. These responses modulate
AT crosstalk with other tissues and maintain the homeostasis of energy
(Mora and Pessin, 2002).
Hence AT, with its adipokine repertoire and various receptors is at the
crossroads of multiple signal exchanges (Lee et al., 2009). For instance,
lipogenesis involves AT along with liver; glucose homeostasis involves AT,
pancreas, liver and muscle (Rosen and Spiegelman, 2006). Apart from the
crosstalk between AT and other organs, there is crosstalk between different
cells within the AT which contributes to the overall impact of this organ
(Lee et al., 2009). Thus, adipocyte biology can no longer be considered in
isolation and must take into account a complex interplay amongst several
33
tissues (Mora and Pessin, 2002). However, despite the large number of
molecules secreted by adipocytes and the large repertoire of receptors on
adipocytes, knowledge of pathways and mechanisms involved in energy
homeostasis are limited. Deregulation of any of these processes can disrupt
homeostasis of body weight and result in manifestation of obesity.
iii.
Adipose Tissue Classification:
In mammals, adipocytes are present in many different sites throughout
the body and generally occur in areas of loose connective tissue (Sethi and
Vidal-Puig, 2007). The two largest depots are subcutaneous AT (SAT) and
visceral AT (VAT) (Figure 8). SAT is located beneath the skin; it
corresponds to more than 80% of total body fat. VAT is located around
internal organs, corresponding to about 10 - 20% of the total body fat (Oka
et al., 2009). VAT is often referred to as abdominal fat, intra-abdominal fat
or organ fat.
Anatomically, it is present mainly in the mesentery and
omentum. Transcriptome and metabolome profiling technologies have
revealed depot-specific differences in these types of AT (Lefebvre et al.,
1998). Notably, VAT is substantially more metabolically active as compared
34
to SAT (Hajer et al., 2008), (Ibrahim, 2010). The classical perception of AT
as a storage place of FFA has been replaced over the last years by the notion
that the AT has a central role in lipid and glucose metabolism and produces
a large number of hormones and cytokines, e.g. TNF-α, IL-6, adiponectin,
leptin, and PAI-1.
In addition to the commonly recognized VAT and SAT depots, most
mammals also have structural AT that provide mechanical padding but
contribute relatively little to energy homeostasis. Examples include the fat
pads of the heels, fingers and toes, and the periorbital fat supporting the eyes
(Ouchi et al., 2011). Additionally, adipocytes are also found in the bone
marrow, lungs and the adventitia of major blood vessels (Ouchi et al., 2011).
35
Figure 8: Major AT depots. Subcutaneous AT (SAT) occurs beneath the
skin and visceral AT (VAT) occurs around intra-abdominal organs (Image
Source:
http://www.bloomtofit.com/danger-below-visceral-fat-vs-
subcutaneous-fat)
36
There is an ongoing debate concerning the role of each of the fat
depots in obesity. Different AT depot have different morphological
characters, replicative potential, metabolic response, function and adipokine
profile (Rosen and Spiegelman, 2006). In general, visceral adipocytes are
smaller than subcutaneous adipocytes in both lean men and women.
Interestingly, while this size difference is lost in obese men, it persists in
obese women (Lafontan and Girard, 2008).
VAT, as compared to SAT, is more cellular, vascular, innervated and
contains larger number of inflammatory and immune cells. Additionally,
VAT has lesser pre-adipocyte differentiation capacity and greater percentage
of large adipocytes (Ibrahim, 2010). Apart from structural differences, VAT
adipocytes are metabolically more active, more sensitive to lipolysis and
more insulin-resistant than SAT adipocytes (Ibrahim, 2010), (Hajer et al.,
2008). In humans, VAT has a greater capacity to generate FFA and has more
glucose uptake than SAT (Ibrahim, 2010). VAT has been shown to contain
more glucocorticoid and androgen receptors than SAT (Ibrahim, 2010). The
higher lipolytic activity in visceral adipocytes triggers the rapid mobilization
of FFA from VAT. This could be attributed mainly to the regional variation
37
in the action of the major lipolysis-regulating hormones, catecholamines and
insulin. Relative to VAT, the lipolytic effect of catecholamines is weaker
and the anti-lipolytic effect of insulin is more pronounced in SAT. This
depot-specific variation is shown to involve the function of different
adrenoceptor subtypes (in particular, α2 and β3) and the function of the
insulin receptor. In addition to initiation of lipolysis, inhibition of lipolysis is
also different among the depots. Inhibition of lipolysis is more pronounced
in SAT than in VAT (Arner, 1998).
The relative sizes of VAT and SAT depots are shown to be
differentially associated with metabolic risks. Even though greater
cardiovascular and metabolic risks are reported to be associated with VAT
accumulation, increased volume of SAT has also been associated with
insulin resistance and metabolic disorders (Fox et al., 2007), (Sam et al.,
2008). These differences raise a number of questions such as - What is the
mechanism behind regional fat distribution? Why is it more dangerous to
accumulate fat in the visceral region? Does the direct access of VAT to the
liver contribute to observed association of NAFLD with obesity?
38
There are two possibilities explaining why VAT maybe a greater
contributor to glucose intolerance, insulin resistance, lipid abnormalities and
liver damage. The first view is the ‘Portal View’, initially proposed by
Bjorntorp (Bjorntorp, 1990). VAT drains directly into the portal circulation;
thus, it has direct access to the liver. Additionally, the higher lipolytic
potential of VAT and lower sensitivity to inhibition of lipolysis enables
rapid mobilization of FFA into portal circulation. A high concentration of
“portal” FFA may have undesirable effects on the liver, including
stimulation of gluconeogenesis, increased triglyceride synthesis and
inhibition of insulin breakdown (Arner, 1998). Specifically, VAT-derived
FFA have been shown to alter several liver functions such as glucose uptake,
insulin sensitivity, reduced mitochondrial fatty acid oxidation (Girard and
Lafontan, 2008). Furthermore, increase in portal FFA when accompanied
with an increased rate of hepatic fatty acid uptake from plasma and de novo
fatty acid synthesis (Fabbrini et al., 2010), can manifest as hepatic steatosis.
The second approach to explain the association of VAT with altered
glucose homeostasis, lipid homeostasis and liver damage is the ‘Endocrine
View’. VAT, in addition to its metabolic differences from SAT, also differs
39
in its protein and adipokine profile. For example, genes for angiotensinogen
(blood pressure regulation), complement factors, and fatty acid-binding
protein 4 (involved in fatty acid trapping in adipocytes), are expressed at
higher levels in VAT than in SAT (Hajer et al., 2008). Leptin, however, is
mainly produced by human SAT, while TNF-α is produced equally by both
fat depots (Hajer et al., 2008). IL-6 levels in SAT, increase with BMI and
decline with weight loss (Fried et al., 1998). In morbidly obese individuals,
VAT, produces more IL-6 and IL-8 than SAT (Fried et al., 1998). The
release of VAT proteins into portal circulation raises the potential of altered
gene expression and metabolic profile of the liver. For instance, IL-6 levels
have been found to be elevated in the portal vein of obese patients (Fontana
et al., 2007) and these peptides have multiple effects on liver (Kershaw and
Flier, 2004), (Trayhurn and Wood, 2004).
The observed increase in the infiltration rate of monocytes into VAT
versus SAT (Hajer et al., 2008), adds support to the hypothesis. A study in
extremely obese patients indicated that visceral fat is the main contributor of
IL-6 in plasma (Hajer et al., 2008). In-turn, IL-6 is an inducer of liver Creactive protein (CRP) production and the proteins involved in hemostasis
40
(PAI-1, fibrinogen, tissue plasminogen activator). IL-6 also contributes to
dyslipidemia via transcriptional repression of microsomal TG transfer
protein which controls the hepatic assembly of apolipoprotein B (ApoB)containing lipoproteins (Navasa et al., 1998). It is therefore conceivable that
VAT directly influences liver function because of direct portal draining of
VAT derived adipokines into the liver.
Both VAT and SAT are innervated by the autonomic nervous system,
with different motor neurons separately serving each depot and firing under
the control of neuro-endocrine feedback (Hajer et al., 2008). Stimulation of
the parasympathetic nervous system leads to an anabolic state with
decreased lipolysis, while stimulation of the sympathetic nervous system
leads to a catabolic state with reduced adipogenesis and stimulated lipolysis
(Hajer et al., 2008). However, at present it is unclear whether and how these
different modes of neural innervations lead to functional differences in AT
(Hajer et al., 2008).
The increased prevalence of excessive visceral obesity and obesityrelated risk factors is closely associated with the rising incidence of
41
cardiovascular diseases and T2D. This clustering of vascular risk factors
such as blood pressure, circulating triglyceride levels, glucose, insulin
resistance, low-grade inflammation indicated by CRP levels is often referred
to as MS. The close relationship between an increased quantity of visceral
fat, metabolic disturbance and cardiovascular diseases has led to increased
interest in VAT. In addition, VAT has direct access to the liver via portal
circulation (Shoelson et al., 2007). These differences are of importance when
considering the role of various AT depots in obesity and related disorders
(Sethi and Vidal-Puig, 2007) and have resulted in an intense endeavour to
unravel the specific endocrine functions of this depot.
In addition to AT classification based on location, another widely
accepted classification of AT refers not to its location; but to its function
(Fantuzzi, 2005), (Sethi and Vidal-Puig, 2007). There are two known forms
of AT, white AT (WAT) and brown AT (BAT), which perform essentially
opposite functions. These two forms of adipose also differ at the
morphological and molecular levels (Frühbeck et al., 2009).
42
The WAT has adipocytes with a single, but large droplet of TGA and
a few mitochondria; whereas the brown fat cell has numerous small lipid
droplets, abundant mitochondria and is supported by rich vascularization
which gives BAT the characteristic brown color (Enerbäck, 2009).
The principle function of WAT is as an storage organ for TGA that
can be released in the form of FFA in times of energy requirements (Figure
6) (Sethi and Vidal-Puig, 2007). BAT, on the other hand, is characterized
mostly by its remarkable capacity to oxidize fat, producing heat but not
synthesizing ATP from ADP (Arch, 1989). BAT plays a central role in both
non-shivering and diet-induced thermogenesis (Frühbeck et al., 2009).
White Adipose Tissue (WAT):
WAT represents the majority of the AT in humans (Fantuzzi, 2005).
Approximately 60% - 85% of the weight of WAT is composed of lipids,
with 90% - 99% being TGA (Qureshi and Abrams, 2007). Small amounts of
FFA, diglycerides, cholesterol and phospholipids are also present (Qureshi
and Abrams, 2007). Quantitatively, FA are the major secretory product of
43
WAT, reflecting the role of this tissue as a fuel reservoir (Trayhurn and
Beattie, 2001). WAT also stores cholesterol and is involved in the
metabolism of steroid hormones (Trayhurn and Beattie, 2001). This tissue
however, does not synthesize steroid hormones de novo but rather expresses
enzymes which are involved in the conversion of both glucocorticoids and
sex hormones, which are subsequently released in circulation (Trayhurn and
Beattie, 2001). In WAT, estrone is converted to estradiol and
androstenedione to testosterone, while androgens can be aromatized to be
converted to estrogens (Trayhurn and Beattie, 2001). The enzyme
lipoprotein lipase was in effect the earliest recognized secretory protein
product of adipocytes. Other secreted proteins from WAT directly involved
in lipid and lipoprotein metabolism include cholesteryl ester transfer protein
(CETP) and apolipoprotein E (ApoE) (Trayhurn and Beattie, 2001). CETP
plays an important role in the accumulation of cholesteryl ester by AT
(Trayhurn and Beattie, 2001). ApoE is highly expressed in AT where it
plays an important role in modulating adipocyte triglyceride metabolism,
triglyceride mass and adipocyte size (Huang et al., 2006).
44
Apart from the above mentioned proteins, a wide range of protein
factors are secreted from WAT, such as leptin, adiponectin, resistin, visfatin,
IL-6, IL-8, TNF–α, ASP, PAI-1 and so on (Trayhurn and Beattie, 2001),
(Pattrick et al., 2009). These soluble molecules include cytokines as well as
proteins directly involved in lipid metabolism, complement system and in
vascular homeostasis (Trayhurn and Beattie, 2001).
There are several key challenges in the continuing investigation of the
secretory functions of WAT such as identification of the complete repertoire
of proteins secreted from adipocyte, establishing specific roles for each of
these proteins, quantifying their impact on adipocyte function and
understanding the response to major changes in AT mass.
Brown Adipose Tissue (BAT):
In contrast to WAT, BAT adipocytes stores TGA in multilocular
droplets. This pool of TGAs is available for quick access and aimed at heat
production
through
mitochondrial
“uncoupling”
of
oxidative
phosphorylation of FFA (non shivering thermogenesis) (Sethi and Vidal45
Puig, 2007). Despite its small contribution to whole body weight (2% in
rodents, substantially less in humans), BAT has a unique and extraordinary
ability to generate heat by non-shivering thermogenesis. It is estimated that
during maximal activation, BAT is able to produce as much heat as the rest
of the body through markedly enhanced rates of fatty acid and glucose
oxidation (Cannon and Nedergaard, 2004). Thus, BAT can potentially
constitute a significant FFA and TGA-clearing organ especially under
sympathetic activation with possible implications in the treatment of
dyslipidemia, hyperglycemia, and IR.
i.
BAT and Thermogenesis:
The regulation of body temperature is essential in homeotherms and
brings into play the importance of thermogenic mechanisms, which are
triggered by a change in internal or external temperature. Thermogenic
processes are not only induced by exposure to cold or heat, and by
hibernation in some species, but also by a variety of pathophysiological
situations that result in changes in body temperature. Such situations include
fasting, food intake, physical exercise, hypo- and hyperthyroidism, alcohol
46
consumption, infection and malignant tumors (Ricquier and Bouillaud,
2000).
In animals, the oxidation of food substances is coupled to the
reduction of nicotinamide adenine dinucleotide (NAD) to nicotinamide
adenine dinucleotide hydride (NADH). The NADH generated is then
oxidized by the mitochondrial electron transport chain (oxidative
phosphorylation) which in turn, is coupled to the setting up of a proton
gradient across the inner mitochondrial membrane. The process of oxidative
phosphorylation involves electron transport between the redox carriers
(Complex I-V) present within the mitochondria (Figure 9). The energy
generated during this process is used to transport protons out across the inner
mitochondrial membrane generating an electrochemical gradient. The
electrochemical gradient is then used to push back protons via ATP synthase
into the mitochondrial matrix. This process allows ATP synthase to generate
ATP.
47
Figure 9: Electron Transport Chain (ETC). ETC couples electron transfer
between spatially separated redox carriers (Complex I-IV). These redox
reactions occur with the transfer of proton (H+) across the inner membrane
of mitochondria into the intermembrane space, creating a transmembrane
48
proton or electrochemical gradient. ATP synthase uses this gradient to
generate ATP.
However, in the presence of uncoupling protein-1 (UCP1), the protons
re-enter the mitochondria via UCP1, instead of ATP synthase. As a result of
this altered path, energy is dissipated as heat (Figure 10). Non-shivering
thermogenesis in mammals is attributed to the presence of this process in
specialized cells referred to as brown adipocytes (Ricquier and Bouillaud,
2000). Non-shivering thermogenesis is especially important in hibernating
animals and in newborn humans who are susceptible to heat loss (Enerbäck,
2010).
49
50
Figure 10: UCP1 mediated thermogenesis. Uncoupling protein-1 (UCP1)
functions to uncouple electron transport from ATP synthesis by providing an
alternative route for protons to re-enter mitochondria. Thus, the proton
motive force is dissipated as heat.
Thermogenesis is a facultative process. This means that, even if the
tissue is present and is well differentiated (including having high levels of
UCP1), the tissue stays inactive in warm surroundings. The tissue, however
has the potential to be acutely activated (in a matter of minutes) in a cold
environment (Nedergaard et al., 2007). The main regulator of this activity is
nor-epinephrine which acts primarily through β-adrenoreceptors (Arch,
1989).
51
Given the importance of thermogenesis in dissipating energy as heat,
a number of studies in animal models have been carried out to determine its
role in obesity. BAT is recognized to be present and active in human
newborns and is responsible for their successful maintenance of body
temperature without shivering (non-shivering thermogenesis) (Nedergaard et
al., 2007), (Sethi and Vidal-Puig, 2007). The general consensus was that
BAT is rapidly lost postnatally, with adult humans possessing no more than
vestigial amounts of BAT (Nedergaard et al., 2007). A number of unrelated
studies in the area of nuclear medicine, since 2002 have challenged this
view. Based on enhanced uptake of Fludeoxyglucose18F (18F-FDG) during
positron emission tomography (PET), the presence of BAT has been shown
in various regions of adult humans (Nedergaard et al., 2007), (Frühbeck et
al., 2009), (Virtanen et al., 2009).
However, whether enhanced FDG uptake on a PET scan serves as a
true indicator of BAT prevalence remains to be verified. When the same
patient was examined several times, the presence or absence of BAT by
FDG uptake was not reproducible (Nedergaard et al., 2007). The
irreproducibility goes both ways; i.e., an uptake into BAT observed initially
52
may show “resolution” on a later occasion, or enhanced FDG uptake may
develop in a patient who had not previously demonstrated enhanced FDG
uptake. In one remarkable study, 33 women were successively examined
five times during an anticancer treatment. Apparently, only six of these
women did not demonstrate BAT labeling on any of these occasions and
only one woman demonstrated it on all five occasions (Nedergaard et al.,
2007). This discrepancy in BAT detection could be attributed to the
prevailing temperature at the time of PET scan. A number of studies since
then have shown temperature as a important factor in activation of BAT
(Au-Yong et al., 2009), (Virtanen et al., 2009), (Van Marken Lichtenbelt et
al., 2009). The variability in BAT detection could be a combination of the
acute character of the regulation of BAT activity and specific efforts made
during routine PET scans to ensure that patients are thermally comfortable
during the procedure (Frühbeck et al., 2009).
Interestingly the presence and the relative amounts of BAT in humans
show sexual dimorphism. The prevalence of detectable BAT has been shown
to be higher in women than men with a ratio of 2.4:1.0 (Cypess et al., 2009).
The presence of BAT also shows age-dependent variation. Two very recent
53
reports have also shown BAT to be more common in adolescents (40%) than
in adults and interestingly this presence was not associated with sexual
dimorphism (Drubach et al., 2011), (Gilsanz et al., 2012).
Notably, another very recent study found supraclavicular fat in
humans to contain scattered cells with multilobular lipid droplets and
detectable expression of BAT-related mRNAs, even though these subjects
had no detectable BAT by a PET/CT scan (Lee et al., 2011). This opens up
the possibility of the presence of BAT remnants in different locations that
can remain undetected by PET scan.
The topographic distribution of BAT in humans is slightly different
compared to rodents. Humans are born with BAT located mainly around the
neck, between the shoulder blades and the large blood vessels of the thorax.
In young adults, the major locations of BAT are (1) cervical/axillary, (2)
perirenal/adrenal, and (3) along the large blood vessels and trachea, and as
finger-like projections following the intercostal arteries (Enerbäck, 2009)
(Figure 11). However, unlike in rodents, human BAT depots are often
54
interspersed within the WAT depots. This, coupled with the facultative
nature of thermogenesis complicates the detection of BAT.
Figure 11: Location of BAT in Adults and Infants. The predominant
locations based on results from postmortem and PET/CT studies: (A)
55
thyroid/tracheal, (B), mediastinal, (C) paracervical/ supraclavicular, (D)
parathoracical, (E) supra and perirenal. In the infant, a kite-shaped thin layer
of BAT has been described (in light brown) (Image Source: Enerbäck,
2009).
ii.
BAT and Obesity Prevention:
In the postprandial state, AT takes up glucose and FFA and stores it as
TGA by the process of lipogenesis. Energy deficiency or fasting, however,
may initiate lipolysis by triggering secretion of hormones, including
epinephrine, nor-epinephrine, and glucagon. This, in turn, leads to the
release of FFA and glycerol, thus, supplying other tissues with metabolites
and energy substrates. Thus, WAT functions largely as an energy source for
energy consuming tissues such as the brain, heart, BAT and skeletal muscle.
56
In accordance with this role is the notable difference in the mitochondrial
system. Energy consuming tissues possess well-developed mitochondrial
systems in contrast to energy storing WAT. Several investigations point
towards a role for the deficiencies in mitochondria and mitochondrial dense
tissues in metabolic disorders associated with obesity like IR and T2D.
Among the mitochondria dense tissues, BAT is particularly attractive
as an interventional target. The anti-obesity functions of BAT are apparent
from studies in rodents. Surgical de-nervation or excision of interscapular
BAT was soon followed by an abnormal increase in the amounts of WAT
(Hansen and Kristiansen, 2006a). Likewise, suppression of UCP1 expression
via the manipulation by the UCP1 promoter translated into a 60 - 70%
reduction in BAT mass, with animals becoming sensitive to both cold and
diet-induced obesity (Cannon and Nedergaard, 2004). On the other hand,
transgenic mice expressing UCP1 from the relatively strong adipose-specific
promoter of the fatty acid binding protein 4 (also known as aP2) were
protected against genetic and diet-induced obesity (Kozak and AnunciadoKoza, 2008). Targeted disruption of cell-death inducing DFF45-like effector
A (CIDEA), which inhibits the uncoupling activity of the UCP1 protein,
57
results in leaner mice that are resistant to diet-induced obesity (Zhou et al.,
2003). In most rodent models, susceptibility to obesity correlates with
decreased BAT activity, whereas resistance to obesity correlates with
increased BAT function or the induction of brown-adipocyte-like gene
expression in WAT (Cannon and Nedergaard, 2004). Additional examples of
relevance of BAT in the prevention of obesity comes from the genetically
obese mouse models like the ob/ob and db/db strains where BAT is
dysfunctional (Hansen and Kristiansen, 2006).
iii.
BAT Differentiation and Transdifferentiation:
The differentiation of white adipocytes has been dissected in much
greater detail as compared to that of brown adipocytes, in the past 20 years.
PPARγ and the C/EBP family of transcription factors have been found to be
the common transcriptional factors regulating the terminal differentiation
processes of both brown and white adipocytes. Although, both PPARγ and
C/EBPα induce UCP1 gene transcription; UCP1 can still be induced in their
absence (Hondares et al., 2011). Additionally, Rb (retinoblastoma protein),
p107, TIF2 (transcriptional intermediary factor 2), 4E-BP1 (4E-binding
58
protein1), and RIP140 have also been shown to induce a white adipocyte
specific phenotype (Tsukiyama-Kohara et al., 2001), (Picard et al., 2002),
(Hansen et al., 2004), (Leonardsson et al., 2004), (Scimè et al., 2005),
(Hansen and Kristiansen, 2006), (Powelka et al., 2006).
The molecular players tipping the scales towards brown cell type are
much more discreet. Several transcriptional factors and coregulators have
been found to promote the embryonic development and acquisition of the
BAT-specific gene expression profile, including UCP1, PRDM16 (PRD1BF1-RIZ1 homologous domain-containing 16) and PGC-1α (PPARG
coactivator 1A). PRDM16, is a zinc-finger protein, highly enriched in BAT
as compared to WAT (Seale et al., 2008). It is shown to control a brown
fat/skeletal muscle switch (Seale et al., 2008) by binding to and promoting
the transcriptional activity of PPARα. In a complex with C-terminal-binding
protein-1 (CtBP-1) and CtBP-2, PRDM16 is recruited onto the promoters of
WAT-specific genes such as resistin. Its binding with PGC-1α and PPARαcoactivator-1B (PGC-1β) displaces CtBP and allows the activation of BATlineage specific genes, such as PGC1A (Kajimura et al., 2008). Importantly,
when paired with C/EBPα, PRDM16 is also capable of brown fat program
59
induction in naive fibroblastic cells, including skin fibroblasts (Kajimura et
al., 2009).
PGC-1α is a transcriptional coactivator involved in activation of BAT
thermogenesis and induction UCP1 gene expression by co-activating
transcription factor PPARα. Forced expression of PGC-1α induces UCP1
expression (i.e. white to brown transdifferentiation) in the WAT (Tiraby et
al., 2003). Additionally, PPARδ acts downstream of PGC-1α and is
necessary for expression of UCP1 (Pan et al., 2009).
In mice, the adipocytes within the WAT depot may be induced to
transdifferentiate into BAT cells under physiological conditions in vivo,
mediated by beta (3) - adrenergic receptors (Himms-Hagen et al., 2000),
(Granneman et al., 2003). Under treatment with selective agonist of β3adrenoceptors, a transient pro-inflammatory response is induced in the
murine WAT. After three days of treatment, inflammation wanes as the
oxidative capacity of WAT adipocytes increase in parallel with the
appearance of BAT-like multilocular cells that seem to be derived from
mature adipocytes (Granneman et al., 2005). Similar changes are observed
60
after cold acclimatization (Barbatelli et al., 2010). Whether human WAT
depots contain certain proportion of BAT or BAT-like cells that are capable
of thermogenic activation, is currently unknown.
With this in mind we attempted to detect genes involved in
thermogenesis, differentiation and mitochondrial biogenesis and functioning
in VAT depots. The expression of UCP1, PGC1A, PPARD, SIRTs and
NAMPT was assessed in VAT among obese and non-obese controls.
Additionally, the expression of KCNRG – a negative regulator of the
voltage-gated K+ channels (KV1 channels) was assessed in the same cohort.
Gastric Tissue
The stomach is an important organ of the gastrointestinal tract. It is a
hollow muscular organ that performs the second phase of digestion after
food mastication. It has 3 main motor functions: storage, mixing, and
emptying (Kong and Singh, 2008). It stores food during the process of
digestion, which can take as long as 1-2 hours in humans (Maton, Anthea et
61
al., 1993). During this period, stomach helps in chemical and mechanical
breakdown of food with the aid of digestive enzymes such as pepsin. The
stomach environment provides the acidic pH needed for the functioning of
these enzymes and the emulsification of fat, while gastric motility mediated
by smooth muscular contractions aid in mixing acids and enzymes with the
food as well as sending partially digested food to the duodenum (Luciani et
al., 1913). The process of releasing partially digested food (chyme) into the
duodenum is known as gastric emptying (Kong and Singh, 2008). This
feature of the stomach has recently garnered attention with studies revealing
the role of rapid gastric emptying in obesity and delayed gastric emptying in
diabetes mellitus (Kong and Singh, 2008). Further, studies involving
bariatric surgery in humans and model animals also suggest that the stomach
plays important but unclear role in energy homeostasis.
i.
Structure:
The stomach has two orifices: the cardiac orifice which is continuous
with the lower end of the oesophagus; and the pyloric orifice which opens
into the duodenum. The entire stomach can be divided into two regions: the
62
larger cardiac part can be further divided into the fundus and the body and
smaller pyloric part subdivided into the pyloric antrum and the pyloric
canal (Priya et al., 2008) (Figure 12). The stomach wall consists of three
layers (Schuenke et al., 2010) (Figure 12). The mucosal layer is composed
of simple columnar epithelium forming small depressions at the luminal
surface called gastric pits, which contains longitudinally organized cells and
is referred to as the glands (Figure 13) (Schuenke et al., 2010). The mucosa
layer consists of two sub-layers: the lamina propria and muscularis mucosae.
Lamina propria is the connective tissue which lines the bottom of gastric pits
with the gastric glands (Figure 13). Below the lamina propria is the layer of
smooth muscle fibers called muscularis mucosae. Below the mucosa is the
submucosa, the second layer composed of loose connective tissue and
finally the muscularis externa, the third layer composed of smooth muscle
fibers which aid in gastric motility (Priya et al., 2008).
A number of physiologically active substances originate in the
mucosal layer of the stomach. For instance, the chief cells secrete
pepsinogen which is activated at acidic pH to produce pepsin, while gastrin
is secreted by the G cells, and ghrelin by A/X cells (Figure 13) (Kahle and
63
Leonhardt,, 1993). Evidence supporting the role of these physiologically
active compounds in energy homeostasis lends further support to the
potentially unexplored role of stomach as a peripheral tissue in obesity and
metabolic syndrome.
Figure 12: Regions within the Stomach and Gastric Tissue. Structure of
gastric wall showing three layers: mucosa, submucosa and muscularis
64
externa
(Image
Source:
http://www.britannica.com/EBchecked/media/68634/Structures-of-thehuman-stomach-The-stomach-has-three-layers).
65
Figure 13: Gastric Glands Structure. Structure of gastric glands from
body of stomach showing different cells and their location (Image Adapted
From: (Karam, 2010).
ii.
Stomach as an Endocrine Organ:
The stomach deals with food during the process of digestion; thereby,
its role in signaling the nutrient status would not be surprising (Weigt and
Malfertheiner, 2009). Indeed, number of recent studies show the importance
of hormones and receptors of the stomach and the intestine in the regulation
of food intake and satiety signals (Ravussin and Bogardus, 2000), (Gale et
al., 2004), (Cammisotto et al., 2005), (Badman and Flier, 2005), (Murphy
and Bloom, 2006), (Yamada et al., 2008).
66
Stomach is recognized as a source of important paracrine peptides
such as leptin (Bado et al., 1998), (Baratta, 2002), ghrelin (Badman and
Flier, 2005), obestatin, motilin, peptide YY, glucagon, glucagon-like
peptide-1, glucagon-like peptide-2 (GLP-2) and oxyntomodulin (Murphy
and Bloom, 2006). These peptides are secreted by several distinct types of
endocrine cells, including histamine-producing ECL cells, somatostatinproducing D cells, gastrin-producing G cells, ghrelin and obestatin
producing A/X cells and serotonin-producing EC cells (Sakata and Sakai,
2010). Endocrine cells of the stomach release respective peptides in response
to metabolic and chemical cues such as FFA and glucose (Badman and Flier,
2005). Stomach-derived peptides have a wide range of functions, ranging
from regulating blood glucose levels, exocrine secretion, adipocyte function,
gastrointestinal motility and regulating the release of other hormones
(Murphy and Bloom, 2006).
One of the most significant advances in revealing the role of
peripheral tissues in energy homeostasis was the discovery of the hormone
leptin. This hormone modulates energy homeostasis via crosstalk with AT,
gastric tissue and the central nervous system (Figure 14) (Cinti S et al.,
67
2000), (Cammisotto et al., 2005), (Ahima, 2008), (Cammisotto et al., 2010).
Leptin provides feedback on the size of adipose depots to the centrally
located receptors, thus contributing to the control of satiety and body weight
homeostasis (Bado et al., 1998), (Gale et al., 2004). Leptin expression was
initially discovered in WAT. Over time, it has been found to be expressed in
several other tissues. After AT, gastric mucosa is the next largest source of
leptin (Mix et al., 2000), (S Cinti et al., 2001), (Baratta, 2002), (Cammisotto
et al., 2005), (Cammisotto et al., 2010). Both leptin and its receptors have
been found in the lower half of the stomach glands. They are localized to the
pepsinogen granules of chief cells, parietal cells (Mix et al., 2000),
(Cammisotto et al., 2005) as well as apical and basolateral membranes of
enterocytes (Cammisotto et al., 2005). It has been reported that the stomach
can produce and store leptin and release it in response to food intake in both
human and rats (Mix et al., 2000). Consequently, a role has been suggested
for gastric leptin in the short-term control of food intake (Picó et al., 1994).
Co-expression of leptin and it's receptor suggests that gastric leptin may
have a paracrine and/or autocrine function in the human stomach (Mix et al.,
2000). In the brain, leptin signals satiety to the hypothalamus, causing
reduced food intake and increased energy expenditure, while in the stomach
68
it reduces gastrin and acid secretion, and increases gastric mucosal cell
proliferation (Marks and Neill, 2007).
The discovery of gastric leptin was soon followed by the discovery of
another satiety hormone – ghrelin. It was first isolated from rat and human
stomach and characterized as a natural ligand for the GH secretagogue
(GHS) receptor (GHS-R) (Kojima and Kangawa, 2005), (Liew et al., 2006).
It is a member of the novel family of gut–brain regulatory peptides and is
produced by the G cells in the fundus (Liew et al., 2006). Ghrelin is a 28amino acid peptide with an n-octanoyl ester at its third serine residue. The noctanoyl group at serine 3 of the ghrelin molecule is essential for the
endocrine role of ghrelin, because the unacylated form of ghrelin, desoctanoyl ghrelin, does not bind the GHSR type 1a (GHS-R1a) and is devoid
of any endocrine activity (Kojima and Kangawa, 2005). However,
unacylated/des-octanyl ghrelin (UAG) has other biologically role; it is able
to share with ghrelin the antiproliferative effects on human breast and
prostate cancer lines (Gauna et al., 2005). It also has negative inotropic
effects (force or energy of muscle contraction) on papillary muscle, and can
stimulate bone marrow adipogenesis, although the signal transduction
69
mechanism(s) for these effects has not yet been determined (Gauna et al.,
2005).
A gradient of ghrelin production occurs in the gastrointestinal tract
with the highest ghrelin expression and peptide levels in the mucosal layer
of the fundus and the lowest levels in the colon (Asakawa et al., 2001).
Further, ghrelin circulates in human blood at a considerable plasma
concentration. In addition, its receptor GHS-R is present in the
hypothalamus, heart, lung, pancreas, intestine, and AT. These findings
indicate possible endocrine and paracrine roles of ghrelin (Asakawa et al.,
2001) similar to circulating leptin.
Ghrelin plays an important role in the regulation of both appetite and
body weight (Liew et al., 2006). It exerts a powerful effect on the secretion
of growth hormone (GH) in the stomach and also stimulates energy
production by signaling directly to the hypothalamic regulatory nuclei that
control energy homeostasis (Inui, 2001), (Gauna et al., 2004).
70
Besides GH-releasing activity, ghrelin has an influence on obesity
through appetite modulation (Figure 14) (Holdstock et al., 2003) as it is
released in response to fasting (Asakawa et al., 2001). Plasma ghrelin levels
are negatively correlated with BMI, plasma leptin, insulin, and glucose
levels (Gauna et al., 2004). Although the influence of ghrelin on glucose
output by liver has not been directly assessed yet, a possibility that ghrelin
has a direct peripheral action on liver is supported by an in vitro study by
Gauna et al. In this study, ghrelin was able to activate the intracellular
signaling of the insulin receptor and to reverse the inhibitory effect of insulin
on the expression of key gluconeogenic enzymes at the transcriptional level
in both human and rat hepatoma cells (Gauna et al., 2004).
Additionally, ghrelin increases arcuate expression of neuropeptide Y
(NPY), which in turn acts through Y1 receptors to increase food intake and
decrease energy expenditure (Asakawa et al., 2001). Ghrelin also has a role
in motivated reward-driven behaviors via activation of the "cholinergicdopaminergic reward link" (Dickson et al., 2011). This reward link
comprises a dopamine projection from the ventral tegmental area (VTA) to
the nucleus accumbens together with a cholinergic input, arising primarily
71
from the laterodorsal tegmental area. The central ghrelin signaling system
thus interfaces neurobiological circuits involved in both food and chemical
drugs related reward (Dickson et al., 2011).
In 2005, another hormone, obestatin, was identified as a peptide
hormone derived from pre-proghrelin in the stomach. Obestatin has been
shown to be involved in energy homeostasis, gastrointestinal motility,
memory, sleep and cell proliferation (Sakata and Sakai, 2010). Contrary to
the appetite-stimulating effects of ghrelin, treatment of rats with obestatin
suppressed food intake, inhibited jejunal contraction, and reduced gain in
body-weight (Zhang et al., 2005c). Similar to ghrelin, obestatin depends on
post-translational modification by amidation of the C-terminus for its
biological activity.
72
Figure 14: Energy Homeostasis. Interplay of 2 major bioactive peptidesleptin and ghrelin from gastric tissue and AT in energy homeostasis (Image
adapted from: Holdstock et al., 2003).
Peptides from gastric tissue such as motilin, leptin and obestatin, are also
believed to participate in the physiologic
73
mechanisms regulating
gastrointestinal motility and gastric emptying (Asakawa et al., 2001), (Kong
and Singh, 2008). Both mechanisms mentioned above play an important role
in regulating food intake and body weight balance (Asakawa et al., 2001).
Thus, by secreting either satiety-inducing/antikinetic peptides or the
orexigenic/prokinetic peptides, gastric tissue may contribute to the energy
homeostasis in an additional way.
Liver
The liver is a central player in the whole body energy homeostasis by
its ability to metabolize glucose and FFA. When energy intake is abundant,
carbohydrates are preferentially burnt to generate ATP. The surplus glucose,
is first used to replenish glycogen stores, followed by conversion to FFA
(lipogenesis) by liver for use in the transport and storage of TGA in the AT
(Large et al., 2004). Although the AT functions as a major reservoir of TGA,
the liver is also accumulates significant quantities of TGA in conditions
associated with prolonged positive energy or impaired fatty acid
metabolism. In fasted states, when glucose availability and insulin levels are
low, there is a depletion of hepatic glycogen stores and a reduction in de
74
novo FFA production by liver. Additionally, TGAs stored in the AT are
hydrolyzed to FFA (Figure 6) and mobilized into plasma to reach the liver.
In the liver, they undergo oxidation, converted to ketone bodies to be used as
fuel by extrahepatic tissues (Reddy and Rao, 2006).
Thus, several pathways may lead to hepatic steatosis such as 1) excess
dietary TGA associated with overeating that reach the liver as chylomicron
particles from the intestinal enterocytes and overwhelm the flux of TGA
between liver and the fat reservoir - the AT; 2) increased TGA synthesis in
the liver from excess of plasma FFA by de novo lipogenesis; 3) excess FFA
influx into the liver from dysregulated lipolysis of TGA in the AT; 4)
diminished export of VLDL, LDL from the liver; and 5) reduced oxidation
of FFA (Reddy and Sambasiva Rao, 2006), (Dowman et al., 2010). Thus,
there is a close interplay between liver, AT as well as the GI tract in
regulating FFA flux.
The liver is a unique organ with respect to its anatomical location and
and its cellular composition. Liver also has a central role in inflammatory
responses and homeostatic functions such as regulation of blood glucose and
75
lipids. The anatomical location of liver allows continuous blood supply not
only from the arterial system (hepatic arteries) but also from the
gastrointestinal tract via the portal vein (Tacke et al., 2009).
Liver comprises of metabolically active hepatocytes, non-hepatocytic
parenchymal cells, and various immune cell populations (Tacke et al., 2009),
(Malhi et al., 2010). In addition to intrahepatic immune cells, liver is in
constant contact with circulating immune cells via its sinusoidal network.
Blood flow through liver allows circulating immune cells and cytokines to
come in contact with a variety of intrahepatic cell populations (Tacke et al.,
2009). Cytokines are key mediators of the crosstalk between intrahepatic
immune cells and hepatocytes (Oo et al., 2010). They function by activating
effector functions of immune cells, as well as hepatocytic intracellular
signaling pathways. While, the cytokines are primarily derived from resident
Kupffer
cells
and
liver-infiltrating
monocyte-derived
macrophages,
cytokines released by the AT into portal circulation may have a major
influence in intrahepatic signaling.
76
Cytokine action is generally mediated by the interaction of cellular
receptors, which signal internally to the nucleus, and external factors, which
are able to bind these receptors. In many instances, hepatic damage might
result from an imbalance between damaging and protective signals that are
very tightly regulated under physiologic conditions.
During liver
inflammation, the rate of lymphocyte recruitment via sinusoidal endothelium
increases and retention within the liver by localization at sites of
inflammation or at epithelial surfaces results in persistent inflammation (Oo
et al., 2010).
Several studies have shown a mileux of cytokines necessary for liver
damage (Tacke et al., 2009). Among the mileux of cytokines, 2 well studied
cytokines are TNF-α and IL-6. TNF-α is itself shown to activate a number of
pathways (Schwabe and Brenner, 2006). It is not only a mediator of
hepatotoxicity but also contributes to the restoration of functional liver mass
by driving hepatocyte proliferation and liver regeneration. Although TNF-α
may act as a potent activator of both proinflammatory and proapoptotic
pathways, these signaling pathways interact in a complex network at several
77
levels, and activation of one pathway often depends on the inactivation of
another pathway (Schwabe and Brenner, 2006).
Obesity is accompanied with a system wide modulation of
inflammation and activation of immune pathways, including hepatic immune
and homeostatic processes (Bertola et al., 2010a) In obesity and associated
NAFLD, the AT has shown to be a substantial contributor of systemic TNFα,
MCP1,
IL-6,
and
adiponectin
which
modify
the
hepatic
inflammatory/immune system. While, a number of cytokines are known to
be released by the AT during obesity and altered cytokines have been
profiled in NAFLD, the inflammatory pathways involved in the progression
of normal liver to steatosis, and then NASH are not well characterized. Little
is known about which pathways are dominant in specific liver cell
populations in vivo, how these pathways interact functionally, and what
impact they have on expression and persistence of disease.
78
CHAPTER 2: Specific Aims
1.
To find the best set of housekeeping genes for adipose tissue and non-
malignant gastric tissue:
Expression of common reference genes varies with tissue type as well
as physiological state (Bustin, 2000). Thus, for the proposed gene expression
studies in AT and gastric tissue, the first step was to shortlist and validate
candidate reference genes suitable for qRT-PCR profiling experiments.
Eight commonly used reference genes were shortlisted for further validation.
VAT from obese and lean individuals and gastric tissue samples from
patients with and without gastritis was used to find the best set of reference
genes for respective tissues using three different statistical algorithms
GeNorm, NormFinder and BestKeeper. To validate the results and determine
the confidence in the results of analysis, three algorithms based on different
principles were selected. Additionally, the effect of sample size on the
performance of each of the algorithm was also explored.
79
2.
