IMIDAZOLINE RECEPTORS IN INSULIN SIGNALING AND
METABOLIC REGULATION
by
Zheng Sun
Submitted in partial fulfillment of the requirements for the Degree of Doctor
of Philosophy
Thesis Advisor: Paul Ernsberger, Ph.D.
Department of Nutrition
CASE WESTERN RESERVE UNIVERSITY
January 2007
CASE WESTERN RESERVE UNIVERSITY
SCHOOL OF GRADUATE STUDIES
We hereby approve the dissertation of
______Zheng Sun candidate for the Ph.D. degree *.
Henri Brunengraber (signed)______(chair of the committee)
Bryan Roth ______
Laura Nagy ______
Jonathan Whittaker ______
Paul Ernsberger ______
______
(date) ______09/07/2006
*We also certify that written approval has been obtained for any proprietary material contained therein. Table of Contents
Table of Contents 1
List of Figures 4
Acknowledgements 6
Abbreviations Used 7
Abstract 8
Chapter 1. Literature review
Imidazoline ligands 10
Imidazoline receptors 11
Identification of imidazoline receptor subtypes 12
Cellular mechanisms of I1-imidazoline receptors 14
Molecular identity of I1-imidazoline receptor 17
PC12 pheochromocytoma cells as a model system 30
I1-imidazoline receptor agonists as therapeutic agents for the insulin
resistance syndrome 31
Insulin and Akt (PKB) cell signaling 33
Metabolic Syndrome X 37
SHROB as animal model for Metabolic Syndrome 40
Phenotypic features of the SHROB model 43
Obesity 43
Hypertension 44
Hyperlipidemia 45
Retinal abnormalities 47
1 Glucose metabolism 47
Insulin and insulin signaling 48
Conclusion 50
Summary 50
Chapter 2. Research design
Introduction 52
Specific aims 53
Chapter 3. Materials and methods
Materials 57
Plasma membrane isolation for binding assays 58
[125I]p-Iodoclonidine radioligand binding assays 58
Cell culture and transfection 59
Cell experiments 60
Akt activation assay in PC12 cells 62
Ruthenium red staining 63
Animals 63
Chronic drug treatment 64
Adipocyte isolation and insulin application 64
Western blot procedure 65
Glucose uptake assay 66
Statistical methods 68
Chapter 4. Manuscripts from the implementation of the research plan
Manuscript 1. 69
2 Manuscript 2. 105
Chapter 5. Discussion and significance
The significance of the present study on IRAS 133
Antisense and imidazoline binding studies 134
Antisense and I1-imidazoline cell signaling study 137
Possibility of IRAS as a subunit of imidazoline receptor 143
IRAS could also act as a scaffolding protein 145
Comparison of adipocyte glucose uptake results to previous studies 147
Impact of insulin concentration 147
SHR as control for SHROB 149
Insulin degradation in the experiments 151
SHROB as an animal model for human syndrome X 152
Akt activation studies 154
Chronic I1-R Activation 156
Acute I1R Activation 158
Clinical significance 159
Chapter 6. Future Studies 161
Bibliography 164
3 List of Figures
Figure 1. Working model of the I1R signaling pathway 16
Figure 2. Functional domains illustration of human IRAS protein 21
Figure 3. Sequence alignment of EST106158 and human IRAS 28
Figure 4. Sequence alignment of EST106158 and predicted rat IRAS 29
Figure 5. A simplified illustration of Akt(PKB) cellular function 36
Figure 6. Domain map of the rat IRAS gene and sequence comparison to
human and mouse 95
125 Figure 7. Saturation kinetics of [ I]-p-iodoclonidine binding to I1R in PC12
98
Figure 8. Effect of antisense treatment on IRAS protein expression 99
Figure 9. IRAS antisense inhibits I1R signaling 101
Figure 10. Effect of antisense treatment on basal activation of ERK1/2 103
Figure 11. Insulin induced ERK activation in PC12 cells 104
Figure 12. Representative Western blot showing phosphorylated and total Akt
immunoreactivity 125
Figure 13. Time course of insulin activation of Akt in lean SHR, SHROB and in
SHROB treated with moxonidine for 21d in vivo 126
Figure 14. Dose response curves for insulin activation of Akt in lean SHR,
SHROB and in SHROB treated with moxonidine for 21d in vivo 127
Figure 15. Basal Akt activation is not affected by phenotype or pharmacotherapy
128
4 Figure 16. Dose-response curve for insulin activation of [3H]-2-deoxy-D-glucose
uptake 129
Figure 17. Treatment with moxonidine alone in vitro does not affect Akt
activation 130
Figure 18. Representative blot showing the time course of Akt activation by
insulin with and without moxonidine pretreatment 131
Figure 19. Effect of in vitro moxonidine treatment on insulin activation of Akt in
adipocytes from SHR and SHROB 132
5 Acknowledgements
Helpful Lab Colleagues
• Paul Ernsberger, Ph.D (Advisor)
• Janean Johnson
• Anna Saal
• Ryan Strachan
Collaborators
• Laura Nagy, Ph.D
• Becky Sebastian
• Chung-Ho Chang, Ph.D
Thesis Committee
• Henri Brunengraber, M.D.
• Laura Nagy, Ph.D
• Jonathan Whittaker, Ph.D.
• Bryan Roth, M.D., Ph.D.
• Paul Ernsberger, Ph.D
Personal support
• Jing Han, Ph.D. Candidate
6 Abbreviations Used
2-DG 2-deoxy-D-glucose
α2AR alpha-2 adrenergic receptor
ERK Extracellular Regulated Kinase
I1-R I1-imidazoline receptor
IRAS Imidazoline Receptor Antisera-Selected
IRBP Imidazoline Receptor Binding Peptide
IRS Insulin Receptor Substrate (4 subtypes)
JNK cJun N-Terminal Kinase
MEK MAPK Kinase (phosphorylates ERK)
NGF Nerve Growth Factor
PAK p21-activated Kinase
PKB protein kinase B
PI3-K phosphoinositol-3-kinase
PIX PAK-interacting exchange factor
Rac Rho-family protein
SHR spontaneously hypertensive rat
SHROB spontaneously hypertensive rat obese strain
7 Abstract
The I1-imidazoline receptor is a novel target of drug development for
hypertension and insulin resistance. This thesis focused on the molecular basis
for I1-imidazoline binding and cell signaling and the mechanisms linking this
signaling protein to regulation glucose metabolism. IRAS is a gene candidate for the I1-imidazoline receptor. To investigate the possibility that IRAS is the I1-
imidazoline receptor, antisense oligo-nucleotides directly against the initiation site
of IRAS sequence were designed and transfected into PC12 cells. Antisense
transfection for 48h reduced specific imidazoline radioligand binding to plasma
membrane fractions by about 50%, with parallel drops in IRAS protein expression
as detected by Western blot. Furthermore, transfection with antisense caused
functional impairment of I1-imidazoline receptor signaling. Imidazoline agonist
induced ERK1/2 activation was significantly inhibited with antisense transfection
without affecting basal ERK level or ERK activation by growth factors. These
findings strongly suggested that IRAS encodes an I1-imidazoline receptor or at
least an important subunit of it.
The mechanism of insulin sensitizing effect from imidazolines was studied in
the SHROB rat, an animal model for human metabolic syndrome. Insulin induced
Akt activation was found to be severely impaired in isolated adipocytes from
SHROB compared to their lean SHR littermates. In addition, insulin induced
glucose uptake in these cells from SHROB were also similarly resistant to
stimulation by insulin. Chronic treatment of SHROB with the imidazoline agonist
moxonidine partially restored both Akt activation and glucose uptake stimulated
8 by insulin in isolated abdominal adipocytes without affecting basal Akt activation level. However, acute in vitro moxonidine administration did not yield similar effects, nor did moxonidine affect basal Akt level in adipocytes from either
SHROB or SHR. These results implicate adipose tissue as a locus of insulin resistance in this model of metabolic syndrome, and impairment of insulin signaling through Akt may contribute to this defect. Chronic oral treatment with I1- imidazoline receptor agonists such as moxonidine significantly normalizes insulin resistance in adipose tissue of SHROB in both insulin cell signaling and glucose metabolism. These changes in adipose tissue very likely contribute to the overall insulin sensitizing effect of imidazoline ligands on SHROB.
9 CHAPTER 1. LITERATURE REVIEW
Imidazoline Ligands
The discovery of imidazoline compounds greatly preceded the concept of
imidazoline receptors and can be traced back to Switzerland in 1939 (Hartmann
& Isler, 1939). The first two imidazoline drugs were tolazoline, an α-adrenoceptor
antagonist possessing vasodilating properties, and naphazoline, an α2-
adrenoceptor/I1-imidazoline receptor agonist, still in daily use as an over the
counter medication for topical application to relieve nasal congestion. Another
milestone in the history of imidazoline drugs was the discovery of clonidine in
1962, initially named St155, which was first developed as a nasal decongestant
but serendipitously was found to lower blood pressure, and subsequently
became the prototype for centrally acting antihypertensive drugs (Hoefke and
Kobinger, 1966;Ernsberger et al., 1987).
Clonidine is still in use, mainly in the form of a patch to be worn on the skin for continuous control of blood pressure (Klein et al., 1985). Clonidine is also widely used for certain psychiatric disorders (Hieble et al., 1991), including posttraumatic stress disorder (Harmon and Riggs, 1996), Tourette’s syndrome
(Leckman et al., 1991), autism (Jaselskis et al., 1992;Posey and McDougle,
2001), attention deficit disorder (Olfson, 2004) and opiate withdrawal (Agren,
1986). Clonidine may also have cognitive enhancing actions (Jackson and
Buccafusco, 1991). Clonidine has also been used to treat metabolic diseases such as glycogen storage disease (Asami et al., 1996), diabetic neuropathy
(Fedorak et al., 1985;Duby et al., 2004), and unrelated conditions such as
10 glaucoma (Robin, 1995), the defective thermoregulation of Shapiro’s syndrome
(Walker et al., 1992) and sickle cell anemia (Chang et al., 1983). Clonidine is
also a potent analgesic, equal to morphine when delivered directly into the spinal
cord (Eisenach et al., 1989). The traditional assumption has been that all of these
diverse effects are the result of a specific action solely on a single molecule, the
α2-adernergic receptor. However, one might speculate that at least one other molecular site of action might be involved.
After clonidine, hundreds of imidazoline compounds were developed, including idazoxan, efaroxan, rilmenidine, and moxonidine. Among those, rilmenidine (Laubie et al., 1985) and moxonidine (Armah et al., 1988) are two representative I1-imidazoline receptor agonists, which are considered as second-
generation centrally-acting antihypertensives and are currently being widely used
for clinical treatment of hypertension in Europe. Even more, clinical trials have
been and continue to be conducted for treatment of insulin resistance
complicated by hypertension (Reid, 2000;Doggrell, 2005).
Imidazoline Receptors
Since many imidazoline compounds have significant α2-adrenergic receptor
binding affinity, it had been assumed that the blood pressure lowering effects of clonidine and its relatives were mediated through α2-adrenergic receptors
(Laubie and Schmitt, 1977). Moreover, pretreatment with α2-adrenergic receptor
blockers such as yohimbine will attenuate the blood pressure lowering effects of
imidazolines.
11 The concept of an imidazoline receptor was first proposed by Bousquet in
1984, based on the in vivo central sympathoinhibiting action not only of clonidine,
but also clonidine analogs that act as antagonists rather than agonists at α2- adrenergic receptors (Bousquet et al., 1984). This hypothesis was soon expanded on by observations from various groups. Atlas & Burstein reported isolation of a clonidine-displacing substance from calf brain that was neither a catecholamine nor a peptide (Atlas and Burstein, 1984). Later, [3H]-clonidine was
found to bind to specific imidazoline-binding sites, in addition to α2-adrenergic
receptors, in the ventral medulla oblongata, which is the region of the brain
where clonidine acts to lower blood pressure (Ernsberger et al., 1987). The two
new antihypertensive ligands moxonidine and rilmenidine showed higher affinity
for I1-imidazoline sites than for α2AR, as well as higher selectivity for I1-
imidazoline sites than clonidine (Bock et al., 1999;Ernsberger, 2000). Also, a
positive correlation between the hypotensive actions of clonidine-like drugs and
the affinity at the imidazoline-binding sites, instead of α2AR binding sites, was
observed in the ventral medulla (Ernsberger et al., 1987). Up to now, there have
been over 1000 reports supporting I1-imidazoline receptor to be a novel and
distinct receptor with specific functions, with more supporting reports coming out
every year.
Identification of Imidazoline Receptor Subtypes
It is now widely accepted that imidazoline receptors are heterogeneous and at
least 2 subtypes, the I1 and I2, and possibly a third I3 subtype, have been
classified. The I1-imidazoline receptor (I1R) subtype is characterized as mediating
12 the sympathoinhibitory effects of a group of agents which act in the brainstem to
lower blood pressure, including clonidine, rilmenidine and moxonidine
(Ernsberger et al., 1995;Ernsberger et al., 1997). In addition to the central
nervous system where they modulate blood pressure, I1-imidazoline binding sites
were also found in peripheral tissues, including liver, kidney, and platelets
(MacKinnon et al., 1993;Wikberg et al., 1992;Molderings et al., 1993;Piletz et al.,
1994). In non-neuronal tissues, they are often found in epithelial cells lining
passages in the prostate gland (Felsen et al., 1994), lung and airways (Liedtke et
al., 1993;Piletz et al., 1999), and kidney (Allan et al., 1993;Ernsberger et al.,
1995;Greven and Bronewski-Schwarzer, 2001;Jovanovska et al., 2004).
The I2-imidazoline receptor, originally referred to as the idazoxan-preferring
receptor, is characterized by high affinity for some imidazolines (i.e. idazoxan
and naphazoline) as well as guanidines such as guanabenz and low affinity for the centrally acting antihypertensive ligands such as clonidine, moxonidine and rilmenidine (Ernsberger et al., 1995;Limon-Boulez et al., 1996). I2-imidazoline
binding sites are widely distributed in both CNS and peripheral tissues, and are
mainly located on the outer membrane of the mitochondria, strongly linked to the
activity of monoamine oxidase A and B (Limon-Boulez et al., 1996). Recent
autoradiographic studies of I2R in mice with genetic knockouts of monoamine oxidase A and B suggest the existence of non-mitochondrial I2R (Anderson et al.,
2006), but this hypothesis remains speculative.
The putative third imidazoline receptor, I3R, has only been identified in
pancreatic beta cells only so far (Chan et al., 1994;Prell et al., 2004;Squires et al.,
13 2004). The activation of I3R leads to increased insulin secretion, possibly through
a mechanism involving ATP-sensitive potassium channels. However, I1R and I3R
share similar ligand binding specificity, with the primary difference being the
relative efficacy of different agents (Eglen et al., 1998). For example, efaroxan is
a selective I1R antagonist but an I3R agonist, while it lacks affinity for I2R.
Cellular Signaling Mechanisms of I1-imidazoline Receptors
The knowledge of I1-imidazoline receptors cell signaling events is still limited
and the topic needs more intense investigation. So far the predominant cellular
model for investigation of I1R intracellular transmission mechanisms has been
PC12 rat pheochromocytoma cells. The main reason is that these cells express a
high density of I1-imidazoline binding sites in their plasma membranes and lack
α2-adrenergic receptors, which recognize many I1-imidazoline ligands (Separovic
et al., 1996). Moxonidine stimulates phosphotidylcholine-specific phospholipase
C (PC-PLC), thereby causing release of choline phosphate and diacylglyceride
(DAG, a second messenger). The release of DAG can be blocked by both
efaroxan, an I1R antagonist, and D609, an inhibitor of PC-PLC but not by
SK&F86466, a specific α2-adrenergic receptor antagonist, as expected
(Separovic et al., 1996). The I1-imidazoline receptor mediated PC-PLC activation
was also observed in rat ventral medulla oblongata in vivo, as either efaroxan or
D609 could block the blood pressure lowering action of moxonidine when locally
applied to this region of the brain (Separovic et al., 1997). Accumulation of
cellular DAG normally leads to the activation of protein kinase C (PKC). Both the
βII and ζ isoforms of PKC are activated with moxonidine stimulation in PC12 cells.
14 The activation of PKC-βII is from DAG accumulation, while the activation of PKC-
ζ is DAG insensitive and is possibly through arachidonic acid, which is also
released in response to imidazoline agonists (Ernsberger, 1998;Edwards et al.,
2001). A PKC-ζ relocation from cytosol to plasma membrane following moxonidine stimulation can also been observed. On the other hand, blockade of
PKC with chelerythrine actually potentiated the action of I1R to inhibit the release
of the inhibitory neurotransmitter γ-aminobutyric acid in the brain (Tanabe et al.,
2006).
Two cell signaling elements in the mitogen-activated protein kinase (MAPK)
cascade, ERK and JNK, were also found to be activated by moxonidine and
clonidine treatment in PC12 cells (Edwards et al., 2001). Activation of ERK and
JNK followed similar time courses with peaks at 90 min and phosphorylation level
up to 300% compared to basal level. These activations could be blocked by
efaroxan, indicating an I1-imidazoline receptor mediated mechanism. Another
observation from these experiments was that two-day treatment of PC12 cells
with the I1/ α2-agonist clonidine increased cell number by up to 50% in a dose
related manner. Together with the MAKP signaling stimulation, these findings
suggested a slight mitogenic action of I1R activation. Later, Dontenwill et al
reported that over expression of human IRAS protein, which is an I1-imidazoline
receptor gene candidate and will be discussed in detail later, led to prolonged cell
survival against known apoptotic stimuli in PC12 cells (Dontenwill et al., 2003b).
These studies suggested that I1-imidazoline receptor may also play a role in cell
growth or in cell survival. Figure 1 summarizes the current signaling model of the
15 I1R.
Figure 1. Working model of the I1R signaling pathway.
Rilmenidine, Moxonidine D609
Efaroxan PO4 Phosphocholine
1
2 PC- PO4 Ph PLC c os e ho ph in li at ol ne id az yl id r - -Im to I 1 cep Re D 3 i 4 g Arachidonic ly c Li 6 e pases acid r PKC-β id II e MAPK Eicos anoids Cascade ζ PKC- 5
Legend for Figure 1. Stimulation of the I1R in PC12 cells leads to activation of
PC-PLC, thus resulting in increased formation of DAG from phosphatidylcholine and the release of phosphocholine. DAG also serves as a substrate for the release of arachidonic acid into the extracellular space by the action of DAG lipase. These effects can be blocked by either the I1-R antagonist efaroxan or the
PC-PLC inhibitor D609. DAG and arachidonic acid then activate PKCβII and
PKCζ, which further promote phosphorylation of two members of the MAPK family
of kinase cascades, ERK and JNK.
16 Molecular Identity of I1-imidazoline Receptor
The molecular structure of I1-imidazoline receptor has not yet been identified despite intense efforts over the last 20 years (Bousquet et al., 2001). In the search for this receptor protein, most progress was made with two antibodies.
One was raised against a 70 kDa imidazoline ligand binding protein purified from solubilized bovine adrenal chromaffin cell membranes by ligand affinity chromatography (Wang et al., 1993). This became known as the imidazoline receptor related protein or “IRP antibody” and was extensively used in Western blot studies by groups around the world (Escriba et al., 1994;Escriba et al.,
1995;Garcia-Sevilla et al., 1996;Ruggiero et al., 1998). The other antibody was actually an secondary antibody raised against purified polyclonal anti-idazoxan antibodies (Bennai et al., 1996). An anti-antibody such as this is referred to as an anti-idiotype antibody (Greenspan, 1993;Linthicum et al., 1988;Vaux and Fuller,
1991). In some cases, antibodies against a ligand can used to generate anti- idiotype antibodies that recognize the receptor. In one example, patients with high levels anti-insulin antibodies will eventually develop anti-idiotype antibodies which bind to insulin receptors (Shoelson et al., 1986). By the same principal, antibodies against the imidazoline ligand idazoxan can be used to generate anti- idiotype antibodies that bind to imidazoline receptors (Bennai et al., 1996). One theory is that the amino acid sequence forming the binding pocket in the idiotype region of an antibody will have some homology with the amino acids forming the binding pocket of a receptor (Linthicum et al., 1988).
17 Multiple immunoreactive proteins from different tissues with a range from 29
to 95 kD were detected by one or both of these two antibodies (Szabo, 2002). At
first, most of the immunoreactive bands of putative imidazoline binding protein
were between 42-45 kDa. A 42 kDa peptide sensitive to clonidine was identified
with the anti-idiotypic antibody from human brain membranes (Greney et al.,
1994), while similar peptides detected in extracts from rat and bovine brains were
45 and 43 kDa respectively (Escriba et al., 1995;Heemskerk et al., 1998). Affinity
purification of proteins from rat brain also yielded a major species at 45 kDa
(Escriba et al., 1995). Bands at 29 and 45 kDa were increased in parallel in
postmortem samples of brains of Alzheimer patients relative to matched controls
(Garcia-Sevilla et al., 1998).
Detection of an 85 kDa band of imidazoline binding protein was also common
in human tissues including brain and platelets (Zhu et al., 2003). One group has
suggested that the 30 and 45 kDa proteins are related to the I2-imidazoline
subtype, since their expression is modulated by I2-selective antagonists, whereas the 43 and 85 kDa proteins may be related to the I1R (Garcia-Sevilla et al., 2004).
However, another group has shown that increasing the number of protease
inhibitors in the lysis medium increases the proportion of immunoreactivity found
at 85 kDa (Ivanov et al., 1998). These authors suggest that the main protein product in cell lystates is 85 kDa in size and the smaller bands represent
proteolytic fragments.
Now with the identification of IRAS as a much larger protein, even the 85 kDa
species appears to be a fragment of full-length IRAS due to proteolytic
18 processing. In the present experiments, immunoreactivity to antibodies raised
against the IRAS sequence did not identify any proteins larger than 85 kDa, and
only the 85 kDa band showed a decrease in response to antisense directed
against the IRAS gene. Ruthenium red staining identified a number of bands at
higher molecular weights, but the staining density was not affected by antisense, which only affected staining the region around 85 kDa. The IRAs protein has a
number of tryptic cleavage sites. Conceivably, IRAS may be cleaved to an 85
kDa fragment which is then active in binding and cell signaling, whereas the full
length protein is mainly involved in protein transport from endosomes.
A major breakthrough came in 1998, when a single 85 kDa protein was
shown to be recognized by both the antibody to the partially purified bovine protein and the anti-idiotype antibody in a variety of tissues (Ivanov et al., 1998).
Furthermore, the immunoreactivity of the 85 kDa band was linearly correlated with I1R radioligand Bmax values determined in parallel in nine rat tissues. Thus it was considered to be a candidate for the I1R, and those smaller protein bands
observed in previous studies were believed to be proteolytic fragments of this 85
kDa protein (Ivanov et al., 1998). Following this discovery, a truncated cDNA
clone was isolated from a human hippocampal lambda cDNA expression library
by relying on the selectivity of the two antisera directed against candidate I1-
imidazoline receptor proteins. The full length 5131-base pair molecule that was
subsequently isolated was then named imidazoline receptor-antisera-selected
(IRAS) gene, which encoded a 1504-amino acid 167 kDa protein. The amino acid sequence of IRAS shares no relationship with any other known imidazoline
19 binding proteins, but certain sequences of IRAS are consistent with signaling motifs found in cytokine receptors, as previously suggested from the cell signaling events of I1R, especially the activation of PC-PLC (Andrieu et al.,
1996;Cai et al., 1993;Cheng et al., 1999;Cowen et al., 1996;Galve-Roperh et al.,
1996;Jackowski et al., 1997;Laviada et al., 1995;Li et al., 1998;Machleidt et al.,
1996). However, the sequence of IRAS suggests that it is unlikely to be a G- protein coupled receptor as previously assumed (Piletz et al., 2000).
Since then, many studies have been conducted in order to identify the molecular functions and cellular mechanisms of this receptor candidate. IRAS transfection into Chinese hamster ovary (CHO) or pheochromocytoma (PC-12) cell lines were found to induce high affinity, specific imidazoline binding capacity
(Piletz et al., 2003). Dontenwill et al reported that PC12 and COS7 cells stably transfected with human IRAS gene exhibited prolonged cell survival protection against apoptotic stimuli, suggesting an anti-apoptotic effect of IRAS (Dontenwill et al., 2003b). Nischarin, which now is believed to be the truncated form (PX domain missing) of mouse IRAS, was found to bind preferentially to the cytoplasmic domain of the integrin α5 subunit of the fibronectin receptor and thus inhibit cell migration. Observations suggested that nischarin play a negative role in cell migration by antagonizing the actions of Rac on cytoskeletal organization and cell movement (Alahari et al., 2000;Alahari, 2003).
IRAS was also found to associate with insulin receptor substrate 4 (IRS-4) in human embryonic kidney 293 cells. The domains of IRAS and IRS-4 responsible for the association of these two proteins were also identified. In addition, IRS-1,
20 IRS-2 and IRS-3 were all found to associate with IRAS, just with lower affinity
relative to IRS-4 (Sano et al., 2002). Since insulin receptor substrates are key
components in signaling from the insulin receptor, consequently any proteins that
interact with them are expected to participate in insulin signaling. Thus, this
discovery may lead to the cellular mechanism of insulin sensitizing effect from
certain imidazoline compounds. The PX domain missing in nischarin was found
to be essential for association of the protein with phosphatidylinositol 3-
phosphate-enriched endosomal membranes. Both the PX and coiled-coil
domains were required for the localization of IRAS into endosomes in HK293
cells (Lim and Hong, 2004).
Figure 2. Functional domains illustration of human IRAS protein
Human IRAS/nischarin/I1R Imidaz- olines
PI3P ruthenium red receptor fibronectin
L L L L L Ser- Coiled integrin α5 Pro- PX R R R R R Coil R R R R R rich binding site rich
14 130 286 410 488 546 630 695 825 1043 1107 1504
rac PAK IRS-1 to -4
Legend for Figure 2. The major domains and motifs of the human IRAS
protein as described in the literature, with amino acid numbers from the protein
sequence. Abbreviations (from N-terminus): PX = phox homology domain, PI3P
= phosphatidyinositol 3-phosphate, LRR = leucine rich repeat. The PX domain
21 usually anchors to plasma and endoplasmic membranes through an interaction
with PI3P. The potential imidazoline binding region and ruthenium red binding
region are possibly overlapping within the coiled coil domain. The integrin α5 binding site also mediates interactions with rac and PAK. The insulin receptor substrate (IRS) proteins binding site resides at the C-terminal region of IRAS.
Within proteins that contain them, PX domains usually consist of 100-130 amino acids. The main cellular function of PX domains is to mediate lipid-protein
interaction with phosphorylated derivatives of phosphoinositides (PIP’s). Most
proteins that contain a PX domain are sorting nexins (SNXs) that function in the endosomes in a similar way as sorting nexin 1 (SNX1). Including IRAS, about 30
SNXs have been identified in mammalian cells so far, among which some have been found to play an role in membrane trafficking of plasma membrane
receptors (Worby and Dixon, 2002). The N-terminal PX domain in the human
IRAS can specifically associate with phosphatidylinositol 3-phosphate (PI3P) and
is required for targeting the protein to the early/sorting and recycling endosomes
(Lim and Hong, 2004). Thus it was suggested that IRAS could be a receptor that be anchored to the intracellular surface of the plasma membrane via its PX domain.
An acidic domain within human IRAS protein (aa 623–700) was found to share high similarity with some regions in ryanodine receptors that can bind to ruthenium red. Staining experiments confirmed the ruthenium red-binding ability of human IRAS. Furthermore, it was found that ruthenium red dye competitively
22 blocks specific imidazoline binding to I1-sites of PC12 plasma membranes,
suggesting this region could also serve as the imidazoline binding domain (Piletz
et al., 2000). Additional functions of this region of IRAS may exist. An overlapping
portion of this region (aa 630-695) is also a predicted coiled-coil region. The
presence of this region was found to be essential for the dimerization of IRAS
and its localization to endosomes (Lim and Hong, 2004). Thus this seems to be a critical region in the IRAS protein and worth further investigation for its through
structure related cellular mechanisms.
Residing in aa 286-410 of the human IRAS are 5 leucine-rich repeats (LRRs).
LRR was first noticed in the leucine-rich u2-glycoprotein (LRG) of human serum
(Takahashi et al., 1985). LRRs are protein interaction modules with 20-29 amino
acids. Each repeat consists of an 11 residue fragment, and always show a consensus sequence of LxxLxLxxN/CxL, followed by a variable stretch of 9-18 residues (Kajava, 1998). Several families of LRR-containing proteins are abundant in the brain (Chen et al., 2006). LRR proteins participate in many important cellular processes including hormone–receptor interactions, enzyme inhibition, cell adhesion and cellular trafficking. The exact function of the LRRS in
IRAS is still not clear. However, in most cases LRRs are implicated in protein- protein interactions (Kobe and Kajava, 2001). LRRs occur in a subfamily of G- protein coupled receptors (Hsu et al., 2000) and in insulin and growth factor
receptors (Edwards et al., 2001). These suggest that the LRRs in IRAS could
possibly mediate protein-protein interactions in IRAS’s cell signaling.
Proline-rich sequences play an indispensable role in mediating protein-protein
23 interactions. A superfamily of PRD (Proline-Recognition Domains) has been
identified to be able to recognize and interact with proline-rich regions. Among
these the SH3 domain is the most dominant (Saito et al., 2004). Since receptors
with a proline-rich region can generally assemble onto multi-element signaling
complexes (for example, the beta3-adrenergic receptor or the EGF receptor) and thereby modulate signal transduction, human IRAS might interact with SH3
regions on tyrosine kinases in the process of interacting with insulin receptor
substrates or other related elements. Interestingly, the predicted rat IRAS
sequence shows a much shorter and limited proline-rich region relative to human and murine sequences (Figure 6A). Whether this discrepancy could lead to any change in cellular functions of IRAS is still unknown and definitely worth investigation. However, given that the functional effects of imidazoline agonists are similar in humans and rats, whether in the intact organism or in isolated cells
(Ernsberger et al., 1995), it is likely that the primary functions of IRAS are independent of the proline-rich domain.
The serine-rich domain (aa 488-546) and proline-rich domain (aa 1043-1107) in human IRAS are very similar to those found in some cytokine receptors, i.e., the β–subunit of interleukin-2 receptors (IL2Rβ). It has been shown with
mutational analysis that these domains are required for the regular cell signal transduction functions of these cytokine receptors (Bazan, 1990;Lock et al.,
1998). These suggest that IRAS may mimic some properties of cytokine
receptors.
At present, IRAS is still the only I1-imidazoline receptor gene candidate.
24 However, only the human IRAS full length sequence has been published so far
(Piletz et al., 2000). The full amino acid sequence is listed below (GenBank
AAC33104.1).
1 matartfgpe reaepakear vvgselvdty tvyiiqvtdg shewtvkhry sdfhdlhekl
61 vaerkidknl lppkkiigkn srslvekrek dlevylqkll aafpgvtprv lahflhfhfy
121 eingitaala eelfekgeql lgagevfaig plqlyavteq lqqgkptcas gdaktdlghi
181 ldftcrlkyl kvsgtegpfg tsniqeqllp fdlsifkslh qveishcdak hirglvaskp
241 tlatlsvrfs atsmkevlvp easefdewep egttlegpvt aviptwqalt tldlshnsis
301 eidesvklip kiefldlshn gllvvdnlqh lynlvhldls ynklsslegl htklgniktl
361 nlagnllesl sglhklyslv nldlrdnrie qmeevrsigs lpclehvsll nnplsiipdy
421 rtkvlaqfge rasevclddt vttekeldtv evlkaiqkak evksklsnpe kkggedsrls
481 aapcirpsss pptvapasas lpqpilsnqg imfvqeeala sslsstdslt pehqpiaqgc
541 sdslesipag qaasddlrdv pgavggaspe haepevqvvp gsgqiiflpf tcigytatnq
601 dfiqrlstli rqaierqlpa wieaanqree gqgeqgeeed eeeeeeedva enryfemgpp
661 dveeeegggq geeeeeeeed eeaeeerlal ewalgadedf llehirilkv lwcflihvqg
721 sirqfaaclv ltdfgiavfe iphqesrgss qhilsslrfv fcfphgdlte fgflmpelcl
781 vlkvrhsent lfiisdaanl hefhadlrsc fapqhmamlc spilygshts lqeflrqllt
841 fykvaggcqe rsqgcfpvyl vysdkrmvqt aagdysgnie wasctlcsav rrsccapsea
901 vksaaipywl lltpqhlnvi kadfnpmpnr gthncrnrns fklsrvplst vlldptrsct
961 qprgafadgh vlellvgyrf vtaifvlphe kfhflrvynq lraslqdlkt vviaktpgtg
1021 gspqgsfadg qpaerrasnd qrpqevpaea lapapvevpa papaaasasg paktpapaea
1081 stsalvpeet pveapapppa eapaqypseh liqatseenq ipshlpacps lrhvaslrgs
1141 aiielfhssi aeveneelrh lmwssvvfyq tpglevtacv llstkavyfv lhdglrryfs
1201 eplqdfwhqk ntdynnspfh isqcfvlkls dlqsvnvglf dqhfrltgst pmqvvtcltr
1261 dsylthcflq hlmvvlssle rtpspepvdk dfysefgnkt tgkmenyeli hssrvkftyp
1321 seeeigdltf tvaqkmaepe kapalsilly vqafqvgmpp pgccrgplrp ktllltssei
1381 flldedcvhy plpefakepp qrdryrlddg rrvrdldrvl mgyqtypqal tlvfddvqgh
1441 dlmgsvtldh fgevpggpar asqgrevqwq vfvpsaesre klisllarqw ealcgrelpv
1501 eltg
25 The identified full length IRAS sequences from mouse and rat are still not available from GenBank. Murine nischarin has been widely accepted as a truncated version of IRAS protein because it lacks the N-terminal PX domain
(Lim and Hong, 2004). Also, many mouse EST clones that encode this N- terminal extension of Nischarin are available. Thus, to compare the protein sequences of IRAS among mouse, rat and human, we manually combined the mouse nischarin sequence (AF315344) with mouse mKIAA0975 protein
(BAC65694) to obtain an artificial “full length mouse IRAS amino acid sequence”.
