Unité Mixte de Recherche 7079, CNRS-Paris VI, Centre de
Recherches Biomédicales des Cordeliers, Paris, France
Membrane-associated semicarbazide-sensitive amine oxidase (SSAO) is
mainly present in the media of aorta and in adipose tissue. Recent
works have reported that SSAO activation can stimulate glucose
transport of fat cells and promote adipose conversion. In this study,
the murine 3T3-L1 preadipose cell line was used to investigate SSAO
regulation by tumor necrosis factor-
(TNF-), a cytokine that is
synthesized in fat cells and known to be involved in obesity-linked
insulin resistance. SSAO mRNA and protein levels, and enzyme activity
were decreased by TNF- in a dose- and time-dependent manner, without
any change of SSAO affinity for substrates or inhibitors. SSAO
inhibition caused by TNF- was spontaneously reversed along the time
after TNF- removal. The decrease in SSAO expression also occurred in
white adipose tissue of C57BL/6 mice treated with mTNF-. Overall, we
demonstrated that reduction in SSAO expression induced by the cytokine
had marked repercussions on amine-stimulated glucose transport, in a
dose- and time-dependent manner. This effect was more pronounced than
the inhibiting effect of TNF- on insulin-stimulated glucose
transport. Moreover, the peroxisome proliferator-activated receptor
agonists thiazolidinediones did not reverse either TNF-
effect on amine-sensitive glucose transport or the inhibition of SSAO
activity, whereas they antagonized TNF- effects on insulin-sensitive
glucose transport. These results demonstrate that TNF- can strongly
down-regulate SSAO expression and activity, and through this mechanism
can dramatically reduce amine-stimulated glucose transport. This
suggests a potential role of this regulatory process in the
pathogenesis of glucose homeostasis dysregulations observed during
diseases accompanied by TNF- overproduction, such as cachexia or obesity.
 |
Introduction |
Copper-containing
amine oxidases form a specific family of enzymes that deaminate some
aromatic or aliphatic amines to generate ammonia, hydrogen peroxide,
and the corresponding aldehydes. An original member of this group is a
membrane-bound amine oxidase, highly inhibited by semicarbazide, and
often referred to as the "tissue-bound" semicarbazide-sensitive
amine oxidase (SSAO) (Lyles, 1996
; Jalkanen and Salmi, 2001). This
enzyme readily oxidizes exogenous (e.g., benzylamine) or endogenous
(e.g., methylamine and aminoacetone) primary amines. SSAO, which is
identical to vascular adhesion protein-1, has been cloned in different
species (Zhang and McIntire, 1996; Morris et al., 1997; Bono et al.,
1998; Smith et al., 1998; Moldes et al., 1999). SSAO activity and
transcripts have been detected in a variety of tissues and cell types
(Lyles, 1996), but they are prominently expressed in vascular smooth
muscle cells (Lyles, 1996), endothelial cells of lymph venules
(Jalkanen and Salmi, 2001), and in white and brown adipocytes (Barrand
and Callingham, 1982; Raimondi et al., 1991).
The physiological and pathophysiological roles of SSAO are still
unclear and depend on the cell type on which the enzyme is expressed
(Lyles and Pino, 1998; Jalkanen and Salmi, 2001; El Hadri et al.,
2002). In fat cells, it has been recently documented that SSAO
expression is strongly induced during preadipocyte differentiation (Moldes et al., 1999; Fontana et al., 2001) and that SSAO chronic activation promotes terminal adipocyte maturation (Fontana et al.,
2001; Mercier et al., 2001). Furthermore, SSAO is not exclusively present at the plasma membrane of adipocytes, but is also detectable in
vesicles containing the insulin-sensitive glucose transporter GLUT4
(Morris et al., 1997; Enrique-Tarancon et al., 1998). The acute effect
of insulin on glucose transport can be mimicked by SSAO substrates
(Enrique-Tarancon et al., 1998, 2000; Marti et al., 1998; Fontana et
al., 2001; Morin et al., 2001) through the release of hydrogen peroxide
(Enrique-Tarancon et al., 1998; Marti et al., 1998). SSAO substrates
can also promote IRS-1 and -3 tyrosine phosphorylation, and stimulate
phosphatidylinositol 3-kinase activity and GLUT4 translocation to the
plasma membrane (Enrique-Tarancon et al., 2000). A recent study (Marti
et al., 2001) has underlined the potential therapeutic interest of SSAO
in the control of glycemia: an acute or chronic administration of the
synthetic SSAO substrate benzylamine in combination with low doses of
vanadate enhances glucose tolerance and reduces hyperglycemia in
streptozotocin-induced diabetic rats. Interestingly, several studies
have reported that SSAO activity is increased in serum from diabetic
(Boomsma et al., 1999; Meszaros et al., 1999) or obese (Meszaros et
al., 1999) patients. Taken together, these observations suggest that
adipocyte SSAO could play a significant role in the control of glucose homeostasis.
TNF- is a cytokine involved in the clinical and metabolic
disturbances observed in obesity-linked insulin resistance: TNF- is
overexpressed in adipose tissue of obese rodents or humans, and
administration of TNF- to animals induces insulin resistance, whereas neutralization of TNF- improves insulin sensitivity
(Hotamisligil, 2000; Moller, 2000). Otherwise, results from
knockout mice deficient in TNF- or its receptors indicate that the
cytokine can regulate in vivo insulin sensitivity and is involved at
least partly in the onset of obesity-associated insulin resistance
(Hotamisligil, 2000; Moller, 2000). However, the exact contribution of
TNF- to the pathophysiology of insulin resistance observed in human obesity remains controversial (Moller, 2000).
Multiple cellular and molecular mechanisms could account for the
metabolic effects of TNF- on adipose tissue. Thus, the cytokine potently suppresses the expression of genes encoding proteins involved
in fatty acid uptake and lipogenesis. TNF- also inhibits preadipocyte differentiation (Torti et al., 1985) and provokes apoptosis (Prins et al., 1997). TNF- stimulates lipolysis and increases free fatty acids through different mechanisms. It is also
well known that in adipocytes, TNF- strongly inhibits
insulin-stimulated glucose transport (Stephens and Pekala, 1991;
Szalkowski et al., 1995), through different alterations in insulin
signaling pathway (Hotamisligil, 2000; Moller, 2000).
Considering the TNF--induced alterations in adipocyte hexose
transport, and as regards to the promoting effect of SSAO activation on
fat cell glucose uptake, the aim of our study was to investigate the
impact of TNF- on the expression and function of this
membrane-associated amine oxidase. Using the murine preadipose 3T3-L1
cell line, we demonstrate that TNF- decreases SSAO expression
without altering adipocyte differentiation level. This effect is also
observed in white adipose tissue of TNF--injected mice. Overall,
this TNF--induced down-regulation of SSAO expression is involved in a dramatic reduction in amine-stimulated glucose transport.
| |
Materials and Methods |
Cell Culture.
Stocks of murine 3T3-L1 preadipocytes were
maintained as described previously (Moldes et al., 1999). For
experiments, cells were seeded at a density of
104/cm2 in plastic culture
dishes (Falcon, Cowley, UK) and were grown in DMEM supplemented with
10% fetal calf serum (Biomedia, Boussens, France), 100 units/ml
penicillin, and 50 µg/ml streptomycin (Invitrogen, Carlsbad, CA) in a
10% CO2-humidified atmosphere. Adipocyte
differentiation was initiated by administration at confluence of
methylisobutylxanthine (100 µM), dexamethasone (0.25 µM), and
insulin (1 µg/ml) for 48 h, then cells were refed by DMEM
containing 10% fetal calf serum and 1 µg/ml insulin. Using this
protocol, more than 95% of the cells acquired an adipocyte morphology
at day 8 after confluence. After washing, mature adipocytes were kept
for 16 h in a defined medium consisting of DMEM/Ham's F-12 medium
(2:1, v/v) and 0.1% bovine serum albumin. Thereafter, cells were
maintained either in the absence or in the presence of mTNF-
(Sigma-Aldrich, St. Louis, MO) at concentrations and for periods of
time mentioned under Results.
