Physiological roles of CIDEs CIDEA deficient mice exhibit lean phenotype and are obesity resistant 2

Chapter 1

1





Chapter 1
Introduction to c
ell death-inducing DFF45-
like e
ffectors (CIDEs)
Chapter 1

2
1.1 Apoptosis and DNA fragmentation
Cell death is an invariable phenomenon of animal development, and it often
continues into adulthood (Raff, 1998). Based on the characteristic ways cells look when
they die in different circumstances, it was proposed in 1972 that normal cell death, as
well as some pathological death, are, in fact, cellular suicide (Kerr et al., 1972). That is,
the cells activate an intracellular death programme and kill themselves in a controlled
way - a process now known as programmed cell death, or apoptosis. Apoptotic cells
shrink and are rapidly phagocytosed by neighbouring cells and macrophage before there
is any leakage of their contents. This process is so efficient that it is difficult to find an
apoptotic cell that is not already been phagocytosed in typical tissue sections (Raff,
1998).
The mechanisms of how apoptosis is initiated and executed remained unclear
until the molecular identification of the key components of this intracellular suicide
program. The prototypical apoptotic process can be divided into three phases: an
induction phase, the nature of which depends on the specific death-inducing signals, an

effector phase, during which the central executioner is activated and the cells become
committed to die, and a degradation phase, during which cells acquire the biochemical
and morphological features of end-stage apoptosis (Green and Kroemer, 1998). The
pattern of this cell suicide programme is evolutionarily conserved from worms to
humans (Ellis et al., 1991; Nagata, 1997; Steller, 1995). There are three major
components in the program: the Bcl-2 family proteins, the caspases, which belong to a
family of cysteine proteases that cleave after aspartic acid residues; and the Apaf-
1/CED-4 protein that relays the signals integrated by Bcl-2 family proteins to caspases
(Adams and Cory, 1998). Death signals originating from the death receptors (such as
Chapter 1

3
TNF and Fas), or the mitochondria trigger the activation of caspase 8 or caspase 9,
respectively. These activated initiator caspases in turn activate the downstream
executioner caspases 3, 6 and 7. The cross-talk between the two pathways is mediated by
BID, which upon cleavage by caspase 8 activates the mitochondrial pathway of cell
death (Adams and Cory, 1998; Colussi and Kumar, 1999; Slee et al., 1999). A schematic
stepwise representation for caspase activation during apoptosis is illustrated in Figure 1.

Apoptotic stimuli
De xmethaso ne
Death receptors
Mitochondria
Pro caspase8 or caspase 10
Active caspase 8 or caspase 10
cytochrome c
Bcl-2
Apoptosome
( Apaf-1, procaspase 9, cytochrome c)
Avtive caspase 9

Pro-effector caspases
Active effector caspase 3,6,7
Vital substrates like DFF
Apoptosis
BAX
BID

Figure 1 Apoptotic pathways and caspase activation.
Death receptors can activate the initiator caspase-8. Activation of caspase-9 by
several apoptotic stimuli requires release of cytochrome c into the cytosol from the
mitochondria.
Chapter 1

4
Upon apoptotic stimulation, initiator caspases such as caspase -8 and -9 are
recruited to their respective adaptors, such as FADD and Apaf-1. Through homophilic
interactions that result in oligomerization of the initiator caspases, activation is achieved
in presumably an autocatalytic fashion. Subsequently, activated initiator caspases
proteolytically activate downstream effector caspases such as caspase-3 and -7, which in
turn carry out the final destruction of the apoptotic cell (Nagata, 1997).
The downstream events of apoptosis are characterized by morphological changes
that include mitochondrial damage, nuclear membrane breakdown, DNA fragmentation,
chromatin condensation, and formation of apoptotic bodies (Takahashi, 1999). DNA
fragmentation is often considered a hallmark of apoptosis,

a result of the activation of
endonucleases that cleave DNA

between nucleosomes (Wyllie et al., 1980).
1.2 DNA fragmentation factor (DFF)

