Content uploaded by Terry P Combs
Author content
All content in this area was uploaded by Terry P Combs on Jul 28, 2015
Content may be subject to copyright.
A Transgenic Mouse with a Deletion in the Collagenous
Domain of Adiponectin Displays Elevated Circulating
Adiponectin and Improved Insulin Sensitivity
TERRY P. COMBS, UTPAL B. PAJVANI, ANDERS H. BERG, YING LIN, LINDA A. JELICKS,
MATHIEU LAPLANTE, ANDREA R. NAWROCKI, MICHAEL W. RAJALA, ALBERT. F. PARLOW,
LAURELLE CHEESEBORO, YANG-YANG DING, ROBERT G. RUSSELL, DIRK LINDEMANN,
ADAM HARTLEY, GLYNN R. C. BAKER, SILVANA OBICI, YVES DESHAIES, MARIAN LUDGATE,
LUCIANO ROSSETTI,
AND PHILIPP E. SCHERER
Departments of Cell Biology (T.P.C., U.B.P., A.H.B., Y.L., A.R.N., M.W.R., L.C., Y.Y.D., P.E.S.), Physiology and Biophysics
(L.A.J.), Pathology (R.G.R.), Neuroscience (A.H.), Medicine (S.O., L.R.), and Molecular Pharmacology (S.O., L.R.); Institute
for Animal Studies (R.G.R.); Department of Medicine, Division of Endocrinology, Diabetes Research and Training Center
(S.O., L.R., P.E.S.), Albert Einstein College of Medicine, Bronx, New York 10461; Department of Anatomy and Physiology,
Laval Hospital Research Center, School of Medicine, Laval University (M.La., Y.-Y.D.), Que´bec, Canada; National Hormone
and Peptide Program, Harbor-University of California-Los Angeles Medical Center (A.F.P.), Torrance, California 90509;
Institute for Virology, Technical University Dresden (D.L.), 01307 Dresden, Germany; and Department of Medicine,
University of Wales College of Medicine (G.R.C.B., M.Lu.), Heath Park, Cardiff, United Kingdom CF14 4XN
Adiponectin is a plasma protein expressed exclusively in ad-
ipose tissue. Adiponectin levels are linked to insulin sensitiv-
ity, but a direct effect of chronically elevated adiponectin on
improved insulin sensitivity has not yet been demonstrated.
We identified a dominant mutation in the collagenous domain
of adiponectin that elevated circulating adiponectin values in
mice by 3-fold. Adiponectinemia raised lipid clearance and
lipoprotein lipase activity, and suppressed insulin-mediated
endogenous glucose production. The induction of adiponectin
during puberty and the sexual dimorphism in adult adiponec-
tin values were preserved in these transgenic animals. As a
result of elevated adiponectin, serum PRL values and brown
adipose mass both increased. The effects on carbohydrate and
lipid metabolism were associated with elevated phosphoryl-
ation of 5ⴕ-AMP-activated protein kinase in liver and elevated
expression of peroxisomal proliferator-activated receptor
␥
2,
caveolin-1, and mitochondrial markers in white adipose
tissue. These studies strongly suggest that increasing endog-
enous adiponectin levels has direct effects on insulin sensi-
tivity and may induce similar physiological responses as pro-
longed treatment with peroxisomal proliferator-activated
receptor
␥
agonists. (Endocrinology 145: 367–383, 2004)
A
DIPOSE TISSUE SECRETES several circulating proteins
that have dramatic effects on carbohydrate and lipid
metabolism (1, 2). Two of these peptides, adiponectin and re-
sistin, have recently received considerable attention. Resistin
has been implicated as a negative regulator of insulin signaling,
whereas the adipocyte-specific protein adiponectin has been
implicated as a potent insulin sensitizer (3, 4).
Many studies have found a correlation between insulin
sensitivity and elevated adiponectin levels in both humans
and experimental animals. For example, type II diabetic pa-
tients have lower adiponectin levels than nondiabetic con-
trols, and chronic calorie restriction regimens lead to ele-
vated adiponectin (5). Furthermore, in rhesus monkeys
predisposed to type II diabetes, the levels of circulating adi-
ponectin decrease before the onset of hyperglycemia (6).
Pharmacological studies have shown that recombinant
adiponectin suppresses endogenous glucose production by
raising hepatic insulin sensitivity, an effect also seen in pri-
mary hepatocyte cultures (4, 7). Glucose-lowering effects in
vivo have also been seen with the globular domain of adi-
ponectin (8).
Two reports recently described mouse models with a dis-
ruption of the adiponectin locus (9, 10). Mice lacking adi-
ponectin exhibit earlier onset of diet-induced insulin resis-
tance. Even on a normal laboratory chow diet, both
adiponectin ⫺/⫺ and ⫺/⫹ mice display moderate insulin
resistance and impaired free fatty acid clearance. However,
the effects of chronically elevated serum adiponectin have
not been explored to date due to the limited supply of re-
combinant full-length material.
These observations suggest that low plasma adiponectin con-
tributes to the pathogenesis of insulin resistance and type 2
diabetes. Therefore, elevating adiponectin values can be viewed
as a potentially therapeutic objective. Furthermore, transgenic
(Tg) methods using conventional strategies of overexpression
have also failed to yield a Tg mouse with endogenously ele-
vated adiponectin. Yamauchi et al. (11) expressed a truncated
analog of adiponectin in Tg mice that only contains the globular
domain. However, this fragment clearly lacks the structural
Abbreviations: AMPK, 5⬘-AMP-activated protein kinase; BAT, brown
adipose tissue; CMV, cytomegalovirus; GDI, GDP dissociation inhibitor;
LPL, lipoprotein lipase; MCS, multiple cloning site; NEFA, nonesterified
fatty acids; PPAR
␥
, peroxisomal proliferator-activated receptor
␥
; Tg,
transgenic; TG, triglycerides; UCP1, uncoupling protein-1; WAT, white
adipose tissue; Wt, wild type.
Endocrinology is published monthly by The Endocrine Society (http://
www.endo-society.org), the foremost professional society serving the
endocrine community.
0013-7227/04/$15.00/0 Endocrinology 145(1):367–383
Printed in U.S.A. Copyright © 2004 by The Endocrine Society
doi: 10.1210/en.2003-1068
367
integrity of circulating adiponectin, and it is thus unlikely that
it shares the same physiological properties as the native full-
length oligomeric complex. Chronic exposure to a class of per-
oxisomal proliferator-activated receptor
␥
(PPAR
␥
) ligands
known as thiazolidinediones can also increase serum adiponec-
tin levels (12), but this class of drugs clearly induces pleiotropic
effects in multiple tissues.
We discovered that the expression of adiponectin lacking
a portion of the collagenous domain has a dominant and
specific effect on the secretion of native adiponectin. This
mechanism was studied in vitro and was used to generate a
Tg mouse that exhibited a chronic 3-fold induction of adi-
ponectin in serum. This mouse model is unique in that it
displays elevated levels of the native oligomeric adiponectin
complexes and preserves the effects seen with sexual mat-
uration and sexual differentiation. Adiponectin in this model
is therefore under the same developmental control as adi-
ponectin in wild-type (Wt) mice, except that the basal set-
point is approximately 3 times higher in the former. Not only
do these mice provide insight into the biogenesis of this
fat-specific secretory protein, but they also demonstrate a
significant improvement in insulin sensitivity that can be
achieved through a relatively modest elevation of the serum
levels of this protein.
Materials and Methods
Inducible expression of the adiponectin deletion construct in
3T3-L1 adipocytes
3T3-L1 fibroblasts were infected with the retroviral construct pcz-
TetDL011 (13) (Lindemann, D., unpublished observations) containing
the adiponectin deletion construct. This is a regulatable retroviral vector
based on the rtTA/Tet system (14) and murine leukemia virus. In this
vector, an EGFP-Zeocin fusion protein and the rtTA genes are consti-
tutively expressed from a bicistronic encephalomyocarditis virus inter-
nal ribosome entry site (EMCV IRES) element containing mRNA driven
by the viral long terminal repeat promoter. Adiponectin cDNA is ex-
pressed from a tet-regulatable cytomegalovirus (CMV) minimal pro-
moter located in opposite transcriptional orientation upstream of the 3⬘
long terminal repeat. Cells were infected, selected for Zeo resistance, and
then differentiated according to standard protocols (15). Induction of the
deletion construct was initiated afterd8ofdifferentiation with various
levels of doxycyclin (0 –1
m).
Generation of adiponectin Tg mice
The open reading frame encoding murine adiponectin was subcloned
into the multiple cloning site (MCS) of pBluescript. The region encoding
amino acids 57–95, containing 13 collagen repeats, was deleted using
site-directed mutagenesis. The 5.4-kb aP2 enhancer/promoter sequence
(provided by Bruce M. Spiegelman, Dana-Farber Cancer Institute, Bos-
ton, MA) was introduced into the MCS region upstream of the cDNA.
In addition, the simian virus 40 splice and polyadenylase sequences
were introduced into the MCS region downstream of the cDNA. For
expression of the full-length protein, the complete open reading frame
was inserted into the MCS of pCB7 (gift from J. Casanova, Massachusetts
General Hospital, Boston, MA). This vector provides expression from a
CMV promoter and contains a human GH terminator sequence. The
entire cassettes were excised from both vectors and purified for pro-
nuclear injection into fertilized one-cell murine zygotes from FVB mice
(Taconic Farms, Germantown, NY). The embryos were implanted in
CD-1 mice (Charles River Breeding Laboratories, Wilmington, MA) at
the Transgenic Mouse Facility of the Albert Einstein College of Medicine.
Transgenic mice were identified by slot-blot analysis using genomic
DNA isolated from the tip of the tail. A fragment containing the simian
virus 40 region was radiolabeled and used for hybridization by slot-blot
analysis.
Laboratory animals
Mice were housed in groups of two to five in filter-top cages and were
given free access to water and Teklad rodent diet no. 8604 (Harlan
Teklad, Indianapolis, IN). The colony was maintained in a pathogen-free
Association for the Accreditation of Laboratory Animal Care-accredited
facility at the Albert Einstein College of Medicine under controlled
environment settings (22–25 C; 40–50% humidity; 12-h light, 12-h dark
cycle with lights on from 0600 –1800). The transgene was propagated in
a pure FVB background by mating Tg males with Wt females.
Total RNA isolation, RT-PCR analysis, and Northern
blot analysis
Tissue was dissected immediately after sacrifice, snap-frozen in liq-
uid nitrogen, and stored at ⫺80 C. Total RNA was isolated from frozen
tissue (⫺80 C) with TRIzol (Life Technologies, Gaithersburg, MD) ac-
cording to the manufacturer’s protocol. cDNA was prepared from 4
g
RNA using Superscript II, and random hexamers (Invitrogen, Carlsbad,
CA). The PCR products were analyzed by gel electrophoresis on 1%
agarose and scanned under UV light. Northern blot analysis was per-
formed as described previously (16).
