|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
up-regulates apelin expression in human and mouse adipose tissue



* INSERM, U586, Unité de recherches sur les obésités, Université Paul Sabatier, Institut Louis Bugnard IFR31, Toulouse, France; and
Service de diabétologie, Maladies métaboliques et nutrition, CHU de Nancy, Hôpital J. dArc, Nancy, France
1Correspondence: FIR 31, Institut Louis Bugnard, BP 84 225 INSERM U586, Toulouse F-31432, Cedex 4, France. Email: isabelle.castan{at}toulouse.inserm.fr
ABSTRACT
We have recently identified apelin as a novel adipokine up-regulated by insulin and obesity. Since obesity and insulin resistance are associated with chronically elevated levels of both insulin and TNF
, the present study was performed to investigate a putative regulation of apelin expression in adipocytes by TNF
. Herein, we report a tight correlation between apelin and TNF
expression in adipose tissue of lean and obese humans. Apelin regulation by TNF
was further studied in cultured explants of human adipose tissue. The endogenous expression of TNF
in adipocytes isolated from the explants was accompanied by a 69 h subsequent increase of apelin expression in adipocytes. This increase was reversed by inhibiting TNF
expression with 100 µM isobutylmethylxanthine. In different mouse models of obesity, expression of both TNF
and apelin was also significantly increased in adipocytes of obese mice. Furthermore, short-term exposure to an i.p. injection of TNF
in C57Bl6/J mice induced an increase of apelin expression in adipose tissue as well as apelin plasma levels. Finally, a direct positive effect of TNF
has been shown in differentiated 3T3F442A adipocytes on apelin expression and secretion. The signaling pathways of TNF
for the induction of apelin were dependent of PI3-kinase, c-Jun NH2-terminal kinase (JNK), and MAPK but not PKC activation. All together, these findings suggest that apelin might be a candidate to better understand potential links between obesity and associated disorders such as inflammation and insulin resistance.Daviaud, D., Boucher, J., Gesta, S., Dray, C., Guigne, C., Quilliot, D., Ayav, A., Ziegler, O., Carpene, C., Saulnier-Blache, J.-S., Valet, P., Castan-Laurell, I. TNF
up-regulates apelin expression in human and mouse adipose tissue.
Key Words: obesity adipocyte adipokine
APELIN IS A BIOACTIVE peptide originally identified from bovine stomach extracts as the endogenous ligand of the orphan G protein-coupled receptor, APJ (1)
. Apelin is derived from a prepropeptide of 77-amino acid that is cleaved into a 55-amino acid fragment and then into shorter forms. The physiologically active form is thought to be apelin-36, the most biologically effective being the pyroglutamylated form of apelin-13 (1)
. In rodents, mRNA expression of both apelin and APJ has been described in the central nervous system and peripheral tissues (for review see 2
, 3
). The expression of apelin in the pituitary gland (4
, 5)
, as well as the presence of apelin in the bloodstream (6
, 7)
is consistent with an endocrine function of this peptide. The origin of blood apelin and the regulations of its input to the bloodstream are of interest. Although little is known about the physiological roles of apelin, several studies have shown that apelin is involved in the regulation of cardiovascular function (8
9
10)
and fluid homeostasis (11
, 12)
. Very recently, it was shown that apelin inhibited insulin secretion in mice (13)
.