The factor that is responsible for DNA fragmentation and nuclear condensation
in apoptosis has been purified (designated DNA fragmentation factor, or DFF) using an
experimental system in which DNA fragmentation was triggered in vitro by activated
caspase-3 (Liu et al., 1996; Enari et al., 1998; Liu et al., 1997; Sakahira et al., 1998).
DFF is a heterodimeric protein, composed of the 40 kDa caspase activated subunit
(DFF40/CAD, CPAN) and its 45 kDa inhibitor subunit (DFF45/ICAD) (Enari et al.,
1998; Liu et al., 1997). DFF had been originally purified from a cytoplasmic fraction
(Liu et al., 1997), and it was postulated that DFF40 undergoes translocation to the
nucleus upon caspase-3 cleavage of DFF45/ICAD. Cleavage of DFF45/ICAD by
caspase 3 at two different sites releases DFF40/CAD from the complex and triggers
DNA fragmentation and chromatin condensation (Enari et al., 1998; Liu et al., 1998).
DFF45/ICAD mutants carrying point mutations at either or both caspase-recognition
Chapter 1

5
sites are competitive inhibitors, and cannot be removed upon caspase-3 cleavage to
activate DFF40/CAD (Sakahira et al., 1999). Purified DFF40/CAD released from the
complex with DFF45/ICAD forms homo-oligomers that are the enzymatically active
forms of the nuclease (Liu et al., 1997).
DFF40/CAD consists of two domains with distinct functions. Its C-terminal part
exhibits deoxyribonuclease activity, whereas the N-terminal CIDE-N domain has a
regulatory function (Inohara et al., 1998). Similarly, DFF45/ICAD contains an N-
terminal CIDE-N domain that is sufficient to inhibit DNA fragmentation (Inohara et al.,
1998). The CIDE-N domain represents

a 75 amino acid fold consisting of a twisted five-
stranded β sheet with two α -helices arranged in an α/β roll (Lugovskoy et al., 1999).
The structure of the CIDE-N domain is illustrated in Figure 2.

Figure 2 Solution structure of CIDE-N domain (Lugovskoy et al., 1999).

DFF45 plays a dual role as both an inhibitor and a chaperon of DFF40. The
expression of DFF40 in various systems in the absence of co-expressed DFF45 results in
Chapter 1

6
inactive aggregates. This suggests that DFF45 is required for the proper folding of the
active DFF40 during its synthesis (Enari et al., 1998; Liu et al., 1997; Liu et al., 1998).
The CIDE-N domain of DFF45 is apparently required for its chaperon function, by
associating with the CIDE-N domain of DFF40 (Freida L., May 1997; Lugovskoy et al.,
1999). A combination of hydrophilic and hydrophobic interactions contributes to the
formation of this complex and determines the specificity of the interactions (Zhou et al.,
2001).
DFF45(nascently translated)
DFF40(nascently translated)
Aggregated & misfo lded DFF40
DFF40/45 complex(folding intermediate )
DFF40/45 complex(monomeric )
DETD DAVD (caspase-3 cleavage sites)
Cytoplasm
Nucleus
Caspase-3
DFF45 fragments
Active DFF40 (oligomeric) active nuclease
DNA fragmentation

Figure 3 Mechanisms of activation of DFF.
DFF45 can be cleaved by caspase-3 at DETD and DAVD site. Both DFF45 and
DFF40 are synthesized in the cytoplasm and functions as mutual chaperons. The
inactive DFF40/45 complex is transfered to nucleus. Upon an apoptotic stimulus,
DFF45 is cut by activated caspase-3, which releases DFF40 and form active

homodimers. The activated DFF40 serves as a nuclease to cause DNA fragmentation.
The molecular structure and mechanism of regulation of DFF are summarized in
Figure 3. Human DFF40 is a basic protein with a pI of 9.3 and is composed of 345
Chapter 1

7
amino acids, while DFF45 is an acidic protein with a pI of 4.5 and composed of 331
amino acids. As illustrated in Figure 3, DFF40 and DFF45 are synthesized in the
cytoplasm and serve as each other’s folding chaperone. The catalytically inactive
complex DFF40/DFF45 is then targeted to the nucleus. When an apoptotic stimulus
triggers the activation of caspase-3, it cleaves DFF45, thereby releasing DFF40 from the
complex. Released DFF40 forms catalytically active homo-oligomers that degrade
chromosomal DNA (Zhou et al., 2001).
1.3 Physiological roles of DFF
The physiological significance of DFF in triggering internucleosomal cleavages
during apoptosis has been unequivocally demonstrated. DFF40 and DFF45 mRNAs and
proteins are expressed in most tissues and cell lines. Cell lines expressing high levels of
DFF40 and DFF45 quickly undergo DNA fragmentation upon an apoptotic stimulus
(Mukae et al., 1998). Thymocytes and splenocytes from mice that lack functional DFF45
gene exhibit neither DNA laddering nor chromatin condensation when exposed to
apoptotic stimuli (Wu et al., 1999). Although such mice appeared normal, thymocytes
from these mice are more resistant to apoptosis than that from wild-type animals (Wu et
al., 1999).
The degradation of genomic DNA into nucleosomal units is one of the best
characterized biochemical hallmarks of apoptosis, and has served as the biochemical
basis for commonly used techniques to detect apoptotic cells (e.g., TUNEL (Terminal
deoxytransferase-mediated deoxy uridine nick end-labelling) assays). One of the
physiological roles of apoptosis is to remove harmful cells (i.e. cancer or virally
infected) from an organism. It was hypothesized that mechanisms of DNA breakdown
during apoptosis are developed to prevent transfer of potentially “incorrect DNA” (e.g.,