Pulse-chase analysis of adiponectin and resistin secretion in
3T3-L1 adipocytes
Fully differentiated d 8 3T3-L1 adipocytes were pulse-labeled for 10
min in the presence of [
35
S]methionine and cysteine (NEN Life Science
Products, Boston, MA). The labeling medium was then exchanged and
washed four times with medium containing excess unlabeled methio-
nine and cysteine, and the cells were kept in that medium supplemented
with 10% fetal calf serum for up to 24 h. Tissue culture supernatants were
harvested, and cells were lysed at the indicated time points. Both su-
pernatants and cellular lysates were immunoprecipitated with antire-
sistin and antiadiponectin antibodies and analyzed by SDS-PAGE.
Measurement of specific protein levels in tissue by
Western blot
Protein was extracted by immersing frozen tissue in cold buffer [20
mm Tris-HCl (pH 7.5), 5 mm EDTA, 10 mm KCl, and 1 mm phenyl-
methylsulfonylfluoride] and sonicating for 20 sec. The homogenate was
centrifuged at4Ctoseparate lipids, lysate, and insoluble material.
Lysate infranatant was removed and mixed with sodium dodecyl sulfate
to a final concentration of 2% (w/v) and heated to 95 C. After 5 min,
lysates were cooled to room temperature and centrifuged again. The
supernatant was assayed for protein concentration by bicinchoninic acid
(protein assay kit, Pierce Chemical Co., Rockford, IL), and 50
g were
assayed for adiponectin, caveolin-1, or GDP dissociation inhibitor (GDI)
content by Western blot as previously described. For protein extractions
from liver, the tissue was homogenized in a buffer containing 1% Triton
X-100, 60 mm octyl-glucoside, 150 mm NaCl, 20 mm Tris (pH 8.0), 2 mm
EDTA, 50 mm sodium fluoride, 30 mm sodium pyrophosphate, 100 mm
sodium orthovanadate, and 1 mm phenylmethylsulfonylfluoride and
sonicated for 20 sec. The resulting homogenate was centrifuged at 4 C,
and the supernatant was assayed for protein content by bicinchoninic
acid and processed for Western blotting as described above. The anti-
bodies against GDI were a gift from Dr. Perry Bickel (Washington
University, St. Louis, MO).
Measurement of adiponectin levels in plasma by Western
blot analysis
Blood samples were obtained from the tail vain of unrestrained con-
scious animals using heparinized capillary tubes (Fisher Scientific, Fair-
lawn, NJ). Plasma was stored at ⫺20 C for future measurement. Adi-
ponectin in plasma (3
l) was measured by Western blot analysis. After
SDS-PAGE (12%), proteins were transferred to BA83 nitrocellulose
(Schleicher & Schuell, Inc., Keene, NH). Blots were exposed to
125
I-
labeled rabbit antibody against murine adiponectin and analyzed with
a PhosphorImager (Molecular Dynamics, Sunnyvale, CA) using Image-
Quant 1.2 software. Each gel contained four standards of purified mouse
adiponectin at four different concentrations to ensure linearity and re-
368 Endocrinology, January 2004, 145(1):367–383 Combs et al. • Transgenic Overexpression of Adiponectin
producibility of the signal. Inter- and intraassay variations for adiponec-
tin were less than 10%, and the limit of detection was approximately 5
ng adiponectin (12).
Velocity sedimentation for size separation of
adiponectin complexes
Sucrose gradients [5–20% in 10 mm HEPES (pH 8) and 125 mm NaCl]
were poured stepwise (5%, 10%, 15%, and 20%) in 2-ml, thin-walled,
ultracentrifuge tubes (BD Biosciences, Mountain View, CA) and allowed
to equilibrate overnight at 4 C. After layering of the plasma sample on
top [diluted 1:10 with 10 mm HEPES (pH 8) and 125 mm NaCl in the case
of serum], the gradients were spun at 55,000 rpm for4hat4CinaTLS55
rotor in a TL-100 table-top ultracentrifuge (Sorvall, Kendro Laboratory
Products, Newton, CT). Gradient fractions (150
l) were retrieved se-
quentially from the top of the gradient and analyzed by quantitative
Western blot analysis. This protocol was described in more detail pre-
viously (17).
Percent fat mass
The percent fat mass was determined on the basis of the low partition
coefficient of H
2
O. A known amount of
3
H
2
O was injected into the ip
cavity and allowed to equilibrate over 8 h. The distribution of tracer in
the aqueous compartment was used to calculate the contribution of the
lipid stores to the total body weight. The calculation was made under
the assumption that 7.5% of the aqueous volume is occupied by non-
lipids that nevertheless exclude the tracer.
Indirect calorimetry
Animals undergoing indirect calorimetry were acclimated to the re-
spiratory chambers for 1 d before the gas exchange measurements. Mice
were individually housed in the calorimeter cages, and data on gas
exchanges, activity, and food intake were collected for 2–3 d. Indirect
calorimetry was performed with a computer-controlled Oxymax open
circuit calorimetry system (Columbus Instruments, Columbus, OH).
Our instrument comprises 4 respiratory chambers with a stainless steel,
elevated, wire floor. Each chamber is equipped with water bottle, food
tray connected to a balance, and activity monitor. Oxygen consumption
and carbon dioxide production were measured for each mouse at 6-min
intervals. Outdoor air reference values were measured after every 10
measurements. Instrument settings were: gas flow rate, 0.5 liter/min;
settle time, 240 sec; and measure time, 60 sec. Gas sensors were calibrated
daily with primary gas standards containing known concentrations of
O
2
,CO
2
,N
2
(Tech Air, Danbury, CT). A mass flow meter was used to
measure air flow. A limited diffusion metal-air battery was used as an
oxygen sensor. CO
2
was measured with a spectrophotometric sensor.
The respiratory quotient was calculated as the ratio of CO
2
production
over O
2
consumption. The energy expenditure (EE) was calculated as
follows: EE ⫽ (3.815 ⫹ 1.232 ⫻ VCO
2
/VO
2
) ⫻ VO
2
, where VCO
2
is the
expired CO
2
volume (ml/kg䡠h), and VO
2
is the expired O
2
volume
(ml/kg䡠h).
Magnetic resonance imaging
All images were obtained using a 9.4-Tesla imaging system. Mice
were first anesthetized by ip injection with a ketamine/xylazine mixture
(0.1 ml/20 g body weight). To quantitatively assess whole body fat and
water, each mouse was subjected to a 16-scan pulse-acquire sequence in
a 40-mm 1H coil, and spectra, including the water and fat peaks, were
obtained and analyzed. For imaging, several datasets of eight slices of
2-mm thickness spanning the whole body were obtained. Imaging was
conducted using a 35-mm 1H coil and a routine spinecho pulse sequence
(18-msec echo time, 600-msec repetition time, and four-signal averages
per scan).
Light and transmission electron microscopy
Adipose tissue was fixed with 2.5% glutaraldehyde/1.2% acrolein in
sym-collidine buffer, postfixed with 1% osmium tetroxide, followed by
1% uranyl acetate, dehydrated through a graded series of ethanol con-
centrations, and embedded in LX112 resin (LADD Research Industries,
Burlington VT). One-micron-thick sections were cut on an Ultracut UCT
(Reichert, Analytical Instruments, Buffalo, NY), stained with 1% tolu-
idine blue in 1% sodium borate, and viewed on an Axiophot light
microscope (Carl Zeiss, Inc., New York, NY). Ultrathin sections were cut
on a Reichert Ultracut UCT, stained with uranyl acetate, followed by
lead citrate, and viewed on a 1200EX transmission electron microscope
(JEOL, Inc., Peabody, MA) at 80 kV. For light microscopy of the orbital
fat, the entire skull was formalin-fixed, paraffin-embedded, sectioned,
and stained with hematoxylin and eosin. For electron microscopy, ad-
ipose tissue was dissected from the intraconal region of the orbit and
frozen under high pressure using an EMpact High Pressure Freezer
(Leica Microsystems, Solms, Austria). The frozen samples were trans-
ferred to a Leica EM AFS Freeze Substitution Unit, freeze-substituted in
1% osmium tetroxide in acetone, brought from ⫺90 C to room temper-
ature over 2–3 d, rinsed in acetone, and embedded in LX112 epoxy resin
(LADD Research Industries). Ultra thin sections of 70 –80 nm were cut
on a Reichart Ultracut UCT, stained with uranyl acetate, followed by
lead citrate, and viewed on a JEOL 1200EX transmission electron micro-
scope at 80 kV. Photographs were taken with a Zeiss digital imaging
system.
Triglyceride clearance
Mice were weighed at 1000 h and were given 15
l olive oil/g body
weight by gastric gavage. Approximately 20
l blood were collected at
0, 2, 4, and 12 h and assayed for triglycerides (TG; Sigma-Aldrich Corp.,
St. Louis, MO) and nonesterified free fatty acids (WAKO Diagnostic
Products, Richmond, VA). The mice were denied access to food during
the course of the study.
Measurement of tissue lipoprotein lipase (LPL) activity
Enzyme activity in adipose tissue was performed as previously de-
scribed (18). Briefly, tissue homogenates were incubated with a substrate
mixture containing [carboxyl-
14
C]triolein, and nonesterified fatty acids
(NEFA) released by LPL were separated and counted. LPL activity was
expressed as microunits (1 mU ⫽ 1
mol NEFA released/h of incubation
at 28 C). Data are expressed as LPL activity per gram of tissue. The
interassay coefficient of variation (4.1%) was using bovine skim milk as
a standard source of LPL.
Measurement of glucose flux
Adult (7- to 10-month-old) male Tg mice and their Wt littermates
(30–35 g) were anesthetized with chloral hydrate (400 mg/kg body
weight, ip) and catheterized through the right internal jugular vein. The
venous catheter was used for infusion while blood samples were col-
lected from the tail vein. Each animal was monitored for food intake and
weight gain for 3– 4 d after surgery to ensure complete recovery. Eu-
glycemic insulin (1 or 4 mU/kg䡠min) clamps were performed in con-
scious catheterized mice as previously described (7). Food was removed
for 5 h before beginning the in vivo studies. A solution of glucose (10%)
was infused at a variable rate as required to maintain euglycemia (6 mm).