The mKIAA0975 protein sequence came from a protein Blast analysis using the first 270 aa sequence of human IRAS as a template. Three mouse sequences
(BAC65694, BAC39757 and BAC29286) showed the highest homology (92%) and were identical to each other in the corresponding region (the missing portion in mouse IRAS was only 153 aa). The combined sequence consists of 1599 amino acids. This is somewhat larger than the human IRAS (1504 aa).
The rat IRAS sequence is the least clearly identified among human, mouse and rat. The only full sequence available from GenBank is XM_240330, which is a sequence predicted by automated computational analysis from annotated genomic sequence NW_047469. This predicted “Rattus norvegicus nischarin” sequence does contain the critical N-terminal PX domain and has a total of 1473 amino acids, which is the smallest among the three species.
With the CLUSTAL W (1.83) multiple sequence alignment program, we compared the “IRAS/Nischarin” sequences from human, mouse and rat discussed above, and the aligned result is shown in Figure 6A. Grey shaded
26 characters mean identical sequence. Bold and Italian characters indicate conserved amino acids among the three species.
The overall homology of the rat IRAS protein sequence is 83% relative to the human IRAS and 88% relative to the mouse sequence; while the overall homology of the rat IRAS nucleotide sequence is 92% relative to human and
93% relative to mouse (nucleotide sequences are not shown). Both rat and mouse IRAS protein sequences show similar functional domains in similar positions along the whole length, with the only one exception that rat sequence shows a much shorter and much less apparent proline-rich region in the corresponding aa 1017-1063 region. The significance of this difference is still not clear.
However, this predicted “rat IRAS/nischarin sequence” may not completely correspond to the real-world rat IRAS due to discrepancies with other data. It has been observed and confirmed by different research groups via western-blot that immunoreactive bands of imidazoline receptor protein in rat tissues or cell lines show an apparent molecular weight of 210 kDa (Dontenwill et al., 2003b;Piletz et al., 2000). Considering that the human IRAS has 1504 amino acids with molecular weight of 167 kDa, this predicted rat IRAS/nischarin with 1473 amino acids must be too small to be the 210 kDa band. Posttranslational modification cannot be ruled out, but glycosylation and myristilation consensus sequences are not present in rat IRAS.
Another conflict comes from Genbank EST 106158, which is an expressed sequence tag from rat PC12 cell cDNA clone RPCAX76. EST106158 contains
27 343 nucleotides and has been considered a partial sequence of rat IRAS gene
that encodes the 210 kDa immunoreactive protein in PC12 cells (Dontenwill et al.,
2003a). EST106158 and human IRAS cDNA share 79% identical nucleotides over the region as shown below in Figure 3, and the initial ATG codons are marked with bold and red. The initial ATG in EST106158 starts at nucleotide 32, and the ATG in human IRAS sequence starts at nucleotide 28, corresponding to nucleotide 126 in EST106158. Thus the rat IRAS protein should have about 31 additional amino acids at the N-terminal compared to human IRAS, which is consistent with the previously reported finding that the imidazoline binding protein
in rat tissues is larger than that in human (210 kDa vs. 167 kDa).
Figure 3. Sequence alignment of EST106158 and human IRAS
EST_106158 CCCTACCCACAGGCCCTNACTCTTGTTTTTGATGACACGCAGGGCCACGACCTCATGGGG 60 Human_IRAS ------
EST_106158 AGTGTCACCCTGGACCACTTCGGGGAGATGCCGGGTGGCCCTGGTAGGGCTGGGCAGGGC 120 Human_IRAS ------TGCGGCGGTGGCGGCGG--AGACC 22 ** *** * * ** ** *
EST_106158 CGGGAGGTGCAGTGGCAGGTGTTTGTGC-CCTGAGCGAGAGGCCGANCCTGCCAAGGAGG 179 Human_IRAS CGAACATGGCGACCGCGCGCACCTTCGGGCCCGAGCGGGAAGCCGAGCCGGCCAAGGAAG 82 ** ** ** * * * ** ***** ** ***** ** ******** *
EST_106158 CGCNCGTTGTGGGTTCCGAGCTCGTGGANACGTACACGGTATATGTCATACAGGTTACCG 239 Human_IRAS CGCGCGTCGTGGGCTCGGAGCTTGTGGACACTTATACGGTTTACATCATCCAGGTCACTG 142 *** *** ***** ** ***** ***** ** ** ***** ** **** ***** ** *
EST_106158 ATGGCAACCATGAGTGGACGATCAAACACCGCTATAGTNATTTTCATGACCTCCATGAAA 299 Human_IRAS ATGGCAGCCATGAGTGGACAGTAAAGCACCGCTACAGCGACTTCCATGACCTGCATGAAA 202 ****** ************ * ** ******** ** * ** ******** *******
EST_106158 AGCTTGTTGCTGAGAGNAAAATTNANAAAACTCTACTTCCACCC------343 Human_IRAS AGCTCGTTGCAGAGAGAAAGATTGATAAAAACCTGCTTCCGCCCAAAAAGATAATTGGGA 262 **** ***** ***** ** *** * **** ** ***** ***
Legend for Figure 3. Sequence alignment of EST106158 and human IRAS.
The initiating ATG codons in both sequences are underlined. Asterisks indicate
28 identical nucleotides. The two sequences share 79% identical nucleotides over
the region.
The comparison between EST106158 and the predicted rat IRAS/nischarin
nucleotide sequence is shown in Figure 4. The initial ATG codon in EST106185
probably occurs earlier than in this predicted sequence. This could be further
evidence supporting the idea that this predicted sequence is not the actual rat
IRAS gene.
Figure 4. Sequence alignment of EST106158 and predicted rat IRAS
EST106158 CCCTACCCACAGGCCCTNACTCTTGTTTTTGATGACACGCAGGGCCACGACCTCATGGGG 60 Rat__IRAS ------
EST106158 AGTGTCACCCTGGACCACTTCGGGGAGATGCCGGGTGGCCCTGGTAGGGCTGGGCAGGGC 120 Rat__IRAS ------A 1
EST106158 CGGGAGGTGCAGTGGCAGGTGTTTGTGCCCTGAGCGAGAGGCCGANCCTGCCAAGGAGGC 180 Rat__IRAS TGGCGGCTGCGACACTCAGC--TTCGGCCCTGAGCGAGAGGCCGAGCCTGCCAAGGAGGC 59 ** * *** * ** ******************* **************
EST106158 GCNCGTTGTGGGTTCCGAGCTCGTGGANACGTACACGGTATATGTCATACAGGTTACCGA 240 Rat__IRAS GCGCGTTGTGGGTTCCGAGCTCGTGGACACGTACACGGTATATGTCATACAGGTTACCGA 119 ** ************************ ********************************
EST106158 TGGCAACCATGAGTGGACGATCAAACACCGCTATAGTNATTTTCATGACCTCCATGAAAA 300 Rat__IRAS TGGCAACCATGAGTGGACGATCAAACACCGCTATAGTGATTTTCATGACCTCCATGAAAA 179 ************************************* **********************
EST106158 GCTTGTTGCTGAGAGNAAAATTNANAAAACTCTACTTCCACCC------343 Rat__IRAS GCTTGTTGCTGAGAGAAAAATTGACAAAACTCTACTTCCACCCAAAAAGATAATTGGAAA 239 *************** ****** * ******************
Legend for Figure 4. Sequence alignment of EST106158 and human IRAS.
The initiating ATG codon in EST106158 is underlined. Asterisks indicate identical nucleotides.
PC12 Pheochromocytoma Cells as a Model System 29 PC12 rat pheochromocytoma cells are the most popular cellular model for
investigations of I1-imidazoline receptor cell signaling pathways because these
cells constitutively express I1-imidazoline receptors but not α2-adrenergic
receptors, as confirmed with radioligand binding and molecular approaches
(Dontenwill et al., 2003b;Separovic et al., 1996). Recent studies revealed that
stimulation of I1-imidazoline receptors in PC12 cells with the imidazoline agonist
moxonidine induced more than a 2-fold increase in the proportion of ERK1/2 and
JNK in the dually phosphorylated active form (Edwards et al., 2001). As reviewed
earlier, moxonidine is a selective I1-imidazoline receptor agonist which has been
used as a clinical treatment for hypertension in Europe (Ernsberger, 2000).
The family of extracellular signal-regulated kinase (ERK)/mitogen-activated
protein kinase (MAPK), which contains protein Ser/Thr kinases, is widely
conserved among eukaryotes. These pathways allows the cell to respond to
divergent extracellular stimuli by controlling a multiple responses ranging from cell growth to apoptosis (Schramek, 2002;Schramek, 2002). Activated MAPKs
lead to phosphorylation of several substrates in PC12 cells including various
transcription factors (Cowley et al., 1994). In PC12 cells, ERK plays an essential
role in differentiation to a neuronal phenotype in response to nerve growth factor.
I1-imidazoline Receptor Agonists as Therapeutic Agents for the Insulin
Resistance Syndrome
Insulin resistance is a state wherein the normal circulating levels of insulin
secreted by the endocrine pancreas are insufficient to maintain glucose homeostasis. Insulin resistance is closely associated with cardiovascular
30 diseases as well being the major factor leading to type2 diabetes (Esler et al.,
2001). Glucose is the main sourse of energy utilized by cells in the human body.
However, most cells can not efficiently use glucose without the stimulation of
insulin which is a hormone secreted by pancreas. When insulin resistance occurs,
there is an impaired sensitivity of the cells to insulin in those insulin target tissues
such as muscle liver and adipose tissue, thus the blood glucose can not be
utilized by tissues as efficiently and quickly as before. Since the normal amount
of insulin is not enough to stimulate these tissues to take in and utilize blood
glucose as usual, to prevent the blood glucose level from getting too high, the
pancreas must produce additional insulin as compensation. Many people with
insulin resistance have high levels of blood glucose (hyperglycemia) and high
levels of insulin (hyperinsulinemia) circulating in their blood at the same time. If
this situation continues, insulin resistance may lead to type 2 diabetes (Liashko
and Dreval', 1973;Livingstone and Gould, 1995). The causes of insulin resistance are still not very clear but are undoubtedly complicated. Both genetic background
(i.e. defect in a specific group of genes) and environmental elements (i.e. obesity
and physical inactivity) are believed to affect the risk of developing insulin
resistance. The cellular mechanism of insulin resistance is also complicated.
Although accelerated insulin degradation can also cause insulin resistance, in most cases it is due to a decreased number of insulin receptors and/or postreceptor failure in peripheral cells (Nunes et al., 2000;Murphy and Nolan,
2000).
31 Although some imidazoline receptor ligands such as moxonidine and
rilmenidine have mainly been in clinical use for hypertension as centrally acting
agents for over 10 years especially in some European countries, more and more
recent studies revealed their potential beneficial effect on insulin resistance. A
retrospective analysis of 228 patients in a placebo-controlled hypertension trial
indicated a reduction in fasting glucose levels with chronic moxonidine treatment
(Schachter, 1999). In patients with both obesity and hypertension, moxonidine was found to not only reduce blood pressure by inhibiting sympathetic nervous system activity but also improve insulin resistance and reduce the plasma levels of leptin(Sanjuliani et al., 2006). Rilmenidine showed similar beneficial effects on plasma lipid and blood glucose control in patients with hypertension and metabolic syndrome (Anichkov et al., 2005). Results from studies carried out on animals are consistent with those from human patients. Both acute and chronic moxonidine treatments lead to improved glucose tolerance in SHROB rats with metabolic syndrome, probably through I1-imidazoline receptors(Velliquette and
Ernsberger, 2003b;Velliquette and Ernsberger, 2003a). Thus imidazoline agents
could be new targets for drug development against insulin resistance and even
metabolic syndrome. Investigations on the mechanisms of insulin sensitizing
effects from these agents would significantly facilitate the applications of them.
Insulin and Akt (PKB) Cell Signaling
Insulin is known to be the most potent anabolic hormone to stimulate the
uptake and storage of carbohydrates, fatty acids, and amino acids into glycogen,
fat, and protein, respectively. Insulin binding induces tyrosine
32 autophosphorylation of the insulin receptor and further catalyzes the tyrosine
phosphorylation of insulin receptor substrates (IRS-1 to IRS-4) as well as several
other intracellular substrates including GAB-1 and Shc (Holgado-Madruga et al.,
1996;Sasaoka et al., 1994).
In addition to tyrosine phosphorylation sites, IRS-1 also contains serine phosphorylation sites. Multiple factors have been fond to be able to induce the serine phosphorylation of IRS-1, including non-esterified fatty acids, tumor necrosis factor α and hyperinsulinemia (Rui et al., 2001). It has been found that serine phosphorylated IRS-1 becomes a poor substrate for tyrosine phosphorylation by the insulin receptor. More and more studies support the idea that serine phosphorylation could be a negative-feedback mechanism that uncouples the IRS-1proteins from their upstream and downstream partners and blocks insulin signal transduction under physiological conditions (Marchand-
Brustel et al., 2003).
Tyrosine autophosphorylated insulin receptor also activates Shc, which recruits the Grb2–SOS (Son of Sevenless) complex. SOS is an exchange- promoting protein which activates ras. Through the downstream elements raf and
MAPKK, insulin can activate the MAPK cascade, which is involved in cell growth and regulation of gene expression (Marchand-Brustel et al., 2003). This growth
promoting pathway coupled to insulin receptors may be independent of the
pathways regulating glucose uptake and metabolism.
The tyrosine phosphorylation of IRS family members generates docking
sites for several downstream proteins which all contain a SH2 domain. Among
33 them PI3-K is the predominant partner. PI3-K catalyzes the phosphorylation of
phosphatidylinositols on the three position. As one example,
phosphatidylinositol(4,5)-bisphosphate is converted into
phosphatidylinositol(3,4,5)-triphosphate on the inner side of plasma membrane.
This is considered to be a critical step in insulin cell signaling since inhibition of
PI3-K with the selective inhibitor wortmannin or by transfection with dominant
negative PI3-K mutants can completely prevent most actions of insulin (Okada et
al., 1994;Cheatham et al., 1994).
The increase in phosphatidylinositol(3)-phosphates results in the recruitment
and/or activation of several pleckstrin homology (PH) domain-containing proteins
including Akt and PDK-1. Akt is a 57kDa Ser/Thr kinase that plays a key role in the insulin induced PI3K-Akt pathway (Whiteman et al., 2002). Interactions
between PDK-1 and Akt cause conformational changes in the Akt molecule
inducing phosphorylation of Thr308 (Mora et al., 2004). However, this is not
sufficient for activation of Akt. Another protein kinase is required for the
phosphorylation of Ser473 on Akt. This protein is sometimes called PDK2 and the
question of its existence is still under debate, however recent studies suggest
mTOR (mammalian target of rapamycin) play this role (Sarbassov et al., 2005).
The Thr/Ser dually phosphorylated Akt then relocates from inner plasma
membrane to cytosol or into the nucleus and where it phosphorylates multiple
substrates and thus controls many insulin related cellular functions.
An important function of Akt is the regulation of glucose metabolism,
including glucose uptake, gluconeogenesis, and glycogen synthesis. For
34 example, the glucose transporter 4 (GLUT4) is the insulin-responsive transporter
in skeletal muscle and adipose tissue. With insulin stimulation, intracellular
vesicles containing GLUT4 translocate and fuse with the plasma membrane.
When located in the plasma membrane, GLUT4 can transport glucose from the
extracellular environment into the cell cytosol. A dominant negative form of Akt,
when overexpressed in adipocytes, blocks insulin-stimulated GLUT4
translocation (Zhou et al., 1997;Kohn et al., 1996). Akt also phosphorylates and
inhibits glycogen synthase kinase 3β (GSK-3β). Since GSK-3β usually
phosphorylates glycogen synthase and inhibits its activity, Akt activation leads to
glycogen storage in liver and muscle (Cross et al., 1995). The process of glucose
incorporation into glycogen is a major component of glucose disposal from the
plasma.
As a key element in the cell signaling of the anabolic hormone insulin, Akt
also regulates synthesis of various proteins. Akt translocates into the nucleus
and regulates transcription by phosphorylating targets such as the initiation factor complexes along with several other regulatory proteins including mTOR/FRAP and p70 S6K (Schmelzle and Hall, 2000).
Akt also exerts a strong anti-apoptotic effect on cells. Numerous targets of
Akt signaling have been identified to contribute to the prosurvival effect of insulin.
Caspases (cysteine–aspartic acid proteases) are responsible for cleavage of cellular proteins resulting in apoptosis of the cell. Akt phosphorylates caspase 9 and thus inhibits this activation to promote cellular survival (Cardone et al., 1998).
Akt also phosphorylates thus inactivates Bad, which antagonizes pro-survival Bcl
35 proteins such as Bcl-2 and Bcl-XL in mitochondria. This step is believed to maintain the integrity and functionality of the mitochondria for the production of
ATP (Yamaguchi and Wang, 2001).
Figure 5. A simplified illustration of Akt(PKB) cellular function
Legend for Figure 5: Akt(PKB) is a key element in insulin cell signaling. With
insulin binding, the tyrosine autophosphorylated insulin receptor phosphorylates
IRS, which then recruits and activates PI3-K. PI3-K catalyzes the transformation
of PIP2 into PIP3 on the inner side of plasma membrane. Akt and PDK1 are
recruited by PIP3, and PDK1 as well as PDK2 (which is possibly mTOR)
36 phosphorylate the Ser and Thr sites on Akt. Dually phosphorylated Akt gets activated and thus regulates glucose metabolism, promotes cell survival and regulates synthesis of various proteins in the nucleus.
Metabolic Syndrome X
Metabolic syndrome is a rising clinical challenge and is classified as a disease entity by the Center for Disease Control. It is referred to as the multi- metabolic and insulin resistance syndrome and is a precursor to type II diabetes.
Furthermore, it is also considered as a major risk factor for coronary heart disease (Haffner, 2006). More and more attention has been drawn onto this disease during the past few years ever since it was recognized to be tightly linked to the development of both cardiovascular disease and type 2 diabetes.
Cardiovascular disease has been and continues to be the major cause of death in the US (Podrid and Myerburg, 2005). Type 2 diabetes has affected 7% of whole US population in the year 2005 and this number is still growing. Medical cost and indirect expenditure caused by diabetes were estimated to be over $130 billion in the year 2002 (Babu and Fogelfeld, 2006). Thus it becomes very important to prevent the incidence of both cardiovascular disease and type 2 diabetes. Since metabolic syndrome may be considered as a precursor of these two major clinical entities, identification of metabolic syndrome and understanding its pathogenesis could significantly help to recognize the risk for
37 developing both diabetes and heart disease thus start therapeutic actions to
prevent these two diseases.
Metabolic syndrome X was first defined by Reaven in 1988 as a cluster of
metabolic disorders including hyperlipidemia, hyperinsulinemia, impaired glucose
tolerance and hypertension in which insulin resistance was hypothesized to be
the common aetiological factor for these disorders. At the same time, an
increased risk for atherosclerosis in the patients with these disorders was noticed
and the effect of both genetic and environmental factors (physical exercise and
obesity) on the severity of insulin resistance was emphasized (Reaven, 1988).
However, the efforts to more accurately define metabolic syndrome and to better understand the relationships between these clustered disorders have continued to intensify. For example, Kaplan grouped central adiposity, impaired glucose tolerance, hyperlipidemia and hypertension together as “the deadly quartet” to emphasize their importance in the development of cardiovascular disease one year after Reaven first proposed the syndrome. This was also the first time for
central obesity (increase of the splachnic and subcutaneous fat depots of the
abdominal region) to be considered as an important factor and a typical
component of metabolic syndrome (Kaplan, 1989). Later, many active
researchers in the field of diabetes used “insulin resistance syndrome” to describe these disorders. Based on accumulating evidence, they believed that insulin resistance, as well as obesity plays an important role in the cause and
development of the rest of disorders (DeFronzo and Ferrannini, 1991;Haffner et al., 1992).
38 Not only does the concept of metabolic syndrome continue to expand with the addition of more nd more related abnormalities, but the definitions of it by various organizations are always different and sometimes even contradictory. For example, the European Group for Study of Insulin Resistance (EGIR) excluded diabetes as a criterion in their definition of metabolic syndrome published in 1999
(Balkau and Charles, 1999) while The National Cholesterol Education Program
Adult Treatment Panel III (ATP III) did not (2001). However, the major components in these different definitions are still similar and in most of them three or more of these disorders are required for the diagnosis of Metabolic
Syndrome X with the presence of impaired glucose regulation or diabetes and/or insulin resistance considered to be mandatory (Kuzuya et al., 2002;Einhorn et al.,
2003).
According to the Third Report of the National Cholesterol Education Program
(NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood
Cholesterol in Adults, the prevalence of syndrome X in the adult human population in USA has been about 25% (Ford et al., 2002). Although a strong correlation between the incidence of metabolic syndrome and body weight in both adults and adolescents, the causality of Metabolic Syndrome X is still not totally clear yet (Park et al., 2003).
The ultimate goal for the clinical treatment on metabolic syndrome is to prevent both cardiovascular disease and the onset of diabetes. Although the pharmaceutical industry has recognized the metabolic syndrome X to be a novel target for drug development and put efforts on finding a single drug or a drug
39 combination for reducing multiple risk factors, so far the only drugs approved for treatment are still those that target the individual risk factors, i.e. anti- hypertensive drugs, lipid-lowering agents and weight-loss agents (Babu and
Fogelfeld, 2006). Thus, both the identification of potential drug candidates for treating multiple risk factors, and the establishment of appropriate animal models for human metabolic syndrome X required for testing these agents has become necessary and important.
SHROB as Animal Model for Metabolic Syndrome X
The obese spontaneously hypertensive rat (SHROB) or Koletsky rat strain is a potential animal model for human syndrome X. It arose from a novel naturally occurring mutation in 1969 (Koletsky, 1973). A female spontaneously hypertensive rat (SHR) descended from early breeding stock provided by the NHLBI colony in
1968 was mated with a male Sprague-Dawley rat and the resulting offspring that remained hypertensive (systolic blood pressure > 150 mmHg) were backcrossed to
SHR. The NHLBI colony of SHR was obtained as an inbred strain from Kyoto
University around 1967, where the strain was first characterized beginning in 1963
(Okamoto and Aoki, 1963). After several generations, an abnormal phenotype was observed among some of the litters, which was obesity. In addition to obesity and hypertension, these obese rats were also found to have other abnormal phenotypic characteristics, including hyperlipidemia, hyperinsulinemia, and proteinuria with fulminant kidney disease (Koletsky et al., 1995).
40 The obese genotype is sex independent and represents a homozygous
recessive trait originally designated as f (Koletsky, 1973) and later reclassified as
fak (Koletsky, 1975a). The mutant gene responsible for SHROB is recessive
since its characteristic effect is not present when paired with its dominant (lean)
allele (Fak). Male and female SHROB are infertile, thus the only way to get
SHROB is from the offspring of lean SHR and the recessive fak/fak genotype can
only be inherited when both parents are heterozygous that carry the same
recessive allele (Fak/fak). Averagely only one out of four puppies from
heterozygous SHR parents are SHROB, and the rest are either heterozygous or homozygous SHR. Heterozygous SHR (Fak/fak) are indistinguishable from
homozygous wild type lean littermates (Fak/Fak) except through breeding of
obese offspring (only heterozygous SHR can reproduce SHROB) or restriction
fragment length polymorphism identification by PCR. Thus in most studies with
SHROB as animal model, both heterozygous (Fak/fak) and homozygous (Fak/Fak)
lean SHR were used together as control.
The fak mutation in the SHROB was later found to be a nonsense mutation of
the leptin receptor gene, resulting in a premature stop codon in the leptin
receptor extracellular domain at amino acid 763 (Takaya et al., 1996). Due to this
early stop codon, all isoforms of leptin receptor are absent in these rats (Ishizuka et al., 1998b). Thus the mutation faK truncates all forms of the leptin receptor and
eliminates all possible downstream events triggered by leptin. Since leptin is an
adipocyte-derived hormone that acts as a major regulator for food intake and
energy homeostasis (Chan and Mantzoros, 2005), this naturally occurring knock-
41 out of leptin receptor results in a total loss in leptin action and thus causes
profound obesity and insulin resistance in SHROB.
In contrast, the Zucker fatty (fa) rat has a missense mutation (Glu269Pro),
which reduces the functionality of leptin receptor but not its presence (Sakurai-
Yamashita et al., 1997). Differences have been reported between the Zucker
fatty rat and the SHROB. The most important is blood pressure. The SHROB is
consistently hypertensive while the Zucker rat, according to most reports, is
normotensive or at best has only slightly elevated blood pressure (Bunag and
Barringer, 1988). Both SHROB (Koletsky et al., 1995;Abramowsky et al.,
1984;Koletsky, 1975a) and Zucker fatty rats (Kasiske et al., 1985) develop renal
disease. However, the renal disease is more rapid, severe and extensive in
SHROB, leading to mortality in SHROB but not in Zucker rats. Furthermore,
SHROB are somewhat heavier than Zucker fatty rats, particularly for females
(17%) (Yen et al., 1977). These differences between Zucker fatty rats and
SHROB may reflect similar mutations on differing genetic backgrounds (SHR versus Sherman and Merck stock rats) or the effects of a nonsense mutation of
the leptin receptor in SHROB versus a missense mutation in the Zucker. The
Zucker diabetic fatty (ZDF) differs from standard Zucker fatty rats in regards to blood glucose, with the ZDF but not the Zucker developing hyperglycemia and attendant complications (Friedman et al., 1991). Diabetes in the ZDF develops from a combination of obesity and susceptibility genes, possibly related to glucose transport (Friedman et al., 1991). The SHROB, despite being more obese than the ZDF or the standard Zucker, is highly resistant to the
42 development of diabetes even when fed a diabetogenic diet (Ernsberger et al.,
1999b).
Phenotypic Features of the SHROB Model
Obesity
SHROB weigh nearly the same as their lean littermates in the first 4 weeks before
weaning. At 4-6 weeks of age, SHROB of both sexes are still not overweight but start
to show a rounded contour of the lower trunk, resembling the outline of a light bulb;
this trait is the earliest to distinguish SHROB from lean SHR (Kumar et al.,
2001;Ernsberger et al., 1996b). After this period, SHROB gain weight rapidly (3 to 5
g/d). Food intake in young adult SHROB is increased by approximately 40%
compared with lean SHR littermates (Ernsberger et al., 1994). Mature SHROB
routinely reach peak weights between 750 and 1000g, which could be double or triple
relative to the body weights of SHR. However, at the period of stable body weight,
SHROB only consume only 20% more food than SHR do (Velliquette et al., 2005).
Thus, food intake per unit of body weight is actually reduced in SHROB. Obese
males are slightly but not significantly heavier than females at all ages. Lean SHR, in
contrast, show a marked sex difference in body weight, as is typical for rodent
species. Thus, SHROB females weigh about 3 times as much as SHR females,
whereas SHROB males weigh twice as much as SHR males (Ernsberger et al., 1994).
SHROB express enormous deposition of fat throughout the body. Retroperitoneal deposits within the abdomen and the sex-specific epididymal and myometrial and the mesenteric fat pads are enlarged by up to l9-fold in female and 12-fold in male
SHROB relative to SHR controls. However the most notable difference occurs in the
43 subscapular depot, which exceeds 30g in SHROB but is absent in SHR littermates.
At least part of the accumulated fat is likely endogenously synthesized, since the animals are fed standard chow containing only 5% fat by weight.
In a recent study in our lab the distribution of abdominal visceral and subcutaneous adipose tissue in SHROB and SHR animals was mapped using magnetic resonance imaging (MRI) (Wan et al., 2005). In SHROB, the total volume of adipose tissue was 275 ml, of which 113 ml (41%) was visceral and 161 ml (59%) was subcutaneous while in lean SHR the total adipose tissue was only 22.8 ml which could be divided into 13.7 ml (60%) of visceral and 9.1 ml (40%) of subcutaneous.
Thus SHROB contain more than 10 times of adipose tissue relative to lean SHR although the fold difference in body weight is only between 2 and 3, indicating a profound obesity of SHROB.
Hypertension
SHROB and their SHR littermates develop high blood pressure spontaneously before weaning around 30d of age. In both SHROB and SHR, systolic blood pressure reaches hypertensive levels (>150mm Hg) prior to weaning (<28d of age) and then rises progressively between 2 and 4 months.
High blood pressure is maintained until shortly before death. Unlike lean SHR where males show significantly higher pressure than females, SHROB show no sex difference in blood pressure (Kumar et al., 2001). Thus, similar to findings in humans, severe obesity removes the protective effect of female gender on blood pressure.
44 SHROB have slightly lower blood pressures than their lean littermates
throughout the life span (peak systolic pressure about 200 mmHg vs. 220 mmHg).
Blood pressure rises more gradually with age in SHROB relative to SHR. This
suggests that obesity does not contribute to hypertension in this model.
Furthermore, caloric restriction and weight loss do not normalize blood pressure in SHROB and may even increase it (Ernsberger et al., 1994;Koletsky and
Puterman, 1976;Ernsberger et al., 1994). The lack of leptin receptor and leptin
signaling may explain the reduced blood pressure of SHROB relative to SHR
littermates since leptin is known to activate sympathetic nervous system activity
and thus increase vasoconstriction (Mark et al., 1999). The mechanism(s)
responsible for hypertension in SHROB and their lean littermates are not known,
but presumably are polygenic in origin which means that multiple genes are
involved in the regulation, as has been shown in genetic marker studies for other
SHR substrains (Schork et al., 1995).
Hyperlipidemia
Hyperlipidemia is also an important component in the disorders associated
to metabolic syndrome X. Both male and female SHROB uniformly develop
hyperlipidemia characterized by a marked rise in plasma triglycerides and a
smaller rise in plasma total cholesterol relative to age-matched lean SHR
(Velliquette et al., 2006b;Kumar et al., 2001). Serum lipid levels in SHROB are
elevated as early as 5 weeks of age, and continue to rise throughout life until the
last few weeks of life when the animals are in a terminal decline (Koletsky,
1975a). Plasma triglyceride and total cholesterol values are not elevated above
45 the normal range in lean littermates of either sex. Lipid profiles are not different between sexes in either SHR or SHROB. Increased hepatic synthesis was observed on SHROB and this could possibly be a mechanism for elevated cholesterol in these animals (Tan et al., 1976;Velliquette et al., 2006a).
Triglyceride levels decline during weight loss on a very low calorie diet, and promptly rise again during regain of lost weight (Ernsberger et al., 2005).
Although the amount of adipose tissue in SHROB is over 10 times more than that in SHR, free fatty acids (FFA) following an overnight fast are elevated by only about
25% in SHROB relative to SHR (1.81 ± 0.09 (N = 50) compared to 1.45 ± 0.05 (N =
51). However FFA levels in both SHROB and SHR are more than double the levels typically found in normotensive control rat strains such as Sprague-Dolly (Velliquette et al., 2002b).
Although plasma FFA is only marginally elevated in SHROB relative to SHR,
SHROB show an abnormal regulation of FFA levels following a glucose challenge.
FFA levels fell at 30 and 60 min after a glucose challenge in SHR, as expected from the switch from a fasted to fed state (Velliquette et al., 2002c). In contrast, FFA levels fail to decline following a glucose load in SHROB. This failure to suppress circulating
FFA in response to a test meal has been reported in humans with type 2 diabetes
(Iannello et al., 1998), and is thought to reflect adipocyte insulin resistance.
Retinal Abnormalities
SHROB have a much shorter lifespan than their SHR or other rat models.
Most SHROB die with a short time frame between 7 and 13 months of age, while their SHR siblings show <5% mortality at these ages and . The shortened
46 lifespan of SHROB is primarily a result of kidney disease, protein wasting and ultimately renal failure. The SHROB retina shows signs of neovascularization and progressive capillary dropout (Khosrof et al., 1995). At 3 months of age, SHROB show arteriolar sclerosis, mild dilatation of larger vessels, increased capillary tortuosity and formation of primitive vascular tufts (Benetz et al., 1996). This retinopathy occurred in the absence of hyperglycemia, implicating other factors in its pathogenesis, such as hypertension, hyperlipidemia and hyperinsulinemia
(Benetz et al., 1996;Huang et al., 1995;Khosrof et al., 1995).
Glucose Metabolism
SHROB and SHR both show normal levels of fasting glucose. However SHROB show a much lower capability of blood glucose homeostasis following a glucose challenge relative to SHR. Oral glucose tolerance testing in SHROB demonstrated a sustained post-challenge elevation in circulating glucose from 60 to 360 min compared to lean SHR littermates. The area under the curve for glucose was far greater for SHROB than for SHR, suggestive of decreased rate of glucose disposal
(Velliquette and Ernsberger, 2003b).
Glycosylated hemoglobin reflects both fasting and postprandial glucose control over time. Despite normal fasting glucose, glycosylated hemoglobin was significantly increased in the plasma SHROB at 6 to 8 months of age (3.91 ± 0.21%; N=13) versus matched SHR controls (3.62 ± 0.13%; N=10; P<0.05, t-test). Although statistically significant, a net increase of less than one part in ten probably has no physiological importance. A very low calorie diet actually causes a further deterioration of glucose tolerance acutely, followed by a sustained improvement of
47 glucose tolerance during regain of lost weight, apparently from increased glucose disposal (Velliquette et al., 2005).
In contrast to normal fasting glucose, fasting insulin levels are elevated by 44-fold in SHROB relative to SHR. In response to an oral glucose challenge, insulin rose markedly in both SHROB (6.0-fold) and SHR (4.8-fold) rats for up to 360 min. The levels of C-peptide during fasting are very high in SHROB and rise to a similar extent as insulin in response to an oral glucose load. This implies that enhanced insulin secretion, rather than reduced insulin turnover, is the primary contributor to hyperinsulinemia (Velliquette and Ernsberger, 2003b).
Insulin and Insulin Signaling
Since SHROB rats have severe insulin resistance, pancreatic islets are greatly enlarged in these animals compared with SHR littermates as a compensation for elevated demand for insulin(Koletsky, 1975b). As a result, the fasting insulin levels are elevated by 44-fold in SHROB, in the presence of normal fasting glucose, indicating a condition of profound insulin resistance but still non-diabetic. Although the fasting plasma insulin level of SHROB is already extremely higher than SHR, with an oral glucose challenge of 6 g/kg , insulin can still rise markedly in an even larger fold scale than SHR (6.0-fold in SHROB and 4.8-fold in SHR) for up to 360 min. Thus the peak plasma insulin level in SHROB could be more than 55-fold higher than lean
SHR. The hyperinsulinemia is mainly due to enhanced insulin secretion from pancreas instead of reduced insulin degradation since the circulating levels of C- peptide during fasting are also very high in SHROB and rise to a similar extent as insulin in response to an oral glucose load.