To study expression of the genes known to be expressed in BAT within
intra-abdominal AT depots of the patients with various pathological
conditions:
Anti-obesity function of BAT has been reported in rodents. Thus, it is
plausible that BAT may have a similar role in adult humans and thus,
obesity can be correlated with the loss of BAT activity. However, until now,
no research has been done on the direct detection of BAT cells within adult
human VAT depots. This study aimed to assess expression of genes
involved in transcriptional regulation of UCP1 gene and BAT differentiation
in VATs from adult humans. Thus, these genes were expected to show BAT
specific expression. Additionally, a novel gene KCNRG, discovered in our
laboratory (Ivanov et al., 2003) was also profiled to examine its role in
obesity. VAT from obese and lean patients was used to accomplish these
goals.
3.
To study the expression of obesity-related and inflammatory genes in
gastric tissue samples of patients with various pathological conditions:
80
The process of energy homeostasis involves the interplay of multiple
tissues and hormones (Dobrin et al., 2009). Although, the chief player in this
homeostasis process is AT, the role of gastric mucosa (Cammisotto et al.,
2010) cannot be ruled out. Studies have shown the presence several
hormone-secreting cells in the gastric tissue and their role in appetite. The
potential influence of gastrointestinal tract on liver via the portal vein (Tacke
et al., 2009) makes the study of gastric mucosa in obesity all the more
alluring. Gastric tissue may be one of the forgotten tissues involved in the
pathogenesis of obesity and associated systemic conditions, especially
NAFLD. A number of molecules involved in inflammation and energy
homeostasis have been shown to have an altered expression in AT of obese
patients. To assess if this alteration is also seen in gastric tissue, a panel of
inflammatory and obesity related genes (Qiagen, USA) was examined by
qPCR.
81
CHAPTER 3: Validation of Reference Genes
A. Validation of Reference Genes in VAT
Background:
The increasing prevalence of obesity worldwide has drawn research
on AT into the spotlight. The AT is a complex and highly active tissue with
important metabolic and endocrine functions. It not only plays a central role
in energy balance but also functions as an endocrine organ secreting various
adipokines and cytokines (Gómez-Ambrosi et al., 2004), (Catalán et al.,
2007). On the basis of its location, AT is divided into two main regions:
subcutaneous and visceral fat (Bosello and Zamboni, 2000), (GómezAmbrosi et al., 2004).
82
Accumulation of excessive visceral fat (visceral obesity) is associated
with an array of metabolic perturbations including T2D, IR, NAFLD,
NASH, cardiovascular disease, hypertension and hyperlipidemia together
referred to as MS (Linder et al., 2004), (Shoelson et al., 2007). However, the
role of visceral obesity in MS is yet to be fully elucidated (Furukawa et al.,
2004). Furthermore, a causal relationship between IR and MS has not been
shown conclusively; Obesity seemingly causes IR, on the other hand IR
appears to aggravate and propagate the adverse effects of obesity (Grundy,
2004). This somewhat co-dependent and circular relationship is difficult to
untangle and has generated a multitude of clinical and research publications.
Another area of disagreement involves NAFLD, a common condition
affecting about 70% of obese and overweight individuals and increasingly
being recognized as a major cause of liver-related morbidity and mortality
(Hamaguchi et al., 2005), (Vernon et al., 2011). The pathological picture of
NAFLD encompasses a spectrum of liver injury ranging from simple hepatic
steatosis to more severe manifestations, including NASH, which can
progress to fibrosis, cirrhosis, and liver failure (Hanley et al., 2005), (Rafiq
and Younossi, 2011). Studies have reported frequent association of MS and
83
diabetes in patients with NASH, which can progress to NAFLD (Hanley et
al., 2005), (Kim and Kim, 2003), (Younossi, 2008). It has even been
suggested that hepatic steatosis itself may be the primary cause of IR and
MS in obesity (Garg and Misra, 2002). However, it is still unclear whether
NAFLD is a cause or a consequence of IR (Tilg and Hotamisligil, 2006) and
if MS precedes NAFLD or is a result of NAFLD (Hamaguchi et al., 2005).
Many NAFLD centered studies involve profiling of VAT for the production
of various soluble mediators of inflammation and their release into
circulation, due to the close proximity of this fat compartment to liver.
Real-time PCR (qRT-PCR) is the routine method for studying
changes in relative gene expression in different tissues and experimental
conditions. The popularity of this technique is attributed to its high
sensitivity and specificity (De Jonge et al., 2007). However, variations in
amount of starting material, enzymatic efficiency and presence of inhibitors
can lead to quantification errors. Hence, the need for accurate data
normalization is vital (Vandesompele et al., 2002). Among several known
strategies for data normalization (Huggett et al., 2005), the use of reference
84
genes as an internal control is the most common approach (De Jonge et al.,
2007).
An ideal reference gene is one which is consistently expressed at the
same level in all samples under investigation regardless of tissue type,
disease state, medication or experimental conditions, and exhibits expression
levels comparable to that of the target gene (Suzuki et al., 2000). 18S, Beta
Actin (ACTB), Glyceraldehyde-3-phosphate dehydrogenase (GAPDH), Beta2-microglobulin (B2M), RNA polymerase II (RPII or POLR2A), Tyrosine-3
monooxygenase/Tryptophan-5 monooxygenase activation protein, zeta
polypeptide
(YWHAZ),
Ubiquitin
C
(UBC)
and
Hypoxanthine
phosphoribosyl transferase 1 (HPRT1) are some of the most commonly used
reference genes in qRT-PCR studies (Catalán et al., 2007), (Thellin et al.,
1999), (De Jonge et al., 2007). However, numerous studies have shown that
expression of these common reference genes vary with tissue type as well as
physiological state (Bustin, 2000), (Thellin et al., 1999). This variation can
potentially explain the often encountered divergence between studies and
more seriously, may ultimately result in misinterpretation of data (Suzuki et
al., 2000). The suitability of a particular reference gene thus, depends on the
85
system being investigated and the inherent experimental conditions
(Schmittgen and Zakrajsek, 2000), (Thellin et al., 1999), (Suzuki et al.,
2000).
Recent studies have shown differences in reference gene expression in
omental fat tissue between lean and obese patients. In addition, there is
strong evidence to suggest that obesity and T2D exert a detectable influence
on reference gene expression in SAT and VAT (Gómez-Ambrosi et al.,
2004), (Catalán et al., 2007). In light of these findings, it is crucial for
studies involving VAT to validate the stability of the reference genes being
used.
Increasing concerns about normalization using ideal reference genes
has led to the development of several mathematical algorithms aimed at
determining the stability of reference genes’ expression (Jo Vandesompele et
al., 2009). In 2002, Vandesompele et al. have developed the software
GeNorm that addresses the critical issues of reference gene validation and
ranks candidate reference genes according to their expression stability using
raw, non-normalized expression levels (Vandesompele et al., 2002). Pfaffl et
86
al. have developed similar software- BestKeeper, that takes into account Ct
values of candidate reference genes instead of relative quantities (Pfaffl et
al., 2004). This software employs a statistical algorithm wherein the Pearson
correlation coefficient for each candidate reference gene pair is calculated
along with the probability of correlation significance of the pair. Andersen et
al. used a model-based evaluation strategy which, ranks candidate genes
with minimal inter- and intra-group variation and developed the software
NormFinder (Andersen, 2004). In all the three softwares, the top ranked
genes are recommended for further use in the similar experimental systems
as endogenous controls.
In this study, GeNorm (Vandesompele et al., 2002), NormFinder
(Andersen, 2004) and BestKeeper (Pfaffl et al., 2004) were used to validate
candidate reference genes suitable for qRT-PCR profiling experiments using
VAT samples from obese and lean individuals with and without diabetes.
87
Results:
To determine the expression stability of eight selected reference
genes, RNA expression levels were measured in 19 VAT samples (10 obese
VAT and 9 lean VAT samples) and cross-validated using three popular
algorithms GeNorm v3.4 (Vandesompele et al., 2002), NormFinder
(Andersen, 2004) and BestKeeper (Pfaffl et al., 2004). Genes encoding for
18S, Beta Actin (ACTB), Glyceraldehyde-3-phosphate dehydrogenase
(GAPDH),
Beta-2-microglobulin
phosphoribosyl
(B2M),
transferase1
Hypoxanthine
(HPRT1),
Tyrosine
guanine
3-
monooxygenase/tryptophan 5-monooxygenase activation protein, zeta
polypeptide (YWHAZ), Ubiquitin C (UBC) and RNA polymerase II (RPII, or
POLR2A) were selected according to previously published studies that relied
on these genes as reference controls (Thellin et al., 1999), (Kim and Kim,
2003), (Catalán et al., 2007). For each tissue sample, expression stability of
each gene was calculated using the mean Ct values. The input data for
BestKeeper algorithm was raw Ct values, while the analysis using GeNorm
88
and NormFinder converted raw Ct values to relative quantities using the
comparative Ct method (Vandesompele et al., 2002).
GeNorm Analysis:
The initial investigation of raw non-normalized data of 5 obese and 4
lean VAT samples (N = 9) allowed sorting of genes ranked on the basis of
their expression stability (M) from least stable to most stable (18S →
YWHAZ → UBC → B2M → GAPDH → HPRT1 → RPII and ACTB). The
respective individual M values compared to the other candidate genes were
0.581, 0.501, 0.446, 0.421, 0.377, 0.295, 0.24 and 0.24 (Figure 15a).
Successive elimination of the least stable genes based on the highest M
values led to the identification of ACTB and RPII as the two most stable
reference genes. To determine the effect of sample size, analysis was done
with additional 5 obese and 5 lean VAT samples (N = 19). Yet again, ACTB
and RPII were found to be the most stable genes, but the respective rankings
of the least stable and intermediate genes were different. When genes were
sorted from least to the most stable, the UBC gene expression was shown to
be the least stable (UBC → 18S → B2M → YWHAZ → HPRT1 → GAPDH
89
→ RPII and ACTB). Their respective individual M-values compared to the
other candidate genes were 1.43, 1.17, 0.88, 0.69, 0.59, 0.56, 0.44 and 0.44
(Figure 15b). Thus, an increase in the sample size did not alter the ranking of
the most stable genes, indicating the robustness of GeNorm performance on
AT gene expression data. Additionally, the pair-wise variation calculated
between two sequential normalization factors (NFn and NFn+1) for all genes
indicated the sufficiency of these two reference genes for accurate
normalization (Figure 16a, b).
90
Figure 15: Gene expression stability M of candidate reference genes in
visceral adipose tissue as calculated by GeNorm software. The program
91
proceeds with stepwise exclusion of genes with relatively higher variable
expression among the samples. The expression stability measure (M) is the
average of the stability values of the remaining genes. The lower the M, the
more stable the gene in the subset. a) n = 9, b) n = 19.
92
Figure 16: Determination of optimal number of reference genes for
normalization by pairwise variation analysis using GeNorm software.
Bar values indicate the magnitude of the change in normalization factor after
the inclusion of an additional reference gene. A large variation indicates that
the added gene has a significant effect and should probably be included in
the calculation of normalization factor. a) n = 9, b) n = 19.
93
NormFinder Analysis:
The analysis of gene expression of candidate reference genes in the
two subgroups: 5 obese visceral fat and 4 lean visceral fat tissue samples (N
= 9) using NormFinder, found HPRT1 and GAPDH as the two genes with
the lowest stability values (Table 1). Further manual inspection for genes
with lowest inter-group variations showed that, although UBC ranked third,
it had the highest intra-group variation after 18S. The next two genes in the
stability ranking, B2M and ACTB had similar inter-group variation but B2M
had slightly higher intra-group variation. RPII was ranked below B2M, and
was found to have a slightly higher inter-group variation and minimal intragroup variation. Thus, using assistance of the process of elimination,
GAPDH, HPRT1, ACTB, B2M and RPII were considered as the best
candidate reference genes (Figure 17a). To assess the robustness of this
model, sample size was increased to include additional 5 obese and 5 lean
VAT samples (N=19). The ranking, however led to a different set of stable
genes. Genes were now ranked similar to GeNorm with RPII and ACTB as
94
the most stable (Table 2). Sorting the remaining genes on basis of inter- and
intra- group variability highlighted GAPDH, YWHAZ, and HPRT1 as the
next most stable genes (Figure 17b). Thus, NormFinder rankings vary with
an increase in sample size and with larger sample number the results are in
agreement with GeNorm analysis.
Table 1: Comparison of highly ranked gene lists identified by all three
algorithms (n=9).
95
Table 2: Comparison of highly ranked gene lists identified by all three
algorithms (n=19).
96
97
98
Figure 17: Determination of the most stable reference genes using
NormFinder. Two groups considered were - lean and obese patient tissues.
Bars represent inter-group variances, while error bars represent the average
of intra-group variance. Ideal reference gene is supposed to have inter-group
variation as close to zero as possible and error bars as small as possible. a) n
= 9, b) n = 19
99
BestKeeper Analysis:
Unlike GeNorm and NormFinder, input data for analysis by
BestKeeper were raw Ct values of each gene. Initial analysis of the data with
9 samples (N = 9) calculated variations (SD (± Ct) and CV (%Ct)) for all the
candidate reference genes in the samples, and showed the overall stability in
gene expression. None of the candidate reference genes under study showed
a SD value higher than 1 indicating that all of the genes under study were
suitable to be considered for selection as reference genes.
However, further data processing using pairwise correlations and
regression analysis assessed the inter-gene relations and eliminated 18S, as
this gene had the highest variation (CV = 4.36) and least correlation (r =
0.434). The lowest variation was seen for the gene YWHAZ (CV = 0.96).
However, YWHAZ demonstrated only a weak correlation to BestKeeper
index as compared to other candidates (r = 0.653). Therefore, both 18S and
100
YWHAZ were excluded from further analysis. Subsequent elimination
singled out UBC and B2M as genes with low correlation with the
BestKeeper index. The analysis of the remaining four genes (HPRT1, ACTB,
GAPDH and RPII) showed a strong and significant correlation (0.914 < r <
0.960) between their expression levels and the BestKeeper index (p < 0.001).
To further assess the consistency and reliability of the BestKeeper
index, sample integrity of all four tightly correlated genes was investigated.
The InVar values of all samples were found to have low CP variation as well
as low x-fold expression regulation. Ranking the four tightly correlated
genes on the basis of variation from the most stable to the least stable was as
follows: GAPDH → HPRT1 → ACTB → RPII, and yielded the best genes
for defining a robust standardizing index. Amongst these four genes, ACTB
and RPII were the most highly correlated (0.975 < r < 0.981) to the
BestKeeper index (Table 3).
101
Table 3: BestKeeper correlation analysis (n=9).
In order to determine the influence of sample size on robustness of
this algorithm, additional 5 obese and 5 lean VAT samples (N = 19) were
included in the analysis. Further data processing using pairwise correlation
and regression analysis assessed the inter-gene relations and eliminated
UBC, as the gene with the highest variation (CV = 10.93) and least
102
correlation (r = 0.350). 18S gene showed similarly a high variation (CV =
12.99). Therefore, both UBC and 18S were excluded from further analysis.
Subsequent sequential elimination based on low coefficient of correlation
singled out B2M and HPRT1 as genes with low correlation with the
BestKeeper index. The analysis of the remaining four genes (YWHAZ,
GAPDH, ACTB and RPII) showed a strong and significant correlation (0.831
< r < 0.904) between their expression levels and the BestKeeper index (p <
0.001) (Table 4).
Table 4: BestKeeper correlation analysis (n=19).
103
Ranking the four tightly correlated genes from the most stable to the
least stable on the basis of their variation was as follows: YWHAZ →
GAPDH → ACTB → RPII, and yielded the best genes for defining a robust
standardizing index. Amongst these four genes, ACTB and RPII were the
most highly correlated (0.975 < r < 0.981) to the BestKeeper index (Table
2).
104
Discussion and Conclusion:
To eliminate non-biological variation, gene expression analysis
involving qRT-PCR requires a reliable normalization strategy. Among the
several approaches proposed, use of reference genes is currently the
preferred way of normalization (Jo Vandesompele et al., 2009). The use of
non-optimal reference genes is known to lead to erroneous results (Dheda et
al., 2005). The studies evaluating the expression levels of the reference
genes themselves, particularly, for GADPH and ACTB (Bustin, 2000),
showed considerable variation in different tissues and experimental
conditions. Specifically, in the omental and subcutaneous fat depots, a
variation in expression of these reference genes was found to be dependent
on the presence of obesity and T2D (Catalán et al., 2007). These findings
necessitate the need to validate reference genes for studies of human visceral
adipose samples.
105
Scientifically, the validation of reference genes presents a circular
problem: assessing stability of expression of a given gene cannot be
achieved without using another gene as a reference. Several algorithms have
been proposed to address this conundrum (Jo Vandesompele et al., 2009).
GeNorm software (Vandesompele et al., 2002) is one of the most popular
algorithms for validating candidate reference genes with low variability. It
utilizes two parameters to quantify the reference gene expression stability:
M (average expression stability) and V (pairwise variation). A low M value
is indicative of a more stable expression, hence, increasing the suitability of
a particular gene as a reference gene. Another feature of GeNorm is that it
does not require a normal distribution of data. However, co-regulation of
candidate genes does seem to influence the efficiency of this algorithm due
to the use of pair-wise comparisons. To minimize this risk, the eight
candidate reference genes selected for this analysis were chosen on the basis
of the difference in their physiological functions - cytoskeleton (ACTB),
carbohydrate
metabolism (GAPDH),
signaling
pathways
(YWHAZ),
transcription (RPII or POLR2A), metabolic salvaging of nucleotides
(HPRT1), protein synthesis (18S) and protein degradation (UBC).
106
The GeNorm algorithm determines expression stability (M) via a pairwise comparison of one candidate reference gene and all other candidate
genes independent of the level of gene expression for each sample. An
identical expression ratio of two reference genes in all samples is used as an
indicator of expression stability. Thus, GeNorm analysis is independent of
variation in amounts of starting material among the samples. According to
this analysis, ACTB and RPII represent the best combination of reference
genes for visceral adipose tissue among lean and obese patients (Figure 15a),
while HPRT1 and GAPDH were ranked third and fourth, respectively. After
completion of this step, a pairwise variation (V) was calculated between two
sequential normalization factors (NFn and NFn+1) for all genes. A large
variation indicates that the added gene has a significant effect and should be
included for calculation for a reliable normalization factor. Figure 16a, b
show that further inclusion of additional reference genes did not
significantly change the pair-wise variation and that the use of two reference
genes is sufficient for accurate normalization. An advantage of the GeNorm
algorithm is that it is minimally affected by expression intensity of the
candidate genes. In addition, since the approach is based on multiple pairwise comparisons, the need for larger sample size is mitigated. This finding
107
was reinforced by experimental observation that an increase in sample size
did not dramatically alter the final results. When sample size was increased
from 9 to 19, ACTB and RPII were again found to be the most stable genes
with GAPDH and HPRT1 being ranked the next two best genes (Figure
15b).
Housekeeping genes, in addition to their basic functions, exert
pleiotropic effects on other cellular systems, decreasing the value of the
function-based predictions of co-regulation. To overcome this problem
Anderson et al., proposed a model based approach incorporated into the
software NormFinder. This algorithm ranks candidate reference genes
according to the least estimated intra- and inter-group variation, which
serves as an effective method to overcome the influence of co-regulation.
Although NormFinder takes into account the heterogeneity in the tested
samples, and attempts to distinguish between stability and bias, this modelbased approach is self-restricted by the importance it places on overall
expression intensity of each candidate gene. A close inspection of the
analysis of the results produced by NormFinder showed that it is biased
towards candidate reference genes that have overall similar expression
108
values (in terms of Ct). Consequently, the robustness of this method is
linked to the sample size.
The ultimate objective of NormFinder is to identify candidate
reference genes(s) with an inter-group variation as close to zero as possible,
while at the same time having small intra-group variation. When the genes
were ranked solely by their stability values, GAPDH and HPRT1 appeared
to be the best combination of endogenous controls (Table 1). Further
examination of the results reveals that although UBC was ranked third, it had
a large intra-group variation (Figure 17a); therefore, UBC gene was
eliminated from further consideration, while the next most stable reference
genes, ACTB, B2M and RPII, remained under the scrutiny.
The same genes, namely GAPDH, HPRT1, ACTB and RPII, were
ranked as the most stable both by GeNorm and NormFinder softwares.
However, the best combinations of two genes proposed by these two
algorithms were different. This variation was expected based on the vastly
different approaches used by each of the analysis softwares and dependence
of robustness of NormFinder on sample size. In GeNorm, gene expression
109
stability (M) based on the expression ratio of the two genes (pairwise
comparison) is the most important criterion for evaluating a reference gene,
while NormFinder focuses on genes with the least intra- and inter- group
variations. Thus, in cases when two genes show high expression variation
while their ratio (M) remains unchanged, there will be discordance in
ranking by these two algorithms.
Further, NormFinder gains in robustness as the number of samples is
increased. This was confirmed by increasing sample size to 19. Ranking of
genes from most stable to least stable revealed ACTB and RPII as the best
combination of genes and this was in accordance with GeNorm results
(Figure 17b).
In contrast to the previous study by Catalan et al., in VAT samples
both the algorithms highlighted 18S as one of the least stable gene. This was
not unexpected, as several arguments against the use of rRNA as reference
genes have been previously put forth. The strongest argument against its use
in real time data analysis is its high abundance compared to other target
110
mRNA which hinders accurate subtraction from the baseline value
(Vandesompele et al., 2002).
In order to compare the GeNorm and NormFinder results with an
independent ranking method, the dataset was also analyzed with the
BestKeeper tool (Pfaffl et al., 2004). In this approach, ideal reference genes
are expected to have stable expression that is indicated by low variation in
the tissue under consideration (Pfaffl et al., 2004). With BestKeeper,
stability (SD) and relationship to the BestKeeper index (r and p values) are
the two criteria for evaluating the stability of reference genes. This algorithm
uses a pair-wise correlation analysis for all pairs of candidate genes based on
the raw Ct values and calculates the geometric mean of the best suited ones.
Based on low CV and high coefficient of correlation (r) to the BestKeeper
index, ACTB and RPII followed by GAPDH and HPRT1 were ranked as the
top four genes (Table 3). High correlation coefficient is an indicator of stable
expression of the reference genes in VAT. Again, 18S was ranked as the
least stable and excluded from further analysis. Robustness of the algorithm
was assessed by increasing sample size (n = 19). Sample size was found to
have minimal effect on the results. The same two genes ACTB and RPII
111
were identified as the most stable followed by YWHAZ and GAPDH (Table
4).
Overall, the BestKeeper results were in line with the NormFinder
data, and with minor differences, the GeNorm data, indicating the reliability
of the validation for reference genes in the present study (Table 2).
Regardless of the algorithm used, all the three software ranked the same set
of genes as the most stable. This also increased the confidence in the results
of the analyses.
In conclusion, we recommend ACTB and RPII as stable reference
genes most suitable for gene expression studies of human VAT. The use of
these genes as a reference pair may further enhance the robustness of qRTPCR-based expression analysis in this model system.
112
B. Validation of Reference Genes in Stomach
Background:
Until recently, majority of the molecular studies involving gastric
tissue have been related to stomach malignancies such as gastric carcinoma.
However, a number of studies point at the important endocrine role of the
stomach. In particular, the importance of hormones and receptors of stomach
in the regulation of both food intake and satiety was convincingly
demonstrated in various studies using rodent model systems (Ravussin and
Bogardus, 2000), (Gale et al., 2004), (Badman and Flier, 2005), (Murphy
and Bloom, 2006), (Yamada et al., 2008). Stomach has been shown to
express a number of inducible genes encoding for various soluble molecules,
for example, leptin (Bado et al., 1998), ghrelin (Badman and Flier, 2005),
113
peptide YY, glucagon, glucagon-like peptide-1, glucagon-like peptide-2
(GLP-2), oxyntomodulin (Murphy and Bloom, 2006) and other paracrine
factors that influence blood glucose levels, exocrine secretion, adipocyte
function, gastrointestinal motility and other physiological responses
(Murphy and Bloom, 2006). Being a major organ of the gastrointestinal
tract, stomach is well situated to influence energy homeostasis. Importantly,
the gastrointerstinal tract drains into the portal vein and thus, has direct
access to the liver.
Obesity is an underlying cause for a number of serious health
problems, including T2D, atherosclerosis, IR, NAFLD and polycystic ovary
syndrome (Yamada et al., 2008). The need to explore the role of gastric
tissue in energy homeostasis and pathogenesis of obesity and associated comorbidities is augmented by a surge in bariatric surgery procedures in the
past 5 year; the number of such procedures has reached 220,000 a year in
U.S. alone (Buchwald and Oien, 2009). These numbers are likely to increase
with recent evidence that bariatric surgery leads to both diabetes resolution
(Gill et al., 2010), (Nandagopal et al., 2011) and substantial improvement in
cardiovascular outcomes (Hofsø et al., 2010). In both cases, the connection
114
between surgically induced changes in molecular circuitry of gastric tissue
warrants further molecular studies.
Studies on common and complex disorders such as obesity and
gastritis require expression profiling of several genes, often performed
simultaneously (Logan et al., 2009). The high sensitivity, wide dynamic
range and high specificity of real-time PCR (qRT-PCR) make it both
suitable and popular technique for gene expression profiling (Logan et al.,
2009), (De Jonge et al., 2007). However, with accurate data normalization is
crucial for correct interpretation of qRT-PCR data (Vandesompele et al.,
2002), the use of reference genes as an internal control being the most
common approach for normalization (De Jonge et al., 2007). The suitability
of a particular reference gene depends on the tissue or cellular system being
investigated (Thellin et al., 1999), (Suzuki et al., 2000), (Schmittgen and
Zakrajsek, 2000).
Notably, the human gastric tissue is among the least investigated for a
stability of reference gene expression. So far, only one systematic study of
this kind was performed (Rho et al., 2010), wherein, six reference genes
115
were profiled in 6 stomach cancer cell lines and 20 pairs of normal and
tumor stomach tissue samples taken from the same individual. However
important to the field, this study included only malignant tissue samples or
adjacent tissue samples likely affected by aberrant DNA methylation
previously demonstrated within “an epigenetic field for cancerization”
commonly found around gastric carcinomas (Ushijima, 2007).
To validate candidate reference genes suitable for qRT-PCR profiling
experiments using gastric tissue from morbidly obese individuals with
various kinds of accompanying non-malignant gastric diseases, GeNorm
(Vandesompele et al., 2002), NormFinder (Andersen et al., 2004) and
BestKeeper (Pfaffl et al., 2004) software packages were employed. The
reference genes top ranked by all three software were then recommended for
further use as endogenous controls in qRT-PCR experiments profiling nonmalignant stomach tissues of human origin.
116
Results:
In this study expression of six reference genes most commonly used
in qRT-PCR studies of human tissues (Baranova et al., 2003), (Carl Wittwer
et al., 2004), (Mehta et al., 2010) namely 18S , GAPDH, B2M, HPRT1,
YWHAZ, UBC and RPII were profiled. The expression stability of these
genes was evaluated in 23 fundic gastric tissue samples. For each tissue
sample, expression stability of each gene was calculated using the mean of
triplicate Ct values. The input data for GeNorm and NormFinder analysis
required conversion of raw Ct values to relative quantities using the
comparative Ct method, while the BestKeeper algorithm input required raw
Ct values (Jo Vandesompele et al., 2009).
GeNorm Analysis:
The software GeNorm, which ranks candidate reference genes
according to their relative quantities, was developed by Vandesompele et al.
(Vandesompele et al., 2002). This software allows sorting of genes ranked
117
on the basis of their expression stability from the least stable to most stable.
Using this software, the following rankings were observed: 18S → UBC →
RPII → YWHAZ → GAPDH and HPRT1. The M values determined for
these genes in gastric tissue were 4.90, 4.60, 4.30, 4.09 and 3.9, respectively
(Figure 18). HPRT1 and GAPDH were highlighted as the two most stable
reference genes by sequential elimination of the least stable genes based on
their M values. Based on pairwise variation calculated between two
sequential normalization factors (NFn and NFn + 1), GeNorm determined
that the simultaneous use of these two reference genes is enough for the
studies in this particular type of tissue (Figure. 19).
118
Figure 18: Gene expression stability M of candidate reference genes in
gastric tissue calculated by GeNorm software. The program proceeds with
stepwise exclusion of genes with relatively higher variable expression
among the samples. The expression stability measure (M) is the average of
the stability values of the remaining genes. The lower the M, the more stable
the gene in the subset (n = 23).
119
Figure 19: Determination of optimal number of reference genes for
normalization by pairwise variation analysis using GeNorm software.
Bar values indicate the magnitude of the change in normalization factor after
the inclusion of an additional reference gene. A large variation indicates that
the added gene has a significant effect and should probably be included for
calculation of the normalization factor (n=23).
120
NormFinder Analysis:
In contrast to GeNorm, the NormFinder software employs a modelbased evaluation strategy (Andersen et al., 2004). This software ranks
candidate genes based on their minimal inter and intra-group variations. In
other words, in addition to the stability of genes within the sample sets, this
software also evaluates the stability of genes across the groups or cohorts of
samples. In this study, the tissue samples (n = 23) were divided into cohorts
of gastritis (n = 11) and non-gastritis (n = 12). GAPDH and HPRT1 with a
stability value of 0.356 were the top two genes ranked on the basis of
combined least inter and intra-group variation (Table 5). Further, manual
inspection for genes with lowest inter- and intra-group variations showed
that, while 18S had the second lowest inter-group variation after GAPDH, it
had the highest intra-group variation and was hence found to be the least
stable (Figure 20). HPRT1, on the other hand had an inter-group variation
121
greater than 18S, but it had the lowest intra-group variation. Hence, HPRT1
was considered stable by the NormFinder algorithm (Figure 20).
Thus, the results produced by NormFinder were found to be in
agreement with GeNorm analysis.
Table 5: Comparison of highly ranked genes by all three software
(n=23).
122
123
Figure 20. Determination of the most stable reference genes using
NormFinder. Two groups considered were - gastritis obese and non-gastritis
obese patient tissues (n = 23). Bars represent inter-group variances, while
error bars representing the average of intra-group variance. Ideal reference
gene has inter-group variation as close to zero as possible and as small error
bars as possible.
124
BestKeeper Analysis:
The BestKeeper algorithm was selected as an independent ranking
method allowing indirect comparison of GeNorm and NormFinder results
(Pfaffl et al., 2004). Unlike GeNorm and NormFinder, input data for analysis
by BestKeeper was raw Ct values of each gene. According to this algorithm,
stable reference genes are selected using two main criteria: low variation
(SD) and high correlation (r and p values) in the tissue under consideration
(Pfaffl et al., 2004).
Analysis of the calculated variations (SD (± Ct) and CV (%Ct)) for all
the candidate reference genes in all 23 samples allowed ranking of the
relative stability in gene expression. Among the 6 putative reference genes,
18S (CV = 22.73) and UBC (CV = 8.42) had the highest variation and were
eliminated. The next step involved the estimation of inter-gene relations of
all possible reference gene pairs on the basis of the Pearson correlation
coefficient (r) and significance (p) values. Among the remaining four genes,
125
HPRT1 and GAPDH expression levels were found to be highly correlated
(0.506 < r < 0.807) to the BestKeeper index with a high significance (p <
0.02) (Table 6). BestKeeper index is used to describe the relation between
each of the contributing candidate reference genes and the index based on
Pearson correlation coefficient (r), coefficient of determination (r 2), and
probability (p) value.
Table 6: BestKeeper correlation analysis (n=23).
126
Thus, the results from all the three algorithms were in agreement. For
studies on non-malignant gastric tissue samples collected from obese
patients, HPRT1 and GAPDH genes were identified as the most stable
reference gene pair, while 18S was found to be the least stable reference
gene (Table 5).
Discussion and Conclusion:
The differences in the levels of the gene expression measured by qRTPCR include both true biological variation and technical variation which are
usually abundant due to the variance in the amount of starting material and
presence
of
inhibitors
that
influence
enzymatic
efficiencies
(Jo
Vandesompele et al., 2009). To account for the influence of non-biological
variation on the results of experiment, qPCR data needs to be normalized to
some internal or external control. Currently, the preferred way of
normalization is the use of reference genes (Jo Vandesompele et al., 2009).
127
The drawback of this approach is a requirement of the validated stable
expression of the reference genes for each experimental condition under
study. However, this approach by itself is a circular problem, due to absence
of any available external standard that allows one to evaluate the stability of
the reference gene (Jo Vandesompele et al., 2009). Three algorithms,
GeNorm, NormFinder and BestKeeper, were developed to address this
challenge (Jo Vandesompele et al., 2009). Each of these popular algorithms
utilizes different methods to arrive at the ideal reference gene or their
combination (Jo Vandesompele et al., 2009). First of these algorithms,
implemented in GeNorm software (Vandesompele et al., 2002) quantifies the
reference gene expression stability utilizing two parameters: Average
expression stability (M) and pairwise variation (V). When M value is lower,
expression is more stable and a suitability of a particular gene as a reference
gene increases. In this algorithm, expression stability (M) is determined
through a pairwise comparison of one candidate reference gene to all other
candidate genes with no regard to the level of gene expression for each
sample. When two reference genes demonstrate an identical expression ratio,
a conclusion of relative expression stability is made. Thus, GeNorm analysis
is independent of variation in amount of starting material between samples
128
and does not require a normal distribution of data. The results of GeNorm
algorithm are minimally affected by expression intensity of the candidate
genes providing for another advantage of this software. In addition, since the
approach is based on multiple pairwise comparisons, the need for large
sample size is mitigated (Carl Wittwer et al., 2004), (Mehta et al., 2010).
However, co-regulation of candidate genes influences results of the GeNorm
analysis; therefore, it is essential to use candidate reference genes differing
in their physiological function. GeNorm analysis of the six reference genes
highlighted GAPDH and HPRT1 as the best combination of reference genes
for the non-malignant samples of stomach tissue (Figure 18), while 18S and
UBC were found to have the least stability. In order to determine the ideal
number of reference genes, for two sequential normalization factors (NFn
and NFn + 1), pairwise variation (V) calculations were carried out. This
calculation showed that adding more reference genes to the list did not
increase the pair-wise variation significantly (V3 = 1.05) and that the use of
two reference genes is sufficient for accurate normalization in given set of
tissue samples (Figure 19). NormFinder algorithm deals with co-regulated
reference genes, thus, overcoming the limitation of GeNorm. NormFinder
software implements a model based approach that ranks candidate reference
129
genes according to their least estimated intra and inter group variation.
Indeed, NormFinder takes into account the heterogeneity in the tested
samples, but it places importance on the overall expression intensity of each
candidate gene, therefore, being biased towards genes that have overall
similar expression values (in terms of Ct). Consequently, the robustness of
this method and the sample size are linked together. According to
NormFinder algorithm, an ideal reference gene(s) has inter- and intra-group
variations as close to zero as possible. When the genes were ranked solely
by their stability values, GAPDH and HPRT1 were highlighted as the best
combination of endogenous controls (Table 5). Interestingly, 18S had the
lowest inter-group variation after GAPDH; HPRT1 was ranked next after
18S (Figure 20) and was more stable. However, among 18S and HPRT1, the
high intra-group variation of 18S compared to HPRT1 (Figure 20) resulted in
18S being considered to be the least stable and, therefore, eliminated from
further consideration. It is of note that both tested algorithms highlighted
18S as one of the least stable gene. This was not unexpected, as several
arguments against the use of rRNA as reference genes have been previously
discussed, including difficulty with accurate subtraction of 18S expression
levels from the baseline values due to its high abundance (Vandesompele et
130
al., 2002). Importantly, the same pair of genes, HPRT1 and GAPDH, was
ranked as the most stable genes by both GeNorm and NormFinder softwares.
In order to compare the GeNorm and NormFinder results with an
independent ranking method, the data was analyzed with another
algorithm—BestKeeper (Pfaffl et al., 2004). This algorithm determines the
expression stability of the reference genes based both on stability of
expression and relationship to the BestKeeper index (r and p values). Based
on low CV and high coefficient of correlation (r) to the BestKeeper index,
GAPDH and HPRT1 were ranked as the most stable genes (Table 5). Again,
18S was ranked as the least stable and was excluded from further analysis.
In this study, the results of the three algorithms were in agreement
with each other. In a recent study by Rho et al. (Rho et al., 2010) on a mixed
set of malignant and adjacent non-malignant tissue, substantially different
set of reference genes was highlighted (Rho et al., 2010). With GeNorm,
HPRT1 and RPL29 pair were considered as the most stable, while 18S and
GAPDH were among the least stable genes. NormFinder analysis on the
other hand found B2M and RPL29 to be the best combination of genes.
Marked difference in the stability of GAPDH expression in study of Rho et
131
al. and in our experiment could be attributed to tumor-specific increase in
GAPDH expression that has been previously demonstrated for other
malignant tissues (Cicinnati et al., 2008) to the degree that it was proposed
as biomarker in certain cancers (Tokunaga et al., 1987). On the other hand, a
number of studies confirmed the reliability of GAPDH as a reference gene in
non-malignant human cells and tissues (Zainuddin et al., 2010), indicating
that the discrepancy between two studies on gastric samples is not an artifact
but a reflection of differences in the nature of samples studied. However, it
remains difficult to say with certainty, whether all uncovered differences in
gene expression stabilities are due to different sets of genes used, or to
inherent difference in malignant and non-malignant gastric samples. The
current study on the other hand, involves non-malignant gastric tissue and
same set of reference genes were found to be the most stable using 3
different algorithms, which allow us to recommend HPRT1/GAPDH pair as
the best reference set of genes for qRT-PCR studies in non-malignant
conditions of the stomach. However, it is ideal to use different sets of
reference genes for malignant and non-malignant conditions of stomach
lining.
132
In conclusion, we recommend GAPDH and HPRT1 as stable reference
genes most suitable for gene expression studies of non-malignant human
gastric tissue. The use of these genes as a reference pair may further enhance
the robustness of qRT-PCR in this model system.
133
C.Validation of reference genes in HepG2
Background:
In the 21st century, the growing incidence of obesity is accompanied
with increased risk for NAFLD (Fabbrini et al., 2010), (Vernon et al., 2011).
The worldwide prevalence of NAFLD is estimated to range from 6% to
35%, with over 30% prevalence in US (Vernon et al., 2011). Pathologically,
NAFLD is a spectrum of disorders ranging from steatosis (fatty liver) to
NASH which may progress to fibrosis and cirrhosis (Matteoni et al., 1999).
It is of general consensus that steatosis has a slow progression, while NASH
can exhibit histological progression and can develop into fibrosis (37% 41%) and cirrhosis (1% - 5%) (Adams et al., 2005), (Ekstedt et al., 2006).
The progression of NAFLD is thought to be dependent on multiple parallel
134
hits involving insulin resistance, FFA, oxidative stress, apoptosis, and
circulating cytokines (Kim and Younossi, 2008). Although, little is known
about the factors determining the progression of NAFLD, emerging data
suggest that hepatocyte apoptosis and altered cytokine profile may be
important players in NAFLD progression (Feldstein et al., 2003), (Ramalho
et al., 2006), (Wieckowska et al., 2006).
The liver has multiple phenotypically distinct cell types such as
kupffer cells, cholangiocytes, stellate cells and hepatocytes, with hepatocytes
being the most prominent cells (Tacke et al., 2009), (Malhi et al., 2010).
Hepatocytes can undergo programmed cell death via the extrinsic as well as
the intrinsic pathways. Among these, however, extrinsic signals predominate
in pathological inflammation (Canbay Ali et al., 2005). The extrinsic
pathway is initiated by ligand binding to the death receptors, which then
recruit adaptor proteins. These adaptor proteins eventually activate caspases
resulting in apoptosis (Peter and Krammer, 1998), (Rust and Gores, 2000),
(Delhalle et al., 2003). Among the death receptors (DRs) specifically
expressed in the liver, the most important and well characterized receptor is
Fas/CD95 (Faubion and Gores, 1999), (Rust and Gores, 2000). Fas/CD95 is
135
expressed by hepatocytes, activated stellate cells and kupffer cells (Faubion
and Gores, 1999) but hepatocytes have the highest constitutive expression of
Fas/CD95. Apoptosis mediated by Fas/CD95 has been shown to be
particularly prominent in the liver (Galle et al., 1995). Intriguingly, besides
inducing apoptosis, Fas/CD95 is also capable of initiating inflammatory
signaling cascades, such as stimulating the synthesis of pro-inflammatory
mediators in various tissues and cells (Faouzi et al., 2001), (Ribeiro et al.,
2004), (Ponton et al., 1996), (Park et al., 2003), (Ma et al., 2004). Studies
have implicated nuclear factor-κB (NF-κB) in the inflammatory signaling of
Fas/CD95 (Figure 21) (Guicciardi and Gores, 2009). However, the role of
NF-κb in Fas mediated apoptosis is debated. For instance, Kühnel et al.,
demonstrated the NF-κB involvement in Fas-mediated apoptosis of
hepatocytes (Kühnel et al., 2000). Contrary to this, Ravi et al demonstrated
that inhibition of NF-κB by Fas ligation is essential for apoptosis in dendritic
cells (Ravi et al., 1998).
NF-κB is a transcription factor with a number of inflammatory
cytokine and chemokine targets genes (Pahl, 1999). Importantly it is an
essential survival factor in several physiological conditions such as
136
embryonal liver development and liver regeneration. It is also a main
mediator of the cellular response to a variety of extracellular stress stimuli
(Kühnel et al., 2000). Thus, it is highly plausible that under conditions of
altered cytokine profile NF-κB is activated. NF-κB may then contribute to
inflammation and/or apoptosis by promoting expression of target cytokines,
chemokines and FasR. In this study, we attempt to determine the effects of
the 2 chemokines CCL21 and CCL4 respectively on NF-κB mediated
signaling using Hep G2 in vitro system. Hep G2 (human liver carcinoma cell
line) is a perpetual cell line which serves as a suitable model system for in
vitro study of polarized hepatocytes.
137
Figure 21: Schematic representation of Fas mediated cellular apoptosis
and role of anti-apoptotic proteins. Fas ligand binding to the Fas death
receptor activates receptors, which then recruits adaptor protein FADD. This
adaptor protein often via NF-κB activates caspases resulting in apoptosis.