Mice received a constant infusion of HPLC-purified [
3
H]glucose (0.1
mCi/min; NEN Life Science Products, Boston, MA) or insulin (1 or 4
mU/min䡠kg body weight). Thereafter, plasma samples were collected to
determine glucose levels as well as [
3
H]glucose specific activity (at 40,
50, 60, 70, 80, and 90 min). Consecutive samples were used for assess-
ment of plasma insulin levels. Steady state conditions for both plasma
glucose concentration and specific activity were achieved by 40 min in
these studies. In the final stage, mice were anesthetized (60 mg pento-
barbital/kg body weight, iv), the abdomen was quickly opened, and the
liver was freeze-clamped in situ with aluminum tongs cooled in liquid
nitrogen. The time between the injection of anesthesia and the freeze-
clamping of tissue samples never exceeded 60 sec. Tissue samples were
stored at ⫺80 C for further analysis. The euglycemic clamp protocol was
approved by the institutional animal care and use committee of the
Albert Einstein College of Medicine.
High fat feeding and oral glucose tolerance tests
A cohort of mice was maintained on a high fat diet from weaning at
3 wk to 6 months of age (59% of calories derived from fat; D12492
Combs et al. • Transgenic Overexpression of Adiponectin Endocrinology, January 2004, 145(1):367–383 369
Research Diets, Harlan Teklad, Madison, WI). Fasted mice (5 h) were
given an oral glucose load (2.5 mg glucose/g body weight using a
solution of 10% glucose in physiological saline). Blood was collected at
0, 30, 60, and 120 min and used for glucose (Sigma-Aldrich Corp.) and
insulin (Linco Research, St. Charles, MO) measurements. Access to food
was denied during the course of the study.
Statistical analysis
Results are shown as the mean ⫾ sem. Statistical analysis was per-
formed by unpaired t test (nonparametric), assuming unequal variance
unless otherwise indicated. Significance was accepted at P ⬍ 0.05.
Results
Adiponectin overexpression
It has been extremely difficult to elevate adiponectin in
serum by simple overexpression. Relatively high serum con-
centrations, a short half-life, and feedback inhibition on en-
dogenous production/clearance may contribute to the com-
plexity. We developed a novel Tg strategy to elevate
adiponectin in the circulation using a deletion mutant of
adiponectin under control of the adipose-specific enhancer/
promoter of the aP2 gene. This deletion mutant lacks 13 of
22 Gly-X-Y repeats in the collagenous domain (Fig. 1A). The
strategy was initially developed in tissue culture to elucidate
the biogenesis of the adiponectin complex.
Adiponectin forms homotrimers upon translocation into
the lumen of the endoplasmic reticulum. The assembly of
circulating isoforms of adiponectin begins with a highly sta-
ble interaction between the carboxyl-terminal globular do-
mains of three adiponectin subunits and proceeds to the
collagenous domain (19). Subsequently, pairs of trimers are
linked through the formation of a disulfide bond at Cys
39
(17). Pairs of trimers ultimately assemble into higher order
structures that contain 12 or 18 adiponectin subunits. A de-
letion mutant of adiponectin lacking 13 of 22 Gly-X-Y repeats
in the collagenous domain (⌬Gly-adiponectin) interferes
with the folding of individual trimer units.
The effects on adiponectin secretion can be inverted based on
the expression of full-length to ⌬Gly-adiponectin. Low ⌬Gly-
adiponectin expression increases the secretion of full-length
adiponectin, whereas high levels lower the secretion of full-
FIG. 1. A, Structural domains of adiponectin
and ⌬Gly-adiponectin; the GlyXY motifs in the
collagenous domain are shown in white and
hatched bars. B, Inducible expression of the de-
letion construct in 3T3-L1 adipocytes. Cells
were differentiated according to the standard
differentiation protocol. After 8 d, the expres-
sion of the deletion construct was induced by the
addition of various levels of doxycyclin (0 –1
M). After 72 h, medium was replaced with fresh
serum-free medium containing the equivalent
amount of doxycyclin. After 18 h, the cell lysates
and supernatants were analyzed for adiponec-
tin expression by Western blot analysis under
reducing conditions. Note that a longer expo-
sure was required to detect the expression of the
intracellular deletion construct. C, Pulse-chase
study in 3T3-L1 adipocytes. Cells were pulsed
with [
35
S]Cys/Met as described in Materials and
Methods. Intracellular and extracellular mate-
rials were assayed for the presence of adiponec-
tin and resistin by immunoprecipitation. D,
Treatment of 3T3-L1 adipocytes with the pro-
teasomal inhibitor MG132. At the indicated
times after initiation of the treatment with 10
M MG132, cells were lysed and analyzed for
intracellular levels of adiponectin by Western
blot analysis.
370 Endocrinology, January 2004, 145(1):367–383 Combs et al. • Transgenic Overexpression of Adiponectin
length adiponectin. This was shown in a dose-dependent in-
duction of the mutant protein in 3T3-L1 adipocytes by Western
blot analysis under reducing conditions (Fig. 1B). Note that a
greater exposure period was needed to detect the intracellular
expression of ⌬Gly-adiponectin compared with full-length adi-
ponectin, and that ⌬Gly-adiponectin is not secreted under any
circumstances. This effect can be exclusively achieved at the
level of secretion, as endogenous adiponectin mRNA levels are
not altered by ⌬Gly-adiponectin (data not shown). This is con-
sistent with our previous observations that point to a critical
posttranscriptional component in the relationship of sexual di-
morphism to adiponectin levels (20). This suggested that a pool
of endogenous adiponectin within the secretory pathway of
adipocytes could be effectively released through low expres-
sion of ⌬Gly-adiponectin.
More detailed analysis of the secretion kinetics for adi-
ponectin using a pulse-chase study revealed that near-max-
imal levels of secretion could be achieved within 3 h after
initiation of the chase. The remaining material was retained
intracellularly and slowly degraded (Fig. 1C). This is in stark
contrast to the secretion kinetics of resistin, another adipo-
cyte-specific secretory protein that could be almost quanti-
tatively secreted under the same conditions. The inability of
the remaining intracellular adiponectin pool to be released is
not a reflection of a high level of misfolding, as a number of
treatments (e.g. treatment with

3
agonists or reducing
agents) can effectively tap into this intracellular pool and
release it (Berg, A., and U. Pajvani, unpublished observa-
tions). This suggests that under normal conditions, a futile
cycle is in place that is characterized by high level production
of adiponectin, followed by degradation of nearly 50% of the
newly synthesized material. Degradation most likely in-
volves the cytosolic proteasomal complex, because treatment
of cells with MG132, a proteasomal inhibitor, leads to a
significant accumulation of intracellular adiponectin, as
shown by Western blot analysis in 3T3-L1 cells (Fig. 1D). In
the absence of ⌬Gly-adiponectin, normal basal levels of full-
length adiponectin are secreted. Upon low level induction of
⌬Gly-adiponectin, a small amount of mixed trimers is
formed, blocking the futile degradation cycle of the Wt pro-
tein. This significantly increases the secretion of full-length
adiponectin. At higher levels of the mutant protein, the de-
letion construct effectively integrates into all trimers formed,
thereby quantitatively blocking the secretion of almost all of
the full-length adiponectin. The net result of transgene ex-
pression is therefore a more efficient secretion of endogenous
Wt adiponectin complexes.
Based on our observations in tissue culture, we expressed
⌬Gly-adiponectin under the control of a fat cell-specific aP2
promoter in mice. Tg mice with Wt adiponectin under con-
trol of the more ubiquitously expressed CMV promoter were
also generated in parallel. The results from three and four
founder lines are shown for both constructs (Fig. 2B). In the
case of CMV-adiponectin, a much larger number of founders
was analyzed (⬎30), but they are not shown. RT-PCR anal-
ysis of gonadal fat with transgene-specific primers showed
three lines expressed ⌬Gly-adiponectin, but only line 2 had
increased circulating levels in serum (Fig. 2C). Because ⌬Gly-
adiponectin females are infertile, the line was propagated
through Tg males and Wt females that yielded a Tg birth rate
of 50%.
Although several founders expressed the CMV-adiponec-
tin transgene, line 3 had increased serum concentrations of
adiponectin. Interestingly, both line 2 of aP2-⌬Gly-adiponec-
tin and line 3 of CMV-adiponectin mice developed an un-
usual phenotype at midlife that included expansion of the
interscapular region and bilateral exophthalmia (Fig. 2D). An
increase in serum adiponectin correlated with the appear-
ance of this phenotype. The expression of either aP2-⌬Gly-
adiponectin or CMV-adiponectin without an increase in
serum adiponectin did not produce expansion of the inter-
scapular region and exophthalmia in old age (24–30 months).
Line 2 from aP2-⌬Gly-adiponectin was chosen for further
analysis because it showed the greatest elevation of serum
adiponectin. A survey of transgene expression by RT-PCR in
various fat pads and other tissues (Fig. 3A) revealed that
⌬Gly-adiponectin expression under the aP2 promoter over-
lapped endogenous adiponectin expression and is thus lim-
ited to adipose tissue. Adiponectin levels in various fat pads
(interscapular, inguinal, mammary, and epididymal) were
elevated in both male and female Tg mice. Interestingly, pe-
rimetrial (uterine) fat differed from all other fat pads examined
in that a reduction of adiponectin was observed in the Tg
animals (Fig. 3B). As the clearance of adiponectin was not mark-
edly different in Wt and mutant mice (data not shown), the net
increase in serum in Tg mice was primarily due to increased
secretion of the protein from a range of adipose pads.
Mice exhibit a marked sexual dimorphism with respect to
serum adiponectin levels (20). Males and females displayed
a dramatic increase in adiponectin at puberty, and subse-
quently, adult females exhibited 2- to 3-fold higher levels
than adult males. As the expression of ⌬Gly-adiponectin
allowed endogenous adiponectin to escape degradation, the
early rise in serum levels and the sexual dimorphism were
both preserved in ⌬Gly-adiponectin mice at an overall higher
level. Both male and female ⌬Gly-adiponectin mice exhibited
serum adiponectin levels 3-fold above those in their age- and
sex-matched Wt littermates (Fig. 4A).
Adiponectin circulates mainly in two high order forms
(17). The relative distribution of these forms may have func-
tional implications. Tg mice displayed a similar distribution
of these high order forms as their Wt littermates. Although
the levels of adiponectin were increased in Tg mice, the size
profile of high order complexes was similar to that of Wt mice
(Fig. 4B). This strongly suggests that the complexes gener-
ated in Tg mice fold properly and resemble Wt protein in
their oligomeric make-up.
Male Tg mice displayed similar body weights as their Wt
littermates up to 1 yr (Fig. 4C). Tg females, however, started
to weigh significantly more than their Wt littermates around
16 wk of age and continued to be heavier until 1 yr of age
when they displayed a small, but significant, 15% increase in
overall body mass (Fig. 4C).