48 Some tissue-specific data about insulin resistance are also available. The rate of
insulin-stimulated 3-O-methylglucose transport was reduced 68% in isolated
epitrochlearis muscles from the SHROB compared to SHR. Some upper stream
elements in insulin cell signaling were found to be impaired in SHROB. Insulin-
stimulated tyrosine phosphorylation of the insulin receptor β-subunit and IRS-1, in intact skeletal muscle of SHROB, was reduced by 36% and 23%, respectively, compared to SHR, due primarily to 32% and 60% decreases in insulin receptor and
IRS-1 protein expression, respectively. The levels of p85 regulatory subunit of phosphatidylinositol-3-kinase and the glucose transporter GLUT4 were reduced by
28% and 25% in SHROB muscle compared to SHR. In the liver of SHROB, the
insulin induced tyrosine phosphorylation level of IRS-1 was not changed, but insulin
receptor β-subunit phosphorylation was decreased by 41% compared to SHR,
probably due to a 30% reduction in insulin receptor expression levels. Reduced
expression of IRS-1 may in particular represent part of the molecular basis for insulin
resistance in this model. (Ernsberger et al., 1999a;Friedman et al., 1998;Ishizuka et
al., 1998a;Friedman et al., 1997a;Friedman et al., 1997c;Ernsberger et al., 1996a).
Defects of these insulin cell signaling elements apparently contribute to the overall
severe insulin resistance of muscle and liver in SHROB, however whether these are
primary causes or there are other elements that play more important roles in the
cellular mechanism of insulin resistance in SHROB still remain unclear. Also, there
are still little or no data available about cellular mechanisms of insulin resistance in
other insulin target cell types of SHROB, such as adipocytes, thus further
49 investigations on insulin signaling and cellular responses in target tissues of SHROB
are definitely necessary.
Conclusion
The SHROB rat is a unique animal model expressing multiple abnormal
phenotypic features including genetic obesity, spontaneous hypertension,
hyperinsulinemia and hyperlipidemia. These features closely resemble those found in
the human "Syndrome X”. The SHROB also has a spontaneous and progressive
nephrotic syndrome, which is a potential model for human diabetic and hypertensive
nephropathies as well as a retinopathy that resembles diabetes. Previous studies
toward the cellular mechanisms of insulin resistance in muscle and liver tissue of
SHROB have revealed some defects in insulin cell signaling elements. Further
studies directed at understanding the mechanisms and interactions between the
multiple abnormal phenotypes and their underlying genotypes will lead to better
understanding of these commonly intertwined clinical problems.
Summary
This introduction has reviewed the history and pharmacology of imidazolines, the
properties and subtypes of imidazoline receptors, and the cell signaling pathways
linked to the I1R subtype of imidazoline receptors. Previous work on the molecular nature of the I1R was reviewed along with new insights gained from recent entries to online gene banks. Next select features of the insulin receptor signaling cascade were reviewed, with an emphasis on features that come into play in the current research. Molecular, cellular, physiological and medical aspects of insulin resistance and the associated metabolic syndrome X were given coverage. Finally, the
50 characteristics of the animal model of insulin resistance and metabolic syndrome X used in these studies, the SHROB rat, were summarized.
51 CHAPTER 2. RESEARCH DESIGN
Introduction
Ever since IRAS was isolated as the gene candidate for the I1-imidazoline receptor, considerable effort has been focused on characterizing its cellular functions and the molecular properties in a handful of laboratories around the world. This very large protein does seem to conduct diverse cellular functions that are beyond what had been attributed to the I1-imidazoline receptor through past research. Surprisingly, most of these recent studies on the IRAS gene failed to take into account that IRAS is a receptor gene candidate. Thus, the paradigm of over expressed IRAS protein has been a dominating theme in IRAS related studies. This is unusual, in that over-expression of a receptor in the absence of agonist stimulation is not expected to have marked effects. Almost no ligand- receptor interactions and/or imidazoline induced cellular signaling changes were considered so far in these studies. Also, no reports of the consequences of under expression of IRAS on cellular behavior can be found, which could be important evidence for characterizing the function of IRAS as I1-imidazoline receptor or any
other functional cellular protein.
Metabolic syndrome X (a cluster of abnormalities including hypertension,
insulin resistance and glucose intolerance, hyperlipidemia and obesity) afflicts
25% of the U.S. population. Imidazoline agonists like moxonidine and rilmenidine
have been widely used for clinical treatment for hypertension. Various studies
have also suggested that imidazoline agonists may also be effective in treating
insulin resistance. However, the cellular mechanisms of the insulin sensitizing
52 effect from imidazoline ligands are still unknown. Also the contribution of each
insulin responsive tissue to the imidazoline induced insulin sensitivity increase is
still unclear.
My dissertation focuses on answering two critical questions regarding the
molecular basis of I1-imidazoline receptor and the cellular mechanism of
imidazoline induced insulin sensitizing effects. First, does the IRAS gene really
encode the I1-imidazoline receptor? More evidence is required to demonstrate
whether IRAS is the functional I1R protein that a large number of research groups have been searching for all these years. Second, how does imidazoline treatment improve insulin sensitivity and what tissue/cell level changes are involved in these actions?
Specific Aims
My first goal was to characterize the cellular function of the IRAS gene as possibly encoding an I1-imidazoline binding site. These studies were conducted
mainly with PC12 rat pheochromocytoma cells. PC12 cells are the most popular
cellular model in the investigations of I1-imidazoline receptor cell signaling
pathways because these cells constitutively express I1-imidazoline receptors but
no α2-adrenergic receptors, as confirmed with radioligand binding and molecular
approaches. Thus in this model I1-imidazoline receptors induced cellular actions from moxonidine would not be disturbed by α2-adrenergic receptors which have
moderate binding affinity with some imidazoline ligands.
I used an antisense oligo-nucleotide strategy to inhibit the IRAS expression to
probe whether the regular I1R-related cellular actions were also affected. Since
53 previous studies showed that over expressed IRAS induced high affinity I1- imidazoline binding, I hypothesized that inhibited IRAS expression would be associated with reduced plasma membrane imidazoline binding and altered cell signaling events. I chose activation ERK1/2 by dual phosphorylation of tyrosine and serine residues, a key step in the MAPK signaling cascade, as the indicator of I1-imidazoline receptor cell signaling.
In the first step, I transfected PC12 cells with antisense oligo-nucleotides and
isolated plasma membranes from these cells for radioligand binding assays of
specific imidazoline binding. If the initial hypothesis was correct, we would expect
that specific imidazoline binding, and in particular the density of these sites within
plasma membrane fractions, should decrease with antisense treatment relative to
controls that were transfected with unrelated cDNA. However, binding assays can be affected by other factors and are not definitive. Thus, it was necessary to
assess imidazoline receptor expression by independent means. We therefore
also measured immunoreactivity to a specific antibody to IRAS. We predicted
that this immunoreactivity would be decreased to the same extent as the density
of radioligand binding sites. Finally, we used a selective stain to assess the amount of IRAS protein using another means. We predicted a decline in ruthenium red reactivity selectively within the expected molecular weight range.
We next tested the possible impairment that antisense treatment might cause
in cell signaling events linked to I1- imidazoline receptors. We expected that
moxonidine induced ERK1/2 activation would be significantly inhibited with
antisense treatment. At the last, negative control and positive control
54 experiments were conducted to test the possibilities that antisense transfection itself might have lowered the basal ERK1/2 activation level or have damaged the
ERK1/2 activation machinery.
My second goal was to determine whether the insulin sensitizing effect from imidazoline treatment improved insulin cell signaling in adipose tissue and whether direct acute treatment on adipocytes could lead to similar effect. In all these studies, the SHROB rat model of metabolic syndrome X was used. The
SHROB is the leading animal model of human metabolic syndrome X, which is characterized by hypertension, insulin resistance and glucose intolerance, hyperlipidemia and obesity. Lean SHR littermates were used as controls. In order to carry out these experiments, I first compared insulin induced Akt activation in adipocytes from SHROB and SHR in both time and dose dependent manners.
Consistent with severe whole body insulin resistance in this rat model, isolated adipocytes from SHROB showed extremely impaired insulin induced Akt activation. Thus a platform for testing therapeutic intervention in insulin resistance in adipocytes was established. Next, I tested the possible insulin sensitivity change from chronic imidazoline in vivo treatment in adipocytes from
SHROB. We hypothesized that chronic moxonidine treatment would significantly improve insulin induced Akt activation in adipocytes from SHROB relative to untreated rats. Then I measured insulin stimulated glucose uptake in adipocytes from all the 3 groups (untreated SHROB, chronic moxonidine treated SHROB and SHR control). We had several predictions for the results. We expected the glucose uptake in adipocytes from SHROB to be significantly impaired relative to
55 SHR, and for chronic moxonidine treatment in vivo to partially restore insulin sensitivity ex vivo. I also tested the possible insulin sensitizing effect from acute in vitro moxonidine treatment directly on adipocytes using different concentrations and durations and orders of treatment. At the last, I designed negative control experiments to evaluate the possibility that moxonidine might directly alter the Akt activation levels from either chronic or acute treatments.
In these studies, I relied on multiple techniques from pharmacology, physiology, cell biology and molecular biology. I intended to clarify the molecular basis of I1-imidazoline receptor, and analyze the cell level mechanism of insulin sensitizing effect from imidazoline treatments. The present findings should help to unveil the mechanisms of I1-imidazoline receptor action, and facilitate the application of imidazoline ligands as treatment for insulin resistance.
56 CHAPTER 3. Materials and Methods
Materials
RPMI 1640 medium and DMEM/F12 medium and Horse serum were obtained
from GIBCO (Gaithersburg, MD, USA). Fetal bovine serum, calf serum and rat
tail collagen were obtained from Upstate Biotechnology (Lake Placid, NY, USA).
Oligonucleotides were synthesized by Sigma Genosys (The Woodlands, TX).
Moxonidine was provided by Solvay Pharmaceutical (Hannover, Germany).
Efaroxan and naphazoline and human recombinant insulin were obtained from
Sigma-Aldrich (St. Louis, MO). Phosphospecific anti-Akt rabbit polyclonal
antibody was obtained from Cell Signaling Technology (Beverly, MA, USA).
Rabbit anti-human-IRAS #1209, which is directed against amino acids 783 to 809
of human IRAS, was provided by Dr. John Piletz at Jackson State University
(Jackson, MS) (Piletz et al., 2000). Anti-phospho ERK1/2 antibody was
purchased from Cell-Signaling (Beverly, MA). Anti-ERK1/2 was purchased from
Sigma-Aldrich (St. Louis, MO). Anti-actin was purchased from Chemicon
(Temecula, CA). Anti-rabbit and anti-mouse IGG were both purchased from
Santa Cruz (Santa Cruz, CA). Anti-Akt rabbit antibody was purchase from Sigma
Aldrich (St Louis, MO, USA). [125I]p-Iodoclonidine was obtained from Perkin-
Elmer (Boston, MA) and stored at -20°C in ethanol. Protein assay reagents for
dye binding (Bradford) and bicinchoninic acid methods were obtained from
Pierce (Rockford, IL, USA). All other chemicals were from Sigma Chemical Co.
(St Louis, MO, USA) or Fisher (Pittsburgh, PA, USA) and were of analytical grade.
57 Plasma Membrane Isolation for Binding Assays
Cells harvested in 75-cm2 flasks were centrifuged at 1,000 x g for 5 min and
cell pellets were flash frozen and stored at -70oC. To isolate plasma membrane
fraction, PC12 cells were thawed and transferred into 15 ml plastic tubes (catalog
#05-527-45, Fisher, Pittsburgh, PA) before being homogenized in a polytron
(Tecmar Tissuemizer, 60% power for 15 s) in HEPES-buffered isotonic sucrose,
and centrifuged at 4000 x g for 5 min at 4oC to remove unbroken cells and nuclei
(P1). The supernatant was centrifuged at 75,000 x g for 25 min, and the resulting
P2 pellet was resuspended by vortexing in 2 ml of 50 mM Tris-HCl buffer, pH 7.7, containing 5 mM EDTA. After recentrifugation at 75,000 x g for 25 min, the pellet was resuspended in Tris-HCl alone and centrifuged. These washed crude membrane pellets were resuspended in 0.32 M sucrose and combined in a gradient tube. The P2 membranes were layered on top of a 0.85 M sucrose layer
(5 ml) that was itself layered over a 1.2 M sucrose layer (5 ml). These three-part
gradients were centrifuged at 100,000 x g for 90 min, and plasma membranes
were collected at the interface between 0.85 M and 1.2 M sucrose.
[125I]p-Iodoclonidine Radioligand Binding Assays
Plasma membranes were slowly thawed on ice and resuspended in Tris-
HEPES buffer (5.0 mM; pH 7.7 at 25°C, containing 0.5 mM EDTA, 0.5 mM EGTA, and 0.5 mM MgCl2) at a concentration of about 0.1 mg protein/ml. Membrane
aliquots (100 µl) were mixed with serial dilutions of [125I]p-iodoclonidine in water
(50 µl). Nonspecific binding was defined in the presence of 10 µM naphazoline
(Sigma-Aldrich, ST. Louis, MO). Incubations were performed in deep well 96-well
58 plates (Beckman Macrowell), initiated by the addition of membrane, and carried out for 30 min with high agitation in a 22°C incubator. Incubations were stopped by vacuum filtration using a cell harvester (Brandel M12) connected to a vacuum pump rated at 120 l/min (Edwards EM8). Samples were filtered over sheets of glass fiber filter paper (Schleicher & Schuell #34), which were preincubated for 4 h at 4°C in 0.03% polylethyleneimine to reduce nonspecific binding. Each sample well was washed four times with 4 ml ice-cold assay buffer, an operation completed in less than 12 sec. Individual filters were placed in 12 X 75mm disposable glass tubes, and counted at 80% efficiency (Packard Instruments,
Downers Grove, IL). Protein was assayed following the manufacturer’s instructions for the Coomassie (Bradford) Protein Assay Kit (Produce #23200,
Pierce, Rockford, IL).
Cell Culture and Transfection
PC12 cells (passage 59 to 70) were grown in either 75cm2 flasks (10 ml volume) or 6-well plates (2 ml volume) in RPMI 1640 (Product #0-040-CV,
Mediatech, Herndon, VA) supplemented with 10% horse serum (heat inactivated, product #26050-088, Invitrogen, Carlsbad, CA) and 5% fetal bovine serum
(product #16000-044, Invitrogen, Carlsbad, CA) at 37°C and in an atmosphere of humidified 5% CO2. Media were changed every other day. Cells were switched to collagen type-I coated 75cm2 flasks or 6-well plates 3 days before any cell experiment.
For antisense oligonucleotide transfection experiments, cells were grown to
70% confluence before being transfected with 6µg/ml oligo-DNA using the
59 GenePorterTM 2 transfection kit (Gene Therapy Systems, San Diego, CA)
according to the manufacturer’s instructions. Cells were incubated with antisense
or sense oligo-DNA for 48h before being harvested for binding assay or being
switched into serum-free medium 2 h prior to imidazoline ligand treatment for tests of ERK1/2 activation.
Cell Experiments
PC12 cells transfected with antisense or sense oligonucleotides were preincubated with serum-free RPMI1640 medium for 2 h just before moxonidine
(Solvay Pharmaceuticals, Hannover, Germany) application. Cells were treated with various doses of moxonidine (0.1nM -1µM) for 90 min. Moxonidine was dissolved in 20% DMSO as a 1mM stock solution before being serially diluted into working solutions with serum-free RPMI 1640 according to final moxonidine concentrations in cell incubations (the v/v ratio of working solution to cell culture was 1:100). Upon the end of each treatment with moxonidine or vehicle (DMSO; final concentration 0.02%), media were removed immediately and cells were rinsed with ice cold PBS buffer (normal strength, Ca2+ and Mg2+ free) twice
before adding lysis buffer (NaCl 150mM, Tris-HCl pH 7.4 50mM, EDTA 5mM,
EGTA 2mM, sodium orthovanadate 1mM, PMSF 1mM, NaF 50mM, Triton-X 100
1%, NP-40 1%, plus Protease Inhibitor Cocktail (Boehringer Mannheim, one tablet per 50ml buffer). The 6-well plates were then frozen at –70oC for at least
30 min before being thawed slowly on ice. Thawed cell fragments then were
removed from the plate with plastic scrapers, followed by centrifugation at
12,000g for 15min at 4oC to remove insoluble debris. Supernatants were
60 collected and total protein concentration of each resulting supernatant was
measured in a 5 µl aliquot with the bicinchoninic acid (BCA) Protein Assay Kit
from Pierce (Rockford, IL, USA). Samples were divided into aliquots containing
25µg of total protein in double strength Laemmli buffer in a total volume of 50µl.
Aliquots were subjected to 10% SDS-PAGE gel according to instructions of Bio-
Rad Mini-PROTEAN 3 system (Bio-Rad, Hercules, CA) and electrophoretically
transferred to a nitrocellulose membrane for immunodetection. The blots were
blocked with 5% fat-free powder milk dissolved in TBS-T buffer (Tris-HCl 20 mM,
NaCl 125 mM, Tween-20 0.1%, pH 7.5) before being incubated overnight with
anti-active ERK1/2 (product #9101,Cell Signaling Inc, Danvers, MA) at a dilution
of 1:10,000 in 5% BSA (product #A6003, Sigma-Aldrich), and anti-ERK1/2
(product #M5670, Sigma-Aldrich) antibodies (1:1,000) in 5% BSA at 4oC with
gentle shaking. IRAS expression was determined with anti-IRAS #1209
polyclonal antibody at a dilution of 1:1,000. Total actin was measured as loading control by using monoclonal anti-actin at 1:10,000 dilution (Chemicon, CA). After repeated washes in TBS-T buffer for 1 h, blots were incubated with horseradish
peroxidase-conjugated donkey-anti-rabbit or donkey-anti-mouse (Santa-Cruz, CA)
in 5% non-fat milk at room temperature for 1 h. Then blots were washed 3 x 10
min with TBS-T before being visualized according to instructions of Pierce ECL
Western Blotting Substrate (product #32106, Pierce, Rockford, IL), digitized
(ScanMaker 4700) and quantified by densitometry as grey scale optical density multiplied by area in pixels (UN-Scan-It, Silk Scientific, Orem, UT). Data were expressed as the ratio of the densitometric signal for phospho-ERK1/2 to that of
61 total ERK1/2. Data were further normalized to the value of vehicle controls or
baseline. The densitometric signal for IRAS immunoreactivity was calculated as a
ratio to that for total actin.
Akt Activation Assay in PC12 Cells
PC12 cells were grown in RPMI 1640 with 10% horse serum, 5% fetal bovine
serum, 100 U/ml penicillin, and 100 mg/ml streptomycin. Cells were transferred to rat tail collagen coated 6-well culture plates 7 days before experiments were performed, and were switched to low-serum medium (1% horse serum and 0.5% fetal bovine serum) for 16 h prior to imidazoline treatment. Right before application of an imidazoline agonist, cells were switched to serum-free RPMI
1640. Various concentrations of moxonidine (1 nM to 10µM) or rilmenidine
(0.5µM; both from Sigma, Saint Louis, MO) were applied to the cells in the presence or absence of efaroxan (10µM; Sigma) for 10 min. In the efaroxan experiments, the cells were pretreated for 10 min before addition of agonist. After treatment, cells were washed with ice cold phosphate buffered saline (pH 7.4) and removed from the plate by scraping and then collected by low speed centrifugation. Cells were then homogenized in lysis buffer followed by centrifugation at 12,000g at 4oC for 15min. Equal amount of protein (25µg) from
the resulting supernatants were subjected to SDS-PAGE and electrophoretically
transferred to a nitrocellulose membrane for immunodetection with anti-active Akt
(Cell Signaling, Inc.), and anti-Akt (Sigma) antibodies. The anti-active Akt
recognizes active forms of Akt (pho-Ser473) and the anti-Akt recognizes all forms
of Akt. Results were expressed as a ratio of active/total Akt immunoreactivity.
62 Ruthenium Red Staining
This procedure was slightly modified from a previous method (Chen and
MacLennan, 1994). Briefly, after protein gel electrophoresis and transfer onto
nitrocellulose membranes, blots were washed twice for 10 min each in 50mM
Tris buffer (pH 8.0) containing 60 mM KCl and 5mM MgCl2 with gentle shaking
before being incubated with 30 µM ruthenium red in the same buffer for 15 min.
The blots were then washed 3 more times for 15 min each and wrapped in
RealviewTM clean film while wet before being scanned with a Microtek
ScanMaker 4700 scanner for quantification. Blots with ruthenium red staining
would then be washed 3 x 10 min before being blocked with 5% non-fat milk and incubated with monoclonal anti-actin as described above in “Cell Experiments” part for total actin detection as loading control.
Animals
Experiments were performed on homozygous male and female SHROB
(fak/fak) (body weight of 489 ± 21g). Age and sex matched lean SHR littermates
(Fak/fak or Fak/Fak) were used in these experiments as well (body weight of 375 ±
11g, P < 0.001 versus SHROB). Animals were housed individually and were provided standard food (Teklad formula 8664) and water ad libitum. Animals
were on a 12:12h light-dark cycle (lights on from 7:00 to 19:00) and were
maintained at a constant temperature of 21ºC. These procedures were carried
out with the approval of the Case Western Reserve University Institutional Animal
Care and Use Committee.
Chronic Drug Treatment
63 Moxonidine was delivered to SHROB and SHR rats orally by adding it to
powdered rat chow prior to pelleting (93.8 mg/kg of chow) at a dose estimated to
provide 4 mg/kg/d. Rats were provided with moxonidine-laced or control chow
and water ad libitum. At day 21 of drug treatment, SHROB consumed 20.1 ± 1.9
g chow and had a body weight of 489 ± 21 g. Thus, the ingested dose of
moxonidine was 3.9 ± 0.3 mg/kg/d. SHR consumed 18.2 ± 1.1 g chow and had a
body weight of 375 ± 11 g. Thus, the ingested dose of moxonidine was 4.5 ± 0.3
mg/kg/d. In Akt activation experiments, rats were fasted for 16 h before sacrifice.
For glucose uptake experiments, rats were not fasted, because adipocytes take up glucose mainly in the postprandial state. Animal sacrifices took place at the
same time each day (10:00 to 11:00).
Adipocyte Isolation and Insulin Application
SHR or SHROB were fasted 18 h, and then anesthesia was induced with
ether and maintained with isoflurane delivered through a nose cone. The
abdomen was cleaned with 70% ethanol, and an abdominal incision was made.
Gonadal fat tissue (4g) was taken for each experiment. Adipose tissue was
briefly rinsed with Krebs Ringer Bicarbonate buffer supplemented with HEPES
(Krebs’ buffer with MgSO4 1.2 mM, CaCl2 2.5 mM, glucose 5.5 mM, Na2CO3 25 mM, HEPES 10 mM, pH7.4, bubbled with 95%O2-5%CO2 for 15 minutes) before
being minced with scissors. The minced fat tissue was then digested in 20 ml of
above Krebs Ringer Bicarbonate buffer supplemented with HEPES plus 0.1%
collagenase type VIII (Sigma; catalog# C2139) by shaking in plastic flask at 80
rpm for 1h at 37oC (total digestion volume would be about 25 ml). Then the
64 adipocytes were filtered through a 250 um mesh and rinsed at least 3 times with
PBS pH7.4 containing 1 mM sodium pyruvate, 0.1% BSA, 20 I.U./ml penicillin and 20 µg/ml streptomycin. The adipocyte suspension from the entire fat depot was equally distributed into eight 50ml plastic tubes with a volume of 2 ml each, and pre-incubated at 37oC with shaking at 80rpm in a water bath for 1h before insulin was applied. In the acute in vitro moxonidine treatment experiments, cell suspensions were treated with 100nM moxonidine or vehicle (0.002% DMSO) for
90 min before insulin (product #I2643, Sigma-Aldrich) was applied. In the insulin dose-dependence experiments, various concentrations of insulin (0.1nM, 1nM,
10nM, 100nM and1µM) or vehicle (0.6µM acetic acid) were directly applied to cell suspensions. The mixtures were then incubated for 10 min with shaking before being harvested. In the insulin time-course experiments, 100nM insulin or vehicle was applied into cell suspensions for multiple time lengths from 3 min to 90 min with tubes shaking in a water bath before cells were harvested. Cells were harvested by placing the tubes in ice for 2 min, then infranatants were removed
(cell volume ≈ 0.5 ml), and 0.5ml Laemmlli buffer with 5% β-mercaptoethanol was added before immersion in a boiling water bath for 10 min. The aqueous phase in each tube was stored at -70°C for later assay by Western blotting.
Western Blot Procedure
The aliquots were electrophoresed on a 10% SDS-PAGE gel and electrophoretically transferred to a nitrocellulose membrane for immunodetection.
The blots were incubated overnight with anti-active Akt(Ser473) (product #9271,
Cell-Signaling, Danvers, MA) at a dilution of 1:5,000 in 5% BSA. After repeated
65 washes with TBS-T buffer, blots were incubated with horseradish peroxidase- conjugated donkey-anti-rabbit in 5% milk at room temperature. Blots were visualized with enhanced chemiluminescence (Pierce, Rockford, IL), digitized
(ScanMaker 4700) and quantified by densitometry as grey scale multiplied by pixels (UN-Scan-It, Silk Scientific, Orem, UT). All blots were then stripped with stripping buffer (Bio-Rad, Hercules, CA) as instructions described before probing with anti-Akt antibody (product #P1601, Sigma) at a dilution of 1:1,000. Data were expressed as the ratio of the densitometric signal for phospho-Akt to that of total Akt. Data were further normalized to the value of untreated controls or baseline.
Glucose Uptake Assay
Gonadal fat tissue (4g from SHR, 8g from SHROB) was taken from non- fasted rats for each experiment as described in “Adipocyte Isolation and Insulin
Application”. Each fat depot was rinsed twice with 10 ml Hanks’ buffer containing
1% BSA and 25mM HEPES at pH 7.4 for 5 min each, then minced with scissors into cubes of roughly 2 mm before being digested in 20 ml Hanks’/BSA/HEPES containing 1mg/ml collagenase type-IV (product #C5138, Sigma-Aldrich) for 50 min in 37oC waterbath with shaking in a polypropylene flask at 80 rpm. All BSA used in glucose uptake assay was RIA grade (product #7888, Sigma-Aldrich).
Then the whole digestion mixture was filtered through nylon mesh (250 µm, Sefar
America Inc.) into a new 50 ml plastic centrifuge tube and adipocytes would float onto the surface of filtered cell mixture. Adipocytes were allowed to float for 5 min before being purified by removing the infranatent and adding 15 ml wash buffer
66 (PBS at pH7.4 containing 1mM sodium pyruvate, 0.1% BSA, 25mM HEPES,
2.5mM MgCl2 and 2.5mM CaCl2) for at least 3 times until the infranatent was clear. Purified adipocytes were suspended in 10 ml volume. Samples of 10 µl of the cell suspension were counted under a microscope with a hemocytometer to ensure a concentration of 106 cells per ml by adding or removing wash buffer
accordingly. Adipocytes were then equally distributed into six 20 ml plastic tubes shaking at 100 RPM in 37oC water bath, each containing 106 cells in a total
volume of 1ml. The six tubes were labeled as “0.1nM”, “1nM”, “10nM”, “100nM”,
“basal” and “non specific” respectively. At time 0, various concentrations of
insulin in 100 µl of wash buffer were applied to the cell suspensions respectively
to make each tube contain 0.1nM, 1nM, 10nM or 100nM final insulin levels
according to the labeling while only 100 µl pure wash buffer was added into the
“basal” tube. At minute 20, cytochalasin-B (product #250225, Calbiochem,
Madison, WI) was added into the “non specific” tube along with 100 µl wash
buffer to get a final concentration of 10 µM for blocking all GLUT-specific glucose
transport. At minute 27, 150 µl of cell suspension from each tube was gently
transferred into a 300µl fine centrifuge tube with on top of a cushion of 70µl of an
oil mixture (57% mineral oil and 43% silicon oil v/v). Each sample was transferred
in triplicate. At minute 30, Glucose uptake was initiated by adding 50µl of 2.5mM
2-DG dissolved in wash buffer with 0.33 µCi of [3H]-2- deoxy-D-glucose (specific
activity = 2.64 Ci/Mol) into the cell suspensions. At minute 33, glucose uptake
was stopped by centrifugation for 10 s at 5,000g to separate cells from media with oil mixture layer. The cell fraction was removed from the rest of medium by
67 cutting through the middle of the oil section with a razor blade, and the top of
tube containing the cells was put into a scintillation vial with 4ml of scintillation
fluid. After a 3 s vortex, vials were counted for [3H]-2-DG incorporation for 5 min
and specific glucose uptake was defined as basal 2-DG incorporation minus
incorporation with cytochalasin-B. Results were expressed as nmol of [3H]-2-DG
incorporated per 105 cells per 3 min reaction time.
Statistical Methods
Results are presented as means ± standard error of the mean. Group means
were compared by using t-tests for simple comparisons or the Newman-Keuls
test for post-hoc analyses following analysis of variance. Time course data were
analyzed by two-way analysis of variance with repeated measures. Saturation
binding data were analyzed by nonlinear curve fitting to a hyperbolic function:
Y=Bmax*X/(Kd+X), where Bmax is the maximal binding, and Kd is the concentration
of ligand required to reach half-maximal binding with no ligand depletion. Dose
response data were analyzed by nonlinear curve fitting to the sigmoidal logistic
function: Y=Bottom + (Emax-Bottom)/(1+10^((LogEC50-X))), where X is the
logarithm of concentration, Y is the response, and Y starts at Bottom and goes to
Emax with a sigmoid shape. EC50 is the concentration of test agent hat produces
50% of the maximal response (Emax). All analyses were carried out using Prism
4.0 (GraphPad Software, San Diego, CA) as suggested by the software authors
(Motulsky and Ransnas, 1987).
68 CHAPTER 4. MANUSCRIPTS FROM THE IMPLEMENTATION OF THE
RESEARCH PLAN
Manuscript 1:
Submitted to The Journal of Neurochemistry
Title:
IDENTIFICATION OF IRAS/NISCHARIN AS AN I1-IMIDAZOLINE RECEPTOR
Abstract:
The I1-imidazoline receptor (I1R) is a proposed target for drug action relevant to blood pressure and glucose control. The IRAS (Imidazoline Receptor
Antisera-Selected) gene has several characteristics of an I1R. To test the
contribution of IRAS to I1R binding capacity and cell signaling function, an
antisense probe was evaluated in PC12 rat pheochromocytoma cells, a well-
established model for I1R action. The density of I1R was significantly reduced by
antisense compared to control transfection (Bmax=400±16 versus 691±29 fmol/mg
protein), without significantly affecting binding affinity (Kd=0.30±0.04 versus
0.39±0.05nM). Thus, IRAS expression is necessary for high affinity binding to I1R.
Western blots with polyclonal anti-IRAS showed reduced IRAS expression in the major 85 kDa band relative to an actin reference, paralleling the reduction in binding site density. To determine whether reduced IRAS expression attenuated
I1R cell signaling, PC12 cells transfected with antisense or sense oligo-DNA were
69 treated with moxonidine, an I1R agonist, then cell lysates were analyzed by
Western blot. Dose-dependent activation of ERK was attenuated without
affecting the potency of the agonist. In contrast, ERK activation by insulin was unchanged. The IRAS gene is likely to encode an I1R or a functional subunit.
70 Introduction
Non-adrenergic receptors for imidazolines were first proposed based on
functional studies in the brainstem (Bousquet et al., 1984). Radioligand binding
studies subsequently identified high-affinity sites which recognized clonidine and
other imidazoline adrenergic agents, but not catecholamines or other non-
imidazoline adrenergics (Ernsberger et al., 1987). These imidazoline binding
sites of the I1 subtype (I1R) were localized to plasma membrane fractions
(Ernsberger and Haxhiu, 1997;Piletz et al., 1991) and specifically to synaptic plasma membranes (Heemskerk et al., 1998). This contrasts to mitochondrial I2R,
which reside within the monoamine oxidase protein (Raddatz et al., 1995). The
subcellular localization of a putative I3R in pancreatic β-cells is not yet known
(Eglen et al., 1998).
Cell signaling responses were linked to this putative receptor, and could
be induced by selective agonists and blocked by selective antagonists. These
studies have chiefly used the PC12 rat pheochromocytoma cell model. Cellular
responses to imidazoline agonists included hydrolysis of phosphatidylcholine into
diacylglyceride and phosphocholine (Separovic et al., 1996;Separovic et al.,
1997;Zhang et al., 2001), release of arachidonic acid and prostaglandin E2
(Ernsberger, 1998;Ernsberger, 1999;Prell et al., 2004), activation of the βII and ζ
isoforms of protein kinase C (Edwards et al., 2001;Edwards and Ernsberger,
2003), and conversion of ERK-1, ERK-2 and c-jun kinases to their active forms
(Zhang et al., 2001;Edwards et al., 2001;Prell et al., 2004). Agonists at I1R also
alter gene expression, increasing expression of phenylethanolamine-N-methyl
71 transferase (Evinger et al., 1995) and the dual specificity phosphatase MKP-2
(Edwards and Ernsberger, 2003). Treatment with I1R ligands increases cell numbers in PC12 cell cultures (Edwards et al., 2001) probably by inhibiting apoptosis and prolonging survival (Dupuy et al., 2004;Dontenwill et al., 2003b).