Anti-apoptotic proteins like BIRC bind and inhibit caspases, thus regulating
apoptosis. It is possible that presence of chemokines may activate NF-κB
mediated apoptosis and/or inflammatory gene expression.
qPCR is a powerful and efficient means of rapidly comparing patterns
of gene expression between different experimental conditions. However, the
first step in the analysis of differential RNA expression of target genes by
qPCR is identification of optimal reference genes for subsequent
normalization (Vandesompele et al., 2002). As discussed previously, several
138
factors including variation in amount of starting material such as well to well
variability notorious in cell based assays often leads to quantification errors.
An ideal reference gene is one which is consistently expressed at the
same level in all samples under investigation regardless of the experimental
conditions and exhibits expression levels comparable to that of the target
gene (Suzuki et al., 2000). 18S, ACTB, GAPDH, B2M, RPII, YWHAZ, UBC
and HPRT1 are some of the most commonly used reference genes in RTPCR studies (Thellin et al., 1999), (Catalán et al., 2007), (De Jonge et al.,
2007). However, numerous studies have shown that expression of these
common reference genes vary with sample type as well as experimental
condition (Thellin et al., 1999), (Bustin, 2000). The suitability of a particular
reference gene thus, depends on the system being investigated and the
inherent experimental conditions (Thellin et al., 1999), (Schmittgen and
Zakrajsek, 2000), (Suzuki et al., 2000). In light of these findings, it is crucial
to validate the stability of the reference genes across the experimental
conditions being tested.
139
Increasing concerns about normalization using ideal reference genes
has led to the development of several mathematical algorithms aimed at
determining the stability of reference genes’ expression (Jo Vandesompele et
al., 2009) with GeNorm (Vandesompele et al., 2002), BestKeeper (Pfaffl et
al., 2004) and NormFinder (Andersen et al., 2004) being highly cited.
Reference gene validation studies discussed previously have shown
similar performance of these algorithms in ranking the most stable gene (s)
(Mehta et al., 2010), (Birerdinc et al., 2012). Among the three most cited
softwares, NormFinder offers the added ability to validate reference gene
across several experimental groups. Thus, for this study prior to evaluating
the role of CCL21 and CCL4 respectively using HepG2 cells, NormFinder
(Andersen et al., 2004) was used to determine the most stable reference gene
(s).
Results
140
In this study the expression stability of the eight reference genes most
commonly used in qRT-PCR studies was evaluated in HepG2 cells across 5
different time points and 2 different chemokine (CCL21, CCL4)
concentrations (30ng/ul, 100ng/ul). The samples were grouped based on type
of chemokine treatment, concentration of chemokine as well as time points.
The basis for dividing the samples into these groups was that each of
the factors (chemokine, concentration and time points) was expected to alter
the reference gene expression and was thus, considered as a grouping
variable. Thus, for every chemokine there were 10 groups (5 time points * 2
concentrations). For the analysis, three biological replicates were used and
the expression stability of each gene was calculated using the mean of
technical replicates (n=3). The input data for analysis required conversion
of raw Ct values to relative quantities using the comparative Ct method
(Andersen et al., 2004), (Jo Vandesompele et al., 2009).
The algorithm performs stability test by comparing genes in all
possible pair-wise combinations among the different groups. Based on interand intra- group variability, a stability value (ρ) is generated (Table 7).
141
Analysis shows UBC has the lowest stability value (0.549) and was thus
considered the most stable. UBC is closely followed by B2M (0.675) and
GAPDH (0.857). Notably, the classic housekeeping genes ACTB (1.760) and
RPII (1.900) were among the least stable genes under these experimental
conditions. HPRT1 was found to have the most variable expression with
stability value of 2.046 (Table 7).
Table 7: Comparison of reference gene stability ranked based on intraand inter-group variation by NormFinder.
142
143
Further, manual inspection for genes with the lowest intra-group
variations (Figure 22) showed that UBC (0.519) and B2M (0.656) had the
lowest intra-group variability as indicated by error bars. These genes also
had the lowest inter-group variability (Figure 22). GAPDH (2.68) was
ranked third based on relatively larger intra-group variability (Figure 22).
HPRT1 (20.48), RPII (18.0) and ACTB (10.42), on the other hand had the
highest intra-group variability (Figure 22). Thus, based on both intra- and
inter-group variability UBC and B2M were found to be the best two
reference genes (Figure 22) for this study, while UBC was the single most
reference gene with stability value of 0.549 (Table 7).
144
Figure 22: Determination of the most stable reference genes using
NormFinder. Bars represent inter-group variances, while error bars
represent the average of intra-group variance. Ideal reference gene is
supposed to have inter-group variation as close to zero as possible and error
bars as small as possible.
145
Methods:
Tissue Sample Collection:
The proposed study was conducted in collaboration with the
Translational Research Institute (TRI) of INOVA Health System (Falls
Church, VA). Human VAT was collected during bariatric surgery or other
intra-abdominal surgery. Human gastric tissue was collected during bariatric
surgery. The samples were flash frozen in liquid nitrogen and stored at -80
0
C. The samples were de-identified in compliance with HIPAA regulations
and a repository of specimens was assembled and made accessible for the
proposed study. Additionally, the clinical data associated with the samples
in the repository was made available.
HepG2 Cell culture:
146
HepG2 cells (human liver carcinoma cell line-ATCC No.HB-8065)
were cultured in DMEM/F-12 with L-glutamine (manufacturer) in 75-cm2
culture flasks (Corning, USA). Cells were plated in replicates at
recommended density, 36 h before chemokine treatment.
For gene expression studies, cells were seeded in 24 well plates
(Corning, USA) in density ~ 4.5 x 10 5 cells/well. Cells were treated
independently with 100ng/ul and 30 ng/ul of chemokines CCL21 and CCL4
(Prospec, USA). Chemokines were added at 2 hrs, 4 hrs, 6 hrs and 8 hrs and
controls were treated with media lacking chemokines. At the end of 8 hours,
the cells were subjected to total RNA extraction as described below.
RNA Extraction from Tissues:
Total RNA was extracted from VAT and gastric tissue using mirVana
kit (Ambion). The first step of the mirVana miRNA Isolation Kit procedure
was to disrupt samples in a denaturing lysis buffer to disrupt the tissue and
the cells and release contents into solution. The lysis buffer contains
guanidinium thiocyanate, which immediately inactivates RNases. Next,
samples were subjected to acid-Phenol:Chloroform extraction which
147
provides a robust front-end purification that removes majority of the
genomic DNA along with proteins and lipids. After organic extraction, the
final purification was accomplished via a glass-fiber filter. Ethanol was
added to samples, and they were passed through a filter cartridge containing
a glass-fiber filter which immobilizes the RNA. The filter was then washed a
few times, and finally the RNA was eluted in sterile water.
For total RNA extraction, 100 mg of human tissue samples were
excised while in liquid nitrogen and added to a tube containing 10 volumes
of Lysis/Binding Buffer per tissue mass (~1 mL). Using a homogenizer
(Power Gen 125) the tissue was immediately homogenized. Excision was
performed in liquid nitrogen to ensure minimal thawing of tissue, which
would otherwise compromise the integrity of RNA and sample. Following
homogenization, 1/10 volume of miRNA Homogenate Additive was added
to each tube. After inverting several times to ensure uniform mixing, the
tube was incubated on ice for 10 min. At the end of incubation period, equal
volume of acid-Phenol:Chloroform was added to the tube, taking care to add
the bottom phase of acid-Phenol:Chloroform. The upper phase is the
aqueous buffer. The mixture was then vortexed for 30–60 sec and the
148
samples centrifuged (Eppendorf, USA) for 5 min at maximum speed (10,000
x g) at room temperature. This step ensures separation of the aqueous phase
from the organic phase. This process ensured that DNA and proteins
partition into the interphase and lower organic phase, while RNA partitions
into upper aqueous phase. Aqueous (upper) phase was then carefully
transferred to a new tube without disturbing the lower phase.
To the aqueous phase, 1.25 volumes of 100% ethanol at room
temperature was added and mixed thoroughly by pipetting. The
lysate/ethanol mixture was then passed through a glass fiber filter cartridge
by adding 700 μL of mixture on the cartridge placed on a fresh tube. For
samples with larger than 700 μL volume the mixture was applied in
successive applications to the same filter. After each addition, tube was
centrifuged (Eppendorf, USA) for ~15 sec at RCF 10,000 x g (typically
10,000 rpm) to pass the mixture through the filter. Flow-through was
discarded after each addition and process repeated until all of the
lysate/ethanol mixture was passed through the filter. miRNA Wash Solution
1 (700 μL) (working solution mixed with ethanol) was applied to the filter
cartridge and centrifuged (Eppendorf, USA) for ~5–10 sec, with the flow –
149
through being discarded. Wash Solution 2/3 (500 μL) (working solution
mixed with ethanol) was then added to the filter and again centrifuged
(Eppendorf, USA) for 5- 10 sec. This step was repeated with a second 500
μL aliquot of Wash Solution 2/3, discarding flow-through each time. After
discarding the flow-through from the last wash, the filter cartridge was
replaced in the same collection tube and the assembly spun for 1 min to
remove residual fluid from the filter. This ensures RNA yields with high
purity, free from all the salts and ethanol which can otherwise interfere with
downstream analysis procedures.
To elute total purified RNA, filter cartridge was placed on a fresh
collection tube. Preheated RNase free water (30 μL) was applied to the
center of the filter. After incubation at room temperature for 5 min, the filter
was spun down for ~20–30 sec at maximum speed to recover the RNA. This
step was repeated with 20 μL of preheated RNase free water to obtain a total
of 50 μL of total RNA. RNA quality and quantity assessment was carried
out as described below.
RNA extraction HepG2 Cells:
150
After appropriate chemokine stimulation as described above total
RNA was extracted using Qiagen RNeasy kit (Qiagen, USA) according to
manufacturer’s protocol. The RNeasy technology combines the selective
binding properties of a silica-based membrane with the speed of microspin
technology. A specialized high-salt buffer system allows up to 100 μg of
RNA longer than 200 bases to bind to the RNeasy silica membrane.
Biological samples are first lysed and homogenized in the presence of a
highly
denaturing
guanidine-thiocyanate–containing
buffer,
which
immediately inactivates RNases to ensure purification of intact RNA.
Ethanol is added to provide appropriate binding conditions, and the sample
is then applied to an RNeasy Mini spin column, where the total RNA binds
to the membrane and contaminants are efficiently washed away. Highquality RNA is then eluted.
To extract total RNA from treated cells, cells were grown in 96 well
plates and were lysed at the end of the treatment duration by direct lysis
using RLT buffer. The RLT contained beta-mercaptoethanol (β-ME) in a
ratio of 10 μl β-ME per 1 ml of RLT Buffer, to ensure inactivation of
cellular RNases. For the lysis, entire volume of cell culture medium was
151
removed and cells were washed twice in 200 uL of PBS, followed by
addition of 350 uL of RLT buffer. Incomplete removal of cell-culture
medium would interfere with lysis and dilute the lysate, thus affecting the
downstream conditions for binding of RNA to the RNeasy membrane. To
lyse and homogenize the cells, the mixture was mixed by pipetting
repeatedly using 1mL micropipette tip. The lysate was then transferred to a
1.5 mL microcentrifuge tube. To this lysate, 1 volume of freshly prepared
70% ethanol was added and mixed by pipetting. The mixture (700 uL) was
then transferred to the RNeasy spin column and centrifuged (Eppendorf,
USA) at ≥10,000 rpm for 15 sec. The flow-through was discarded and the
process was repeated for the remaining lysate mixture. The filter was then
washed by addition of 700 uL RW1 buffer, followed by centrifugation at
≥10,000 rpm for 15 sec. The filter was then washed with 500 uL of RPE
buffer, twice, each time centrifuging (Eppendorf, USA) at ≥10,000 rpm for
15 sec and ≥10,000 rpm for 2 min respectively. The flow-through was
discarded each time during the washing steps.
To elute total purified RNA, the RNAeasy column was placed on a
fresh collection tube. Preheated RNase free water (30 μL) was applied to the
152
center of the filter. After incubation at room temperature for 5 min, the filter
was spun down for ~20–30 sec at maximum speed to recover the RNA. This
step was repeated with 20 μL of preheated RNase free water to obtain a total
of 50 μL of total RNA. The quality of RNA was assessed and 1 ug of total
RNA was used for cDNA synthesis using RT2 first strand kit
(SABiosciences, USA) as discussed below.
Assessing RNA quality and quantity:
To determine the quantity and purity of the extracted RNA,
absorbance at 260 nm (A260) and 280 nm (A280) was measured using
GeneQuant 1300 spectrophotometer (GE, USA). A260/A280 ratio between
1.8 - 2.1 indicated high quality RNA, while A260 determined the yield of
total RNA.
In order to assess the integrity of RNA, 1% agarose gel
electrophoresis was carried out. 400 mg of agarose was mixed with 40 mL
of 1X TAE (Fisher Scientific) and heated to dissolve agarose in the buffer.
After the solution cooled to 50 0C, 4 μL of ethidium bromide (10mg/ml) was
added and gently swirled for uniform distribution. The solution was then
poured in a casting tray ensuring no bubbles are formed and a desired comb
placed. The gel was then allowed to solidify at room temperature for 30 min.
153
After loading the RNA samples in equal concentrations (500 ng- 250 ng)
with 1X loading dye and in the presence of a 1 kb ladder, gel electrophoresis
was carried out at 100 V for 1 hour in 1X TAE. After separation, the gel was
visualized under UV (Bio-Rad) and results were documented.
Intact total RNA was expected to have sharp, clear 28S and 18S
rRNA bands. The 28S rRNA band would be approximately twice as intense
as the 18S rRNA band. This 2:1 ratio (28S:18S) serves as a good indicator of
integrity of RNA. Partially degraded RNA will appear as a smear and lack
the sharp rRNA bands, or will not exhibit the 2:1 ratio of high quality RNA.
Completely degraded RNA will appear as a very low molecular weight
smear.
First Strand cDNA Synthesis:
After assessing the quality and quantity of total RNA, cDNA
synthesis (RT² First Strand Kit (Qiagen, USA) was carried out on the same
day of total RNA extraction to prevent loss of RNA during storage. Random
hexamers and oligo-dT primed the reverse transcription reactions and a
reverse transcriptase synthesized the cDNA product.
154
The first step in cDNA synthesis was the genomic DNA elimination
step. This was done by addition of genomic DNA elimination mixture to
total RNA isolated from tissues (VAT, gastric tissue) and Hep G2 cells (560
ng and 1 ug of total RNA, respectively) followed by incubation at 42 0C for
5 min. The mixture was then placed on ice immediately for at least one
minute.
1. Genomic DNA Elimination Mixture:
For each sample, the following were combined in a sterile PCR tube:
Total RNA
25.0 ng to 5.0 ug
GE (5X gDNA Elimination Buffer)
2.0 μL
RNase free H2O to a final volume of
10 μL
The next step was reverse transcription using random hexamers and
oligo-dT primers. This was done by addition of RT cocktail to the above
mixture, followed by incubation at 42 0C for exactly 15 min in a PCR
machine (Bio-Rad). The reaction was then immediately stopped by heating
155
at 95 0C for 5 minutes. Sterile water (91 µL) was then added to each 20 µL
of cDNA synthesis reaction and mixed well. The cDNA was then stored at 20 0C until further use.
2. RT Cocktail:
RT Cocktail
1 reaction
BC3 (5X RT Buffer 3)
4 µL
P2 (Primer and External Control Mix)
1 µL
RE3 (RT Enzyme Mix 3)
2 µL
RNase free H2O
3 µL
Final Volume
10 µL
Reference Gene Selection and Primer Design and Primer Sequences:
Candidate reference genes previously reported as housekeeping genes
were selected from literature as follows: 18S, B-Actin (ACTB),
glyceraldehyde-3-phosphate
dehydrogenase
(GAPDH),
beta-2-
microglobulin (B2M), hypoxanthine guanine phosphoribosyl transferase1
156
(HPRT1),
tyrosine
3-monooxygensae/tryptophan
5-
monooxygenase
activation protein, zeta polypeptide (YWHAZ), ubiquitin C (UBC) and RNA
polymerase II (RPII, or POLR2A).
ACTB, GAPDH, B2M and HPRT1 were obtained from Real Time
Primers (Real Time Primers, USA). Primers for the remaining genes, RPII,
YWHAZ and UBC, were custom designed to span exon-exon boundaries
using Oligo Perfect Designer software (Invitrogen, USA) and synthesized
commercially (Invitrogen, USA). All primers were confirmed using the
NCBI Blast tool (Ye et al., 2012) against all available mRNA sequences to
ensure specificity. Gene accession numbers as well as primer sequences are
listed in Table 8.
Quantitative real-time PCR Protocol:
Quantitative real-time PCR was performed in a 96 well format in the
Bio-rad CFX96 Real Time System (Bio-Rad Laboratories, USA). The realtime PCR mixtures consisted of 5 μL cDNA corresponding to ~560 ng total
RNA for tissues or 1 μL cDNA corresponding to 1 ug total RNA from Hep
G2 cells, 0.1 uM of Real-Time primers or 0.2 nM of Invitrogen primer and
157
1× Sso Fast Evagreen Supermix ( Bio-Rad, USA ) in a final volume of 15
μL.
The assay included “no template” and “RT minus” controls to detect
reagent contamination and presence of genomic DNA respectively. The
thermal profile of the RT-PCR procedure was repeated for 50 cycles : 1) 95
0
C for 10 min; 2)10s denaturation at 950C, 40 s annealing at 55 0C for Real
Time primers and 60 0C for Invitrogen primers (amplification data collected
at the end of each amplification step); 3) dissociation curve consisting of 10
s incubation at 95 0C, 5 s incubation at 65 0C, a ramp up to 95 0C (Bio-rad
CFX96 Real Time System, USA). Melting curves were used to validate
product specificity. All samples were amplified in triplicates from the same
total RNA preparation and the mean value was used for further analysis. To
further ensure specificity of primers the products of the qPCR were run on
agarose gels to confirm single product amplification.
158
Table 8: Primer sequences of reference genes.
Target
Gene Accession
Gene
Number
B2M
NM_004048.2
Tm
55 0C
Primer Sequence
F-5’GTGCTCGCGCTACTCTCTCT
R-5’TCAATGTCGGATGGATGAAA
GAPDH
NM_002046.2
55 0C
F-5’ACAGTCAGCCGCATCTTCTT
R-5’GACAAGCTTCCCGTTCTCAG
ACTB
NM_001101.2
55 0C
F-5’CTCTTCCAGCCTTCCTTCCT
159
F-5’AGCACTGTGTTGGCGTACAG
HPRT1
NM_000194.1
55 0C
F-5’AAGCTTGCTGGTGAAAAGGA
R-5’AAGCAGATGGCCACAGAACT
YWHAZ
NM_001135702.1
55 0C
F-5’ACTTTTGGTACATTGTGGC
R-5’CCGCCAGGACAAACCAGT
18S
NR_003286.2
60 0C
F-5’AGGAATTCCCAGTAAGTGCG
R-5’GCCTCACTAAACCATCCAA
RP II
NM_000937.3
60 0C
F-5’CTTCACGGTGCTGGGCATT
160
R-5’GTGCGGCTGCTTCCATAA
UBC
XM_002344708.1
60 0C
F-5’CCTGGTGCTCCGTCTTAGAG
R-5’TTTCCCAGCAAAGATCAACC
161
Determination of reference gene expression stability:
To assess the stability of expression of candidate reference genes
across all samples three different statistical algorithms - GeNorm, v3.4,
NormFinder v0.953 and BestKeeper v1, were used according to developer's
recommendations.
GeNorm Analysis:
162
To determine the stability of a candidate reference gene among a
panel of reference genes, a Visual Basic Application (VBA) for Microsoft
Excel - termed GeNorm was written by Vandesompele et al.,
(Vandesompele et al., 2002). The application automatically calculates the
gene-stability measure (M) for all candidate genes in a given set of samples.
The gene-stability measure (M) is defined as the “average pairwise variation
of a particular candidate gene with all other candidate genes” (Equation 3).
Genes with the lowest M values have the lowest variation across the samples
and hence the most stable expression.
163
For M value determination, raw non-normalized expression values are
used. The principle of stability measure (M) is that the expression ratio (V jk)
of any two ideal reference genes (j and k) (Equation 3) is identical in all
samples, regardless of the experimental condition or cell type. In this way,
variation of the expression ratios of a reference gene, for instance gene z
compared to all other reference genes under investigation, for instance gene
a, gene b and gene c reflects the fact that one (or both) of the gene(s) is (are)
not constantly expressed. Thus, increasing variation in ratio corresponds to
decreasing expression stability. The program automates the calculation of
pairwise variation (Vjk) of every possible combination of a reference gene
with other genes (Equation 2) based on log-transformed expression ratios
(Equation 1). Based on average of pairwise variations of a reference gene, M
is calculated (Equation 3). Comparison of M for all reference genes allows
elimination of the worst-scoring reference gene (highest M value).
Assuming that the reference genes are not co-regulated, stepwise
exclusion of the gene with the highest M value results in a combination of
two constitutively expressed housekeeping genes that have the most stable
expression in the tested samples.
164
For any given tissue sample (m), real-time RT-PCR gene-expression
levels aij of n reference genes are measured. For every combination of two
internal control genes j and k, an array Ajk of m elements is calculated which
consist of log2-transformed expression ratios aij/aik (Equation 1):
(∀j, k ∈ [1, n] and j ≠ k) (Vandesompele et al., 2002):
(1)
Pairwise variation Vjk for the reference genes j and k is defined as the
standard deviation of the Ajk elements (Equation 2).
(2)
165
The gene-stability measure Mj for reference gene j is the arithmetic
mean of all pairwise variations Vjk (Equation 3).
(3)
In iterative steps of exclusion, genes with the lowest expression
stability (i.e. the highest Mj value) are removed. This procedure is repeated
until only the genes with the lowest M j values and hence the most stable
expression remains. The normalization factor is then calculated based on the
geometric mean of the final optimal set of reference genes.
166
In order to measure expression levels accurately, normalization by
multiple reference genes instead of one is required. GeNorm recommends
the minimal use of the three most stable internal control genes for
calculation of an RT-PCR normalization factor (NFn, n = 3), and stepwise
inclusion of more control genes until the (n + 1)th gene has no significant
contribution to the newly calculated normalization factor (NFn + 1). However,
it is impractical to use, for example, 4 reference genes when only a few
target genes need to be studied, or when only minimal amounts of RNA are
available. If all the reference genes under consideration are relatively stable
in expression, the normalization factor will not significantly change whether
or not more genes are included.
167
Thus, in order to determine the ideal number of genes for
normalization, GeNorm calculates, the pairwise variation Vn/n + 1 between the
two sequential normalization factors (NFn and NFn + 1) for all samples within
the same tissue panel (with aij = NFn,i and aik = NFn + 1,i, n the number of
genes used for normalization, and i the sample index; normalization factor
(NFn + 1). A large variation means that the added gene has a significant effect
and should preferably be included for calculation of a reliable normalization
factor. Normalization factors were calculated for the three most stable
control genes (NFn=3, lowest M value) and for the remaining additional
genes by stepwise inclusion of the most stable remaining reference gene
(NFn + 1). Pairwise variations (Vn/n+1) were subsequently calculated for every
series of NFn and NFn
+ 1
normalization factors, reflecting the effect of
adding an (n + 1)th gene.
NormFinder Analysis:
168
The NormFinder Visual Basic application for Microsoft Excel was
used as an alternative algorithm to GeNorm algorithm for determining
suitable reference genes in AT. NormFinder uses a mathematical model
based approach to describe the expression values measured by RT-PCR,
separate analysis of the sample subgroups, estimation of both the intra- and
the inter-group expression variation, and calculation of a candidate gene
“stability value”. The model uses log transformed gene expression (yigj) for
gene i in the jth sample within the group g. (Equation 4):
(4)
169
The three components of the model being: the general expression
level for candidate gene i within group g (αig), the amount of mRNA in the
sample j (βgj) and the random variation caused by biological and
experimental factors (εigj). The objective is to find the two genes with the
least intra- (σig2) and inter- (αig) group expression variation across all sample
groups. Confidence intervals on the inter-group variances are indicated by
averages of intra-group variances and represented as error bars on intergroup variances. The program algorithm combines the intra-group (σ2ig) and
inter-group (zig - θg, g = 1,...,G) variation and expresses it as a stability value
(ρig) for each investigated gene where γ2 is the variance of expression levels
(αig) (Andersen et al., 2004):
(5)
170
This expression effectively combines multiple sources of variation,
and indicates the overall systemic error per gene. Therefore, the top ranked
gene (which has the smallest stability value, hence the smallest combined
variation) is the candidate reference gene most stably expressed in the
sample set being investigated. However, since the systemic error value (ρig)
is calculated with the null assumption that expression levels of each gene
will be group independent, further manual inspection of inter and intra-group
variability was performed. As can be derived from the mathematical
expression of the model, this approach gains in robustness as the number of
samples is increased.
BestKeeper Analysis:
171
BestKeeper is a Excel based spreadsheet software that calculates the
gene expression variation for all individual housekeeping genes based on
crossing points (CP), as defined by the number of cycles necessary to reach
the selected threshold fluorescence (Wittwer et al., 2004). Initial analysis of
the raw CP data determines the expression stability of reference genes. To
achieve this, the program calculates descriptive statistics for each reference
gene: the geometric mean (GM), arithmetic mean (AM), minimal (Min) and
maximal (Max) value, standard deviation (SD) and coefficient of variance
(CV) across all samples, including control and all treatment groups.
Based on the inspection of calculated variations (SD and CV), the
stability of genes is determined. According to the variability observed,
reference genes are ranked from the most stable expression: exhibiting
lowest variation, to the least stable one with the highest variation. All the
remaining stably expressed reference genes are combined into an index BestKeeper index (Equation 6) for the respective sample using the geometric
mean of each candidate gene's CP values, where z is the total number of
reference genes used (Pfaffl et al., 2004).
172
(6)
BestKeeper calculates the relationship between each gene by
pair-wise correlation analyses, assigning each combination a Pearson
correlation coefficient (r) and a probability (p) value. The highly correlated
genes are combined into BestKeeper index and used to define the
relationship between each contributing reference gene and the index. The
relation is defined by Pearson correlation coefficient (r), coefficient of
determination (r2) and p-value. This serves as an estimate of inter-gene
relations and indicates the degree of contribution for each reference gene.
173
Since occurrences of outliers among samples can interfere with the
accuracy of the analysis, BestKeeper analyses sample integrity (InVar)
(Equation 7) for differences in respective CP values (n) and for the average
CP value of each reference gene (m) (Pfaffl et al., 2004).
(7)
BestKeeper then tests each reference gene sample integrity value by
subjecting it to the following analysis (Equation 8) (Pfaffl et al., 2004):
(8)
Samples with efficiency corrected intrinsic variation within 3 fold over or -under expression are considered acceptable. Hence the BestKeeper
software seeks to eliminate outliers and thereby increases the reliability and
consistency of the BestKeeper index.
174
CHAPTER 4: Gene expression of Brown and White Adipose
Tissue
Background:
Gene expression programs of BAT and WAT are highly similar. Both
brown and white adipogenesis programs are regulated by several common
transcriptional regulators such as peroxisome proliferator-activated receptor
gamma (PPARG) and members of CCAAT/enhancer binding protein
(C/EBP) family of transcription factors (Rosen et al., 2000). These
regulators control the production of proteins and processes that are common
to both cell types, including lipogenesis, insulin-dependent glucose
transport, mitochondrial biogenesis, β-oxidation as well as secretion of
adipokines (Rosen et al., 2000), (Farmer, 2006), (Gesta et al., 2007),
(Farmer, 2008). However, notably the genes involved in mitochondrial
175
biogenesis and function are often expressed at higher levels in BAT relative
to WAT (Unami et al., 2004).
The molecular players tipping the scales towards a brown cell type are
much more discreet. One emergent molecule is a zinc-finger protein
PGC1A, which is highly enriched in BAT as compared to WAT. This coactivator along with peroxisome proliferator-activated receptor delta
(PPARD) and PPARG, activates expression of gene UCP1 (Figure 23). In
fact, UCP1 is the only human gene reported to display exclusively brown
adipocyte-specific expression (Cannon and Nedergaard, 2004). This
mitochondria-specific uncoupling protein disconnects electron transport
chain and adenosine triphosphate production, thus, facilitating non-shivering
thermogenesis.
Several RNA expression studies have determined a number
of
molecular players crucial for BAT thermogenesis and function (Table 9)
(Hansen and Kristiansen, 2006), (Gesta et al., 2007). In this study, an
attempt was made to detect and quantify human BAT by quantitative
profiling of selected genes encoding BAT-specific molecules (Figure 23).
176
The expression levels of respective genes were determined in the visceral
adipose tissue of morbidly obese subjects and compared with the levels
expressed in lean subjects.
Table 9: A list of gene profiled in VAT of morbidly obese and non-obese
subjects and their roles in energy homeostasis.
Genes
UCP1
Function
Expressed only in BAT. Uncouples proton flow and ATP synthesis
(Thermogenesis). Present in inner membrane of mitochondrial.
177
PGC1A Transcriptional coactivator. Interacts and regulates PPARG,
CREB, nuclear respiratory factors (NRFs). Direct link between
external physiological stimuli and the regulation of mitochondrial
biogenesis. Regulates genes involved in energy metabolism;
maybe also involved in controlling blood pressure, regulating
cellular cholesterol homoeostasis, and the development of obesity.
SIRT2
NAD+ deacetylase present in the cytoplasm. Involved in cell cycle,
oxidative stress, epigenetic gene regulation. Induced in cold
response and calorific restriction.
SIRT3
NAD+ deacetylase present in mitochondrial matrix. Involved in
cell cycle, oxidative stress. Induced in cold response and calorific
restriction.
178
NAMPT Catalyzes rate limiting condensation of nicotinamide with 5phosphoribosyl-1-pyrophosphate to yield nicotinamide
mononucleotide. Thought to be involved in many important
biological processes, including metabolism, stress response and
aging.
PPARD Nuclear hormone receptor. Potent inhibitor of PPARG and
PPARA
KCNRG Negative regulator of voltage gated potassium channels.
179
Figure 23: Key regulators of the BAT functioning and thermogenesis:
Nuclear factor PPARD and PPARG induce the expression of UCP1 in the
presence co-activator PGC1A. In turn, PGC1A is regulated by deacetlyases
SIRT2 and SIRT3. The NAD needed for SIRT activity is provided by
NAMPT also known as visfatin.
180
UCP1:
The uncoupling protein 1 (UCP1), or thermogenin, is a 33 kDa innermembrane mitochondrial protein that is exclusively present in brown
adipocytes. It was identified in 1976-1977, and purified in 1982, while its
cDNA was cloned in 1984 (Nicholls and Locke, 1984), (Cannon and
Nedergaard, 1985). UCP1 is responsible for the unique thermogenic ability
of BAT (Figure 10).
UCP1 proteins are highly conserved across multiple mammalian
species (Stuart et al., 1999) indicating their biological importance. Indeed, it
is a highly regulated proton transporter over the mitochondrial inner
membrane. When stimulated, UCP1 mediates the passive re-entry of protons
into the mitochondrial matrix: this uncouples proton gradient from ATP
synthesis and dissipates proton gradient as heat, thus, decreasing the ATP
yield of oxidative phosphorylation. Notably, lipolysis and thermogenesis are
closely connected. Activation of lipolysis such as under stimulation by nor181
epinephrine triggers the release of fatty acids from endogenous triglycerides.
The released FFA not only serve as substrates for mitochondrial β-oxidation
but, also interact directly with UCP1 in mitochondria and activate it (Figure
24). Activated UCP1 mediates the passive re-entry of protons into the
mitochondrial matrix, thus producing heat (Palou et al., 1998). Therefore,
lipolysis, mobilizes FFA which serve as substrate for mitochondrial βoxidation and signal for activation of thermogenesis (Figure 24).
The UCP1 gene is mainly regulated at the transcriptional level (Palou
et al., 1998a) by nor-epinephrine (Figure 24) (Cannon et al., 1996). Norepinephrine activates the adrenoceptor β3, which is expressed in both white
and brown adipocytes. This adrenoceptor β3 receptor is coupled via
stimulatory G-proteins to adenylyl cyclase. The key downstream targets of
the adenylyl cyclase include hormone-sensitive lipase and cAMP-responseelement-binding protein (CREB). On one hand, phosphorylation of lipase by
adenyl cyclase triggers lipolysis, thereby providing FFA to activate UCP1
(Figure 24). On the other hand, CREB - also a downstream target of adenyl
cyclase stimulates transcription of the UCP1 gene (Palou et al., 1998)
(Figure 24). In addition, in primary brown adipocytes differentiated in
182
culture, nor-epinephrine stabilizes UCP1 mRNA (Picó et al., 1994). This
suggests that nor-epinephrine mediated increase in UCP1 gene expression is
accompanied by the activation of the constitutive pool of UCP1 protein, thus
augmenting active thermogenesis in more than one way.
Other activators of UCP1 transcription include triiodothyronine (T3),
which acts in concert with nor-epinephrine and retinoic acid and stimulates
UCP1 gene transcription through a mechanism independent from the
adrenergic pathway (Palou et al., 1998). Thus, the regulation of UCP1
expression and activity is highly complex, with multiple factors contributing
to UCP1 induction.
183
Figure 24: Regulation of UCP1 gene expression and activity by norepinephrine and free fatty acids.
184
PPARD:
Peroxisomal proliferator activated receptor-δ (PPARD), is a member
of the nuclear hormone receptor superfamily PPAR (Barish, 2006). PPARD
was identified in 1992 (Schmidt et al., 1992). It has since shown to be an
important player in BAT and WAT functioning. Several studies implicated
PPARD in transcriptional activation of genes involved in fatty acid transport
and oxidation, mitochondrial biogenesis (Barish, 2006), (Ravnskjaer et al.,
2010), (Wan et al., 2010) (Figure 25). Transgenic expression of activated
form of PPARD in adipose tissue produced lean mice resistant to obesity,
hyperlipidemia and tissue steatosis (Wang et al., 2003). It thus seems that
PPARD signaling stimulates decrease in the triglyceride stores through an
increase in fatty acid catabolism capacity.
Importantly, PPARD is shown to induce expression of thermogenic
gene UCP1 upon its own activation (Puigserver et al., 1998). The
thermogenic function of PPARD is similar to that of PGC1A (Wu et al.,
1999). Additionally, PPARD has been shown to be bound to PGC1A both in
cultured cells and in tissue extracts (Wang et al., 2003). It appears that the
185
metabolic effects of PPARD may be mediated through its activation by
PGC1A – a nuclear transcriptional co-activator.
Figure 25: Tissue specific role of PPARD in thermogenesis and energy
homeostasis.
186
PGC1A:
PGC1A – a 90kDa nuclear transcriptional co-regulator is considered
as the master regulator of BAT differentiation and functioning (Farmer,
2008). It has been studied extensively for its role in thermogenesis.
Puigserver et al., showed the expression of PGC1A to be exclusive to
murine BAT compared to WAT and to be strongly induced by cold in BAT
and skeletal muscle (Puigserver et al., 1998). Additionally, the authors
observed a change in the function of WAT from energy storing tissue to a
tissue with thermogenic capacity upon ectopic expression of PGC1A. This
observation was further supported by over-expression and loss-of-function
studies demonstrating the regulation of thermogenesis by PGC1A
(Puigserver et al., 1998), (Lin et al., 2005).
187
The transcriptional changes mediated by PGC1A are attributed to its
interaction with several transcription factors. PGC1A was first identified as a
PPARG-interacting protein from murine brown fat (Puigserver et al., 1998).
PGC1A is now known to interacts with nuclear receptors like NRF-1, NRF2, thyroid hormone (TR), PPARA and PPARD (Semple et al., 2003), (Lin et
al., 2005), (Hondares et al., 2011). As discussed above, interaction of
PGC1A with PPARD in BAT leads to induction of brown fat-specific UCP1
expression (Figure 26) (Hondares et al., 2011).
188
Figure 26: PGC1A and regulation of UCP1 expression.
189
Apart from its specific role in transcriptional activation of UCP1 , a
number of independent studies have also supported the role of PGC1A in
mitochondrial biogenesis and functioning (Leone et al., 2005), (Lin et al.,
2004), energy homeostasis (Lin et al., 2004), hepatic gluconeogenesis and
ketogenesis (Herzig et al., 2001), (Yoon et al., 2001), (Rhee et al., 2003) and
fatty acid beta-oxidation (Semple et al., 2003) (Figure 27). For instance,
PGC1A increases transcription mediated by nuclear respiratory factors
NRF1 and NRF2, leading to the increased expression of nuclear-encoded
mitochondrial transcription factor A (mtTFA). mtTFA translocates to the
mitochondrion, where it stimulates mitochondrial biogenesis as manifested
by the stimulation of mitochondrial DNA replication and mitochondria gene
expression. PGC1A also mediates the transcription of nuclear-encoded
mitochondria subunits of the electron transport chain complex such as ßATP synthase, cytochrome c, and cytochrome c oxidase IV (Liang and
Ward, 2006). Hence, by simultaneously co-activating several transcription
factors, PGC1A can rapidly and synchronously modulate a transcriptional
program that governs energy metabolism.
190
PGC1A has been shown to be a key coordinator of gene regulatory
responses in not only BAT, but in other tissues with high energy demands as
well. PGC1A is highly expressed in tissues kidney, brain, heart and skeletal
muscle (Puigserver et al., 1998), (Knutti and Kralli, 2001), (Semple et al.,
2003), (Pan et al., 2009b), which are also characterized by a high
mitochondrial density. For a tissue like WAT to function as an energy
storage organ and given the role of PGC1A in mitochondrial biogenesis, it is
apparent that PGC1A expression must be low or suppressed (Cantó and
Auwerx, 2009). In accordance, studies show that levels of PGC1A
expression in WAT are low (Puigserver et al., 1998), (Semple et al., 2003)
and that the ectopic expression of PGC1A in WAT increases its thermogenic
ability (Puigserver et al., 1998). PGC1A is also known to co-activate
transcription factors HNF-4α and FOXO1, thus driving the expression of
genes involved in gluconeogenesis (Schwer and Verdin, 2008).
Given the involvement of PGC1A in several processes of energy
metabolism in mitochondrial dense tissues and the role of PGC1A in
mitochondrial biogenesis, we attempted to assess whether the expression of
191
PGC1A in visceral adipose depot could serve as a marker of presence of
mitochondrial rich BAT interspersed in VAT.
Figure 27: Role of PGC1A in BAT and energy homeostasis.
192
An ERR (estrogen-related receptor) also belongs to the group of
nuclear receptors co-activated by PGC1A. This nuclear receptor is
responsible for activation of SIRT3 transcription (Giralt et al., 2011). As
discussed below, SIRT3 is an important mitochondrial protein involved in
fatty acid oxidation.
SIRTs:
Sirtuins (SIRTs) have emerged as the key metabolic sensors that
directly link environmental signals to mammalian metabolic homeostasis
and stress responses (Li and Kazgan, 2011). SIRT genes encode a family of
nicotinamide
adenine
dinucleotide
(NAD+)–dependent
deacetylases
involved in chromatin remodeling, cellular metabolism and lifespan
regulation (Gasser and Cockell, 2001), (Dali‐Youcef et al., 2007).
Eukaryotic SIRTs have been divided into four broad phylogenetic groups
based on sequence similarities. SIRT1, SIRT2, and SIRT3 composing class
193
I, SIRT4 constituting class II, SIRT5 forming class III, while SIRT6 and
SIRT7 are forming class IV (Roy A., 2000). However, there is no obvious
correlation between this classification and the specific biological functions
of the SIRTs. SIRTs can be also classified based on their intracellular
localizations (Michishita et al., 2005). SIRT1, SIRT3, SIRT6, and SIRT7,
are nuclear proteins, with distinct subnuclear localizations. SIRT2 is
generally localized in the cytoplasm, but moves to the nucleus during the
G2/M phase and binds chromatin (Vaquero et al., 2006). SIRT3, SIRT4, and
SIRT5 are present in the mitochondria. Recent studies suggest that SIRT3
can also be a nuclear protein that transfers to the mitochondria during
cellular stress (Scher et al., 2007). Because SIRTs require the coenzyme
nicotinamide adenine dinucleotide (NAD+) as a substrate, NAD+, NADH,
or the ratio NAD+/NADH modulate SIRTs activity, linking it to the energy
status of the cell .
SIRT2: Mammalian SIRT2 is the most abundant sirtuin in adipocytes
(Jing et al., 2007) where it shows a predominantly cytoplasmic distribution
and enters nuclei upon activation (Vaquero et al., 2006). SIRT2 deacetylates
histones, thus playing a critical role in transcriptional silencing (Perrod et al.,
194
2001), (North et al., 2003). SIRT2 activity is mediated by members of the
forkhead transcription O (FOXO) subfamily of transcription factors such as
FOXO1 (Jing et al., 2007). SIRT2 deacetylates FOXO1, thereby enabling
the binding of FOXO1 to PPARG. This represses PPARG transcriptional
activity and subsequently adipogenesis and UCP1 gene expression response
(Jing et al., 2007) (Figure 28).
195
Figure 28: Role of SIRT2 in regulation of thermogenic gene expression.
The regulation of SIRT2 is not well understood. It has been shown to
respond to starvation and energy expenditure (Jing et al., 2007), (Wang et
al., 2007), (North et al., 2003). For example, animal studies have shown
196
increased levels of SIRT2 proteins in WAT of C57BL/6 mice subjected to
calorific restriction (Wang et al., 2007). Notably, while short-term food
deprivation induces murine Sirt2 expression in both WAT and BAT, cold
exposure elevates Sirt2 expression only in BAT (Wang and Tong, 2009).