Chronic elevation of adiponectin does not affect food intake,
but leads to an increase in adiposity in females
At 6–8 months of age, we determined the percent body fat
content in males and females. There was a significant in-
Combs et al. • Transgenic Overexpression of Adiponectin Endocrinology, January 2004, 145(1):367–383 371
crease in percent total body fat by that age in the females,
which reached approximately 30% of their total body weight,
in contrast to 18% in their Wt siblings (Fig. 5A). The males
did not display any significant differences at that age (data
not shown). These differences in the females could not be
explained by a difference in food intake within the accuracy
with which these experiments can be performed (Fig. 5B). We
were also unable to observe any changes with respect to the
average 24-h energy expenditure (Fig. 5C), nor were there
differences with respect to the respiratory quotient in either
the resting or the active state (Fig. 5D).
Chronic elevation of adiponectin leads to remodeling of
select fat depots
Even though a detailed pathological examination of the
mice did not reveal any gross morphological changes, it
became apparent that Tg female mice developed a large
tissue mass in the interscapular region by 12 months (Fig. 6A)
that was particularly striking in a cross-sectional view (Fig.
6B). Magnetic resonance imaging strongly suggested that the
expansion consists primarily of lipid mass (Fig. 6B). A his-
tological analysis at the light microscopic (Fig. 6C) and elec-
tron microscopic (Fig. 6D) levels clearly identified the ex-
panding tissue mass as adipose tissue. Interscapular brown
adipose tissue (BAT) could be identified in both Wt and Tg
mice on the basis of its darker brown appearance. Rather than
the classical multilobular appearance of the lipid droplets as
they appeared in Wt mice, the Tg BAT had a more unilocular
appearance.
Another fat pad that selectively hyperproliferates is the
orbital (intraconal) fat pad. This was evident in 100% of
12-month-old females. Orbital fat proliferation pushed the
orbit away from the shallow bony orbit, producing axial
proptosis. The proptosis progressed asymmetrically, but
eventually affected both eyes (Fig. 7A). High resolution mag-
netic resonance imaging (Fig. 7B) and histological analysis
(Fig. 7C) clearly indicated the intraconal fat pad as the pri-
mary cause of the proptosis. There was no indication of
FIG. 2. A, Tg expression constructs for 1) truncated adiponectin lacking seven GlyXY repeats (⌬Gly-adiponectin) under the adipocyte-specific
aP2 promoter (left), or 2) full-length adiponectin under the ubiquitous active CMV promoter (right). B, Plasma adiponectin levels of Wt
adiponectin resulting from Tg expression of aP2-⌬Gly (left) or CMV-adiponectin (right). 䡺, Female Wt mice; f, female Tg littermates. The levels
are indicated as arbitrary densitometric units. C, Transgene expression assessed by nonquantitative RT-PCR using primers that recognize the
endogenous Wt adiponectin and the ⌬Gly construct (left) or primers specific for the CMV-adiponectin transgene (right). D, Phenotypic
appearance of Tg mice: expansion of interscapular tissue in 12-month-old mice carrying the aP2-⌬Gly transgene (left) or the CMV-adiponectin
transgene (right).
372 Endocrinology, January 2004, 145(1):367–383 Combs et al. • Transgenic Overexpression of Adiponectin
immune infiltration or inflammatory activity within the mus-
cle fibers, but individual orbital muscles separated into mul-
tiple fibers by the expansion of adipose tissue. At the orbital
apex, the tendinous ring insertion of the ocular muscles di-
lated as fat prolapsed. Chronic exposure keratopathy due to
proptosis ultimately ulcerated the entire cornea and resulted
in globe perforation with hemorrhagic ulcerated plaque
clinging to the sclera beneath, followed by complete globe
shrinkage and fibrosis (phthisis). The intraconal fat further
expanded between the straps of ocular muscles. The prop-
tosis worsened progressively and mice were usually eutha-
nized by 15 months of age.
This selective expansion of the orbital adipose tissue is
reminiscent of pathological changes associated with some
cases of Graves’ disease (21). However, our results suggest
that it is unlikely that overexpression of adiponectin triggers
an autoimmune reaction toward the TSH receptor as ob-
served in Graves’ patients, because TSH receptor antibodies
were assayed using three different methods that measure 1)
simple binding to the TSH receptor; 2) TSH receptor agonist
activity, resulting in elevated intracellular cAMP; and 3) in-
hibition of binding of the endogenous TSH receptor ligand,
TSH. No evidence for a humoral autoimmune reaction to the
TSH receptor was observed (Table 1).
Both Wt and Tg orbital adipocytes have multiple mitochon-
dria and express the BAT-specific marker protein uncoupling
protein-1 (UCP1; Fig. 7D). The primary source of expanding fat
in Tg mice therefore appears to originate from BAT-like tissue,
as judged by low level UCP1 expression even in Wt fat (Fig. 7E).
This adipose mass expanded laterally, eventually wrapping
around the entire neck. With age, the adipose tissue stretched
even further and included regions of the head, including the
submandibular fat pad (see three-dimensional nuclear mag-
netic resonance movie of Wt and Tg published as supplemental
data on The Endocrine Society’s Journals Online web site,
http://endo.endojournals.org).
Macroscopic appearance and weight were comparable be-
tween Wt and Tg in all other adipose pads examined. The
expansion of the interscapular and orbital adipose tissues
was significantly delayed in males.
Chronic elevation of adiponectin affects PRL levels
Serum levels of adiponectin are under complex hormonal
control (20). To test whether chronic adiponectin overex-
pression feeds back on any factors known to affect adiponec-
tin levels under normal conditions, we determined serum
levels of a number of hormones and metabolites. Table 2
shows that levels of TNF
␣
and glucocorticoids were normal;
the pituitary hormones GH and TSH were also comparable
in Wt and Tg. Surprisingly, despite overall increased fat mass
in Tg females, leptin levels were comparable as well. In
FIG. 3. A, Tissue profile for mRNA ex-
pression of adiponectin, the adiponectin
transgene (⌬Gly-adiponectin), and

-actin
expression in female ⌬Gly-adiponectin
mice (Tg) and their Wt littermates (Wt)
by RT-PCR analysis. B, Western blot
analysis for adiponectin levels in adipose
tissue from 4-month-old female (left) and
male (right) Tg mice and their Wt litter-
mates. BAT and inguinal and mammary/
epididymal WAT were used; representa-
tive extracts from two mice per group are
shown. Equal protein loading was con-
firmed by blotting for the internal stan-
dard GDI.
Combs et al. • Transgenic Overexpression of Adiponectin Endocrinology, January 2004, 145(1):367–383 373
contrast, Tg males and females displayed a 2-fold increase in
PRL levels. PRL exerts a strong negative influence on serum
adiponectin levels in Wt mice (20). The chronic elevation of
PRL levels in Tg animals may therefore reflect a compensa-
tory mechanism to lower the elevated concentration of adi-
ponectin in serum.
Improved clearance of TG in female Tg adiponectin mice
Several groups have proposed an involvement of adi-
ponectin in TG and free fatty acid metabolism (8). To test
whether there is an effect on TG clearance, we gavaged a
group of Wt and Tg mice with olive oil and monitored TG
and free fatty acid levels in plasma during the subsequent
12 h. Tg females displayed dramatically different clear-
ance rates than Wt mice. Plasma TG and free fatty acid
(NEFA; Fig. 8, A and B) levels before the gavage were
vastly overlapping. However, subsequent to the gavage,
plasma levels of TG and NEFAs were consistently lower
in the Tg animals, never approaching the peak levels found
in Wt females. These observations suggest that chronically
increased levels of adiponectin in females have a positive
impact on TG clearance and are consistent with acute
pharmacological effects that lead to a decrease in plasma
TG upon injection of recombinant adiponectin. The more
modest increases in serum adiponectin levels in males are
not sufficient to cause any changes in lipid clearance,
because the kinetics of TG and NEFA disposal are vastly
overlapping between Wt and Tg males. A partial expla-
nation for this effect is revealed upon analysis of adipose
tissue lipoprotein lipase activity (Fig. 8C). The Tg females
displayed significantly elevated LPL activity per gram of
tissue in the fed state in a number of different fat pads.
Similar trends were observed in the fasted state (not
shown). Consistent with the improved clearance of TG that
could only be observed in females, the increased lipopro-
tein lipase activity was observed in females only. Addi-
tional metabolic parameters were analyzed in these mice,
but did not reveal any significant differences (Table 3).
Chronic increase in serum adiponectin triggers
transcriptional changes in adipose tissue
To describe the changes in the signaling modules involved
in insulin signal transduction more fully we analyzed adi-
pose tissue extracts by Western blot analysis. We previously
reported that mice with a genetic deletion of the caveolin-1
locus display a reduction of serum adiponectin by almost
90% (22). Caveolin-1 is a critical regulator of signal trans-
duction and has also been implicated as a modulator of
insulin receptor signaling. The association between adi-
ponectin and caveolin-1 levels is preserved in this Tg model
as well. Adipose pads with increased production of adi-
ponectin [BAT and mammary white adipose tissue (WAT)
are shown as representative examples] showed increased
caveolin-1 levels, whereas metrial fat that showed reduced
levels of adiponectin in the Tg mice also had reduced levels
of caveolin-1 (Fig. 9A).
Finally, as many of the phenotypes described for the Tg
mice are consistent with increased PPAR
␥
activation, we
tested whether we would detect increased PPAR
␥
2 levels in
Tg mice. At least in a subset of white fat pads (Fig. 9B shows
abdominal fat as a representative example), PPAR
␥
2 mRNA
levels were markedly increased, particularly in females. Fur-
FIG. 4. A, Circulating adiponectin levels
in female (left) and male (right) ⌬Gly-
adiponectin (Tg) mice (F) and their Wt
littermates (E) measured in a longitudi-
nal 1-yr study by Western blot analysis.
B, Size distribution of circulating adi-
ponectin in 4- to 6-month-old female (left)
and male (right) Tg mice (F) and their Wt
littermates (E) using velocity sedimen-
tation in 5–20% sucrose gradients. The
numbers reflect arbitrary densitometric
units per milliliter of plasma. C, Change
in body weight of female (left) and male
(right) Tg mice (F) and their Wt litter-
mates (E) during a longitudinal 1-yr
study. Each group represents 8 –10 mice.
*, P ⬍ 0.05, Wt vs. Tg mice.
374 Endocrinology, January 2004, 145(1):367–383 Combs et al. • Transgenic Overexpression of Adiponectin
thermore, resistin mRNA levels were slightly induced com-
pared with control levels, and a very significant induction of
mRNA levels of several mitochondrial markers (the ADP/
ATP translocator and the dicarboxylate transporter are
shown) was detected.