Chapter 1

8
activated oncogenes or viral genes) to another cell or to reduce the possibility of
autoimmune responses (Widlak, 2000). It has been shown that cells can die as a result of
apoptotic stimuli without internucleosomal DNA cleavage (Cohen et al., 1994;
Oberhammer et al., 1993). However, although apoptotic cell death can be disconnected
from DNA breakdown, this event is beneficial for the efficient removal of potentially
toxic cell debris from the organism.
1.4 CIDE family
To identify potential DFF45-related genes, Inohara and colleagues searched the
expressed sequence tag (EST) database of Genbank for clones with homology to DFF45
and identified a novel family of cell death inducing-DFF45-like effector (CIDE) proteins
(Inohara et al., 1998). These CIDEs share an N-terminal domain that is homologous to
the CIDE-Ns of DFF40/CAD and DFF45/ICAD (Inohara et al., 1998). CIDE proteins
are highly homologous between human and mouse. The mammalian family members
Cidea, Cideb, and FSP27 share homology with each other both in the CIDE-N and
CIDE-C domains as shown in Figure 4A. The CIDE-C domain show apoptosis inducing
function (Inohara et al., 1998).
1
335
32 94
1
239
56 116
135 203
1
219
48
108

121 188
1 219
49
108
125
193
CIDE-N
CIDE-C
DFF45
Cidea
Cideb
FSP27

Figure 4A Schematic structure of mouse Cidea, Cideb, FSP27 and DFF45.

Chapter 1

9
The amino acid sequence alignments of CIDEs are shown below in Figure 4B.

Figure 4B Sequence alignments of CIDEs family members.
The accession numbers for hDFF45, hCIDE-A, mCIDE-A, FSP27, hCIDE-B and
mCIDE-B are NP_004392, NP_938031, NP_034174, NP_848460, NP_055245, and
NP_034024 respectively. Red color characters indicate conserved in all molecules and
blue color characters indicate conserved in more than two molecules. Alignment was
performed using Clustalw program. Conserved residues are listed in the line labeled
as “consensus”. hDFF45, hCIDE-A and hCIDE-B are from humans, mCIDE-A,
mCIDE-B and FSP27 are from mice.
Chapter 1

10
It has been shown that ectopic overexpression of many CIDEs causes apoptosis
(Inohara et al., 1998). The C-terminal domain of Cidea is necessary and sufficient for
cell killing, whereas its N-terminus is required for DFF45 to inhibit Cidea induced
apoptosis (Inohara et al., 1998). Thus it has been postulated that N-terminal domains
play a regulatory role in mediating CIDE-induced apoptosis by associating with other
CIDE-N functional domain containing proteins (Inohara et al., 1998).
Cideb, on the other hand, is localized to mitochondria when overexpressed
ectopically and can form homo- or heterodimers with other CIDE family members
(Chen et al., 2000). Furthermore, the C-terminal region of Cideb, which shares
homology with Cidea and FSP27, is responsible for Cideb induced cell death,
mitochondrial localization and dimerization (Chen et al., 2000).
Unlike the ubiquitous DFF45 and DFF40, both Cidea and Cideb are expressed
in a more restricted manner and show pronounced tissue specificity. Expression of
human Cidea was detected in heart, skeletal muscle, brain, lymph node, thymus,
appendix, bone marrow, placenta, kidney and lung (Inohara et al., 1998). Human
Cideb was detected in adult and fetal liver as well as in spleen, peripheral blood
lymphocyte and bone marrow (Inohara et al., 1998). Cidea, but not Cideb mRNA, was
expressed in 293T embryonic kidney, MCF-7 breast carcinoma and SHEP
neuroblastoma cells (Inohara et al., 1998).
Although Cidea and Cideb activate apoptosis when ectopically expressed, and
appear to function as positive effectors of the apoptotic pathway, it is still not clear
which cell death-signaling cascade Cidea, Cideb or FSP27, might be involved.
Overexpression of Cidea or Cideb alone induced caspase-independent cell death that
Chapter 1