Animals.
Four-week-old male C57BL/6 mice received a single
intraperitoneal injection of 1 µg of recombinant mTNF-.
Forty-eight hours after TNF- administration, animals were killed and
epididymal fat pads were quickly excised and immediately frozen in
liquid nitrogen until preparation of tissue extracts. Experiments were undertaken according to the Guidelines for the Care and the Use of
Experimental Animals.
Cell and Tissue Extracts, Biochemical Determinations, and Enzyme
Assays.
3T3-L1 cells were washed twice in PBS, harvested, and
homogenized in 25 mM Tris, pH 7.5, 1 mM EDTA (20 strokes in a Dounce homogenizer, pestle B). A fraction of the homogenate was stored at
80°C. The remaining fraction was centrifuged at 10,000g
for 10 min at 4°C, and the supernatant was kept at 80°C until
use. The 10,000g pellet was resuspended in the
homogenization buffer and stored at 80°C. Aliquots of homogenates,
supernatants, and resuspended pellets were used to determine protein
content by the method of Lowry, using bovine serum albumin as a
standard. Triglyceride concentration was determined with the Infinity
triglycerides kit (Sigma Diagnostics, St. Louis, MO).
For preparation of tissue extracts, the epididymal fat depot was
homogenized in a buffer consisting of
KH2PO4 1 mM, pH 7.8, 250 mM
sucrose. After a centrifugation at 600g for 5 min at 4°C, the pellet and the fat cake were discarded, and the supernatant was
kept at 80°C until SSAO activity measurement.
SSAO activity was tested by measurement of hydrogen peroxide production
by the fluorometric method of Matsumoto et al. (1982). Briefly, 25 µg
of cell homogenates was preincubated for 30 min at 37°C, in a final
volume of 100 µl containing 40 mM sodium phosphate, pH 7.4, 1 mM
homovanillic acid, 1 mM sodium azide, and 1 mM pargyline to inhibit
monoamine oxidase A and B. When indicated, an SSAO inhibitor was
preincubated with cell extract for 30 min at 37°C before the addition
of the substrate. Incubation was initiated by the addition of the SSAO
substrate (benzylamine, methylamine, tyramine, or 
-phenylethylamine)
and was carried out in duplicate for 1 h at 37°C. Reaction was
stopped with 1 mM semicarbazide, and 1.2 ml of 0.1 N NaOH was added.
Fluorescence intensity was measured with excitation at 323 nm and
emission at 426 nm. As blank tests, assays were incubated in parallel
without substrate addition. Preliminary experiments were performed to
ensure that SSAO activity was tested at the initial rate of reaction.
Apparent kinetic constants for SSAO substrates were determined within
the following concentration ranges: 3 to 300 µM for benzylamine, 0.1 to 10 mM for methylamine, 0.1 to 10 mM for tyramine, and 30 µM to 3 mM for -phenylethylamine. To determine IC50
and Ki values for amine oxidase
inhibitors, SSAO activity was determined in the presence of 30 µM
benzylamine as a substrate. The concentration ranges of inhibitors used
were as follows: 1 µM to 1 mM semicarbazide, 1 to 100 µM
aminoguanidine, 0.1 to 100 nM phenelzine, 1 to 100 nM hydrazine, 3 to
100 nM hydralazine, and 30 nM to 3 µM benserazide. Kinetic parameters
were determined using the nonlinear regression analysis curve fitting
procedure of the ENZFITTER program (Biosoft-Elsevier, Cambridge, UK).
Monoamine oxidase activity was tested under conditions identical to
those for SSAO, except that pargyline was omitted and replaced by 1 mM
semicarbazide, and tyramine or serotonin was used as substrate.
Polyamine oxidase activity was also tested with the same procedure,
using N1-acetyl spermidine as a substrate.
G3PDH activity was assayed by recording the initial rate of oxidation
of NADH at 340 nm at 25°C (Mercier et al., 2001). The standard
mixture contained 50 mM triethanolamine-HCl buffer, pH 7.5, 1 mM EDTA,
0.13 mM -NADH, 1 mM dihydroxyacetone phosphate, 1 mM
2-mercaptoethanol, and variable amounts of 10,000g cell supernatants.
Adenylyl cyclase activity was measured on 10,000g pellet
fractions as described previously (El Hadri et al., 1997).
All enzyme substrates and inhibitors were from Sigma-Aldrich.
Western Blot Analysis.
Cells were washed three times with
cold PBS, harvested, and homogenized in 1 ml of HES buffer (20 mM
Hepes, pH 7.4, 1 mM EDTA, 250 mM sucrose) supplemented with a cocktail
of anti-proteases (Complete; Roche Diagnostics, Indianapolis, IN).
Homogenates were centrifuged at 200,000g for 90 min at
4°C. The resulting pellets were resuspended in 30 mM Hepes buffer, pH
7.4, and stored at 80°C.
SDS-polyacrylamide gel electrophoresis was performed with a mini
protean III apparatus (Bio-Rad, Hercules, CA). Proteins (30 µg/lane)
were boiled for 5 min at 95°C, separated on a 7.5%
polyacrylamide-SDS gel, and electroblotted onto 0.45-µm
polyvinylidene difluoride membrane (Immobilon-P; Millipore
Corporation, Bedford, CA) in 0.1% SDS, 192 mM glycine, and 25 mM Tris,
pH 8.3. The membrane was dipped in a methanol bath and then blocked
with 5% fish gelatin, 0.1% Tween 20 in TBS buffer (20 mM Tris, 137 mM
NaCl, pH 7.6) for 45 min at room temperature. After an initial washing
in TBS buffer containing 0.1% Tween 20, the membrane was incubated
overnight at 4°C with primary antibodies (rabbit polyclonal antibody
against bovine lung SSAO, dilution 1:2000; Lizcano et al., 1998) in TBS buffer containing 2% fish gelatin and 0.05% Tween 20. After four washes with 0.1% Tween 20-TBS, the membrane was incubated with horseradish peroxidase-conjugated anti-rabbit IgG (dilution 1:20,000 in
TBS buffer containing 0.05% Tween 20 and 2% fish gelatin), for 1 h at room temperature. The membrane was then washed in 0.1% Tween
20-TBS. Detection of immune complex was performed using an enhanced
chemiluminescence detection kit for Western blot (Amersham Biosciences
UK, Ltd., Little Chalfont, Buckinghamshire, UK) on a X-OMAT-AR Kodak
film (Eastman Kodak, Rochester, NY).
RNA Analysis.
Total RNA was extracted from 3T3-L1 adipocytes
and from epididymal white adipose tissue with the RNA-PLUS kit
procedure (QBIOgene, Illkirch, France).