A gene candidate for the I1R protein has been proposed (Piletz et al.,
1999;Piletz et al., 2000;Piletz et al., 2003). Antiserum prepared against partially
purified bovine I1R protein (Wang et al., 1993) was used along with an anti-
idiotype antibody (Bennai et al., 1996) to clone a human mRNA encoding a 1,504
amino acid protein, which was named the imidazoline receptor antisera selected
(IRAS) gene (Ivanov et al., 1998). IRAS is structurally unrelated to any known G-
protein coupled receptors, including the α2-adrenergic receptors (α2AR) which
bind many of the same imidazoline compounds. The lack of an N-terminal signal
for insertion of IRAS into membranes raised a question of how IRAS could be a
membrane protein. IRAS now is known to be anchored to the intracellular surface
of the plasma membrane by virtue of possessing a phosphoinositide-3-
phosphate (PI3) binding domain in its N-terminus region, known as the PX
domain (Worby and Dixon, 2002;Xu et al., 2001). In addition, IRAS binds
selectively to the α5 subunit of the fibronectin receptor (Alahari et al.,
2000;Alahari, 2003;Alahari et al., 2004;Lim and Hong, 2004) which also anchors it to the inner surface of the plasma membrane.
IRAS possesses several domains involved in protein-protein interactions, including leucine rich repeats and a coiled-coil domain in addition to the PX domain and integrin binding region. This protein is thus distinct from classic cell
72 surface receptors. We hypothesized that binding of low molecular weight ligands such as imidazolines to IRAS might trigger cell signaling events. To test this hypothesis, we generated antisense to IRAS and used it to transfect PC12 cells.
Most of these findings were initially reported as an abstract (Sun et al., 2004).
73 MATERIALS AND METHODS
Materials
Rabbit anti-human antibody directed against IRAS was provided by Dr.
John Piletz at Jackson State University (Jackson, MS) (Piletz et al., 2000). We used antibody #1209, which is directed against amino acids 783 to 809 of human
IRAS, located within the region that interacts with alpha5-integrin and rac (Piletz et al., 2000) (Figure 6). This region is 83% homologous with rat IRAS (Figure 6).
Antibody #1209 has already been shown to recognize the rat IRAS gene product in PC12 cells (Piletz et al., 1999;Piletz et al., 2000).
Oligo-DNAs were synthesized by Sigma Genosys (The Woodlands, TX).
Anti-phospho ERK1/2 antibody was purchased from Cell-Signaling (Beverly, MA).
Anti-ERK1/2 was purchased from Sigma-Aldrich (St. Louis, MO). Anti-actin was purchased from Chemicon (Temecula, CA). Anti-rabbit and anti-mouse were both purchased from Santa Cruz (Santa Cruz, CA). Efaroxan and naphazoline were obtained from Sigma-Aldrich (St. Louis, MO). Moxonidine was provided by
Solvay Pharmaceutical (Hannover, Germany). [125I]p-Iodoclonidine was obtained from Perkin-Elmer (Boston, MA) and stored at -20°C in ethanol.
Antisense Synthesis
The rat IRAS sequence (Figure 6) was determined from and is highly homologous with the human sequence reported previously (Piletz et al., 2000). A
20-nucleotide sequence covering the initiating ATG codon in the rat sequence was selected as the target for the antisense. This sequence shows perfect homology with human IRAS. The antisense sequence was the following 20-mer:
74 5’-GCCCTGCGTGTCATCAAAAA-3’. The complimentary 20-mer sense
sequence, 5’-TTTTTGATGACACGCAGGGC-3’, was used as a negative control.
As an additional negative control, we used a non-relevant oligonucleotide which
targets angiotensin converting enzyme, with the sequence 5’-
CCACCGTGTTCTTCGACATTG-3’. This sequence was selected as a control
because angiotensin converting enzyme is not expressed in rat PC12 cells
(Chang et al, unpublished observations). The 3’-end of each oligonucleotide was
modified to a phosphothiolate for improved stability.
Cell Culture and Transfection
PC12 cells were grown in either 70cm2 flasks or 6-well plates in RPMI
1640 supplemented with 10% horse serum and 5% fetal bovine serum at 37°C
and in an atmosphere of humidified 5% CO2. Media were changed every other
day, and cells were grown to 70% confluence before being transfected with
6µg/ml oligo-DNA using the GenePorterTM 2 transfection kit (Gene Therapy
Systems, San Diego, CA) according to the manufacturer’s instructions. Cells
were incubated with antisense or sense oligo DNA for 48h before being
harvested for binding assay or being switched into serum-free medium for tests of ERK1/2 activation.
Plasma Membrane Isolation and [125I]p-Iodoclonidine Binding Assays
Harvested cells were centrifuged at 1,000 x g for 5 min and flash frozen.
Cell pellets were homogenized and membrane fractions isolated as previously
described (Separovic et al., 1996). Radioligand binding assays with [125I]p-
iodoclonidine were performed as previously described (Ernsberger et al., 1995).
75 Briefly, membranes were slowly thawed and resuspended in Tris-HEPES buffer
(5.0 mM; pH 7.7 at 25°C, containing 0.5 mM EDTA, 0.5 mM EGTA, and 0.5 mM
MgCl2) at a concentration of about 0.1 mg protein/ml. Incubations were initiated
by the addition of membrane and were carried out for 30 min at 22°C.
Nonspecific binding was defined in the presence of 10 µM naphazoline.
Incubations were stopped by vacuum filtration using a cell harvester (Brandel
M18) connected to a vacuum pump rated at 120 l/min (Edwards EM8). Samples
were filtered over sheets of glass fiber filter paper (Schleicher & Schuell #34),
which were preincubated for 4 h at 4°C in 0.03% polylethyleneimine to reduce
nonspecific binding. Each sample well was washed four times with 4 ml ice-cold
assay buffer, an operation completed in less than 12 sec. Individual filters were
placed in glass 12X75mm tubes, and counted at 80% efficiency (Packard
Instruments, Downers Grove, IL). At around the Kd concentration of radioligand, total binding was 5900 ± 540 dpm and nonspecific was 3200 ± 350 dpm, yielding specific counts of 2700 ± 230 dpm.
Ruthenium Red Staining
This procedure was slightly modified from a previous method (Chen and
MacLennan, 1994). Briefly, blots were washed twice for 10 min each in 50mM
Tris buffer (pH 8.0) containing 60 mM KCl and 5mM MgCl2 before being
incubated with 30 µM ruthenium red in the same buffer for 15 min. The blots
were then washed 3 more times for 15 min each before being scanned with a
Microtek ScanMaker 4700 scanner for quantitation.
Assay of ERK Activation
76 PC12 cells transfected with antisense or sense oligo-DNA were
preincubated with serum-free RPMI1640 medium for 2h just before moxonidine
application. Moxonidine or vehicle (DMSO; final concentration 0.2%) was applied
for 90min. After treatment, media were removed and cells were washed with ice
cold phosphate buffered saline (pH 7.4) twice. Cells were lysed by application of
0.3ml of lysis buffer as previously described (Edwards et al., 2001). The 6-well
plates were then frozen at –75oC for 30 min. Thawed cell fragments then were
scraped from the plate, followed by centrifugation at 12,000g for 15min at 4oC to
remove insoluble debris. Total protein concentration of each resulting
supernatant was measured with BCA Protein Assay Kit from Pierce (Rockford, IL,
USA). Samples were divided into aliquots containing 25µg of total protein in double strength Laemmli buffer in a total volume of 50µl. Aliquots were subjected to 10% SDS-PAGE gel and electrophoretically transferred to a nitrocellulose membrane for immunodetection. The blots were incubated overnight with anti- active ERK1/2 (Cell Signaling Inc) at a dilution of 1:10,000 in 5% BSA, and anti-
ERK1/2 (Sigma) antibodies (1:1,000) in 5% BSA at 4oC. IRAS expression was
determined with anti-IRAS 1209 polyclonal antibody at a dilution of 1:1,000
(courtesy of J.E. Piletz). Total actin was measured as loading control by using
monoclonal anti-actin at 1:10,000 dilution (Chemicon, CA). After repeated
washes in 5% reconstituted skim milk, blots were incubated with horseradish
peroxidase-conjugated donkey-anti-rabbit or donkey-anti-mouse (Santa-Cruz, CA)
in 5% milk at room temperature. Blots were visualized with enhanced
chemiluminescence (Pierce, Buckinghamshire, UK), digitized (ScanMaker 4700)
77 and quantified by densitometry as grey scale optical density multiplied by pixels
(UN-Scan-It, Silk Scientific, Orem, UT). Data were expressed as the ratio of the densitometric signal for phospho-ERK1/2 to that of total ERK1/2. Data were further normalized to the value of controls or baseline. The densitometic signal for IRAS immunoreactivity was calculated as a ratio to that for total actin.
Statistical Analysis All values are given as means ± standard error.
Group means were compared by using t-tests for simple comparisons or the
Newman-Keuls test for post-hoc analyses. All analyses were carried out using
Prism 4.0 (GraphPad Software, San Diego, CA).
78 RESULTS
Rat IRAS amino acid sequence. The predicted rat IRAS sequence was
derived from an annotated genomic sequence (XM_240330) and can be divided
into a number of distinct domains (Figure 6A). The overall homology of the amino
acid sequence (Figure 6B) is 83% relative to human IRAS (Piletz et al., 2000)
and 88% relative to the corrected full-length mouse sequence (Lim and Hong,
2004). The overall homology of the nucleotide sequence is 92% relative to
human and 93% relative to mouse. The mouse IRAS gene product first described
as “nischarin” (Alahari et al., 2000) is truncated and missing the initial PX domain
which may serve as a membrane anchor (Lim and Hong, 2004).
Among the possible domains of rat IRAS, the PX domain has the highest
homology to human and mouse (85% and 91% respectively). The ruthenium red
binding domain, which is also predicted to be the imidazoline ligand binding area,
has 81% and 92% homology to human and mouse respectively. The proline-rich
region in human IRAS (amino acid 1043-1107) has no corresponding sequence
in rat IRAS. Because binding and signaling properties of human, mouse and rat
I1R are similar, the function of the proline-rich region is not clear.
The 20 mer sequence targeted by antisense is completely homologous to
human IRAS and 80% homologous to the mouse. The sequence is also found in
EST106158, whose expression has been confirmed in rat PC12 cells (Dontenwill et al. 2003a). Thus, the targeted region has been confirmed as an expressed sequence.
79 Effect of antisense oligo-DNA on the specific I1-imidazoline binding
in plasma membrane fractions from PC12 cells. Saturation binding assays
confirmed the presence of a high density of specific high-affinity binding sites for
125 [ I]p-iodoclonidine on plasma membrane fractions of PC12 cells (Bmax = 691 ±
29 fmol/mg; Kd = 0.39 ± 0.05 nM). As previously reported, specific binding was inhibited by imidazolines such as oxymetazoline, naphazoline and moxonidine
but not by epinephrine and other non-imidazoline adrenergic ligands (data not
shown). Treatment with antisense oligo-DNA for 48h significantly reduced the
density of binding sites (Bmax = 400 ± 16 fmol/mg) assayed in parallel with control
cells transfected with either sense or non-relevant oligo-DNA (Figure 7A). The
125 binding affinity for [ I]p-iodoclonidine remained unchanged (Kd = 0.30 ± 0.04
nM). Thus, antisense treatment reduced the total specific I1-imidazoline binding
sites on the plasma membrane of PC12 cells, while not affecting binding affinity.
In the Scatchard plot of the binding data (Figure 7B), antisense and control data
points generated straight lines, consistent with nonlinear curve fitting analysis
showing that there was only one class of high-affinity I1-imidazoline binding
existing in either antisense treated or control cell membranes.
Inhibition of IRAS protein expression in plasma membrane of PC12 cells by antisense treatment. To test whether the decreased I1-imidazoline
binding capacity from antisense treatment was accompanied with a reduction in
the expression of IRAS protein in PC12 cells, PC12 cells with either antisense or
sense control oligo-DNA treatment were harvested and total cell lysates were
subjected to Western blot analysis for IRAS. The IRAS protein was detected with
80 polyclonal antibody raised against IRAS (Ab1209) (Piletz et al., 2000) and
normalized to the immunoreactivity for total actin detected after stripping each
blot. After 48 h exposure to antisense, total relative expression of IRAS in PC12
cells was significantly reduced by about 50% compared to control cells treated
with sense oligo-DNA according to densitometric scans (Figure 8A) and as can
be seen on a representative blot (Figure 8B), in close agreement with the
reduction in total I1-imidazoline specific binding capacity. The protein bands
immunoreactive to IRAS polyclonal antibody were mainly detected at 85kDa
(Figure 8B), consistent with previous findings in human brain, human platelets
and rat tissues (Piletz et al., 2000;Dontenwill et al., 2003a;Zhu et al., 2003).
Ruthenium staining of IRAS. A ruthenium binding region is present in
the coiled-coil domain of IRAS (Figure 6A). Furthermore, ruthenium red
competes for specific [125I]p-iodoclonidine binding in PC12 cell membranes
(Piletz et al., 2000). Moreover, an 85-kDa candidate IR1 protein band has been
confirmed on blots of rat brain membranes stained with ruthenium red. For these
reasons, we compared the ruthenium staining density between antisense and
sense followed by total actin western-blot as loading control. Many bands and
molecular weight regions were positively stained (blots not shown). The
normalized ruthenium red staining density within the 80 to 90 KDa region from
antisense transfected PC12 cells was 74.9 ± 0.3% of that from sense control. In contrast, the region below 80 kDa was 92.3 ± 1.4% of the sense control and the region above 90 kDa was 98.6 ± 4.2% of the sense control. Thus, ruthenium red
81 staining was reduced in the area around 85 kDa without an overall decrease in
ruthenium red staining proteins.
Effect of antisense treatment on ERK1/2 activation in response to
moxonidine. It has been previously demonstrated that moxonidine leads to
ERK1/2 activation in PC12 cells through activation of specific I1-imidazoline
receptors. (Edwards et al., 2001). Thus, to investigate the effects of antisense
treatment on the I1-imidazoline receptor signaling, we tested moxonidine induced
ERK1/2 activation. Since peak ERK1/2 activation occurs at 90 min (Edwards and
Ernsberger, 2003), this period of incubation was selected. Moxonidine
concentrations from 0.1 nM to 1.0 µM, as well as vehicle, were added to PC12
cells preincubated for 30 min in serum-free medium (Figure 9). The dually
phosphorylated active form of ERK-1/2 and total ERK1/2 were detected with
selective antibodies and the ratio optical densities was determined. This ratio
was normalized to the vehicle condition (0 dose). In both groups, the ERK1/2
activation at 90 min was elevated with the increasing concentrations of
moxonidine. The extent and concentration-dependency of ERK1/2 activation are
in close agreement with previous results from this laboratory (Edwards et al.,
2001) and others (Zhang et al., 2001). In control PC12 cells transfected with
sense oligo-DNA the estimated peak activation of ERK1/2 (Emax) reached 197 ±
33% over vehicle control, whereas in cells transfected with antisense oligo-DNA
the level of activation only reached 44 ± 16% above basal level (Figure 9A). The
EC50 values for sense and antisense curves in the same figure were in
82 approximate agreement (1.3 ± 0.4 nM for control vs 0.24 ± 0.51 nM for antisense treated), similar to the unchanged affinities noted in the binding assay (Figure 7).
Basal ERK1/2 phosphorylation level with or without antisense treatment. Conceivably, decreased IRAS protein expression from antisense treatment in PC12 cells might lead to a nonspecific loss of ERK1/2 expression or function in the basal state. These nonspecific changes might result from the antisense treatment or the transfection procedure itself. To test this possibility, we compared the basal ERK1/2 phosphorylation level of PC12 cells with 48 h incubation with either antisense or vehicle control. Western-blotting results showed there was no significant difference in basal ERK1/2 activation level induced by antisense treatment (Figure 10).
Effect of IRAS antisense on ERK activation by insulin. The decreased moxonidine induced ERK activation in antisense transfected PC12 cells could be due to either decreased IRAS expression or impaired ERK signaling machinery.
Thus we compared the insulin induced ERK activation between antisense transfected and untreated control PC12 cells. PC12 cells were either incubated with antisense for 48 h as in other experiments or without any treatment. Cells were exposed to insulin or moxonidine (both at 100 nM) for 5 min, harvested and lysed for assay of ERK activation. Control cells showed marked activation of ERK by insulin (Figure 11). The effect of moxonidine was modest at this 5 min timepoint (ratio of 1.65 ± 0.25 relative to basal, P < 0.05 by t-test), comparable to our previously reported data on the timecourse of ERK activation by moxonidine
83 showing a small initial response at 5 min building to a peak activation at 90 min
(Edwards et al., 2001).
Antisense treatment did not alter ERK activation by insulin (ratio relative to basal: 7.3 ± 0.64 in the antisense condition vs. 8.2 ± 0.94 in untreated control cells). Moxonidine failed to induce ERK activation in cells transfected with IRAS antisense (a ratio of 1.08 ± 0.11). Thus, the response to this imidazoline agonist was abolished at 5 min as shown in Figure 11.These results suggested that the impairment of ERK signaling machinery in PC12 cells by IRAS antisense transfection does not reflect a global depression of ERK responsiveness, or nonspecific effects on cell signaling by the transfection procedure itself, as a completely unchanged response to insulin was observed.
84 DISCUSSION
Endogenously expressed I1R on native PC12 cell membranes are
substantially down-regulated by treatment with antisense oligonuceotides
directed against the IRAS gene. Thus, native I1R are dependent on the
production of IRAS mRNA. Furthermore, IRAS immunoreactivity was decreased
in parallel with the decline of I1R density. A cell signaling event downstream of
I1R stimulation, namely the activation of ERK, was dramatically reduced after
antisense treatment. The reduction in the cell signaling response was nearly
twice as great as the reduction in binding sites, as expected for an event several
steps downstream of the receptor. We have shown that receptor occupancy
leads successively to increases in phosphatidylcholine hydrolysis, diacylgyceride
mass, arachidonic acid release and protein kinase C activity, all of which are required for ERK activation (Edwards et al., 2001;Separovic et al.,
1997;Separovic et al., 1996). Phosphatidylcholine hydrolysis is triggered within
15s, while PKC activation and arachidonic acid release require 5 to 15 min, and
ERK and JNK activation do not peak until 90 min. The fall in blood pressure triggered by imidazoline agents also requires several minutes to develop, and
persists for up to 2h (Buccafusco et al., 1995;Haxhiu et al., 1994).
Prior to the present study, the evidence linking IRAS to I1R was mainly
indirect. First, the gene was isolated using antibodies to an affinity-purified
imidazoline binding protein fraction (Ivanov et al., 1998). Second, Northern blot
estimates of IRAS mRNA expression were correlated with radioligand binding
density in native tissues (Piletz et al., 1999). Third, transfection of IRAS cDNA
85 into Chinese Hamster Ovary (CHO) cells leads to the appearance of high affinity
I1-like binding sites without appearance of α2AR sites or I2 sites (Piletz et al.,
2000). The sites encoded by IRAS cDNA were appropriately labeled in
membrane fractions with [125I]p-iodoclonidine, and competition for binding was
observed for selective I1R ligands such as moxonidine and naphazoline but not
for adrenergic compounds. Furthermore, levels of I1 binding sites and IRAS
immunoreactivity covaried in the transfected CHO cells. One limitation of these
prior studies is that the maximum density of transfected receptors was small,
comparable to native expression in other cells, and not of the order seen with transfected G-protein coupled receptors. Fourth, PC12 cells stably transfected with human IRAS show elevated levels of IRAS immunoreactivity and I1R Bmax for
[125I]p-iodoclonidine binding (Dontenwill et al., 2003b;Dontenwill et al., 2003b).
Fifth, these IRAS-transfected PC12 cells show an attenuation of the normal ERK
activation in response to NGF (Piletz et al., 2003), similar to moxonidine-treated
native PC12 cells(Edwards and Ernsberger, 2003). In the latter case, inactivation
of ERK appears to result from rapid induction of the dual specificity phosphatase
MKP2. Finally, while this manuscript was in preparation it was reported that
antisense directed against the mouse Nischarin gene when transfected in PC12
cells decreased immunoreactivity to IRAS/Nischarin antibodies in both Western
blots and by immunofluorescence (Zhang and Abdel-Rahman, 2006). The
activation of ERK by the I1R agonist rilmenidine was also decreased by a factor
of four. The present report confirms and extends these findings.
86 In agreement with previous studies, IRAS immunoreactivity was located
primarily in an 85 kDa band, despite the predicted 167 kDa mass of the IRAS
gene product (Dontenwill et al., 2003b;Zhu et al., 2003;Sano et al., 2002;Ivanov
et al., 1998;Ivanov et al., 1998). Very little immunoreactivity can be detected at
167 kDa, even in the presence of extensive protease inhibitor cocktails. The
cellular processing of the IRAS protein is not understood, although different
forms may exist in the endoplasmic reticulum and the plasma membrane (Lim
and Hong, 2004).
We previously suggested (Piletz et al., 2000), based on a dye interference
assay, that the I1 binding domain might be a glu/asp-rich region near the middle
of IRAS. This domain encodes a ruthenium red binding site, and this dye labels
125 IRBP and competes with [ I]p-iodoclonidine binding to the native I1R (Piletz et
al., 2000). This putative imidazoline binding domain lies within a coiled-coil
domain motif, which has been shown to mediate protein-protein interactions and homodimerization of IRAS (Lim and Hong, 2004). In the present study, we found widespread ruthenium red staining at multiple molecular weight regions, but only in the 80-90 kDa region was staining density reduced. This suggests that the 85- kDa active fragment of IRAS contains the ruthenium red binding site.
The coiled-coil domain lies immediately upstream of a peptide region of
IRAS that has since been found to specifically bind to a membrane spanning region of the α5 subunit of the fibronectin receptor, which might participate in anchoring IRAS to the plasma membrane. Indeed, the imidazoline binding site may only be formed when IRAS is complexed to the fibronectin receptor or other
87 partner proteins. This might explain why transfection of human IRAS cDNA into
three cell lines (COS-7, HEK293 and Sf9 cells) has proven unable to induce I1R, even though the 167 kDa version of the human protein was evidently synthesized
(Sano et al., 2002;Piletz et al., 2000).
In this study, we demonstrated that antisense oligonucleotides impair
protein expression of IRAS in PC12 cells, and this is accompanied with
significant loss of 42% of specific binding sites to imidazolines on plasma
membrane fractions. While it may appear modest, this reduction is greater than
that achieved in nearly two-thirds of published reports of antisense knockdown of
neuronal receptors (Van Oekelen et al., 2003). Through Western-blotting, we
then found that the expression of IRAS protein in antisense treated cells was also
reduced by about 50% compared to control. This apparent agreement between
attenuated protein expression level and impaired binding capacity suggested a
possible role of IRAS in the I1R binding. We had further hypothesized that the
attenuated binding capacity caused by antisense should also lead to impaired cell signaling associated with I1-imidazoline receptors. As expected, the peak
activation of ERK1/2 induced by moxonidine in PC12 cells was significantly
lowered from about 300% to 150% following antisense treatment compared to
sense control. The basal level of ERK1/2 activation was unchanged. This further
suggested that IRAS is required not only for I1-imidazoline binding but also for
cell signaling events associated with I1R. As a positive control we tested the
ability of insulin to activate ERK in PC12 cells with or without antisense
transfection, and the results suggested that overall ERK signaling was not
88 significantly altered by transfection. In addition, the ERK activation induced by
100 nM moxonidine at 5 min in control PC12 cells (1.65 ± 0.25 fold of basal) are
fairly consistent with previously published data from our lab at 10 min of
treatment (1.28 ± 0.07 fold of basal at 10 min) (Edwards et al., 2001). The activation of ERK by I1R agonists peaks at around 90 min of exposure (Edwards
et al., 2001;Zhang et al., 2001). This delayed response may reflect the presence of multiple intermediate signaling steps, which include hydrolysis of phosphatidylcholine to liberate diacylglyceride and subsequent activation of protein kinases C, both of which are required for I1R activation of ERK (Edwards
et al., 2001). This cascade of signaling events may also explain how a two-fold reduction in I1R binding sites and immunoreactivity can translate into a four-fold reduction in the maximum activation of ERK.
Although IRAS is large protein by itself (167kDa), the I1R may consist of
multiple proteins in a complex with this scaffolding protein. The fact that over-
expression of IRAS leads to elevated specific I1-sites binding in some cell lines
(CHO and PC12) while not in some others (COS-7, HEK293, and Sf9 cells)
(Piletz et al., 1999;Piletz et al., 2000;Piletz et al., 2003) suggested that IRAS
itself may not function alone in mediating ligand binding and cell signaling events.
The present results indicate that the expression level of IRAS is critical for
maintaining the proper function of I1-imidazoline receptors, both the binding of
imidazoline ligands and the initiation of cell signaling cascades. Thus it is strongly
suggested that IRAS may be the I1-imidazoline receptor itself, or at least a functional subunit of it.
89 Footnotes
*This work was supported by HL44514 from the National Institutes of
Health.
90 Figure Legends
Figure 6. Domain map of the rat IRAS gene and sequence comparison to
human and mouse. Panel A: The major domains and motifs of the IRAS protein
as described in the literature, with amino acid numbers from the rat sequence.
Abbreviations (from N-terminus): PX = phox homology domain (Lim and Hong,
2004), PI3P = phosphatidyinositol 3-phosphate, LRR = leucine rich repeat,
(Ceulemans et al., 1999). The LRR-cap was identified in IRAS by Ceulemans
(Ceulemans et al., 1999). The interaction of rac and PAK with IRAS/Nischarin
was shown by Alahari (Alahari, 2003). The binding of insulin receptor substrate
(IRS) proteins to the C-terminal region IRAS was also recently reported (Sano et al., 2002). Panel B: Comparison of aligned amino acid sequences of rat, human and mouse IRAS proteins. Conversed sequences are shaded, conservative substitutions are in italics and non-conservative substitutions are in plain font.
The rat IRAS amino acid sequence is predicted Rattus norvegicus Nischarin
(XM_240330). The human IRAS sequence is adopted from I1R candidate protein
(AAC33104) (Piletz et al. 2000). The mouse IRAS sequence is an integrated
combination of mouse Nischarin sequence (AF315344) (Lim and Hong 2004)
and mouse mKIAA0975 protein (BAC65694). The alignment was carried out with
the CLUSTAL W (1.83) multiple sequence alignment program.
125 Figure 7. Saturation kinetics of [ I]-p-iodoclonidine binding to I1R in PC12
cells transfected with IRAS antisense compared with control. PC12 cells were
transfected for 48 h at a dose of 6µg/ml with IRAS antisense or control cDNA,
either sense or non-relevant sequence directed against a gene not expressed in
91 PC12 cells (angiotensin converting enzyme). The two controls gave equivalent
results and so the data were combined. Transfected cells were harvested and
plasma membranes were isolated for saturation binding assays. (A) Plasma
membranes were incubated with a series of eight concentrations of [125I]-p-
iodoclonidine ranging from 30 pM to 1.7 nM in the presence and absence of 10
µM naphazoline to define nonspecific binding. (B) Scatchard transformation of
the data. Scatchard plots were used for display purposes and not for data
analysis. Values represent the mean ± standard error for six experiments. Curves
were fit by nonlinear analysis with GraphPad Prism (SanDiego, CA). Nonlinear
curve-fitting was used to obtain Bmax and Kd values and the standard error of the
estimate (S.E.E.).
Figure 8. Effect of antisense treatment on IRAS protein expression. PC12 cells were transfected with antisense or sense oligo-nucleotides for 48 h before being harvested. Lysates were immunoblotted for IRAS protein using a polyclonal antibody. Blots were stripped and reprobed for actin. Panel A: Data are expressed as the ratio of optical density of IRAS to total actin, and were normalized to control. The data represent six experiments in each group. Panel B:
Representative blots of IRAS protein (top panel) expression level in PC12 cells with or without antisense transfection, relative to actin (bottom panel). Each lane represents a different flask of PC12 cells. The blot shows three control and three antisense treated flasks of cells.
Figure 9. IRAS antisense inhibits I1R signaling. PC12 cells were
transfected with IRAS antisense or control for 48h before being incubated in
92 serum-free medium for 2h. Then various concentrations of moxonidine or vehicle
only were applied for 90 min. Cells were lysed and total cell lysates were
subjected to Western-blot for ERK activation analysis. Panel A: The relative ERK
activation level was expressed as the ratio of optical density of phosphorylated
ERK1/2 to total ERK1/2. All data were normalized to vehicle treated samples.
Each group included at least five experiments. Panel B: Representative blots showing phosphorylated ERK-1 and ERK-2 immunoreactivity (upper and lower bands, respectively) and total ERK-1 and ERK-2 immunoreactivity following stripping and re-probing of the same blots.
Figure 10. Effect of antisense treatment on basal activation of ERK1/2.
PC12 cells were transfected with antisense or control oligo-nucleotides for 48 h before being harvested. Detergent lysates of the cells were immunoblotted for phosphorylated and total ERK1/2. The results were expressed as the ratio of optical density of phosphorylated to total ERK1/2. The data represent eight experiments in each group. The error bar for the control indicates the standard error of the optical density values.
Figure 11. Insulin induced ERK activation in PC12 cells, and the early phase of moxonidine activation. Flasks of PC12 cells were either transfected with antisense oligo-nucleotides or left undisturbed for 48 h prior to testing. All cells were then treated with 100 nM insulin, 100 nM moxonidine or vehicle for 5 min respectively before being harvested for measurement of ERK activation as described for Figures 3 and 4. The dotted line represents basal (unstimulated) level. The standard error bars in the control condition reflect the standard error in
93 the absorbance ratio in controls across experiments. The data represent six experiments in each group.