So far, few data are available concerning apelin production and regulation in apelin-expressing cells. Apelin has recently been described by our group as a novel adipokine, produced and secreted by human and mouse isolated mature adipocytes. Apelin mRNA levels found in isolated adipocytes were close to those of the so-called stroma vascular fraction (containing other cell types present in adipose tissue) or organs known to express apelin such as kidney and heart (14)
. We showed that a strong relationship exists between adipocyte-secreted apelin and insulin. Indeed, a direct action of insulin in the regulation of apelin expression in adipocytes was demonstrated in vivo and in vitro in 3T3F442A differentiated adipocytes. In addition, adipocyte apelin mRNA levels as well as plasma apelin concentration were increased in various mouse models of obesity associated with hyperinsulinemia (14)
. Since obesity and insulin resistance are associated with chronically elevated levels of both insulin and TNF
, the present study was performed to investigate a putative regulation of apelin expression in human and mouse adipocytes by TNF
. A tight correlation has been observed between apelin and TNF
mRNA levels in different anatomical location of adipose tissues from lean to severe obese subjects. To determine whether TNF
exerts a direct effect on apelin expression in adipocytes, further experiments were performed in C57Bl6/J mice and 3T3F442A mouse adipocyte cell line. Even though TNF
has been shown to inhibit insulin receptor signaling and insulin metabolic functions (15)
, the present work clearly describes a TNF
-dependent up-regulation of apelin expression in human and mouse adipose tissue.
MATERIALS AND METHODS
Human samples
Human adipose tissue was collected according to the guidelines of the Ethical Committee of Toulouse-Rangueil and Nancy J. dArc Hospitals. All subjects gave their informed consent to participate to the study, and investigations were performed in accordance with the declaration of Helsinki as revised in 2000 (http://www.wma.net/e/policy/b3.htm). Human subcutaneous (s.c.) adipose tissue samples were obtained from 20 patients [40.3±2.2 yr old, Body Mass Index (BMI): 27.3±1.2 kg/m2] undergoing abdominal lipectomy for plastic surgery. No clinical data from these patients were available. Human intra-abdominal adipose tissue samples were obtained from 23 morbidly (grade III) obese subjects (41.3±2.2 yr old, BMI: 45.7±2.9 kg/m2) before a bariatric surgery. Tissue samples were immediately frozen in liquid nitrogen and stored at 80°C.
Animals
Mice were handled in accordance with the principles and guidelines established by the National Institute of Medical Research (INSERM). FVB/n and C57Bl6/J female mice were obtained from Charles River laboratory (lArbresle, France). Mice were housed conventionally in a constant temperature (2022°C) and humidity (5060%) animal room and with a 12 h lightdark cycle (lights on at 8:00 am). All mice had free access to food and water throughout the experiment. For the in vivo effect of TNF
, 15-wk-old fasted female C57Bl6/J mice received either an i.p. injection of 2 µg/mouse TNF
(SIGMA) or PBS (in a same vol) at 8:00 a.m. They were sacrificed 8 h later; blood and different organs (heart, kidney and adipose tissue) were taken.
Diets
C57Bl6/J mice were assigned to chow, low-fat diet (LFD) or high-fat diet (HFD (SAFE, France). Energy contents of the specific diets were (% kcals): 20% protein, 70% carbohydrate, and 10% fat for LFD; 20% protein, 35% carbohydrate, and 45% fat for HFD. The main source of fat in HFD was lard (20 g/100g of food). C57Bl6/J (5 wk old) were fed a HFD or LFD for 8 wk. In 8-week-old FVB/n mice, obesity was induced by a single i.p. injection (0.5 g/kg) of gold thioglucose (GTG) (RDI, Flanders, NJ). All mice were sacrificed at 13 wk of age.
Isolation of adipocytes from adipose tissue
Mouse intra-abdominal and s.c. adipose tissues were dissected immediately after sacrifice, minced in 5 ml of Dulbeccos modified Eagles medium (DMEM; Life Technologies, Inc., Invitrogen, Paisley, UK) supplemented with 1 mg/ml collagenase (SIGMA) and 1% BSA for 30 min at 37°C under shaking. Digestion was followed by filtration through a 150 µm screen, and the floating adipocytes were separated from the medium containing the stroma-vascular fraction (SVF). Adipocytes were washed twice in DMEM and further processed for RNA extraction using the RNeasy mini kit (Qiagen, GmbH, Hilden, Germany). Human s.c. adipose tissue explants were prepared and cultured as described previously (16)
, and total RNAs of isolated adipocytes were extracted.