For real time RT-PCR analysis of SSAO gene expression, total RNA was
first digested for 15 min at 37°C with 0.1 unit of RNase-free DNase I
(RQ1 DNase; Promega, Madison, WI)/µg of nucleic acid in 40 mM
Tris-HCl, pH 7.9, 10 mM NaCl, 6 mM MgCl2, and 10 mM CaCl2. After phenol/chloroform extraction and
ethanol precipitation, RNA (0.25-1 µg/sample) was reverse
transcribed with Moloney murine leukemia virus-reverse transcriptase
(200 units/µg) (Invitrogen) in the presence of 10 µM random
hexanucleotides (Amersham Biosciences UK, Ltd.), 400 µM of each dNTP
in a final volume of 40 µl consisting of 50 mM Tris-HCl, pH 8.3, 75 mM KCl, 3 mM MgCl2, and 10 mM dithiothreitol. After a 1-h incubation at 42°C, Moloney murine leukemia virus-reverse transcriptase was heat-inactivated. To ensure that subsequent amplification did not derive from contaminant genomic DNA, a control without MMLV-RT was included in parallel for each RNA sample.
Reverse transcribed SSAO mRNAs were amplified on ICycler thermal cycler
(Bio-Rad), using the SYBR green fluorescence method, and were analyzed
with the ICycler IQ real-time detection system software. PCR assay was
performed under a final volume of 25 µl, starting from 12.5 to 50 ng
of reverse transcribed total RNA, in the presence 250 nM of each sense
and antisense oligonucleotide, 125 µM of each dNTP, 2.5 mM
MgCl2, 0.5 unit TaqDNA polymerase in
its recommended buffer, including 1× SYBR green (Eurogentec, Seraing,
Belgium). To verify that fluorescence generated by SYBR green
incorporation into double-stranded DNA was not overestimated by
contaminations resulting from residual genomic DNA amplification and/or
from primer dimer formation, controls without reverse transcriptase, and without DNA template and reverse transcriptase were included in
each experiment. Moreover, RT-PCR products were analyzed on agarose
gels stained with ethidium bromide and in a postamplification fusion
curve to ensure that a single amplicon was obtained. Ribosomal 18S RNA
levels that remain stable during TNF- exposure were used to
normalize the initial amounts of cDNA. To measure PCR efficiency, serial dilutions of reverse transcribed RNA (0.4-12 ng) were amplified in the presence of SSAO and 18S RNA primers and then cycle threshold (CT) was plotted against the initial
amounts of reverse transcribed RNA, and the slope of the resulting
curve allowed calculation of PCR efficiency. For all experiments, PCR
efficiencies were close to one (0.956 ± 0.055, n = 19), indicating a doubling of DNA at each PCR cycle, as theoretically
expected. Thus, taking into account standardization with 18S RNA, it
was possible to compare the relative levels of SSAO transcripts between
the different experimental conditions. Quantification of SSAO mRNA was
carried out by comparison of the number of cycles required to reach
reference and target threshold values (delta-delta
CT method, see
http://www.bio.com/newsfeature, Introduction to real-time
PCR by M. Tevfik Dorak). Sequences of the sense and antisense
oligonucleotides were 5'-CCCCCTGCCCTATTACCG-3' and
5'-AAAACCCAGCCCCTGAGAGA-3' for SSAO; and 5'-GGGAGCCTGAGAAACGGC-3' and
5'-GGGTCGGGAGTGGGTAATTT-3' for 18S ribosomal RNA.
Determination of 2-Deoxyglucose Uptake.
For determination of
hexose transport, cells were seeded in 24-well plates, and were grown,
differentiated, and exposed to TNF- as mentioned above. When the
thiazolidinedione rosiglitazone was tested for its ability to reverse
TNF- effect on glucose transport, it was added into the culture
medium together with the cytokine. The uptake of glucose was determined
using [1,2-3H]deoxyglucose
([3H]DOG) (ICN Pharmaceuticals, Costa Mesa,
CA), a nonmetabolizable analog of glucose. Mature 3T3-L1 adipocytes
were rinsed three times with Krebs-Ringer buffer containing 12 mM
Hepes, pH 7.4 and then preincubated for 2 h at 37°C in 500 µl
of the Krebs-Ringer buffer containing 12 mM Hepes, pH 7.4. When
indicated, semicarbazide (1 mM) was added during this preincubation
period 90 min before measurement of hexose transport. Sodium
orthovanadate (100 µM), insulin (100 nM), or benzylamine (300 µM)
was introduced 60 min before [3H]DOG
incorporation test. Glucose uptake measurement was initiated by
addition of [3H]DOG (0.2 mM, 0.5 µCi/well).
After 5 min at 37°C, the experiment was terminated by rapidly
aspirating the radioactive incubation buffer and rinsing the cells
three times with ice-cold PBS. The cells were solubilized with 1% SDS,
and radioactivity was determined by scintillation counting. To
determine noncarrier-mediated glucose transport, 10 µM cytochalasin B
was used in parallel wells. Noncarrier-mediated glucose uptake
accounted for less than 3% of total glucose transport. Preliminary
experiments were performed to ensure that
[1,2-3H]deoxyglucose uptake was tested at the
initial rate of the transport. The results were expressed as picomoles
of glucose transported per minute per well.
Statistical Analyses.
Results are presented as means ± S.E. The statistical comparison of data between groups was assessed
with analysis of variance using the STATVIEW software.
| |
Results |
TNF-, through multiple effects on lipid and carbohydrate
metabolism, has been implicated in the pathogenesis of obesity-linked insulin resistance (Hotamisligil, 2000; Moller, 2000). Because recent
studies have shown that SSAO activation can promote hexose uptake and
contribute to glucose homeostasis (Enrique-Tarancon et al., 1998, 2000;
Marti et al., 1998, 2001), it was of interest to evaluate the possible
action of TNF- on adipocyte SSAO expression and its consequences on
SSAO-mediated glucose transport.
TNF- Down-Regulates SSAO Expression in Vitro and in Vivo.
SSAO regulation by TNF- was first investigated in vitro, using the
murine preadipose 3T3-L1 cell line as a model. Although SSAO
transcripts and enzyme activity are virtually absent in 3T3-L1 preadipocytes, there is a dramatic increase in SSAO mRNA levels and
enzyme activity when cells undergo adipose conversion (Moldes et al.,
1999).
In a first set of experiments, 3T3-L1 mature adipocytes were exposed to
various TNF- concentrations for 24 or 48 h. Results shown in
Fig. 1A indicate that SSAO mRNA levels
were dramatically decreased in the presence of increasing
concentrations of TNF-. Repression of SSAO transcript abundance was
statistically detectable at a cytokine concentration of 10 pM and was
maximal at 1 nM (8-fold reduction in SSAO mRNA levels compared with
control), giving a half-maximal effect between 100 and 300 pM. These
TNF--induced changes in SSAO mRNA levels were followed by
corresponding variations in SSAO protein (Fig. 1B). Western blot
analysis of protein extracts from control or cytokine-exposed 3T3-L1
adipocytes show a slight reduction in SSAO protein levels at 100 pM
TNF- (90 ± 4.5% of control), with a maximal effect at 1 nM of
the cytokine (37.5 ± 3.5% of control). In agreement with
variations in SSAO mRNA and protein levels, SSAO enzyme activity was
significantly decreased at 100 pM and was maximally inhibited at 1 nM
TNF- (25 ± 3.7% of control value), giving a half-maximal
effect at about 200 pM (Fig. 1C).
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 1.
Dose-dependent effect of TNF- on SSAO expression
and activity. 3T3-L1 mature cells were exposed to various TNF-
concentrations for 24 h (mRNA) or 48 h (Western blot and
enzyme activity analyses), in a DMEM/Ham's F-12 medium supplemented
with 0.1% bovine serum albumin. A, SSAO mRNA expression by real-time
RT-PCR was carried out as described under Materials and
Methods. Results were normalized with 18S RNA expression and
are expressed in percentage of control cell levels according to the
delta-delta CT comparative method. Results
represent the mean ± S.E. of four to five separate experiments
performed in duplicate. B, Western blot analysis of SSAO protein.