11
shows not only nuclear condensation and DNA fragmentation, but also membrane
blebbing (Toh, S.Y. & Li, P., personal communication, unpublished data). Unlike
DFF45/DFF40, no translocation of Cidea or Cideb from mitochondria to the nucleus

was observed (Toh, SY & Li, P., personal communication, unpublished data). All these
data suggest that the mechanisms of function of CIDE members are different from that
of DFF45. The specific tissue expression pattern and the mitochondrial subcellular
localization also suggest that these proteins could have a completely different in vivo
physiological role from DFF.

In order to get an idea of the in vivo function of CIDEs members, we investigated
the tissue distribution pattern of them in mice and surprisingly discovered that Cidea is
specifically expressed in brown adipose tissue, a fat storage tissue functions in heat
generation. In the following sections, the biochemistry and physiology of obesity and
adipose tissue function will be discussed. This is necessary because Cidea turned out to
have significant roles on the regulation of body fat composition and lipid metabolism.
1.5 Background information on obesity and adipose tissue
1.5.1 Obesity
Excess total body fat, specifically adipose tissue, manifests obesity. It is the
most prevalent disorder, affecting one-third of the middle-aged population in the
Western world (Kuczmarski et al., 1994). It is also responsible for the increased
incidence of diseases such as coronary artery disease, hypertension, stroke and diabetes
and clinical conditions such as hyperlipidermia (reviewed by Lowell and Spiegelman,
2000; Pi-Sunyer, 2003; Spiegelman and Flier, 2001). Obesity develops only if energy
Chapter 1

12
intake, in the form of feeding, chronically exceeds total body energy expenditure.
Energy expenditure results from physical activities, basal metabolism, and adaptive
thermogenesis (Lowell and Spiegelman, 2000). Adaptive thermogenesis is a metabolic
response to enviromental changes, such as exposure to cold and alterations in diet
(Lowell and Spiegelman, 2000). The components of energy expenditure that can be
readily altered, i.e., physical activity and adaptive thermogenesis, are of particular
interest in the control of obesity (Spiegelman and Flier, 2001).

1.5.2 Body weight regulation
From an evolutionary standpoint, it is believed that there must be a homeostatic
system balancing food intake and energy expenditure (Friedman, 2000). In preparation
for food shortage, this system appears to guard against changes in body weight by
storing chemical energy in a compact form, fat. This system should theoretically
consist of at least three components: an afferent signal reporting the state of the body’s
fat stores, a brain receiver analogous to a thermostat, where fat-store reports are
compared with a set-point, and a downstream effector system using behavior and
metabolic processes to correct deviations from the set-point.
Single-gene obesity syndromes had been known in rodents since the early
1950s, when the obese (ob) mouse was described. In the 1970s, a second model, the
diabetes (db) mouse, was identified (Coleman, 1978; Hummel et al., 1966). The two
animals share a phenotype characterized by over eating, low energy expenditure, early
onset obesity, insulin resistance, and susceptibility to diabetes mellitus.
The gene mutated in ob mice was identified in 1994 as Lep (Zhang et al.,
1994), located on murine chromosome 6. The gene product, a 16 kD protein consisting
Chapter 1

13
of 167 amino acid, is called leptin (Greek for “thin”). The peptide is produced by
adipose tissue, circulates in the blood and reports nutritional information to key
regulatory centers in the hypothalamus. Increased body fat is associated with increased
levels of leptin, which then acts to reduce food intake and increase energy expenditure.
A decrease in body fat leads to a decreased level of leptin, which stimulates food
intake and reduces energy expenditure.
The gene mutated in db was cloned in 1995 (Chen et al., 1996), and is located
on murine chromosome 4. The gene is expressed in brain, lung, kidney, muscle,
pancreas, and adipose tissue. The gene product was identified as the leptin receptor
(lepr). It is a membrane-spanning receptor that responds to extracellular leptin by
initiating intracellular gene activation.