Given the fact that Sirt2 expression is upregulated under calorific restriction
and cold exposure in murine BAT and is involved in up-regulation of
thermogenesis and mitochondrial β-oxidation, it is highly plausible that
SIRT2 may be downregulated in VAT under conditions of positive energy
balance resulting in decreased thermogenesis. Thus, we attempted to
determine the expression of human SIRT2 in VAT of obese and lean
patients.
SIRT3: SIRT3 is another member of sirtuins that groups along with
SIRT1 and SIRT2 into class I based on sequence similarity (Roy A., 2000).
Like other molecules of the same family, functionally, it is classified as
predominantly mitochondrial protein (Michishita et al., 2005). However, it is
suspected that it may function in a nucleus before it moves into the
mitochondria upon activation (Scher et al., 2007). The biological
significance of SIRT3 in the nuclear subcellular compartment is not yet
197
established (Scher et al., 2007). Interestingly, in humans, SIRT3 is a soluble
protein in the mitochondrial matrix (Schwer et al., 2002), while in rodents
Sirt3 is found mostly in the inner mitochondrial membrane (Shi et al., 2005),
pointing at possibility that rodent and human SIRT3 proteins may have
different targets and functions.
In BAT tissue of rodents, Sirt3 has been shown to play a vital role in
the control of adaptive thermogenesis. Similar to SIRT2, both cold exposure
and calorific restriction has been shown to induce murine Sirt3 expression in
BAT in vivo (Shi et al., 2005). Importantly, while PGC1A positively
regulates gene expression of SIRT3 (Kong X et al., 2010), SIRT2 activates
PGC1A activity (Krishnan et al., 2012). Thus, it is expected that under
conditions that stimulate the expression and activity of SIRT2 and PGC1A
such as cold exposure or calorific restriction, gene expression of SIRT3 will
also be upregulated (Figure 28). In support of this hypothesis, studies have
shown similar patterns of increase in mRNA levels of both SIRT3 and UCP1
mRNAs in brown adipose after a 12-h period of cold exposure. Notably,
overexpression of SIRT3 in HIB1B brown adipocytes increased the
expression of PGC1A, UCP1 and a number of other mitochondria-related
198
genes (Shi et al., 2005). Thus, there appears to be a positive feedback loop
between SIRT3 and PGC1A. At present, the relevance of these observations
in vivo is not known.
Studies in murine models show additional roles of Sirt3 in fatty acid
oxidation as it activates deacetylating enzymes such as long-chain acyl-CoA
dehydrogenase and acetyl-CoA synthase (Hirschey et al., 2010), (Giralt et
al., 2011). In support of these recent observations, Sirt3 knockout mice
(Sirt3KO) have been shown to have hyperacetylated mitochondrial proteins
(Lombard et al., 2007) and are susceptible to diet-induced obesity. In a
recent study, Hirschey et al., 2011 demonstrated that a high fat diet (HFD)
induced down-regulation of Sirt3 (Hirschey et al., 2011). The subsequent
hyperacetylation of mitochondrial proteins in the liver was associated with
increased susceptibility to obesity, insulin resistance, hyperlipidemia, and
steatohepatitis.
The role of SIRT3 in regulating the expression of mitochondrial
energy metabolism genes and its positive regulation by SIRT2 and PGC1A,
lead us to believe that in the presence of mitochondrial rich BAT and
199
absence of positive energy balance, SIRT3 expression will be upregulated in
VAT of lean subjects.
NAMPT (visfatin):
Nicotinamide phosphoribosyltransferase (NAMPT), also known as
pre-B cell colony enhancing factor and visfatin is a highly conserved
enzyme in all animals (Moschen et al., 2007), (Luk T et al., 2008). NAMPT
was originally identified in 1994 as a modulator of B cell differentiation
produced by lymphocytes and acting on lymphocyte maturation and
inflammatory regulation (Samal et al, 1994). Later, it was also referred to as
Visfatin, due to the observation that it is predominantly produced in VAT . It
is secreted in VAT through a non-classical secretory pathway (Revollo et al.,
2007). Since then it has been shown that NAMPT is ubiquitously expressed
in all tissues. In mice, among the tissues examined, brown adipose tissue
(BAT), liver, and kidney showed the highest levels of intracellular Nampt
(iNampt) protein expression, while heart showed an intermediate level of
iNampt. White adipose tissue (WAT), lung, spleen, testis, and muscle
contain low levels of iNampt, while in the brain and in the pancreas iNampt
200
is undetectable (Revollo et al., 2007). In humans, apart from leukocytes and
adipose tissue, NAMPT is also expressed in hepatocytes (Garten et al., 2010)
and muscles (Costford et al., 2010).
Modern understanding of the NAMPT function is that it is a
pleotropic
protein
that
plays
an
important
role
in
oxidative
metabolism/cellular energetics. NAMPT controls the rate limiting step in the
nicotinamide adenine dinucleotide (NAD) biosynthetic pathway. NAD is
synthesized from three major precursors—tryptophan, nicotinic acid, and
nicotinamide (Shin-ichiro Imai, 2009). In mammals, NAMPT catalyses the
first step in NAD biosynthesis: conversion of nicotinamide to nicotinamide
mononucleotide (NMN). NMN is subsequently converted to NAD by
nicotinamide/nicotinic acid mononucleotide adenylyltransferase (NMNAT)
(Shin-ichiro Imai, 2009). By virtue of its role in NAD biosynthesis, NAMPT
can regulate cellular levels of NAD. It can thus impact not only cellular
energy levels but also NAD-dependent enzymes such as SIRTs. Most
notably, NAD is a substrate of enzymes involved in post-translation
processing. Thus, NAMPT can be an important enzyme for drug discovery.
201
Apart from its role in NAD biosynthesis, NAMPT is believed to
function as a pro-inflammatory cytokine. NAMPT expression is upregulated in a variety of acute and chronic inflammatory conditions
including sepsis, acute lung injury, rheumatoid arthritis, inflammatory bowel
disease, and myocardial infarction. It plays a key role in the persistence of
inflammation through its capacity to inhibit neutrophil apoptosis (Luk T et
al., 2008). NAMPT stimulates the p38 mitogen-activated protein kinase (p38
MAPK) and extracellular signal-regulated kinase (ERK) pathways,
augmenting the production of IL-1β, TNF-α and IL-6 (Tilg and Moschen,
2008). In turn, these factors stimulate chemotactic activity of human
monocytes (Moschen et al., 2007). The pro-inflammatory activity of
NAMPT was shown to be regulated by negative feedback by TNF-α, IL-6,
growth hormone, and β-adrenergic receptor agonists, while glucocorticoids
had an opposite effect in 3T3-L1 cells (Kralisch et al., 2005a), (Kralisch et
al., 2005b).
Reports on the role of NAMPT in obesity related phenotypes are
ambiguous. Some authors show that the circulating levels of NAMPT in
patients with obesity and T2D are higher than that in normal controls
202
(Haider et al., 2006), (El-Mesallamy et al., 2011). Some others (Oki et al.,
2007), however, point that the levels of NAMPT correlate to the level of
serum inflammatory markers like IL-6 and CRP, but not insulin resistance.
Overall, NAMPT seems to be involved in both inflammation and pancreatic
function. However, further studies are required to understand its
physiological functions in VAT with regard to obesity-linked metabolic
disorders.
Yang et al have demonstrated that the levels of NAMPT increase in
response to fasting (Yang et al., 2007), indicating its energy sensing role. It
is highly plausible that, in obesity, high levels of circulating FFA and
positive energy balance may trigger suppression of NAMPT expression. In a
study by Pagano et al., mRNA expression of NAMPT was analyzed in SAT
(subcutaneous adipose tissue) and VAT of 30 lean and 39 obese subjects
(Pagano et al., 2006). NAMPT mRNA levels in SAT are significantly lower
in obese subjects, compared to that in normal-weight controls, while VAT
NAMPT mRNA levels positively correlated with body mass index.
Similarly, in an independent study (Berndt et al., 2005) of 189 subjects,
there was a significant correlation between visceral NAMPT gene expression
203
levels and BMI (r2 = 0.06, P = 0.001) , whereas no significant association
between BMI or percent body fat and subcutaneous NAMPT mRNA
expression was registered (both P > 0.5). It is important to note that the
reports on differences in NAMPT mRNA levels in SAT vs VAT are
conflicting (Berndt et al., 2005), (Pagano et al., 2006), (Terra et al., 2011).
However, all studies concur on a positive correlation of VAT NAMPT
mRNA and BMI.
To clarify these ambiguities, levels of NAMPT mRNA were
determined in human VAT in morbidly obese subjects and lean subjects.
KCNRG:
KCNRG (K+ Potassium CN Channel RG Regulator, GeneID:
283518), was first cloned by Ivanov et al. (Ivanov et al., 2003). KCNRG
encodes a soluble protein, whose N-terminus bears sequence homology to
the T1 (tetramerization) domain of several voltage-gated potassium
channels, including members of the Shaker family of potassium channels.
Heterologous expression of KCNRG has been shown to reduce whole-cell
204
potassium currents in the prostate cancer cell line LNCaP (Ivanov et al.,
2003) and also reduce cellular proliferation rates in the hepatic cell line
Hep3B (Cho et al., 2006). It has been suggested that KCNRG functions by
interfering with the assembly of the T1 domain of Kv channels (Ivanov et
al., 2003) or interfering with the surface expression of the receptors (Usman
and Mathew, 2010), thereby reducing the number of functional channels and
hence whole-cell K+ currents. While the role of voltage gated potassium
channels has not been extensively investigated in obesity and adipose tissue,
Xu et al., have reported that Kv1.3 channels knockdown protects mice from
diet induced obesity (Xu et al., 2004). The authors observed that while food
intake did not differ significantly between knockout mice and controls, basal
metabolic rate was significantly higher in knockout animals (Xu et al.,
2004). Thus, we hypothesize that KCNRG -a negative regulator of Kv
channel also plays a role in regulating energy homeostasis and aimed at
determining the gene expression of KCNRG in VAT of morbidly obese
subjects as compared to lean controls.
205
Results:
Expression levels of the genes encoding multiple factors important
for BAT differentiation, thermogenesis and energy homeostasis were
profiled in VAT in morbidly obese (BMI >40) and non-obese patients (BMI
< 30) (Table 10).
206
Table 10: Demographic and clinical data.
Obese
Lean
(Average ± SD)
(Average ± SD)
5
5
Age (years)
40.00 ± 9.0
52.00 ± 18.0
ALT (U/L)
42.00 ± 13.98
27.66 ± 8.0
AST (U/L)
30.00 ± 7.7
29.3 ± 5.8
BMI
45.3 ± 7.1
21.8 ± 2.9
Clinical Parameter
N
207
Diabetes presence
60% (3)
20%(1)
Gender (Females)
80 % (4)
40 % (2)
Race (Caucasian)
80 % (4)
80 % (4)
Within the omental depots of morbidly obese subjects as compared to
lean subjects, PGC1A gene expression trends toward down-regulation (-1.1
fold, NS) (Table 11). In active BAT and in mitochondrial rich tissues,
PGC1A is reported to have higher expression (Farmer, 2008). Thus, the
observed down-regulation of PGC1A could be attributed to the absence of
mitochondria-rich brown adipose cells within the VAT of obese subjects or
208
could also be an indicator of mitochondrial dysfunction in VAT of morbidly
obese subjects. Additional studies are warranted to determine the
mitochondrial functioning in VAT of obese subjects. Under conditions of
activation, PPARD/PGC1A increase the expression BAT specific genes,
including UCP1 (Hondares et al., 2011) (Figure 23). Thus, it can be
expected that down-regulation of PGC1A gene will be accompanied by
decrease in expression of UCP1 gene. In line with this premise, UCP1 gene
expression was downregulated in VAT of morbidly obese individuals (-1.6
folds, NS) (Table 11) as compared to VAT of lean subjects. Another protein
regulating the activity of PGC1A is SIRT2. It is encoded by gene SIRT2 and
was also downregulated in VAT of obese subjects (-1.2 folds, NS) (Table
11). Our data are in agreement with previous findings that, under conditions
of positive energy balance, a decrease in expression of Sirt2 is observed
(Jing et al., 2007), (Wang et al., 2007), (North et al., 2003).
As opposed to PGC1A, UCP1 and SIRT2, the SIRT3 gene showed a
trend toward up-regulation (1.4 folds, NS) (Table 11) in VAT of morbidly
obese subjects as compared to that of lean subjects. These results contradict
observations made using BAT samples collected from cold-exposed rodents,
209
where a similarity of mRNA expression patterns for Ucp1 and Sirt3 was
observed. These differences may be attributed to the different cellular
localization of SIRT3 reported in rodents and humans (Schwer et al., 2002),
(Shi et al., 2005). As rodent and human SIRT3 proteins are likely to have
different intracellular targets, they might as well be under control of
different sets of transcriptional regulators. Another gene with observed
expression pattern contrary to that reported in literature was PPARD. The
PPARD expression is up-regulated (1.4 folds, NS) (Table 11) in VAT of
morbidly obese subjects. While PPARD was previously reported to upregulate the expression of UCP1 (Puigserver et al., 1998), our studies
showed that PPARD trends toward over-expression, while UCP1 trends
towards down-regulation. In our opinion, in our study cohort an upregulation of PPARD may serve as an indicator of increased triglyceride
hydrolysis and fatty acid oxidation due to pre-surgery dieting attempts
common in morbidly obese subjects undergoing bariatric surgery.
In addition to nuclear factors important in BAT differentiation and
functioning, the gene expression levels for two other targets, NAMPT and
KCNRG, was also profiled.
210
NAMPT is a pleotropic protein that plays an important role in cellular
metabolism via regulation of NAD levels. In VAT samples collected from
obese subjects, expression of NAMPT was down-regulated 5-folds (p <
0.007) (Table 11). These findings are in accordance with previously
observed inverse relation of NAMPT expression levels with the state of
energy balance (Yang et al., 2007b). Notably, the expression of SIRT2 was
also down-regulated in the cohort of obese patients, albeit this downregulation was not statistically significant. This is interesting in the light of a
recent study demonstrating role of a transcription factor FOXO1 in upregulation of NAMPT (Tao et al., 2011). To bind DNA, FOXO1 needs to be
activated by SIRT2 deacetylase that is in turn, stimulated by NAMPT
through regulation of cellular levels of NAD. Thus, the observed downregulation of both SIRT2 and NAMPT in conditions of morbid obesity hints
at a potential role of this pathway in energy homeostasis.
KCNRG is an ER associated protein that serves as negative regulator
of the voltage-gated K+ channels (KV1 channels) by reducing their surface
expression (Usman and Mathew, 2010b). The role of KCNRG has never
been investigated in VAT. Human KCNRG encodes two protein isoforms
211
KCNRG-L (mRNA isoform B encoding 272 aa long protein) and KCNRG-S
(mRNA isoform A encoding 229aa long protein) differing in their C-ends
and possessing common N-end of 191 aa (Birerdinc et al., 2010) (Figure 29).
The C-end difference is due to an out-of-frame insertion resulting in a
truncated 229 aa protein from KCNRG-S transcript. Human mRNA
isoforms encoding the two KCNRG proteins are reported to be co-expressed
in the same set of tissues (Birerdinc et al., 2010).
For real-time PCR profiling of KCNRG isoforms two primer pairs
previously designed in the laboratory (Birerdinc et al., 2010) were used.
KCNRGV1 primer pair is capable of amplifying both mRNA isoforms with
product sizes 107 basepairs (KCNRG-L) and 129 basepairs (KCNRG-S)
respectively while KCNRGV2 primer pair specifically amplified KCNRG-S
specific
product.
Primer
sequences
TTTTCCCTCCTCAGATGACC-3′
were
as
follows:
V1
and
5′5′-
TCCAGTTTGGTTATCAGTAGTGC-3′; V2 5′-CCTGGTTTTCCAGTGTG
-3′,
and
5′-GCTGAGGCAGGAGAATCACT-3′.
Interestingly,
only
KCNRG-S transcript showed a significant down-regulation (10-folds, p <
0.03) (Table 11) expression in VAT samples from morbidly obese subjects
212
as compared to that from lean subjects. Blocking KV1 potassium current by
KCNRG may potentially produces an increase in basal metabolic rate. The
highly significant difference in gene expression of KCNRG mRNA isoform B
encoding for KCNRG-S protein that is observed in obese as compared to
non-obese subjects warrants additional studies on role of this gene in human
adipose.
213
214
Figure 29: Genomic organization of KCNRG gene locus. Scheme
represents the structure of the mRNA isoforms and the encoded protein
isoforms of the human KCNRG genes.
215
Table 11: Clinical and Gene Expression Data
Obese (N=5)
Lean (N=5)
Fold
Average ± SD
p
Average ± SD
Regulation value
BMI
45 ± 7.1
21.8 ± 2.9
NA
0.01
Age
40 ± 9
52 ± 18
NA
0.28
80%
40%
NA
Gender
(Females)
216
Diabetes
60%
20%
NA
UCP1
0.02 ± 0.02
0.04 ± 0.04
-1.6
0.8
PGC1A
0.12 ± 0.07
0.12 ± 0.12
-1.1
1
PPARD
0.06 ± 0.05
0.04 ± 0.01
1.4
0.6
SIRT2
0.18 ±0.12
0.23 ± 0.07
-1.2
0.5
SIRT3
0.05 ±0.04
0.04 ± 0.03
1.4
0.8
NAMPT
0.15 ± 0.12
0.8 ± 0.4
-5.0
0.007
217
KCNRGV1
0.004 ± 0.002
0.008 ± 0.004
-2.0
0.09
KCNRGV2
0.001 ± 0.001
0.01 ± 0.005
-10.0
0.03
Discussion and Conclusion:
218
Figure 30: Gene expression profile in visceral AT of obese subjects
compared to lean subjects
219
The qRT-PCR based profiling of gene expression demonstrated trends
toward up-regulation of PPARD (Figure 30), an important transcription
factor involved in activation of UCP1, fatty acid oxidation and triglyceride
hydrolysis and SIRT3- that deacetylates substrates important for stimulation
of β-oxidation. On the other hand, expression of PGC1A that encodes a
coactivator of various transcription factors including PPARD, shows downregulation and is accompanied by suppression of mRNA levels of SIRT2 and
UCP1 (Figure 30). Thus, in VAT of morbidly obese subjects we observed a
concerted down-regulation of BAT specific gene expression along with
factors regulating fatty acid oxidation (Figure 31). Notably, this study shows
for the first time the expression of BAT specific UCP1 mRNA in VAT of
human subjects and is trending toward down-regulation in VAT of obese
subjects. These finding may be employed for development of novel methods
to detect functional brown adipocytes within visceral adipose depots of
humans.
220
Figure 31: Potential role of nuclear factors in thermogenesis and
lipolysis in omental adipose tissue as it relates to morbid obesity.
221
Further, our study explored potential roles of NAMPT and KCNRG in
obesity. As discussed previously, a study by Yang et al., has reported a
positive regulatory loop between NAMPT and SIRT2 (Yang et al., 2007b).
Here we observed concomitant down-regulation of NAMPT and SIRT2
expression in AT of obese subjects as compared to lean subjects (Figure 31),
thus pointing at importance of this energy sensing network in obesity.
Potassium channel suppressor KCNRG has not been previously studied in
the context of obesity. The significant 10-fold down-regulation of
KCNRGV2 (KCNRG-S isoform) calls for additional studies on the role of
voltage gated potassium channel regulators in energy homeostasis.
The study, though limited by the small sample size, shows for the first
time the altered gene expression of UCP1 and nuclear factors involved in
thermogenesis and lipolysis in VAT of morbidly obese human subjects.
Additional studies with a larger sample size are warranted.
Methods:
222
Sample Collection:
The proposed study was conducted in collaboration with the
Translational Research Institute (TRI) of INOVA Health System (Falls
Church, VA).
Human adipose tissues were collected during bariatric
surgery or other intra- abdominal surgery and flash frozen in liquid nitrogen
and stored at -80 0C. The samples were de-identified in compliance with
HIPAA regulations. Clinical data associated with the samples in the
repository are available.
RNA Extraction:
Total RNA was extracted from VAT using mirVana kit (Ambion).
The first step of the mirVana miRNA Isolation Kit procedure was to disrupt
samples in a denaturing lysis buffer. The lysis buffer disrupts the cells and
releases the contents into solution. The lysis buffer contains guanidinium
thiocyanate which immediately inactivates RNases. Next, samples were
subjected to acid-Phenol:Chloroform extraction which provides a robust
223
front-end purification that removes majority of the genomic DNA along with
proteins and lipids. After organic extraction, the final purification was
accomplished via a glass- fiber filter. Ethanol was added to samples, and
they were passed through a filter cartridge containing a glass-fiber filter
which immobilizes the RNA. The filter was then washed a few times, and
finally the RNA is eluted with a low ionic-strength solution.
Briefly, for total RNA extraction 100 mg of human VAT samples,
were excised while in liquid nitrogen and added to a tube containing 10
volumes of Lysis/Binding Buffer per tissue mass (~1 mL). Using a
homogenizer (Power Gen 125, Fisher, USA) the tissue was immediately
homogenized. Performing the excision in liquid nitrogen ensured minimal
thawing of tissue, which would otherwise compromise the integrity of RNA
due to RNA degradation. Following homogenization, 1/10 volume of
miRNA Homogenate Additive was added to each tube. After inverting
several times to ensure uniform mixing, the tube was incubated on ice for 10
min. At the end of incubation period, equal volume of acid-Phenol:
Chloroform was added to the tube, taking care to add the bottom phase of
acid-Phenol:Chloroform. The upper phase is the aqueous buffer. The
224
mixture was then vortexed for 30 - 60 sec and the samples centrifuged
(Eppendorf, USA) for 5 min at maximum speed (10,000 x g) at room
temperature. This helped to separate the aqueous and organic phases. This
process ensures that the DNA and proteins partition into the interphase and
lower organic phase while RNA partitions into upper aqueous phase.
Aqueous (upper) phase was carefully transferred to a new tube without
disturbing the lower phase.
To the aqueous phase, 1.25 volumes of 100% ethanol at room
temperature was added and mixed thoroughly by pipetting. The
lysate/ethanol mixture was then passed through a glass fiber filter cartridge
by adding 700 μL of mixture on the cartridge placed on a fresh tube. For
samples with larger than 700 μL volume the mixture was applied in
successive applications to the same filter. After each addition, tube was
centrifuged (Eppendorf, USA) for ~15 sec at RCF 10,000 x g (typically
10,000 rpm) to pass the mixture through the filter. Flow-through was
discarded after each addition and process repeated until all of the
lysate/ethanol mixture was passed through the filter. miRNA Wash Solution
1 (700 μL) (working solution mixed with ethanol) was applied to the filter
225
cartridge and centrifuged (Eppendorf, USA) for ~5–10 sec, with the flow –
through being discarded. Wash Solution 2/3 (500 μL) (working solution
mixed with ethanol) was then added to the filter and again centrifuged
(Eppendorf, USA) for 5- 10 sec. This step was repeated with a second 500
μL aliquot of Wash Solution 2/3, discarding flow-through each time. After
discarding the flow-through from the last wash, the filter cartridge was
replaced in the same collection tube and the assembly spun for 1 min to
remove residual fluid from the filter. This ensures RNA yields with high
purity, free from all the salts and ethanol which can otherwise interfere with
downstream analysis procedures.
To elute total purified RNA, filter cartridge was placed on a fresh
collection tube. Preheated RNase free water (30 μL) was applied to the
center of the filter. After incubation at room temperature for 5 min, the filter
was spun down for ~20–30 sec at maximum speed to recover the RNA. This
step was repeated with 20 μL of preheated RNase free water to obtain a total
of 50 μL of total RNA. RNA quality and quantity assessment was carried
out as described below.
226
Assessing RNA quality and quantity:
To determine the quantity and purity of the extracted RNA,
absorbance at 260 nm (A260) and 280 nm (A280) was measured using
GeneQuant 1300 spectrophotometer (GE, USA). A260/A280 ratio between
1.8-2.1 indicates high quality RNA, while yield of total RNA was estimated
by A260.
In order to assess the integrity of RNA, 1% agarose gel
electrophoresis was carried out. 400 mg of agarose was mixed with 40 mL
of 1X TAE (Fisher Scientific) and heated to dissolve agarose in the buffer.
After the solution cools to 50 0C, 4 μL of ethidium bromide (10mg/ml) was
added and gently swirled for uniform distribution. The solution was then
poured in a casting tray ensuring no bubbles are formed and a desired comb
placed. The gel was then allowed to solidify at room temperature for 30 min.
After loading the RNA samples in equal concentrations (250 ng - 500 ng)
with 1X loading dye and in the presence of a 1 kb ladder, gel electrophoresis
was carried out at 100 V for 1 hour in 1X TAE. After separation, the gel was
visualized under UV (Bio-Rad) and results were documented.
227
Intact total RNA was expected to have sharp, clear 28S and 18S
rRNA bands. The 28S rRNA band would be approximately twice as intense
as the 18S rRNA band. This 2:1 ratio (28S:18S) is a good indication that the
RNA is completely intact. Partially degraded RNA will appear as a smear
and lack the sharp rRNA bands, or will not exhibit the 2:1 ratio of high
quality RNA. Completely degraded RNA will appear as a very low
molecular weight smear.
First Strand cDNA Synthesis:
After assessing the quality and quantity of total RNA, cDNA
synthesis (RT² First Strand Kit , Qiagen) was carried out on the same day of
total RNA extraction to prevent loss of RNA during storage. The procedure
also contains a genomic DNA elimination step prior to reverse transcription.
Random hexamers and oligo-dT primed the reverse transcription reactions
and a reverse transcriptase synthesized the cDNA product.
The first step in cDNA synthesis was the genomic DNA elimination
step. This was done by addition of genomic DNA elimination mixture to 560
228
ng of total RNA followed by incubation at 42 0C for 5 min. The mixture was
then placed on ice immediately for at least one minute.
1. Genomic DNA Elimination Mixture:
For each 560 ng of total RNA sample the following were combined in
a sterile PCR tube:
Total RNA
25.0 ng to 5.0 ug
GE (5X gDNA Elimination Buffer)
2.0 μL
RNase free water to a final volume of
10 μL
The next step in cDNA synthesis was reverse transcription using
random hexamers and oligo-dT primers. This was done by addition of RT
cocktail to the above mixture, followed by incubation at 42 0C for exactly 15
229
min in a PCR machine (Bio-Rad). The reaction was then immediately
stopped by heating at 95 0C for 5 minutes. Sterile water (91 µL) was then
added to each 20-µL of cDNA synthesis reaction and mixed well. The
cDNA was then stored at -20 0C until further use.
2. RT Cocktail:
RT Cocktail
1 reaction
BC3 (5X RT Buffer 3)
4 µL
P2 (Primer and External Control Mix)
1 µL
RE3 (RT Enzyme Mix 3)
2 µL
230
RNase free H2O
3 µL
Final Volume
10 µL
Selection of BAT specific genes and primer design:
Gene expression of several genes that are integral to the
formation and functioning of brown AT including: UCP1, PGC1A, PPARD,
SIRT2, SIRT3 and NAMPT were shortlisted. Additionally, KCNRG
transcript variants were also targeted. Custom primers were designed for
each of these genes and verified for specificity using NCBI BLAST (Ye et
al., 2012) to span exon-exon boundaries, and to be target-specific. Gene
accession numbers as well as primer sequences are listed in Table 12.
Housekeeping genes validated in the previous study were used as the
reference genes for the purpose of normalization.
231
Table 12: Primer sequences of BAT specific genes.
Target
Gene Accession
Gene
Number
UCP1
NM_021833.4
Tm
Primer Sequence
55 0C F-5’TCTCTCAGGATCGGCCTCTA
R-5’GCCCAATGAATACTGCCACT
PGC1A
NM_013261.3
55 0C F-5’GTGAAGACCAGCCTCTTTGC
232
R-5’TCACGTCTCCATCTGTCAGC
PPARD
NM_001171818.1 55 0C F-5’CTATCCGTTTTGGTCGGATG
R-5’CGATGTCGTGGATCACAAAG
SIRT2
NM_001193286.1 55 0C F-5’TCGCAGAGTCATCTGTTTGG
R-5’GGAGTAGCCCCTTGTCCTTC
SIRT3
NM_001017524
55 0C F-5’CTGTACAGCAACCTCCAGCA
R-5’GCAAAGGTTCCATGAGCTTC
NAMPT
NM_005746
55 0C F-5’TATCCACCCAACACAAGCAA
233
R-5’GACAAAGCCCTCAGGAACAG
KCNRGV1 NM_173605.1,
60 0C F-5’TTTTCCCTCCTCAGATGACC
NM_199464.2
R5’TCCAGTTTGGTTATCAGTAGTGC
KCNRGV2 NM_173605.1
60 0C F-5’CCTGGTTTTCCAGTGTG
R-5’GCTGAGGCAGGAGAATCACT
Quantitative Real Time PCR:
234
Quantitative real-time PCR was performed in a 96 well format in the
Bio-rad CFX96 Real Time System (BioRad Laboratories, USA). The realtime PCR mixtures consisted of 5 μL cDNA corresponding to ~560 ng total
RNA, 0.1 uM of Real-Time primers or 0.2 nM of Invitrogen primer and 1×
Sso Fast Evagreen Supermix ( BioRad, USA ) in a final volume of 15 μL.
The assay included no template and RT minus controls to detect reagent
contamination and presence of genomic DNA. The thermal profile of the
RT-PCR procedure was repeated for 50 cycles: 1) 95°C for 10 min; 2)10 s
denaturation at 95°C, 40 s annealing at 55°C or 60°C (amplification data
collected at the end of each amplification step); 3) dissociation curve
consisting of 10 s incubation at 95°C, 5 s incubation at 65°C, a ramp up to
95°C. (Bio-rad CFX96, USA). Melt curves were used to validate product
specificity. All samples were amplified in triplicates from the same total
RNA preparation and the mean value was used for further analysis.
Data Analysis:
235
All the Ct values greater that 40 were changed to N/A and were
considered a negative call. Average Ct of technical triplicates was
considered for calculations. Previously validated reference genes were used
for normalization. Normalized Δ Ct for each gene was calculated using the
formula:
Δ Ct = CtGOI – CtAVG HKG
Relative expression for each gene was calculated using the formula:
Efficiency ^
- (Δ Ct)
Fold-change for each gene was determined using:
Fold Change = Δ CtExpt / Δ CtControl
For the fold regulation, negative inverse of fold change less than 1
(meaning that the gene is down regulated) is calculation. Thus, changing the
fractional number into a whole number. It is important to note that the
236
values of the numbers are the same, only the representation of the value has
changed.
Non-parametric Mann-Whitney test was done to determine the
significance of fold difference between groups ( p < 0.05 was considered as
significant).
237
CHAPTER 5: Gene Expression in Stomach
A. Inflammatory panel
Background:
The process of regulation of the size of fat stores is a complex
process. It involves both central and peripheral tissues (Flier and MaratosFlier, 1998), (Ahima, 2008), (Dobrin et al., 2009) and over 50 secreted
molecules, such as the adipocytic leptin and adiponectin (Baranova et al.,
2006), (Morton and Schwartz, 2011), gastric ghrelin (Gurriarán-Rodríguez et
al., 2011), (Estep et al., 2011) and intestinal cholecystokinin (Chaudhri et al.,
238
2008). Many of these molecules also play a role in various diseases
associated with obesity, particularly, non-alcoholic fatty liver disease
(NAFLD) (Jarrar et al., 2007), (Estep et al., 2011) .
The association of NAFLD with obesity, particularly visceral obesity
has long been recognized (Machado et al., 2006). Although a number of
pathways, such as enhanced oxidative stress, increased susceptibility to
apoptosis, and insulin resistance have been implicated in the pathogenesis of
NAFLD (Kim and Younossi, 2008); little is known about the triggers of the
progression of steatosis to NASH, fibrosis and ultimately cirrhosis. Not all
individuals with fatty liver progress to inflammatory NASH and fibrosis.
Additionally, not all obese patients develop fatty liver or NASH. One
explanation for this maybe the contribution of non-adipose peripheral tissues
to the obesity related NAFLD. Given that the stomach is one of the central
organs of the digestive tract relaying satiety signals to the hypothalamus
(Woods, 2004), (Cummings and Overduin, 2007) and is a source of peptides
with critical roles in energy homeostasis, we hypothesize that gastric tissue
may participate in the development of obesity related NAFLD or its
progression to NASH.
239
Until recently, studies on obesity-related diseases have focused
primarily on adipose tissue. However, adipokines and cytokines released by
expanded adipose tissue into circulation exert their action upon many
peripheral targets. These targets in turn, produce and release other bioactive
molecules, thus, adding to the obesity-specific hormonal milieu that
augments the development of metabolic syndrome and associated
conditions. Stomach is one of the neglected peripheral players in the obesity
phenotype. Fortunately, the discovery of ghrelin and its role in human
metabolism has intensified the studies of hypothalamic control of the
appetite and its contribution to obesity and other weight-related disorders
(Kojima et al., 1999). In 2005, it was found that the ghrelin-encoding gene
also encodes obestatin, which unlike ghrelin, is involved in appetite
suppression (Zhang et al., 2005a). In addition to ghrelin and obestatin, the
stomach is the second largest source, after adipose tissue, of the appetite
inhibiting peptide - leptin (S Cinti et al., 2001), (Cammisotto et al., 2005),
(Cammisotto et al., 2010). The interaction of food with stomach is known to
stimulate the release of various peptides. These peptides then act in an
autocrine and paracrine fashion to regulate food intake (Woods et al., 1998).
240
Despite these observations, studies on the role of stomach in obesity-related
disorders, such as NAFLD, are rare.
In this study, we aim to show that stomach in obese subjects is under
inflammatory stress and furthermore, is actively contributing to the systemic
inflammation itself, thus influencing physiological states of other organs
such as the liver. To investigate this, we performed comparative expression
profiling for 84 genes encoding inflammatory cytokines, chemokines and
their receptors in the stomach tissue of morbidly obese subjects (Appendix
Table 1)
Results:
Clinical and demographic data for patients are summarized in Table
13. All the patients were obese with histologically proven NAFLD. For gene
expression analysis, the following comparisons were performed: Mild or
241
without Hepatic Inflammation vs Advanced Hepatic Inflammation; Mild
Steatosis vs Advanced Steatosis; NASH vs without NASH; Fibrosis
presence vs no Fibrosis; Gastritis vs. no Gastritis.
Table 13: Demographic and clinical characteristics. Patient cohorts
profiled for expression of inflammation and immunity related genes (values
given as %).
Demographic or Clinical
Inflammatory Signal
Parameter
Panel (N=20)
242
BMI (> 40)
100%
Gender (Females)
75%
Race (Caucasian)
80%
Age (>=40)
65%
Advanced Inflammation (score ≥ 3)
50%
NASH
65%
Advanced Steatosis (grade ≥ 3)
60%
243
Gastritis
45%
Fibrosis
75%
Advanced Fibrosis
5%
Steatosis with advanced inflammation
50%
NASH with advanced inflammation
30%
244
Gene expression differences between patients with mild and advanced
inflammation:
When cohorts with mild and advanced liver inflammation were
compared, expression levels for CCL4, CCR5, CXCL2, CXCL6, IFNA2,
IL19, IL1F8 and IL8 were significantly increased (p < 0.05) (Table 14).
Among these cytokines, CCL4, CCR5, IFNA2, IL1F8 and IL8 were also
independently and significantly correlated with hepatic inflammation scores
(p < 0.05) (Table 15). CCL21 and CCL3, on the other hand, were found to
be significantly correlated (p < 0.05) with hepatic inflammation scores, but
did not show significant differential expression in the group-wise
comparisons (p > 0.05) (Table 15).
Gene expression differences between patients with advanced steatosis
and mild or without steatosis:
In patients with advanced steatosis (score ≥ 3), CXCL14, IL1F10 and
IL8RB had significant differential expression (p < 0.05) as compared to
245
those with mild steatosis (score ≤ 2) (Table 14). IL8RB and IL1F10 were
also positively correlated with degree of steatosis (p < 0.05) (Table 15).
Gene expression differences between patients with NASH and without
NASH:
Patients with presence of NASH as compared to those without NASH
showed significant differential expression of CCR3, CCR9, IL1RN, IL8RA
and IL9 (p < 0.05) (Table 14). Correlation coefficient analysis showed some
of the differentially expressed genes, namely CCR3, CCR9, IL1RN, IL8RA
and IL9 to be also positively correlated with NASH (p < 0.05) (Table 15).
Additionally, IL8RB, CXCL12 and CCL1 were also positively and
significantly correlated with NASH (p < 0.05) (Table 15).
Gene expression differences between patients with and without fibrosis:
246
In patients with fibrosis, only CCL17 was significantly upregulated (p
< 0.05) (Table 14). A different set of genes, SCYE1, IL1RN and C5, however
were positively correlated with severity of fibrosis (p < 0.05) (Table 15).
Gene expression differences between patients with and without gastritis:
In patients with gastritis, CCL23, CCL7, CCR4, CCR8, CXCL12 and
CXCL5 were upregulated in cohort of patients compared with those without
gastritis (p < 0.05) (Table 14). These genes, along with CCL18, CXCL13 and
LTB were also independently positively correlated with the presence of
gastritis (p < 0.05) (Table 15).
247
Table 14: List of inflammatory genes significantly upregulated in
gastric tissues of patients with the following pathological conditions:
Advanced Liver Inflammation (score ≥ 3) (N = 10), Advanced Steatosis
(score ≥ 3) (N = 8), NASHǂ (N = 13), Fibrosisǂ (N = 15), Gastritisǂ (N = 9).
ǂComparison was performed to groups of patients without the condition
listed.
Genes
Fold Change
p < 0.05
Advanced liver Inflammation (Score ≥ 3)
CCL4
2.32
< 0.03
CXCL2
2.82
< 0.03
248
CCR5
3.16
< 0.02
IFNA2
3.91
< 0.02
IL19
3.48
< 0.02
IL1F8
4.03
< 0.02
CXCL6
4.3
< 0.03
IL8
4.82
< 0.02
Advanced Steatosis (Grade ≥ 3)
249
IL8RB
1.56
< 0.02
CXCL14
1.77
< 0.03
IL1F10
2.24
< 0.04
NASH ǂ
CCR9
2.96
< 0.04
CCR3
3.44
< 0.04
IL1RN
3.95
< 0.02
250
IL9
9.49
< 0.02
IL8RA
15.04
< 0.04
Fibrosis ǂ
CCL17
4.81
< 0.03
Gastritis ǂ
CXCL12
1.69
< 0.03
CCR8
2.66
< 0.04
251
CCR4
3.36
< 0.01
CCL23
4.3
< 0.01
CCL7
4.7
< 0.03
CXCL5
1.65
< 0.006
Table 15. Correlations between inflammatory gene expression with the
spectrum of NAFLD
252
Genes
Correlation
p < 0.05
Degree of Liver Inflammation
CCL3
0.45
0.041
CCL4
0.47
0.035
IL8
0.45
0.042
CCR5
0.48
0.031
IL8RB
0.50
0.025
253
CCL21
0.50
0.024
IFNA2
0.51
0.024
IL19
0.51
0.021
IL1F8
0.53
0.014
Severity of Steatosis
IL1F10
0.45
0.044
IL8RB
0.49
0.026
254
NASH
CXCL12
-0.44
0.049
CCL1
0.45
0.046
CCR3
0.46
0.040
CCR9
0.46
0.040
IL5
0.44
0.049
IL8RA
0.47
0.033
255
IL8RB/CXCR2
0.44
0.049
IL1RN
0.53
0.015
IL9
0.53
0.015
Severity of Fibrosis
C5
0.46
0.039
SCYE1
0.48
0.030
IL1RN
0.53
0.016
256
Gastritis
LTB
0.44
0.050
CXCL13
0.44
0.050
CCL18
0.44
0.048
CCR8
0.47
0.032
CXCL12
0.49
0.026
CCL7
0.48
0.028
257
CCL23
0.54
0.012
CCR4
0.54
0.012
CXCL5
0.63
0.003
Total Cholesterol (mg/dL)
Triglyceride
0.74
<0.0001
BCL6
-0.55
0.01
C3
-0.52
0.017
258
CCR3
-0.53
0.016
CCR9
-0.33
0.04
CEPBP
0.22
0.04
CXCL11
-0.51
0.019
Triglycerides (mg/dL)
CCL17
0.45
0.042
CEBPB
0.55
0.012
259
CX3CR1
0.51
0.19
IL13RA1
-0.48
0.03
IL17C
0.53
0.016
LTB
0.47
0.035
Fasting Glucose (mg/dL)
IL1F7
0.54
260
0.012
Multiple regression analysis for assessment of the relationship between
between gene expression of target genes, clinical factors and advanced
inflammation, NASH and fibrosis:
To assess the relationship between advanced inflammation (dependent
variable) and expression of target genes and clinical factors (independent
variables), a single equation multivariate regression model was generated. In
this model, four independent variables - CCL21, CCR5, ALT and age were
found to be related with advanced inflammation. CCL21 (β coefficient
=0.0012 ± 0.0003, p < 0.0007) and CCR5 (β coefficient =0.0004 ± 0.0001,p
< 0.0064) were the most strongly and positively related (Table 13) to the
presence of advanced inflammation. These four variables explain 66% of the
variance in the inflammation phenotype (R 2 =0.66). Multivariate regression
model for NASH generated statistically significant model (p < 0.002) with
CCR3, CXCL12, IL1RN, IL8RA, IL8RB and IL5. CXCL12 (β coefficient = 0.0001 ± 0.0000, p>0.07) and IL5 (β coefficient = -0.0003 ± 0.0001, p>0.07)
were negatively related with presence of NASH. This model explained 75%
261
of the variance in NASH phenotype (R 2 = 0.75). The regression model of
advanced fibrosis (p < 0.006) included only IL1RN (p < 0.006) as sole
component and was able to explain for only 34% of the variance in fibrosis
(R2 = 0.34). Interestingly, none of the genes showing differential regulation
(p < 0.05) or significantly correlated with the steatosis cohort, were able to
be related with degree of steatosis, hence, no models were generated.