Euglycemic insulin clamps reveal dramatically improved
hepatic insulin action in Tg adiponectin mice
To determine whether chronic increases in circulating lev-
els of adiponectin improve glucose tolerance by increasing
glucose uptake, decreasing glucose production, or both, we
used the insulin clamp technique on a cohort of 7- to 10-
month old male and female mice that were fed a standard
laboratory chow diet. Insulin infusion rates were selected to
raise plasma insulin levels similarly in all groups, and a
variable infusion of a 10% glucose solution was started and
periodically adjusted to maintain the plasma glucose con-
centration at approximately 5 mm for the rest of the study.
Similar to the acute administration of recombinant adiponec-
tin, the rate of disappearance of glucose was not altered in
Tg mice, indicating that whole body glucose uptake was not
affected. However, with either low (1 mU/kg䡠min) or high (4
mU/kg䡠min) insulin, glucose production in Tg females was
lower than that in Wt mice, indicating greater insulin sen-
sitivity in the Tg animals (Fig. 10, A and B). This effect became
even more striking with hyperinsulinemia, which demon-
strated a marked decrease in the rates of glucose production
in Tg animals compared with Wt mice in both females and
males (Fig. 10B). Therefore, both male and female Tg mice
exhibit increased hepatic insulin sensitivity. The relevant
metabolic parameters for these experiments are indicated in
Table 4. From that table it is apparent that the Tg mice
displayed a marked decrease in postabsorptive plasma in-
sulin concentrations compared with Wt mice. Although fe-
male Tg mice do show initial signs of increased interscapular
fat mass at this age, it is important to note that the males do
not display any detectable differences in fat distribution be-
tween Wt and Tg animals, emphasizing that the improve-
ment in hepatic insulin sensitivity is not a consequence of the
interscapular fat hyperproliferation.
Kadowaki and colleagues (23) have recently shown that
full-length adiponectin activates 5⬘-AMP-activated protein
kinase (AMPK) in liver. We therefore examined the activa-
tion state of AMPK in livers isolated from female animals
used for the hyperinsulinemic clamp studies. Although ab-
solute AMPK levels were not different, and basal phospho-
AMPK levels were very low in Wt and Tg animals, Tg an-
imals displayed a prolonged activation of AMPK that could
still be visualized at the end of the clamp studies (Fig. 10C),
consistent with our own observations in primary hepatocytes
upon treatment with adiponectin (not shown) and in vivo
results by other groups (23), further corroborating that ele-
vated adiponectin levels will lead to activation of AMPK.
Elevated circulating adiponectin improves oral
glucose tolerance
Our previous studies suggested that pharmacological lev-
els of recombinant adiponectin dramatically improve hepatic
insulin sensitivity in Wt and type I and type II diabetic mice.
To further test whether the constitutively elevated adiponec-
tin levels in our Tg mice have an impact on insulin sensitivity,
we subjected a cohort of high fat-fed Wt and Tg males to an
oral glucose tolerance test. The presence of the transgene did
not prevent weight gain during a high fat diet for 6 months.
A modest increase in body weight was observed in female Tg
compared with their Wt littermates, similar to chow-fed an-
imals (Fig. 11A). The Tg females responded with a signifi-
cantly improved glucose excursion during an oral glucose
tolerance test, indicating improved insulin sensitivity com-
FIG. 5. Comparison of energy homeostasis between female Wt and
⌬Gly-adiponectin (Tg) mice at 8 months of age. Percent body fat (A)
and daily food intake (B) of 8-month-old individually housed Wt and
mice and their Tg littermates. The percent body fat was determined
by equilibration of
3
H
2
O. Energy expenditure (C) and respiratory
quotient (D) in the resting and active states were obtained by indirect
calorimetry in 8-month-old female Tg mice (f) and their Wt litter-
mates (s). Gas exchanges were measured in mice individually housed
in metabolic chambers, whereas activity was monitored with an in-
frared light movement sensor. Resting values for energy expenditure
were calculated from measurements obtained during times when
activity sensors recorded less than 20 events in a 20-min interval. Gas
exchanges values were the mean of measurements made every 20 min
for a 24-h period after a 24- to 48-h acclimation period. Each group
represents 8–10 mice. *, P ⬍ 0.05, active is significantly higher than
resting.
Combs et al. • Transgenic Overexpression of Adiponectin Endocrinology, January 2004, 145(1):367–383 375
pared with Wt littermates (Fig. 11B). Although glucose dis-
posal was comparable between Wt and Tg, the males not
only displayed reduced insulin levels in the basal state, but
also achieved normal glucose disposal with significantly
lower insulin levels. Female Tg demonstrated improved glu-
cose clearance compared with Wt littermates and required
less insulin to achieve these improved clearance kinetics (Fig.
11C). This indicates that the Tg animals are vastly resistant
to diet-induced decreases in insulin sensitivity.
Discussion
We have studied the effects of a life-long elevation in
circulating adiponectin. To achieve this, we employed a Tg
approach that improves the efficiency with which endoge-
nous adiponectin is secreted. In parallel, we have generated
a large number (⬎30) of additional Tg founder lines that
express the full-length protein under the control of the ubiq-
uitously expressed CMV promoter. However, only 1 of these
lines displayed approximately 3-fold increased serum adi-
ponectin levels, and we were unable to propagate the line
due to fertility problems. Interestingly, though, as the first
generation of these mice aged, they also displayed hyper-
proliferation of the interscapular fat pad and the exophthal-
mos. Again, this phenomenon was not seen in other lines that
effectively expressed the CMV-driven transgene, but was
strictly dependent on a net increase in serum adiponectin
levels.
Epidemiological studies in humans and animal models
describe a close association between insulin sensitivity and
adiponectin levels within a 2- to 3-fold range from baseline.
Consistent with the effects of a pharmacological 2- to 3-fold
elevation of adiponectin after the injection or infusion of
recombinant adiponectin (4, 7), congenital elevation of adi-
ponectin led to improved hepatic insulin sensitivity. Female
Tg mice exhibit a significant increase in the rate of oral lipid
clearance, consistent with the phenotype reported for mice
carrying a genetic deletion of the adiponectin locus, which
display delayed clearance of free fatty acids after injection of
a TG emulsion. A recent study by Matsuzawa and colleagues
(9) in adiponectin
⫺/⫺
mice focuses on a mechanism involv
-
ing free fatty acid uptake and oxidation by muscle in line
with original suggestions by Lodish and colleagues (8). Al-
though this represents a potential mechanism, we are also
intrigued by the effects of adiponectin overexpression on
serum TG clearance. It will be interesting to determine
whether adiponectin has additional effects on lipolysis or
lipoprotein clearance by the liver, perhaps via the same
mechanisms as those responsible for the liver glucose effects.
Additionally, we reported (22) that mice lacking caveolin-1
display dramatically reduced adiponectin levels coupled
with severely elevated TG and free fatty acid levels, espe-
cially in the postprandial state. In light of the findings pre-
sented here, this suggests that the postprandial hypertriglyc-
eridemia may be functionally related to the reduced
adiponectin levels in the caveolin-1
⫺/⫺
mice. It is intriguing
that the lack of caveolin-1 triggers reduced adiponectin levels
in serum, whereas the Tg up-regulation of adiponectin in Wt
mice causes increased adipose caveolin-1 expression. The
dramatically increased lipoprotein lipase activity that we
FIG. 6. A, Comparison of a Wt mouse (left) and a ⌬Gly-adiponectin
animal (right) at 12 months of age. B, Whole body sagittal series of
14-month-old female Wt (left) and Tg (right) mice obtained by T1
weighted magnetic resonance imaging (MRI). C and D, Light and
electron micrographs of interscapular BAT of a 14-month-old female
Wt (left)andaTg(right) mouse exhibiting hypertrophy of interscap-
ular adipose tissue. Note the increased lipid droplet size in the Tg
adipose tissue.
376 Endocrinology, January 2004, 145(1):367–383 Combs et al. • Transgenic Overexpression of Adiponectin
observe in female adipose tissues analyzed could contribute
to increased clearance of a postprandial TG load. Similar
increases in LPL activity have recently been reported in
PPAR
␥
agonist-treated rats and linked to improvements in
postprandial TG clearance (18).
Humans and mice display a sexual dimorphism for cir-
culating adiponectin values, with female levels increased
compared with those in males. This sexual dimorphism is
preserved in Tg mice, with the transgene effectively ampli-
fying the host’s endogenous levels. Due to overexpression,
Tg male mice have levels of adiponectin approximately equal
to those in Wt females. The serum concentrations found in
Tg males are therefore physiological and readily found in Wt
mice, albeit in a different (female) hormonal milieu. Female
Tg have levels approximately 3-fold above those in Wt fe-
males. Relatively small changes in serum adiponectin con-
centrations can have a significant effect on insulin sensitivity.
An example is seen in heterozygous mice (⫹/⫺) with only
one functional allele for adiponectin. These mice have ap-
proximately 40% of the Wt adiponectin serum levels, but
TABLE 1. Antibodies to the thyrotropin receptor (TSHR)
12- to 16-month-old females
Wt Tg
TSHR IgGs 0.24 ⫾ 0.4 0.22 ⫾ 0.5
TBII 117 ⫾ 3 104 ⫾ 5
TSAB 1.0 ⫾ 0.1 1.2 ⫾ 0.2
TOTAL T
4
(mg/dl)
2.5 ⫾ 0.4 1.8 ⫾ 0.2
TSH (mU/ml) 4.3 ⫾ 0.4 4.5 ⫾ 0.5
CHO cells stably transfected with the human TSHR and flow
cytometry were used to measure TSHR IgGs. Mouse serum (diluted
1 in 50) was tested on CHO cells with and without TSHR and the
histograms compared using Kolmogorov Smirnof statistics to obtain
a “d” value. Two monoclonal antibodies against the TSHR had d val-
ues of 0.5 and 0.67, whereas controls with just the second FITC
conjugate and no first antibody had a value of 0.24.
Thyrotropin binding inhibiting Igs (TBII) used CHO cells that
express TSHR to compare the binding of I-125 TSH in the presence
of each individual mouse serum (all in triplicates that agree to within
10%). Results are expressed as a percentage and
SE from the average
of the counts of the 10 female Wt mice used to define our 100%, i.e.
no inhibition of I-125 TSH binding. Note that with an excess TSH the
binding drops to 25%. Thyroid stimulating antibodies (TSAB) used
CHO cells with the TSHR and a cAMP responsive luciferase reporter.
Results are expressed as a stimulation index (S.I.) calculated from the
ratio of light in the presence of test serum to light in the presence of
a euthyroid pool (all tested as duplicates that agree to within 12%).