94
Mouse PAQYPSE Human PAQYPSE Rat PAQYPSE Mouse LTSAEAPAAAEAPAAAEAPAAAEAPAAAEAPAAAEAPAAAEAPAPAEAPAAAEAPAA Human ---GGSPQGSFADGQPAERRASNDQRPQEVPAEALAPAPVEVPAPAPAAASASGPAK Rat ------SARPR Mouse SCTQPRGAFADGHVLELLVGYRFVTAIFVLPHEKFHFLRVYNQLRASL Human SCTQPRGAFADGHVLELLVGYRFVTAIFVLPHEKFHFLRVYNQLRASL Rat SCTQPRGAFADGHVLELLVGYRFVTAIFVLPHEKFHFLRVYNQLRASL Mouse LLTFYKVAGGSQERSQGCFPVYLVYSDKRMVQTPAGDYSGNIEWASCTLCSAVRRSCCAPSEAVKSAAIPYWLLLTSQHLNVIKADFNPMPNRGTHNCRNRNSFKLSRVPLSTVLLDPTR Human LLTFYKVAGGCQERSQGCFPVYLVYSDKRMVQTAAGDYSGNIEWASCTLCSAVRRSCCAPSEAVKSAAIPYWLLLTPQHLNVIKADFNPMPNRGTHNCRNRNSFKLSRVPLSTVLLDPTR Rat LLTFYKVAGGSQERSQGCFPVYLVYSDKRMVQTAAGDYSGNIEWASCTLCSAVRRSCCAPSEAVKSAAIPYWLLLTSQHLNVIKADFNPMPSRGTHNCRNRNSFKLSRVPLSTVLLDPTR 95 Mouse VQGSIRQFAACLVLTDFGIAVFEIPHQESRGSSQHILSSLRFVFCFPHGDLTEFGFLMPELCLVLKVRHSENTLFIISDAANLHEFHADLRSCFAPQHMAMLCSPILYGSHT Human VQGSIRQFAACLVLTDFGIAVFEIPHQESRGSSQHILSSLRFVFCFPHGDLTEFGFLMPELCLVLKVRHSENTLFIISDAANLHEFHADLRSCFAPQHMAMLCSPILYGSHT Rat VQGSIRQFAACLVLTDFGIAVFEIPHQESRGSSQHILSSLRFVFCFPHGDLTEFGFLMPELCLVLKVRHSENTLFIISDAANLHEFHADLRSCFAPQHMAMLCSPILYGSHT Mouse TNQDFIQRLSTLIRQAIERQLPAWIEAANQREEA Human TNQDFIQRLSTLIRQAIERQLPAWIEAANQREEG Rat TNQDFIQRLSTLIRQAIERQLPAWIEAANQREEA Mouse LPPAPCIRPGGSPP-AAPASASLPQPILSNQGIMFVQEEALASSLSSTDSLPPEDHRPIA Human LSAAPCIRPSSSPPTVAPASASLPQPILSNQGIMFVQEEALASSLSSTDSLTPE-HQPIA Rat LPTAPCIRPSSSPPTAVPTSASLPQPILSNQGIMFVQEEALASSLSSTDSLPPD-DRPIA Mouse TLNLAGN Human TLNLAGN Rat TLNLAGN Mouse PTLAT Human PTLAT Rat PTLAT Mouse YEVNG Human YEING Rat YPALG Mouse MA Human MA Rat MA
A T A ATL AR- ATL M L M V I L SVRFSA SVRFSA SVRFSA T T G S T S R H R F L F A A S FGPEREAEPAKEARVVGSELVDTYTVY FGPEREAEPAKEARVVGSELVDTYTVY FGPEREAEPAKEARVVGSELVDTYTVY LIQ LIQ LIQ LESLSGLHKLYSLVN LESLSGLHKLYSLVN LERLSGLHKLYSLVN ALAEELFEKGEQLLGAGEVFAIRPLQLYA ALAEELFEKGEQLLGAGEVFAIGPLQLYA PLCLALT--GEQLLGAGEVFAIRPLQLYA S A S TSEENQIPSHLPVCPSL TSEENQIPSHLPACPSL TSEENQIPSHLPVCPSL T T A SMKEVLAPEASEFDEWEPEGT SMKEVLVPEASEFDEWEPEG- SMKEVLVPEASEFDEWEPEG- V L L DLRDNRIEQ DLRDNRIEQ DLRDNRIEQ Q R Q H H H I V I ARLRGRAII ASLRGSAII ARLRGRAII LD ME LD A T T V I V H Q H EV EV EV TLGGPVTA TLEGPVTA TLGGPVTA IQVTDGNHEWT IQVTDGSHEWT IQVTDGNHEWT GEQGEEE-EEEEEEEDVAENRYFEMGPPDAEEEEGSGQGEEDEE GEQGEEEDEEEEEEEDVAENRYFEMGPPDVEEEEGGGQGEEEEE GEQGEEE--EEEEEEDVAESRYFEMGPPDAEEEEGSGQGEEDEE K R K I V I SIGSLPCLE SIGSLPCLE SIGNLPCLE TEQLQQGKPTCASGDAKTDLGHILDFTCRLKYLKVSGTEGPFGTSNI TEQLQQGKPTCASGDAKTDLGHILDFTCRLKYLKVSGTEGPFGTSNI TEQLQQGKPTCASGDAKTDLGHILDFTCRLKYLKVSGTEGPFGTSNI D E D LFHNSIAEVENEELRHL LFHSSIAEVENEELRHL LFHSSIAEVENEELRHL I V V IPTWQALTTLDLSHNSICEIDESVKLIPKIEYLDLSHNG IPTWQALTTLDLSHNSISEIDESVKLIPKIEFLDLSHNG IPTWQALTTLDLSHNSISEIDESVKLIPKIEYLDLSHNG I V I RLT HVS HVA KHRYSDFHDLHEKLVAERKIDKSLLPPKKIIGKNSRSLVEKRE KHRYSDFHDLHEKLVAERKIDKNLLPPKKIIGKNSRSLVEKRE KHRYSDFHDLHEKLVAERKIDKTLLPPKKIIGKNSRSLVEKRE Q Q K DLKTVVI DLKTVVI DLKTVVI LLNNPLSIIPDYRTKVL LLNNPLSIIPDYRTKVL LLNNPLSIIPDYRTKVL S A A KNPSAKPRNQPAKSRASAEQRLQETPADAPAPAAVPPTASAPAPAEALAPDLAPVQAPGEDRG 1077 KTPGT------1019 KNP------1017 A T T EAPAAAEAP PAPAEASTS QAPMRVKLL L M L WSSVVFYQTPGLEVTACVLLS WSSVVFYQTPGLEVTACVLLS WSSVVFYQTPGLEVTACVLLS R Q Q ACS GCS ACS D D N S S S L L M S A S ES ES GS A A S QFGERASE QFGERASE QFGERASE AAEAPASA LVPEETPV IR-----A I I L P P P A A T GQVA GQAA GQVA I V I SD SD AE E E Q CLDDVATTEKELDTVEVLKAIQKAK CLDDTVTTEKELDTVEVLKAIQKAK CLDDVATTEKELDTVEVLKAIQKAK APAPNQAPAPARGPAPARGPAPAGGPAPAGGPAPAEALAQAEV 1197 APAP------PPAEA 1102 APVE------AQAEV 1042 DLRD DLRD DLRD D E D -EAERAEAGDDLEIIKLCLH77 E-DEEAEEERLALEWALGADEDFLLEHIRILKVLWCFLIH 717 EEDEEAEEERLALEWALGADEDFLLEHIRILKVLWCFLIH 717 E-DEEAEEERLALEWALGADEDFLLEHIRILKVLWCFLIH 713 S T T KAVYFILHDGLRRYFSEPLQ------1299 KAVYFVLHDGLRRYFSEPLQ------1204 KAVYFILHDGLRRYFSEPLQGPPCVQVVLAPCQAAFRL 1162 V V L L L V PGAVGGVSP PGAVGGASP PGAVGGVSP RVVDNLQHLYNLVHLDLSYNKLSSLEG LVVDNLQHLYNLVHLDLSYNKLSSLEG LVVDNLQHLYNLVHLDLSYNKLSSLEG K Q R EQLLPFDLSIFKSLHQVE EQLLPFDLSIFKSLHQVE EQLLPFDLSIFKSLHQVE R K K DLEVYLQTLLT DLEVYLQKLLA DLEVYLQTLLK D E D HAEPEVQVVPGSGQIIFLPFTCIGYTA 599 HAEPEVQVVPGSGQIIFLPFTCIGYTA 597 HAEPEVQVVPGSGQIIFLPFTCIGYTA 596 D E D T A T VKSKLSNTEKKAGEDFR 480 VKSKLSNPEKKGGEDSR 478 VKSKLSSTEKKVGEDFR 477 FPDV FPGV FPDV I I M SHCDAKH SHCDAKH SHCDAKH A T A PRVLAHFLHFH PRVLAHFLHFH PRVLAHFLHFH V L V HTKLGN HTKLGN HTKLGN I I V T S S RGLV RGLV RGLV LQEFLRQ 837 LQEFLRQ 837 LQEFLRQ 833 T A T V I V K240 SK SK 239 SK 238 K 360 K 358 K 357 L F L 120 119 120 957 957 3 95 Rat Mouse FDDTQGHDLMGSVTLDHFGE Human FDDVQGHDLMGSVTLDHFGE Rat FDDTQGHDLMGSVTLDHFGE Mouse RVKFTYPSEEE Human RVKFTYPSEEE RVKFTYPSEEE Mouse ------DFWHQKNTDYNNSPFH Human ------DFWHQKNTDYNNSPFH Rat DAVNRSLFSTTDFWHQKNTDYNNSPFH Figure 6A V I V GDLT GDLT GDLT Y F Y IVAQKMA TVAQKMA VVAQKMA M V M PGGPGRVGQGREVQWQVFVPSAESREKLISLLARQWEALCGRELPVELTG 1599 PGGPARASQGREVQWQVFVPSAESREKLISLLARQWEALCGRELPVELTG 1504 PGGPGRAGQGREVQWQVFVPSAESREKLISLLARQWEALCGRELPVELTG 1473 D E D PAKNPALSILLY PEKAPALSILLY PAKNPALSILLY V I I SQCFVLKLSDLQSVNVGLFDQ SQCFVLKLSDLQSVNVGLFDQ SQCFVLKLSDLQSVNVGLFDQ I V I QAFQVVTPHLGRGRGPLRPKTLLLTS QAFQVGMPPPGCCRGPLRPKTLLLTS QAFQVITPQLGRGRGPLRPKTLLLTS Y H Y FRLTGS FRLTGS FRLTGS S T S PTQVVTCLTRDSYLTHCFLQHLM PMQVVTCLTRDSYLTHCFLQHLM PTQVVTCLTRDSYLTHCFLQHLM A S A EIFLLDEDY EIFLLDEDC EIFLLDEDY I V I HYPLPEFAKEPPQRDRYRLDDGRRVRDLDRVLMGYYPYPQALTLV 1528 HYPLPEFAKEPPQRDRYRLDDGRRVRDLDRVLMGYQTYPQALTLV 1433 HYPLPEFAKEPPQRDRYRLDDGRRVRDLDRVLMGYNPYPQALTLV 1402 L V L VLSSLERTPSPEP VLSSLERTPSPEP VLSSLERTPSPEP V V I DKDFYSEFG DKDFYSEFG DKDFYSEFG D N D KNTGKMENYELIHSS 1408 KTTGKMENYELIHSS 1313 KNTGKMENYELIHSS 1282
96
Figure 6B
97 Figure 7
98 Figure 8A
99 Figure 8B
100 Figure 9A
101 Figure 9B
102 Figure 10
103 Figure 11 104 Manuscript 2:
Submitted to The Journal of Pharmacology and Experimental Therapeutics
Revisions pending
Title:
MARKED INSULIN RESISTANCE IN SHROB RAT ADIPOCYTES IS
AMELIORATED BY IN VIVO TREATMENT WITH MOXONIDINE
Abstract
The obese spontaneously hypertensive rat (SHROB) is a model of marked insulin resistance with normoglycemia. We sought to determine whether insulin resistance extends to adipocytes and the impact of an insulin sensitizing imidazoline, moxonidine. Gonadal adipocytes were isolated from SHROB and lean SHR littermates. In lean SHR adipocytes, Akt activation by insulin (100 nM) peaked at 3 min at 25-fold, whereas SHROB adipocytes showed only 4-fold activation. In dose response experiments, the maximum response (Emax) was
markedly reduced 18.8 ± 2.3 vs. 3.7 ± 0.8. Insulin sensitivity was also attenuated,
with higher concentrations required for responses (ED50: 3.5 ± 0.5 vs. 29 ± 3.8
nM). Glucose uptake as determined with [3H]-2-deoxyglucose was also less
responsive to insulin in SHROB relative to lean SHR. Moxonidine had little or no
effect when applied acutely in vitro, but adipocytes isolated from SHROB treated with moxonidine in vivo showed significantly improved responses to insulin, both in terms of Akt activation and facilitation of glucose uptake. Chronic but not acute
105 moxonidine treatment partially restores insulin sensitivity in SHROB adipocytes, suggesting an indirect action of this agent.
106 Introduction
SHROB rats are markedly insulin resistant, showing a greater than 20-fold elevation fasting insulin levels in the presence of normal fasting glucose
(Friedman et al., 1997b;Velliquette et al., 2005). At the cellular level, defects in insulin action have been noted in skeletal muscle and the liver, as indicated by reduced insulin receptor protein, and reduced insulin-induced tyrosine phosphorylation of the insulin receptor and its substrate protein IRS-1. In adipocytes and in skeletal muscle, the stimulation of glucose transport by insulin is impaired. However, insulin signaling in adipocytes in the SHROB model has not yet been characterized.
Akt (PKB) is a 57kD Ser/Thr kinase that plays a key role in the insulin induced
PI3K-Akt pathway (Hanada et al., 2004;Osaki et al., 2004). The binding of insulin to its receptors leads to the phosphorylation of PI3K, which then phosphorylates phosphatidylinositols at the 3 position. Then Akt is recruited to the inner side of plasma membrane due to the interaction between its PH domain and the PIP3 produced by PI3K. The Thr308 and Ser473 on Akt are then phosphorylated by
PDK1/2 and mTOR (Hresko and Mueckler, 2005). Once activated, Akt regulates many cellular functions related to insulin action (Hanada et al., 2004). As a key element in insulin signaling, Akt could be an efficient indicator for cellular insulin response. Here we measured the phosphorylation level of Akt in corresponding to insulin stimulation as an indicator of insulin sensitivity in isolated rat adipocytes.
Moxonidine is a centrally acting sympatholytic agent that was unexpectedly found to possess insulin sensitizing actions through largely unknown
107 mechanisms in humans (Starchina et al., 2005;Haenni and Lithell, 1999) and in
animal models (Ernsberger et al., 1996b;Ernsberger et al., 1999c;Henriksen et al.,
1997;Yakubu-Madus et al., 1999). Similar results have been obtained for another
imidazoline agonist, rilmenidine (Anichkov et al., 2005;Pele-Tounian et al.,
1998;Velliquette and Ernsberger, 2003a). Moxonidine is a selective agonist at I1-
imidazoline receptors, while also activating α2-adrenergic (Ernsberger et al.,
1993). Whereas both imidazoline and α2-adrenergic receptors contribute to
sympatholytic actions, only the imidazoline component improves glucose
metabolism (Velliquette and Ernsberger, 2003a;Velliquette and Ernsberger,
2003a). Most studies have been carried out with chronic treatment, but acute
improvements in glucose tolerance and insulin secretion can be detected under
blockade of α2-adrenergic receptors (Velliquette and Ernsberger, 2003a).
The cellular mechanisms for the insulin sensitizing action of moxonidine are
only partially unknown. Chronic treatment with moxonidine increases the
expression of insulin receptor and insulin receptor substrate-1 (IRS-1) in muscle
and liver (Ernsberger et al., 1999b). The impact of moxonidine treatment on other steps in the insulin signaling cascade or in other cell types are not known. In the present study, we focused on insulin signaling in adipocytes through the Akt
phosphorylation step, a possible site of insulin resistance in human diabetes type
2 (Karlsson et al., 2005).
108 Methods
Materials. Moxonidine (free base) was provided by Solvay Pharmaceuticals
(Hannover, Germany). Insulin and other chemicals were obtained from Sigma-
Aldrich, St.Louis, MO.
Animal Procedures. Adult male and female SHR and SHROB were used in
these studies. Animals were housed in pairs and were provided food (Teklad
8664; Madison, WI) and water ad libitum. Animals were on a 12:12-h light/dark
cycle (lights on from 7:00 AM to 7:00 PM) and were maintained at a constant
temperature of 21oC. These procedures were carried out with the approval of the
Case Western Reserve University Animal Care and Use Committee.
Adipocyte Isolation and Incubation. SHR or SHROB were fasted 18 h.
Anesthesia was induced with ether and maintained with isoflurane, and gonadal
fat tissue (4g) was quickly removed and rinsed. Adipose tissue was minced
before being digested for 1h with shaking in 20 ml Hanks’ buffer containing 1%
BSA and 25 mM HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) at
pH 7.4 with 1.0 mg/ml collagenase and 200 nM adenosinase. Then the
adipocytes were filtered through a 250 um nylon mesh (Sefar America) and
rinsed with phosphate buffered saline (pH 7.4) containing 1 mM sodium pyruvate,
0.1% BSA, 20 I.U penicillin and 20 µg/ml streptomycin. Cells (0.5 ml packed volume) were distributed to microcentrifuge tubes and pre-incubated at 37oC with
shaking for 1 h before insulin was applied. At the end of each experiment, tubes were placed in ice for 2 min, infranatants were removed, and 0.5 ml Laemmli
109 buffer with mercaptoethanol was added before boiling for 10 min. The aqueous phase was stored at -70°C for later Western-blot analysis.
Western Blot Procedure. Aliquots containing 20 :g protein were subjected to
SDS-PAGE on a 10% polyacrylamide gel and electrophoretically transferred to a nitrocellulose membrane for immunodetection. The blots were incubated overnight with anti-active Akt (phosphoserine473) at a dilution of 1:5,000 in 5%
BSA. After repeated washes in 5% reconstituted milk, blots were incubated with horseradish peroxidase-conjugated donkey-anti-rabbit in 5% milk at room temperature. Blots were visualized with enhanced chemiluminescence (Pierce,
Buckinghamshire, UK), digitized (ScanMaker 4700) and quantified by densitometry as grey scale multiplied by pixels (UN-Scan-It, Silk Scientific, Orem,
UT). All blots were completely stripped with a stripping buffer (Bio-Rad, Hercules,
CA) prior to another round of immunoblotting with anti-Akt antibody (1:1,000).
Data were expressed as the ratio of the densitometric signal for phospho-Akt to that of total Akt. Data were further normalized to the value of untreated controls or baseline.
Chronic Moxonidine Treatment. Moxonidine was dissolved in 20% (v/v) ethanol and mixed with powered rat chow (identical formulation to standard chow) before pelleting. SHROB were administrated moxonidine orally for 21 days at a dose of 4 mg/kg/day as described before (Velliquette and Ernsberger, 2003a).
During a 10d run-in period, body weight and food intake were monitored to assure accurate dosing 4 mg/kg/d. SHROB were provided food and water ad libitum. After 21 days of treatment, SHROB were fasted 18 h before sacrifice.
110 Given the 1h half life of moxonidine (He et al., 2000), it was expected that no
drug would be present at the time of tissue harvesting.
Acute Moxonidine Treatment and Insulin Stimulation. Moxonidine (100
nM) or 0.1% citric acid vehicle was applied into each tube of floating adipocytes
90 min before insulin stimulation. Insulin was applied into the cell incubations at a
concentration of 100 nM for various time lengths from 0 to 90 min (in time-course
experiments) or at various concentrations from 0.0 to 1.0 µM (in dose-response
experiments) for 10 min.
Glucose Uptake Assay Gonadal adipose tissue was taken from non-fasted
rats for each experiment at the same of day (9am). Following collagenase
digestion as described above, the adipocytes were filtered through mesh and
rinsed with wash buffer (phosphate buffered saline pH 7.4 containing 1mM
sodium pyruvate, 0.1% BSA, 25mM HEPES, 2.5mM MgCl2 and 2.5mM CaCl2) at
least 3 times. Cell suspensions were equally distributed into 20ml plastic tubes
shaking at 100 RPM in a 37oC water bath, each containing 106 cells in a total
volume of 1ml. Various concentrations of insulin (100nM, 10nM, 1nM or 0.1nM)
or vehicle (wash buffer) were applied to the cell suspensions for 30 min before
exposure to [3H]-2-deoxy-D-glucose. Nonspecific uptake was determined in the
presence of 10µM cytochalasin-B added 10 min before [3H]-2-deoxy-D-glucose to block glucose transport. At the end of incubation, 150µl of cell suspension from each tube was transferred into a 300µl elongated centrifuge tube on top of 70µl mineral oil. Uptake was initiated by adding 50µl 2.5mM [3H]-2-deoxy-D-glucose
containing a total of 0.33µCi of 3H labeling. In 3 min reaction was stopped by
111 spinning 10 s at 5,000g to separate cells from media. The cell fraction was
removed from the rest of medium by slicing through the oil layer, and the top
portion of the tube containing adipocytes was put into a scintillation vial with 4ml
of scintillation fluid (EcoScint A, National Diagnostics, Atlanta, GA). Vials were
counted in a scintillation counter for 5 min and specific glucose uptake was
defined as [3H]-2-deoxy-D-glucose incorporation minus incorporation in the
presence of cytochalasin-B. The rate of uptake was expressed as pmol per 105 cells per 3 min. Assays were carried out with triplicate cell aliquots and the results averaged.
Statistics. Results are presented as means ± standard error of the mean.
Dose response curves were analyzed by nonlinear curve fitting to a logistic equation (Prism 4.0, GraphPad Software, San Diego, CA). Groups were compared by one or two-way analysis of variance followed by Newman-Keuls tests.
112 Results:
SHROB adipocytes show profound insulin resistance.
A representative set of Western blots is shown in Figure 12. Immunoreactivity
to the phosphospecific antibody is shown in the top image, and the
immunoreactivity for total Akt protein is shown in the corresponding image below,
which was obtained form the same blot after stripping. Note the large sustained
increase in phosphospecific immunoreactivity, whereas total Akt is relatively
constant, indicating equal loading of the lanes. Adipocytes isolated from SHROB
show an equivalent amount of total Akt immunoreactivity, but the increase in
phosphospecific immunoreactivity elicited by insulin at each time point is
noticeably less.
The level of Akt phosphorylation in gonadal adipocytes from lean SHR
increased 26.6±3.5 fold compared to basal level in response to 3 min exposure
to 100 nM insulin (Figure 13). The level of Akt activation then slowly declined
over time and but was still elevated 14.2±5.4 fold at 90 min, the last time point
tested. In contrast, adipocytes from SHROB showed a greatly attenuated
response to 100 nM insulin stimulation, with a peak activation of only 3.9±1.4 fold
at 3 min. This small response was largely maintained at 90 min (3.3± 1.4 fold
activation).
This marked attenuation of Akt responses to insulin in adipocytes from
SHROB was confirmed in dose-response experiments (Figure 14). SHR and
SHROB adipocytes showed large differences in maximum response (Emax) at 10 min of insulin stimulation: 18.8 ± 2.3 vs. 3.7 ± 0.8 (expressed as fold increase).
113 Also consistent with reduced insulin sensitivity, the ED50 for insulin was lower in
SHR than in SHROB by nearly a full log unit (3.5 ± 0.5 vs. 29 ± 3.8 nM). Thus,
higher insulin concentrations were required to induce even a relatively small
response in SHROB rat adipocytes.
Chronic moxonidine treatment enhances insulin sensitivity in adipocytes from SHROB To investigate whether chronic moxonidine improves insulin sensitivity in the adipose tissue of SHROB, an animal model of insulin resistance, we treated
SHROB moxonidine orally at a dose of 4 mg/kg/day for 21 days. The adipocytes from treated SHROB showed significantly enhanced Akt response to 100 nM insulin (Figure 13): the peak fold activation at 3 min increased to 11.4±1.5 after
treatment, which was more than twice the response in control SHROB treated
with vehicle alone. The activation of Akt did not change over time between 3 and
90 min of treatment, similar to control SHROB adipocytes.
In dose-response experiments, after chronic moxonidine treatment the
maximum response (Emax) expressed as fold increase at 10 min rose to 7.5±0.6
compared to 3.7±0.8 from untreated SHROB (Figure 14). The ED50 of also fell to
2.6±0.6 nM from 29±3.8 nM. Thus, moxonidine treatment in vivo increased the
maximum response to insulin as well increasing the sensitivity to low
concentrations insulin.
Since all of the data are expressed as a ratio of phosphospecific to total
immunoreactivity, it is possible that changes in unstimulated basal Akt
phosphorylation may have contributed to the apparent effect of drug treatment.
To evaluate this possibility, we compared the ratio of phosphospecific to total Akt
114 immunoreactivity for adipocytes incubated in the absence of insulin. As shown in
Figure 15, adipocytes from SHR, SHROB and SHROB treated with moxonidine
all showed identical levels of unstimulated baseline Akt activation. Thus, the
insulin resistance of SHROB adipocytes and the insulin sensitizing action of
chronic in vivo moxonidine were not mediated by changes in baseline activation.
Insulin induced glucose uptake in SHR and SHROB adipocytes and the effect
of moxonidine treatment
To test whether changes in insulin sensitivity detected at the level of Akt
signaling correlate with functional changes relevant to metabolism, we assayed
adipocyte glucose uptake. First, we compared insulin responses in adipocytes
from SHR and SHROB without drug treatment (Figure 16A). SHR adipocytes
showed a basal glucose uptake of 1.4 ± 0.17 nmol 2-deoxy-D-glucose per 105 cells in 3 min, and SHROB adipocytes showed similar results with (1.4 ± 0.16 nmol per 105 cells in 3 min). Insulin induced a concentration dependent increase
in glucose uptake up until 10 nM, with 100 nM having less effect in the SHROB groups (indicated by descending dotted lines). Following stimulation for 30 min
with 10 nM insulin, glucose uptake in SHR adipocytes increased to 5.3 ± 0.60
nmol/105 cells/3 min, while SHROB adipocytes could only reach 3.1 ± 0.28
nmol/105 cells/3 min (p < 0.05, t-test with Bonferonni correction). These data are consistent with insulin resistance in adipocytes of SHROB.
We then tested adipocytes from chronic moxonidine treated (21 days,
4mg/kg/day) SHROB and SHR to see whether the insulin sensitizing effect detected with Akt activation was reflected in glucose uptake. As expected and
115 consistent with Akt activation results, the glucose uptake in SHROB adipocytes
stimulated with 10 nM insulin was increased to 3.9 ± 0.41 nmol/105 cells/3 min by
21 day chronic oral administration of moxonidine while the basal glucose uptake
remained almost unchanged: 1.3 ± 0.16 nmol/105 cells/3 min. However, in contrast to Akt activation, the EC50 for insulin in adipocytes from the three groups
were quite close to each other: the LogEC50 was close to -9.0 for all three groups
(Figure 16A).
We also compared adipocyte glucose uptake between control SHR and SHR
treated with moxonidine for 21 d (Figure 16B). Results for the untreated control
SHR were similar to the previous experiment. In contrast to SHROB, SHR treated with moxonidine showed no difference relative to vehicle treated controls in either basal uptake or insulin (10 nM) stimulated uptake.
Effect of acute moxonidine treatment.
To test for a direct effect of moxonidine on Akt activation, we tested the effect of treatment with 100 nM moxonidine for 90 min. This concentration and duration of treatment has previously been shown to trigger multiple cell signaling events in other cell types (Edwards et al., 2001;Edwards and Ernsberger, 2003). Akt activation in adipocytes treated with moxonidine alone was normalized to the control group treated with vehicle alone. Both SHR and SHROB adipocytes
showed no change in basal Akt activation level in response 90 min acute in vitro
treatment with moxonidine (Figure 17). Thus, moxonidine does not directly
activate Akt.
116 Given that in vivo treatment with moxonidine facilitated insulin signaling, we sought to test the possible insulin sensitizing effect from acute treatment of adipocytes with moxonidine in vitro. Isolated adipocytes from SHR or SHROB were preincubated with or without 100nM moxonidine for 90 min before being exposed to 100 nM insulin for 0 to 90 min. A representative blot for SHR adipocytes is shown in Figure 18 and the averaged results for SHR and SHROB are shown in Figures 19A and 19B, respectively. Adipocytes from SHR with moxonidine preincubation showed a shifted insulin response time course with a slightly facilitated maximum fold response at 5 min: 36.5±7.4 vs. 29.0±5.3, but a significantly reduced insulin response at 90 min: 4.4±0.77 vs. 11.3±2.6 (Figure
19A). No significant differences were found between 10 min to 60 min. However, adipocytes from SHROB did not show any change in insulin response by acute moxonidine treatment. The peak fold activation of pretreated and untreated adipocytes from SHROB were very similar (7.8±1.0 vs. 8.1±1.5), and there were no significant difference at any other time points through 90 min (Figure 19B).
Similar negative results were obtained with concurrent treatment with moxonidine and insulin (data not shown) Thus, acute exposure to moxonidine in vitro does not reproduce the effects of chronic treatment in vivo, implying a mechanism not involving direct and immediate cellular actions.
117 Discussion
Previous studies have already demonstrated that SHROB express severe
insulin resistance and glucose intolerance at the level of the whole body and in
liver and skeletal muscle at the level of the insulin receptor and IRS-1
(Ernsberger, 1999;Friedman et al., 1997b). In the present study we isolated
adipocytes from SHROB and SHR to test insulin responses in the absence of the
physiological milieu. We assayed Akt activation, which is a key element in insulin
cell signaling downstream of both the insulin receptor and IRS-1. Since the
results were obtained from isolated cells, the insulin response was independent
of influence from blood hormones, and should reflect the function of adipose
tissue only. Results of this study showed that adipose tissue, one of the major
sites of insulin action, expresses profound insulin resistance with impaired Akt
activation resulting from insulin stimulation. Insulin induced uptake of glucose into
adipocytes was also impaired, and this cellular defect might contribute to the
whole body insulin resistance syndrome.
The insulin sensitizing effect of chronic moxonidine treatment was observed
in both animal and human experiments. The insulin sensitizing effect of chronic
moxonidine treatment was observed in both animal and human experiments
(Jacob et al., 2004;Velliquette and Ernsberger, 2003b;Velliquette and Ernsberger,
2003b). Results of this study were quite consistent with those studies in that chronic moxonidine treatment shows significant beneficial effects on the sensitivity and responsiveness to insulin. As in other recent studies, moxonidine
118 had no effect on body weight or fat depot size in either SHR or SHROB, ruling
out any effect due to loss or gain of weight.
Not only chronic treatment but also acute application of moxonidine has been
shown to improve glucose tolerance in SHROB within 15 min of injection
(Velliquette and Ernsberger, 2003b). Thus we asked the question that whether
moxonidine works directly on the insulin responsive tissues or cells or whether it
might work through other organ systems such as the central nervous system or
the endocrine pancreas in a short time period such as 15 min. Although definite
conclusions can not yet be made, the results of the present study suggested that the effect of acute moxonidine treatment are not mediated by direction action on adipocytes. The action of acute moxonidine may be indirect. Moxonidine may alter adipocyte gene expression or act indirectly by affecting another organ, such as the liver.
There is a slight discrepancy in the results that in the moxonidine pretreatment experiments (Figure 16) the peak fold Akt activations in SHR and
SHROB were both higher than the corresponding 5 min levels in Figure 13. This appeared to be a consequence of the 90 min pretreatment period, which reduced the basal Akt activation as a result of additional exposure to serum-free medium.
Also, while cell signaling responses to insulin were dose dependent up to 1.0 µM, the activation of glucose transport fell off at 100 nM insulin for some experimental groups. This might reflect desensitization of insulin signaling pathways in vitro from prolonged exposure to higher concentrations. Differences in the
119 susceptibility of SHR and SHROB adipocytes to insulin desensitization should be examined in future studies.
In conclusion, insulin resistance in SHROB adipocytes persists upon isolation and challenge with insulin in vitro. The phosphorylation of Akt is a step in the insulin signaling cascade that shows much stronger evidence of insulin resistance than tyrosine phosphorylation of the insulin receptor or IRS-1 we have previously described (Ernsberger et al., 1999c;Friedman et al., 1997b;Velliquette et al., 2005). Thus, the SHROB adipocyte is a potential in vitor model of insulin resistance. In vivo treatment with moxonidine for 21d increased the sensitivity of
SHROB adipocytes to insulin, despite withdrawal of treatment 18h prior to tissue harvesting and extensive washing during the isolation process. This suggests a durable effect of the treatment on adipocytes such as a change in gene expression. This conclusion is supported by the lack of an acute effect of moxonidine on insulin-dependent activation of Akt. Finally, moxonidine treatment in vivo also facilitated the glucose uptake response of adipocytes to insulin, suggesting that facilitated Akt signaling may have consequences of insulin sensitivity in the whole organism.
FOOTNOTES
*Supported by HL44514 from the NIH
1Submitted in partial fulfillment of the requirements for a doctorate in Nutrition from Case Western Reserve University School of Medicine.
120 Figure Legends
Figure 12. Representative Western blot showing phosphorylated Akt
immunoreactivity (rows labeled Akt-PO4) and total Akt immunoreactivity (rows labeled Akt) following stripping and re-probing of the same blots. Blots were scanned and digital images processed by taking the ratio of optical density in the
Akt-PO4 band to the corresponding Akt band. Numbers along the bottom indicate time of exposure to 100 nM insulin, from 0 to 90 min. Zero time controls were run in duplicate.
Figure 13. Time course of insulin activation of Akt in lean SHR, SHROB and in
SHROB treated with moxonidine for 21d in vivo. Epididymal adipocytes were
isolated from age and sex-matched rats from each group. Insulin (100nM) was
applied to incubated adipocytes for various lengths of time lengths up to 90 min
before cells were lysed in a boiling water bath. Cell lysates were immunobloted
for phosphorylated Akt using anti-active Akt antibody. Blots were stripped and
reprobed for total Akt. Data are expressed as the ratio of optical density of
phosphorylated Akt to total Akt. All data were normalized to vehicle treated
samples processed in parallel on the same gel. Each group included at least six
rats.
Figure 14. Dose response curves for insulin activation of Akt in lean SHR,
SHROB and in SHROB treated with moxonidine for 21d in vivo. Epididymal
adipocytes were exposed to various concentrations of insulin for 10 min prior to
lysis by boiling. Cell lysates were immunobloted for phosphorylated Akt using
anti-active Akt antibody. Blots were stripped and reprobed for total Akt. Data are
121 expressed as the ratio of optical density of phosphorylated Akt to total Akt. All
data were normalized to vehicle treated samples processed in parallel. Each
group included at least six rats.
Figure 15. Basal Akt activation is not affected by phenotype or
pharmacotherapy. Relative Akt activation in the absence of insulin was compared
between adipocytes from SHR, SHROB and SHROB treated with moxonidine for
21d. Data were expressed as a ratio with SHR adipocytes run in parallel in each
experiment. The standard error bars reflects between subject variability for each
group. None of the differences approached statistical significance.
Figure 16. Dose-response curve for insulin activation of [3H]-2-deoxy-D-
glucose uptake. Adipocytes were isolated and incubated with or without various
concentrations of insulin for 30 min. The transport blocker cytochalasin-B (10 µM)
was applied 10 min before the end of insulin incubation in some cell aliquots to
allow determination of non-specific uptake. Cells were exposed to 2.5mM [3H]-2-
deoxy-D-glucose for 3 min prior to isolation by centrifugation through an oil
barrier. Data are expressed as nmol [3H]-2-deoxy-D-glucose as amount of
glucose transported in 3 min per 105 cells. Curves were obtained by nonlinear
curve-fitting to a logistic equation. Panel A: Glucose uptake comparison among
adipocytes from SHR (N=7), SHROB (N=4) and SHROB with chronic moxonidine
treatment (N=6). Nonspecific uptake did not differ between groups (SHR: 1.6 ±
0.48, SHROB: 1.4 ± 0.27. SHROB + moxonidine: 1.6 ± 0.31, all in nmol/105 cells/3min). *Significantly different from value for SHR group. Panel B: Glucose uptake comparison between SHR and SHR with chronic moxonidine treatment.
122 The control SHR group included four rats and the treated SHR group included
three rats.
Figure 17. Treatment with moxonidine alone in vitro does not affect Akt
activation. Adipocytes were incubated with 100 nM moxonidine for 90 min before
cells were lysed. Cell lysates were immunobloted for phosphorylated Akt using
anti-active Akt antibody. Blots were stripped and reprobed for total Akt. Data are expressed as the ratio of optical density of phosphorylated Akt to total Akt. All data were normalized to control, which was arbitrarily defined as 1 in both SHR and SHROB group. Each group included at least five rats.
Figure 18. Representative blot showing the time course of Akt activation by insulin with and without 90 min pretreatment with 100 nM moxonidine. The presence or absence of moxonidine pre-exposure is indicated by + or –, respectively, below the band. The bottom row of numbers indicates the number of minutes of exposure to 100 nM insulin (0 to 90). Cell lysates were immunoblotted for phosphorylated Akt using anti-active Akt antibody. Blots were then stripped and reprobed for total Akt.
Figure 19. Effect of in vitro moxonidine treatment on insulin activation of Akt in adipocytes from SHR and SHROB. Adipocytes were isolated from lean SHR
(Panel A) or from SHROB (Panel B) and half of the cells from each animal were incubated with 100 nM moxonidine for 90 min before insulin was applied while to other half was incubated in parallel with vehicle alone. Insulin (100 nM) was applied to both moxonidine pretreated and control adipocytes for various lengths
123 of time from 3 min up to 90 min before cells were lysed. Each group included at least six rats. *Significant difference for vehicle group by Newman-Keuls test.
124 Time (min)
Figure 12
125
Figure 13
126
Figure 14.
127
Figure 15
128
Figure 16
129
Time (min)
Figure 17
130
Figure 18
131
Figure 19
132 CHAPTER 5. DISCUSSION AND SIGNIFICANCE
The Significance of the Present Study on IRAS
The molecular basis for a putative I1-imidazoline receptor (I1R) has been a major interest in related investigations ever since the proposal of the existence of such a receptor in 1984 (Bousquet et al., 1984). The early efforts trying to purify an I1R protein did not successfully lead to sufficient quantities or purities of a protein for amino acid sequencing (Wang et al., 1992). However, the partially purified protein was successfully used to generate antibodies that labeled neurons in specific brain regions but not in others, suggesting that the imidazoline receptor immunoreactivity related protein might play a role in brain function (Escriba et al., 1994). The cross-reactivity this antibody displayed with a second independently developed anti-idiotype antibody against the imidazoline receptor was used in the isolation of a cDNA clone.