Table 16: Multiple regression analysis: Best fitting multiple linear
regression models showing the relationship between predictor variables and
the predicted clinical parameter. Regression coefficient β represents Slope
estimate ± Standard Error of the estimate (SE) (p <0.05 is considered
significant).
262
R2
Independent
P-value
P-values of
Regression
Group
of the
independent
Variable
coefficient β
entire
variables
model
(Intercept)
-0.5745 ± 0.8110 0.4896
CCL21
0.0012 ± 0.0003
0.0007
CCR5
0.0004 ± 0.0001
0.0064
Advanced liver
inflammation
0.66
<0.001
AGE
0.0288 ± 0.0162
0.0964
ALT
0.0267 ± 0.0136
0.0688
263
NASH
(Intercept)
0.3545 ± 0.1988
0.0980
CCR3
0.0003 ± 0.0002
0.0626
CXCL12
-0.0001 ± 0.0000 0.0724
IL1RN
0.0003 ± 0.0001
IL5
-0.0003 ± 0.0001 0.0683
IL8RA
0.0655 ± 0.0215
0.0092
IL8RB
0.0004 ± 0.0002
0.0532
264
0.0810
0.75 <0.002
(Intercept)
0.5750 ± 0.2471
0.0318
Advanced
0.35 <0.006
Fibrosis
IL1RN
0.0007 ± 0.0002
0.0063
Discussion and Conclusion:
Liver is one of the major organs involved in lipid metabolism
(Nguyen et al., 2008). However, it has a limited ability to store lipids
(Petersen et al., 2005) with excess lipid buildup resulting in the development
of Non-Alcoholic Fatty Liver Disease (NAFLD). One of the critical thrusts
in the studies of the progression of NALFD has been the search for factors
265
that may influence the progression of steatosis to NASH and liver failure.
According to the multiple hit model of NAFLD, many hits may act in
parallel, particularly gut-derived and adipose tissue–derived factors leading
to inflammatory conditions of the liver. Inflammation, a central player in
NAFLD may then augment the probability of progression of steatosis to
NASH and fibrosis (Tilg and Moschen, 2010).
In the past decade, WAT compartment was considered as a major
source for inflammatory cytokines and chemokines circulating in the blood
of obese patients (Baranova et al., 2007), (Estep et al., 2009), (Samaras et
al., 2010). In addition to adipose, it was suggested that other tissues,
particularly, stomach and intestine may overproduce various soluble
molecules that can contribute to overall inflammatory background and
influence the function of distant organs (Baranova et al., 2007). Our study,
for the first time, showed that various soluble molecules are overproduced in
gastric tissue from morbidly obese patients with various types of histological
changes in the livers. Remarkably, there was a substantial overlap in genes
with significant differential expression (p < 0.05) and genes with significant
correlation (p < 0.05) to the same histological characteristic of NAFLD
266
(Figure 32). Further, distinct and notably, non-overlapping sets of soluble
molecule encoding genes change their expression along with various
features of NAFLD (Figure 33). Importantly, an overlap between sets of
genes significantly correlating (p < 0.05) with specific histological
characteristic of NAFLD or presence of gastritis was minimal (Figure 33).
IL8RB/CXCR2 is a notable exclusion with its over-expression correlating
with steatosis, diagnosis of NASH as well as fibrosis.
267
Figure 32: Venn diagram depicting overlap in results of group
comparison and correlation analysis in different cohorts. There is
substantial overlap in genes with significant differential expression in group
comparisons and genes with significant correlation to the same histological
character.
268
Figure 33: Venn diagram depicting results of the statistical analysis.
Sets of genes significantly correlating (p < 0.05) with specific histological
characteristic of NAFLD or presence of gastritis overlap only minimally.
269
IL8RB/CXCR2 is a receptor for the IL8 chemokine that plays an
important role in liver inflammation, regeneration and repair (Kuboki et al.,
2008), (Zimmermann et al., 2011) as well as in the neutrophil accumulation
in other inflammatory conditions (Zernecke et al., 2008), (Chou et al., 2010).
Increased levels of the gastric expression of IL8RB indicates that in morbidly
obese patients with NASH-associated inflammation, IL8 activation is not
limited to hepatic macrophages as had been shown before (Zimmermann et
al., 2011), but is a system-wide feature. It is plausible, that IL8RB present on
the resident gastric macrophages cells or on neutrophils activates the
neutrophils locally upon its binding to IL8. In turn, activated neutrophils
may then release additional chemokines and/or may enter the liver through
portal circulation and influence the progression of NAFLD (Figure 34). This
premise is also supported by our observation that the expression of IL8, the
270
ligand of IL8RB, positively correlates with advanced hepatic inflammation
(Table 15). Circulatory IL8 levels are reported to increase under oxidative
stress and, in turn, stimulate further increase in levels of oxidant stress
mediators by local recruitment of inflammatory cells (Gómez-Quiroz et al.,
2003) (Figure 34). As an expanding adipose tissue of obese individuals
releases increased levels of IL8 (Jarrar et al., 2007), (Estep et al., 2009), it
may trigger increased expression of gastric IL8 and its receptor IL8RB.
Additionally, studies have shown that free fatty acids (FFA), also increased
in obese individuals, influence expression of IL8 in various peripheral
tissues (Andoh et al., 2000), (Böni-Schnetzler et al., 2009). Thus, the paired
increase in levels of IL8 and its receptor found in the gastric tissue of obese
may act to activate local as well as circulating, thus contributing towards
vicious cycle of inflammation and influencing progression of NAFLD.
271
Figure 34. Inflammation-related genes in stomach and obesity
associated Non-Alcoholic Fatty Liver Disease (NAFLD). In obesity,
increased levels of inflammatory molecules such as IL-1F8 may alter gene
expression in stomach by activating NF-kB. NF-kB is known to activate
gene expression of a number of downstream inflammatory molecules
272
including CCL4, IL8 and CCL21. These inflammatory molecules may then
regulate their own expression in positive feedback loop, thus further
exacerbating inflammatory profile. These molecules can potentially activate
local immune cells and attract additional immune cells. Oxidative stress
triggered by activated immune cells can further add to existing
inflammation. The entry of secreted inflammatory molecules and activated
immune cells into portal circulation may contribute to NAFLD.
The expression levels of anti-inflammatory receptor IL1RN, an
antagonist of IL1A and IL1B, was positively correlated both with the
presence of NASH and with fibrosis (Table 15). In the regression model
predicting advanced fibrosis, IL1RN was the only significant component that
explained 34% of the variance in fibrosis. Additionally, IL1RN significantly
contributed to the regression model predicting NASH (Table 16). These
273
observations are in agreement with a recent report on association of serum
IL1Ra levels and liver IL1RN expression with NASH (Pihlajamäki et al.,
2012). IL1Ra is expressed and secreted by a number of immune cells such as
monocytes, macrophages, and neutrophils as well as epithelial cells and
hepatocytes (Perrier et al., 2006). As its expression is regulated by proinflammatory cytokines, IL1RN is considered to be an acute phase protein
(Gabay et al., 1997) with levels elevated in many inflammatory conditions
(Arend et al., 1998). We hypothesize that increased levels of circulating
and/or local pro-inflammatory cytokines upregulate
gastric
IL1RN
expression either directly or via activated leukocytes (Figure 34). Once
upregulated, IL1Ra may stimulate its own gastric expression by a positive
feedback loop (Figure 34). This mechanism is supported by studies showing
elevated circulating IL1RN in patients with obesity (Gabay et al., 1997) and
NAFLD (Somm et al., 2006), (Pihlajamäki et al., 2012).
Many genes differentially expressed in the gastric tissue of patients
with advanced forms of NAFLD encode chemokines that were previously
shown to play important roles in various inflammatory conditions. For
example, expression levels of both CCL4 chemokine and its receptor CCR5
274
showed significant up-regulation in advanced hepatic inflammation (Table
14) and a positive correlation with the severity of the hepatic inflammation
(p < 0.05) (Table 11). In the multivariate regression model, CCR5 also was
one of the strongest predictors of the severity of hepatic inflammation (Table
13). CCL4 attracts natural killer cells, monocytes and a variety of other
immune cells (Andreas Beilhack and Rockson, 2003). The increased
expression of CCL4 and CCR5 in gastric tissue could be attributed to local
immune cells activated in response to upstream regulators like IL1F8
(Figure 34). In the present study, IL1F8 is also upregulated in stomach tissue
of patients with advanced liver inflammation (Table 14). CCR5 has been
implicated in NASH (Bertola et al., 2010) and hepatic fibrosis (Seki et al.,
2009). Both of these conditions develop almost exclusively in a proinflammatory environment. While the role of CCL4/CCR5 in the
pathogenesis of NAFLD remains to be sketched out, these collective
findings make it an attractive target for further investigation.
It is of importance to note that the sets of inflammatory genes
differentially expressed in the gastritis and the liver pathology cohorts
overlapped minimally (Figure 33). Among the 9 genes whose expression
275
levels correlated with the degree of gastritis (p < 0.05), only CXCL12 was
also found to be associated with NASH. Moreover, expression levels of
CXCL12 positively correlated with the presence of gastritis (r = 0.49, p =
0.02), but negatively with the presence of NASH (r = -0.44, p = 0.04) (Table
15). Therefore, our data indicate that, in obese subjects, the pathogenenic
inflammatory patterns accompanying underlying gastritis differ from the
cytokine and chemokine dependent pathways contributing to the
development of liver pathologies.
The complex interaction of cytokines, chemokines and their receptors
highlighted in this study suggests that the stomach is an integral player in
obesity associated NAFLD. It seems that in obesity, an increase in
inflammatory responses of adipose tissue is paralleled by similar increase in
the tissues involved in satiety response such as gastric tissue. Activated
immune cells embedded in the gastric tissue may recruit additional immune
cells or be released in circulation, and hence amplify the inflammatory
response and promote the development and progression of NAFLD (Figure
34). With stomach being increasingly recognized as an important endocrine
organ in energy homeostasis, it is likely that altered inflammatory profile
276
also influences its endocrine function. This, in turn, may trigger a cascade of
metabolic dysfunction culminating in NAFLD (Figure 34). It remains to be
determined if the complex interaction of inflammatory molecules in gastric
tissue lies upstream or downstream of the intricate network of inflammatory
signaling, which is the hallmark of NAFLD. However, it is evident now that
the stomach is a critical component of metabolic dysfunction, and cannot be
overlooked.
To summarize, in this study we demonstrate an altered pattern of gene
expression of cytokines and chemokines in the gastric tissue of individuals
with obesity and varying degrees of hepatic inflammation and different
forms of NAFLD. Soluble inflammatory molecules produced by the stomach
appear to contribute to obesity related NAFLD. Although the causal links
between these signaling events remains to be determined, we propose that
gastric tissue, in particular, the fundic tissue as being an integral player in
the signaling pathways associated with both obesity and NAFLD.
277
B. Obesity Panel
Background:
Energy balance is a highly regulated process with a plethora of
hormones, cytokines and neurotransmitters regulating the process. Despite
fluctuations in daily food/energy intake, there is surprisingly little variation
in day to day weight (Keesey and Powley, 2008). This is attributed to the
process of energy homeostasis that integrates signals from the peripheral
organs with those from central organs (Ahima and Flier, 2000).
Additionally, energy homeostasis also integrates signals from other external
regulators of food intake, such as the presence of food, habits or social
behavior (Gale et al., 2004).This system wide crosstalk among various
organs suggests a highly regulated and redundant network balances appetite
and satiety. However, in an environment of positive energy balance this
homeostatic process appears to be overwhelmed resulting in excess fat
deposition.
278
The classical fat and energy storage organ is AT (Trayhurn and
Beattie, 2001). Notably, under conditions of chronic positive energy, the
storage capacity of this organ is eventually exceeded, resulting in ectopic fat
deposition (Heilbronn et al., 2004). Ectopic fat refers to the fat deposited at
sites other than adipose tissue such as liver (Ryysy et al., 2000), pancreas
(Pinnick et al., 2008), skeletal muscle (Goodpaster and Kelley, 1998) and
even the cardiovascular system (Montani et al., 2004). Ectopic fat can affect
the tissue in three possible ways: (1) First, the size of fat surrounding organs
(periorgan fat) may increase substantially resulting in simple physical
compression (2) Second, periorgan fat cells may secrete various locally
acting substances (3) Third, lipid accumulation in non-adipose cells may
lead to cell dysfunction or cell death, a phenomenon known as lipotoxicity
(Montani et al., 2004).
Given the effect of chronic positive energy on multiple tissues,
multiple peripheral signals are expected to be altered. For instance, the local
effect of ectopic fat within skeletal muscle, pancreas and liver, has been
documented in several studies. Increased intramyocellular lipid is shown to
cause defects in insulin signaling in muscle by reducing insulin-stimulated
279
muscle glucose transport activity and glycogen synthesis (Goodpaster and
Kelley, 1998). Infiltration of fat in and around the pancreas referred to as
pancreatic lipomatosis has been shown to affect insulin secretion (Pinnick et
al., 2008). Ectopic storage of fat in the liver is also of pathophysiological
importance. A well known hepatic manifestation of obesity is non-alcoholic
fatty liver disease (NAFLD). Studies have shown hepatic fat to be associated
with insulin resistance and impaired suppression of hepatic glucose
production by insulin (Ryysy et al., 2000), (Seppälä-Lindroos et al., 2002).
Interestingly, studies exploring the participation of stomach in energy
homeostasis and the effects of dysregulated energy homeostasis on stomach
function, are scarce (Inui et al., 2004), (Badman and Flier, 2005). In fact, the
GI tract is the largest endocrine organ of the body, measured either in terms
of the number of endocrine cells it contains or in terms of the number of
hormones those cells secrete (Rehfeld, 1998). Genes for over 30 hormones
and prohormones are expressed in the GI tract, and it secretes, including
products of alternative splicing, some 100 or more active peptides (Rehfeld,
2004). The stomach is an important organ of the gastrointestinal system
(GI). Aside from its obvious role in the digestion and absorption of nutrients,
280
the stomach is involved in important endocrine pathways that communicate
with regulators of energy homeostasis (Inui et al., 2004), (Badman and Flier,
2005). One of the best examples of endocrine role of stomach is: stomachderived ghrelin, obestatin and leptin (Date et al., 2000), (Nakazato et al.,
2001), (Wren et al., 2001), (Lee et al., 2002). Ghrelin has an appetite
stimulating role (Kojima et al., 1999) while, obestatin (Zhang et al., 2005b)
and leptin (Picó et al., 2003) have an appetite suppressing function.
Although, the association between obesity and ectopic fat deposition
in various tissues has received intensive attention, the effect of obesity on
stomach and the role of stomach in metabolic dysfunction have been largely
overlooked. In this study, we explore this relationship by gene expression
profiling of energy metabolism associated genes in the stomach tissue of
morbidly obese patients with NAFLD.
Results:
281
Gene expression profiling of 84 energy metabolism associated genes
(Appendix Table 2) in the stomach tissue of morbidly obese patients with
NAFLD (N=24) was carried out. Clinical and demographic data for patients
are summarized in Table 17. All patients were morbidly obese with
histologically proven NAFLD.
Gene expression differences between patients with mild/no steatosis and
advanced steatosis:
Group comparison of advanced steatosis (Grade ≥ 3) vs mild or no
steatosis (Grade < 3) found CPE (-1.88, p<0.05) and IL1B (-2.5, p=0.05) to
have decreased expression in subjects with severe steatosis. Notably, no
linear correlation was found between the genes and severity of steatosis.
Gene expression differences between patients with no NASH and
presence of NASH:
282
Among patients with NASH compared to those without NASH, IL1R1
(1.99, p=0.04), OPRM1 (2.56, p=0.02), SIGMAR1 (3.13, p=0.03), THRB
(1.94, p=0.02) and ZFP91 (3.09, p=0.01) were upregulated. Pearson’s
analysis also showed positive correlation of IL1R1 (r = 0.42, p = 0.03),
OPRM1 (r = 0.43, p = 0.034), SIGMAR1 (r = 0.51, p=0.009), THRB (r =
0.40, p = 0.05) and ZFP91 (r = 0.43, p = 0.03) with presence of NASH.
Gene expression differences between patients with mild/no hepatic
inflammation and advanced hepatic inflammation:
In patients with advanced hepatic inflammation (score ≥ 3) vs mild
inflammation, 21 genes (p < 0.05) were found to have increased gene
expression (fold change: 2.1 – 6.9) (Table 1). Among these genes ADRA2B,
CNR1 and LEP were also found to be mildly (r = 0.5) but significantly
correlated (p < 0.05) with the degree of hepatic inflammation. In addition to
these genes, IL1A and OPRM1 were also correlated with the degree of
hepatic inflammation (r = 0.5, p < 0.05).
283
Gene expression differences between patients with mild/no hepatic
fibrosis and advanced hepatic fibrosis:
Comparison of subjects with advanced fibrosis (score ≥ 3) vs mild or
no fibrosis (score < 3) showed the NTS (6.7, p < 0.05) and OPRK1 (5.6, p <
0.05) genes to be greater than 5 fold upregulated in subjects with severe
fibrosis. Correlation analysis found GHR and IL1A to be mildly but
significantly correlated with the severity of fibrosis (r = 0.5, p < 0.05)
Gene expression differences between patients with no gastritis and
presence of gastritis:
To determine the influence of gastritis on gene expression, group
comparison was also carried out in subjects with gastritis compared to those
without gastritis. A different set of genes were found to be differentially
expressed. The expression of C3 (-2.12, p < 0.05) was downregulated while
the expression of DRD1 (2.6, p < 0.05) was upregulated. Pearson’s
284
correlation shows DRD1 (r = 0.46, p = 0.02) is correlated with the presence
of gastritis.
Table 17: Clinical and Demographic Data
Demographic or Clinical Parameter
% (N=24)
BMI > 40
83 (N=20)
Gender (Females)
79 (N=19)
Race (caucasian)
67 (N=16)
Age (≥40)
67 (N=16)
285
Advanced Inflammation (score ≥ 3)
54 (N=13)
NASH
63 (N=15)
Advanced Steatosis
42 (N=10)
Gastritis
46 (N=11)
Any Fibrosis
79 (N=19)
Steatosis with advanced inflammation
96 (N=23)
NASH with advanced inflammation
75 (N=18)
286
Table 18: List of obesity related genes significantly upregulated in
gastric tissues of patients with the following pathological conditions:
Advanced Liver Inflammation (score ≥ 3) (N = 13), Advanced Steatosis
(score ≥ 3) (N = 10), NASHǂ (N = 15), Fibrosisǂ (N = 19), Gastritisǂ (N =
11). ǂComparison was performed to groups of patients without the condition
listed.
Genes
Fold Regulation
P value
Advanced Steatosis (grade ≥ 3)
CPE
-1.8
287
0.04
IL1B
-2.5
0.05
NASH
IL1R1
1.99
0.04
OPRM1
2.65
0.02
SIGMAR1
3.13
0.03
THRB
1.94
0.02
ZFP91
3.09
0.01
288
Advanced Hepatic Inflammation (score ≥ 3)
ADCYAP1
5.5
0.04
ADRA2B
2.1
0.02
BDNF
3.5
0.03
CNR1
5.1
0.001
CNTFR
3.2
0.02
GALR1
2.5
0.04
289
GH2
5.1
0.01
GRPR
4.1
0.004
IAPP
2.5
0.03
LEP
2.3
0.04
LEPR
2.3
0.02
MC3R
4.1
0.02
NMB
2.4
0.04
290
NMU
3.9
0.004
NMUR1
6.9
0.03
PPARGC1A
4.1
0.006
PRLHR
4.2
0.02
RAMP3
2.5
0.02
SIGMAR1
2.3
0.01
SSTR2
3.2
0.04
291
UCN
4.9
0.04
Fibrosis presence
NTS
6.7
0.02
OPRK1
5.6
0.01
Gastritis presence
C3
-2.1
0.04
DRD1
2.6
0.02
292
Correlation Analysis with Clinical Variables:
BMI was positively correlated with weight (kg) (r = 0.82, p<0.05),
and CPE and negatively correlated with ZFP91 (Table 19). Fasting glucose
levels were positively correlated with AGRP, NMB, THRB, TNF and UCP1.
Serum triglyceride levels were positively correlated with 14 genes (Table
19) including ADIPOQ and APOA4 and negatively correlated with CPD
(Table 19).
293
Table 19: Correlations between obesity gene expression and spectrum of
NAFLD.
Gene
Correlation (r)
p value
NASH
IL1R1
0.42
0.03
OPRM1
0.43
0.034
SIGMAR1
0.51
0.009
294
THRB
0.40
0.05
ZFP91
0.43
0.03
Degree of Inflammation
ADRA2B
0.45
0.02
CNR1
0.43
0.03
LEP
0.40
0.04
IL1A
0.43
0.03
295
OPRM1
0.42
0.03
Fibrosis
GHR
0.42
0.03
IL1A
0.48
0.01
Gastritis
DRD1
0.46
0.02
GHRL
0.41
0.04
BMI
296
CPE
0.47
0.01
ZFP91
-0.48
0.01
Fasting Glucose (mg/dL)
AGRP
0.52
0.008
NMB
0.41
0.04
THRB
0.55
0.005
TNF
0.42
0.04
297
UCP1
0.65
0.0005
Serum Triglycerides (mg/dL)
ADCYAP1R1
0.48
0.01
ADIPOQ
0.50
0.01
APOA4
0.44
0.02
CNTFR
0.41
0.04
CALCA
0.44
0.02
298
DRD2
0.49
0.01
GCGR
0.43
0.03
GH2
0.41
0.04
GLP1R
0.56
0.003
IL6
0.42
0.03
NMUR1
0.48
0.01
NTRK2
0.40
0.04
299
PPARGC1A
0.42
0.04
PRLHR
0.43
0.03
CPD
-0.42
0.03
Discussion and Conclusion:
Obesity is characterized by excessive adipose tissue mass as well as
the storage of fat in ectopic sites such as muscle and liver (Unger, 2003).
300
These ectopic sites are, however, are not adapted for lipid storage and even
modest increases in lipid concentration manifests as tissue dysfunction
(lipotoxicity) (Snel et al., 2012). A number of studies have reported altered
orexigenic and anorexigenic gene expression profile in adipose tissue
(Gómez-Ambrosi et al., 2004), (Dahlman et al., 2005), (Shankar et al.,
2010). In this study, we see altered expression of energy homeostasis genes
in the stomach tissue of morbidly obese patients with NAFLD.
CPE (-1.88, p<0.05) a gene involved in the regulation of energy
expenditure and IL1B (-2.5, p=0.05) an anorectic gene were both found to
have decreased expression in subjects with severe steatosis as compared to
those with no/mild steatosis. Carboxypeptidase E (CPE) an important
exopeptidase encoded by CPE, is involved in the post-translational
processing of many of the prohormones and neuropeptides that have central
roles in energy homeostasis (Figure 35) (Lacourse et al., 1997), (FriisHansen et al., 2001), (Naggert et al., 1995), (Rovere et al., 1996). For
instance, TRH, GCG, BDNF, PMCH and CCK are implicated in feeding
behavior, while PRKAA1 and INS are involved in fatty acid homeostasis.
CPE (carboxypeptidase E) is found in the secretory granules of endocrine
301
and neuroendocrine cells and is expressed in many tissues, including the
brain, pituitary, pancreas, and adrenals. It is strongly implicated in obesity
(Chen and Garg, 1999) via its role in the maturation of a variety of peptide
hormones associated with the control of body fat mass and nutrient
partitioning (review:Fricker and Leiter, 1999). A study by Plum et al.,
showed a 50% decrease in fasting Cpe mRNA in the arcuate nucleus of mice
rendered obese by a high-fat diet (HFD) (Plum et al., 2009). Additionally,
overexpression of Cpe resulted in decreased appetite and reduced locomotor
activity indicating an additional anorexigenic role (Plum et al., 2009).
Therefore, decreased CPE expression in the stomach tissue of subjects with
advanced steatosis may be indicative of failed appetite suppression. This in
turn, may contribute to a positive energy balance and contribute to ectopic
fat deposition.
302
Figure 35: Role of CPE in processing of propeptides involved in energy
homeostasis. (GCG- glucagon; GAST- gastrin; CCK- cholecystokinin;
NTS- neurotensin; BDNF-; PMCH-pro-melanin-concentrating hormone;
INS- insulin; PRKAA1-protein kinase, AMP-activated, alpha 1 catalytic
subunit; TRH-thyrotrpoin releasing hormone (STRING was used to
determine known protein-protein interactions. The filter used was
‘experimental evidence’ of medium confidence)
303
Interleukin 1 beta (IL-1β) on the other hand, is an inflammatory
cytokine with anorexigenic function (Reyes and Sawchenko, 2002),
(Kornman, 2006), (DeBoer et al., 2009). IL1β knockout or deficient mice
have been shown to be susceptible to mature - onset obesity (Matsuki et al.,
2003), (García et al., 2006). IL1β is also known to inhibit gene expression of
orexigenic ghrelin (Asakawa et al., 2001). Thus, the decreased expression of
CPE and IL1B may indicate a downregulation of the anorexigenic signaling
pathways in obese patients with advanced steatosis as compared to those
with mild/no steatosis. Interestingly, IL-1 genes are reported to be among the
first genes to be activated in response to tissue injury (Kornman, 2006). The
observed decrease in expression of IL1B in obese patients with advanced
steatosis may indicate a complex response in pre-existing low grade chronic
inflammation characteristic of obesity.
Obesity is now well known to be associated with low grade
metainflammation induced by a metabolic surplus, with the liver being one
of the primary organs affected. In our study, 21 genes in the stomach of
obese subjects (BMI ≥ 30) were found to have increased gene expression
304
(fold change: 2.1 – 6.9) (Table 1) among patients with advanced hepatic
inflammation vs those with no/mild hepatic inflammation. Amongst these,
ADRA2B, CNR1, LEP were also significantly correlated (r = 0.5, p < 0.05)
with the degree of hepatic inflammation.
Leptin (LEP) is an inflammatory cytokine, whose production is
acutely increased during inflammation in response to other inflammatory
cytokines such as TNF-α and IL-1β. Although, WAT is the major site of
LEP expression (Zhang et al., 1994), LEP has also been detected in the
gastric epithelium and in the glands of the gastric fundic mucosa in rats
(Bado et al., 1998). There is a large body of evidence implicating adipose
derived leptin in the scheme of NAFLD. In a study by Chitturi et al., LEP
levels in patients with biopsy proven NASH were twice those found in
controls matched for BMI, gender, and age (Chitturi et al., 2002). Our study
shows similar trends in the stomach tissue. LEP is differentially expressed in
the stomach in advanced hepatic inflammation and is positively correlated
with the degree of hepatic inflammation in morbidly obese patients with
NAFLD (Table 18). LEP also functions as an anorexigenic hormone and has
been shown to stimulate gene expression of neurotensin (NTS) (Sahu, 1998).
305
Thus, it is plausible that stomach derived LEP induces expression of NTS,
which in turn plays a role in liver fibrosis.
NTS is a 13 amino acid neuropeptide found in the central nervous
system and the gastrointestinal tract. NTS is not only anorexigenic but also
has a wide spectrum of actions on the gut-liver axis, influencing bile acid
secretion, enterohepatic circulation, intestinal motility, blood flow, secretion,
nutrient absorption and immune response (Assimakopoulos et al., 2004),
(Assimakopoulos et al., 2011). Notably, NTS has been shown to activate
diverse intracellular signals in hepatocytes, including their mitogenic
induction (Hasegawa et al., 1994). Thus, NTS serves as an anorexigenic
neurotransmitter, as well as a gastrointestinal hormone which is involved in
the regulation of liver regeneration (Hasegawa et al., 1994). Our study
shows increased expression of NTS in the stomach of subjects with advanced
fibrosis as compared to those with no/mild fibrosis, similar to the reported
role of NTS in liver regeneration.
Another important gene with differential expression in the gastric
tissue of morbidly obese patients with advanced hepatic inflammation and a
306
positive correlation with the degree of hepatic inflammation is CNR1/CB1.
CNR1 receptor of endocannabinoid system is located throughout the
gastrointestinal tract (Pacher et al., 2006). It is implicated in de novo fatty
acid and cholesterol synthesis pathways, in both liver and adipose tissue
(Zhao et al., 2010). Activation of this receptor in the liver increases de novo
synthesis of fatty acids in mice (Pagano et al., 2007). Additionally,
genetically obese Zucker rats have been shown to have up-regulated Cnr1 in
the adipose tissue (Bensaid et al., 2003). Further, CNR1 antagonists improve
metabolic parameters by exerting an anorexigenic effect and reducing
inflammation (Zhao et al., 2010).
Interestingly, CNR1 is also implicated in the pathophysiology of acute
and chronic liver disease, including inflammation, fibrogenesis and steatosis.
CNR1, which is faintly expressed in normal livers has been shown to be
substantially up-regulated in liver injury and in liver cirrhosis of various
etiologies (Izzo and Camilleri, 2008). Thus, the observed expression of
gastric CNR1 in morbidly obese patients with advanced hepatic
inflammation in the present study maybe attributable to the involvement of
307
gastric CNR1 in fatty acid metabolism and hepatic inflammation under
conditions of positive energy balance.
The anorexigenic effect of administering general opioid receptor
antagonists has been well documented (Czyzyk et al., 2012). However, the
role of endogenous opioids and opioid receptor systems in energy balance
has only recently been understood. Few studies have focused on this aspect
and there are conflicting results. The mu-opioid receptor gene (OPRM1,
located on chromosome 6q24) is one of four genes whose protein products
bind to endogenous opioids. OPRM1 is a member of the large family of
transmembrane G-protein-coupled receptors. It is expressed in both the brain
and the periphery (Heber and Carpenter, 2011). OPRM1 is believed to
modulate the efficiency of energy storage during high-fat diets via the
regulation of energy partitioning (Zuberi et al., 2008), (Xu et al., 2009)
playing a crucial role in energy homeostasis. Its genetic deletion has been
shown in one study to provide protection from diet-induced obesity (Zuberi
et al., 2008), while another study reported increased body weight in adult
rodents attributed to decreased hypothalamic NPY (Han et al., 2006). The
OPRM1 gene has also been implicated in the susceptibility to diabetes (Sale
308
et al., 2004), (Gallagher et al., 2006) by inducing insulin resistance (Cheng
et al., 2003). Insulin, in turn has anti-inflammatory actions and insulin
resistance by default, would lead to enhanced circulating levels of
proinflammatory cytokines (Dandona et al., 2004), (Dandona et al., 2007),
(Heber and Carpenter, 2011), subsequently triggering hepatic inflammation.
The observed positive correlation of gastric OPRM1 with the degree of
hepatic inflammation is in agreement with the previous study showing
altered inflammatory cytokine profiles in the stomach of morbidly obese
patients
with
NAFLD
(Rohini
Mehta
Inflammation(accepted)).
309
et
al.,2013
Mediators
of
Figure 36: Differentially expressed genes in obese patients with NAFLD.
Differentially expressed genes (p < 0.05) are unique and non-overlapping
among the cohorts analyzed. Genes showing fold up-regulation are indicated
by  symbol, while genes showing a fold down-regulation are indicated by 
symbol
310
Thus, gastric tissue is yet another tissue involved in obesity and the
spectrum of NAFLD. Importantly, the gene expression pattern of gastric
tissue is unique and non – overlapping with the histological character of
NAFLD in obese patients (Figure 33 and Figure 36).
Methods:
Samples:
This study was approved by Internal Review Board. After informed
consent, 24 morbidly obese NAFLD patients undergoing laparoscopic sleeve
gastrectomy were included for the study. For each patient, a large number of
clinical and laboratory variables were available (Table x and Table 17).
Other chronic liver diseases were excluded by negative serology for hepatitis
B and C, no history of toxic exposure and no other cause of chronic liver
disease. Excessive alcohol consumption (>10 grams/day in women and >20
311
grams/day in men) was also excluded. No patients were taking TZDs or
medication for gastritis. From each patient, discarded gastric tissue during
sleeve gastrectomy was obtained and nap frozen with liquid nitrogen. Every
gastric sample was also evaluated histologically for the presence of gastritis.
As noted, samples were flash frozen in liquid nitrogen, placed in -80°C.
Gene expression profiling experiments were performed using fundic samples
collected from remaining sleeve gastrectomy specimens. Samples were
profiled for expression levels of 84 genes encoding inflammatory cytokines,
chemokines, their receptors and other components of inflammatory cascades
(Appendix Table 1) and 84 genes related to obesity and energy homeostasis
(Appendix Table 2) using RT² Profiler PCR Arrays (Qiagen, USA)
respectively.
For each patient, a liver biopsy was also performed and read by the
central hepatopathologist. Before histopathological evaluation, each liver
biopsy specimen was formalin-fixed, sectioned, and stained with
hematoxylin-eosin and Masson trichrome. The slides were reviewed
following a predetermined histologic grading system; the extent of steatosis
was graded as an estimate of the percentage of tissue occupied by fat
312
vacuoles as follows: 0 = none, 1 = <5%, 2 = 6-33%, 3 = 34-66%, 4 = >66%.
Other histological features evaluated in H & E sections included portal
inflammation, lymphoplasmacytic lobular inflammation, polymorphonuclear
leukocyte inflammation, Kupffer cell hypertrophy, apoptotic bodies, focal
parenchymal necrosis, glycogen nuclei, hepatocellular ballooning, and
Mallory-Denk bodies. Patients who had hepatic steatosis (with or without
non-specific inflammation) or NASH were considered to have NAFLD.
NASH was defined as steatosis, lobular inflammation, and ballooning
degeneration with or without Mallory Denk bodies, and with or without
fibrosis.
RNA extraction and reverse transcription:
Total RNA was extracted from fundic gastric tissue samples (N=20) using
RNeasy kit (Qiagen, USA) according to manufacturer’s instructions. To
determine the quantity and purity of the extracted RNA, absorbance at 260
nm (A260) and 280 nm (A280) was measured by the GeneQuant1300
spectrophotometer (GE Healthcare, USA). RNA of A260/A280 ratio
between 1.8 - 2.1 was considered of high purity. RNA integrity was
313
confirmed by gel electrophoresis using 1% agarose with ethidium bromide.
RNA with sharp, clear 28S and 18S ribosomal RNA (rRNA) bands and the
intensity of 28S rRNA band approximately twice as intense as the 18S
rRNA band were used as parameters to evaluate the integrity of total RNA.
560 ng of extracted total RNA was reverse transcribed using RT2 first strand
kit (Qiagen, USA). According to manufacturer's protocol, total RNA was
treated to eliminate genomic DNA. Both random hexamers and oligo-dT
primers were used to prime reverse transcription performed as recommended
by enzyme manufacturer (Invitrogen, USA).
Quantitative Real Time PCR analysis:
Quantitative real-time PCR was performed in 96 well PCR format
using Bio-Rad CFX96 Real Time System (BioRad Laboratories, USA) with
a ramp speed of 1 0C/sec. ‘Inflammatory cytokines and receptor’ RT²
Profiler PCR Arrays (Qiagen, USA) and ‘Obesity and energy homeostasis’
RT² Profiler PCR Arrays (Qiagen, USA) were used to simultaneously
examine the mRNA levels of 96 genes encoding for (1) inflammatory
cytokines, their receptors and intracellular components of inflammatory
314
cascades (PAHS-011Z) along with five housekeeping genes (2) orexigenic,
anorexigenic and energy expenditure related genes and their receptors
(PAHS-017Z) along with five housekeeping genes respectively. The realtime PCR mixtures consisted of 1 μL cDNA, 7.5 μL of RT PCR master mix
(Qiagen, USA) in a final volume of 25 μL. The thermal profile of the RTPCR procedure was repeated for 50 cycles: 1) 95°C for 10 min; 2) 10 s
denaturation at 95°C, 15 s annealing at 60°C (amplification data collected at
the end of each amplification step); 3) dissociation curve consisting of 10 s
incubation at 95°C, 5 s incubation at 65°C, a ramp up to 95°C (Bio-Rad
CFX96 Real Time System, USA). Melt curves were used to validate product
specificity.
Data Analysis:
All the Ct values greater than 40 were changed to N/A and were
considered to be a negative call. Ct values of control wells (genomic DNA
control, reverse transcriptase control, positive PCR control) were examined
for each plate. Ct
GDC
(Ct of Genomic DNA control) with a value greater
than 35 was considered as too low to affect gene expression profiling results.
315
The RT2 First Strand Kit’s built-in external RNA control was used as control
to help monitor the efficiency of reverse transcription. Another built-in
external PCR reaction control consisted of DNA spiked in PPC well which
would control for PCR efficiency. Using the following formula, influence of
inhibition was calculated:
Δ Ct = AVG (CtRTC) – AVG (CtPPC)
A value of ΔCt < 5 was considered as an indication of no apparent
inhibition. To control for interplate variability CtPPC was compared. Average
CtPPC should be 20 ± 2 on each PCR Array and should not vary by more than
two cycles between PCR Arrays being compared. The average Ct value of
previously
validated
reference
genes
(HKG)
was
considered
for
normalization. Normalized Δ Ct for each pathway focused gene in each plate
was calculated using the formula:
Δ Ct = CtGOI – CtAVG HKG
Relative expression was calculated for each gene using the formula:
316
Efficiency ^
- (Δ Ct)
Fold-change for each gene was calculated as:
Fold Change = Δ CtExpt / Δ CtControl
Statistical Analysis:
The study was designed to uncover the changes in gene expression in the
stomach of patients with more advanced forms of NAFLD as compared to
these with less advanced forms. The comparisons were performed between
the following different cohorts:
1. Mild
or
No
Hepatic
Inflammation
vs.
Advanced
Inflammation;
2. Mild Steatosis vs Advanced Steatosis;
3. Histologic NASH vs. NAFLD without histologic NASH;
4. Hepatic Fibrosis vs. NAFLD without histologic fibrosis;
317
Hepatic
To assess the significance of gene expression differences between the
compared
groups,
univariate
analysis
was
performed
using
the
nonparametric Mann-Whitney test. To determine whether two variables covary, and to measure the strength of any relationship between, Spearman’s
and Pearson’s Correlation analysis was carried out to find out significantly
correlated genes. Multiple regression analysis was also used to understand
which among the independent variables (target gene expression and clinical
variables) are related to the dependent variable (pathological condition), and
to explore the forms of these relationships.
318
APPENDIX TABLE 1
Gene Description
Symbol
Biological Function
Genes involved in inflammatory response
ATP-binding cassette, subfamily F (GCN20), member ABCF1
1
B-cell CLL/lymphoma 6
BCL6
Complement component 3
C3
Complement component 4A
C4A
(Rodgers blood group)
CCAAT/enhancer
binding CEBPB
319
Protein
plays
a
role
in
enhancement of
protein
synthesis and the inflammation
process and may be regulated by
tumor necrosis factor-alpha.
Is a zinc finger transcription
factor and acts as a sequencespecific repressor of transcription.
Shown
to
modulate
the
transcription
of
STARTdependent IL-4 responses of B
cells.
Plays a central role in the
activation of complement system
and is required for classical and
alternative complement pathways
activation. It is a mediator of local
inflammatory process.
Is part of the classical activation
pathway and a mediator of local
inflammatory process.
Is a bZIP transcription factor and
protein (C/EBP), beta
is important in the regulation of
genes involved in immune and
inflammatory responses. It has
been shown to bind to regulatory
regions of several acute-phase and
cytokine genes.
Level of this protein in plasma
increases greatly during acute
C-reactive
protein,
CRP
phase response to tissue injury,
pentraxin-related
infection, or other inflammatory
stimuli.
Inhibits generation of IL-1-beta
by interacting with caspase-1 and
Caspase recruitment domain
CARD18 preventing its association with
family, member 18
RIP2. Down-regulates the release
of IL1B.
Receptor for interleukin-1 alpha
(IL-1A), beta (IL-1B), and
interleukin-1 receptor antagonist
Interleukin 1 receptor, type I IL1R1
protein (IL-1RA). Binding to the
agonist leads to the activation of
NF-kappa-B.
This protein inhibits the activities
of interleukin 1, alpha (IL1A) and
Interleukin
1
receptor
interleukin 1, beta (IL1B), and
IL1RN
antagonist
modulates a variety of interleukin
1
related
immune
and
inflammatory responses.
Receptor for extracellular ATP >
UTP and ADP. Is a receptor for
leukotriene
B4,
a
potent
Leukotriene B4 receptor
LTB4R
chemoattractant
involved
in
inflammation
and
immune
response.
Fifth component of complement,
Complement component 5
C5
which plays an important role in
inflammatory and cell killing
320
Chemokine (C-X-C motif)
CXCR2
receptor 2
Toll interacting protein
TOLLIP
processes. It is a mediator of local
inflammatory process. C5a also
stimulates the locomotion of
polymorphonuclear
leukocytes
(chemokinesis) and direct their
migration toward
sites
of
inflammation
(chemotaxis).
Defects in this gene have also
been linked to a susceptibility to
liver fibrosis.
Receptor for interleukin 8 (IL8)
and transduces the signal through
a G-protein activated second
messenger system.
Mediates
neutrophil migration to sites of
inflammation.
Ubiquitin-binding protein that
interacts with several Toll-like
receptor (TLR) signaling cascade
components.
Chemokine Genes and Receptors
Chemokine
ligand 1
(C-C
motif)
Chemokine
ligand 11
(C-C
motif)
Chemokine
ligand 13
(C-C
motif)
CCL1
CCL11
CCL13
321
Is a CXC subfamily of cytokine,
secreted by activated T cells.
Displays chemotactic activity for
monocytes but not for neutrophils.