We used the average of the female Wt mice and an S.I. of 1.5 or above
is considered positive. Two animals, a transgenic female and male,
had weakly positive TSAB. Note that a known human TSAB had S.I.
of 3.9, whereas TSH had 20.15.
A total of four to six mice were tested in each case.
FIG. 7. A, Frontal view of 11-month-old ⌬Gly-adiponectin (Tg) and
Wt mice with asymmetric exophthalmos. B, MRI series of the head
(coronal sections) of a 14-month-old female Wt mouse (top)andaTg
mouse (bottom). Bright tissue indicates selective hyperplasia of ret-
roorbital adipose tissue in the Tg. C, Light microscopy of the entire
orbital cavity after hematoxylin/eosin staining in a 14-month-old fe-
male Wt mouse (left) and a Tg animal (right). D, Intraconal (retro-
orbital) adipocytes seen by transmission electron microscopy showing
a high number of mitochondria in both Wt and Tg. E, UCP1 and

-actin expression in 12-month-old female Tg and Wt mice exhibiting
selective hyperplasia of interscapular and intraconal (eye) adipose
tissue by nonquantitative RT-PCR analysis. Note the presence of
UCP1 in fat pads from Wt mice.
Combs et al. • Transgenic Overexpression of Adiponectin Endocrinology, January 2004, 145(1):367–383 377
they have a significantly reduced insulin sensitivity (10). Wt
females with higher circulating adiponectin levels are more
insulin sensitive than Wt males (20). Acute injections that
alter circulating levels by as little as 2-fold have a profound
impact on insulin sensitivity (4, 7).
Finally, a number of epidemiological studies in humans
that draw strong positive correlations between insulin sen-
sitivity and adiponectin levels focus on values that vary by
less than 2- to 3-fold from baseline. Along with the sexually
dimorphic expression pattern, we observed various sexually
dimorphic phenotypes in a number of different areas. By
midlife, the female Tg mouse exhibits a moderate increase in
FIG. 8. Circulating TG (A, top) and free
fatty acids (B, bottom) in 6- to 8-month-old
female (left) and male (right) ⌬Gly-
adiponectin mice (F) and their Wt litter-
mates (E) after oral lipid challenge. Mice
were kept on a standard chow diet. C, Ad-
ipose tissue LPL activity in the fed state in
gonadal, perirenal, abdominal, and brown
adipose tissue. Note the marked increase
in the female Tg animals. Each group rep-
resents five animals. *, P ⬍ 0.05, Wt vs. Tg
mice.
TABLE 2. Serum metabolites in 6- to 8-month-old female and male Wt and ⌬GLY-Adiponectin (Tg) mice where n ⫽ 5–10 per group in a
controlled environment
6- to 8-month-old females 6- to 8-month-old males
Wt Tg Wt Tg
Fed glucagon (mEq/liter) 110 ⫾ 10 105 ⫾ 14 85 ⫾ 13 109 ⫾ 14
Fasted glucagon (mEq/liter) 229 ⫾ 42 289 ⫾ 31 182 ⫾ 34 230 ⫾ 26
Fed corticosterone (mg/dl) 30 ⫾ 428⫾ 318⫾ 320⫾ 2
Fasted corticosterone (mg/dl) 140 ⫾ 15 155 ⫾ 18 118 ⫾ 7 127 ⫾ 12
Prolactin (mEq/liter) 2.4 ⫾ 1.5 *10.9 ⫾ 4.1 7.5 ⫾ 2.2 *11.8 ⫾ 2.0
GH (mEq/liter) 3.7 ⫾ 0.6 5.2 ⫾ 2.2 3.3 ⫾ 0.75 5.5 ⫾ 1.2
Leptin (ng/dl) 4.1 ⫾ 0.4 3.5 ⫾ 0.5 3.3 ⫾ 0.75 3.9 ⫾ 0.5
TNF
␣
(ng/ml) 8.1 ⫾ 1.3 6.7 ⫾ 1.2 3.9 ⫾ 0.6 3.5 ⫾ 0.3
378 Endocrinology, January 2004, 145(1):367–383 Combs et al. • Transgenic Overexpression of Adiponectin
body weight due to a dramatic increase in interscapular
adipose mass. The expansion of interscapular and orbital
adipose tissues is also apparent in males older than 20
months. We previously showed that estrogen represses se-
rum adiponectin levels (20) and that ovariectomy results in
a significant increase in adiponectin levels in Wt mice. In-
terestingly, consistent with our hypothesis that the expan-
sion of interscapular and orbital adipose pads is a direct
consequence of increased serum adiponectin levels, ovari-
ectomy accelerated the remodeling of fat tissue in the Tg by
3–4 months (data not shown).
Based on the reports by Fruebis et al. (8) that increased
levels of the globular domain of adiponectin lead to increased

-oxidation in muscle, we suspected that overexpression of
adiponectin might lead to a lean phenotype. This is clearly
not the case. There is overall an increase in fat mass in females
and relatively normal levels in males. Indirect calorimetry
confirmed that there is no evidence for an increased meta-
bolic rate in these mice. Importantly, the studies by Fruebis
et al. (8) focused primarily on the effects of the globular
domain, whose pharmacological activities become increas-
ingly more established. However, this globular domain, a
putative proteolytic fragment, is difficult to visualize in vivo,
and it is not clear at this stage whether this fragment is a bona
fide processing intermediate or represents a pharmacological
antagonist only. We have also not been able to detect any
specific adiponectin degradation products in the Tg mice
with either amino- or with carboxyl-terminal-specific anti-
bodies. We cannot exclude the possibility that there is a
rate-limiting step along the activation pathway that we do
not overcome by mere overexpression of the full-length
product. Nevertheless, the overproduced material in Tg mice
is, by all possible criteria established to date, identical to
endogenous adiponectin observed under normal conditions.
Adiponectin in serum of Tg mice forms complexes of similar
molecular weight as adiponectin in Wt mice, with discrete
trimer-dimers and a high molecular weight complex. How-
ever, based on the data presented, it is clear that overex-
pression of full-length adiponectin is sufficient to cause
marked improvement of insulin sensitivity in liver. Yamau-
chi et al. (11) recently reported overexpression of the globular
adiponectin fragment from a liver-specific promoter. When
analyzed in a leptin-deficient ob/ob background, these mice
displayed increased expression of enzymes involved in

oxidation and energy dissipation (such as uncoupling pro-
teins), but did not display decreased adipose mass due to a
compensatory increase in food intake. Even though these
effects of the globular domain of adiponectin are well es-
tablished, data from a number of different laboratories in-
dicate the fundamentally different nature of this globular
domain compared with the full-length protein found in vivo.
Consistent with a recent report by Kadowaki and col-
leagues (23) demonstrating activation of AMPK by full-
length adiponectin in the liver, we observed increased acti-
vation of AMPK in our Tg mice at the end of the clamp
studies. This further corroborates the functional link between
increased adiponectin levels in serum and AMPK activation
in the liver.
Yokota et al. (24) reported that bacterially produced re-
combinant adiponectin blocked fat cell formation in long-
term bone marrow cultures and inhibited the differentiation
FIG. 9. A, Western blot analysis of caveolin-1 and adiponectin ex-
pression in fat pads isolated from Tg and Wt mice. Protein (50
g)
isolated from the respective fat pads was used for analysis. The figure
shows a representative Western from three independent experi-
ments. B, Northern blot analysis of female and male abdominal ad-
ipose tissues for PPAR
␥
, resistin, the mitochondrial ADP/ATP trans-
locator, and the mitochondrial dicarboxylate carrier mRNA levels.
Blots were also probed with

-actin as a loading control. The figure
shows a representative Northern blot from three independent exper-
iments.
TABLE 3. Hormone levels in 6- to 8-month-old female and male Wt and ⌬GLY-Adiponectin (Tg) mice where n ⫽ 5–10 per group in a
controlled environment
6- to 8-month-old females 6- to 8-month-old males
Wt Tg Wt Tg
Fed glucose (mEq/liter) 162 ⫾ 7 159 ⫾ 6 175 ⫾ 8 180 ⫾ 7
Fasted glucose (mEq/liter) 118 ⫾ 3 125 ⫾ 4 135 ⫾ 4 130 ⫾ 3
Fed fatty acids (mEq/liter) 110 ⫾ 10 105 ⫾ 14 85 ⫾ 13 109 ⫾ 14
Fasted fatty acids (mEQ/liter) 210 ⫾ 23 195 ⫾ 24 220 ⫾ 19 230 ⫾ 16
Fed cholesterol (mg/dl) 152 ⫾ 17 140 ⫾ 10 147 ⫾ 12 135 ⫾ 11
Fasted cholesterol (mg/dl) 120 ⫾ 9 111 ⫾ 12 115 ⫾ 8 110 ⫾ 7
Fed lactate (mg/dl) 45 ⫾ 840⫾ 339⫾ 542⫾ 7
Combs et al. • Transgenic Overexpression of Adiponectin Endocrinology, January 2004, 145(1):367–383 379
of cloned stromal preadipocytes. We have no evidence for an
inhibitory activity on differentiation based on the expression
of other differentiation-related marker proteins, such as re-
sistin (3, 16), whose expression is, in fact, slightly elevated in
fad pads from Tg animals. On the contrary, increased levels
of adiponectin cause expansion of interscapular and orbital
pads. The hyperproliferative response of the interscapular
adipose tissue is not unique and has been reported previ-
ously. Tg overexpression of serum response element-binding
protein-1c (SREBP-1c) in adipose tissue leads to development
of a large fat mass in the interscapular region consisting of
an enlarged, pale, bilobed fat pad highly similar in our mice
(25, 26). These mice, in contrast to ours, are severely insulin
resistant due to a generalized lipodystrophy, lack of leptin,
and the presence of fatty livers.
Acute treatment of mice with thiazolidinediones has se-
lective proliferative effects on interscapular brown adipo-
cytes as well and increases UCP1 expression (27), eventually
resulting in a very significant increase in adipose tissue in the
interscapular region. These observations were recently also
reported in both rats and monkeys treated with the potent
PPAR
␥
agonist darglitazone that resulted in hyperprolifera-
tion of a number of fat pads, in particular the dorsal thoracic
and the interscapular brown adipose pads (28). It is note-
worthy that thiazolidinediones also trigger a significant and
sustained increase in circulating adiponectin levels, similar
in magnitude to the increases achieved in our Tg model (12).
Additional similarities to chronic PPAR
␥
activation are
found at the level of adipose tissue LPL induction that can
be observed upon PPAR
␥
agonist treatment (18) as well as
in mice chronically overexpressing adiponectin, as demon-
strated here. Interestingly, Wilson-Fritch and colleagues (29)
recently reported that treatment of 3T3-L1 adipocytes with
the PPAR
␥
agonist rosiglitazone leads to a dramatic up-
regulation of a host of nuclear-encoded mitochondrial pro-
teins, consistent with other observations made in WAT of rats
and dogs that describe significant effects of rosiglitazone
treatment on mitochondrial number and morphology (30).