Cell culture
The mouse preadipose cell line 3T3F442A was grown at 37°C in 7% CO2 in DMEM containing 25 mM glucose, 100 U/ml penicillin, 100 µg/ml streptomycin (culture medium), and 10% donor calf serum. For differentiation, confluent preadipocytes were cultured in culture medium supplemented with 10% fetal calf serum and 50 nM insulin for 8 to 10 d, after which more than 90% of the cells had accumulated fat droplets. Total RNAs were prepared using the RNeasy mini kit (Qiagen).
Real-time RT-polymerase chain reaction (PCR)
Total RNAs (1 µg) were reverse-transcribed using random hexamers and Superscript II reverse transcriptase (Invitrogen). The same reaction was performed without Superscript II (RT-) to estimate DNA contamination. Real-time PCR was performed starting with 12.5 ng cDNA, and both sense and antisense oligonucleotides in a final vol of 25 µl using the SYBR green TaqMan Universal PCR Master Mix (Applied Biosystems, Warrington, UK). Fluorescence was monitored and analyzed in a GeneAmp 7000 detection system instrument (Applied Biosystems). Analysis of the 18S rRNA was performed in parallel using the rRNA control Taqman Assay Kit (Applied Biosystem, Warrington, UK) to normalize gene expression. Results are expressed as 2(Ct18SCtgene)(1(1/2(CtgeneCtRT)) ), where Ct corresponds to the number of cycles needed to generate a fluorescent signal above a predefined threshold. Oligonucleotide primers were designed using the Primer Express software (Applied Biosystems, Warrington, UK). All primers used were validated for PCR efficiency.
Apelin assay
Apelin was quantified using the nonselective apelin immunoassay kit (Phoenix Pharmaceuticals Belmont, CA) following the manufacturers instructions.
Statistical analysis
Results are expressed as means ± SEM. Statistical differences between two groups were evaluated using Students t test. Correlations were analyzed using the nonparametric Spearman rank test. The concentration of significance was set at P < 0.05.
RESULTS
The relationship between obesity, TNF
, and apelin expression in humans
TNF
and apelin expression were investigated in s.c. abdominal fat pad from lean or moderately obese humans (BMI ranged from 20.5 to 36.5 kg/m2, n=20). A linear correlation (r=0.790, slope 0.085, P <0.01) was found between TNF
and apelin mRNA levels (Fig. 1
A). A similar correlation (r=0.994, slope 0.097, P < 0.01) was found in the intra-abdominal adipose tissue of massively obese subjects (BMI ranged from 39.2 to 59.3 kg/m2, n=23), (Fig. 1B
). Thus, it clearly appears that a parallel increase of TNF
and apelin expression occurred in adipose tissue, whatever the anatomical location, the highest correlation being observed in morbid grade III obesity. No significant correlation with apelin has been found for other adipokines such as leptin, adiponectin or the plasminogen activator inhibitor type-1 factor (PAI-1) (not shown).
|
To further study apelin regulation by TNF
in human adipose tissue in vitro, explants of s.c. adipose tissue were prepared and maintained in primary culture for 48h as described previously (16)
. Adipocytes were then isolated from explants and both TNF
and apelin mRNAs were quantified. As shown in Fig. 1C
, the rise in TNF
expression started 3 h after the beginning of primary culture and was significantly different after 6 h culture from cells at time zero. Apelin expression in isolated adipocytes from cultured explants increased significantly only beyond the 12th h after the beginning of culture. Furthermore, explants were cultured in the presence of isobutylmethylxanthine (IBMX), a phosphodiesterase inhibitor known to inhibit TNF
secretion and expression in adipocytes from cultured explants (16)
. In the present study, 100 µM IBMX, attenuated the rise of TNF
as well as apelin mRNA in adipocytes. Taken together, these data strongly suggest the involvement of TNF
produced during the culture in the up-regulation of apelin in human adipocytes.