It binds to the chemokine receptor
CCR8.
Is a CC cytokine. Displays
chemotactic
activity
for
eosinophils, but not mononuclear
cells or neutrophils. Binds to
CCR3.
Is a CC cytokine with chemotactic
activity
for
monocytes,
lymphocytes,
basophils
and
Chemokine
ligand 15
(C-C
motif)
Chemokine
ligand 16
(C-C
motif)
Chemokine
ligand 17
(C-C
motif)
CCL15
CCL16
CCL17
Chemokine (C-C motif)
ligand 18 (pulmonary and CCL18
activation-regulated)
322
eosinophils, but not neutrophils.
The cytokine encoded by this
gene is chemotactic for T cells
and monocytes and induces Nacetyl-beta-D-glucosaminidase
release in monocytes. It induces
changes in intracellular calcium
concentration in monocytes and is
thought to act through the CCR1
receptor.
Shows chemotactic activity for
lymphocytes and monocytes but
not neutrophils. Induces a calcium
flux in THP-1 cells that were
desensitized by prior expression
to RANTES. The expression of
this gene is upregulated by IL-10.
Displays chemotactic activity for
T
lymphocytes,
but
not
monocytes or granulocytes. The
product of this gene binds to
chemokine receptors CCR4 and
CCR8. This chemokine plays
important roles in T cell
development in thymus as well as
in trafficking and activation of
mature T cells.
Displays chemotactic activity for
naive T cells, CD4+ and CD8+
T
cells
and
nonactivated
lymphocytes,
but
not
for
monocytes or granulocytes. This
chemokine attracts naive T
lymphocytes toward dendritic
cells and activated macrophages
in lymph nodes. It may play a role
in both humoral and cell-mediated
Chemokine
ligand 19
(C-C
motif)
Chemokine
ligand 2
(C-C
motif)
Chemokine
ligand 20
(C-C
motif)
Chemokine
ligand 21
(C-C
motif)
Chemokine
ligand 23
(C-C
motif)
Chemokine
ligand 24
(C-C
motif)
CCL19
CCL2
CCL20
CCL21
CCL23
CCL24
323
immunity responses.
Play a role in normal lymphocyte
recirculation and homing. It also
plays an important role in
trafficking of T cells in thymus,
and in T cell and B cell migration
to secondary lymphoid organs. It
specifically binds to chemokine
receptor CCR7.
A
CC
chemokine
with
chemotactic
activity
for
monocytes and basophils but not
for neutrophils or eosinophils.
Involved in formation and
function of the mucosal lymphoid
tissues by attracting lymphocytes
and dendritic cells towards
epithelial
cells.
C-terminal
processed forms have been shown
to be equally chemotactically
active for leukocytes.
Chemotactic
in
vitro
for
thymocytes and activated T cells,
but not for B cells, macrophages,
or neutrophils. May also play a
role in mediating homing of
lymphocytes
to
secondary
lymphoid organs. It is a high
affinity functional ligand for
chemokine receptor 7 (CCR7).
Chemotactic activity on resting T
lymphocytes and monocytes,
lower activity on neutrophils and
no activity on activated T
lymphocytes.
The cytokine encoded by this
gene displays chemotactic activity
Chemokine
ligand 25
(C-C
motif)
Chemokine
ligand 26
(C-C
motif)
Chemokine
ligand 3
(C-C
motif)
Chemokine
ligand 4
(C-C
motif)
Chemokine
ligand 5
(C-C
motif)
CCL25
CCL26
CCL3
CCL4
CCL5
324
on resting T lymphocytes, a
minimal activity on neutrophils,
and is negative on monocytes and
activated T lymphocytes. Binds to
CCR3.
Chemotactic activity for dendritic
cells, thymocytes, and activated
macrophages but is inactive on
peripheral blood lymphocytes and
neutrophils. The product of this
gene binds to chemokine receptor
CCR9.
Displays chemotactic activity for
normal
peripheral
blood
eosinophils and basophils. The
product of this gene is one of
three related chemokines that
specifically activate chemokine
receptor CCR3.
Also known as MIPmonokine with inflammatory and
chemokinetic properties. Binds to
CCR1,
CCR4
and
CCR5
receptors.
Monokine with inflammatory and
chemokinetic properties. Binds to
CCR5. The processed form MIP1-beta retains the abilities to
induce
down-modulation
of
surface
expression
of
the
chemokine receptor CCR5.
Chemoattractant
for
blood
monocytes, memory T-helper
cells and eosinophils. Causes the
release of histamine from
basophils
and
activates
eosinophils. Binds to CCR1,
Chemokine
ligand 7
(C-C
motif)
Chemokine
ligand 8
(C-C
motif)
Chemokine
receptor 1
(C-C
motif)
Chemokine
receptor 2
(C-C
motif)
Chemokine
receptor 3
(C-C
motif)
CCL7
CCL8
CCR1
CCR2
CCR3
325
CCR3, CCR4 and CCR5.
This gene encodes monocyte
chemotactic protein 3, a secreted
chemokine
which
attracts
macrophages during inflammation
and metastasis. The protein is an
in vivo substrate of matrix
metalloproteinase 2, an enzyme
which degrades components of
the extracellular matrix.
Chemotactic factor that attracts
monocytes,
lymphocytes,
basophils and eosinophils. May
play a role in neoplasia and
inflammatory host responses.
The ligands of this receptor
include macrophage inflammatory
protein 1 alpha (MIP-1 alpha),
regulated on activation normal T
expressed and secreted protein
(RANTES),
monocyte
chemoattractant protein 3 (MCP3), and myeloid progenitor
inhibitory factor-1 (MPIF-1).
Subsequently transduce a signal
by increasing the intracellular
calcium ions level.
This gene encodes two isoforms
of a receptor for monocyte
chemoattractant
protein-1.
Receptor
mediates
agonistdependent calcium mobilization
and inhibition of adenylyl cyclase.
Binds to eotaxin, eotaxin-3, MCP3, MCP-4, RANTES and MIP-1
delta. Subsequently transduces a
signal
by
increasing
the
Chemokine
receptor 4
(C-C
motif)
Chemokine
receptor 5
(C-C
motif)
Chemokine
receptor 6
(C-C
motif)
Chemokine
receptor 7
(C-C
motif)
CCR4
CCR5
CCR6
CCR7
326
intracellular calcium ions level.
It is a receptor for the CC
chemokine - MIP-1, RANTES,
TARC and MCP-1. Functions as a
chemoattractant homing receptor
on
circulating
memory
lymphocytes and the activity of
this receptor is mediated by G(i)
proteins.
The ligands of this receptor
include monocyte chemoattractant
protein 2 (MCP-2), macrophage
inflammatory protein 1 alpha
(MIP-1
alpha),
macrophage
inflammatory protein 1 beta
(MIP-1 beta) and regulated on
activation normal T expressed and
secreted protein (RANTES).
Transduces a signal by increasing
the intracellular calcium ion level.
The gene is preferentially
expressed by immature dendritic
cells and memory T cells. Binds
to
MIP-3-alpha/LARC
and
subsequently transduces a signal
by increasing the intracellular
calcium ions level. important for
B-lineage maturation and antigendriven B-cell differentiation, and
it may regulate the migration and
recruitment of dentritic and T
cells during inflammatory and
immunological responses.
Expressed in various lymphoid
tissues and activates B and T
lymphocytes. It has been shown
to control the migration of
memory T cells to inflamed
tissues, as well as stimulate
dendritic cell maturation.
Role in regulation of monocyte
chemotaxis and thymic cell
apoptosis. More specifically, this
Chemokine (C-C motif)
receptor may contribute to the
CCR8
receptor 8
proper positioning of activated T
cells
within
the
antigenic
challenge sites and specialized
areas of lymphoid tissues.
The specific ligand of this
receptor is CCL25. Subsequently
Chemokine (C-C motif)
transduces a signal by increasing
CCR9
receptor 9
the intracellular calcium ions
level. Role in the thymocytes
recruitment and development.
Receptor
for
chemokine
Chemokine (C-X3-C motif)
Fractalkine involved in the
CX3CR1
receptor 1
adhesion and migration of
leukocytes.
CXC chemokine with chemotactic
Chemokine (C-X-C motif)
activity for neutrophils. May play
ligand 1 (melanoma growth CXCL1 a role in inflammation and exerts
stimulating activity, alpha)
its effects on endothelial cells in
an autocrine fashion.
Stimulation of monocytes, natural
Chemokine (C-X-C motif)
killer and T-cell migration, and
CXCL10
ligand 10
modulation of adhesion molecule
expression.
Encoded protein induces a
chemotactic response in activated
T-cells and is the dominant ligand
Chemokine (C-X-C motif)
CXCL11 for CXC receptor-3.
Induces
ligand 11
calcium release in activated Tcells. IFN-gamma is a potent
inducer of transcription of this
327
gene.
Chemoattractant active on Tlymphocytes, monocytes, but not
neutrophils. Activates the C-X-C
chemokine receptor CXCR4 to
Chemokine (C-X-C motif)
induce a rapid and transient rise in
CXCL12
ligand 12
the level of intracellular calcium
ions and chemotaxis. Acts as a
positive regulator of monocyte
migration and a negative regulator
of monocyte adhesion.
Preferentially
promotes
the
migration of B lymphocytes
(compared to T cells and
Chemokine (C-X-C motif)
macrophages), apparently by
CXCL13
ligand 13
stimulating calcium influx into,
and
chemotaxis
of,
cells
expressing Burkitt's lymphoma
receptor 1 (BLR-1).
Displays chemotactic activity for
monocytes
but
not
for
lymphocytes, dendritic cells,
neutrophils or
macrophages. It
Chemokine (C-X-C motif)
CXCL14 has been implicated that this
ligand 14
cytokine is involved in the
homeostasis of monocyte-derived
macrophages rather than in
inflammation.
Produced by activated monocytes
Chemokine (C-X-C motif)
CXCL2 and neutrophils and expressed at
ligand 2
sites of inflammation.
Chemotactic
activity
for
neutrophils. May play a role in
Chemokine (C-X-C motif)
CXCL3 inflammation and exert its effects
ligand 3
on endothelial cells in an
autocrine fashion.
Chemokine (C-X-C motif) CXCL5 Potent chemotaxin involved in
328
ligand 5
neutrophil activation. Produced
concomitantly with interleukin-8
(IL8) in response to stimulation
with either interleukin-1 (IL1) or
tumor
necrosis
factor-alpha
(TNFA).
Chemokine (C-X-C motif)
ligand
6
(granulocyte CXCL6
chemotactic protein 2)
Chemokine (C-X-C motif)
CXCL9
ligand 9
Chemokine
receptor 1
(C
motif)
XCR1
Chemokine (C-X-C motif)
CXCR1
receptor 1
Chemotactic
granulocytes.
for
neutrophil
Thought to be involved in T cell
trafficking.
Receptor for chemokines SCYC1
and
SCYC2.
Subsequently
transduces a signal by increasing
the intracellular calcium ions
level.
Receptor for interleukin 8 (IL8).
Causes activation of neutrophils
via a G-protein.
Cytokine Genes, Cytokine Receptors
Interferon, alpha 2
IFNA2
Interleukin 10
IL10
329
Produced by macrophages in
response to viral infection.
Cytokine produced primarily by
monocytes and to a lesser extent
by
lymphocytes.
pleiotropic
effects in immunoregulation and
inflammation. It down-regulates
the expression of Th1 cytokines,
MHC class II Ags, and
costimulatory
molecules
on
macrophages. It also enhances B
cell survival, proliferation, and
antibody
production.
This
cytokine can block NF-kappa B
activity, and is involved in the
regulation of the JAK-STAT
signaling pathway.
Mediates the immunosuppressive
signal of interleukin 10, and thus
inhibits
the
synthesis
of
Interleukin 10 receptor,
IL10RA proinflammatory
cytokines.
alpha
Activation of this receptor leads
to tyrosine phosphorylation of
JAK1 and TYK2 kinases.
Coexpression of this and IL10RA
proteins has been shown to be
Interleukin 10 receptor, beta IL10RB
required for IL10-induced signal
transduction.
Immunoregulatory
cytokine
produced primarily by activated
Th2
cells.
Down-regulates
Interleukin 13
IL13
macrophage activity. Critical in
regulating inflammatory and
immune responses.
Binds with low affinity to
interleukin-13 (IL13). Mediate the
Interleukin 13 receptor,
signaling processes that lead to
IL13RA1
alpha 1
the activation of JAK1, STAT3
and STAT6 induced by IL13 and
IL4.
Stimulates the release of tumor
necrosis factor alpha and IL-1Interleukin 17C
IL17C
beta from the monocytic cell line
THP-1.
Produced
by
activated
macrophages, IL-1 stimulates
thymocyte
proliferation
by
inducing IL-2 release, B-cell
Interleukin 1, alpha
IL1A
maturation and proliferation, and
fibroblast growth factor activity.
IL-1 proteins are involved in the
inflammatory response, being
330
Interleukin 1, beta
Interleukin
1
member 10 (theta)
Interleukin
antagonist
36
IL1B
family,
receptor
IL1F10
IL36RN
Interleukin 36, alpha
IL36A
Interleukin 37
IL37
Interleukin 36, beta
IL36B
331
identified
as
endogenous
pyrogens, and are reported to
stimulate
the
release
of
prostaglandin and collagenase
from synovial cells.
Produced
by
activated
macrophages.
Mediator of the
inflammatory response, and is
involved in a variety of cellular
activities,
including
cell
proliferation, differentiation, and
apoptosis.
Participate in a network of
interleukin 1 family members to
regulate adapted and innate
immune responses. Binds soluble
IL-1 receptor type 1.
Inhibit the activation of NFkappaB induced by interleukin 1
family, member 6 (IL1F6).
Immune
and
inflammatory
response
Suppressor
of
innate
inflammatory
and
immune
responses involved in curbing
excessive
inflammation.
Suppresses,
or
reduces,
proinflammatory
cytokine
production, including IL1A and
IL6, as well
as
CCL12,
CSF1,
CSF2,
CXCL13, IL1B, IL23A and
IL1RN,
but
spares
antiinflammatory cytokines.
Stimulates
production
of
interleukin-6 and interleukin-8 as
well as a number of matrix
Interleukin 36, gamma
IL36G
Interleukin 22
IL22
Interleukin
5
(colonystimulating
factor, IL5
eosinophil)
Interleukin 5 receptor, alpha IL5RA
Interleukin 8
IL8
Interleukin 9
IL9
Interleukin 9 receptor
IL9R
Lymphotoxin alpha (TNF
LTA
superfamily, member 1)
332
metalloproteases.
Function as an agonist of NFkappa B activation through the
orphan
IL-1-receptor-related
protein 2.
Contributes to the inflammatory
response
Main regulator of eosinopoiesis,
eosinophil
maturation
and
activation.
Receptor for interleukin-5.
CXC chemokine, major mediators
of the inflammatory response.
Secreted by several cell types, it
functions as a chemoattractant
that
attracts
neutrophils,
basophils, and T-cells, but not
monocytes.
also involved in
neutrophil activation.
Stimulates cell proliferation and
prevents apoptosis. Regulator of a
variety of hematopoietic cells.
Activates
different
signal
transducer and activator (STAT)
proteins.
Member of the tumor necrosis
factor family, produced by
lymphocytes.
Binds
to
TNFRSF1A/TNFR1,
TNFRSF1B/TNFBR
and
TNFRSF14/HVEM.
In
its
heterotrimeric form with LTB
binds to TNFRSF3/LTBR. Also
mediates a large variety of
inflammatory,
immunostimulatory, and antiviral
responses, is involved in the
Lymphotoxin beta (TNF
superfamily, member 3)
Macrophage
migration
inhibitory
factor
(glycosylation-inhibiting
factor)
Aminoacyl tRNA synthetase
complex-interacting
multifunctional protein 1
Secreted phosphoprotein 1
Tumor necrosis factor
CD40 ligand
formation of secondary lymphoid
organs during development and
plays a role in apoptosis.
Inducer of the inflammatory
response system and involved in
LTB
normal development of lymphoid
tissue.
Regulates
the
function of
macrophages in host defense.
MIF
Counteracts the anti-inflammatory
activity of glucocorticoids.
Induced by apoptosis. Involved in
the
control of angiogenesis,
inflammation, and wound healing.
AIMP1
Negatively regulates TGF-beta
signaling. Involved in glucose
homeostasis through induction of
glucagon secretion.
Enhancing
production
of
interferon-gamma
and
SPP1
interleukin-12
and
reducing
production of interleukin-10.
Mainly secreted by macrophages.
Regulation of a wide spectrum of
biological processes including cell
TNF
proliferation,
differentiation,
apoptosis, lipid metabolism, and
coagulation.
Mediates B-cell proliferation in
the absence of co-stimulus as well
CD40LG
as IgE production in the presence
of IL-4.
B2M
Housekeeping Genes
Beta-2-microglobulin
Hypoxanthine
HPRT1
phosphoribosyltransferase 1
Ribosomal protein L13a
RPL13A
Glyceraldehyde-3-phosphate GAPDH
333
Housekeeping Genes
Housekeeping Genes
Housekeeping Genes
dehydrogenase
Actin, beta
ACTB
334
Housekeeping Genes
APPENDIX TABLE 2
Gene Description
Symbol
Function
Genes Related to Energy Expenditure
Stimulates
adenylate
cyclase and subsequently
increases the cAMP level
in target cells.
May regulate the release
of
adrenocorticotropin,
luteinizing
hormone,
growth
hormone,
prolactin,
epinephrine,
Adenylate
cyclase
and catecholamine. May
activating polypeptide 1 ADCYAP1R1
play
a
role
in
(pituitary) receptor type I
spermatogenesis
and
sperm motility. Causes
smooth
muscle
relaxation and secretion in
the gastrointestinal tract
Adiponectin, C1Q and
Encoded
protein
collagen
domain ADIPOQ
circulates in the plasma
containing
and is involved with
Adenylate
cyclase
activating polypeptide 1 ADCYAP1
(pituitary)
335
Adiponectin receptor 1
ADIPOR1
Adiponectin receptor 2
ADIPOR2
Adrenergic,
beta-1-, ADRB1
336
metabolic and hormonal
processes. Involved in the
control of fat metabolism
and insulin sensitivity,
with direct anti-diabetic,
anti-atherogenic and antiinflammatory activities.
Stimulates
AMPK
phosphorylation
and
activation in the liver and
the
skeletal
muscle,
enhancing
glucose
utilization and fatty-acid
combustion. Antagonizes
TNF-alpha by
negatively regulating its
expression in various
tissues such as liver and
macrophages, and also by
counteracting its effects.
Inhibits endothelial NFkappa-B
signaling
through
a
cAMPdependent pathway.
Receptor for adiponectin.
Mediates
increased
AMPK, PPARA ligand
activity,
fatty
acid
oxidation and glucose
uptake by adiponectin.
Receptor for adiponectin.
Mediates
increased
AMPK, PPARA ligand
activity,
fatty
acid
oxidation and glucose
uptake by adiponectin.
Beta-adrenergic receptor,
receptor
Complement component
C3
3
Carboxypeptidase D
CPD
Carboxypeptidase E
CPE
Peroxisome proliferatorPPARA
activated receptor alpha
Peroxisome proliferatorPPARG
activated receptor gamma
337
mediates
the
catecholamine-induced
activation of adenylate
cyclase through the action
of G proteins. This
receptor
binds
epinephrine
and
norepinephrine.
Central role in the
activation
of
the
complement
system.
Derived from proteolytic
degradation
of
complement C3, C3a
anaphylatoxin
is
a
mediator
of
local
inflammatory process.
Regulatory
B-type
carboxypeptidase
belonging
to
metallocarboxypeptidase
family of enzymes.
Involved
in
the
biosynthesis of peptide
hormones
and
neurotransmitters,
including insulin.
Nuclear
transcription
factor, key regulator of
lipid
metabolism.
Regulates
the
peroxisomal
betaoxidation pathway of fatty
acids.
Nuclear
transcription
factor.
Regulator
of
adipocyte differentiation
Peroxisome proliferatoractivated
receptor
PPARGC1A
gamma, coactivator 1
alpha
Uncoupling protein 1
(mitochondrial,
proton UCP1
carrier)
Protein
tyrosine
phosphatase, non-receptor PTPN1
type 1
338
and glucose homeostasis.
Controls the peroxisomal
beta-oxidation pathway of
fatty acids. Implicated in
the
pathology
of
numerous
diseases
including
obesity,
diabetes, atherosclerosis
and cancer.
Interacts with PPARg.
Transcriptional
coactivator for steroid
receptors and nuclear
receptors. Regulate the
activities
of,
cAMP
response element binding
protein
(CREB)
and
nuclear respiratory factors
(NRFs). Can regulate key
mitochondrial genes that
contribute to the program
of
adaptive
thermogenesis.
Mitochondrial uncoupling
proteins, expressed only
in brown adipose tissue, a
specialized tissue which
functions to produce heat.
Negative regulator of
insulin
signaling
by
dephosphorylating
the
phosphotryosine residues
of insulin receptor kinase.
dephosphorylate
epidermal growth factor
receptor kinase, as well as
JAK2 and TYK2 kinases.
Sigma
non-opioid
SIGMAR1
intracellular receptor 1
Play an important role in
the cellular functions of
various tissues associated
with
the
endocrine,
immune, and nervous
systems. Functions in
lipid transport from the
endoplasmic reticulum.
Also
regulates
ion
channels
like
the
potassium channel and
could
modulate
neurotransmitter release
Thyroid
hormone
receptor,
beta
(erythroblastic leukemia THRB
viral (v-erb-a) oncogene
homolog 2, avian)
Nuclear hormone receptor
for triiodothyronine.
Orexigenic Genes
Adrenergic,
receptor
alpha-2B-,
ADRA2B
Agouti related protein
AGRP
homolog (mouse)
Cannabinoid receptor 1
CNR1
(brain)
Galanin prepropeptide
GAL
339
Mediate
the
catecholamine-induced
inhibition of adenylate
cyclase through the action
of G proteins.
Antagonist
of
the
melanocortin-3
and
melanocortin-4 receptor.
Plays a role in weight
homeostasis.
Involved in cannabinoidinduced CNS effects.
Acts
by
inhibiting
adenylate cyclase.
Contracts smooth muscle
of the gastrointestinal and
Galanin receptor 1
GALR1
Ghrelin/obestatin
prepropeptide
GHRL
Growth
hormone
GHSR
secretagogue receptor
Melanin-concentrating
hormone receptor 1
MCHR1
Hypocretin
receptor 1
HCRTR1
Neuropeptide Y
(orexin)
NPY
340
genitourinary
tract,
regulates growth hormone
release,
modulates
insulin release, and may
be involved in the control
of adrenal secretion.
Receptor
for
GAL,
inhibits adenylyl cyclase
via a G protein of the
Gi/Go family.
Encodes ghrelin-obestatin
preproprotein. Ghrelin has
an
appetite-stimulating
effect, induces adiposity
and stimulates gastric acid
secretion. Obestatin has
appetite-reducing effect
resulting in decreased
food intake. May reduce
gastric emptying activity
and jejunal motility.
Receptor for ghrelin.
Stimulates
growth
hormone secretion.
G
protein-coupled
receptor
family
1.
Inhibit
cAMP
accumulation
and
stimulate
intracellular
calcium flux.
G-protein
coupled
receptor involved in the
regulation of feeding
behavior.
Implicated in the control
of
feeding
and
in
secretion
of
Neuropeptide Y receptor
NPY1R
Y1
Nuclear
receptor
subfamily 3, group C,
NR3C1
member 1 (glucocorticoid
receptor)
Opioid receptor, kappa 1
OPRK1
Opioid receptor, mu 1
OPRM1
gonadotrophin-release
hormone
Receptor for neuropeptide
Y and peptide YY.
Function both as a
transcription factor and as
a regulator of other
transcription factors. It is
involved in inflammatory
responses,
cellular
proliferation,
and
differentiation in target
tissues.
Inhibits neurotransmitter
release
by
reducing
calcium ion currents and
increasing potassium ion
conductance. Receptor for
dynorphins. May play a
role in arousal and
regulation of autonomic
and
neuroendocrine
functions.
Inhibits neurotransmitter
release
by
reducing
calcium ion currents and
increasing potassium ion
conductance.
Receptor
for
betaendorphin.
Anorectic Genes
Apolipoprotein A-IV
APOA4
341
May have a role in
chylomicrons and VLDL
secretion and catabolism.
Required for efficient
Attractin
ATRN
Brain-derived
neurotrophic factor
BDNF
Bombesin-like receptor 3
BRS3
Calcitonin-related
polypeptide alpha
CALCA
Calcitonin Receptor
CALCR
CART prepropeptide
CARTPT
342
activation of lipoprotein
lipase.
Involved in the initial
immune cell clustering
during
inflammatory
response. May regulate
chemotactic activity of
chemokines. May play a
role
in melanocortin
signaling pathways that
regulate
energy
homeostasis.
May play a role in the
regulation
of
stress
response and in the
biology
of
mood
disorders.
Mediates its action by
association
with
G
proteins that activate a
phosphatidylinositolcalcium
second
messenger system.
Involved
in
calcium
regulation and acts to
regulate
phosphorus
metabolism.
Involved in maintaining
calcium homeostasis and
in regulating osteoclastmediated bone resorption.
Satiety factor inhibits
both
normal
and
starvation-induced
feeding and completely
blocks
the
feeding
response induced by
Cholecystokinin
Cholecystokinin
receptor
Colipase, pancreatic
CCK
A
CCKAR
CLPS
Ciliary
neurotrophic
CNTFR
factor receptor
Corticotropin releasing
CRHR1
hormone receptor 1
Dopamine receptor D1
DRD1
343
neuropeptide
Y
and
regulated by leptin in the
hypothalamus.
Induces
gall bladder
contraction
and
the
release of pancreatic
enzymes in the gut.
In the central and
peripheral nervous system
this receptor regulates
satiety and the release of
beta-endorphin
and
dopamine.
Cofactor of pancreatic
lipase and has a biological
activity as a satiety signal.
Plays a critical role in
neuronal cell survival,
differentiation and gene
expression.
Single
nucleotide
polymorphisms in this
gene may be associated
early onset of eating
disorders.
Essential
for
the
activation
of
signal
transduction
pathways
that
regulate
diverse
physiological processes
including
stress,
reproduction,
immune
response and obesity.
Stimulates
adenylyl
cyclase and activates
cyclic
AMP-dependent
protein kinases.
Dopamine receptor D2
DRD2
Glucagon
GCG
Glucagon receptor
GCGR
Growth hormone 1
GH1
Growth hormone 2
GH2
Growth hormone receptor GHR
Glucagon-like peptide 1
GLP1R
receptor
Gastrin-releasing peptide
GRP
344
Inhibits adenylyl cyclase
activity.
Key role in glucose
metabolism
and
homeostasis.
Receptor plays a central
role in regulating the level
of blood glucose by
controlling the rate of
hepatic
glucose
production and insulin
secretion.
Important role in growth
control. Stimulate the
liver and other tissues to
secrete
IGF-1.
Also
stimulates amino acid
uptake
and
protein
synthesis.
Important role in growth
control. Stimulate the
liver and other tissues to
secrete
IGF-1.
Also
stimulates amino acid
uptake
and
protein
synthesis.
On
ligand
binding,
couples
to
the
JAK2/STAT5 pathway.
Receptor
activity
mediated by G proteins
which activate adenylyl
cyclase. Role in delaying
gastric emptying and
regulating appetite.
Regulate
release
of
gastrointestinal hormones
Gastrin-releasing peptide
GRPR
receptor
Hypocretin
(orexin)
HCRT
neuropeptide precursor
Histamine receptor H1
HRH1
5-hydroxytryptamine
(serotonin) receptor 2C
HTR2C
Islet amyloid polypeptide
IAPP
345
smooth
muscle
cell
contraction, and epithelial
cell proliferation.
Mediates its action by
association
with
G
proteins that activate a
phosphatidylinositolcalcium
second
messenger system.
Encodes a hypothalamic
neuropeptide
precursor
protein that gives rise to
two
mature
neuropeptides, orexin A
and
orexin
B,
by
proteolytic
processing.
Role in feeding behavior,
metabolism,
and
homeostasis
In peripheral tissues, the
H1 subclass of histamine
receptors mediates the
contraction of smooth
muscles,
increase
in
capillary permeability as
well
as
mediating
neurotransmission in the
central nervous system.
Receptor for serotonin,
receptor mediates its
action by association with
G proteins that activate a
phosphatidylinositolcalcium
second
messenger system.
Selectively
inhibits
insulin-stimulated glucose
Interleukin 1, alpha
IL1A
Interleukin 1, beta
IL1B
Interleukin
type I
IL1R1
1
receptor,
Interleukin 6 (interferon,
IL6
beta 2)
346
utilization and glycogen
deposition in muscle,
while
not
affecting
adipocyte
glucose
metabolism.
Produced by activated
macrophages,
IL-1
stimulates
thymocyte
proliferation by inducing
IL-2
release,
B-cell
maturation
and
proliferation,
and
fibroblast growth factor
activity. IL-1 proteins are
involved
in
the
inflammatory response.
Produced by activated
macrophages,
IL-1
stimulates
thymocyte
proliferation by inducing
IL-2
release,
B-cell
maturation
and
proliferation,
and
fibroblast growth factor
activity. IL-1 proteins are
involved
in
the
inflammatory response.
Receptor for interleukin-1
alpha (IL-1A), beta (IL1B), and interleukin-1
receptor
antagonist
protein
(IL-1RA).
Binding to the agonist
leads to the activation of
NF-kappa-B.
Primarily produced at
sites of acute and chronic
Interleukin 6 receptor
IL6R
Insulin
INS
Insulin receptor
INSR
Leptin
LEP
Leptin receptor
LEPR
Melanocortin 3 receptor
MC3R
347
inflammation, where it is
secreted into the serum
and
induces
a
transcriptional
inflammatory
response
through interleukin 6
receptor, alpha.
Activation leads to the
regulation of the immune
response,
acute-phase
reactions
and
hematopoiesis.
Insulin decreases blood
glucose concentration. It
increases
cell
permeability
to
monosaccharides, amino
acids and fatty acids. It
accelerates glycolysis, the
pentose phosphate cycle,
and glycogen synthesis in
liver.
Binds insulin and has a
tyrosine-protein
kinase
activity.
Plays a major role in the
regulation
of
body
weight.
On
ligand
binding,
mediates
signaling
through
JAK2/STAT3.
Involved in the regulation
of fat metabolism.
A
G-protein-coupled
receptor for melanocytestimulating hormone and
adrenocorticotropic
Neuromedin B
NMB
Neuromedin B receptor
NMBR
Neuromedin U
NMU
Neuromedin U receptor 1
NMUR1
Neurotrophic
tyrosine
NTRK2
kinase, receptor, type 2
Neurotensin
NTS
Neurotensin receptor 1 NTSR1
348
hormone.
Activation
mediated by G proteins
which activate adenylate
cyclase.
Stimulates smooth muscle
contraction
involved in the regulation
of
many
biological
functions
including
sensory
transmission,
thermoregulation,
feeding, pituitary, gastric
and pancreatic secretion.
Stimulates
muscle
contractions of specific
regions
of
the
gastrointestinal tract.
Mediate
diverse
biological
effects
including contraction of
the uterus, peripheral
vasoconstriction causing
hypertension
and
activation of the HPA
axis resulting in decreased
food intake via the CRH
system.
Tyrosine-protein kinase
receptor for brain-derived
neurotrophic
factor
(BDNF), neurotrophin-3
and neurotrophin-4/5.
Role in endocrine or
paracrine role in the
regulation
of
fat
metabolism.
Activates
a
(high affinity)
Proopiomelanocortin
POMC
Prolactin
releasing
PRLHR
hormone receptor
Peptide YY
PYY
Receptor (G proteincoupled)
activity RAMP3
modifying protein 3
Sortilin 1
SORT1
349
phosphatidylinositolcalcium
second
messenger system.
Wide
range
of
physiological functions,
including pigmentation,
energy
homeostasis,
inflammation,
immunomodulation,
steroidogenesis
and
temperature control.
Implicated in lactation,
regulation of food intake
and
pain-signal
processing.
Gut
peptide
inhibits
exocrine
pancreatic
secretion,
has
a
vasoconstrictory
action
and inhibits jejuna and
colonic mobility
Transports the calcitonin
gene-related peptide type
1 receptor (CALCRL) to
the plasma membrane.
Acts as a
receptor for
adrenomedullin
(AM)
together with CALCRL.
Sorting receptor in the
Golgi compartment and as
a clearance receptor on
the
cell
surface.
Probably required in
adipocytes
for
the
formation of specialized
storage
vesicles
containing the glucose
Somatostatin
SST
Somatostatin receptor 2
SSTR2
Tumor necrosis factor
TNF
Thyrotropin-releasing
hormone
TRH
Urocortin
UCN
Zinc finger protein 91
ZFP91
homolog (mouse)
350
transporter GLUT4.
Inhibits the release of
somatotropin. Couples to
adenylyl cyclase, PLC,
K+
channels,
Ca2+
channels and others.
Coupled to inhibition of
adenylyl
cyclase.
In
addition it stimulates
phosphotyrosine
phosphatase and PLC.
Secreted by macrophages.
Binds
to
receptors
TNFRSF1A/TNFR1 and
TNFRSF1B/TNFBR.
Involved in the regulation
of wide spectrum of
biological
processes
including
cell
proliferation,
differentiation, apoptosis,
lipid metabolism, and
coagulation.
Responsible
for
the
regulation and release of
thyroid-stimulating
hormone, as well as
prolactin.
Stimulates the secretion
of
adrenocorticotropic
hormone (ACTH).
Belongs to zinc finger
family
of
proteins,
regulator of the noncanonical
NF-kappaB
pathway in lymphotoxinbeta receptor signaling.
Beta-2-microglobulin
Hypoxanthine
phosphoribosyltransferase
1
Ribosomal protein L13a
Glyceraldehyde-3phosphate dehydrogenase
Actin, beta
B2M
May also play an
important role in cell
proliferation and/or antiapoptosis.
Housekeeping Gene
HPRT1
Housekeeping Gene
RPL13A
Housekeeping Gene
GAPDH
Housekeeping Gene
ACTB
Housekeeping Gene
351
REFERENCES
Adams, L.A., Lymp, J.F., St. Sauver, J., Sanderson, S.O., Lindor, K.D., Feldstein, A., and
Angulo, P. (2005). The Natural History of Nonalcoholic Fatty Liver Disease: A Population-Based
Cohort Study. Gastroenterology 129, 113–121.
Ahima, R.S. (2008). Revisiting leptin’s role in obesity and weight loss. J Clin Invest 118, 2380–
2383.
Ahima, R.S., and Flier, J.S. (2000). Adipose Tissue as an Endocrine Organ. Trends in
Endocrinology & Metabolism 11, 327–332.
Alkhouri, N., Gornicka, A., Berk, M.P., Thapaliya, S., Dixon, L.J., Kashyap, S., Schauer, P.R.,
and Feldstein, A.E. (2009). Adipocyte apoptosis: A link between obesity, insulin resistance and
hepatic steatosis. Journal of Biological Chemistry.
Andersen, C.L. (2004). Normalization of Real-Time Quantitative Reverse Transcription-PCR
Data: A Model-Based Variance Estimation Approach to Identify Genes Suited for Normalization,
Applied to Bladder and Colon Cancer Data Sets. Cancer Research 64, 5245–5250.
Andersen, C.L., Jensen, J.L., and Ørntoft, T.F. (2004). Normalization of Real-Time Quantitative
Reverse Transcription-PCR Data: A Model-Based Variance Estimation Approach to Identify
Genes Suited for Normalization, Applied to Bladder and Colon Cancer Data Sets. Cancer Res 64,
5245–5250.
Andoh, A., Takaya, H., Araki, Y., Tsujikawa, T., Fujiyama, Y., and Bamba, T. (2000). Mediumand Long-Chain Fatty Acids Differentially Modulate Interleukin-8 Secretion in Human Fetal
Intestinal Epithelial Cells. The Journal of Nutrition 130, 2636 –2640.
Andreas Beilhack, and Rockson, S.G. (2003). Immune Traffic: A Functional Overview.
Lymphatic Research and Biology 1, 219–234.
Arch, J.R.S. (1989). The brown adipocyte β-adrenoceptor. Proceedings of the Nutrition Society
48, 215–223.
Arend, W.P., Malyak, M., Guthridge, C.J., and Gabay, C. (1998). INTERLEUKIN-1
RECEPTOR ANTAGONIST: Role in Biology. Annual Review of Immunology 16, 27–55.
Arner, P. (1998). Not all fat is alike. Lancet 351, 1301–1302.
352
Arner P (2001). Regional differences in protein production by human adipose tissue. Biochemical
Society Transactions 29, 72–75.
Asakawa, A., Inui, A., Kaga, O., Yuzuriha, H., Nagata, T., Ueno, N., Makino, S., Fujimiya, M.,
Niijima, A., Fujino, M.A., et al. (2001). Ghrelin is an appetite-stimulatory signal from stomach
with structural resemblance to motilin. Gastroenterology 120, 337–345.
Assimakopoulos, S.F., Vagianos, C.E., Zervoudakis, G., Filos, K.S., Georgiou, C., Nikolopoulou,
V., and Scopa, C.D. (2004). Gut regulatory peptides bombesin and neurotensin reduce hepatic
oxidative stress and histological alterations in bile duct ligated rats. Regulatory Peptides 120,
185–193.
Assimakopoulos, S.F., Tsamandas, A.C., Alexandris, I.H., Georgiou, C., Vagianos, C.E., and
Scopa, C.D. (2011). Stimulation of oval cell and hepatocyte proliferation by exogenous bombesin
and neurotensin in partially hepatectomized rats. World J Gastrointest Pathophysiol 2, 146–154.
Badman, M.K., and Flier, J.S. (2005). The Gut and Energy Balance: Visceral Allies in the
Obesity Wars. Science 307, 1909 –1914.
Bado, A., Levasseur, S., Attoub, S., Kermorgant, S., Laigneau, J.-P., Bortoluzzi, M.-N., Moizo,
L., Lehy, T., Guerre-Millo, M., Le Marchand-Brustel, Y., et al. (1998). The stomach is a source
of leptin. Nature 394, 790–793.
Baranova, A., Hammarsund, M., Ivanov, D., Skoblov, M., Sangfelt, O., Corcoran, M., Borodina,
T., Makeeva, N., Pestova, A., Tyazhelova, T., et al. (2003). Distinct organization of the candidate
tumor suppressor gene RFP2 in human and mouse: multiple mRNA isoforms in both species- and
human-specific antisense transcript RFP2OS. Gene 321, 103–112.
Baranova, A., Gowder, S.J., Schlauch, K., Elariny, H., Collantes, R., Afendy, A., Ong, J.P.,
Goodman, Z., Chandhoke, V., and Younossi, Z.M. (2006). Gene expression of leptin, resistin,
and adiponectin in the white adipose tissue of obese patients with non-alcoholic fatty liver disease
and insulin resistance. Obesity Surgery 16, 1118–1125.
Baranova, A., Randhawa, M., Jarrar, M., and Younossi, Z.M. (2007). Adipokines and
melanocortins in the hepatic manifestation of metabolic syndrome: nonalcoholic fatty liver
disease. Expert Review of Molecular Diagnostics 7, 195–205.
Baratta, M. (2002). Leptin--from a signal of adiposity to a hormonal mediator in peripheral
tissues. Med. Sci. Monit. 8, RA282–292.
Barbatelli, G., Murano, I., Madsen, L., Hao, Q., Jimenez, M., Kristiansen, K., Giacobino, J.P.,
Matteis, R.D., and Cinti, S. (2010). 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, E1244–E1253.
Barish, G.D. (2006). PPAR : a dagger in the heart of the metabolic syndrome. Journal of Clinical
Investigation 116, 590–597.
353
Bensaid, M., Gary-Bobo, M., Esclangon, A., Maffrand, J.P., Fur, G.L., Oury-Donat, F., and
Soubrié, P. (2003). The Cannabinoid CB1 Receptor Antagonist SR141716 Increases Acrp30
mRNA Expression in Adipose Tissue of Obese fa/fa Rats and in Cultured Adipocyte Cells. Mol
Pharmacol 63, 908–914.
Berndt, J., Klöting, N., Kralisch, S., Kovacs, P., Fasshauer, M., Schön, M.R., Stumvoll, M., and
Blüher, M. (2005). Plasma Visfatin Concentrations and Fat Depot–Specific mRNA Expression in
Humans. Diabetes 54, 2911 –2916.
Bertola, A., Bonnafous, S., Anty, R., Patouraux, S., Saint-Paul, M.-C., Iannelli, A., Gugenheim,
J., Barr, J., Mato, J.M., Le Marchand-Brustel, Y., et al. (2010a). Hepatic Expression Patterns of
Inflammatory and Immune Response Genes Associated with Obesity and NASH in Morbidly
Obese Patients. PLoS ONE 5, e13577.
Bertola, A., Bonnafous, S., Anty, R., Patouraux, S., Saint-Paul, M.-C., Iannelli, A., Gugenheim,
J., Barr, J., Mato, J.M., Le Marchand-Brustel, Y., et al. (2010b). Hepatic Expression Patterns of
Inflammatory and Immune Response Genes Associated with Obesity and NASH in Morbidly
Obese Patients. PLoS ONE 5, e13577.
Birerdinc, A., Nohelty, E., Marakhonov, A., Manyam, G., Panov, I., Coon, S., Nikitin, E.,
Skoblov, M., Chandhoke, V., and Baranova, A. (2010). Pro-apoptotic and antiproliferative
activity of human KCNRG, a putative tumor suppressor in 13q14 region. Tumour Biol 31, 33–45.
Birerdinc, A., Mehta, R., Alhussain, R., Afendi, A., Chandhoke, V., Younossi, Z., and Baranova,
A. (2012). Selection of reliable reference genes for qRT-PCR analysis in human non-cancerous
gastric tissue. Molecular Biology 46, 153–160.