FIG. 10. Rate of glucose uptake and
glucose production in 6- to 8-month-old
female (A and B) and male (B) ⌬Gly-
adiponectin mice (f) and their Wt lit-
termates (s) during clamps at low (A)
and high (B) insulin levels. C, Phospho-
AMPK and AMPK levels in liver ex-
tracts isolated from Wt and Tg female
mice killed either under basal condi-
tions (basal) or at the end of a hyperin-
sulinemic clamp study (4 mU/kg/min).
In all cases, 50
g liver protein were
used for analysis. *, P ⬍ 0.05, Wt vs. Tg
mice.
TABLE 4. Insulin clamp parameters in 6 – 8 month old female and male Wt and ⌬GLY-Adiponectin (Tg) mice.
Insulin (mg/kg䡠h) Sex
N Body weight (g) Basal insulin (ng/ml) Liver lipid (% weight)
Wt Tg Wt Tg Wt Tg Wt Tg
1.0 F 6 6 28.2 ⫾ 1.5 32.7 ⫾ 2.0 1.26 ⫾ 0.31 *0.45 ⫾ 0.14 N/A N/A
4.0 F 3 3 32.1 ⫾ 3.4 34.8 ⫾ 1.6 1.43 ⫾ 0.28 *0.67 ⫾ 0.21 1.2 ⫾ 0.2 1.6 ⫾ 0.4
4.0 M 5 5 38.7 ⫾ 3.6 37.8 ⫾ 0.8 1.64 ⫾ 0.31 *0.89 ⫾ 0.35 1.4 ⫾ 0.5 1.5 ⫾ 0.1
F, Female; M, Male.
380 Endocrinology, January 2004, 145(1):367–383 Combs et al. • Transgenic Overexpression of Adiponectin
These increased levels of mitochondrial proteins may explain
in part the ability of PPAR
␥
agonists to increase

oxidation,
which results in an increased clearance of free fatty acids that,
in turn, can improve insulin sensitivity (31). Finally, even in
humans there is an apparent propensity for hyperprolifera-
tion of a similar fat pad in the context of human immuno-
deficiency virus-mediated lipodystrophies, particularly in
individuals undergoing highly aggressive antiretroviral
therapy (HAART) (32, 33). There is, therefore, an increased
propensity for the adipose tissue located in the interscapular
region to undergo proliferation under a number of different
conditions. Mechanistically, however, we do not know the
link between serum response element-binding protein-1c
overexpression in adipose tissue, thiazolidinedione-in-
creased serum adiponectin levels, and human immunodefi-
ciency virus protease inhibitors.
Orbital fat proliferation is a unique phenomenon among
animal models. Only one other mouse model, the TSH re-
ceptor-induced model of Graves’ ophthalmopathy, has dis-
played a similar phenotype, albeit on a much less dramatic
scale (34). In this respect, our Tg mice recapitulate some
phenotypic changes associated with Graves’ exophthalmia.
However, the Tg mice do not show evidence of hyperthy-
roidism or TSH receptor autoantibodies. It is therefore un-
likely that the underlying causes of Graves’ disease are di-
rectly linked to adiponectin overexpression. It is noteworthy
that in a small cohort of Graves’ patients displaying exoph-
thalmos, we found serum adiponectin levels elevated by a
factor of 2–3 above average levels and are currently testing
whether the increased serum adiponectin is due to the hy-
perthyroidism per se (Combs, T. P., G. Baker, E. Ludgate, and
P. E. Scherer, unpublished observations). Nevertheless, this
suggests that the hypertrophy of the orbital fat pad observed
in a subset of Graves’ patients may be directly linked to the
elevation of serum adiponectin, which, in turn, is secondary
to a hormonal change induced by the autoimmune reaction.
Consistent with that, a recent case has been reported in which
a type 2 diabetic patient, upon initiation of treatment with the
PPAR
␥
agonist pioglitazone, experienced exacerbation of his
thyroid eye disease, which had been stable and inactive for
more then 2 yr. This patient displayed a massive expansion
of the orbital fat pad (35).
This is the first time that orbital proliferation has been
observed in mice to such a significant extent. Therefore, this
mouse model offers a tool to study noninflammatory fat
proliferation as often seen in Graves’ ophthalmopathy.
FIG. 11. A, Female and male body weights in
Wt and ⌬Gly-adiponectin (Tg) mice after 6
months of a high fat diet initiated at weaning.
B and C, Circulating glucose and insulin in
6-month-old high fat-fed female (left) and male
(right) Tg mice (F) and their Wt littermates (E)
after an oral glucose challenge. High fat feeding
was initiated at weaning; mice were analyzed at
6 months of age. *, P ⬍ 0.05, Wt vs. Tg mice.
Combs et al. • Transgenic Overexpression of Adiponectin Endocrinology, January 2004, 145(1):367–383 381
It is surprising that the female Tg mice display such dramatic
improvements in insulin sensitivity despite their increased ad-
ipose mass. Again, this phenomenon resembles the effects of
thiazolidinedione treatment, which is known to lead to in-
creases in serum adiponectin levels, insulin sensitivity, fat cell
differentiation, and fat mass. Importantly, though, we ruled out
the possibility that the improvement in insulin sensitivity is a
secondary effect of the interscapular hyperproliferative fat pad,
because we observed similarly striking improvements in insu-
lin sensitivity in young male mice that showed no evidence of
adipose depot redistribution.
In a very recent paper, Yamauchi and colleagues (36) re-
ported the cloning of two receptors for adiponectin. Inter-
estingly, one of these receptors shows preferential binding to
the full-length native version of adiponectin and is predom-
inantly expressed in the liver. The other receptor displays
preferential binding to the globular trimeric version of adi-
ponectin and shows a more ubiquitous tissue distribution,
with strong expression in muscle. This is fully consistent with
previous reports that show a differential effect of the two
ligands on liver and muscle, respectively. It will be interest-
ing to determine whether these receptors play any role in the
differential adipose tissue distribution observed in the Tg
mouse model presented here.
In summary, this is the first mouse model in which adi-
ponectin levels are chronically elevated. This elevation is
within physiological levels and accurately recapitulates the
changes in endogenous adiponectin levels during matura-
tion of the animal. These elevated serum adiponectin levels
dramatically improve insulin sensitivity and at a later age
lead to the hypertrophy of select fat pads. It is not clear
whether this hypertrophy is the consequence of an autocrine
effect of adiponectin or an indirect effect of improved sys-
temic insulin sensitivity or is due to a hormonal change
secondary to adiponectin overexpression. Consistent with
our acute pharmacological studies, chronic elevation of full-
length adiponectin in serum had a positive effect on insulin
sensitivity, primarily in the liver. As several groups have
implicated adiponectin as a possible mediator of thiazo-
lidinedione action (12, 37–39), it is striking that chronic ele-
vation of adiponectin by means of Tg overexpression causes
effects that are vastly overlapping with those of chronic
treatment with thiazolidinediones. This further strengthens
the relationship between the effects of thiazolidinedione
treatment and the induction of adiponectin as a prerequisite
step to achieve clinical improvement in insulin sensitivity.
Acknowledgments
We are thankful to Dr. Rebecca Bahn (Mayo Clinic, Rochester, MN)
for expert advice on various aspect of Graves’ ophthalmopathy, Dr.
James F. Nelson (University of Texas Health Science Center, San An-
tonio, TX) for help with the measurement of corticosterone, and Carolyn
Marks and Leslie Cummins (AECOM) for help with the EM imaging.
Received August 18, 2003. Accepted October 7, 2003.
Address all correspondence and requests for reprints to: Dr. Philipp
E. Scherer, Department of Cell Biology, Albert Einstein College of Med-
icine, 1300 Morris Park Avenue, Bronx, New York 10461. E-mail:
scherer@aecom.yu.edu.
This work was supported by NIH National Research Service Award
DK-61228 (to T.P.C.), NIH Hormones/Membrane Interactions Training
Grant T32-DK-07513-15 (to T.P.C.), an American Diabetes Medical Sci-
entist Training Grant (to A.H.B.), NIH Medical Scientist Training Grant
T32-GM-07288 (to U.B.P. and M.W.R.), a postdoctoral fellowship from
the Swiss National Science Foundation (to A.R.N.), a grant from the
Wales Office for Research and Development (to G.B. and M.L.), the Core
Laboratories of the Albert Einstein Diabetes Research and Training
Center, NIH Grants R01-DK-45024 and R01-DK-48321 (to L.R.), and NIH
Grants R01-DK-55758 and R03-EY-014935 (to P.E.S.).
References
1. Havel, PJ 2002 Control of energy homeostasis and insulin action by adipocyte
hormones: leptin, acylation stimulating protein, and adiponectin. Curr Opin
Lipidol 13:51–59
2. Rajala MW, Scherer PE 2003 Minireview: the adipocyte: at the crossroads of
energy homeostasis, inflammation, and atherosclerosis. Endocrinology 144:
3765–3773
3. Steppan CM, Bailey ST, Bhat S, Brown EJ, Banerjee RR, Wright CM, Patel
HR, Ahima RS, Lazar MA 2001 The hormone resistin links obesity to diabetes.
Nature 409:307–312
4. Berg AH, Combs T, Du X, Brownlee M, Scherer PE 2001 The adipocyte-
secreted protein Acrp30 enhances hepatic insulin action. Nat Med 7:947–953
5. Hotta K, Funahashi T, Arita Y, Takahashi M, Matsuda M, Okamoto Y,
Iwahashi H, Kuriyama H, Ouchi N, Maeda K, Nishida M, Kihara S, Sakai
N, Nakajima T, Hasegawa K, Muraguchi M, Ohmoto Y, Nakamura T,
Yamashita S, Hanafusa T, Matsuzawa Y 2000 Plasma concentrations of
a novel, adipose-specific protein, adiponectin, in type 2 diabetic patients.
Arterioscler Thromb Vasc Biol 20:1595–1599
6. Hotta K, Funahashi T, Bodkin NL, Ortmeyer HK, Arita Y, Hansen BC,
Matsuzawa Y 2001 Circulating concentrations of the adipocyte protein adi-
ponectin are decreased in parallel with reduced insulin sensitivity during the
progression to type 2 diabetes in rhesus monkeys. Diabetes 50:1126–1133
7. Combs TP, Berg AH, Obici S, Scherer PE, Rossetti L 2001 Endogenous glucose
production is inhibited by the adipose-derived protein Acrp30. J. Clin. Inv.