While efforts to identify the ob and db mutations were underway, three other
mutations capable of causing obesity in mice were being investigated. The mouse
mutants are called Tubby, Fat, and Yellow. The genes mutated are designated Tub
(Noben-Trauth et al., 1996), Cpe (carboxypeptidase E, an enzyme that participates in
the biosynthesis of hormones such as insulin and melanocortin and neuropeptides.)
(Naggert et al., 1995b), Agouti (Galbraith and Wolff, 1974), respectively. All mutants
share syndromes of increased body weight associated with increased food intake, and
decreased autonomic activation. As noted, all five genes have now been cloned and all
five have human homologs. All are well conserved, because the amino-acid sequences
of the expressed proteins have changed relatively little during mammalian evolution.
Of the five genes identified through their involvement in mouse obesity
syndromes, lep fits the description for a signal reporting the state of the body’s fat
stores, while lepr appears to be the signal’s brain receptor, as shown in Figure 5. In
Chapter 1

14
varying ways, the other genes participate in or reveal the existence of brain signaling
pathways, probably downstream from the leptin receptor, that create the brain’s
capacity for the regulation of body weight.

Figure 5 Leptin’s physiologic function as a signal reporting the body’s status of
energy store in adipose tissue.
Leptin and its receptor are essential components in thecomplex genetic wiring
diagram underlying energy homeostasis and body weight. The long, signaling-
competent isoform of the leptin receptor (LR) shows high expression peaks in the
feeding centers of the hypothalamus, consistent with leptin being the afferent signal
informing the central nervous system of the body fat status. To date, six splice variants
of the LR have been identified. The long isoform or Ob-Rb (further referred to as
LRlo) consists of 1162 amino acids and is the only LR isoform with clearly
demonstrated signaling capability. Four short isoforms (Ob-Ra, Ob-Rc, Ob-Rd, and

Ob-Rf ; Ob-Ra is further referred to as LRsh) with shortened intracellular tails have
been identified. A secreted isoform can be generated either by alternative splicing (Ob-
Re) or by ectodomain shedding, and may be involved in modulating leptin activity.
The LR was first cloned from a mouse choroid plexus cDNA library using an
expression cloning strategy by Tartaglia and co-workers. Based on sequence
Chapter 1

15
homology, this receptor belongs to the class I cytokine receptor family, which typically
contains a so-called CRH (cytokine receptor homology) domain.This structure consists
of two barrel-like domains, each approximately 100 amino acids in length, which
resemble the fibronectin type III (FN III) fold. Two conserved disulfide bridges are
found in the N-terminal domain, while a WSXWS motif is characteristic for the C-
terminal part. The LR shares highest sequence similarity with the granulocyte colony-
stimulating factor (G-CSF) receptor and the glycoprotein 130 (gp130) family
receptors, including gp130, the leukemia inhibitory factor (LIF) and oncostatin M
(OSM) receptors. Moreover, structural superposition shows that also leptin is
structurally most similar to G-CSF and cytokines of the gp 130 family, such as
interleukin-6 (IL-6).
Like all other class I cytokine receptors, the LR lacks any intrinsic kinase
activity, and uses cytoplasmic-associated kinases of the JAK family. Leptin binding
results in formation of a receptor complex leading to cross-phosphorylation and
activation of the JAKs. These activated JAKs then rapidly phosphorylate tyrosine
residues in the cytosolic domain of the receptor. Such phosphorylated residues provide
binding sites for signaling molecules including members of the STAT family.Leptin
signaling occur typically through the JAK/STAT pathway.
The genes implicated in the mouse syndromes are not the only ones that could
potentially affect body weight. Other hormornal regulators of body weight known
include insulin, neuropeptide Y (NPY), uncoupling proteins (UCP1, 2, 3,) melanocyte
concentrating hormone (MCH), and glucagon-like peptide (GLP), etc. (reviewed by

Robinson et al., 2000). It is likely there are other regulatory proteins yet to be
discovered. The idea of simply including or excluding each gene as the cause of a
Chapter 1

16
person’s obesity may be misleading because obesity can be caused by a variety of
genetic and environmental components. The mouse syndromes are a dramatic
demonstration of the complexity inherent in weight regulation. Therapies that target
energy intake or expenditure alone may initially produce weight loss, but the existence
of a homeostatic feedback loop would be predicted that acts to resist further weight
loss and limit efficacy. Hence, it seems unlikely that most obesity is caused by the
function or malfunction of a single gene.
Chapter 1