Samples (30 µg/lane) were subjected to electrophoresis, blotted onto
a polyvinylidene difluoride membrane, hybridized with an anti-SSAO
primary antibody (dilution 1:2000), and then with a horseradish
peroxidase-conjugated anti-rabbit-IgG (dilution 1:20,000). The
autoradiography is representative of four independent experiments. C,
SSAO enzyme activity toward benzylamine was tested on cell homogenates
(25 µg/assay) by a fluorometric detection of hydrogen peroxide
production. Results are expressed as percentage of control SSAO
activity and represent the mean ± S.E. of 4 to 10 independent
experiments performed in duplicate. SSAO enzyme activity of control
cells was 37.7 ± 4.4 nmol/h/mg. *, p < 0.05; **, p < 0.01; TNF--treated versus
control cells.
|
|
The time dependence of SSAO down-regulation by TNF- was also
studied. For that purpose, 3T3-L1 mature adipocytes were exposed for 1 to 72 h to a constant concentration of TNF- (0.5 nM). Note that
whatever the time exposure to TNF-, all cell extracts were prepared
at the same time point, implicating that the first cells treated by the
cytokine correspond to the 72-h condition. The same procedure was used
for all subsequent time-dependent experiments. SSAO mRNA expression was
significantly decreased after a 3-h exposure to TNF- (60% of
control value) (Fig. 2A), and the maximal
effect was observed after a 24-h treatment by the cytokine and
represented 25% of control levels. SSAO protein expression (Fig. 2B)
was moderately decreased (53% of control protein level) after a 12-h
exposure to TNF-. Maximal reduction in protein levels was obtained
after 48 h (39% of control value). A significant reduction in
SSAO enzyme activity was detectable from a 24-h exposure to TNF-
(63% of control value) and continued to decrease until 72 h (34%
of control value) (Fig. 2C). These initial dose- and time-dependent
studies show a good correlation between the TNF--induced
down-regulation of SSAO mRNA and protein levels and enzyme activity,
and suggest that the repression in SSAO gene expression by the cytokine
is responsible for the decrease in SSAO activity.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 2.
Time-dependent effect of TNF- on SSAO expression
and activity. 3T3-L1 mature cells were exposed to 0.5 nM TNF- for 1 to 72 h. A, SSAO mRNA expression was evaluated with a real-time
RT-PCR method. All results were normalized with 18S mRNA expression and
are given in percentage of control untreated cells. Results represent
the mean ± S.E. of three to six independent experiments performed
in duplicate. B, SSAO protein expression by Western blot (see Fig. 1).
The blot shown here is representative of four independent experiments.
C, SSAO enzyme activity was tested on cell homogenates using
benzylamine as a substrate and a fluorometric detection of hydrogen
peroxide production. The results are expressed as percentage of control
SSAO activity and represent the mean ± S.E. of five to nine
independent experiments performed in duplicate. Control SSAO activity
was 31.7 ± 1.9 nmol of H2O2/h/mg of
protein. *, p < 0.05; **,
p < 0.01; ***, p < 0.001; TNF--treated versus control cells.
|
|
To verify that TNF- effect on SSAO was primarily related to a
decrease in Vmax enzyme activity and
was not a consequence of a modification of SSAO affinity toward its
substrates or inhibitors, kinetic parameters of SSAO activity were
analyzed with different SSAO substrates (Table
1) and inhibitors (Table
2). 3T3-L1 mature adipocytes were
cultured in the absence or in the presence of 0.3 nM TNF- for
48 h. This submaximal concentration of TNF- was chosen because
it induced a clear but moderate decrease in SSAO maximal activity, thus
allowing an exact determination of Km
and Ki values for substrates and
inhibitors, respectively. As shown in Table 1,
Km values of SSAO for the four tested
substrates, i.e., benzylamine, methylamine, tyramine, and
-phenylethylamine, were not modified by TNF- treatment. On the
contrary, Vmax values obtained with each
SSAO substrate were reduced by 45 to 56% in TNF--treated cells
compared with Vmax values of control cells. Likewise, the Ki values of SSAO for
several well characterized inhibitors remained unaffected by TNF-
exposure (Table 2). Thus, TNF--induced alterations in SSAO enzyme
activity really correspond to a reduction in
Vmax without change in enzyme affinity for
its substrates or inhibitors.
View this table:
[in this window]
[in a new window]
|
TABLE 1
Kinetic parameters of SSAO activity in 3T3-L1 adipocytes cultured in
the absence or in the presence of TNF-
Cell homogenates were prepared from mature 3T3-L1 adipocytes exposed or
not to 0.3 nM TNF- for 48 h. Enzyme activity towards
benzylamine, methylamine, tyramine, and -phenylethylamine were
performed as described under Materials and Methods, with a
prior incubation with 1 mM of the monoamine oxidase-selective inhibitor
pargyline. Since in control or TNF--treated cells, less than 5% of
amine oxidase activity remained when cells extracts were preincubated
in the presence of 1 mM semicarbazide, it corresponded to the SSAO
activity (not shown). Km and Vmax
values were calculated by the nonlinear regression analysis
curve-fitting procedure of the ENZFITTER program. The percentage of
amine oxidase activity observed in TNF--treated compared with
control cells is also mentioned in the right column. Results represent
the mean ± S.E. of four to five separate experiments performed in
duplicate.
|
|
View this table:
[in this window]
[in a new window]
|
TABLE 2
Inhibition of SSAO activity by different compounds in control and
TNF--exposed cells
Cell extracts were prepared from mature 3T3-L1 adipocytes cultured in
the absence or presence of 0.3 nM TNF- for 48 h. 3T3-L1 cell
homogenates were preincubated for 30 min at 37°C in the presence of
various concentrations of each inhibitor. Pargyline at 1 mM was also
added in the assay to inhibit monoamine oxidase activity. Benzylamine
(30 µM) was then added as substrate for 1 h. IC50 values
were obtained by computer analysis with the ENZFITTER program. Apparent
Ki values of SSAO towards the different SSAO
inhibitors were then calculated by the equation of Cheng and Prusoff.
Results represent the mean ± S.E. of four separate experiments.
|
|
TNF- is known to inhibit adipose differentiation and to induce
adipocyte dedifferentiation (Torti et al., 1985). Because SSAO
represents a marker of terminal adipocyte differentiation, it was of
importance to verify that under our experimental conditions, the
TNF--induced decrease in SSAO expression was not related to a
general reduction in adipocyte maturation level. Thus, G3PDH activity
and triglyceride concentration, two classical biochemical markers of
adipose conversion (Pairault and Green, 1979; Gaillard et al., 1989),
were measured in parallel with SSAO activity in control and
TNF--treated cells. First, 3T3-L1 differentiated adipocytes were
exposed to various concentrations of TNF- for 48 h (Fig.
3A). G3PDH activity was only slightly
inhibited in the presence of the highest TNF- concentration tested.
TNF- at 0.3 nM induced a 6% decrease in G3PDH activity (Fig. 3A),
whereas the corresponding SSAO activity was reduced by 52% (Fig. 1C). The maximal inhibition (27%) in G3PDH activity was observed at 1 nM
TNF-, a concentration at which SSAO activity was decreased by 75%.
Even at the highest TNF- concentrations, cell triglyceride content
remained unchanged (Fig. 3A). Second, 3T3-L1 mature cells were exposed
to 0.5 nM TNF- for various periods of time (6-72 h) (Fig. 3B).