Because of the results of the current study, it is now clear in retrospect that the isolation of the IRAS gene should be considered a major landmark of progress in the I1R research area. Several converging lines of evidence now support the idea that the IRAS gene encodes an I1R receptor protein. This
unique protein shares little identity with any other known sequences of
imidazoline binding proteins in public databases, including I2 binding sites on
monoamine oxidase and the candidate I3 protein which may be a potassium
channel subunit (Monks et al., 1999). Transfecting IRAS cDNA into PC12 cells
and Chinese Hamster Ovary (CHO) cells both induced the appearance of
specific high affinity binding sites with properties identical to those of binding
133 sites in native tissues (Piletz et al., 2000;Piletz et al., 1999). Furthermore, a
significant positive correlation between IRAS mRNA level and I1-R binding
density across rat tissues has been demonstrated by conducting Northern blot
and radioligand binding assays in parallel on identical tissues (Piletz et al., 1999).
The first half of the results presented in the current work was centered on
testing the hypothesis that the IRAS gene encodes a functional I1-imidazoline
receptor. Studies were performed in PC12 rat pheochromocytoma cells, which is
a tumor cell line originally derived from a solid tumor in rats of the New England
Deaconess strain (Greene and Tischler, 1976). This is an excellent experimental
model in which to test I1-imidazoline receptor related hypotheses. For example,
PC12 cells have high affinity specific I1-imidazoline binding confirmed by multiple research groups (Gentili et al., 2003;Greney et al., 2002;Ivanov et al.,
1998;Musgrave et al., 1996;Separovic et al., 1996;Steffen et al., 1995). In
addition, most of the current knowledge of I1-R linked cell signaling events has
come from studies conducted in PC12 cells, including the activation of
phosphatidylcholine-selective phospholipase C (PC-PLC) (Separovic et al., 1996),
PKCβII,PKCζ, and ERK (Edwards et al., 2001;Edwards and Ernsberger, 2003).
Many of the other signaling studies of I1R have been conducted in neurons,
which are closely related to PC12.
Antisense and Imidazoline Binding Studies:
This starting point for this series of experiments was to determine whether
antisense treatment had any effect on imidazoline radioligand binding. The
hypothesis was that antisense transfection should cause reduced expression of
134 IRAS protein. If expression of the IRAS gene is required for specific imidazoline
binding, then this loss in total amount of IRAS should be reflected by impaired
binding capacity. Radioligand binding assays with [125I]p-iodoclonidine has been
an established technique for over 10 years in I1-imidazoline receptor related
investigations (Ernsberger et al., 1995). The advantage of performing this binding
assay with PC12 cells is that these cells have been confirmed to have no α2- adrenergic receptors, which that also bind [125I]p-iodoclonidine and other
imidazoline ligands with high affinity. Only plasma membrane enriched fractions
were responsible for the specific and high affinity radioligand binding, although
IRAS was found to reside both in cytosol as well as membrane fractions. In
multiple cell types, it has been found that more IRAS protein was present in
plasma membrane fractions than in cytosol (Piletz et al., 2000). This is consistent
with the idea that IRAS is a candidate for a plasma membrane residing receptor
and the fact that IRAS has a PX domain being able to interact and associate with
membrane-bound phosphoinositides, as well as serving as a subunit of the
fibronectin receptor by tightly binding to alpha5 integrin. In the present study, we
used partially purified plasma membrane fractions for binding assays and a high
density of binding sites of high affinity were detected, consistent with previously
published results.
A recent study also confirmed that IRAS participates the transportation of
proteins from plasma membrane to endosomes in a mechanism requiring the
presence of PX domain (Lim and Hong, 2004). It is therefore possible that the
IRAS protein may exist in two conformations, once associated with endosomes
135 which does not bind imidazolines and a second conformation associated with the
plasma membrane which does bind imidazolines. The conformation of the IRAS
protein might be altered by attachment to the α5 subunit of the fibronectin
receptor in such a way that exposes the imidazoline binding site. The proposed
binding site for imidazolines is located within the coiled-coil domain, which is
involved in the dimerization of the IRAS protein. One could speculate that
dimerized IRAS would be unable to bind imidazolines, since the coiled-coil
domain would be attached to the analogous domain of a second IRAS molecule.
Interaction of IRAS with the fibronectin receptor or other membrane proteins,
may promote formation of monomers which can bind imidazolines. PIP3,
although it binds to the PX domain, is probably not sufficient to induce the
formation of monomers that can bind imidazolines, since this membrane lipid is
abundant in endosomes as well as in the plasma membrane.
The loss of binding capacity from antisense transfection was about 42% (Bmax
= 691 ± 29 fmol/mg vs Bmax = 400 ± 16 fmol/mg) as shown in Figure 13A. This is
highly consistent with the IRAS expression level immunoreactive assay, in which
antisense transfected cells showed only about 50% total IRAS protein amount
(Figure 8A). The magnitude of this effect is within the general range of effect magnitude reported for antisense transfection in mammalian cells (Gleave and
Monia, 2005;Van Oekelen et al., 2003). Thus, these findings strongly suggest
that the normal expression level of IRAS is necessary for high capacity and high
affinity imidazoline binding. Also, since the Scatchard plot indicated that only one type of I1-R binding site was involved in specific binding across the radioligand
136 concentration range tested, it is not highly likely that IRAS is only an allosteric
regulating protein that adjusts the binding ability of a distinct unknown I1-R
protein.
Antisense and I1-imidazoline Cell Signaling Study
The initial event in imidazoline receptor signaling is the activation of
phosphatidylcholine-selective phospholipase C (PC-PLC), which is a very quick
response since the resultant synthesis of diacylglyceride second messenger
occurs within 15 s after exposure to an agonist (Ernsberger et al., 1995). The
activation of ERK1/2 seems to be much slower. Edwards et al observed an
increased phosphorylation level of ERK as early as 5 min after agonist treatment
in PC12 cells, however with a peak of activation did not occur until 90 min. A
similar activation pattern for JNK was also observed in the same study (Edwards
et al., 2001). This is consistent with the hypothesized mechanism that ERK is
activated by PKCβII and PKCξ, which are activated by diacylglyceride and
arachidonic acid respectively. The activation of ERK by imidazolines is blocked
by the broad spectrum PKC inhibitor staurosporine (Edwards et al.,
2001;Edwards and Ernsberger, 2003). Since MAPK activation can also be blocked by either the receptor antagonist efaroxan or by the PC-PLC inhibitor
D609, PC-PLC is probably the first upstream element of imidazoline cell signaling
cascade that leads to ERK activation. Thus the slow pattern of ERK activation
can be explained with release and accumulation of diacylglyceride and
arachidonic acid. I chose ERK activation over other elements as indicator of I1- imidazoline receptor cell signaling mainly because the effect is robust and
137 reproducible and the time-course curve of ERK activation has been established
(Edwards et al., 2001), and the peak at 90 min is more convenient to repeat.
The maximum ERK1/2 activation without antisense transfection at 90 min was
about 300% of basal level, very similar to what was observed in previous studies.
With antisense transfection, the maximum ERK1/2 activation was reduced by
about half (Emax=1.44 ± 0.16 fold increase vs Emax=2.97 ± 0.33 fold increase, as shown in Figure 9A). Considering that the proportion of IRAS expression remaining in the presence of antisense transfection was about 50%, this 300% to
150% ratio in activation of a very downstream signaling element seems a little low at a first glance. However, if compare the increases of ERK activation over basal level between these two conditions, we have 200% increase without and
50% increase with antisense transfection, thus a 4:1 ratio. With the consideration that imidazoline receptor cell signaling events are generally moderate in intensity relative to powerful stimuli such as growth factors, this antisense induced ERK activation impairment is in fact quite significant.
This is the first study so far in the IRAS related research area to link the binding capacity and cellular signaling of I1-imidazoline receptor in an agonist-
receptor interaction paradigm. My main purpose was not to explore new cellular
functions of IRAS like many other groups are doing, but to establish the role of
this protein as a receptor candidate. Thus the application of imidazoline agonist
has been a major concern throughout the whole study. Although the sequence of
IRAS protein ruled out the possibility of being a G-protein coupled receptor,
contrary to some speculation in the past (Ernsberger et al., 1995), the two
138 potential transmembrane sites predicted within its amino acid sequence suggest
that IRAS could be a unique transmembrane receptor. Another possibility is that
IRAS may attach to the inner side of plasma membrane via its N-terminal PX
domain, and imidazoline ligands have to travel through the plasma membrane
before interacting with the receptor. This is possible with most of the imidazoline
ligands, which are highly hydrophobic and thus cell permeable. In fact, most of
the agents penetrate the brain when given systemically, which requires the
crossing of multiple cell membranes. However, there are other ligands for the
imidazoline receptor which are hydrophilic, most notably the endogenous imidazoline ligand imidazole-4-acetic acid ribotide, which includes an essential phosphate group which is negatively charged (Prell et al., 2004).
Under either situation, the results from my study strongly support the
hypothesis that IRAS is the I1-imidazoline receptor gene, or at least a main part
of a receptor complex. Since IRAS contains a ruthenium binding sequence and
ruthenium red was found to compete with imidazoline binding, it is more likely
that IRAS alone functions as a functional binding protein. Functional binding
proteins can be considered receptors even if they do not mediate transmembrane signaling. Examples include the ryanodine receptor, the inositol triphosphate receptor, the wortmannin receptor, various mitochondrial binding sites such as the I2-imidazoline and the peripheral benzodiazepeine site, as well
as others.
Another way to establish the role of IRAS is to fully understand the functions
of this protein that are not apparently linked to the binding of imidazoline ligands.
139 For example, the truncated mouse copy of IRAS has been named Nischarin, after a Hindu god of destruction. This name came from the observation that it could bind to integrin α5 subunit and thus apparently inhibit cell migration (Alahari et al., 2000). So far, no reports are available to link imidazoline ligands with cell migration. Is this just a side function of IRAS that has nothing to do with imidazoline action? Is the effect of cell migration related to the fact that the
Nischarin gene is truncated, and is missing the crucial PX domain that allows protein sorting functions and localization of the protein to the plasma membrane?
It is possible that the truncated gene used in these studies might have functioned as a dominant negative, and thus the functions attributed to IRAS/Nischarin are actually a result of interference with IRAS/Nischarin. Or is cell migration really an imidazoline related function that had been missed in the previous studies with imidazoline receptors?
Once these questions can be answered and we have sufficient knowledge on these cellular functions of IRAS, the identity of IRAS as I1-imidazoline receptor should no longer be in debate.
Lienhard’s group searched for proteins in HEK293 cells that interact with IRS-
4 by co-immunoprecipitation and serendipitously found IRAS (Sano et al., 2002).
In fact in the same study all 4 forms of insulin receptor substrate showed the ability to form protein complexes with IRAS as demonstrated by immunoprecipitation, it was just that IRS-1, IRS-2 and IRS-3 showed lower affinity than IRS-4. Since IRS family members are important and specific insulin signaling elements, the association with these proteins strongly suggested that
140 IRAS plays a role in cellular insulin actions. Considering multiple studies
including my data on SHROB adipocytes showed that I1-imidazoline receptor
ligands moxonidine and rilmenidine facilitate insulin sensitivity and glucose
tolerance in both human patients and animal models (Velliquette and Ernsberger,
2003b;Bousquet et al., 2000;Haenni and Lithell, 1999;Reid, 2000), this association between IRAS and IRS proteins could be a potential link between imidazoline actions and insulin signaling, if IRAS really encodes an I1-imidazoline receptor or a subunit of it. My antisense study on IRAS strongly supports the role of IRAS as a functional imidazoline receptor, thus the association of IRAS with
IRS proteins could be a potential link between imidazoline actions and insulin signaling. In other words, imidazoline ligands might facilitate insulin sensitivity through the interaction between IRAS and IRS proteins by conformational changes in the protein complex induced by ligand binding. Interestingly, treatment of SHROB with moxonidine increases the level of expression of IRS-1 protein in liver and skeletal muscle, as well as facilitating insulin-induced tyrosine phosphorylation of both IRS-1 and the insulin receptor itself (Ernsberger et al.,
1999b;Ishizuka et al., 1998b). It is not known whether the physical association of
IRAS with IRS proteins is involved in the apparent induction of IRS-1 protein by sustained imidazoline agonist exposure in vivo.
In the Lienhard group’s study, IRAS was found to form a protein complex with
PI3-K and Grb2 through IRS. PI3-K and Grb2 are critical insulin signaling elements recruited and activated by tyrosine phosphorylated IRS, thus inducing the activation of Akt and ERK cascades respectively (Cheatham et al.,
141 1994;Holgado-Madruga et al., 1996). However, over-expression of IRAS in
HEK293 cells does not elevate the binding of PI3-K or Grb2 to IRS, nor does it enhance the insulin induced Akt activation (Sano et al., 2002). This seeming discrepancy in fact reveals an important limitation of their study, which is the lack of imidazoline ligand stimulation. Indeed, it is reasonable for a receptor protein not to be able to elicit normal cellular functions without the specific ligand interaction. In my antisense experiments, imidazoline ligand (moxonidine) always plays a central role in evaluating the cellular functions of IRAS. Our main concern is to investigate IRAS as receptor protein, as that is what it is a candidate for being. Thus it could be fruitful to test the possibility of IRAS induced enhancement in PI3K and Grb2 binding to IRS or facilitated insulin stimulated
IRS tyrosine phosphorylation in the presence of imidazoline ligands.
Within months after IRAS was isolated and identified as I1-imidazoline receptor candidate by the Piletz group (Piletz et al., 2000), a protein named
Nischarin was found to interact with the integrin and thus inhibit cell migration
(Alahari et al., 2000). Later it was found that these two proteins share a very high homology in sequence. Now it has been widely accepted that Nischarin is in fact a truncated version of IRAS (Lim and Hong, 2004) in which Nischarin lacks the
N-terminal PX domain in IRAS.
We have already discussed the question of whether the migration inhibiting effect of this truncated version of IRAS is one of normal functions of IRAS, since changes in cell migration has never been observed in any imidazoline receptor related studies, although it has never been directly tested for either. A recent
142 study demonstrated that the PX domain is critical for association of IRAS with
PI3P enriched endosomes, although the interaction with integrin α5 subunit is not significantly affected by the presence or absence of the PX domain (Lim and
Hong, 2004). This suggests that the missing PX domain in Nischarin could alter the cellular distribution and trafficking thus affect the normal cellular functions of
this protein. In HEK293 cells almost all IRAS is found to be complexed with IRS-4
even under normal conditions, as nearly all IRAS protein can be recovered by
immunoprecipitation with anti-IRS antibodies (Sano et al., 2002). Thus the PX
domain mediated membrane association of IRAS could play a critical role in IRS translocation and activation of its downstream targets including PI3-K and Grb2.
In addition, evidence exists to support the idea that overexpression of integrin- binding proteins may alter the conformation or even disrupt the cytoskeleton (van der and Sonnenberg, 2001). Thus, changes in cell signaling induced by overexpressing such pivotal proteins may not reflect their normal functions within the cell. In other words, the cell migration blocking effect induced by over expression of this truncated version of IRAS might not be one of the normal cellular functions of IRAS, the imidazoline receptor.
Possibility of IRAS as a Subunit of Imidazoline Receptor
My antisense study strongly suggests IRAS be a functional I1-imidazoline receptor. However possibility can not be ignored that the so called imidazoline receptor could be more than one single protein and IRAS is just one subunit of it.
Discrepancies have been observed between IRAS expression and I1-imidazoline
binding capacity in different cell lines. On the one hand, overexpression of IRAS
143 in CHO cells leads to an increase in high affinity I1-imidazoline binding sites
(Piletz et al., 2000). Also in the present study antisense transfection induced
inhibition of IRAS expression in PC12 cells significantly reduced specific I1- imidazoline binding on plasma membrane. On the other hand, IRAS overexpression in COS7 and SF9 cells failed to yield any increase in the density of I1-sites (Piletz et al., 2000;Piletz et al., 2003). One reasonable explanation to
this discrepancy is that the I1-imidazoline receptor resembles a multi-protein complex and IRAS is only one required subunit of it. Some other components are also required for the full function of receptor. Thus if in some cell types such as
COS7 and Sf9 one or more of these other components is not expressed, over expression of the single IRAS is not enough to form a functional I1-imidazoline
receptor. However, although the Sf9 cell lacks endogenous integrins (Miranti et al., 1999), COS7 cells express alpha5 integrin at high levels (Lin et al., 2005) and respond to fibronectin with ERK activation and cell adhesion (Miranti et al., 1999).
Thus, a lack of alpha5 integrin cannot account for the failure of COS7 cells to express imidazoline binding sites when transfected with human IRAS gene. Even more surprisingly, COS7 cells transfected with IRAS show an anti-apoptotic effect similar to the apoptotic effect of overexpressing IRAS in PC12 cells
(Dontenwill et al., 2003b). The presence of functional imidazoline binding is apparently not required for the anti-apoptotic effect of the IRAS protein. One would predict that COS7 cells transfected with IRAS would not be responsive to imidazoline ligands, since these cells are not competent to express imidazoline binding sites.
144 IRAS Could Also Act as a Scaffolding Protein
Since IRAS is a relatively huge protein containing multiple functional domains,
it is also likely that IRAS serves as a scaffolding protein that provides a structural
platform to form functional complexes. One example of scaffolding protein in
insulin signaling is the JNK-associated leucine zipper protein (JLP), which is of
similar size (180 KDa) and contains multiple functional domains to bring together
Max and c-Myc along with JNK and p38MAPK, as well as their upstream kinases
MKK4 and MEKK3 (Lee et al., 2002). The association of IRAS with IRS-4, PI3-K
and Grb2 might be a result of the possible scaffolding protein potency of IRAS. If
this is true, IRAS could be an adaptor that facilitates the interaction of its
associated proteins, and conformational changes induced by imidazoline binding
may play a role of modulating the signaling events within the protein complex.
Interestingly, the integrin α5 subunit that interacts tightly with is a PDZ domain
containing protein. The presence of one or more PDZ domains appears to be a
hallmark feature of many scaffolding proteins (Altschuler et al., 2003). Thus the
size and complicity of the IRAS, IRS-1, PI3-K and Grb2 complex could be significantly higher than we assumed, and an I1-imidazoline receptor complex
with more cellular functions and much more subtle regulatory machinery could be
expected.
There is a proline-rich region (PRR) in human IRAS between amino acid 1047
and 1103. PRR is a functional domain found to interact with various
macromolecules including microtubules, acidic phospholipids, and Src homology
3 (SH3) domains (Okamoto et al., 1997). Does this domain play a role in the
145 cellular functions of IRAS? Is the presence of this domain critical for the regular cellular actions of IRAS? Musgrave and colleagues hypothesized a model for I1- imidazoline receptor action, in which ligand stimulated IRAS binds to integrin α5, thereby serving as a transmembrane scaffolding protein. These authors propose that the PPR in IRAS binds to an SH3 domain containing tyrosine kinase such as
Src, and then activated Src further activates PC-PLC, which has been demonstrated to be critically involved in I1-imidazoline receptor signaling
(Musgrave et al., 2003). In this model, the PPR is definitely necessary for the proper assembling of a functional protein complex. However, from my comparison among IRAS sequences in human, mouse and rat (figure 6), the rat version of the IRAS protein may lack this PPR (the rat sequence has only 3 proline residues corresponding to the 15 prolines in the human sequence between amino acid 1047 and 1103). In contrast, the mouse sequence shows a high homology and even longer proline-rich region relative to human sequence in the corresponding area. This finding suggests that a fully functional PPR may not be necessary for the cellular actions of IRAS and this might compromise the model proposed by Musgrave et al. The rat version used here (XM_240330) is in fact a predicted sequence which still needs further evidence to confirm its identity.
However, if it is true that rat IRAS does lack the PPR, we may have to reconsider the possible role of this domain in IRAS. Further evidence is required to settle this discrepancy.
Dontenwill et al found that over expression of IRAS in PC12 cells prolongs cell survival against apoptotic stimulus (Dontenwill et al., 2003a;Dontenwill et al.,
146 2003b). My findings as well as previous evidence strongly suggest that IRAS
may encode an I1-imidazoline receptor and treatment with imidazoline ligands
improves impaired insulin signaling and glucose uptake. Considering all these
preliminary facts together, I hypothesize that the anti-apoptotic effect of IRAS
could be the result of the possible intersection between I1-imidazoline and insulin
receptor signaling pathways. Akt, a key element in insulin signaling, has been
known to promote cell survival through regulating activity of Bad and caspase proteins (Osaki et al., 2004;Yamaguchi and Wang, 2001). Thus it is possible that
IRAS enhances the activity of the Akt cascade which then promotes cell survival.
The fact that IRAS can bind to insulin receptor substrate proteins supports this possibility (Sano et al., 2002). This anti-apoptotic effect of IRAS does not require the presence of imidazoline ligands (Dontenwill et al., 2003b), though they did not test whether ligands can promote the cell survival to a higher extent.
However since Edwards et al found that two-day treatment of PC12 cells with clonidine increases cell number up to 50% in culture in a dose responsive manner(Edwards et al., 2001), it is likely that the anti-apoptotic effect of IRAS could be activated with imidazoline ligands. In the experiments by Edwards et al, it was not possible to distinguish between proliferative and anti-apoptotic actions.
Comparison of Adipocyte Glucose Uptake Results to Previous Studies
In the insulin stimulated adipocyte glucose uptake experiments, the EC50 of
insulin (insulin concentration required to reach 50% of maximum action) in the
three groups are around 1.5 nM (2.0 ± 0.8 nM in SHR, 1.4 ± 0.6 nM in SHROB
and 1.7 ± 0.6 nM in moxonidine treated SHROB). However in many previous
147 studies the EC50 of insulin stimulated rat adipocyte glucose uptake lies between
0.1 nM and 0.3 nM, including studies carried out with Sprague-Dawley, Wistar
Kyoto and SHR rats (Green, 1986;Caldiz and de Cingolani, 1999). Multiple reasons could be responsible for such a discrepancy, for example difference in the ingredients of incubation buffer, or different insulin incubation time length (45 min vs. 30 min). However I think the different insulin concentrations used could also be an important issue. Since in these previous studies the highest concentration of insulin tested was usually 10 nM or even less, a consequence of narrowed dose range and reduced maximum insulin action could lead to a lower
EC50. Thus, a limitation of the present series of experiments was that they were
biased toward higher insulin concentrations. The main reason we expand the
dose range of insulin to 100 nM is that SHROB adipocytes are severely insulin
resistant which is most readily detectable at an insulin concentration of 100 nM.
Impact of Insulin Concentration
In my studies on insulin stimulated adipocyte glucose uptake, the effect of
100 nM insulin in all groups was less than the effect of 10 nM insulin. This is
reflected by the dotted line in the figure showing glucose uptake, which indicates
a portion of the dose response curve which cannot be fitted by a normal logistic
dose response function. This discrepancy between effect and dose could be due
to a couple of reasons. One reason is that there is spare insulin receptors
present on adipocytes, and the bottle neck for glucose uptake probably lies in the
GLUT4 transporting machinery (Frank et al., 1981). Thus the higher insulin
concentration has no advantage in inducing insulin action. On the other hand,
148 higher insulin concentration may induce insulin resistance in these adipocytes
during in vitro incubation by the phenomenon of receptor desensitization.
Alternatively, excessively high concentrations of insulin might have nonspecific
effects, such as binding to other receptors like IGF-1 receptor, consequently
resulting in impairment in insulin signaling.
Spare receptors are receptors which exist in excess of those required to
produce a full effect. Thus, a maximum biological response can be achieved
when only a fraction of total receptor sites are occupied and the ligand
concentration required to produce one-half maximal cellular action (EC50) is less than the concentration required to achieve one-half maximal binding (Kd). It has
been found that there are spare receptors in rat adipocytes for insulin action
(Kono and Barham, 1971;Caldiz and de Cingolani, 1999;Frank et al., 1981), and
less than10% of the receptors present on adipocytes have to be occupied to
trigger for maximal stimulation of insulin responses (Frank et al., 1981). If
SHROB adipocytes express fewer insulin receptors, then this might explain their
reduced sensitivity. Supporting this concept, our laboratory has previously found
that SHROB liver and skeletal muscle express a lower level of insulin receptor
protein (Friedman et al., 1997b). Moreover, SHR are not entirely normal, as they
have elevated blood pressures and increased levels of several hormones
including catecholamines which could affect metabolism and possibly affect
insulin receptor expression (Bursztyn, 1996). Indeed, adipocytes from SHR have
been reported to be insulin resistant (Caldiz and de Cingolani, 1999).
SHR as Control for SHROB
149 In all my animal and adipocyte experiments, SHR were used as control for
SHROB. In fact SHR are still quite different than those normotensive controls such as Sprague-Dawley or Wistar rats. Other than spontaneous hypertension,
SHR also express high levels of free fatty acids (1.81 ± 0.09 mmol/L in SHROB and 1.45 ± 0.05 mmol/L in SHR) which are more than double the levels typically found in normotensive control rat strains (Velliquette et al., 2002a). Moreover,
SHR are slightly but significantly insulin resistant in relative to Wistar rats
(Swislocki and Tsuzuki, 1993). These aspects make SHR even capable of being an animal model for insulin resistance and essential hypertension in non-obese humans. Thus, our control group was somewhat insulin resistant. This means the level of insulin resistance in the SHROB rat is probably underestimated, and would be even greater if we used a truly normal control such as a Brown Norway rat. However, since SHROB express much more severe insulin resistance than
SHR yet show comparable levels of hypertension and free fatty acids, I believe
SHR is still the most suitable control for investigating of imidazoline receptor mediated insulin sensitizing effect in SHROB. This ensures that the effect of chronic moxonidine treatment not be affected by non-relevant factors such as genetic differences. Interestingly we recently found out that our SHR (SHR/Kol) do not have the CD36 mutation that most other SHR have, which means our rats are genetically closer to the original Japanese SHR from Kyoto than are other
SHR strains that are descended from colonies at the NIH. The defect in CD36 is though to contribute to the insulin resistance of SHR (Aitman et al., 1999). Thus,
150 the insulin resistance in the SHR/Kol substrain used in my studies is probably
less than that reported by others outside of Japan.
Insulin Degradation in the Experiments
In both insulin induced Akt activation and glucose uptake assays, insulin was
applied to adipocyte suspensions in the presence of 0.1% bovine serum albumin
(BSA). BSA in these buffers has multiple protective actions, one of which is to
protect insulin from degraded by proteases released by adipocytes (Sonne,
1985). According to a study published in 1985, the protective effect against
insulin-degrading enzymes is not complete even with high concentration of BSA.
Thus it seems the 0.1% BSA used in my studies may not be enough to totally
inhibit extracellular insulin degradation during incubations. However the apparent
dose dependence of insulin induced Akt activation and glucose transport in the
results of my studies suggest that the influence of insulin degradation may not
play a major role in these experiments. Plus, a BSA concentration between 0.1%
and 1% has been widely applied in glucose uptake assays with isolated rat
adipocytes (Juan et al., 2005;Rachdaoui et al., 2003). A few possibilities may
account for the insignificance of insulin degradation in these incubations. The
incubation buffers are all about pH7.4 and Zn++ free, which probably would impair
the activity of some proteases that require an acidic environment or presence of
Zn++. Also, the relative short time span of incubation (usually within 30 min, no longer than 90 min) may limit the influence of insulin degradation. The possibility still exists that the real insulin concentrations be lower than expected in these incubations due to extracellular as well as intracellular endosome-dependent
151 degradation, and this may cause a shift of dose response curves toward left on X
axis. However even in this case the relative comparisons between different
groups are still valid and Emax data are not shifted. Perifusion rather than static
incubation may solve this problem. We would predict that lower concentrations of
insulin would be effective in a flow through or perifusion type of incubation.
SHROB as an animal model for human syndrome X
Metabolic syndrome X is a complex disorder. Numerous metabolic and
biochemical processes contribute to its development and maintenance. The
second part of my dissertation focused on the effect of chronic and acute I1R activation on insulin sensitizing effects, in adipocytes from the SHROB rat, an animal disease model that closely resembles human metabolic syndrome X.
Cellular mechanisms of both chronic and acute I1R activation were investigated
and compared. The reason I chose SHROB as animal model is that this rat is an excellent experimental model for investigations toward metabolic syndrome X,
and in this model the abnormalities are responsive to treatment with imidazoline
drugs. SHROB have all the major components of metabolic syndrome X:
hypertension, insulin resistance and glucose intolerance, hyperlipidemia and
abdominal obesity. Also, the lean littermates SHR, are ideal genetic and
environmental controls of SHROB for glucose metabolism related studies, even
thought the addition of a second completely normal control group may have been
advantageous.
Muscle, liver and adipose tissue are major insulin-responsive tissues that
contribute to taking in and utilizing blood glucose thus to maintain normal
152 glycemia level. Insulin resistance in either of these tissues could lead to a reduced efficiency of blood glucose clearance following a glucose load or food intake. The present study found the phosphorylation of Akt in response to insulin stimulation is severely impaired in adipocytes from SHROB, and insulin stimulated glucose uptake by these cells were also significantly decreased.
Considering the huge mass of adipose tissue in these animals, the impact of reduction in glucose uptake by fat cells could significantly contribute to the overall insulin resistance and glucose intolerance of SHROB with metabolic syndrome.
However insulin resistance in adipose tissue may not always lead to deterioration of glucose metabolism. In contrast, mice with adipose tissue specific insulin receptor knock-out have reduced fat mass and are protected against obesity and its subsequent metabolic abnormalities including insulin resistance (Bluher et al.,
2002;Bluher et al., 2003).
Does impaired insulin responsiveness in fat tissue worsen or improve glucose metabolism? The answer may depend on whether other tissues of the animal are insulin resistant or not. The adipose tissue insulin receptor knock-out mice that
Bluher et al used in their studies were without metabolic syndrome. Although the fat cells did not respond to insulin at all due to absence of the insulin receptor, other insulin target tissues had normal responsiveness to insulin and could probably take and utilize blood glucose fast enough to compensate for adipose tissue. And since adipocytes could not take enough glucose for triglyceride synthesis, the total body fat mass was kept low, and thus obesity induced insulin resistance was prevented.
153 However, in SHROB the whole body is already extremely insulin resistant.
There might be no compensation that muscle and liver can provide for the defect
in glucose uptake by adipocytes. Thus, insulin resistance in adipose tissue could
directly worsen the whole body glucose tolerance. Also, insulin resistance in
adipose tissue in SHROB does not prevent them from gaining fat mass as it does
in those knock-out mice. This is probably due to two reasons. One is that the
adipocytes in SHROB still have some remaining insulin responsiveness instead
of total loss like in knock-out mice. The other one is that SHROB have
hyperinsulinemia, in which the fasting plasma insulin level is over 40 fold higher
than SHR. This extremely high level of insulin should be enough to stimulate fat
storage in adipose tissue in SHROB. It would be interesting to test whether
adipose tissue specific knock-out of insulin receptor would also lead to improved
whole body insulin resistance in a murine model of metabolic syndrome, such as
the BL6 mouse strain fed a high fat diet (Einhorn et al., 2003).
Akt Activation Studies
To link I1R activation with cellular changes of glucose metabolism, a
reproducible indicator of insulin cell signaling events is necessary. I chose Akt as
this representative index of insulin signaling because it is considered a key
downstream element in the cascade (Whiteman et al., 2002). Insulin signaling
pathways diverge downstream of Akt, ultimately impacting multiple cellular
processes including glucose metabolism, cell survival and selective gene
transcription. Thus any element downstream of Akt may not be able to reflect the entire spectrum of insulin signaling. And since Akt is several steps downstream
154 from the insulin receptor itself in the signaling cascade, it should yield high
sensitivity in terms of activation. Previous studies in our lab indicated that both
the expression level and tyrosine phosphorylation of liver insulin receptor β subunit (IRβ) and insulin receptor substrate 1 (IRS-1) could be enhanced by chronic imidazoline treatment in SHROB(Ernsberger et al., 1999b;Friedman et al.,
1997b). However, in these studies the maximum increase in the proportion of
tyrosine phosphorylated protein for either IRβ or IRS-1 was no more than 3 fold greater than basal, and obese insulin-resistant animals show less response.
Working with a larger experimental signal would be preferable. Thus, we looked at steps further downstream in the hopes of detecting a larger insulin-induced change.
The results showed that Akt activation was an appropriate indicator for evaluating the sensitivity of insulin signaling in these adipocytes. The peak insulin induced Akt phosphorylation could be as high as 25 fold greater than the basal level in the adipocytes from SHR rat that are not affected by severe insulin resistance. In cells from insulin resistant SHROB rats, this peak Akt activation dropped to about 5-fold above basal (Figure 13). Thus Akt phosphorylation seems to be a very responsive indicator for insulin cell signaling tests in rat adipocytes, and the considerable difference between SHROB and SHR in term of insulin induced Akt activation suggests that the impaired glucose metabolism in
SHROB could well be linked to impaired insulin cell signaling at this step of the cascade. Moreover, as discussed in the Introduction, this step in the insulin
155 signaling cascade has also been implicated in human insulin resistance and type
2 diabetes.
Chronic I1-R Activation
Although moxonidine was initially discovered as an antihypertensive agent
which modulates sympathetic nervous system activity and attenuates peripheral
vascular resistance (Ernsberger et al., 1993), soon after that studies showed that
moxonidine also improves glucose tolerance and plasma insulin levels in animal
models and human patients with hypertension and obesity, suggesting that this
imidazoline agonist may also improve insulin sensitivity (Ernsberger et al.,
1996b;Lithell, 1997). More and more studies support the point of view that
moxonidine and other imidazoline agonists exert their insulin sensitizing effects
through I1R but not α2-adrenergic receptors which also interact with these ligands,
albeit with slightly lower affinity (Szabo, 2002;Velliquette and Ernsberger, 2003b).
However, the purpose of my study is not to further explorer the physiological
mechanism and the in vivo effects of these imidazolines but to find a linkage
between whole body glucose metabolism improvement and possible changes at
the cellular level. This knowledge may ultimately facilitate further investigation toward thorough mechanisms of these drugs in their multiple clinical applications.
The main site of action for the insulin sensitizing effect from chronic imidazoline treatment is still unclear. Although it has been demonstrated a long time ago that moxonidine lowers blood pressure mainly through imidazoline receptors in the rostroventrolateral medulla of the brain stem (Ziegler et al., 1996), the glucose metabolism related effects may still be through other sites or by other
156 mechanisms. The lipid lowering action of imidazolines appears to be situated within the liver (Velliquette et al., 2006b). By isolating and investigating cells from various tissues, especially insulin-responsive tissues, it may be possible to analyze whether imidazoline agonists exert effects directly on a certain tissue or cell type to alter glucose homeostasis. These data also bear on the contribution of each insulin responsive tissues to the whole body improvement in glucose metabolism.