17
1.5.3 The liver plays a central role in lipid transport and metabolism
Fat absorbed from the diet and lipids synthesized by the liver and adipose tissue
must be transported between the various tissues and organs for utilization and storage.
Since lipids are insoluble in water, the problem arises of how to transport them in an
aqueous environment such as blood plasma. This is solved by associating nonpolar
lipids (triglycerides and cholesteryl esters) with amphipathic lipids (phospholipids and
cholesterol) and proteins to make water-miscible lipoproteins (Mayes, 1990).
Lipoproteins have been classified into five broad categories on the basis of
their functional and physical properties. They are chylomicrons derived from intestinal
absorption of triglyceride, which transport exogenous (externally supplied, in this case,
dietary) triglycerides and cholesterol from the intestines to the tissues; very low
density lipoproteins (VLDL), intermediate density lipoproteins (IDL) and low-density
lipoproteins (LDL), a group of related particles that transport endogenous (internally
produced) triglycerides and cholesterol from the liver to the tissues (the liver
synthesizes triglyerides from excess carbohydrates); and high-density lipoproteins

(HDL), which transport endogenous cholesterol from the tissues to the liver. The
protein components of lipoproteins are known as apolipoproteins or just apoproteins.
At least nine apoproteins are distributed in significant amounts in the different human
lipoproteins. They are ApoA-I, ApoA-II, ApoB-48, ApoB-100, ApoC-I, ApoC-II,
ApoC-III, Apo-D, Apo-E.
Plasma lipids are composed of four major groups − triglycerides,
phospholipids, cholesterol, and cholesteryl esters. In addition, a much smaller fraction
is in the form of unesterified long-chain fatty acids (LCFA) and free fatty acid (FFA),
which are known to be the most metabolically active of the plasma lipids.
Chapter 1

18
Lipids are transported in lipoprotein complexes. The lipid digestion products
absorbed by the intestinal mucosa are converted by these tissues to triglycerides and
then packaged into lipoprotein particles called chylomicrons. These in turn are released
into the bloodstream via the lymph system for delivery to the tissues. Similarly
triglycerides synthesized by the liver re packaged into very low density lipoproteins
(VLDL) and released directly into the blood. The triglycerides components of
chylomicrons and VLDL are hydrolyzed to FFA and glycerol in the capillaries of
adipose tissue and skeleton muscle by lipoprotein lipase (LPL). The resulting FFAs are
taken up by these tissues while the glycerol is transported to the liver or kidneys. There
it is converted to glycolitic intermediate dihydroxyacetone phosphate by the sequential
actions of glycerol kinase and glycerol-3-phosphate dehydrogenase. Mobilization of
triglycerides stored in adipose tissue involves their hydrolysis to glycerol and FFAs by
hormone sensitive lipase (HSL). The FFAs are released into the bloodstream, where
they bind to albumin, a soluble 66.5-kD monomeric protein that comprises about half
of the bloodstream protein.
Triglyceride is the predominant lipid in chylomicrons and VLDL, whereas
cholesterol and phospholipid are the predominant lipids in LDL abd HDL,
respectively. All plasma lipoproteins are interrelated components of one or more

metabolic cycles that together are responsible for the complex process of plasma lipid
transport. The liver plays a central role in lipid transport and metabolism (Mayes,
1990). The metabolic interaction between liver and other tissues is illustrated in Figure
6.
Chapter 1

19
Exogenous pathway
Endo genous pathwa y
In tes ti ne
Li ver
Extr a hepatic tis s ue
Dietary fat Bile acids
and cholesterol
Chylomicrons Remnants VLDL IDL
HDL
LDL
ApoE
C-II
B-48
ApoE
B-48
ApoE
C-II
B-100
ApoE
B-100
ApoA-I
A-II
Capillaries

Capillaries
LPL
LPL
FFA
Adipose tiss ue, mus cle
FFA
Adipose tiss ue, mus cle
Endogenous
cholesterol
Dietary
cholesterol
Re mn ant
receptor
ApoB-100
LDL
receptor
LDL
receptor
Plasma LCAT
Lec ithin
Choles terol
Acyl
Transferase

Figure 6 Liver plays a central role in lipid transport and metabolism
.
Triglyceride is transported from the intestines in chylomicrons and from the
liver in VLDL. FFA arises from lipolysis of triglyceride in adipose tissue or as a result
of the action of lipoprotein lipase during the uptake of plasma triglyceride into tissues.
1.5.4 Adipose tissue is the main storage site for triglyceride in the body