108:1875–1881
8. Fruebis J, Tsao TS, Javorschi S, Ebbets-Reed D, Erickson MR, Yen FT, Bihain
BE, Lodish HF 2001 Proteolytic cleavage product of 30-kDa adipocyte com-
plement-related protein increases fatty acid oxidation in muscle and causes
weight loss in mice. Proc Natl Acad Sci USA 98:2005–2010
9. Maeda N, Shimomura I, Kishida K, Nishizawa H, Matsuda M, Nagaretani
H, Furuyama N, Kondo H, Takahashi M, Arita Y, Komuro R, Ouchi N, Kihara
S, Tochino Y, Okutomi K, Horie M, Takeda S, Aoyama T, Funahashi T,
Matsuzawa Y 2002 Diet-induced insulin resistance in mice lacking adipo-
nectin/ACRP30. Nat Med 8:731–737
10. Kubota N, Terauchi Y, Yamauchi T, Kubota T, Moroi M, Matsui J, Eto K,
Yamashita T, Kamon J, Satoh H, Yano W, Nagai R, Kimura S, Kadowaki T,
Noda T 2002 Disruption of adiponectin causes insulin resistance and neo-
intimal formation. J Biol Chem 277:25863–25866
11. Yamauchi T, Kamon J, Waki H, Imai Y, Shimozawa N, Hioki K, Uchida S,
Ito Y, Matsui J, Eto K, Komeda K, Tsunoda M, Murakami K, Ohnishi Y,
Yamamura K, Ueyama Y, Froguel P, Kimura S, Nagai R, Kadowaki T 2002
Globular adiponectin protected ob/ob mice from diabetes and apoE deficient
mice from atherosclerosis. J Biol Chem 12:12
12. Combs TP, Wagner JA, Berger J, Doebber T, Wang W-J, Zhang BB, Tanen
M, Berg AH, O’Rahilly S, Savage DS, Chatterjee K, Weiss S, Larson PJ,
Gottesdiener KM, Gertz BG, Charron MJ, Scherer PE, Moller DE 2002
Induction of adipocyte complement-related protein of 30 kilodaltons by
PPAR
␥
agonists: a potential mechanism of insulin sensitization. Endocrinol-
ogy 143:998–1007
13. Lindemann D, Patriquin E, Feng S, Mulligan RC 1997 Versatile retrovirus
vector systems for regulated gene expression in vitro and in vivo. Mol Med
3:466–476
14. Gossen M, Freundlieb S, Bender G, Muller G, Hillen W, Bujard H 1995
Transcriptional activation by tetracyclines in mammalian cells. Science 268:
1766–179
15. Engelman JA, Lisanti MP, Scherer PE 1998 Specific inhibitors of p38 mitogen-
activated protein kinase block 3T3-L1 adipogenesis. J Biol Chem 273:32111–
32120
16. Rajala MW, Lin Y, Ranalletta M, Yang XM, Qian H, Gingerich R, Barzilai
N, Scherer PE 2002 Cell type-specific expression and coregulation of murine
resistin and resistin-like molecule-
␣
in adipose tissue. Mol Endocrinol 16:
1920–1930
17. Pajvani UB, Du X, Combs TP, Berg AH, Rajala MW, Schulthess T, Engel J,
Brownlee M, Scherer PE 2003 Structure-function studies of the adipocyte-
secreted hormone Acrp30/adiponectin: implications for metabolic regulation
and bioactivity. J Biol Chem 278:9073–9085
18. Laplante M, Sell H, MacNaul KL, Richard D, Berger JP, Deshaies Y 2003
PPAR-
␥
activation mediates adipose depot-specific effects on gene expression
and lipoprotein lipase activity: mechanisms for modulation of postprandial
lipemia and differential adipose accretion. Diabetes 52:291–299
19. Shapiro L, Scherer PE 1998 The crystal structure of a complement-1q family
382 Endocrinology, January 2004, 145(1):367–383 Combs et al. • Transgenic Overexpression of Adiponectin
protein suggests an evolutionary link to tumor necrosis factor. Curr Biol
8:335–338
20. Combs TP, Berg AH, Rajala MW, Klebanov S, Iyengar P, Jimenez-Chillaron
JC, Patti ME, Klein SL, Weinstein RS, Scherer PE 2003 Sexual differentiation,
pregnancy, calorie restriction and aging affect the adipocyte-specific secretory
protein Acrp30/adiponectin. Diabetes 52:268–276
21. Ludgate M, Baker G 2002 Unlocking the immunological mechanisms of orbital
inflammation in thyroid eye disease. Clin Exp Immunol 127:193–198
22. Razani B, Combs TP, Wang XB, Frank PG, Park DS, Russell RG, Li M,
Tang B, Jelicks LA, Scherer PE, Lisanti MP 2002 Caveolin-1-deficient mice are
lean, resistant to diet-induced obesity, and show hypertriglyceridemia with
adipocyte abnormalities. J Biol Chem 277:8635–8647
23. Yamauchi T, Kamon J, Minokoshi Y, Ito Y, Waki H, Uchida S, Yamashita S,
Noda M, Kita S, Ueki K, Eto K, Akanuma Y, Froguel P, Foufelle F, Ferre P,
Carling D, Kimura S, Nagai R, Kahn BB, Kadowaki T 2002 Adiponectin
stimulates glucose utilization and fatty-acid oxidation by activating AMP-
activated protein kinase. Nat Med 8:1288–1295
24. Yokota T, Meka CS, Medina KL, Igarashi H, Comp PC, Takahashi M,
Nishida M, Oritani K, Miyagawa J, Funahashi T, Tomiyama Y, Matsuzawa
Y, Kincade PW 2002 Paracrine regulation of fat cell formation in bone marrow
cultures via adiponectin and prostaglandins. J Clin Invest 109:1303–1310
25. Shimomura I, Hammer RE, Richardson JA, Ikemoto S, Bashmakov Y,
Goldstein JL, Brown MS 1998 Insulin resistance and diabetes mellitus in
transgenic mice expressing nuclear SREBP-1c in adipose tissue: model for
congenital generalized lipodystrophy. Genes Dev 12:3182–3194
26. Shimomura I, Hammer RE, Ikemoto S, Brown MS, Goldstein JL 1999 Leptin
reverses insulin resistance and diabetes mellitus in mice with congenital lip-
odystrophy. Nature 401:73–76
27. Burkey BF, Dong M, Gagen K, Eckhardt M, Dragonas N, Chen W,
Grosenstein P, Argentieri G, de Souza CJ 2000 Effects of pioglitazone on
promoting energy storage, not expenditure, in brown adipose tissue of obese
fa/fa Zucker rats: comparison to CL 316,243. Metabolism 49:1301–1308
28. Aleo MD, Lundeen GR, Blackwell DK, Smith WM, Coleman GL, Stadnicki
SW, Kluwe WM 2003 Mechanism and implications of brown adipose tissue
proliferation in rats and monkeys treated with the thiazolidinedione dargli-
tazone, a potent peroxisome proliferator-activated receptor-
␥
agonist. J Phar-
macol Exp Ther 305:1173–1182
29. Wilson-Fritch L, Burkart A, Bell G, Mendelson K, Leszyk J, Nicoloro S,
Czech M, Corvera S 2003 Mitochondrial biogenesis and remodeling during
adipogenesis and in response to the insulin sensitizer rosiglitazone. Mol Cell
Biol 23:1085–1094
30. Toseland CD, Campbell S, Francis I, Bugelski PJ, Mehdi N 2001 Comparison
of adipose tissue changes following administration of rosiglitazone in the dog
and rat. Diabetes Obes Metab 3:163–170
31. Yu C, Chen Y, Cline GW, Zhang D, Zong H, Wang Y, Bergeron R, Kim JK,
Cushman SW, Cooney GJ, Atcheson B, White MF, Kraegen EW, Shulman GI
2002 Mechanism by which fatty acids inhibit insulin activation of insulin
receptor substrate-1 (IRS-1)-associated phosphatidylinositol 3-kinase activity
in muscle. J Biol Chem 277:50230–50236
32. Qaqish RB, Fisher E, Rublein J, Wohl DA 2000 HIV-associated lipodystrophy
syndrome. Pharmacotherapy 20:13–22
33. Gelato MC, Mynarcik DC, Quick JL, Steigbigel RT, Fuhrer J, Brathwaite CE,
Brebbia JS, Wax MR, McNurlan MA 2002 Improved insulin sensitivity and
body fat distribution in HIV-infected patients treated with rosiglitazone: a pilot
study. J Acquired Immune Defic Syndr 31:163–170
34. Many MC, Costagliola S, Detrait M, Denef F, Vassart G, Ludgate MC 1999
Development of an animal model of autoimmune thyroid eye disease. J Im-
munol 162:4966–4974
35. Starkey K, Heufelder A, Baker G, Joba W, Evans M, Davies S, Ludgate M
2003 Peroxisome proliferator-activated receptor-
␥
in thyroid eye disease: con-
traindication for thiazolidinedione use? J Clin Endocrinol Metab 88:55–59
36. Yamauchi T, Kamon J, Ito Y, Tsuchida A, Yokomizo T, Kita S, Sugiyama T,
Miyagishi M, Hara K, Tsunoda M, Murakami K, Ohteki T, Uchida S,
Takekawa S, Waki H, Tsuno NH, Shibata Y, Terauchi Y, Froguel P, Tobe K,
Koyasu S, Taira K, Kitamura T, Shimizu T, Nagai R, Kadowaki T 2003
Cloning of adiponectin receptors that mediate antidiabetic metabolic effects.
Nature 423:762–769
37. Maeda N, Takahashi M, Funahashi T, Kihara S, Nishizawa H, Kishida K,
Nagaretani H, Matsuda M, Komuro R, Ouchi N, Kuriyama H, Hotta K,
Nakamura T, Shimomura I, Matsuzawa Y 2001 PPAR
␥
ligands increase
expression and plasma concentrations of adiponectin, an adipose-derived
protein. Diabetes 50:2094–2099
38. Moore GB, Chapman H, Holder JC, Lister CA, Piercy V, Smith SA,
Clapham JC 2001 Differential regulation of adipocytokine mRNAs by rosig-
litazone in db/db mice. Biochem Biophys Res Commun 286:735–741
39. Yu JG, Javorschi S, Hevener AL, Kruszynska YT, Norman RA, Sinha M,
Olefsky JM 2002 The effect of thiazolidinediones on plasma adiponectin levels
in normal, obese, and type 2 diabetic subjects. Diabetes 51:2968–2974
Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the
endocrine community.
Combs et al. • Transgenic Overexpression of Adiponectin Endocrinology, January 2004, 145(1):367–383 383