The results from chronic moxonidine treatment were entirely consistent with previously observed insulin sensitizing effects on metabolic syndrome X.
Extending previous work on the initial steps of the insulin signaling cascade, insulin induced Akt activation in SHROB adipocytes was partially normalized by chronic moxonidine treatment. Glucose uptake, which is a major process in overall glucose metabolism within adipocytes, was also significantly improved toward normal levels. These results suggested that abdominal adipose tissue responses to insulin may play an important role in the overall insulin resistance in these obese rats, and that chronic imidazoline treatment can significantly normalize glucose metabolism in adipose tissue possibly through an enhancement in insulin signaling through much of its cascade. Furthermore, these changes in SHROB adipose tissue should count for a significant part of the whole body glucose tolerance improvement elicited by chronic imidazoline treatment, in part because adipose tissue makes up a high proportion of the total body mass in this model.
157 My study with SHROB suggests that imidazoline treatment partially restores glucose uptake into adipocytes in rats expressing severe insulin resistance.
Although there has been accumulating evidence that imidazoline ligands such as moxonidine and rilmenidine enhance glucose metabolism and insulin actions in both human patients and experimental animals (Haenni and Lithell,
1999;Velliquette et al., 2006b;Velliquette and Ernsberger, 2003b), the specific effect of these agents on each insulin responsive tissue had remained to be evaluated, especially fat tissue since adipose tissue was considered mainly as storage for energy in the form of lipid and less active than other insulin responsive tissues like liver and skeletal muscle in glucose metabolism. In fact, other than energy storage, adipose tissue is also an important endocrine organ and actively participates body glucose metabolism (Klein et al., 2006;Boden and
Hoeldtke, 2003). Due to the large amount of total fat tissue in human body, moxonidine induced adipocyte glucose uptake could count for a significant part in the whole body glucose tolerance enhancement. Adipose tissue should receive equal attention comparable to liver and muscle in future studies dealing with imidazolines and glucose metabolism.
Acute I1R Activation
Acute imidazoline treatment improved glucose tolerance and the insulin response to an oral glucose load (Velliquette and Ernsberger, 2003b;Stone et al.,
2003). Since both chronic and acute treatment of intact animals with imidazolines seems to elicit similar effects, a reasonable question might be: do these two drug effects share a common mechanism? My study on adipocytes with in vitro acute
158 moxonidine treatment suggests that this is unlikely. Primary freshly isolated
SHROB adipocytes showed no improvement in insulin signaling with 90 min
acute in vitro moxonidine treatment. Although we did not determine whether
these adipocytes have imidazoline binding sites or exhibit imidazoline receptor
linked cell signaling events, the lack of an influence of extracellular imidazolines on insulin stimulation of Akt suggested that in vivo acute imidazoline treatment probably does not elicit effects on adipose tissue directly. Instead, the central nervous system might be required in this procedure since CNS is the where imidazoline receptors elicit hypotensive effects and the acute treatment can take effect within 15 min after injection. Alternatively, the effects on glucose homeostasis may be mediated by the liver, similar to the effects on lipid homeostasis (Velliquette et al., 2006b). However, more evidence is necessary
before the question can be truly resolved.
Clinical Significance
Considerable efforts have been devoted to the search for drugs capable of addressing impaired glucose tolerance and insulin resistance. However, so far no single agent has been found to be suitable for treating all the components of metabolic syndrome X. Although Peroxisome proliferator-activated receptor gamma (PPARγ) agonists are currently the most promising agents dealing with glucose intolerance insulin resistance and hyperlipidemia, hepatic toxicity and weight gain effect may limit the application of these agents in many patients. Also in some cases, negative drug interactions and/or side effects can occur, leading to the addition of pharmacological agents to treat adverse effects of the drug
159 cocktail. Imidazoline agonists like moxonidine and rilmenidine already show the
potency to be effective therapeutic agents for the treatment of metabolic
syndrome X. Studies with these agents, including those presented in this thesis,
have shown the therapeutic benefits in both acute and chronic treatments and
suggested that I1-imidazoline receptor could be a novel drug target for the treatment of metabolic diseases.
160 CHAPTER 6. FUTURE STUDIES
Further investigation into I1-R cell signaling pathways could be a future study
project. So far, the cell signaling events determined to be linked to the I1-R are
still limited to the release of diacylglyceride and arachidonic acid, and the resultant PKC and MAPK activation. MAPK cascade activation is known to induce cell proliferation, which is consistent with later observations. For example, overnight moxonidine treatment in incubated PC12 cells lead to significant higher cell quantity, and over expression of IRAS provided the cells with an anti- apoptotic effect. However, these appear irrelevant to the insulin sensitizing effect of I1-R. Thus we hypothesized that imidazoline and insulin receptor signaling
pathways may intersect. Since Akt is a central element in insulin signaling,
whether imidazoline ligands can trigger Akt activation becomes an interesting
question. My preliminary data, presented at a recent Experimental Biology
meeting, showed that Akt could be stimulated by moxonidine up to 3 fold higher
than the basal level of activation, in both time- and dose-dependent manners.
PC12 cells and H4IIE rat hepatoma cells showed similar results. In PC12 cells,
this activation could be totally blocked with efaroxan, a specific I1-R antagonist.
This specific I1-R mediated Akt activation suggests that I1-imidazoline receptor
may regulate insulin signaling pathways directly, thus possibly contributing to its
insulin sensitizing effect.
More studies need to be done before we can determine this intersection of
cell signaling of these two distinct receptors. I would use specific antagonists of
I1-R and alpha2-adrenergic receptors to rule out any possible nonspecific effects.
161 In addition, radioligand binding studies using plasma membrane fractions from adipose tissues could be carried out. Preliminary data indicated excessive nonspecific binding of the lipophilic radioligand [125I]p-iodoclonidine to crude membranes isolated from rat adipose tissue, probably due to nonspecific binding interactions of triglycerides with the radioligand (data not shown). To get around this problem, highly purified plasma membrane fractions of adipocytes could be prepared by using continuous Percoll gradients or by detergent solubilization followed by polyethyleneglycol precipitation (Marshall et al., 1985).
Antisense strategy could be extended to animal experiments. Since imidazoline agonists normalize blood pressure by mediating sympathetic nervous system through rostral ventrolateral medulla oblongata (RVLM) region,
(Friedman et al., 1997d); (Ernsberger et al., 1997), injecting antisense oligo- nucleotides into RVLM of hypertensive SHR may provide data on identification of
IRAS as imidazoline receptor. The blood lowering effect of moxonidine should be compared on SHR rats with or without antisense injection in RVLM. If moxonidine treatment fails or partially fails to lower blood pressure in rats with antisense injection as in control SHR, this could be further evidence to support IRAS to be an imidazoline receptor. Also, similar radioligand imidazoline binding assay should be conducted with the plasma membranes from brain tissues of RVLM.
Thus the possible reduction of imidazoline sites in RVLM with antisense injection can be evaluated.
One possible follow up of my study on SHROB could be measuring in vivo glucose flux into adipose tissue following oral glucose load with stable isotope
162 tracers. Thus the importance of adipose tissue in imidazoline actions on insulin resistance could be further evaluated in a quantitative manner.
The molecular mechanism of insulin resistance in adipocytes from SHROB could also need further investigation. For example, expression level and insulin induced phosphorylation level of both insulin receptor and insulin receptor substrate could be compared between SHROB and SHR with western-blot.
Similar experiments have been done in liver and muscle tissues of SHROB as mentioned in the literature review section, however not in adipose tissue yet.
Subsequently, the possible effects of chronic moxonidine treatment on the expression and sensitivity to insulin of these proteins could also be measured.
This might lead to more understanding of how imidazoline agonists improve insulin sensitivity in adipocytes of SHROB following prolonged treatment in vivo.
This project could also be extended to more elements in insulin signaling cascade such as PI3-K and Akt.
Ultimately, the goal would be to identify agents that selectively enhance insulin signaling without excessively perturbing normal physiology. The imidazoline receptor, which we now know to be a scaffolding protein involved in cell signaling, could be an ideal target for correcting metabolic abnormalities associated with obesity and insulin resistance.
163 Bibliography
(2001) Executive Summary of The Third Report of The National Cholesterol
Education Program (NCEP) Expert Panel on Detection, Evaluation, And
Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III). JAMA
285: pp 2486-2497.
Abramowsky CR, Aikawa M, Swinehart G L and Snajdar R M (1984)
Spontaneous Nephrotic Syndrome in a Genetic Rat Model. Am J Pathol 117: pp
400-408.
Agren H (1986) Clonidine Treatment of the Opiate Withdrawal Syndrome. A
Review of Clinical Trials of a Theory. Acta Psychiatr Scand Suppl 327: pp 91-113.
Aitman TJ, Glazier A M, Wallace C A, Cooper L D, Norsworthy P J, Wahid F N,
Al Majali K M, Trembling P M, Mann C J, Shoulders C C, Graf D, St Lezin E,
Kurtz T W, Kren V, Pravenec M, Ibrahimi A, Abumrad N A, Stanton L W and
Scott J (1999) Identification of Cd36 (Fat) As an Insulin-Resistance Gene
Causing Defective Fatty Acid and Glucose Metabolism in Hypertensive Rats. Nat
Genet 21: pp 76-83.
Alahari SK (2003) Nischarin Inhibits Rac Induced Migration and Invasion of
Epithelial Cells by Affecting Signaling Cascades Involving PAK. Exp Cell Res 288:
pp 415-424.
164 Alahari SK, Lee J W and Juliano R L (2000) Nischarin, a Novel Protein That
Interacts With the Integrin Alpha5 Subunit and Inhibits Cell Migration. J Cell Biol
151: pp 1141-1154.
Alahari SK, Reddig P J and Juliano R L (2004) The Integrin-Binding Protein
Nischarin Regulates Cell Migration by Inhibiting PAK. EMBO J 23: pp 2777-2788.
Allan DR, Penner S B and Smyth D D (1993) Renal Imidazoline Preferring Sites and Solute Excretion in the Rat. Br J Pharmacol 108: pp 870-875.
Altschuler Y, Hodson C and Milgram S L (2003) The Apical Compartment:
Trafficking Pathways, Regulators and Scaffolding Proteins. Curr Opin Cell Biol 15: pp 423-429.
Anderson NJ, Seif I, Nutt D J, Hudson A L and Robinson E S (2006)
Autoradiographical Distribution of Imidazoline Binding Sites in Monoamine
Oxidase A Deficient Mice. J Neurochem 96: pp 1551-1559.
Andrieu N, Salvayre R and Levade T (1996) Comparative Study of the Metabolic
Pools of Sphingomyelin and Phosphatidylcholine Sensitive to Tumor Necrosis
Factor. Eur J Biochem 236: pp 738-745.
Anichkov DA, Shostak N A and Schastnaya O V (2005) Comparison of
Rilmenidine and Lisinopril on Ambulatory Blood Pressure and Plasma Lipid and
165 Glucose Levels in Hypertensive Women With Metabolic Syndrome. Curr Med
Res Opin 21: pp 113-119.
Armah BI, Hofferber E and Stenzel W (1988) General Pharmacology of the Novel
Centrally Acting Antihypertensive Agent Moxonidine. Arzneimittelforschung 38: pp 1426-1434.
Asami T, Kikuchi T, Asami K and Uchiyama M (1996) Effect of Clonidine on the
Height of a Child With Glycogen Storage Disease Type VI: a 13 Year Follow-Up
Study. Acta Paediatr Jpn 38: pp 524-528.
Atlas D and Burstein Y (1984) Isolation and Partial Purification of a Clonidine-
Displacing Endogenous Brain Substance. Eur J Biochem 144: pp 287-293.
Babu A and Fogelfeld L (2006) Metabolic Syndrome and Prediabetes. Dis Mon
52: pp 55-144.
Balkau B and Charles M A (1999) Comment on the Provisional Report From the
WHO Consultation. European Group for the Study of Insulin Resistance (EGIR).
Diabet Med 16: pp 442-443.
Bazan JF (1990) Structural Design and Molecular Evolution of a Cytokine
Receptor Superfamily. Proc Natl Acad Sci U S A 87: pp 6934-6938.
166 Benetz, B. A., Khosrof, S. A., Huang, S. S., Bedol, D. M., Bardenstein, D. S.,
Koletsky, R. J., and Ernsberger, P. Age of onset of retinal vascular changes in obese SHR. Invest.Ophthalmol.Vis.Sci. 37[3], S695. 1996.
Ref Type: Abstract
Bennai F, Greney H, Vonthron C, Bousquet P and Dontenwill M (1996)
Polyclonal Anti-Idiotypic Antibodies to Idazoxan and Their Interaction With
Human Brain Imidazoline Binding Sites. Eur J Pharmacol 306: pp 211-218.
Bluher M, Kahn B B and Kahn C R (2003) Extended Longevity in Mice Lacking the Insulin Receptor in Adipose Tissue. Science 299: pp 572-574.
Bluher M, Michael M D, Peroni O D, Ueki K, Carter N, Kahn B B and Kahn C R
(2002) Adipose Tissue Selective Insulin Receptor Knockout Protects Against
Obesity and Obesity-Related Glucose Intolerance. Dev Cell 3: pp 25-38.
Bock C, Niederhoffer N and Szabo B (1999) Analysis of the Receptor Involved in the Central Hypotensive Effect of Rilmenidine and Moxonidine. Naunyn
Schmiedebergs Arch Pharmacol 359: pp 262-271.
Boden G and Hoeldtke R D (2003) Nerves, Fat, and Insulin Resistance. N Engl J
Med 349: pp 1966-1967.
167 Bousquet P, Dontenwill M, Greney H and Feldman J (2000) Imidazoline
Receptors in Cardiovascular and Metabolic Diseases. J Cardiovasc Pharmacol
35: pp S21-S25.
Bousquet P, Feldman J and Schwartz J (1984) Central Cardiovascular Effects of
Alpha Adrenergic Drugs: Differences Between Catecholamines and Imidazolines.
J Pharmacol Exp Ther 230: pp 232-236.
Bousquet P, Monassier L and Feldman J (2001) Does It Make Sense to Develop
New Centrally Acting Cardiovascular Drugs? Clin Exp Pharmacol Physiol 28: pp
976-978.
Buccafusco JJ, Lapp C A, Westbrooks K L and Ernsberger P (1995) Role of
Medullary I1-Imidazoline and Alpha 2-Adrenergic Receptors in the
Antihypertensive Responses Evoked by Central Administration of Clonidine
Analogs in Conscious Spontaneously Hypertensive Rats. J Pharmacol Exp Ther
273: pp 1162-1171.
Bunag RD and Barringer D L (1988) Obese Zucker Rats, Though Still
Normotensive, Already Have Impaired Chronotropic Baroreflexes. Clin Exp
Hypertens A 10 Suppl 1: pp 257-262.
Bursztyn M (1996) Does Insulin Have a Role in Hypertension? Blood Press
Suppl 1: pp 64-66.
168 Cai JX, Ma Y Y, Xu L and Hu X T (1993) Reserpine Impairs Spatial Working
Memory Performance in Monkeys: Reversal by the Alpha 2-Adrenergic Agonist
Clonidine. Brain Res 614: pp 191-196.
Caldiz CI and de Cingolani G E (1999) Insulin Resistance in Adipocytes From
Spontaneously Hypertensive Rats: Effect of Long-Term Treatment With Enalapril and Losartan. Metabolism 48: pp 1041-1046.
Cardone MH, Roy N, Stennicke H R, Salvesen G S, Franke T F, Stanbridge E,
Frisch S and Reed J C (1998) Regulation of Cell Death Protease Caspase-9 by
Phosphorylation. Science 282: pp 1318-1321.
Ceulemans H, De Maeyer M, Stalmans W and Bollen M (1999) A Capping
Domain for LRR Protein Interaction Modules. FEBS Lett 456: pp 349-351.
Chan JL and Mantzoros C S (2005) Role of Leptin in Energy-Deprivation States:
Normal Human Physiology and Clinical Implications for Hypothalamic
Amenorrhoea and Anorexia Nervosa. Lancet 366: pp 74-85.
Chan SL, Brown C A, Scarpello K E and Morgan N G (1994) The Imidazoline
Site Involved in Control of Insulin Secretion: Characteristics That Distinguish It
From I1- and I2-Sites. Br J Pharmacol 112: pp 1065-1070.
169 Chang JP, Cook A F and Peter R E (1983) Influence of Catecholamines on
Gonadotropin Secretion in Goldfish, Carassius Auratus. Gen Comp Endocrinol
49: pp 22-31.
Cheatham B, Vlahos C J, Cheatham L, Wang L, Blenis J and Kahn C R (1994)
Phosphatidylinositol 3-Kinase Activation Is Required for Insulin Stimulation of
Pp70 S6 Kinase, DNA Synthesis, and Glucose Transporter Translocation. Mol
Cell Biol 14: pp 4902-4911.
Chen SR and MacLennan D H (1994) Identification of Calmodulin-, Ca(2+)-, and
Ruthenium Red-Binding Domains in the Ca2+ Release Channel (Ryanodine
Receptor) of Rabbit Skeletal Muscle Sarcoplasmic Reticulum. J Biol Chem 269: pp 22698-22704.
Chen Y, Aulia S, Li L and Tang B L (2006) AMIGO and Friends: An Emerging
Family of Brain-Enriched, Neuronal Growth Modulating, Type I Transmembrane
Proteins With Leucine-Rich Repeats (LRR) and Cell Adhesion Molecule Motifs.
Brain Res Brain Res Rev 51: pp 265-274.
Cheng L, Yang C, Hsu L, Lin M T, Jen C J and Chen H (1999) Acute Exercise
Enhances Receptor-Mediated Endothelium-Dependent Vasodilation by Receptor
Upregulation. J Biomed Sci 6: pp 22-27.
170 Cowen DS, Sowers R S and Manning D R (1996) Activation of a Mitogen-
Activated Protein Kinase (ERK2) by the 5-Hydroxytryptamine1A Receptor Is
Sensitive Not Only to Inhibitors of Phosphatidylinositol 3-Kinase, but to an
Inhibitor of Phosphatidylcholine Hydrolysis. J Biol Chem 271: pp 22297-22300.
Cowley S, Paterson H, Kemp P and Marshall C J (1994) Activation of MAP
Kinase Kinase Is Necessary and Sufficient for PC12 Differentiation and for
Transformation of NIH 3T3 Cells. Cell 77: pp 841-852.
Cross DA, Alessi D R, Cohen P, Andjelkovich M and Hemmings B A (1995)
Inhibition of Glycogen Synthase Kinase-3 by Insulin Mediated by Protein Kinase
B. Nature 378: pp 785-789.
DeFronzo RA and Ferrannini E (1991) Insulin Resistance. A Multifaceted
Syndrome Responsible for NIDDM, Obesity, Hypertension, Dyslipidemia, and
Atherosclerotic Cardiovascular Disease. Diabetes Care 14: pp 173-194.
Doggrell SA (2005) Clinical Evidence for Drug Treatments in Obesity-Associated
Hypertensive Patients--a Discussion Paper. Methods Find Exp Clin Pharmacol
27: pp 119-125.
Dontenwill M, Pascal G, Piletz J E, Chen M, Baldwin J, Ronde P, Dupuy L,
Urosevic D, Greney H, Takeda K and Bousquet P (2003a) IRAS, the Human
171 Homologue of Nischarin, Prolongs Survival of Transfected PC12 Cells. Cell
Death Differ 10: pp 933-935.
Dontenwill M, Piletz J E, Chen M, Baldwin J, Pascal G, Ronde P, Dupuy L,
Greney H, Takeda K and Bousquetd P (2003b) IRAS Is an Anti-Apoptotic Protein.
Ann N Y Acad Sci 1009: pp 400-412.
Duby JJ, Campbell R K, Setter S M, White J R and Rasmussen K A (2004)
Diabetic Neuropathy: an Intensive Review. Am J Health Syst Pharm 61: pp 160-
173.
Dupuy L, Urosevic D, Greney H, Quaglia W, Pigini M, Brasili L, Dontenwill M and
Bousquet P (2004) I(1) Imidazoline Receptor-Mediated Effects on Apoptotic
Processes in PC12 Cells. Cell Death Differ.
Edwards L and Ernsberger P (2003) The I(1)-Imidazoline Receptor in PC12
Pheochromocytoma Cells Reverses NGF-Induced ERK Activation and Induces
MKP-2 Phosphatase. Brain Res 980: pp 71-79.
Edwards L, Fishman D, Horowitz P, Bourbon N, Kester M and Ernsberger P
(2001) The I1-Imidazoline Receptor in PC12 Pheochromocytoma Cells Activates
Protein Kinases C, Extracellular Signal-Regulated Kinase (ERK) and C-Jun N-
Terminal Kinase (JNK). J Neurochem 79: pp 931-940.
172 Eglen RM, Hudson A L, Kendall D A, Nutt D J, Morgan N G, Wilson V G and
Dillon M P (1998) 'Seeing Through a Glass Darkly': Casting Light on Imidazoline
'I' Sites. Trends Pharmacol Sci 19: pp 381-390.
Einhorn D, Reaven G M, Cobin R H, Ford E, Ganda O P, Handelsman Y,
Hellman R, Jellinger P S, Kendall D, Krauss R M, Neufeld N D, Petak S M,
Rodbard H W, Seibel J A, Smith D A and Wilson P W (2003) American College of Endocrinology Position Statement on the Insulin Resistance Syndrome.
Endocr Pract 9: pp 237-252.
Eisenach JC, Castro M I, Dewan D M and Rose J C (1989) Epidural Clonidine
Analgesia in Obstetrics: Sheep Studies. Anesthesiology 70: pp 51-56.
Ernsberger P (1998) Arachidonic Acid Release From PC12 Pheochromocytoma
Cells Is Regulated by I1-Imidazoline Receptors. J Auton Nerv Syst 72: pp 147-
154.
Ernsberger P (1999) The I1-Imidazoline Receptor and Its Cellular Signaling
Pathways. Ann N Y Acad Sci 881: pp 35-53.
Ernsberger P (2000) Pharmacology of Moxonidine: an I1-Imidazoline Receptor
Agonist. J Cardiovasc Pharmacol 35: pp S27-S41.
173 Ernsberger P, Damon T H, Graff L M, Schafer S G and Christen M O (1993)
Moxonidine, a Centrally Acting Antihypertensive Agent, Is a Selective Ligand for
I1-Imidazoline Sites. J Pharmacol Exp Ther 264: pp 172-182.
Ernsberger P, Friedman J E and Koletsky R J (1997) The I1-Imidazoline
Receptor: From Binding Site to Therapeutic Target in Cardiovascular Disease. J
Hypertens Suppl 15: pp S9-23.
Ernsberger P, Graves M E, Graff L M, Zakieh N, Nguyen P, Collins L A,
Westbrooks K L and Johnson G G (1995) I1-Imidazoline Receptors. Definition,
Characterization, Distribution, and Transmembrane Signaling. Ann N Y Acad Sci
763: pp 22-42.
Ernsberger P and Haxhiu M A (1997) The I1-Imidazoline-Binding Site Is a
Functional Receptor Mediating Vasodepression Via the Ventral Medulla. Am J
Physiol 273: pp R1572-R1579.
Ernsberger P, Ishizuka T, Liu S, Farrell C J, Bedol D, Koletsky R J and Friedman
J E (1999a) Mechanisms of Antihyperglycemic Effects of Moxonidine in the
Obese Spontaneously Hypertensive Koletsky Rat (SHROB). J Pharmacol Exp
Ther 288: pp 139-147.
Ernsberger P, Ishizuka T, Liu S, Farrell C J, Bedol D, Koletsky R J and Friedman
J E (1999b) Mechanisms of Antihyperglycemic Effects of Moxonidine in the
174 Obese Spontaneously Hypertensive Koletsky Rat (SHROB). J Pharmacol Exp
Ther 288: pp 139-147.
Ernsberger P, Koletsky R J, Baskin J S and Foley M (1994) Refeeding
Hypertension in Obese Spontaneously Hypertensive Rats. Hypertension 24: pp
699-705.
Ernsberger P, Koletsky R J, Collins L A and Bedol D (1996b) Sympathetic
Nervous System in Salt-Sensitive and Obese Hypertension: Amelioration of
Multiple Abnormalities by a Central Sympatholytic Agent. Cardiovasc Drugs Ther
10 Suppl 1: pp 275-282.
Ernsberger P, Koletsky R J, Collins L A and Bedol D (1996a) Sympathetic
Nervous System in Salt-Sensitive and Obese Hypertension: Amelioration of
Multiple Abnormalities by a Central Sympatholytic Agent. Cardiovasc Drugs Ther
10 Suppl 1: pp 275-282.
Ernsberger P, Koletsky R J and Friedman J E (1999c) Molecular Pathology in the
Obese Spontaneous Hypertensive Koletsky Rat: a Model of Syndrome X. Ann N
Y Acad Sci 892: pp 272-288.
Ernsberger P, Meeley M P, Mann J J and Reis D J (1987) Clonidine Binds to
Imidazole Binding Sites As Well As Alpha 2-Adrenoceptors in the Ventrolateral
Medulla. Eur J Pharmacol 134: pp 1-13.
175 Ernsberger, P., Velliquette, R. A., Johnson, J. L., and Koletsky, R. J.
Improvements in glucose tolerance and insulin resistance following weight loss
persist after regain of the lost weight. Obes.Res. 13, A53. 2005.
Ref Type: Abstract
Escriba PV, Ozaita A, Miralles A, Reis D J and Garcia-Sevilla J A (1995)
Molecular Characterization and Isolation of a 45-Kilodalton Imidazoline Receptor
Protein From the Rat Brain. Brain Res Mol Brain Res 32: pp 187-196.
Escriba PV, Sastre M, Wang H, Regunathan S, Reis D J and Garcia-Sevilla J A
(1994) Immunodetection of Putative Imidazoline Receptor Proteins in the Human
and Rat Brain and Other Tissues. Neurosci Lett 178: pp 81-84.
Esler M, Rumantir M, Wiesner G, Kaye D, Hastings J and Lambert G (2001)
Sympathetic Nervous System and Insulin Resistance: From Obesity to Diabetes.
Am J Hypertens 14: pp 304S-309S.
Evinger MJ, Ernsberger P, Regunathan S and Reis D J (1995) Regulation of
Phenylethanolamine N-Methyltransferase Gene Expression by Imidazoline
Receptors in Adrenal Chromaffin Cells. J Neurochem 65: pp 988-997.
Fedorak RN, Field M and Chang E B (1985) Treatment of Diabetic Diarrhea With
Clonidine. Ann Intern Med 102: pp 197-199.
176 Felsen D, Ernsberger P, Sutaria P M, Nejat R J, Nguyen P, May M, Breslin D S,
Marion D N and Vaughan E D, Jr. (1994) Identification, Localization and
Functional Analysis of Imidazoline and Alpha Adrenergic Receptors in Canine
Prostate. J Pharmacol Exp Ther 268: pp 1063-1071.
Ford ES, Giles W H and Dietz W H (2002) Prevalence of the Metabolic
Syndrome Among US Adults: Findings From the Third National Health and
Nutrition Examination Survey. JAMA 287: pp 356-359.
Frank HJ, Davidson M B and Serbin P A (1981) Insulin Binding and Action in
Isolated Rat Hepatocytes: Evidence for Spare Receptors. Metabolism 30: pp
1159-1164.
Friedman JE, de Vente J E, Peterson R G and Dohm G L (1991) Altered
Expression of Muscle Glucose Transporter GLUT-4 in Diabetic Fatty Zucker Rats
(ZDF/Drt-Fa). Am J Physiol 261: pp E782-E788.
Friedman JE, Ishizuka T, Liu S, Bedol D, Koletsky R J and Ernsberger P (1997a)
Metabolic Consequences of a Nonsense Mutation in the Leptin Receptor Gene
(Fak) in the Obese Spontaneously Hypertensive Koletsky Rat (SHROB). Exp Clin
Endocrinol Diabetes 105 (Suppl. 3): pp 82-84.
177 Friedman JE, Ishizuka T, Liu S, Farrell C J, Bedol D, Koletsky R J, Kaung H L and Ernsberger P (1997b) Reduced Insulin Receptor Signaling in the Obese
Spontaneously Hypertensive Koletsky Rat. Am J Physiol 273: pp E1014-E1023.
Friedman JE, Ishizuka T, Liu S, Farrell C J, Bedol D, Koletsky R J, Kaung H L and Ernsberger P (1997d) Reduced Insulin Receptor Signaling in the Obese
Spontaneously Hypertensive Koletsky Rat. Am J Physiol 273: pp E1014-E1023.
Friedman JE, Ishizuka T, Liu S, Farrell C J, Bedol D, Koletsky R J, Kaung H L and Ernsberger P (1997c) Reduced Insulin Receptor Signaling in the Obese
Spontaneously Hypertensive Koletsky Rat. Am J Physiol 273: pp E1014-E1023.
Friedman JE, Ishizuka T, Liu S, Farrell C J, Koletsky R J, Bedol D and
Ernsberger P (1998) Anti-Hyperglycemic Activity of Moxonidine: Metabolic and
Molecular Effects in Obese Spontaneously Hypertensive Rats. Blood Press
Suppl 3: pp 32-39.
Galve-Roperh I, Malpartida J M, Haro A and Diaz-Laviada I (1996) Addition of
Phosphatidylcholine-Phospholipase C Induces Cellular Redistribution and
Phosphorylation of Protein Kinase C Zeta in C 6 Glial Cells. Neurosci Lett 219: pp 68-70.
178 Garcia-Sevilla JA, Escriba P V, Busquets X, Walzer C and Guimon J (1996)
Platelet Imidazoline Receptors and Regulatory G Proteins in Patients With Major
Depression. Neuroreport 8: pp 169-172.
Garcia-Sevilla JA, Escriba P V, Walzer C, Bouras C and Guimon J (1998)
Imidazoline Receptor Proteins in Brains of Patients With Alzheimer's Disease.
Neurosci Lett 247: pp 95-98.
Garcia-Sevilla JA, Ventayol P, Perez V, Rubovszky G, Puigdemont D, Ferrer-
Alcon M, Andreoli A, Guimon J and Alvarez E (2004) Regulation of Platelet Alpha
2A-Adrenoceptors, Gi Proteins and Receptor Kinases in Major Depression:
Effects of Mirtazapine Treatment. Neuropsychopharmacology 29: pp 580-588.
Gentili F, Bousquet P, Brasili L, Dontenwill M, Feldman J, Ghelfi F, Giannella M,
Piergentili A, Quaglia W and Pigini M (2003) Imidazoline Binding Sites (IBS)
Profile Modulation: Key Role of the Bridge in Determining I1-IBS or I2-IBS
Selectivity Within a Series of 2-Phenoxymethylimidazoline Analogues. J Med
Chem 46: pp 2169-2176.
Gleave ME and Monia B P (2005) Antisense Therapy for Cancer. Nat Rev
Cancer 5: pp 468-479.
179 Green A (1986) The Insulin-Like Effect of Sodium Vanadate on Adipocyte
Glucose Transport Is Mediated at a Post-Insulin-Receptor Level. Biochem J 238: pp 663-669.
Greene LA and Tischler A S (1976) Establishment of a Noradrenergic Clonal
Line of Rat Adrenal Pheochromocytoma Cells Which Respond to Nerve Growth
Factor. Proc Natl Acad Sci U S A 73: pp 2424-2428.
Greenspan BJ (1993) Quality Considerations in Inhalation Toxicology. Qual
Assur 2: pp 105-111.
Greney H, Bennai F, Molines A, Belcourt A, Dontenwill M and Bousquet P (1994)
Isolation of a Human Cerebral Imidazoline-Specific Binding Protein. Eur J
Pharmacol 265: pp R1-R2.
Greney H, Urosevic D, Schann S, Dupuy L, Bruban V, Ehrhardt J D, Bousquet P and Dontenwill M (2002) [125I]2-(2-Chloro-4-Iodo-Phenylamino)-5-Methyl-
Pyrroline (LNP 911), a High-Affinity Radioligand Selective for I1 Imidazoline
Receptors. Mol Pharmacol 62: pp 181-191.
Greven J and Bronewski-Schwarzer B (2001) Site of Action of Moxonidine in the
Rat Nephron. Naunyn Schmiedebergs Arch Pharmacol 364: pp 496-500.
180 Haenni A and Lithell H (1999) Moxonidine Improves Insulin Sensitivity in Insulin-
Resistant Hypertensives. J Hypertens Suppl 17: pp S29-S35.
Haffner SM (2006) The Metabolic Syndrome: Inflammation, Diabetes Mellitus, and Cardiovascular Disease. Am J Cardiol 97: pp 3A-11A.
Haffner SM, Valdez R A, Hazuda H P, Mitchell B D, Morales P A and Stern M P
(1992) Prospective Analysis of the Insulin-Resistance Syndrome (Syndrome X).
Diabetes 41: pp 715-722.
Hanada M, Feng J and Hemmings B A (2004) Structure, Regulation and
Function of PKB/AKT--a Major Therapeutic Target. Biochim Biophys Acta 1697: pp 3-16.
Harmon RJ and Riggs P D (1996) Clonidine for Posttraumatic Stress Disorder in
Preschool Children. J Am Acad Child Adolesc Psychiatry 35: pp 1247-1249.
Hartmann M and Isler H (1939) Chemische Konstitution und Pharmakologische
Wirksamkeit von in 2-Stellung Substituierten Imidazolinen. Naunyn
Schmiedebergs Arch Exp Pathol Pharmakol 192: pp 141-154
Haxhiu MA, Dreshaj I, Schafer S G and Ernsberger P (1994) Selective
Antihypertensive Action of Moxonidine Is Mediated Mainly by I1-Imidazoline
181 Receptors in the Rostral Ventrolateral Medulla. J Cardiovasc Pharmacol 24
Suppl 1: pp S1-S8.
He MM, Abraham T L, Lindsay T J, Chay S H, Czeskis B A and Shipley L A
(2000) Metabolism and Disposition of Moxonidine in Fischer 344 Rats. Drug
Metab Dispos 28: pp 446-459.
Heemskerk FM, Dontenwill M, Greney H, Vonthron C and Bousquet P (1998)
Evidence for the Existence of Imidazoline-Specific Binding Sites in Synaptosomal
Plasma Membranes of the Bovine Brainstem. J Neurochem 71: pp 2193-2202.