Bjorntorp, P. (1990). “Portal” Adipose Tissue as a Generator of Risk Factors for Cardiovascular
Disease and Diabetes. Arterioscler Thromb Vasc Biol 10, 493–496.
Böni-Schnetzler, M., Boller, S., Debray, S., Bouzakri, K., Meier, D.T., Prazak, R., Kerr-Conte, J.,
Pattou, F., Ehses, J.A., Schuit, F.C., et al. (2009). Free Fatty Acids Induce a Proinflammatory
Response in Islets via the Abundantly Expressed Interleukin-1 Receptor I. Endocrinology 150,
5218 –5229.
Bosello, O., and Zamboni, M. (2000). Visceral obesity and metabolic syndrome. Obesity Reviews
1, 47–56.
Buchwald, H., and Oien, D. (2009). Metabolic/Bariatric Surgery Worldwide 2008. Obesity
Surgery 19, 1605–1611–1611.
Bustin, S. (2000). Absolute quantification of mRNA using real-time reverse transcription
polymerase chain reaction assays. J Mol Endocrinol 25, 169–193.
Cammisotto, P.G., Renaud, C., Gingras, D., Delvin, E., Levy, E., and Bendayan, M. (2005).
Endocrine and Exocrine Secretion of Leptin by the Gastric Mucosa. Journal of Histochemistry &
Cytochemistry 53, 851 –860.
354
Cammisotto, P.G., Levy, É., Bukowiecki, L.J., and Bendayan, M. (2010). Cross-talk between
adipose and gastric leptins for the control of food intake and energy metabolism. Progress in
Histochemistry and Cytochemistry 45, 143–200.
Canbay Ali, Kip SN, A, K., Gieseler RK, A, N., and Gerken G (2005). Apoptosis and fibrosis in
non-alcoholic fatty liver disease. The Turkish Journal of Gastroenterology 16, 1–6.
Cannon, B., and Nedergaard, J. (1985). The biochemistry of an inefficient tissue: brown adipose
tissue. Essays Biochem. 20, 110–164.
Cannon, B., and Nedergaard, J. (2004). Brown Adipose Tissue: Function and Physiological
Significance. Physiological Reviews 84, 277 –359.
Cannon, B., Jacobsson, A., Rehnmark, S., and Nedergaard, J. (1996). Signal transduction in
brown adipose tissue recruitment: noradrenaline and beyond. Int. J. Obes. Relat. Metab. Disord.
20 Suppl 3, S36–42.
Cantó, C., and Auwerx, J. (2009). PGC-1[alpha], SIRT1 and AMPK, an energy sensing network
that controls energy expenditure. Current Opinion in Lipidology 20.
Carl Wittwer, Hahn, M., and Kaul, K. (2004). Rapid Cycle Real-time PCR: Methods and
Applications : Quantification (Springer).
Catalán, V., Gómez-Ambrosi, J., Rotellar, F., Silva, C., Rodríguez, A., Salvador, J., Gil, M.J.,
Cienfuegos, J.A., and Frühbeck, G. (2007). Validation of endogenous control genes in human
adipose tissue: relevance to obesity and obesity-associated type 2 diabetes mellitus. Horm. Metab.
Res. 39, 495–500.
Chaudhri, O.B., Field, B.C.T., and Bloom, S.R. (2008). Gastrointestinal satiety signals. Int J Obes
32, S28–S31.
Cheng, J.-T., Liu, I.-M., and Hsu, C.F. (2003). Rapid induction of insulin resistance in opioid mureceptor knock-out mice. Neurosci. Lett. 339, 139–142.
Chitturi, S., Farrell, G., Frost, L., Kriketos, A., Lin, R., Fung, C., Liddle, C., Samarasinghe, D.,
and George, J. (2002). Serum leptin in NASH correlates with hepatic steatosis but not fibrosis: a
manifestation of lipotoxicity? Hepatology 36, 403–409.
Cho, Y.G., Kim, C.J., Song, J.H., Rhie, D.J., Park, Y.K., Kim, S.Y., Nam, S.W., Yoo, N.J., Lee,
J.Y., and Park, W.S. (2006). Genetic and expression analysis of the KCNRG gene in
hepatocellular carcinomas. Exp. Mol. Med. 38, 247–255.
Chou, R.C., Kim, N.D., Sadik, C.D., Seung, E., Lan, Y., Byrne, M.H., Haribabu, B., Iwakura, Y.,
and Luster, A.D. (2010). Lipid-Cytokine-Chemokine Cascade Drives Neutrophil Recruitment in a
Murine Model of Inflammatory Arthritis. Immunity 33, 266–278.
355
Cicinnati, V., Shen, Q., Sotiropoulos, G., Radtke, A., Gerken, G., and Beckebaum, S. (2008).
Validation of putative reference genes for gene expression studies in human hepatocellular
carcinoma using real-time quantitative RT-PCR. BMC Cancer 8, 350.
Cinti S, De, M.R., Picó C, E, C., Obrador, A., Maffeis, C., Oliver, J., and Palou, A. (2000).
Secretory granules of endocrine and chief cells of human stomach mucosa contain leptin.
International Journal of Obesity 24, 789–793.
Coppack S W (2005). Adipose tissue changes in obesity. Biochemical Society Transactions 33,
1049–1052.
Costford, S.R., Bajpeyi, S., Pasarica, M., Albarado, D.C., Thomas, S.C., Xie, H., Church, T.S.,
Jubrias, S.A., Conley, K.E., and Smith, S.R. (2010). Skeletal muscle NAMPT is induced by
exercise in humans. American Journal of Physiology - Endocrinology And Metabolism 298, E117
–E126.
Cummings, D.E., and Overduin, J. (2007). Gastrointestinal regulation of food intake. J Clin
Invest 117, 13–23.
Cypess, A.M., Lehman, S., Williams, G., Tal, I., Rodman, D., Goldfine, A.B., Kuo, F.C., Palmer,
E.L., Tseng, Y.-H., Doria, A., et al. (2009). Identification and Importance of Brown Adipose
Tissue in Adult Humans. New England Journal of Medicine 360, 1509–1517.
Czyzyk, T.A., Romero-Picó, A., Pintar, J., McKinzie, J.H., Tschöp, M.H., Statnick, M.A., and
Nogueiras, R. (2012). Mice lacking δ-opioid receptors resist the development of diet-induced
obesity. FASEB J. 26, 3483–3492.
Dahlman, I., Linder, K., Nordström, E.A., Andersson, I., Lidén, J., Verdich, C., Sørensen, T.I.,
and Arner, P. (2005). Changes in adipose tissue gene expression with energy-restricted diets in
obese women. Am J Clin Nutr 81, 1275–1285.
Dali‐Youcef, N., Lagouge, M., Froelich, S., Koehl, C., Schoonjans, K., and Auwerx, J. (2007).
Sirtuins: The “magnificent seven”, function, metabolism and longevity. Ann Med 39, 335–345.
Dandona, P., Aljada, A., O’donnell, A., Dhindsa, S., and Garg, R. (2004). Insulin Is an Antiinflammatory and Anti-atherosclerotic Hormone. Metabolic Syndrome and Related Disorders 2,
137–142.
Dandona, P., Chaudhuri, A., Mohanty, P., and Ghanim, H. (2007). Anti-inflammatory effects of
insulin. Curr Opin Clin Nutr Metab Care 10, 511–517.
Date, Y., Kojima, M., Hosoda, H., Sawaguchi, A., Mondal, M.S., Suganuma, T., Matsukura, S.,
Kangawa, K., and Nakazato, M. (2000). Ghrelin, a Novel Growth Hormone-Releasing Acylated
Peptide, Is Synthesized in a Distinct Endocrine Cell Type in the Gastrointestinal Tracts of Rats
and Humans. Endocrinology 141, 4255–4261.
DeBoer, M.D., Scarlett, J.M., Levasseur, P.R., Grant, W.F., and Marks, D.L. (2009).
Administration of IL-1beta to the 4th ventricle causes anorexia that is blocked by agouti-related
356
peptide and that coincides with activation of tyrosine-hydroxylase neurons in the nucleus of the
solitary tract. Peptides 30, 210–218.
Delhalle, S., Duvoix, A., Schnekenburger, M., Morceau, F., Dicato, M., and Diederich, M.
(2003). An Introduction to the Molecular Mechanisms of Apoptosis. Annals of the New York
Academy of Sciences 1010, 1–8.
Despres, J.-P., and Lemieux, I. (2006). Abdominal obesity and metabolic syndrome. Nature 444,
881–887.
Dheda, K., Huggett, J.F., Chang, J.S., Kim, L.U., Bustin, S.A., Johnson, M.A., Rook, G.A.W.,
and Zumla, A. (2005). The implications of using an inappropriate reference gene for real-time
reverse transcription PCR data normalization. Analytical Biochemistry 344, 141–143.
Dickson, S.L., Egecioglu, E., Landgren, S., Skibicka, K.P., Engel, J.A., and Jerlhag, E. (2011).
The role of the central ghrelin system in reward from food and chemical drugs. Molecular and
Cellular Endocrinology in press.
Dobrin, R., Zhu, J., Molony, C., Argman, C., Parrish, M., Carlson, S., Allan, M., Pomp, D., and
Schadt, E. (2009). Multi-tissue coexpression networks reveal unexpected subnetworks associated
with disease. Genome Biology 10, R55.
Dowman, J.K., Tomlinson, J.W., and Newsome, P.N. (2010). Pathogenesis of non-alcoholic fatty
liver disease. QJM 103, 71–83.
Drubach, L.A., Palmer, E.L., Connolly, L.P., Baker, A., Zurakowski, D., and Cypess, A.M.
(2011). Pediatric Brown Adipose Tissue: Detection, Epidemiology, and Differences from Adults.
The Journal of Pediatrics 159, 939–944.
Ekstedt, M., Franzén, L.E., Mathiesen, U.L., Thorelius, L., Holmqvist, M., Bodemar, G., and
Kechagias, S. (2006). Long-term follow-up of patients with NAFLD and elevated liver enzymes.
Hepatology 44, 865–873.
El-Mesallamy, H.O., Kassem, D.H., El-Demerdash, E., and Amin, A.I. (2011). Vaspin and
visfatin/Nampt are interesting interrelated adipokines playing a role in the pathogenesis of type 2
diabetes mellitus. Metabolism 60, 63–70.
Enerbäck, S. (2009). The Origins of Brown Adipose Tissue. New England Journal of Medicine
360, 2021–2023.
Enerbäck, S. (2010). Human Brown Adipose Tissue. Cell Metabolism 11, 248–252.
Estep, J.M., Baranova, A., Hossain, N., Elariny, H., Ankrah, K., Afendy, A., Chandhoke, V., and
Younossi, Z.M. (2009). Expression of cytokine signaling genes in morbidly obese patients with
non-alcoholic steatohepatitis and hepatic fibrosis. Obesity Surgery 19, 617–624.
Estep, M., Abawi, M., Jarrar, M., Wang, L., Stepanova, M., Elariny, H., Moazez, A., Goodman,
Z., Chandhoke, V., Baranova, A., et al. (2011). Association of Obestatin, Ghrelin, and
357
Inflammatory Cytokines in Obese Patients with Non-alcoholic Fatty Liver Disease. Obesity
Surgery 21, 1750–1757.
Fabbrini, E., Sullivan, S., and Klein, S. (2010). Obesity and nonalcoholic fatty liver disease:
Biochemical, metabolic, and clinical implications. Hepatology 51, 679–689.
Faintuch, J., Ishida, R.K., Jacabi, M., Ribeiro, A.S., Kuga, R., Sakai, P., Barbeiro, H.V., Barbeiro,
D.F., Soriano, F.G., and Cecconello, I. (2007). Increased Gastric Cytokine Production After
Roux-en-Y Gastric Bypass for Morbid Obesity. Arch Surg 142, 962–968.
Fantuzzi, G. (2005). Adipose tissue, adipokines, and inflammation. Journal of Allergy and
Clinical Immunology 115, 911–919.
Faouzi, S., Burckhardt, B.E., Hanson, J.C., Campe, C.B., Schrum, L.W., Rippe, R.A., and Maher,
J.J. (2001). Anti-Fas Induces Hepatic Chemokines and Promotes Inflammation by an NF-κBindependent, Caspase-3-dependent Pathway. Journal of Biological Chemistry 276, 49077 –49082.
Farmer, S.R. (2006). Transcriptional control of adipocyte formation. Cell Metabolism 4, 263–
273.
Farmer, S.R. (2008). Molecular determinants of brown adipocyte formation and function. Genes
& Development 22, 1269 –1275.
Faubion, W.A., and Gores, G.J. (1999). Death receptors in liver biology and pathobiology.
Hepatology 29, 1–4.
Feldstein, A.E., Canbay, A., Guicciardi, M.E., Higuchi, H., Bronk, S.F., and Gores, G.J. (2003).
Diet associated hepatic steatosis sensitizes to Fas mediated liver injury in mice. Journal of
Hepatology 39, 978–983.
Ferré, P., and Foufelle, F. (2007). SREBP-1c Transcription Factor and Lipid Homeostasis:
Clinical Perspective. Horm Res 68, 72–82.
Flier, J.S., and Maratos-Flier, E. (1998). Obesity and the Hypothalamus: Novel Peptides for New
Pathways. Cell 92, 437–440.
Fontana, L., Eagon, J.C., Trujillo, M.E., Scherer, P.E., and Klein, S. (2007). Visceral Fat
Adipokine Secretion Is Associated With Systemic Inflammation in Obese Humans. Diabetes 56,
1010–1013.
Fox, C.S., Massaro, J.M., Hoffmann, U., Pou, K.M., Maurovich-Horvat, P., Liu, C.-Y., Vasan,
R.S., Murabito, J.M., Meigs, J.B., Cupples, L.A., et al. (2007). Abdominal Visceral and
Subcutaneous Adipose Tissue Compartments Association With Metabolic Risk Factors in the
Framingham Heart Study. Circulation 116, 39–48.
Fried, S.K., Bunkin, D.A., and Greenberg, A.S. (1998). Omental and Subcutaneous Adipose
Tissues of Obese Subjects Release Interleukin-6: Depot Difference and Regulation by
Glucocorticoid. JCEM 83, 847–850.
358
Friis-Hansen, L., Lacourse, K.A., Samuelson, L.C., and Holst, J.J. (2001). Attenuated processing
of proglucagon and glucagon-like peptide-1 in carboxypeptidase E-deficient mice. J Endocrinol
169, 595–602.
Frühbeck, G., Becerril, S., Sáinz, N., Garrastachu, P., and García-Velloso, M.J. (2009). BAT: a
new target for human obesity? Trends in Pharmacological Sciences 30, 387–396.
Furukawa, S., Fujita, T., Shimabukuro, M., Iwaki, M., Yamada, Y., Nakajima, Y., Nakayama, O.,
Makishima, M., Matsuda, M., and Shimomura, I. (2004). Increased oxidative stress in obesity and
its impact on metabolic syndrome. J. Clin. Invest. 114, 1752–1761.
Gabay, C., Smith, M.F., Eidlen, D., and Arend, W.P. (1997). Interleukin 1 receptor antagonist
(IL-1Ra) is an acute-phase protein. J Clin Invest 99, 2930–2940.
Gale, S.M., Castracane, V.D., and Mantzoros, C.S. (2004). Energy Homeostasis, Obesity and
Eating Disorders: Recent Advances in Endocrinology. The Journal of Nutrition 134, 295 –298.
Gallagher, C.J., Gordon, C.J., Langefeld, C.D., Mychaleckyj, J.C., Freedman, B.I., Rich, S.S.,
Bowden, D.W., and Sale, M.M. (2006). Association of the mu-opioid receptor gene with type 2
diabetes mellitus in an African American population. Mol. Genet. Metab. 87, 54–60.
Galle, P.R., Hofmann, W.J., H Walczak, Schaller, H., G Otto, W Stremmel, Krammer, P.H., and
L Runkel (1995). Involvement of the CD95 (APO-1/Fas) receptor and ligand in liver damage.
Journal of Experimental Medicine 182, 1223–1230.
García, M.C., Wernstedt, I., Berndtsson, A., Enge, M., Bell, M., Hultgren, O., Horn, M., Ahrén,
B., Enerback, S., Ohlsson, C., et al. (2006). Mature-onset obesity in interleukin-1 receptor I
knockout mice. Diabetes 55, 1205–1213.
Garg, A., and Misra, A. (2002). Hepatic Steatosis, Insulin Resistance, and Adipose Tissue
Disorders. J Clin Endocrinol Metab 87, 3019–3022.
Garten, A., Petzold, S., Barnikol-Oettler, A., Körner, A., Thasler, W.E., Kratzsch, J., Kiess, W.,
and Gebhardt, R. (2010). Nicotinamide phosphoribosyltransferase (NAMPT/PBEF/visfatin) is
constitutively released from human hepatocytes. Biochemical and Biophysical Research
Communications 391, 376–381.
Gasser, S.M., and Cockell, M.M. (2001). The molecular biology of the SIR proteins. Gene 279,
1–16.
Gauna, C., Meyler, F.M., Janssen, J.A.M.J.L., Delhanty, P.J.D., Abribat, T., van Koetsveld, P.,
Hofland, L.J., Broglio, F., Ghigo, E., and van der Lely, A.J. (2004). Administration of Acylated
Ghrelin Reduces Insulin Sensitivity, Whereas the Combination of Acylated Plus Unacylated
Ghrelin Strongly Improves Insulin Sensitivity. J Clin Endocrinol Metab 89, 5035–5042.
Gauna, C., Delhanty, P.J.D., Hofland, L.J., Janssen, J.A.M.J.L., Broglio, F., Ross, R.J.M., Ghigo,
E., and van der Lely, A.J. (2005). Ghrelin Stimulates, Whereas Des-Octanoyl Ghrelin Inhibits,
Glucose Output by Primary Hepatocytes. J Clin Endocrinol Metab 90, 1055–1060.
359
Gavrilova, O., Marcus-Samuels, B., Graham, D., Kim, J.K., Shulman, G.I., Castle, A.L., Vinson,
C., Eckhaus, M., and Reitman, M.L. (2000). Surgical implantation of adipose tissue reverses
diabetes in lipoatrophic mice. J. Clin. Invest. 105, 271–278.
Gesta, S., Tseng, Y.-H., and Kahn, C.R. (2007). Developmental Origin of Fat: Tracking Obesity
to Its Source. Cell 131, 242–256.
Von Gierke E (1906). Ueber fett metabolism. Ver Deutsch Ges Path 10, 182–185.
Gill, R.S., Karmali, S., and Sharma, A.M. (2010). Treating Type 2 Diabetes Mellitus With Sleeve
Gastrectomy in Obese Patients. Obesity.
Gilsanz, V., Smith, M.L., Goodarzian, F., Kim, M., Wren, T.A.L., and Hu, H.H. (2012). Changes
in Brown Adipose Tissue in Boys and Girls during Childhood and Puberty. The Journal of
Pediatrics 160, 604–609.e1.
Giralt, A., Hondares, E., Villena, J.A., Ribas, F., Díaz-Delfín, J., Giralt, M., Iglesias, R., and
Villarroya, F. (2011). Peroxisome Proliferator-activated Receptor-γ Coactivator-1α Controls
Transcription of the Sirt3 Gene, an Essential Component of the Thermogenic Brown Adipocyte
Phenotype. Journal of Biological Chemistry 286, 16958 –16966.
Girard, J., and Lafontan, M. (2008). Impact of visceral adipose tissue on liver metabolism and
insulin resistance. Part II: Visceral adipose tissue production and liver metabolism. Diabetes &
Metabolism 34, 439–445.
Gómez-Ambrosi, J., Catalán, V., Diez-Caballero, A., Martinez-Cruz, L.A., Gil, M.J., GarcíaFoncillas, J., Cienfuegos, J.A., Salvador, J., Mato, J.M., and Frühbeck, G. (2004). Gene
expression profile of omental adipose tissue in human obesity. FASEB J. 18, 215–217.
Gómez-Quiroz, L., Bucio, L., Souza, V., Escobar, C., Farfán, B., Hernández, E., Konigsberg, M.,
Vargas-Vorackova, F., Kershenobich, D., and Gutiérrez-Ruiz, M.C. (2003). Interleukin 8
response and oxidative stress in HepG2 cells treated with ethanol, acetaldehyde or
lipopolysaccharide. Hepatology Research 26, 134–141.
Goodpaster, B., and Kelley, D. (1998). Role of muscle in triglyceride metabolism. [Miscellaneous
Article]. Current Opinion in Lipidology June 1998 9, 231–236.
Granneman, J.G., Burnazi, M., Zhu, Z., and Schwamb, L.A. (2003). White adipose tissue
contributes to UCP1-independent thermogenesis. Am J Physiol Endocrinol Metab 285, E1230–
E1236.
Granneman, J.G., Li, P., Zhu, Z., and Lu, Y. (2005). Metabolic and cellular plasticity in white
adipose tissue I: effects of β3-adrenergic receptor activation. Am J Physiol Endocrinol Metab
289, E608–E616.
Greenberg, A.S., and Obin, M.S. (2006). Obesity and the role of adipose tissue in inflammation
and metabolism. Am J Clin Nutr 83, 461S–465.
360
Grundy, S.M. (2004). Obesity, Metabolic Syndrome, and Cardiovascular Disease. J Clin
Endocrinol Metab 89, 2595–2600.
Guicciardi, M.E., and Gores, G.J. (2009). Life and death by death receptors. The FASEB Journal
23, 1625 –1637.
Gurriarán-Rodríguez, U., Al-Massadi, O., Roca-Rivada, A., Crujeiras, A.B., Gallego, R., Pardo,
M., Seoane, L.M., Pazos, Y., Casanueva, F.F., and Camiña, J.P. (2011). Obestatin as a regulator
of adipocyte metabolism and adipogenesis. Journal of Cellular and Molecular Medicine 15,
1927–1940.
Haider, D.G., Schindler, K., Schaller, G., Prager, G., Wolzt, M., and Ludvik, B. (2006). Increased
Plasma Visfatin Concentrations in Morbidly Obese Subjects Are Reduced after Gastric Banding.
Journal of Clinical Endocrinology & Metabolism 91, 1578 –1581.
Hajer, G.R., van Haeften, T.W., and Visseren, F.L.J. (2008). Adipose tissue dysfunction in
obesity, diabetes, and vascular diseases. European Heart Journal 29, 2959 –2971.
Hamaguchi, M., Kojima, T., Takeda, N., Nakagawa, T., Taniguchi, H., Fujii, K., Omatsu, T.,
Nakajima, T., Sarui, H., Shimazaki, M., et al. (2005). The Metabolic Syndrome as a Predictor of
Nonalcoholic Fatty Liver Disease. Annals of Internal Medicine 143, 722 –728.
Han, W., Hata, H., Imbe, H., Liu, Q.-R., Takamatsu, Y., Koizumi, M., Murphy, N.P., Senba, E.,
Uhl, G.R., Sora, I., et al. (2006). Increased body weight in mice lacking mu-opioid receptors.
Neuroreport 17, 941–944.
Hanley, A.J.G., Williams, K., Festa, A., Wagenknecht, L.E., D’Agostino, R.B., and Haffner, S.M.
(2005). Liver Markers and Development of the Metabolic Syndrome. Diabetes 54, 3140 –3147.
Hansen, J.B., and Kristiansen, K. (2006a). Regulatory circuits controlling white versus brown
adipocyte differentiation. Biochem J 398, 153–168.
Hansen, J.B., and Kristiansen, K. (2006b). Regulatory circuits controlling white versus brown
adipocyte differentiation. Biochem J 398, 153–168.
Hansen, J.B., and Kristiansen, K. (2006c). Regulatory circuits controlling white versus brown
adipocyte differentiation. Biochem. J. 398, 153–168.
Hansen, J.B., Jørgensen, C., Petersen, R.K., Hallenborg, P., De Matteis, R., Bøye, H.A., Petrovic,
N., Enerbäck, S., Nedergaard, J., Cinti, S., et al. (2004). Retinoblastoma protein functions as a
molecular switch determining white versus brown adipocyte differentiation. Proc. Natl. Acad.
Sci. U.S.A. 101, 4112–4117.
Hasegawa, K., Kar, S., and Carr, B.I. (1994). Stimulation of hepatocyte DNA synthesis by
neurotensin. J. Cell. Physiol. 158, 215–222.
Heber, D., and Carpenter, C.L. (2011). Addictive Genes and the Relationship to Obesity and
Inflammation. Mol Neurobiol 44, 160–165.
361
Heilbronn, L., Smith, S.R., and Ravussin, E. (2004). Failure of fat cell proliferation,
mitochondrial function and fat oxidation results in ectopic fat storage, insulin resistance and type
II diabetes mellitus. International Journal of Obesity 28, S12–S21.
Herman, M.A., and Kahn, B.B. (2006). Glucose transport and sensing in the maintenance of
glucose homeostasis and metabolic harmony. J Clin Invest 116, 1767–1775.
Herzig, S., Long, F., Jhala, U.S., Hedrick, S., Quinn, R., Bauer, A., Rudolph, D., Schutz, G.,
Yoon, C., Puigserver, P., et al. (2001). CREB regulates hepatic gluconeogenesis through the
coactivator PGC-1. Nature 413, 179–183.
Himms-Hagen, J., Melnyk, A., Zingaretti, M.C., Ceresi, E., Barbatelli, G., and Cinti, S. (2000).
Multilocular fat cells in WAT of CL-316243-treated rats derive directly from white adipocytes.
Am. J. Physiol., Cell Physiol. 279, C670–681.
Hirschey, M.D., Shimazu, T., Goetzman, E., Jing, E., Schwer, B., Lombard, D.B., Grueter, C.A.,
Harris, C., Biddinger, S., Ilkayeva, O.R., et al. (2010). SIRT3 regulates mitochondrial fatty-acid
oxidation by reversible enzyme deacetylation. Nature 464, 121–125.
Hirschey, M.D., Shimazu, T., Jing, E., Grueter, C.A., Collins, A.M., Aouizerat, B., Stančáková,
A., Goetzman, E., Lam, M.M., Schwer, B., et al. (2011). SIRT3 Deficiency and Mitochondrial
Protein Hyperacetylation Accelerate the Development of the Metabolic Syndrome. Molecular
Cell 44, 177–190.
Hofsø, D., Nordstrand, N., Johnson, L.K., Karlsen, T.I., Hager, H., Jenssen, T., Bollerslev, J.,
Godang, K., Sandbu, R., Røislien, J., et al. (2010). Obesity-related cardiovascular risk factors
after weight loss: a clinical trial comparing gastric bypass surgery and intensive lifestyle
intervention. Eur. J. Endocrinol. 163, 735–745.
Holdstock, C., Engstrom, B.E., Ohrvall, M., Lind, L., Sundbom, M., and Karlsson, F.A. (2003).
Ghrelin and Adipose Tissue Regulatory Peptides: Effect of Gastric Bypass Surgery in Obese
Humans. J Clin Endocrinol Metab 88, 3177–3183.
Hondares, E., Rosell, M., Díaz-Delfín, J., Olmos, Y., Monsalve, M., Iglesias, R., Villarroya, F.,
and Giralt, M. (2011). Peroxisome Proliferator-activated Receptor α (PPARα) Induces PPARγ
Coactivator 1α (PGC-1α) Gene Expression and Contributes to Thermogenic Activation of Brown
Fat. Journal of Biological Chemistry 286, 43112 –43122.
Hossain, P., Kawar, B., and El Nahas, M. (2007). Obesity and Diabetes in the Developing World
— A Growing Challenge. New England Journal of Medicine 356, 213–215.
Huang, Z.H., Reardon, C.A., and Mazzone, T. (2006). Endogenous ApoE Expression Modulates
Adipocyte Triglyceride Content and Turnover. Diabetes 55, 3394 –3402.
Huggett, J., Dheda, K., Bustin, S., and Zumla, A. (2005). Real-time RT-PCR normalisation;
strategies and considerations. Genes Immun. 6, 279–284.
362
Ibrahim, M.M. (2010). Subcutaneous and visceral adipose tissue: structural and functional
differences. Obesity Reviews 11, 11–18.
Inui, A. (2001). Ghrelin: An orexigenic and somatotrophic signal from the stomach. Nat Rev
Neurosci 2, 551–560.
Inui, A., Asakawa, A., Bowers, C.Y., Mantovani, G., Laviano, A., Meguid, M.M., and Fujimiya,
M. (2004). Ghrelin, appetite, and gastric motility: the emerging role of the stomach as an
endocrine organ. FASEB J 18, 439–456.
Ivanov, D.V., Tyazhelova, T.V., Lemonnier, L., Kononenko, N., Pestova, A.A., Nikitin, E.A.,
Prevarskaya, N., Skryma, R., Panchin, Y.V., Yankovsky, N.K., et al. (2003). A new human gene
KCNRG encoding potassium channel regulating protein is a cancer suppressor gene candidate
located in 13q14.3. FEBS Lett. 539, 156–160.
Izzo, A.A., and Camilleri, M. (2008). Emerging role of cannabinoids in gastrointestinal and liver
diseases: basic and clinical aspects. Gut 57, 1140–1155.
Jarrar, M.H., Baranova, A., Collantes, R., Ranard, B., Stepanova, M., Bennett, C., Fang, Y.,
Elariny, H., Goodman, Z., Chandhoke, V., et al. (2007). Adipokines and cytokines in nonalcoholic fatty liver disease. Alimentary Pharmacology & Therapeutics 27, 412–421.
Jing, E., Gesta, S., and Kahn, C.R. (2007). SIRT2 Regulates Adipocyte Differentiation through
FoxO1 Acetylation/Deacetylation. Cell Metabolism 6, 105–114.
Jo Vandesompele, Kubista, M., and Pfaffl, M.W. (2009). Reference Gene Validation Software for
Improved Normalization. In Real-time PCR : Current Technology and Applications, (Horizon
Scientific Press),.
De Jonge, H.J.M., Fehrmann, R.S.N., de Bont, E.S.J.M., Hofstra, R.M.W., Gerbens, F., Kamps,
W.A., de Vries, E.G.E., van der Zee, A.G.J., te Meerman, G.J., and ter Elst, A. (2007). Evidence
Based Selection of Housekeeping Genes. PLoS ONE 2, e898.
Kajimura, S., Seale, P., Tomaru, T., Erdjument-Bromage, H., Cooper, M.P., Ruas, J.L., Chin, S.,
Tempst, P., Lazar, M.A., and Spiegelman, B.M. (2008). Regulation of the brown and white fat
gene programs through a PRDM16/CtBP transcriptional complex. Genes Dev. 22, 1397–1409.
Kajimura, S., Seale, P., Kubota, K., Lunsford, E., Frangioni, J.V., Gygi, S.P., and Spiegelman,
B.M. (2009). Initiation of myoblast/brown fat switch through a PRDM16-C/EBP-β transcriptional
complex. Nature 460, 1154–1158.
Karam, S.M. (2010). A focus on parietal cells as a renewing cell population. World Journal of
Gastroenterology 16, 538–546.
Keesey, R.E., and Powley, T.L. (2008). BODY ENERGY HOMEOSTASIS. Appetite 51, 442–
445.
363
Kershaw, E.E., and Flier, J.S. (2004). Adipose Tissue as an Endocrine Organ. J Clin Endocrinol
Metab 89, 2548–2556.
Kersten, S. (2001). Mechanisms of nutritional and hormonal regulation of lipogenesis. EMBO
Rep 2, 282–286.
Kim, C.H., and Younossi, Z.M. (2008). Nonalcoholic Fatty Liver Disease: A Manifestation of the
Metabolic Syndrome. Cleveland Clinic Journal of Medicine 75, 721–728.
Kim, S., and Kim, T. (2003). Selection of optimal internal controls for gene expression profiling
of liver disease. BioTechniques 35, 456–458, 460.
Kim, J.K., Gavrilova, O., Chen, Y., Reitman, M.L., and Shulman, G.I. (2000). Mechanism of
Insulin Resistance in A-ZIP/F-1 Fatless Mice. J. Biol. Chem. 275, 8456–8460.
Klein, J., Perwitz, N., Kraus, D., and Fasshauer, M. (2006). Adipose tissue as source and target
for novel therapies. Trends in Endocrinology & Metabolism 17, 26–32.
Knutti, D., and Kralli, A. (2001). PGC-1, a versatile coactivator. Trends Endocrinol Metab 12,
360–365.
Kojima, M., and Kangawa, K. (2005). Ghrelin: Structure and Function. Physiological Reviews
85, 495 –522.
Kojima, M., Hosoda, H., Date, Y., Nakazato, M., Matsuo, H., and Kangawa, K. (1999). Ghrelin is
a growth-hormone-releasing acylated peptide from stomach. Nature 402, 656–660.
Kong, F., and Singh, R. p. (2008). Disintegration of Solid Foods in Human Stomach. Journal of
Food Science 73, R67–R80.
Kong X, Wang R, Xue Y, Liu X, Zhang H, Chen Y, Fang F, and Chang Y (2010). Sirtuin 3, a
new target of PGC-1alpha, plays an important role in the suppression of ROS and mitochondrial
biogenesis. PloS One.
Kornman, K.S. (2006). Interleukin 1 genetics, inflammatory mechanisms, and nutrigenetic
opportunities to modulate diseases of aging. Am J Clin Nutr 83, 475S–483S.
Koska, J., Stefan, N., Votruba, S.B., Smith, S.R., Krakoff, J., and Bunt, J.C. (2008). Distribution
of Subcutaneous Fat Predicts Insulin Action in Obesity in Sex-specific Manner. Obesity 16,
2003–2009.
Kozak, L.P., and Anunciado-Koza, R. (2008). UCP1: its involvement and utility in obesity. Int J
Obes 32, S32–S38.
Kralisch, S., Klein, J., Lossner, U., Bluher, M., Paschke, R., Stumvoll, M., and Fasshauer, M.
(2005a). Hormonal regulation of the novel adipocytokine visfatin in 3T3-L1 adipocytes. Journal
of Endocrinology 185, R1 –R8.
364
Kralisch, S., Klein, J., Lossner, U., Bluher, M., Paschke, R., Stumvoll, M., and Fasshauer, M.
(2005b). Interleukin-6 is a negative regulator of visfatin gene expression in 3T3-L1 adipocytes.
American Journal of Physiology - Endocrinology And Metabolism 289, E586 –E590.
Krishnan, J., Danzer, C., Simka, T., Ukropec, J., Walter, K.M., Kumpf, S., Mirtschink, P.,
Ukropcova, B., Gasperikova, D., Pedrazzini, T., et al. (2012). Dietary Obesity-Associated Hif1α
Activation in Adipocytes Restricts Fatty Acid Oxidation and Energy Expenditure Via
Suppression of the Sirt2-NAD+ System. Genes Dev. 26, 259–270.
Kuboki, S., Shin, T., Huber, N., Eismann, T., Galloway, E., Schuster, R., Blanchard, J., Edwards,
M.J., and Lentsch, A.B. (2008). Hepatocyte signaling through CXC chemokine receptor-2 is
detrimental to liver recovery after ischemia/reperfusion in mice. Hepatology 48, 1213–1223.
Kühnel, F., Zender, L., Paul, Y., Tietze, M.K., Trautwein, C., Manns, M., and Kubicka, S. (2000).
NFκB Mediates Apoptosis through Transcriptional Activation of Fas (CD95) in Adenoviral
Hepatitis. Journal of Biological Chemistry 275, 6421 –6427.
Kypreos, K.E., Karagiannides, I., Fotiadou, E.H., Karavia, E.A., Brinkmeier, M.S., Giakoumi,
S.M., and Tsompanidi, E.M. (2009). Mechanisms of obesity and related pathologies: Role of
apolipoprotein E in the development of obesity. FEBS Journal 276, 5720–5728.
Lacourse, K.A., Friis-Hansen, L., Rehfeld, J.F., and Samuelson, L.C. (1997). Disturbed progastrin
processing in carboxypeptidase E-deficient fat mice. FEBS Lett. 416, 45–50.
Lafontan, M., and Girard, J. (2008). Impact of visceral adipose tissue on liver metabolism: Part I:
Heterogeneity of adipose tissue and functional properties of visceral adipose tissue. Diabetes &
Metabolism 34, 317–327.
Large, V., Peroni, O., Letexier, D., Ray, H., and Beylot, M. (2004). Metabolism of lipids in
human white adipocyte. Diabetes Metab. 30, 294–309.
Lee, D.-E., Kehlenbrink, S., Lee, H., Hawkins, M., and Yudkin, J.S. (2009). Getting the message
across: mechanisms of physiological cross talk by adipose tissue. American Journal of
Physiology - Endocrinology And Metabolism 296, E1210 –E1229.
Lee, H.-M., Wang, G., Englander, E.W., Kojima, M., and Greeley, G.H. (2002). Ghrelin, A New
Gastrointestinal Endocrine Peptide that Stimulates Insulin Secretion: Enteric Distribution,
Ontogeny, Influence of Endocrine, and Dietary Manipulations. Endocrinology 143, 185–190.
Lee, P., Ho, K.K.Y., Lee, P., Greenfield, J.R., Ho, K.K.Y., and Greenfield, J.R. (2011). Hot fat in
a cool man: infrared thermography and brown adipose tissue. Diabetes, Obesity and Metabolism
13, 92–93.
Leonardsson, G., Steel, J.H., Christian, M., Pocock, V., Milligan, S., Bell, J., So, P.-W., MedinaGomez, G., Vidal-Puig, A., White, R., et al. (2004). Nuclear receptor corepressor RIP140
regulates fat accumulation. Proc. Natl. Acad. Sci. U.S.A. 101, 8437–8442.
365
Leone, T.C., Lehman, J.J., Finck, B.N., Schaeffer, P.J., Wende, A.R., Boudina, S., Courtois, M.,
Wozniak, D.F., Sambandam, N., Bernal-Mizrachi, C., et al. (2005). PGC-1α Deficiency Causes
Multi-System Energy Metabolic Derangements: Muscle Dysfunction, Abnormal Weight Control
and Hepatic Steatosis. PLoS Biol 3, e101.
Li, X., and Kazgan, N. (2011). Mammalian Sirtuins and Energy Metabolism. International
Journal of Biological Sciences 7, 575–587.
Liang, H., and Ward, W.F. (2006). PGC-1α: a key regulator of energy metabolism. Advances in
Physiology Education 30, 145 –151.
Liew, P.-L., Lee, W.-J., Lee, Y.-C., and Chen, W.-Y. (2006). Gastric Ghrelin Expression
Associated with Helicobacter pylori Infection and Chronic Gastritis in Obese Patients. Obesity
Surgery 16, 612–619.
Lin, J., Wu, P.-H., Tarr, P.T., Lindenberg, K.S., St-Pierre, J., Zhang, C., Mootha, V.K., Jäger, S.,
Vianna, C.R., Reznick, R.M., et al. (2004). Defects in Adaptive Energy Metabolism with CNSLinked Hyperactivity in PGC-1α Null Mice. Cell 119, 121–135.
Lin, J., Yang, R., Tarr, P.T., Wu, P.-H., Handschin, C., Li, S., Yang, W., Pei, L., Uldry, M.,
Tontonoz, P., et al. (2005a). Hyperlipidemic Effects of Dietary Saturated Fats Mediated through
PGC-1β Coactivation of SREBP. Cell 120, 261–273.
Lin, J., Handschin, C., and Spiegelman, B.M. (2005b). Metabolic control through the PGC-1
family of transcription coactivators. Cell Metabolism 1, 361–370.
Linder, K., Arner, P., Flores-Morales, A., Tollet-Egnell, P., and Norstedt, G. (2004).
Differentially expressed genes in visceral or subcutaneous adipose tissue of obese men and
women. Journal of Lipid Research 45, 148 –154.
Logan, J., Edwards, K., and Saunders, N. (2009). Real-time PCR: current technology and
applications (Horizon Scientific Press).
Lombard, D.B., Alt, F.W., Cheng, H.-L., Bunkenborg, J., Streeper, R.S., Mostoslavsky, R., Kim,
J., Yancopoulos, G., Valenzuela, D., Murphy, A., et al. (2007). Mammalian Sir2 Homolog SIRT3
Regulates Global Mitochondrial Lysine Acetylation. Molecular and Cellular Biology 27, 8807 –
8814.
Luciani, L., Camis, M., and Sir Gordon Holmes (1913). Human Physiology: Internal secretion,
digestion, excretion, the skin (Macmillan and co.).
Ma, Y., Liu, H., Tu-Rapp, H., Thiesen, H.-J., Ibrahim, S.M., Cole, S.M., and Pope, R.M. (2004).
Fas ligation on macrophages enhances IL-1R1-Toll-like receptor 4 signaling and promotes
chronic inflammation. Nat Immunol 5, 380–387.
Machado, M., Marques-Vidal, P., and Cortez-Pinto, H. (2006). Hepatic histology in obese
patients undergoing bariatric surgery. Journal of Hepatology 45, 600–606.
366
MacLaren, R., Cui, W., Lu, H., Simard, S., and Cianflone, K. (2010). Association of adipocyte
genes with ASP expression: a microarray analysis of subcutaneous and omental adipose tissue in
morbidly obese subjects. BMC Medical Genomics 3, 3.
Malhi, H., Guicciardi, M.E., and Gores, G.J. (2010). Hepatocyte Death: A Clear and Present
Danger. Physiological Reviews 90, 1165 –1194.
Marchesini, G., Bugianesi, E., Forlani, G., Cerrelli, F., Lenzi, M., Manini, R., Natale, S., Vanni,
E., Villanova, N., Melchionda, N., et al. (2003). Nonalcoholic fatty liver, steatohepatitis, and the
metabolic syndrome. Hepatology 37, 917–923.
Mariman, E., and Wang, P. (2010). Adipocyte extracellular matrix composition, dynamics and
role in obesity. Cellular and Molecular Life Sciences 67, 1277–1292–1292.