G3PDH activity was weakly decreased (20-34% inhibition compared with
control) from 24 to 72 h of TNF- exposure, whereas triglyceride
content of cytokine-treated cells was not significantly modified.
However, SSAO activity was markedly reduced (37-66% inhibition
compared with control) during the same interval (Fig. 2C). Thus, there
was a striking discrepancy between the very poor TNF--induced
inhibition in adipocyte maturation level and the strong inhibition of
SSAO expression and enzyme activity induced by the cytokine. This
observation underlines that the down-regulation of SSAO by TNF-
really corresponds to a specific gene and protein modulation and is not
related to a general dedifferentiating effect.
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 3.
Effect of TNF- on cell triglyceride content and
glycerol-3-phosphate dehydrogenase activity. Two markers of adipose
differentiation, G3PDH activity (open diamonds) and triglyceride
concentration (closed squares) were evaluated on mature 3T3-L1 cells
treated or not with TNF-. G3PDH activity was tested on the
10,000g supernatants whereas triglyceride content was
measured on cell homogenates. A, cells were exposed to various
concentrations of TNF- for 48 h. Results are given as
percentage of control value and represent the mean ± S.E. of six
to eight independent experiments performed in duplicate. Control G3PDH
activity was 727.9 ± 56.6 nmol of NADH/min/mg of protein, and
control triglyceride concentration was 2.19 ± 0.11 mM. B, cells
were exposed for various periods of time to 0.5 nM TNF-. Results are
expressed as percentage of control and represent the mean ± S.E.
of 12 to 18 experiments performed in duplicate. Control G3PDH activity
was 1104.6 ± 54.1 nmol of NADH/min/mg of protein and control
triglyceride concentration was 2.77 ± 0.32 mM. *,
p < 0.05; **, p < 0.01;
***, p < 0.001; TNF--treated versus
control cells.
|
|
To evaluate the capacity of mature adipocytes to spontaneously reverse
the TNF--induced inhibition in SSAO activity, 3T3-L1 adipocytes were
initially cultured for 48 h in the absence or in the presence of 1 nM TNF-, a concentration that potently down-regulated SSAO
expression. Thereafter, cells were rinsed extensively, and TNF- was
omitted in some dishes previously cultured in the presence of the
cytokine. SSAO activity was then tested on cell homogenates at
different intervals (12-72 h) after TNF- removal (Fig.
4). As expected, a 48-h exposure to 1 nM
TNF- caused a dramatic 70% decrease in SSAO activity (time 0 after
TNF- removal). When TNF- was maintained in the culture medium,
SSAO activity continued to be down-regulated, reaching 6% of control
SSAO activity after an additional 72-h exposure to the cytokine. When
TNF- was removed after an initial 2-day treatment, there was an
initial decrease in SSAO activity that likely reflected the slow
turnover of SSAO protein and/or a persistence in TNF-
biological effect. Thereafter, from 48 h after TNF- removal,
there was a marked increase in SSAO activity that reached 60% of the
control value at 72 h.
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 4.
Spontaneous reversibility of TNF- effect on SSAO
activity. In a first step, mature 3T3-L1 cells were exposed (closed
squares and triangles) or not (open diamonds) for 48 h to 1 nM
TNF-. Thereafter, the cells were washed and refed, and the procedure
was pursued in control cells (open diamonds) and in some of
TNF--treated cells (closed squares), whereas a part of the dishes
initially exposed to the cytokine were shifted into a TNF--free
medium (closed triangles). Cells homogenates were prepared at various
intervals (0-72 h) after TNF- removal, and SSAO enzyme activity was
measured with a fluorometric detection of hydrogen peroxide production.
Results represent the mean ± S.E. of two independent experiments
and are expressed as percentage of control value at each time point.
|
|
To test whether TNF- effect on amine oxidases was restricted to
SSAO, the activity of two amine oxidases present in fat cells, monoamine oxidase and polyamine oxidase, was also tested in control and
TNF--treated 3T3-L1 adipocytes. SSAO activity was measured in the
presence of 1 mM pargyline to block any residual monoamine oxidase
activity. Under these conditions, benzylamine, methylamine, tyramine,
and -phenylethylamine oxidase activities were totally inhibited by a
prior incubation with 1 mM semicarbazide, indicating that they
corresponded to a SSAO activity (data not shown). In contrast,
monoamine oxidase activity was measured using serotonin or tyramine as
substrates and in the presence of 1 mM semicarbazide to block SSAO
activity. Under these conditions, serotonin or tyramine oxidase
activities were totally inhibited by a prior incubation with 1 mM
pargyline, indicating that they corresponded to a monoamine oxidase
activity (data not shown). After a 48-h exposure, to 0.3 nM TNF-,
there was a clear and significant decrease in SSAO activity, regardless
of the SSAO substrate used in the assay (Table
3). In contrast, in TNF--treated
cells, monoamine oxidase activity tested in the presence of serotonin
or tyramine was increased by 42 and 22%, respectively. Polyamine
oxidase activity was assayed using
N1-acetyl spermidine as a substrate
and was similar between control and cytokine-exposed 3T3-L1 adipocytes.
View this table:
[in this window]
[in a new window]
|
TABLE 3
Effects of TNF- on several amine oxidase activities present in
mature 3T3-L1 adipocytes
Cell homogenates were prepared from 3T3-L1 adipocytes cultured in the
absence or presence of 0.3 nM TNF- for 48 h. SSAO, monoamine
oxidase (MAO) and polyamine oxidase (PAO) activities were measured as
described under Materials and Methods. SSAO activity was
tested in the presence of 1 mM pargyline to inhibit monoamine oxidase
activity. Monoamine oxidase activity was tested in the presence of 1 mM
semicarbazide to inhibit SSAO activity. Results are expressed in
nanomoles of H2O2 per minute per milligram of protein
and represent the mean ± S.E. of four to six experiments.
|
|
We have previously reported that effectors of the cAMP signaling
pathway, ()-isoproterenol and forskolin, are able to decrease SSAO
mRNA levels and enzyme activity (Moldes et al., 1999). Thus, we
examined whether a chronic exposure to TNF- could modify cAMP production and participate to the cytokine-induced down-regulation in
SSAO activity. Adenylyl cyclase activity was measured in membranes from
control cells and from 3T3-L1 adipocytes treated with 0.5 nM TNF-
for 48 h. As observed in 3T3-F442A adipocytes (El Hadri et al.,
1997), cell exposure to TNF- led to a moderate but nonsignificant 1.5-fold increase in basal adenylyl cyclase activity (0.58 ± 0.04 and 0.88 ± 0.18 pmol of cAMP/min/mg protein in control and
TNF--treated cells, respectively; n = 8;
p = 0.095). As controls and in agreement with TNF-
effect on 3T3-F442A fat cells (El Hadri et al., 1997), we observed that
the cytokine decreased ()-isoproterenol-stimulated adenylyl cyclase
activity and did not modify forskolin-stimulated adenylyl cyclase
activity (data not shown). However, this slight TNF--induced
increase in basal cAMP production did not seem involved in the
down-regulation of SSAO activity by the cytokine. Indeed, the
cAMP-dependent protein kinase selective inhibitor H89 did not prevent
the reduction in SSAO activity caused by the cytokine (data not shown).
Finally, to more completely characterize TNF- effect on SSAO
expression, we investigated the consequences of an in vivo
administration of the cytokine on the expression and activity of the
amine oxidase. Four-week-old male C57BL/6 mice received a single
intraperitoneal injection of 1 µg of TNF-, and animals were
sacrificed 48 h after injection. SSAO mRNA levels were examined in
epididymal white adipose tissue by real time RT-PCR analysis (Fig.