In adult mammals, the major bulk of adipose tissue is a loose association of lipid-
filled cells called adipocytes, which are held in a framework of collagen fibers. In
addition to adipocytes, adipose tissue contains stromal-vascular cells including
fibroblastic connective tissue cells, leukocytes, macrophages, and pre-adipocytes (not yet
filled with lipid), which contribute to the structural integrity. Adipose tissue is therefore
a specialized connective tissue that functions as the major storage site for fat in the form
of triglyceride.
Chapter 1

20
Adipose tissue
Plasma
Insulin
Glucose
Glucose-6-phosphate
Glycolysis
Acetyl-CoA
2 CO
2
Citric
Acid
Cycle
Acyl-CoA
Lipogenesis
β-Oxidation
Glycerol 3-phosphate
Ester ificatio n
TG
HSL
Lipolysis

FFA
(pool 1)
Glycerol
LPL
FFA
Glycerol
FFA
TG(Chylo micro ns, VLDL)
FFA
(pool 2)
ATP
CoA
Acyl-CoA
Synthetase

Figure 7 Key metabolic cycles in the adipose tissue.
TG, triglyceride; FFA, free fatty acids; VLDL, very low density lipoprotein; HSL,
hormone sensitive lipase; LPL, lipoprotein lipase.
The triglyceride stores in adipose tissue continuously undergo lipolysis and
reesterification (see Figure 7). These two processes are not the forward and reverse
phases of the same reaction. Rather, they are entirely different pathways involving
different reactants and enzymes. Many of the nutritional, metabolic and hormonal factors
that regulate the metabolism of adipose tissue act either on the process of esterification
or lipolysis. The resultant of the these two processes determines the amount of the free
fatty acid (FFA) in adipose tissue, which in turn is the source and determinant of the
level of FFA circulating in the plasma. Since the level of plasma FFA has profound
Chapter 1

21
effects on the metabolism of other tissues, particularly liver and muscle, factors

regulating the outflow of FFA in adipose tissue exert an influence on metabolism far
beyond the tissue itself (Voet, 1995).
1.5.5 Adipose tissue as an endocrine organ
Adipose tissue has emerged conceptually as an endocrine organ that is central
to the regulation of energy homeostasis. It can secrete proteins that exert a pleiotropic
effect on metabolism upon diet or environmental stimuli. The proteins are involved in
glucose and fat metabolism and hence can influence insulin resistance. They include
leptin, resistin, adiponectin, acylation-stimulating protein, tumour necrosis factor-alpha
and interleukin-6 (Simpson, 1996).
As mentioned previously leptin is the hormone that reports the status of energy
storage in adipose tissues which is encoded by the ob gene, and is an important
metabolic regulator. Initial work indicates that the leptin gene is expressed only in
white adipose tissue (WAT), but subsequent developments have shown it to be
expressed at lower levels in other forms of adipose tissue, including brown adipose
tissue (BAT). A variety of other tissues (e.g., bone, mammary gland, ovarian follicles,
placenta, stomach, muscle and certain fetal organs, such as the heart, bone and
mucosa) have now been shown to have leptin gene expression. The placenta is also a
site of leptin synthesis in humans, rodents and ruminants (Friedman and Halaas, 1998).
The structure and function difference of BAT and WAT will be stated in section 1.5.7.



Chapter 1

22
1.5.6 Hormones regulate fat mobilization in adipose tissues
Adipocytes express a variaety of receptors that are responsible for hormonal
regulation of lipid metabolism under different physiological requirements. They
include insulin-like growth factor-I (IGF-I), growth hormone (GH), leptin, glucagon,
α-adrenergic and β-adrenergic receptors (Simpson, 1996). Among these receptors, the

β
3
-adrenergic receptor is the one that can be activated by catecholamines leading to
thermogenesis in BAT (Lowell and Flier, 1997).
A principle action of insulin in adipose tissue is to inhibit the activity of the
hormone-sensitive lipase (HSL), reducing the release of both FFA and glycerol.
Simultaneously, insulin can enhance the uptake of glucose into adipose tissues,
promoting lipogenesis. Adipose tissue is thus a major site of insulin action in vivo
(Voet, 1995).
Other hormones accelerate the release of FFA from adipose tissue and raise
plasma FFA concentration by increasing the lipolysis of triglyceride stores (Voet,
1995). These hormones include epinephrine, norepinephrine, glucagon,
adrenocorticotropic hormone, thyroid stimulating hormone, growth hormone and
vasopressin. The hormones that act rapidly in promoting lipolysis, i.e., catecholamines,
do so by stimulating the activity of adenylate cyclase, the enzyme that converts ATP to
cAMP (Voet, 1995). CAMP, by stimulating cAMP-dependent protein kinase (PKA),
phosphorylates HSL. The activated HSL stimulates lipolysis in adipose tissue, raising
blood FFA levels and ultimately activating the β-oxidation pathway in other tissues
such as liver, muscle and BAT (Voet, 1995). In liver, this process leads to the
production of ketone bodies that are secreted into bloodstream for use as an alternative
fuel to glucose by peripheral tissues. PKA, acting in concert with AMP-dependent
Chapter 1