Van Marken Lichtenbelt, W.D., Vanhommerig, J.W., Smulders, N.M., Drossaerts, J.M.A.F..,
Kemerink, G.J., Bouvy, N.D., Schrauwen, P., and Teule, G.J.J. (2009). Cold-Activated Brown
Adipose Tissue in Healthy Men. New England Journal of Medicine 360, 1500–1508.
Marks, A.R., and Neill, U.S. (2007). Science in medicine: the JCI textbook of molecular medicine
(Jones & Bartlett Publishers).
Maton, Anthea, Jean Hopkins, Charles William McLaughlin, Susan Johnson, Maryanna Quon
Warner, David LaHart, and Jill D. Wright (1993). Human Biology and Health (Englewood Cliffs
New Jersey, USA: Prentice Hall).
Matsuki, T., Horai, R., Sudo, K., and Iwakura, Y. (2003). IL-1 Plays an Important Role in Lipid
Metabolism by Regulating Insulin Levels under Physiological Conditions. J Exp Med 198, 877–
888.
Matteoni*, C.A., Younossi*, Z.M., Gramlich‡, T., Boparai§, N., Liu∥, Y.C., and McCullough¶,
A.J. (1999). Nonalcoholic fatty liver disease: A spectrum of clinical and pathological severity.
Gastroenterology 116, 1413–1419.
Mehta, R., Birerdinc, A., Hossain, N., Afendy, A., Chandhoke, V., Younossi, Z., and Baranova,
A. (2010). Validation of endogenous reference genes for qRT-PCR analysis of human visceral
adipose samples. BMC Molecular Biology 11, 39.
Michishita, E., Park, J.Y., Burneskis, J.M., Barrett, J.C., and Horikawa, I. (2005). Evolutionarily
Conserved and Nonconserved Cellular Localizations and Functions of Human SIRT Proteins.
Molecular Biology of the Cell 16, 4623 –4635.
Mix, H., Widjaja, A., Jandl, O., Cornberg, M., Kaul, A., Göke, M., Beil, W., Kuske, M., Brabant,
G., Manns, M.P., et al. (2000). Expression of leptin and leptin receptor isoforms in the human
stomach. Gut 47, 481 –486.
Montani, J.-P., Carroll, J.F., Dwyer, T.M., Antic, V., Yang, Z., and Dulloo, A.G. (2004). Ectopic
fat storage in heart, blood vessels and kidneys in the pathogenesis of cardiovascular diseases.
International Journal of Obesity 28, S58–S65.
367
Mora, S., and Pessin, J.E. (2002). An adipocentric view of signaling and intracellular trafficking.
Diabetes Metab. Res. Rev. 18, 345–356.
Morton, G.J., and Schwartz, M.W. (2011). Leptin and the Central Nervous System Control of
Glucose Metabolism. Physiological Reviews 91, 389 –411.
Moschen, A.R., Kaser, A., Enrich, B., Mosheimer, B., Theurl, M., Niederegger, H., and Tilg, H.
(2007). Visfatin, an Adipocytokine with Proinflammatory and Immunomodulating Properties.
The Journal of Immunology 178, 1748 –1758.
Murphy, K.G., and Bloom, S.R. (2006). Gut hormones and the regulation of energy homeostasis.
Nature 444, 854–859.
Naggert, J.K., Fricker, L.D., Varlamov, O., Nishina, P.M., Rouille, Y., Steiner, D.F., Carroll, R.J.,
Paigen, B.J., and Leiter, E.H. (1995). Hyperproinsulinaemia in obese fat/fat mice associated with
a carboxypeptidase E mutation which reduces enzyme activity. Nature Genetics 10, 135–142.
Nakazato, M., Murakami, N., Date, Y., Kojima, M., Matsuo, H., Kangawa, K., and Matsukura, S.
(2001). A role for ghrelin in the central regulation of feeding. Nature 409, 194–198.
Nandagopal, R., Brown, R.J., and Rother, K.I. (2011). Resolution of Type 2 Diabetes Following
Bariatric Surgery: Implications for Adults and Adolescents. Diabetes Technology & Therapeutics
12, 671–677.
Navasa, M., Gordon, D.A., Hariharan, N., Jamil, H., Shigenaga, J.K., Moser, A., Fiers, W.,
Pollock, A., Grunfeld, C., and Feingold, K.R. (1998). Regulation of microsomal triglyceride
transfer protein mRNA expression by endotoxin and cytokines. J. Lipid Res. 39, 1220–1230.
Nedergaard, J., Bengtsson, T., and Cannon, B. (2007). Unexpected evidence for active brown
adipose tissue in adult humans. American Journal of Physiology - Endocrinology And
Metabolism 293, E444 –E452.
Nguyen, P., Leray, V., Diez, M., Serisier, S., Bloc’h, J.L., Siliart, B., and Dumon, H. (2008).
Liver lipid metabolism. Journal of Animal Physiology and Animal Nutrition 92, 272–283.
Nicholls, D.G., and Locke, R.M. (1984). Thermogenic mechanisms in brown fat. Physiol. Rev.
64, 1–64.
North, B.J., Marshall, B.L., Borra, M.T., Denu, J.M., and Verdin, E. (2003). The Human Sir2
Ortholog, SIRT2, Is an NAD+-Dependent Tubulin Deacetylase. Molecular Cell 11, 437–444.
Oka, R., Miura, K., Sakurai, M., Nakamura, K., Yagi, K., Miyamoto, S., Moriuchi, T., Mabuchi,
H., Yamagishi, M., Takeda, Y., et al. (2009). Comparison of waist circumference with body mass
index for predicting abdominal adipose tissue. Diabetes Research and Clinical Practice 83, 100–
105.
368
Oki, K., Yamane, K., Kamei, N., Nojima, H., and Kohno, N. (2007). Circulating visfatin level is
correlated with inflammation, but not with insulin resistance. Clinical Endocrinology 67, 796–
800.
Oo, Y.H., Shetty, S., and Adams, D.H. (2010). The Role of Chemokines in the Recruitment of
Lymphocytes to the Liver. Dig Dis 28, 31–44.
Ouchi, N., Parker, J.L., Lugus, J.J., and Walsh, K. (2011). Adipokines in inflammation and
metabolic disease. Nat Rev Immunol 11, 85–97.
Pacher, P., Bátkai, S., and Kunos, G. (2006). The Endocannabinoid System as an Emerging
Target of Pharmacotherapy. Pharmacol Rev 58, 389–462.
Pagano, C., Pilon, C., Olivieri, M., Mason, P., Fabris, R., Serra, R., Milan, G., Rossato, M.,
Federspil, G., and Vettor, R. (2006). Reduced Plasma Visfatin/Pre-B Cell Colony-Enhancing
Factor in Obesity Is Not Related to Insulin Resistance in Humans. Journal of Clinical
Endocrinology & Metabolism 91, 3165 –3170.
Pagano, C., Pilon, C., Calcagno, A., Urbanet, R., Rossato, M., Milan, G., Bianchi, K., Rizzuto,
R., Bernante, P., Federspil, G., et al. (2007). The Endogenous Cannabinoid System Stimulates
Glucose Uptake in Human Fat Cells via Phosphatidylinositol 3-Kinase and Calcium-Dependent
Mechanisms. JCEM 92, 4810–4819.
Pahl, H.L. (1999). Activators and target genes of Rel/NF-kappaB transcription factors. Oncogene
18, 6853–6866.
Palou, A., Picó, C., Bonet, M.L., and Oliver, P. (1998a). The uncoupling protein, thermogenin.
The International Journal of Biochemistry & Cell Biology 30, 7–11.
Palou, A., Picó, C., Bonet, M.L., and Oliver, P. (1998b). The uncoupling protein, thermogenin.
The International Journal of Biochemistry & Cell Biology 30, 7–11.
Pan, D., Fujimoto, M., Lopes, A., and Wang, Y.-X. (2009a). Twist-1 is a PPAR δ-inducible,
negative feedback regulator of PGC-1 α in brown fat metabolism. Cell 137, 73–86.
Pan, D., Fujimoto, M., Lopes, A., and Wang, Y.-X. (2009b). Twist-1 Is a PPARδ-Inducible,
Negative-Feedback Regulator of PGC-1α in Brown Fat Metabolism. Cell 137, 73–86.
Park, D.R., Thomsen, A.R., Frevert, C.W., Pham, U., Skerrett, S.J., Kiener, P.A., and Liles, W.C.
(2003). Fas (CD95) Induces Proinflammatory Cytokine Responses by Human Monocytes and
Monocyte-Derived Macrophages. The Journal of Immunology 170, 6209 –6216.
Pattrick, M., Luckett, J., Yue, L., and Stover, C. (2009). Dual role of complement in adipose
tissue. Molecular Immunology 46, 755–760.
Perrier, S., Darakhshan, F., and Hajduch, E. (2006). IL-1 receptor antagonist in metabolic
diseases: Dr Jekyll or Mr Hyde? FEBS Letters 580, 6289–6294.
369
Perrod, S., Cockell, M.M., Laroche, T., Renauld, H., Ducrest, A.-L., Bonnard, C., and Gasser,
S.M. (2001). A cytosolic NAD-dependent deacetylase, Hst2p, can modulate nucleolar and
telomeric silencing in yeast. EMBO J 20, 197–209.
Peter, M.E., and Krammer, P.H. (1998). Mechanisms of CD95 (APO-1/Fas)-mediated apoptosis.
Current Opinion in Immunology 10, 545–551.
Petersen, K.F., Dufour, S., Befroy, D., Lehrke, M., Hendler, R.E., and Shulman, G.I. (2005).
Reversal of Nonalcoholic Hepatic Steatosis, Hepatic Insulin Resistance, and Hyperglycemia by
Moderate Weight Reduction in Patients With Type 2 Diabetes. Diabetes 54, 603 –608.
Pfaffl, M.W., Tichopad, A., Prgomet, C., and Neuvians, T.P. (2004). Determination of stable
housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper –
Excel-based tool using pair-wise correlations. Biotechnology Letters 26, 509–515–515.
Picard, F., Géhin, M., Annicotte, J.-S., Rocchi, S., Champy, M.-F., O’Malley, B.W., Chambon,
P., and Auwerx, J. (2002). SRC-1 and TIF2 control energy balance between white and brown
adipose tissues. Cell 111, 931–941.
Picó, C., Herron, D., Palou, A., Jacobsson, A., Cannon, B., and Nedergaard, J. (1994).
Stabilization of the mRNA for the uncoupling protein thermogenin by transcriptional/translational
blockade and by noradrenaline in brown adipocytes differentiated in culture: a degradation factor
induced by cessation of stimulation? Biochem J 302, 81–86.
Picó, C., Oliver, P., Sánchez, J., and Palou, A. (2003). Gastric leptin: a putative role in the shortterm regulation of food intake. Br. J. Nutr. 90, 735–741.
Pihlajamäki, J., Kuulasmaa, T., Kaminska, D., Simonen, M., Kärjä, V., Grönlund, S., Käkelä, P.,
Pääkkönen, M., Kainulainen, S., Punnonen, K., et al. (2012). Serum interleukin 1 receptor
antagonist as an independent marker of non-alcoholic steatohepatitis in humans. Journal of
Hepatology 56, 663–670.
Pinnick, K.E., Collins, S.C., Londos, C., Gauguier, D., Clark, A., and Fielding, B.A. (2008).
Pancreatic Ectopic Fat Is Characterized by Adipocyte Infiltration and Altered Lipid Composition.
Obesity 16, 522–530.
Plum, L., Lin, H.V., Dutia, R., Tanaka, J., Aizawa, K.S., Matsumoto, M., Kim, A.J., Cawley,
N.X., Paik, J.-H., Loh, Y.P., et al. (2009). The obesity susceptibility gene Cpe links FoxO1
signaling in hypothalamic pro-opiomelanocortin neurons with regulation of food intake. Nature
Medicine 15, 1195–1201.
Ponton, A., Cl�ment, M.-V., and Stamenkovic, I. (1996). The CD95 (APO-1/Fas) Receptor
Activates NF-kappaB Independently of Its Cytotoxic Function. J. Biol. Chem. 271, 8991–8995.
Powelka, A.M., Seth, A., Virbasius, J.V., Kiskinis, E., Nicoloro, S.M., Guilherme, A., Tang, X.,
Straubhaar, J., Cherniack, A.D., Parker, M.G., et al. (2006). Suppression of oxidative metabolism
and mitochondrial biogenesis by the transcriptional corepressor RIP140 in mouse adipocytes. J.
Clin. Invest. 116, 125–136.
370
Priya, R., Singhal Suruchi, N Leelavathy, Rao Vani Vijay, and R Roopa (2008). Basics in Human
Anatomy (Jaypee Brothers Medical Publishers (P) Ltd).
Puigserver, P., Wu, Z., Park, C.W., Graves, R., Wright, M., and Spiegelman, B.M. (1998). A
Cold-Inducible Coactivator of Nuclear Receptors Linked to Adaptive Thermogenesis. Cell 92,
829–839.
Qureshi, K., and Abrams, G.A. (2007). Metabolic liver disease of obesity and role of adipose
tissue in the pathogenesis of nonalcoholic fatty liver disease. World J Gastroenterol 13, 3540–
3553.
Rafiq, N., and Younossi, Z.M. (2011). Interaction of metabolic syndrome, nonalcoholic fatty liver
disease and chronic hepatitis C. Expert Review of Gastroenterology & Hepatology 2, 207–215.
Ramalho, R.M., Cortez-Pinto, H., Castro, R.E., Solá, S., Costa, A., Moura, M.C., Camilo, M.E.,
and Rodrigues, C.M.P. (2006). Apoptosis and Bcl-2 expression in the livers of patients with
steatohepatitis. European Journal of Gastroenterology & Hepatology 18.
Ravi, R., Bedi, A., Fuchs, E.J., and Bedi, A. (1998). CD95 (Fas)-induced Caspase-mediated
Proteolysis of NF-κB. Cancer Research 58, 882 –886.
Ravnskjaer, K., Frigerio, F., Boergesen, M., Nielsen, T., Maechler, P., and Mandrup, S. (2010).
PPARdelta is a fatty acid sensor that enhances mitochondrial oxidation in insulin-secreting cells
and protects against fatty acid-induced dysfunction. J. Lipid Res. 51, 1370–1379.
Ravussin, E., and Bogardus, C. (2000). Energy balance and weight regulation: genetics versus
environment. British Journal of Nutrition 83, S17–S20 M3–M10.1017/S0007114500000908.
Reddy, J.K., and Rao, M.S. (2006). Lipid Metabolism and Liver Inflammation. II. Fatty liver
disease and fatty acid oxidation. Am J Physiol Gastrointest Liver Physiol 290, G852–G858.
Reddy, J.K., and Sambasiva Rao, M. (2006). Lipid Metabolism and Liver Inflammation. II. Fatty
liver disease and fatty acid oxidation. American Journal of Physiology - Gastrointestinal and
Liver Physiology 290, G852 –G858.
Rehfeld, J.F. (1998). The new biology of gastrointestinal hormones. Physiol. Rev. 78, 1087–
1108.
Rehfeld, J.F. (2004). A centenary of gastrointestinal endocrinology. Horm. Metab. Res. 36, 735–
741.
Revollo, J.R., Körner, A., Mills, K.F., Satoh, A., Wang, T., Garten, A., Dasgupta, B., Sasaki, Y.,
Wolberger, C., Townsend, R.R., et al. (2007). Nampt/PBEF/Visfatin Regulates Insulin Secretion
in β Cells as a Systemic NAD Biosynthetic Enzyme. Cell Metabolism 6, 363–375.
Reyes, T.M., and Sawchenko, P.E. (2002). Involvement of the Arcuate Nucleus of the
Hypothalamus in Interleukin-1-Induced Anorexia. J. Neurosci. 22, 5091–5099.
371
Rhee, J., Inoue, Y., Yoon, J.C., Puigserver, P., Fan, M., Gonzalez, F.J., and Spiegelman, B.M.
(2003). Regulation of hepatic fasting response by PPARγ coactivator-1α (PGC-1): Requirement
for hepatocyte nuclear factor 4α in gluconeogenesis. Proceedings of the National Academy of
Sciences 100, 4012 –4017.
Rho, H.-W., Lee, B.-C., Choi, E.-S., Choi, I.-J., Lee, Y.-S., and Goh, S.-H. (2010). Identification
of valid reference genes for gene expression studies of human stomach cancer by reverse
transcription-qPCR. BMC Cancer 10, 240.
Ribeiro, P.S., Cortez-Pinto, H., Sola, S., Castro, R.E., Ramalho, R.M., Baptista, A., Moura, M.C.,
Camilo, M.E., and Rodrigues, C.M. (2004). Hepatocyte Apoptosis, Expression of Death
Receptors, and Activation of NF-[kappa]B in the Liver of Nonalcoholic and Alcoholic
Steatohepatitis Patients. Am J Gastroenterol 99, 1708–1717.
Ricquier, D., and Bouillaud, F. (2000). Mitochondrial uncoupling proteins: from mitochondria to
the regulation of energy balance. J Physiol 529, 3–10.
Rohini Mehta, Aybike Birerdinc, Noreen Hossain, Amir Moazzez, Vikas Chandhoke, Ancha
Baranova, and Zobair M. Younossi (2010). Profiling and Validation of Reference Genes in the
Visceral Adipose Tissue(VAT). DDW.
Rosen, E.D., and Spiegelman, B.M. (2006). Adipocytes as regulators of energy balance and
glucose homeostasis. Nature 444, 847–853.
Rosen, E.D., Walkey, C.J., Puigserver, P., and Spiegelman, B.M. (2000). Transcriptional
Regulation of Adipogenesis. Genes Dev. 14, 1293–1307.
Rovere, C., Viale, A., Nahon, J., and Kitabgi, P. (1996). Impaired processing of brain
proneurotensin and promelanin-concentrating hormone in obese fat/fat mice. Endocrinology 137,
2954–2958.
Roy A., F. (2000). Phylogenetic Classification of Prokaryotic and Eukaryotic Sir2-like Proteins.
Biochemical and Biophysical Research Communications 273, 793–798.
Rust, C., and Gores, G.J. (2000). Apoptosis and liver disease. The American Journal of Medicine
108, 567–574.
Ryysy, L., Häkkinen, A.M., Goto, T., Vehkavaara, S., Westerbacka, J., Halavaara, J., and YkiJärvinen, H. (2000). Hepatic fat content and insulin action on free fatty acids and glucose
metabolism rather than insulin absorption are associated with insulin requirements during insulin
therapy in type 2 diabetic patients. Diabetes 49, 749–758.
S Cinti, R De Matteis, E Ceresi, C Picó, J Oliver, P Oliver, A Palou, A Obrador, and C Maffeis
(2001). Leptin in the human stomach. Gut 49, 155.
Sahu, A. (1998). Evidence suggesting that galanin (GAL), melanin-concentrating hormone
(MCH), neurotensin (NT), proopiomelanocortin (POMC) and neuropeptide Y (NPY) are targets
of leptin signaling in the hypothalamus. Endocrinology 139, 795–798.
372
Sakata, I., and Sakai, T. (2010). Ghrelin Cells in the Gastrointestinal Tract. International Journal
of Peptides 2010.
Sale, M.M., Freedman, B.I., Langefeld, C.D., Williams, A.H., Hicks, P.J., Colicigno, C.J., Beck,
S.R., Brown, W.M., Rich, S.S., and Bowden, D.W. (2004). A Genome-Wide Scan for Type 2
Diabetes in African-American Families Reveals Evidence for a Locus on Chromosome 6q.
Diabetes 53, 830–837.
Sam, S., Haffner, S., Davidson, M.H., D’Agostino, R.B., Feinstein, S., Kondos, G., Perez, A., and
Mazzone, T. (2008). Relationship of Abdominal Visceral and Subcutaneous Adipose Tissue With
Lipoprotein Particle Number and Size in Type 2 Diabetes. Diabetes 57, 2022 –2027.
Samaras, K., Botelho, N.K., Chisholm, D.J., and Lord, R.V. (2010). Subcutaneous and Visceral
Adipose Tissue Gene Expression of Serum Adipokines That Predict Type 2 Diabetes. Obesity 18,
884–889.
Schaffer, J. (2003). Lipotoxicity: when tissues overeat. [Miscellaneous Article]. Current Opinion
in Lipidology June 2003 14, 281–287.
Scher, M.B., Vaquero, A., and Reinberg, D. (2007). SirT3 is a nuclear NAD+-dependent histone
deacetylase that translocates to the mitochondria upon cellular stress. Genes & Development 21,
920 –928.
Schmidt, A., Endo, N., Rutledge, S.J., Vogel, R., Shinar, D., and Rodan, G.A. (1992).
Identification of a New Member of the Steroid Hormone Receptor Superfamily That Is Activated
by a Peroxisome Proliferator and Fatty Acids. Molecular Endocrinology 6, 1634–1641.
Schmittgen, T.D., and Zakrajsek, B.A. (2000). Effect of experimental treatment on housekeeping
gene expression: validation by real-time, quantitative RT-PCR. J. Biochem. Biophys. Methods
46, 69–81.
Schuenke, M., Schulte, E., Schumacher, U., and Ross, L. (2010). Neck and Internal Organs
(THIEME Atlas of Anatomy) (Thieme).
Schwabe, R.F., and Brenner, D.A. (2006). Mechanisms of Liver Injury. I. TNF-α-induced liver
injury: role of IKK, JNK, and ROS pathways. Am J Physiol Gastrointest Liver Physiol 290,
G583–G589.
Schwer, B., and Verdin, E. (2008). Conserved Metabolic Regulatory Functions of Sirtuins. Cell
Metabolism 7, 104–112.
Schwer, B., North, B.J., Frye, R.A., Ott, M., and Verdin, E. (2002). The human silent information
regulator (Sir)2 homologue hSIRT3 is a mitochondrial nicotinamide adenine dinucleotide–
dependent deacetylase. The Journal of Cell Biology 158, 647 –657.
Scimè, A., Grenier, G., Huh, M.S., Gillespie, M.A., Bevilacqua, L., Harper, M.-E., and Rudnicki,
M.A. (2005). Rb and p107 regulate preadipocyte differentiation into white versus brown fat
through repression of PGC-1alpha. Cell Metab. 2, 283–295.
373
Seale, P., Bjork, B., Yang, W., Kajimura, S., Chin, S., Kuang, S., Scimè, A., Devarakonda, S.,
Conroe, H.M., Erdjument-Bromage, H., et al. (2008). PRDM16 controls a brown fat/skeletal
muscle switch. Nature 454, 961–967.
Seki, E., De Minicis, S., Gwak, G.-Y., Kluwe, J., Inokuchi, S., Bursill, C.A., Llovet, J.M.,
Brenner, D.A., and Schwabe, R.F. (2009). CCR1 and CCR5 promote hepatic fibrosis in mice. J
Clin Invest 119, 1858–1870.
Semple, R.K., Crowley, V.C., Sewter, C.P., Laudes, M., Christodoulides, C., Considine, R.V.,
Vidal-Puig, A., and O’Rahilly, S. (2003). Expression of the thermogenic nuclear hormone
receptor coactivator PGC-1[alpha] is reduced in the adipose tissue of morbidly obese subjects. Int
J Obes Relat Metab Disord 28, 176–179.
Seppälä-Lindroos, A., Vehkavaara, S., Häkkinen, A.-M., Goto, T., Westerbacka, J., Sovijärvi, A.,
Halavaara, J., and Yki-Järvinen, H. (2002). Fat Accumulation in the Liver Is Associated with
Defects in Insulin Suppression of Glucose Production and Serum Free Fatty Acids Independent of
Obesity in Normal Men. JCEM 87, 3023–3028.
Sethi, J.K., and Vidal-Puig, A.J. (2007). Thematic review series: Adipocyte Biology. Adipose
tissue function and plasticity orchestrate nutritional adaptation. Journal of Lipid Research 48,
1253 –1262.
Shankar, K., Harrell, A., Kang, P., Singhal, R., Ronis, M.J.J., and Badger, T.M. (2010).
Carbohydrate-Responsive Gene Expression in the Adipose Tissue of Rats. Endocrinology 151,
153–164.
Shi, T., Wang, F., Stieren, E., and Tong, Q. (2005). SIRT3, a Mitochondrial Sirtuin Deacetylase,
Regulates Mitochondrial Function and Thermogenesis in Brown Adipocytes. Journal of
Biological Chemistry 280, 13560 –13567.
Shin-ichiro Imai (2009). Nicotinamide Phosphoribosyltransferase (Nampt): A Link Between
NAD Biology, Metabolism, and Diseases. Current Pharmaceutical Design 15, 20–28.
Shoelson, S.E., Herrero, L., and Naaz, A. (2007). Obesity, Inflammation, and Insulin Resistance.
Gastroenterology 132, 2169–2180.
Sims, E.A.H. (2001). Are there persons who are obese, but metabolically healthy? Metabolism
50, 1499–1504.
Snel, M., Jonker, J.T., Schoones, J., Lamb, H., de Roos, A., Pijl, H., Smit, J.W.A., Meinders,
A.E., and Jazet, I.M. (2012). Ectopic Fat and Insulin Resistance: Pathophysiology and Effect of
Diet and Lifestyle Interventions. International Journal of Endocrinology 2012, 1–18.
Somm, E., Cettour-Rose, P., Asensio, C., Charollais, A., Klein, M., Theander-Carrillo, C., JugeAubry, C., Dayer, J., Nicklin, M., Meda, P., et al. (2006). Interleukin-1 receptor antagonist is
upregulated during diet-induced obesity and regulates insulin sensitivity in rodents. Diabetologia
49, 387–393.
374
Stanley, S., Wynne, K., McGowan, B., and Bloom, S. (2005). Hormonal Regulation of Food
Intake. Physiol Rev 85, 1131–1158.
Stienstra, R., Mandard, S., Patsouris, D., Maass, C., Kersten, S., and Muller, M. (2007).
Peroxisome Proliferator-Activated Receptor {alpha} Protects against Obesity-Induced Hepatic
Inflammation. Endocrinology 148, 2753–2763.
Stuart, J.A., Harper, J.A., Brindle, K.M., and Brand, M.D. (1999). Uncoupling protein 2 from
carp and zebrafish, ectothermic vertebrates. Biochimica Et Biophysica Acta (BBA) Bioenergetics 1413, 50–54.
Suzuki, T., Higgins, P.J., and Crawford, D.R. (2000). Control selection for RNA quantitation.
BioTechniques 29, 332–337.
Luk T, Malam Z, and Marshall JC (2008). Pre-B cell colony-enhancing factor (PBEF)/visfatin: a
novel mediator of innate immunity. Journal of Leukocyte Biology 83, 804–801.
Tacke, F., Luedde, T., and Trautwein, C. (2009). Inflammatory Pathways in Liver Homeostasis
and Liver Injury. Clinic Rev Allerg Immunol 36, 4–12.
Tao, R., Wei, D., Gao, H., Liu, Y., DePinho, R.A., and Dong, X.C. (2011). Hepatic FoxOs
Regulate Lipid Metabolism Via Modulation of Expression of the Nicotinamide
Phosphoribosyltransferase Gene. J. Biol. Chem. 286, 14681–14690.
Terra, X., Auguet, T., Quesada, I., Aguilar, C., Luna, A.M., Hernández, M., Sabench, F., Porras,
J.A., Martínez, S., Lucas, A., et al. (2011). Increased levels and adipose tissue expression of
visfatin in morbidly obese women. The relationship with pro-inflammatory cytokines. Clinical
Endocrinology no.
Thellin, O., Zorzi, W., Lakaye, B., De Borman, B., Coumans, B., Hennen, G., Grisar, T., Igout,
A., and Heinen, E. (1999). Housekeeping genes as internal standards: use and limits. Journal of
Biotechnology 75, 291–295.
Tilg, H., and Hotamisligil, G.S. (2006). Nonalcoholic Fatty Liver Disease: Cytokine-Adipokine
Interplay and Regulation of Insulin Resistance. Gastroenterology 131, 934–945.
Tilg, H., and Moschen, A.R. (2008). Role of adiponectin and PBEF/visfatin as regulators of
inflammation: involvement in obesity-associated diseases. Clinical Science 114, 275–288.
Tilg, H., and Moschen, A.R. (2010). Evolution of inflammation in nonalcoholic fatty liver
disease: The multiple parallel hits hypothesis. Hepatology 52, 1836–1846.
Tiraby, C., Tavernier, G., Lefort, C., Larrouy, D., Bouillaud, F., Ricquier, D., and Langin, D.
(2003). Acquirement of Brown Fat Cell Features by Human White Adipocytes. J. Biol. Chem.
278, 33370–33376.
375
Tokunaga, K., Nakamura, Y., Sakata, K., Fujimori, K., Ohkubo, M., Sawada, K., and Sakiyama,
S. (1987). Enhanced expression of a glyceraldehyde-3-phosphate dehydrogenase gene in human
lung cancers. Cancer Res. 47, 5616–5619.
Trayhurn, P. (2005). Endocrine and signalling role of adipose tissue: new perspectives on fat.
Acta Physiologica Scandinavica 184, 285–293.
Trayhurn, P., and Beattie, J.H. (2001). Physiological role of adipose tissue: white adipose tissue
as an endocrine and secretory organ. Proceedings of the Nutrition Society 60, 329–339.
Trayhurn, P., and Wood, I.S. (2004). Adipokines: Inflammation and the Pleiotropic Role of White
Adipose Tissue. British Journal of Nutrition 92, 347–355.
Tsukiyama-Kohara, K., Poulin, F., Kohara, M., DeMaria, C.T., Cheng, A., Wu, Z., Gingras, A.C.,
Katsume, A., Elchebly, M., Spiegelman, B.M., et al. (2001). Adipose tissue reduction in mice
lacking the translational inhibitor 4E-BP1. Nat. Med. 7, 1128–1132.
Unami, A., Shinohara, Y., Kajimoto, K., and Baba, Y. (2004). Comparison of gene expression
profiles between white and brown adipose tissues of rat by microarray analysis. Biochemical
Pharmacology 67, 555–564.
Unger, R.H. (2003). Minireview: Weapons of Lean Body Mass Destruction: The Role of Ectopic
Lipids in the Metabolic Syndrome. Endocrinology 144, 5159–5165.
Unger, R.H., and Orci, L. (2001). Diseases of Liporegulation: New Perspective on Obesity and
Related Disorders. FASEB J 15, 312–321.
Ushijima, T. (2007). Review: Epigenetic Field for Cancerization. Journal of Biochemistry and
Molecular Biology 40, 142–150.
Usman, H., and Mathew, M.K. (2010a). Potassium channel regulator KCNRG regulates surface
expression of Shaker-type potassium channels. Biochemical and Biophysical Research
Communications 391, 1301–1305.
Usman, H., and Mathew, M.K. (2010b). Potassium channel regulator KCNRG regulates surface
expression of Shaker-type potassium channels. Biochemical and Biophysical Research
Communications 391, 1301–1305.
Vandesompele, J., De Preter, K., Pattyn, F., Poppe, B., Van Roy, N., De Paepe, A., and
Speleman, F. (2002). Accurate normalization of real-time quantitative RT-PCR data by geometric
averaging of multiple internal control genes. Genome Biology 3, research0034.1–
research0034.11.
Vaquero, A., Scher, M.B., Lee, D.H., Sutton, A., Cheng, H.-L., Alt, F.W., Serrano, L.,
Sternglanz, R., and Reinberg, D. (2006). SirT2 is a histone deacetylase with preference for
histone H4 Lys 16 during mitosis. Genes & Development 20, 1256 –1261.
376
Vernon, G., Baranova, A., and Younossi, Z.M. (2011). Systematic review: the epidemiology and
natural history of non-alcoholic fatty liver disease and non-alcoholic steatohepatitis in adults.
Alimentary Pharmacology & Therapeutics 34, 274–285.
Virtanen, K.A., Lidell, M.E., Orava, J., Heglind, M., Westergren, R., Niemi, T., Taittonen, M.,
Laine, J., Savisto, N.-J., Enerbäck, S., et al. (2009). Functional Brown Adipose Tissue in Healthy
Adults. New England Journal of Medicine 360, 1518–1525.
Wan, J., Jiang, L., Lü, Q., Ke, L., Li, X., and Tong, N. (2010). Activation of PPARdelta upregulates fatty acid oxidation and energy uncoupling genes of mitochondria and reduces
palmitate-induced apoptosis in pancreatic beta-cells. Biochem. Biophys. Res. Commun. 391,
1567–1572.
Wang, F., and Tong, Q. (2009). SIRT2 Suppresses Adipocyte Differentiation by Deacetylating
FOXO1 and Enhancing FOXO1’s Repressive Interaction with PPARγ. Molecular Biology of the
Cell 20, 801 –808.
Wang, F., Nguyen, M., Qin, F.X.-F., and Tong, Q. (2007). SIRT2 deacetylates FOXO3a in
response to oxidative stress and caloric restriction. Aging Cell 6, 505–514.
Wang, Y.-X., Lee, C.-H., Tiep, S., Yu, R.T., Ham, J., Kang, H., and Evans, R.M. (2003).
Peroxisome-Proliferator-Activated Receptor δ Activates Fat Metabolism to Prevent Obesity. Cell
113, 159–170.
Weigt, J., and Malfertheiner, P. (2009). Influence of Helicobacter pylori on gastric regulation of
food intake. Current Opinion in Clinical Nutrition & Metabolic Care 12.
Weisberg, S.P., McCann, D., Desai, M., Rosenbaum, M., Leibel, R.L., and Jr., A.W.F. (2003).
Obesity is associated with macrophage accumulation in adipose tissue. J Clin Invest. 112, 1796–
1808.
Wieckowska, A., Zein, N.N., Yerian, L.M., Lopez, A.R., McCullough, A.J., and Feldstein, A.E.
(2006). In vivo assessment of liver cell apoptosis as a novel biomarker of disease severity in
nonalcoholic fatty liver disease. Hepatology 44, 27–33.
Wittwer, C., Hahn, M., and Kaul, K. (2004). Rapid Cycle Real-time PCR: Methods and
Applications : Quantification (Springer).
Woods, S.C. (2004). Gastrointestinal Satiety Signals I. An overview of gastrointestinal signals
that influence food intake. American Journal of Physiology - Gastrointestinal and Liver
Physiology 286, G7 –G13.
Woods, S.C., Seeley, R.J., Porte, D., and Schwartz, M.W. (1998). Signals That Regulate Food
Intake and Energy Homeostasis. Science 280, 1378 –1383.
Wren, A.M., Seal, L.J., Cohen, M.A., Brynes, A.E., Frost, G.S., Murphy, K.G., Dhillo, W.S.,
Ghatei, M.A., and Bloom, S.R. (2001). Ghrelin Enhances Appetite and Increases Food Intake in
Humans. JCEM 86, 5992–5992.
377
Wu, Z., Puigserver, P., Andersson, U., Zhang, C., Adelmant, G., Mootha, V., Troy, A., Cinti, S.,
Lowell, B., Scarpulla, R.C., et al. (1999). Mechanisms Controlling Mitochondrial Biogenesis and
Respiration through the Thermogenic Coactivator PGC-1. Cell 98, 115–124.
Xu, J., Wang, P., Li, Y., Li, G., Kaczmarek, L.K., Wu, Y., Koni, P.A., Flavell, R.A., and Desir,
G.V. (2004). The voltage-gated potassium channel Kv1.3 regulates peripheral insulin sensitivity.
Proc Natl Acad Sci U S A 101, 3112–3117.
Xu, L., Zhang, F., Zhang, D., Chen, X., Lu, M., Lin, R., Wen, H., Jin, L., and Wang, X. (2009).
OPRM1 gene is associated with BMI in Uyghur population. Obesity (Silver Spring) 17, 121–125.
Yamada, T., Oka, Y., and Katagiri, H. (2008). Inter-organ metabolic communication involved in
energy homeostasis: Potential therapeutic targets for obesity and metabolic syndrome.
Pharmacology & Therapeutics 117, 188–198.
Yang, H., Yang, T., Baur, J.A., Perez, E., Matsui, T., Carmona, J.J., Lamming, D.W., SouzaPinto, N.C., Bohr, V.A., Rosenzweig, A., et al. (2007a). Nutrient-Sensitive Mitochondrial NAD+
Levels Dictate Cell Survival. Cell 130, 1095–1107.
Yang, H., Yang, T., Baur, J.A., Perez, E., Matsui, T., Carmona, J.J., Lamming, D.W., SouzaPinto, N.C., Bohr, V.A., Rosenzweig, A., et al. (2007b). Nutrient-Sensitive Mitochondrial NAD+
Levels Dictate Cell Survival. Cell 130, 1095–1107.
Ye, J., Coulouris, G., Zaretskaya, I., Cutcutache, I., Rozen, S., and Madden, T.L. (2012). PrimerBLAST: a tool to design target-specific primers for polymerase chain reaction. BMC
Bioinformatics 13, 134.
Au-Yong, I.T.H., Thorn, N., Ganatra, R., Perkins, A.C., and Symonds, M.E. (2009). Brown
Adipose Tissue and Seasonal Variation in Humans. Diabetes 58, 2583 –2587.
Yoon, J.C., Puigserver, P., Chen, G., Donovan, J., Wu, Z., Rhee, J., Adelmant, G., Stafford, J.,
Kahn, C.R., Granner, D.K., et al. (2001). Control of hepatic gluconeogenesis through the
transcriptional coactivator PGC-1. Nature 413, 131–138.
Younossi, Z.M. (2008). Review article: current management of non-alcoholic fatty liver disease
and non-alcoholic steatohepatitis. Aliment. Pharmacol. Ther. 28, 2–12.
Younossi, Z., Jarrar, M., Nugent, C., Randhawa, M., Afendy, M., Stepanova, M., Rafiq, N.,
Goodman, Z., Chandhoke, V., and Baranova, A. (2008). A Novel Diagnostic Biomarker Panel for
Obesity-related Nonalcoholic Steatohepatitis (NASH). Obesity Surgery 18, 1430–1437.
Zainuddin, A., Chua, K.H., Abdul Rahim, N., and Makpol, S. (2010). Effect of experimental
treatment on GAPDH mRNA expression as a housekeeping gene in human diploid fibroblasts.
BMC Mol. Biol. 11, 59.
Zernecke, A., Shagdarsuren, E., and Weber, C. (2008). Chemokines in Atherosclerosis.
Arteriosclerosis, Thrombosis, and Vascular Biology 28, 1897 –1908.
378
Zhang, J.V., Ren, P.-G., Avsian-Kretchmer, O., Luo, C.-W., Rauch, R., Klein, C., and Hsueh,
A.J.W. (2005a). Obestatin, a Peptide Encoded by the Ghrelin Gene, Opposes Ghrelin’s Effects on
Food Intake. Science 310, 996–999.
Zhang, J.V., Ren, P.-G., Avsian-Kretchmer, O., Luo, C.-W., Rauch, R., Klein, C., and Hsueh,
A.J.W. (2005b). Obestatin, a Peptide Encoded by the Ghrelin Gene, Opposes Ghrelin’s Effects on
Food Intake. Science 310, 996 –999.
Zhang, X., Ding, L., and Sandford, A.J. (2005c). Selection of reference genes for gene expression
studies in human neutrophils by real-time PCR. BMC Mol Biol 6, 4.
Zhang, Y., Proenca, R., Maffei, M., Barone, M., Leopold, L., and Friedman, J.M. (1994).
Positional cloning of the mouse obese gene and its human homologue. , Published Online: 01
December 1994; | Doi:10.1038/372425a0 372, 425–432.
Zhao, W., Fong, O., Muise, E.S., Thompson, J.R., Weingarth, D., Qian, S., and Fong, T.M.
(2010). Genome-wide expression profiling revealed peripheral effects of cannabinoid receptor 1
inverse agonists in improving insulin sensitivity and metabolic parameters. Mol. Pharmacol. 78,
350–359.
Zhou, Z., Yon Toh, S., Chen, Z., Guo, K., Peng Ng, C., Ponniah, S., Lin, S.-C., Hong, W., and Li,
P. (2003). Cidea-deficient mice have lean phenotype and are resistant to obesity. Nat Genet 35,
49–56.
Zimmermann, H.W., Seidler, S., Gassler, N., Nattermann, J., Luedde, T., Trautwein, C., and
Tacke, F. (2011). Interleukin-8 Is Activated in Patients with Chronic Liver Diseases and
Associated with Hepatic Macrophage Accumulation in Human Liver Fibrosis. PLoS ONE 6,
e21381.
Zuberi, A.R., Townsend, L., Patterson, L., Zheng, H., and Berthoud, H.-R. (2008). Increased
adiposity on normal diet, but decreased susceptibility to diet-induced obesity in mu-opioid
receptor-deficient mice. Eur. J. Pharmacol. 585, 14–23.
379
CURRICULUM VITAE
Rohini Mehta was born and brought up in a small town in India. She pursued her
undergraduate degree in biological sciences with majors in biochemistry, botany and
microbiology. After graduating with honors in 2001, she chose to pursue Masters degree
in Microbiology at Maharaja Sayajirao University in Vadodara, India and graduated in
2005. She spent the following 2 years working as a research assistant in CFTRI, India. In
2009, eager to gain international exposure in research, she joined the Ph.D. program in
School of Systems Biology at George Mason University. She was awarded with the
prestigious Presidential Scholarship for the entire duration of her PhD program. An
interest in human health and keenness for mentoring led her to become a part of the
project initiated as a collaboration between School of Systems Biology, George Mason
University and Beatty Research Centre, Inova Fairfax Hospital. As a PhD student she
been an author on atleast 6 publications and has presented over 26 posters and oral
presentations at national and international conferences. She is also the recipient of the
Best Doctorate of Philosophy Dissertation for Academic year 2012-2013 awarded by the
School of Systems Biology. When not working, she enjoys spending time outdoors
birdwatching and camping.
380