5A). There was a marked 51% decrease of
SSAO mRNA levels in TNF--injected mice compared with control
animals. Likewise, SSAO activity was tested in epididymal fat pad
extracts using benzylamine or methylamine as substrates. As shown in
Fig. 5B, SSAO activity of TNF--injected mice was impaired compared
with that of control animals, with a 33 and 34% reduction in
benzylamine and methylamine oxidase activities, respectively. Thus,
TNF- is able to suppress SSAO expression both in vitro and in vivo.
View larger version (11K):
[in this window]
[in a new window]
|
Fig. 5.
TNF- regulates SSAO expression and activity in
vivo. Four-week-old male C57BL/6 mice were treated (filled columns) or
not (open columns) with a single intraperitoneal injection of 1 µg of
TNF-. Forty-eight hours after the injection, mice were sacrificed
and epididymal white adipose tissue was removed to measure SSAO mRNA
expression (A) and enzyme activity (B). A, total RNA was extracted,
digested with DNase I, and gene expression was tested by real-time
RT-PCR. Reverse transcribed mRNAs were amplified with an ICycler
thermal cycler using the SYBR green fluorescence method (see
Materials and Methods). SSAO mRNA expression was
estimated according to the delta-delta CT
method, and normalized to 18S RNA expression. Results are expressed as
percentage of control SSAO mRNA levels and represent the mean ± S.E. of 11 animals in the control group and of nine TNF--injected
mice. B, SSAO enzyme activity was measured on tissue homogenates with a
fluorometric detection of hydrogen peroxide production, using
benzylamine (BZM, 300 µM) or methylamine (3 mM) as substrates. In
control and TNF--treated mice, benzylamine and methylamine oxidase
activities were completely abolished by a prior treatment with 1 mM
semicarbazide, indicating that they were related to the SSAO activity
(data not shown). Enzyme activities are expressed in nanomoles of
H2O2 per hour per milligram of protein and
represent the mean ± S.E. of 9 to 11 samples performed in
duplicate. *, p < 0.05; **,
p < 0.01; ***, p < 0.001; TNF--treated versus control animals.
|
|
Functional Consequences on Glucose Uptake of TNF--Induced SSAO
Down-Regulation.
Recent studies have reported that SSAO activation
in response to its substrates can stimulate glucose transport in
adipocytes (Enrique-Tarancon et al., 1998, 2000; Marti et al., 1998;
Fontana et al., 2001; Morin et al., 2001). Otherwise, TNF- is well
known to inhibit insulin-stimulated glucose transport (Stephens and Pekala, 1991) and to be involved in the pathogenesis of
obesity-associated insulin resistance (Hotamisligil, 2000; Moller,
2000). It was thus consistent to examine whether TNF--induced
down-regulation of SSAO could also be implicated in the modulation of
amine-stimulated glucose transport.
In this aim, preliminary experiments were performed on mature 3T3-L1
adipocytes treated or not with 0.5 nM TNF- for 48 h. [3H]DOG incorporation was then estimated after
a 1-h incubation with 100 nM insulin or 300 µM benzylamine, in the
absence or in the presence of 100 µM sodium orthovanadate for the
same period. In rat or 3T3-L1 adipocytes, low concentrations of sodium
orthovanadate have been shown to be required to activate glucose
transport in response to SSAO substrates (Enrique-Tarancon et al.,
1998; Marti et al., 1998). As shown in Fig.
6A, benzylamine alone did not modify
[3H]DOG uptake. When 3T3-L1 cells were
preincubated in the presence of sodium orthovanadate, at a
concentration (100 µM) that had no significant effect alone, there
was a clear potentiation of benzylamine action on glucose transport.
Under these conditions, benzylamine caused a 2.35-fold increase in
basal [3H]DOG uptake. In contrast,
insulin-stimulated glucose transport was not influenced in the presence
of sodium orthovanadate. As previously documented in several studies
(Stephens and Pekala, 1991; Szalkowski et al., 1995), TNF-
exposure led to a decrease in insulin-stimulated glucose transport.
When cells were treated with 0.5 nM TNF- for 48 h, the
insulin-stimulated [3H]DOG uptake over basal
transport was reduced by 40%. The presence of sodium orthovanadate did
not influence TNF- effect on insulin-activated hexose uptake. Under
the same conditions, amine-stimulated [3H]DOG
uptake was more strongly inhibited by TNF- exposure, because in
response to benzylamine hexose transport over basal was reduced by 66%
in comparison with control cells. This observation suggests that
down-regulation in SSAO expression could play a role in the dramatic
TNF--induced decrease in amine-stimulated glucose transport.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 6.
Characterization of TNF- effect on SSAO-mediated
glucose transport. 3T3-L1 cells were grown in 24-well plates until day
8 after confluence. Thereafter, cells were maintained in DMEM/Ham's
F-12 supplemented with 0.1% bovine serum albumin and treated (filled
columns) or not (open columns) with 0.5 nM TNF- for 48 h. Cells
were then tested for [3H]DOG) (final concentration 0.2 mM, 0.5 µCi/well) incorporation. A, when indicated, 100 µM sodium
orthovanadate (NaV), 100 nM insulin, and 300 µM benzylamine (BZM)
were added 60 min before [3H]DOG uptake measurement.
Results are expressed as percentage of control basal uptake (without
any treatment) and represent the mean ± S.E. of six independent
experiments performed in triplicate. Basal glucose uptake was 86.2 ± 9.0 pmol of [3H]DOG/min/well and is represented with a
dotted line. B, when mentioned, 1 mM semicarbazide (SCZ) was
preincubated with 3T3-L1 cells 90 min before [3H]DOG
incorporation measurement. Sixty minutes before glucose transport test,
100 µM NaV was added in all wells, whereas 300 µM BZM was added in
half wells. Control basal uptake was 74.7 ± 5.8 pmol/min/well and
is represented with a dotted line. Results represent the mean ± S.E. of four independent experiments performed in triplicate. *,
p < 0.05; **, p < 0.01;
BZM- or insulin-stimulated glucose transport versus control basal
uptake.  , p < 0.05; , p < 0.01; TNF--treated versus control cells.
|
|
In subsequent experiments, because sodium orthovanadate was required to
observe benzylamine effect on [3H]DOG uptake,
it was added in all assays, including those testing basal glucose
transport. To ensure that the induction of
[3H]DOG uptake by benzylamine was the
consequence of SSAO activation, semicarbazide was added 30 min before
benzylamine addition (Fig. 6B). Whereas semicarbazide alone had no
effect on glucose transport, it completely abolished
benzylamine-stimulated [3H]DOG uptake in
control cells.
To further characterize the functional consequences of TNF--induced
repression of SSAO expression on glucose transport, 3T3-L1 mature
adipocytes were exposed for 48 h to various concentrations of
TNF- before [3H]DOG uptake measurements
(Fig. 7). In control cells, benzylamine induced a 2.3-fold increase in basal glucose transport.
Benzylamine-stimulated glucose transport was significantly reduced at
0.1 nM TNF- and was totally abolished between 0.5 and 1 nM of the
cytokine, the half-maximal effect being between 0.25 and 0.3 nM. It was
noticeable that in slight contradiction with a previous study
(Cornelius et al., 1990), TNF- had only a weak inducing effect
(
20%) on basal glucose transport. In fact, we observed a more
pronounced induction in basal glucose transport after a shorter (12-h)
exposure to the cytokine (data not shown). Interestingly, it has been
previously reported that basal glucose transport was only slightly
increased in 3T3-L1 adipocytes cultured in the absence of insulin
(Ranganathan and Davidson, 1996).