Henriksen EJ, Jacob S, Fogt D L, Youngblood E B and Godicke J (1997)
Antihypertensive Agent Moxonidine Enhances Muscle Glucose Transport in
Insulin-Resistant Rats. Hypertension 30: pp 1560-1565.
Hieble JP, Sulpizio A C, Edwards R, Chapman H, Young P, Roberts S P,
Blackburn T P, Wood M D, Shah D H, Demarinis R M and . (1991) Additional
Evidence for Functional Subclassification of Alpha-2 Adrenoceptors Based on a
New Selective Antagonist, SK&F 104856. J Pharmacol Exp Ther 259: pp 643-
652.
Hoefke W and Kobinger W (1966) [Pharmacological Effects of 2-(2,6-
Dichlorophenylamino)-2-Imidazoline Hydrochloride, a New, Antihypertensive
Substance]. Arzneimittelforschung 16: pp 1038-1050.
182 Holgado-Madruga M, Emlet D R, Moscatello D K, Godwin A K and Wong A J
(1996) A Grb2-Associated Docking Protein in EGF- and Insulin-Receptor
Signalling. Nature 379: pp 560-564.
Hresko RC and Mueckler M (2005) MTOR.RICTOR Is the Ser473 Kinase for
Akt/Protein Kinase B in 3T3-L1 Adipocytes. J Biol Chem 280: pp 40406-40416.
Hsu SY, Kudo M, Chen T, Nakabayashi K, Bhalla A, van der Spek P J, van Duin
M and Hsueh A J (2000) The Three Subfamilies of Leucine-Rich Repeat-
Containing G Protein-Coupled Receptors (LGR): Identification of LGR6 and
LGR7 and the Signaling Mechanism for LGR7. Mol Endocrinol 14: pp 1257-1271.
Huang SS, Khosrof S A, Koletsky R J, Benetz B A and Ernsberger P (1995)
Characterization of Retinal Vascular Abnormalities in Lean and Obese
Spontaneously Hypertensive Rats. Clin Exp Pharmacol Physiol Suppl 22: pp
S129-S131.
Iannello S, Campione R and Belfiore F (1998) Response of Insulin, Glucagon,
Lactate, and Nonesterified Fatty Acids to Glucose in Visceral Obesity With and
Without NIDDM: Relationship to Hypertension. Mol Genet Metab 63: pp 214-223.
Ishizuka T, Ernsberger P, Liu S, Bedol D, Lehman T M, Koletsky R J and
Friedman J E (1998a) Phenotypic Consequences of a Nonsense Mutation in the
183 Leptin Receptor Gene (Fak) in Obese Spontaneously Hypertensive Koletsky
Rats (SHROB). J Nutr 128: pp 2299-2306.
Ishizuka T, Ernsberger P, Liu S, Bedol D, Lehman T M, Koletsky R J and
Friedman J E (1998b) Phenotypic Consequences of a Nonsense Mutation in the
Leptin Receptor Gene (Fak) in Obese Spontaneously Hypertensive Koletsky
Rats (SHROB). J Nutr 128: pp 2299-2306.
Ivanov TR, Zhu H, Regunathan S, Reis D J, Dontenwill M, Vonthron C, Bousquet
P and Piletz J E (1998) Co-Detection by Two Imidazoline Receptor Protein
Antisera of a Novel 85 Kilodalton Protein. Biochem Pharmacol 55: pp 649-655.
Jackowski S, Xu X X and Rock C O (1997) Phosphatidylcholine Signaling in
Response to CSF-1. Mol Reprod Dev 46: pp 24-30.
Jackson WJ and Buccafusco J J (1991) Clonidine Enhances Delayed Matching- to-Sample Performance by Young and Aged Monkeys. Pharmacol Biochem
Behav 39: pp 79-84.
Jacob S, Klimm H J, Rett K, Helsberg K, Haring H U and Godicke J (2004)
Effects of Moxonidine Vs. Metoprolol on Blood Pressure and Metabolic Control in
Hypertensive Subjects With Type 2 Diabetes. Exp Clin Endocrinol Diabetes 112: pp 315-322.
184 Jaselskis CA, Cook E H, Jr., Fletcher K E and Leventhal B L (1992) Clonidine
Treatment of Hyperactive and Impulsive Children With Autistic Disorder. J Clin
Psychopharmacol 12: pp 322-327.
Jovanovska V, Eitle E and Harris P J (2004) Effects of Rilmenidine on Proximal
Tubular Fluid Absorption in Rats. Clin Exp Pharmacol Physiol 31: pp 348-353.
Juan CC, Chien Y, Wu L Y, Yang W M, Chang C L, Lai Y H, Ho P H, Kwok C F and Ho L T (2005) Angiotensin II Enhances Insulin Sensitivity in Vitro and in Vivo.
Endocrinology 146: pp 2246-2254.
Kajava AV (1998) Structural Diversity of Leucine-Rich Repeat Proteins. J Mol
Biol 277: pp 519-527.
Kaplan NM (1989) The Deadly Quartet. Upper-Body Obesity, Glucose
Intolerance, Hypertriglyceridemia, and Hypertension. Arch Intern Med 149: pp
1514-1520.
Karlsson HK, Zierath J R, Kane S, Krook A, Lienhard G E and Wallberg-
Henriksson H (2005) Insulin-Stimulated Phosphorylation of the Akt Substrate
AS160 Is Impaired in Skeletal Muscle of Type 2 Diabetic Subjects. Diabetes 54: pp 1692-1697.
185 Kasiske BL, Cleary M P, O'Donnell M P and Keane W F (1985) Effects of
Genetic Obesity on Renal Structure and Function in the Zucker Rat. J Lab Clin
Med 106: pp 598-604.
Khosrof, S. A., Huang, S. S., Benetz, B. A., Bardenstein, D. S., Ernsberger, P., and Koletsky, R. J. Characterization of retinal vascular abnormalities in lean and obese SHR: A new model of type 2 diabetic retinopathy.
Invest.Ophthalmol.Vis.Sci. 36[4], S172. 1995.
Ref Type: Abstract
Klein C, Morton N, Kelley S and Metz S (1985) Transdermal Clonidine Therapy in
Elderly Mild Hypertensives: Effects on Blood Pressure, Plasma Norepinephrine and Fasting Plasma Glucose. J Hypertens Suppl 3: pp S81-S84.
Klein J, Perwitz N, Kraus D and Fasshauer M (2006) Adipose Tissue As Source and Target for Novel Therapies. Trends Endocrinol Metab 17: pp 26-32.
Kobe B and Kajava A V (2001) The Leucine-Rich Repeat As a Protein
Recognition Motif. Curr Opin Struct Biol 11: pp 725-732.
Kohn AD, Summers S A, Birnbaum M J and Roth R A (1996) Expression of a
Constitutively Active Akt Ser/Thr Kinase in 3T3-L1 Adipocytes Stimulates
Glucose Uptake and Glucose Transporter 4 Translocation. J Biol Chem 271: pp
31372-31378.
186 Koletsky RJ, Boccia J and Ernsberger P (1995) Acceleration of Renal Disease in
Obese SHR by Exacerbation of Hypertension. Clin Exp Pharmacol Physiol Suppl
22: pp S254-S256.
Koletsky S (1973) Obese Spontaneously Hypertensive Rats--a Model for Study of Atherosclerosis. Exp Mol Pathol 19: pp 53-60.
Koletsky S (1975a) Animal Model: Obese Hypertensive Rat. Am J Pathol 81: pp
463-466.
Koletsky S (1975b) Pathologic Findings and Laboratory Data in a New Strain of
Obese Hypertensive Rats. Am J Pathol 80: pp 129-142.
Koletsky S and Puterman D I (1976) Effect of Low Calorie Diet on the
Hyperlipidemia, Hypertension, and Life Span of Genetically Obese Rats. Proc
Soc Exp Biol Med 151: pp 368-371.
Kono T and Barham F W (1971) Insulin-Like Effects of Trypsin on Fat Cells.
Localization of the Metabolic Steps and the Cellular Site Affected by the Enzyme.
J Biol Chem 246: pp 6204-6209.
Kumar PD, Nartsupha C and Koletsky R J (2001) Delayed Syndrome of
Inappropriate Antidiuretic Hormone Secretion 1 Year After a Head Injury. Ann
Intern Med 135: pp 932-933.
187 Kuzuya T, Nakagawa S, Satoh J, Kanazawa Y, Iwamoto Y, Kobayashi M, Nanjo
K, Sasaki A, Seino Y, Ito C, Shima K, Nonaka K and Kadowaki T (2002) Report of the Committee on the Classification and Diagnostic Criteria of Diabetes
Mellitus. Diabetes Res Clin Pract 55: pp 65-85.
Laubie M, Poignant J C, Scuvee-Moreau J, Dabire H, Dresse A and Schmitt H
(1985) Pharmacological Properties of (N-Dicyclopropylmethyl) Amino-2-
Oxazoline (S 3341), an Alpha-2 Adrenoceptor Agonist. J Pharmacol 16: pp 259-
278.
Laubie M and Schmitt H (1977) Sites of Action of Clonidine: Centrally Mediated
Increase in Vagal Tone, Centrally Mediated Hypotensive and Sympatho-
Inhibitory Effects. Prog Brain Res 47: pp 337-348.
Laviada ID, Baudet C, Galve-Roperh I, Naveilhan P and Brachet P (1995)
Phosphatidylcholine-Phospholipase C Mediates the Induction of Nerve Growth
Factor in Cultured Glial Cells. FEBS Lett 364: pp 301-304.
Leckman JF, Hardin M T, Riddle M A, Stevenson J, Ort S I and Cohen D J (1991)
Clonidine Treatment of Gilles De La Tourette's Syndrome. Arch Gen Psychiatry
48: pp 324-328.
188 Lee CM, Onesime D, Reddy C D, Dhanasekaran N and Reddy E P (2002) JLP: A
Scaffolding Protein That Tethers JNK/P38MAPK Signaling Modules and
Transcription Factors. Proc Natl Acad Sci U S A 99: pp 14189-14194.
Li Y, Maher P and Schubert D (1998) Phosphatidylcholine-Specific
Phospholipase C Regulates Glutamate-Induced Nerve Cell Death. Proc Natl
Acad Sci U S A 95: pp 7748-7753.
Liashko KI and Dreval' A V (1973) [Current Concepts of Insulin Resistance and
Causes of Its Development]. Ter Arkh 45: pp 9-15.
Liedtke CM, Furin J and Ernsberger P (1993) Alpha 2-Adrenergic, but Not
Imidazole, Agonists Activate NaCl Cotransport in Rabbit Tracheal Epithelial Cells.
Am J Physiol 264: pp C568-C576.
Lim KP and Hong W (2004) Human Nischarin/Imidazoline Receptor Antisera-
Selected Protein Is Targeted to the Endosomes by a Combined Action of a PX
Domain and a Coiled-Coil Region. J Biol Chem 279: pp 54770-54782.
Limon-Boulez I, Tesson F, Gargalidis-Moudanos C and Parini A (1996) I2-
Imidazoline Binding Sites: Relationship With Different Monoamine Oxidase
Domains and Identification of Histidine Residues Mediating Ligand Binding
Regulation by H+1. J Pharmacol Exp Ther 276: pp 359-364.
189 Lin CY, Lynch G and Gall C M (2005) AMPA Receptor Stimulation Increases
Alpha5beta1 Integrin Surface Expression, Adhesive Function and Signaling. J
Neurochem 94: pp 531-546.
Linthicum DS, Bolger M B, Kussie P H, Albright G M, Linton T A, Combs S and
Marchetti D (1988) Analysis of Idiotypic and Anti-Idiotypic Antibodies As Models of Receptor and Ligand. Clin Chem 34: pp 1676-1680.
Lithell HO (1997) Considerations in the Treatment of Insulin Resistance and
Related Disorders With a New Sympatholytic Agent. J Hypertens Suppl 15: pp
S39-S42.
Livingstone C and Gould G W (1995) Insulin Resistance in Diabetes Mellitus.
Defective Insulin-Regulatable Glucose Transport Plays an Important Role. Scott
Med J 40: pp 37-39.
Lock P, Abram C L, Gibson T and Courtneidge S A (1998) A New Method for
Isolating Tyrosine Kinase Substrates Used to Identify Fish, an SH3 and PX
Domain-Containing Protein, and Src Substrate. EMBO J 17: pp 4346-4357.
Machleidt T, Kramer B, Adam D, Neumann B, Schutze S, Wiegmann K and
Kronke M (1996) Function of the P55 Tumor Necrosis Factor Receptor "Death
Domain" Mediated by Phosphatidylcholine-Specific Phospholipase C. J Exp Med
184: pp 725-733.
190 MacKinnon AC, Stewart M, Olverman H J, Spedding M and Brown C M (1993)
[3H]p-Aminoclonidine and [3H]Idazoxan Label Different Populations of
Imidazoline Sites on Rat Kidney. Eur J Pharmacol 232: pp 79-87.
Marchand-Brustel Y, Gual P, Gremeaux T, Gonzalez T, Barres R and Tanti J F
(2003) Fatty Acid-Induced Insulin Resistance: Role of Insulin Receptor Substrate
1 Serine Phosphorylation in the Retroregulation of Insulin Signalling. Biochem
Soc Trans 31: pp 1152-1156.
Mark AL, Correia M, Morgan D A, Shaffer R A and Haynes W G (1999) State-of- the-Art-Lecture: Obesity-Induced Hypertension: New Concepts From the
Emerging Biology of Obesity. Hypertension 33: pp 537-541.
Marshall S, Heidenreich K A and Horikoshi H (1985) Stoichiometric Translocation of Adipocyte Insulin Receptors From the Cell-Surface to the Cell-Interior. Studies
Using a Novel Method to Rapidly Remove Detergent and Concentrate Soluble
Receptors. J Biol Chem 260: pp 4128-4135.
Miranti CK, Ohno S and Brugge J S (1999) Protein Kinase C Regulates Integrin-
Induced Activation of the Extracellular Regulated Kinase Pathway Upstream of
Shc. J Biol Chem 274: pp 10571-10581.
191 Molderings GJ, Moura D, Fink K, Bonisch H and Gothert M (1993) Binding of
[3H]Clonidine to I1-Imidazoline Sites in Bovine Adrenal Medullary Membranes.
Naunyn Schmiedebergs Arch Pharmacol 348: pp 70-76.
Monks LK, Cosgrove K E, Dunne M J, Ramsden C A, Morgan N G and Chan S L
(1999) Affinity Isolation of Imidazoline Binding Proteins From Rat Brain Using 5-
Amino-Efaroxan As a Ligand. FEBS Lett 447: pp 61-64.
Mora A, Komander D, van Aalten D M and Alessi D R (2004) PDK1, the Master
Regulator of AGC Kinase Signal Transduction. Semin Cell Dev Biol 15: pp 161-
170.
Motulsky HJ and Ransnas L A (1987) Fitting Curves to Data Using Nonlinear
Regression: a Practical and Nonmathematical Review. FASEB J 1: pp 365-374.
Murphy E and Nolan J J (2000) Insulin Sensitiser Drugs. Expert Opin Investig
Drugs 9: pp 1347-1361.
Musgrave IF, Dehle F C and Piletz J (2003) Assembly of PRR-Containing
Receptors on Scaffolds: a Model for Imidazoline I(1)-Receptor Action. Ann N Y
Acad Sci 1009: pp 413-418.
Musgrave IF, Krautwurst D and Schultz G (1996) Imidazoline Binding Sites and
Signal Transduction Pathways. Clin Exp Pharmacol Physiol 23: pp 990-994.
192 Nunes ML, Liptakova S, Veliskova J, Sperber E F and Moshe S L (2000)
Malnutrition Increases Dentate Granule Cell Proliferation in Immature Rats After
Status Epilepticus. Epilepsia 41 Suppl 6: pp S48-S52.
Okada T, Kawano Y, Sakakibara T, Hazeki O and Ui M (1994) Essential Role of
Phosphatidylinositol 3-Kinase in Insulin-Induced Glucose Transport and
Antilipolysis in Rat Adipocytes. Studies With a Selective Inhibitor Wortmannin. J
Biol Chem 269: pp 3568-3573.
Okamoto K and Aoki K (1963) Development of a Strain of Spontaneously
Hypertensive Rats. Jpn Circ J 27: pp 282-293.
Okamoto PM, Herskovits J S and Vallee R B (1997) Role of the Basic, Proline-
Rich Region of Dynamin in Src Homology 3 Domain Binding and Endocytosis. J
Biol Chem 272: pp 11629-11635.
Olfson M (2004) New Options in the Pharmacological Management of Attention-
Deficit/Hyperactivity Disorder. Am J Manag Care 10: pp S117-S124.
Osaki M, Oshimura M and Ito H (2004) PI3K-Akt Pathway: Its Functions and
Alterations in Human Cancer. Apoptosis 9: pp 667-676.
Park YW, Zhu S, Palaniappan L, Heshka S, Carnethon M R and Heymsfield S B
(2003) The Metabolic Syndrome: Prevalence and Associated Risk Factor
193 Findings in the US Population From the Third National Health and Nutrition
Examination Survey, 1988-1994. Arch Intern Med 163: pp 427-436.
Pele-Tounian A, Wang X, Rondu F, Lamouri A, Touboul E, Marc S, Dokhan R,
Pfeiffer B, Manechez D, Renard P, Guardiola-Lemaitre B, Godfroid J J, Penicaud
L and Ktorza A (1998) Potent Antihyperglycaemic Property of a New Imidazoline
Derivative S-22068 (PMS 847) in a Rat Model of NIDDM. Br J Pharmacol 124: pp
1591-1596.
Piletz JE, Andorn A C, Unnerstall J R and Halaris A (1991) Binding of [3H]-p-
Aminoclonidine to Alpha 2-Adrenoceptor States Plus a Non-Adrenergic Site on
Human Platelet Plasma Membranes. Biochem Pharmacol 42: pp 569-584.
Piletz JE, Halaris A and Ernsberger P R (1994) Psychopharmacology of
Imidazoline and Alpha 2-Adrenergic Receptors: Implications for Depression. Crit
Rev Neurobiol 9: pp 29-66.
Piletz JE, Ivanov T R, Sharp J D, Ernsberger P, Chang C H, Pickard R T, Gold G,
Roth B, Zhu H, Jones J C, Baldwin J and Reis D J (2000) Imidazoline Receptor
Antisera-Selected (IRAS) CDNA: Cloning and Characterization. DNA Cell Biol 19: pp 319-329.
194 Piletz JE, Jones J C, Zhu H, Bishara O and Ernsberger P (1999) Imidazoline
Receptor Antisera-Selected CDNA Clone and MRNA Distribution. Ann N Y Acad
Sci 881: pp 1-7.
Piletz JE, Wang G and Zhu H (2003) Cell Signaling by Imidazoline-1 Receptor
Candidate, IRAS, and the Nischarin Homologue. Ann N Y Acad Sci 1009: pp
392-399.
Podrid PJ and Myerburg R J (2005) Epidemiology and Stratification of Risk for
Sudden Cardiac Death. Clin Cardiol 28: pp I3-11.
Posey DJ and McDougle C J (2001) Pharmacotherapeutic Management of
Autism. Expert Opin Pharmacother 2: pp 587-600.
Prell GD, Martinelli G P, Holstein G R, Matulic-Adamic J, Watanabe K A, Chan S
L, Morgan N G, Haxhiu M A and Ernsberger P (2004) Imidazoleacetic Acid-
Ribotide: an Endogenous Ligand That Stimulates Imidazol(in)e Receptors. Proc
Natl Acad Sci U S A 101: pp 13677-13682.
Rachdaoui N, Sebastian B M and Nagy L E (2003) Chronic Ethanol Feeding
Impairs Endothelin-1-Stimulated Glucose Uptake Via Decreased G Alpha 11
Expression in Rat Adipocytes. Am J Physiol Endocrinol Metab 285: pp E303-
E310.
195 Raddatz R, Parini A and Lanier S M (1995) Imidazoline/Guanidinium Binding
Domains on Monoamine Oxidases. Relationship to Subtypes of Imidazoline-
Binding Proteins and Tissue-Specific Interaction of Imidazoline Ligands With
Monoamine Oxidase B. J Biol Chem 270: pp 27961-27968.
Reaven GM (1988) Banting Lecture 1988. Role of Insulin Resistance in Human
Disease. Diabetes 37: pp 1595-1607.
Reid JL (2000) Rilmenidine: a Clinical Overview. Am J Hypertens 13: pp 106S-
111S.
Robin AL (1995) Questions Concerning the Role of Apraclonidine in the
Management of Glaucoma. Arch Ophthalmol 113: pp 712-714.
Ruggiero DA, Regunathan S, Wang H, Milner T A and Reis D J (1998)
Immunocytochemical Localization of an Imidazoline Receptor Protein in the
Central Nervous System. Brain Res 780: pp 270-293.
Rui L, Aguirre V, Kim J K, Shulman G I, Lee A, Corbould A, Dunaif A and White
M F (2001) Insulin/IGF-1 and TNF-Alpha Stimulate Phosphorylation of IRS-1 at
Inhibitory Ser307 Via Distinct Pathways. J Clin Invest 107: pp 181-189.
Saito K, Kigawa T, Koshiba S, Sato K, Matsuo Y, Sakamoto A, Takagi T,
Shirouzu M, Yabuki T, Nunokawa E, Seki E, Matsuda T, Aoki M, Miyata Y,
196 Hirakawa N, Inoue M, Terada T, Nagase T, Kikuno R, Nakayama M, Ohara O,
Tanaka A and Yokoyama S (2004) The CAP-Gly Domain of CYLD Associates
With the Proline-Rich Sequence in NEMO/IKKgamma. Structure 12: pp 1719-
1728.
Sakurai-Yamashita Y, Yamashita K, Kataoka Y, Himeno A, Niwa M and
Taniyama K (1997) Expression of Endothelin Receptors and Nitric Oxide
Synthase in the Brain of Stroke-Prone Spontaneously Hypertensive Rats With
Cerebral Apoplexy. Brain Res 756: pp 61-67.
Sanjuliani AF, de Abreu V G and Francischetti E A (2006) Selective Imidazoline
Agonist Moxonidine in Obese Hypertensive Patients. Int J Clin Pract 60: pp 621-
629.
Sano H, Liu S C, Lane W S, Piletz J E and Lienhard G E (2002) Insulin Receptor
Substrate 4 Associates With the Protein IRAS. J Biol Chem 277: pp 19439-19447.
Sarbassov DD, Guertin D A, Ali S M and Sabatini D M (2005) Phosphorylation and Regulation of Akt/PKB by the Rictor-MTOR Complex. Science 307: pp 1098-
1101.
Sasaoka T, Draznin B, Leitner J W, Langlois W J and Olefsky J M (1994) Shc Is the Predominant Signaling Molecule Coupling Insulin Receptors to Activation of
197 Guanine Nucleotide Releasing Factor and P21ras-GTP Formation. J Biol Chem
269: pp 10734-10738.
Schachter M (1999) Metabolic Effects of Moxonidine and Other Centrally Acting
Antihypertensives. Diabetes Obes Metab 1: pp 317-322.
Schmelzle T and Hall M N (2000) TOR, a Central Controller of Cell Growth. Cell
103: pp 253-262.
Schork NJ, Krieger J E, Trolliet M R, Franchini K G, Koike G, Krieger E M,
Lander E S, Dzau V J and Jacob H J (1995) A Biometrical Genome Search in
Rats Reveals the Multigenic Basis of Blood Pressure Variation. Genome Res 5: pp 164-172.
Schramek H (2002) MAP Kinases: From Intracellular Signals to Physiology and
Disease. News Physiol Sci 17: pp 62-67.
Separovic D, Kester M and Ernsberger P (1996) Coupling of I1-Imidazoline
Receptors to Diacylglyceride Accumulation in PC12 Rat Pheochromocytoma
Cells. Mol Pharmacol 49: pp 668-675.
Separovic D, Kester M, Haxhiu M A and Ernsberger P (1997) Activation of
Phosphatidylcholine-Selective Phospholipase C by I1-Imidazoline Receptors in
PC12 Cells and Rostral Ventrolateral Medulla. Brain Res 749: pp 335-339.
198 Shoelson SE, Marshall S, Horikoshi H, Kolterman O G, Rubenstein A H and
Olefsky J M (1986) Antiinsulin Receptor Antibodies in an Insulin-Dependent
Diabetic May Arise As Autoantiidiotypes. J Clin Endocrinol Metab 63: pp 56-61.
Sonne O (1985) Increased Inhibitory Potency of Free Fatty Acid-Poor Albumin on the Released and Activity of Insulin-Degrading Enzymes From Isolated Rat
Adipocytes and Hepatocytes. Anal Biochem 151: pp 109-117.
Squires PE, Hills C E, Rogers G J, Garland P, Farley S R and Morgan N G (2004)
The Putative Imidazoline Receptor Agonist, Harmane, Promotes Intracellular
Calcium Mobilisation in Pancreatic Beta-Cells. Eur J Pharmacol 501: pp 31-39.
Starchina I, Parfenov V A, Chazova I E, Sinitsyn V E, Pustovitova T S, Kolos I P and Ustiuzhanin D V (2005) [Cognitive Functions and Emotional State of Post
Stroke Patients on Antihypertensive Therapy]. Zh Nevrol Psikhiatr Im S S
Korsakova Suppl 15: pp 39-44.
Steffen G, Dendorfer A and Dominiak P (1995) Imidazoline Binding Sites on
PC12 Cells and Bovine Chromaffin Cells. Ann N Y Acad Sci 763: pp 157-162.
Stone LS, Fairbanks C A and Wilcox G L (2003) Moxonidine, a Mixed Alpha(2)-
Adrenergic and Imidazoline Receptor Agonist, Identifies a Novel Adrenergic
Target for Spinal Analgesia. Ann N Y Acad Sci 1009: pp 378-385.
199 Sun H, Zhang L, Wang A and Xue Z (2004) Prolonged Hypotensive Effect of
Human Tissue Kallikrein Gene Delivery and Recombinant Enzyme
Administration in Spontaneous Hypertension Rats. Exp Mol Med 36: pp 23-27.
Swislocki A and Tsuzuki A (1993) Insulin Resistance and Hypertension: Glucose
Intolerance, Hyperinsulinemia, and Elevated Free Fatty Acids in the Lean
Spontaneously Hypertensive Rat. Am J Med Sci 306: pp 282-286.
Szabo B (2002) Imidazoline Antihypertensive Drugs: a Critical Review on Their
Mechanism of Action. Pharmacol Ther 93: pp 1-35.
Takahashi N, Takahashi Y and Putnam F W (1985) Periodicity of Leucine and
Tandem Repetition of a 24-Amino Acid Segment in the Primary Structure of
Leucine-Rich Alpha 2-Glycoprotein of Human Serum. Proc Natl Acad Sci U S A
82: pp 1906-1910.
Takaya K, Ogawa Y, Hiraoka J, Hosoda K, Yamori Y, Nakao K and Koletsky R J
(1996) Nonsense Mutation of Leptin Receptor in the Obese Spontaneously
Hypertensive Koletsky Rat. Nat Genet 14: pp 130-131.
Tan E, Butkus A and Koletsky S (1976) Hepatic Cholesterol Metabolism in Vitro in the Obese Spontaneously Hypertensive, Hyperlipemic and Atherosclerotic Rat.
Exp Mol Pathol 25: pp 142-151.
200 Tanabe M, Kino Y, Honda M and Ono H (2006) Presynaptic I1-Imidazoline
Receptors Reduce GABAergic Synaptic Transmission in Striatal Medium Spiny
Neurons. J Neurosci 26: pp 1795-1802.
van der FA and Sonnenberg A (2001) Function and Interactions of Integrins. Cell
Tissue Res 305: pp 285-298.
Van Oekelen D, Luyten W H and Leysen J E (2003) Ten Years of Antisense
Inhibition of Brain G-Protein-Coupled Receptor Function. Brain Res Brain Res
Rev 42: pp 123-142.
Vaux D and Fuller S D (1991) The Use of Antiidiotype Antibodies for the
Characterization of Protein-Protein Interactions. Methods Cell Biol 34: pp 1-38.
Velliquette RA and Ernsberger P (2003a) Contrasting Metabolic Effects of
Antihypertensive Agents. J Pharmacol Exp Ther 307: pp 1104-1111.
Velliquette RA and Ernsberger P (2003b) The Role of I(1)-Imidazoline and
Alpha(2)-Adrenergic Receptors in the Modulation of Glucose Metabolism in the
Spontaneously Hypertensive Obese Rat Model of Metabolic Syndrome X. J
Pharmacol Exp Ther 306: pp 646-657.
Velliquette RA, Friedman J E, Shao J, Zhang B B and Ernsberger P (2005)
Therapeutic Actions of an Insulin Receptor Activator and a Novel Peroxisome
201 Proliferator-Activated Receptor Gamma Agonist in the Spontaneously
Hypertensive Obese Rat Model of Metabolic Syndrome X. J Pharmacol Exp Ther
314: pp 422-430.
Velliquette RA, Koletsky R J and Ernsberger P (2002b) Plasma Glucagon and
Free Fatty Acid Responses to a Glucose Load in the Obese Spontaneous
Hypertensive Rat (SHROB) Model of Metabolic Syndrome X. Exp Biol Med
(Maywood ) 227: pp 164-170.
Velliquette RA, Koletsky R J and Ernsberger P (2002c) Plasma Glucagon and
Free Fatty Acid Responses to a Glucose Load in the Obese Spontaneous
Hypertensive Rat (SHROB) Model of Metabolic Syndrome X. Exp Biol Med
(Maywood ) 227: pp 164-170.
Velliquette RA, Koletsky R J and Ernsberger P (2002a) Plasma Glucagon and
Free Fatty Acid Responses to a Glucose Load in the Obese Spontaneous
Hypertensive Rat (SHROB) Model of Metabolic Syndrome X. Exp Biol Med
(Maywood ) 227: pp 164-170.
Velliquette RA, Kossover R, Previs S F and Ernsberger P (2006a) Lipid-Lowering
Actions of Imidazoline Antihypertensive Agents in Metabolic Syndrome X.
Naunyn Schmiedebergs Arch Pharmacol 372: pp 300-312.
202 Velliquette RA, Kossover R, Previs S F and Ernsberger P (2006b) Lipid-Lowering
Actions of Imidazoline Antihypertensive Agents in Metabolic Syndrome X.
Naunyn Schmiedebergs Arch Pharmacol 372: pp 300-312.
Walker BR, Anderson J A and Edwards C R (1992) Clonidine Therapy for
Shapiro's Syndrome. Q J Med 82: pp 235-245.
Wan, D. P., Johnson, D. H., Johnson, J. L., Koletsky, R. J., Flask, C., and
Ernsberger, P. Magnetic resonance imaging of visceral and subcutaneous fat distribution in genetic versus dietary obesity. Obes.Res. 13, A172. 2005.
Ref Type: Abstract
Wang H, Regunathan S, Meeley M P and Reis D J (1992) Isolation and
Characterization of Imidazoline Receptor Protein From Bovine Adrenal
Chromaffin Cells. Mol Pharmacol 42: pp 792-801.
Wang H, Regunathan S, Ruggiero D A and Reis D J (1993) Production and
Characterization of Antibodies Specific for the Imidazoline Receptor Protein. Mol
Pharmacol 43: pp 509-515.
Whiteman EL, Cho H and Birnbaum M J (2002) Role of Akt/Protein Kinase B in
Metabolism. Trends Endocrinol Metab 13: pp 444-451.
203 Wikberg JE, Uhlen S and Chhajlani V (1992) Evidence That Drug Binding to
Non-Adrenergic [3H]-Idazoxan Binding Sites (I-Receptors) Occurs to Interacting or Interconvertible Affinity Forms of the Receptor. Pharmacol Toxicol 70: pp 208-
219.
Worby CA and Dixon J E (2002) Sorting Out the Cellular Functions of Sorting
Nexins. Nat Rev Mol Cell Biol 3: pp 919-931.
Xu Y, Seet L F, Hanson B and Hong W (2001) The Phox Homology (PX) Domain, a New Player in Phosphoinositide Signalling. Biochem J 360: pp 513-530.
Yakubu-Madus FE, Johnson W T, Zimmerman K M, Dananberg J and Steinberg
M I (1999) Metabolic and Hemodynamic Effects of Moxonidine in the Zucker
Diabetic Fatty Rat Model of Type 2 Diabetes. Diabetes 48: pp 1093-1100.
Yamaguchi H and Wang H G (2001) The Protein Kinase PKB/Akt Regulates Cell
Survival and Apoptosis by Inhibiting Bax Conformational Change. Oncogene 20: pp 7779-7786.
Yen TT, Shaw W N and Yu P L (1977) Genetics of Obesity of Zucker Rats and
Koletsky Rats. Heredity 38: pp 373-377.
Zhang J and Abdel-Rahman A A (2006) Nischarin As a Functional Imidazoline (I1)
Receptor. FEBS Lett 580: pp 3070-3074.
204 Zhang J, El Mas M M and Abdel-Rahman A A (2001) Imidazoline I(1) Receptor-
Induced Activation of Phosphatidylcholine-Specific Phospholipase C Elicits
Mitogen-Activated Protein Kinase Phosphorylation in PC12 Cells. Eur J
Pharmacol 415: pp 117-125.
Zhou L, Chen H, Lin C H, Cong L N, McGibbon M A, Sciacchitano S, Lesniak M
A, Quon M J and Taylor S I (1997) Insulin Receptor Substrate-2 (IRS-2) Can
Mediate the Action of Insulin to Stimulate Translocation of GLUT4 to the Cell
Surface in Rat Adipose Cells. J Biol Chem 272: pp 29829-29833.
Zhu H, Hayes J, Chen M, Baldwin J and Piletz J E (2003) Relationship Between
Platelet Imidazoline Receptor-Binding Peptides and Candidate Imidazoline-1
Receptor, IRAS. Ann N Y Acad Sci 1009: pp 439-446.
Ziegler D, Haxhiu M A, Kaan E C, Papp J G and Ernsberger P (1996)
Pharmacology of Moxonidine, an I1-Imidazoline Receptor Agonist. J Cardiovasc
Pharmacol 27 Suppl 3: pp S26-S37.
205