23
protein kinase (AMPK), also causes the inactivation of acetyl-CoA carboxylase
(ACC), one of the rate determining enzymes of fatty acid synthesis (Voet, 1995). Thus
cAMP-dependent phosphorylation simultaneously stimulates FFA oxidation and
inhibits fatty acid synthesis. The AMPK is the downstream component of a protein
kinase cascade that is activated by a rise in the AMP: ATP ratio. AMPK is switched on
by cellular stresses that either interfere with ATP production (e.g. hypoxia, glucose

deprivation, or ischemia), or by stresses that increase ATP consumption (e.g. muscle
contraction). Hormones that act via G
q
-coupled receptors, and by leptin and
adiponectin, via mechanisms that remain unclear, also activate AMPK. It was reported
that leptin stimulates FFA oxidation by activating AMPK (Minokoshi et al., 2002). A
schema on the leptin and adrenergic signaling pathways in brown adipocytes is shown
in Figure 8.









Chapter 1

24


Figure 8 the mechanism of hormonally induced uncoupling of oxidative
phosphorylation in brown fat mitochondria.

The sympathetic nervous system (SNS), through liberation of norepinephrine in
adipose tissue, plays a central role in the mobilization of FFA by exerting tonic
influences, even in the absence of augmented nervous activity (Lowell and Flier, 1997).
Fatty acyl-CoA
CPT1

Acetyl-CoA
Citric
Acid
Cycle
β-Oxidation
cAMP
R
2
C
2
(inactive PKA)
HSL
Triglyceride
FFA
Brown
adipocyte
Norepinephrine
β-adrenegic
receptor
Leptin
Cold is sensed
by the brain
Sympathetic nerves are activated
A
d
e
n
y
l
c

y
c
l
a
s
e
Activation
ATP
2C
(active PKA)
HSL-P
(active)
R
2
(cAMP)
4
Ob-Rb
α-adrenergic
receptor
AMPK
AMPK-P
AMPKK
ACC
ACC-P
Acetyl-CoA
Malonyl-CoA
Fatty acyl-CoA
Respiratory chain
H
+

H
+
H
+
H
+
ADP
ATP
H
+
H
+
ATP Synthase
UCP1
H
+
Heat
Chapter 1

25
It was established that β-adrenergic stimulation of the tissue leads to rapid hydrolysis of
the multi triglyceride droplets within the brown adipocytes and that the abundant
mitochondria are able to oxidize the resulting fatty acids at an enormous rate to account
for the heat production.
1.5.7 Brown adipose tissue promotes thermogenesis
Adipose tissue is found in mammals in two different forms: white adipose tissue
(WAT) and brown adipose tissue (BAT). As shown in Figure 9, they have dramatically
different morphological appearances. WAT consists of unilocular cells containing a
single large lipid droplet that pushes the cell nucleus against the plasma membrane,
giving the cell a signet-ring shape (Fig. 9 lower picture). Mitochondria are found

predominantly in the thicker portion of the cytoplasmic rim near the nucleus. The large
lipid droplets do not appear to contain any intracellular organelles. Multilocular cells,
typically seen in BAT, contain many smaller lipid droplets (Fig. 9 upper picture). The
brown color of this tissue is derived from the rich vascularization, and densely packed
mitochondria that have extensively developed cristae. These mitochondria vary in size
and may be round, oval, or filamentous in shape. WAT is not as richly vascularized as
BAT, but each adipocyte in WAT is in contact with at least one capillary. This blood
supply provides sufficient support for active metabolism, which occurs in the thin rim of
cytoplasm surrounding the lipid droplet. Blood flow to adipose tissue varies depending
upon body weight and nutritional state, with blood flow increasing during fasting,
presumably to transport the products of lipolysis (FFA and glycerol) to the other tissues.