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 7.
Dose-dependent effects of TNF- on
benzylamine-stimulated glucose transport. Mature 3T3-L1 cells cultured
in 24-well plates were incubated for 48 h with various (0-1 nM)
TNF- concentrations. One hour before [3H]DOG uptake
measurement, cells were all preincubated with 100 µM NaV, and in the
absence (basal transport, open diamonds) or in the presence of 300 µM
benzylamine (closed squares). Results are given as percentage of
control basal uptake and represent the mean ± S.E. of 10 to 12 separate experiments performed in triplicate. Basal glucose transport
was 89.9 ± 8.2 pmol/min/well. *, p < 0.05;
**, p < 0.01; benzylamine-stimulated glucose
transport of TNF--treated versus that of control cells. ,
p < 0.05; , p < 0.01;
basal glucose transport of TNF--treated versus that of control
cells.
|
|
We also examined the time dependence of the TNF--induced reduction
in SSAO-mediated glucose transport. 3T3-L1 mature adipocytes were
exposed for 6 to 72 h to a constant concentration of TNF- (0.5 nM) before testing hexose uptake. In agreement with SSAO activity
modulation, the reduction in benzylamine-stimulated
[3H]DOG uptake was detectable from 24 h
after TNF- exposure and persisted to decrease after 72 h (Fig.
8).
View larger version (11K):
[in this window]
[in a new window]
|
Fig. 8.
Time-dependent effect of TNF- on
benzylamine-stimulated glucose transport. Mature 3T3-L1 cells were
cultured in the absence (open triangles) or in the presence (filled
squares) of 0.5 nM TNF- for 6 to 72 h. Then, 1 h before
glucose transport measurement, cells were preincubated with 100 µM
NaV and in the absence (basal transport) or in the presence of 300 µM
benzylamine. Results are given as percentage of control basal uptake
and represent the mean ± S.E. of 10 to 12 separate experiments
performed in triplicate. The control basal uptake value was 53.4 ± 2.2 pmol/min/well. , p < 0.05; ,
p < 0.01; benzylamine-stimulated glucose uptake of
TNF--treated cells versus that of control cells.
|
|
It has been clearly demonstrated that the inhibiting effects
of TNF- on insulin-stimulated glucose transport can be
counteracted in the presence of thiazolidinediones (Szalkowski et
al., 1995; Peraldi et al., 1997). Thus, we investigated whether the
TNF--induced decrease in amine-stimulated glucose uptake was
reversible in the presence of the insulin-sensitizing compounds
thiazolidinediones. 3T3-L1 adipocytes were cultured for 48 h with
or without 0.5 nM TNF-, and in the absence or in the presence of 1 µM rosiglitazone, a potent peroxisome proliferator-activated receptor
agonist. Then, basal, insulin- and benzylamine-stimulated glucose
uptakes were tested. As expected, almost all of the inhibitory effect of TNF- on insulin-stimulated glucose transport was antagonized by
the concomitant addition of rosiglitazone; 77 ± 6% of the
suppressive action of TNF- on insulin-induced
[3H]DOG uptake was reversed in the presence of
the thiazolidinedione (data not shown). In contrast, rosiglitazone did
not modify the preventing effect of the cytokine on
benzylamine-stimulated glucose uptake (Fig.
9A). Rosiglitazone alone did not alter
basal or benzylamine-mediated glucose transport. In agreement with the
results on glucose uptake, rosiglitazone did not reverse the inhibitory
effect of TNF- on SSAO enzyme activity (Fig. 9B). Rosiglitazone
alone did not change SSAO activity.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 9.
Thiazolidinediones do not reverse TNF- effects on
benzylamine-stimulated glucose transport and SSAO activity. A, mature
3T3-L1 adipocytes grown in 24-well plates were maintained for 48 h
in the absence (open columns) or in the presence (filled columns) of
0.5 nM TNF-, and in the absence () or in the presence (+) of
rosiglitazone (1 µM). Cells were preincubated with 100 µM NaV, and
in the absence (basal transport) or in the presence of 300 µM
benzylamine (BZM) 1 h before [3H]DOG incorporation
measurement. Results are given as percentage of control basal uptake
and represent the mean ± S.E. of six separate experiments carried
out in triplicate. Basal glucose uptake was 71.2 ± 10.0 pmol/min/well. B, mature 3T3-L1 cells were treated (filled columns) or
not (open columns) with 0.5 nM TNF- exposed or not to 1 µM
rosiglitazone for 48 h. Cell homogenates were used to evaluate
SSAO enzyme activity with a fluorometric detection of hydrogen peroxide
production. Results are expressed as percentage of control SSAO
activity (cells without any treatment) and represent the mean ± S.E. of five independent experiments performed in duplicate. Control
SSAO activity was 24.84 ± 0.96 nmol of
H2O2/h/mg of protein. *,
p < 0.05; **, p < 0.01;
benzylamine-stimulated glucose transport versus basal transport. ,
p < 0.05; , p < 0.01;
, p < 0.001; TNF--treated versus control
cells.
|
|
| |
Discussion |
In the present work, we observed that in adipocytes, a chronic
exposure to TNF- led to a potent down-regulation in SSAO enzyme activity. Several data strongly suggest that the inhibitory effect on
enzyme activity is likely related to a decrease in gene and protein
expression. First, the half-maximal efficient TNF- concentration and
the magnitude of maximal repression were similar for mRNA or protein
expression and for enzyme activity. Otherwise, the down-regulation in
SSAO activity occurred in the same period than the decrease in protein
expression and was preceded by the reduction in SSAO mRNA levels.
Finally, our in vivo studies indicate a good correlation between the
reduction in SSAO mRNA steady-state levels and that of enzyme activity.
Identification of the TNF- receptor subtypes and intracellular
signaling pathways as well as the transcriptional or
post-transcriptional mechanisms at the basis of the regulation of SSAO
gene expression by the cytokine will require further investigation.
Interestingly, SSAO protein seems to display a slow turnover. There was
obviously a time lag between the cytokine-induced reduction in SSAO
transcript abundance and the down-regulation in protein expression and
enzyme activity. Furthermore, after TNF- removal from the culture
medium, we observed an additional 24-h interval before the very
progressive restoration in enzyme activity. Although we cannot exclude
that our culture conditions in a defined serum-free medium modify
protein half-life or that the biological effect of TNF- on SSAO
expression is particularly persistent, it is likely that SSAO
displays a physiological slow turnover. Consequently, variations in
SSAO expression could only represent a long-term adaptive mechanism for
regulating the activity of this amine oxidase.
It is noteworthy that SSAO modulation by TNF- is not restricted to
this type of amine oxidase. In contrast to SSAO, monoamine oxidase
activity is moderately induced in response to the cytokine. Monoamine
oxidase activity is largely expressed in adipocytes (Pizzinat et al.,
1999) and is induced during adipose conversion (Fontana et al., 2001).
Studies using monoamine oxidase A- and B-selective inhibitors have
shown that most of this activity was attributable to monoamine oxidase
A expression (Pizzinat et al., 1999). In our study, we did not
delineate whether TNF--induced monoamine oxidase activity was
related to an increase in monoamine oxidase A, or monoamine oxidase B,
or a combination of both. Regardless of the mechanism at the basis of
the up-regulation of monoamine oxidase activity by TNF-, it would be
of interest to investigate whether regulation of SSAO and monoamine
oxidase correspond to independent biological events, or whether an
intracellular cross talk exists between the two types of amine oxidase
and regulates the redox equilibrium of adipocytes in a coordinated manner.
A fundamental question was to determine whether the TNF--induced
down-regulation of SSAO expression could significantly contribute to
the reduction in benzylamine-stimulated glucose transport observed after cytokine exposure. Enrique-Tarancon et al. (2000) have recently shown that SSAO-mediated hexose uptake shares several signaling pathways with the insulin-sensitive glucose transport, such as activation of IRS proteins and phosphatidylinositol 3-kinase and translocation of the glucose transporter GLUT4 to the plasma membrane. Otherwise, numerous studies have reported that TNF- can alter the
expression and/or function of several key steps in the
insulin-sensitive glucose pathway, including alterations in
phosphorylation of the insulin receptor or IRS-1, or decreased GLUT4
expression (Hotamisligil, 2000; Moller, 2000). Thus, the targeting by
TNF- of several signaling molecules common between insulin- and
SSAO-mediated effects may be sufficient to explain all the inhibitory
effect of the cytokine on benzylamine-stimulated glucose transport.
However, several lines of evidence support the view that the
down-regulation of SSAO expression caused by TNF- has a key role in
the decreased hexose transport in response to SSAO substrates. First,
after TNF- exposure, the decrease of benzylamine-stimulated glucose transport is much more pronounced than that of insulin-sensitive glucose uptake, thus indicating that different and/or additional mechanisms contributed to the alterations in amine-activated glucose uptake. Second, the kinetics and dose-response curve of TNF--induced down-regulation in benzylamine-stimulated glucose uptake are consistent with those of the TNF--provoked reduction in SSAO protein expression and enzyme activity. Both amine-stimulated glucose transport and enzyme
activity were decreased from 24 h after cytokine addition and
continued to lower at 72 h. Moreover, after a 48-h treatment by
TNF-, both the inhibitory effects on glucose transport and on SSAO
activity were detectable from 100 pM of the cytokine, with a
half-maximal concentration being in the 200 to 300 pM range. 3)
Finally, experiments performed with the thiazolidinedione rosiglitazone are also very informative. The insulin-sensitizing compounds
thiazolidinediones, through peroxisome proliferator-activated receptor
activation, are known to counteract TNF- effects on
insulin-sensitive glucose transport, adipocyte differentiation, and
protein synthesis (Szalkowski et al., 1995; Peraldi et al., 1997).
However, in our study, rosiglitazone completely failed to antagonize
TNF- effects on benzylamine-stimulated glucose transport. This was
paralleled by the absence of rosiglitazone action to reverse
TNF--induced suppression of SSAO enzyme activity. This indicates
that the down-regulation of SSAO expression by TNF- does not involve
a PPAR-dependent mechanism. Moreover, this strongly suggests that
the main mechanism at the basis of TNF--induced decrease in
benzylamine-stimulated glucose transport is the inhibition of SSAO gene
and protein expression.
Recently, considerable attention has been focused on the possible
involvement of the cytokine in the pathogenesis of obesity-associated insulin resistance (Hotamisligil, 2000; Moller, 2000). TNF-
overproduction has been observed in adipose tissue of several rodent
models of obesity, as well as in obese humans (Hotamisligil, 2000;
Moller, 2000). Exogenous administration of TNF- to animals can
induce insulin resistance, whereas neutralization of endogenous TNF- can restore insulin sensitivity in genetically obese rats displaying high levels of TNF- expression in adipose tissue (Hotamisligil, 2000; Moller, 2000). Moreover, genetic ablation of TNF- or its receptors was reported to improve insulin sensitivity in various models
of insulin resistance, including genetically, chemically, or
diet-induced obese mice (Hotamisligil, 2000; Moller, 2000). The exact
involvement of TNF- in the pathophysiology of insulin resistance in
obese humans and in noninsulin-dependent diabetes mellitus is still a
matter of controversy, but it is reasonable to believe that TNF-, in
addition to other cytokines and metabolic factors, could contribute to
the balance of tissular insulin sensitivity. Several recent works have
reported that SSAO substrates stimulate glucose transport in rodent
(Enrique-Tarancon et al., 1998, 2000; Marti et al., 1998; Fontana et
al., 2001) or human adipocytes (Morin et al., 2001). This
amine-stimulated glucose transport may thus represent an original
mechanism by which circulating or locally produced amines can regulate
glucose homeostasis. Acute administration of the SSAO substrate
benzylamine can thus reduce plasma glucose levels independently to
variations in plasma insulin levels (Enrique-Tarancon et al., 2000;
Marti et al., 2001). However, it must be kept in mind that modulations
of glucose uptake or of plasma glucose levels observed in the presence
of SSAO substrates correspond to pharmacological manipulations. So far,
there is no definitive evidence that establishes a physiological and
fundamental role of SSAO in glucose homeostasis or that demonstrates a
pathophysiological involvement of SSAO in disorders of carbohydrate metabolism.
In agreement with our in vitro studies, TNF--injected mice displayed
a clear reduction in SSAO mRNA levels and enzyme activity. Low levels
of SSAO transcripts and activity have also been reported in white
adipose tissue of the Zucker fa/fa rat (Moldes et al., 1999), a genetic rodent model of obesity that overexpresses TNF- in
fat cells (Hotamisligil, 2000). Thus, the overproduction of TNF-
observed in adipose tissue of rodent models of obesity or in human
obesity could reduce SSAO expression, and in turn may contribute to
disturbances in glucose homeostasis. Studies on the effects of SSAO
substrates on carbohydrate metabolism in animals overexpressing TNF-
in adipose tissue or in obese mice lacking TNF- or its receptors
would be helpful to ascertain the physiological relevance of our observations.
In the present work, we have only investigated the consequences of
TNF--induced SSAO down-regulation on amine-stimulated glucose
transport. Hexose transport activation by SSAO substrates can be
totally prevented in the presence of catalase and antioxidants (Enrique-Tarancon et al., 1998; Marti et al., 1998; Morin et al., 2001), indicating that hydrogen peroxide generated by SSAO was involved
in the stimulation of glucose transport. However, in fat cells, besides
its action on glucose transport, exogenously provided hydrogen peroxide
has been shown to promote glucose oxidation or incorporation into
glycogen, to stimulate lipogenesis or to inhibit lipolysis (Moldes et
al., 1999 and references therein). Thus, through hydrogen peroxide
production, SSAO may mediate pleiotropic functions in adipocytes. For
instance, the antilipolytic effect of SSAO has recently been reported
in human fat cells (Morin et al., 2001). As a consequence, the
reduction in SSAO expression and activity caused by TNF- may
decrease the antilipolytic properties of the amine oxidase. The
resulting increase in free fatty acids may represent an additional
mechanism to alter insulin sensitivity (Arner, 2001).
We thank Dr. M. Unzeta for the generous gift of the polyclonal
antibody against SSAO.
This work was supported by grants from the Centre National de
la Recherche Scientifique and of the Université Paris VI.
N.M. is the recipient of a grant of the Ministère de
l'Enseignement Supérieur et de la Recherche.
SSAO, semicarbazide-sensitive amine oxidase;
IRS, insulin receptor substrate;
TNF-, tumor necrosis factor-;
DMEM, Dulbecco's modified Eagle's medium;
PBS, phosphate buffer
saline;
G3PDH, glycerol-3-phosphate dehydrogenase;
TBS, Tris-buffered
saline;
RT-PCR, reverse transcription-polymerase chain reaction;
PCR, polymerase chain reaction;
DOG, deoxyglucose.
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
in Adipocytes: Functional Consequences on
Glucose Transport
Top
Abstract
Introduction