Regulation of apelin by TNF
in mouse
Since obesity is associated with a rise in adipose tissue TNF
content, we checked in mouse models of obesity the in vivo regulation of apelin by TNF
. Thus, we studied apelin and TNF
expression in adipocytes isolated from C57Bl6/J mice either fed a HFD or LFD, and in mice rendered hyperphagic and obese by gold-thioglucose (GTG) treatment. Both HFD-fed C57Bl6/J and GTG-treated FVB/n mice exhibited a significant increase in body mass (28.8±1.5 and 39.4±3.1 g, respectively) when compared with their control littermates (23.4±0.5 and 23.0±0.4 g, respectively). Apelin plasma levels were increased in both models of obese mice (14)
, while TNF
plasma levels were not significantly different from control mice (not shown). However, as shown in Fig. 2
, both TNF
and apelin expression were significantly increased in isolated adipocytes of obese mice. Interestingly, in a model of obese transgenic mice (17)
where TNF
expression was not significantly increased (18)
in adipocytes, apelin expression (14)
was also not increased. In addition, the expression of other pro-inflammatory adipokines such as PAI-1 or interleukin (IL)-6 (IL-6) known to be altered in obesity and/or diabetes (18)
were not significantly increased (not shown) in HFD-fed C57Bl6/J and GTG-treated FVB/n mice. These results in mouse are in agreement with a coregulation between apelin and TNF
mRNA in adipocytes.
|
To determine whether increased TNF
levels could directly up-regulate apelin expression as depicted in mouse, we investigated the direct effect of i.p. injection of TNF
into C57Bl6/J mice on apelin expression. Eight hours after injection, 2 µg/ mouse TNF
elicited a significant increase of apelin expression in isolated adipocytes compared with control mice (Fig. 3
A). A similar increase was observed in the stroma-vascular fraction (2.34±0.16 in control and 11.90±3.82 arbitrary units in TNF
-treated mice, P<0.05, n=6). No significant variations of apelin mRNA levels were observed in other tissues such as heart (9.1±0.91 in control and 10.96±1.54 arbitrary units in TNF
-treated mice, n=6) and kidney (8.81±1.23 in control and 9.62±0.76 arbitrary units in TNF
-treated mice, n=6) known to express apelin. In addition, a significant increase of apelin plasma levels (Fig. 3B
) was observed in TNF
-treated mice. All together these results suggest that 1) TNF
can up-regulate apelin expression in vivo specifically in adipose tissue; 2) the increased apelin expression in adipose tissue by TNF
could contribute to apelin plasma levels.
|
Apelin expression is directly regulated by TNF
via the MAPK, JNK and PI3-K but not the PKC signaling pathway
To test a direct effect of TNF
on fat cells, 10 d-differentiated adipocytes from the murine 3T3F442A cell line were used. TNF
had no effect on apelin expression in undifferentiated 3T3F442A cells (not shown). The effect of TNF
on apelin expression in 3T3F442A cells was concentration-dependent (Fig. 4
A). The increased expression of apelin was followed by an increased secretion of apelin in the medium (from 0.04 ng/ml in control cells to 0.35 ng/ml in treated cells). Thus, TNF
can promote directly both apelin expression and secretion in adipocytes.
|
Since TNF
is known to signal through various transduction pathways in adipocytes, we tested the potential intracellular targets of TNF
by using specific inhibitors (Fig. 4B
). Differentiated 3T3F442A cells were either preincubated with 20 µM SP 600125 (an inhibitor of JNK) or 10 µM LY294002 (a PI3K inhibitor). Both of them significantly decreased the TNF
-induced apelin expression. Since TNF
has been shown to signal through ERK 1 and ERK 2 activator in 3T3F442A adipocytes (19)
, we also checked the effects of the MAPK inhibitor PD098059. TNF
-induced increase of apelin expression was significantly inhibited by 50 µM of PD098059. The potential involvement of PKC was studied by using 10 nM of GF109203X, a PKC inhibitor. The GF109203X did not inhibit the expression of apelin induced by TNF
.
DISCUSSION
A large number of studies have demonstrated that obese animals as well as humans have increased TNF
mRNA and protein levels in white adipose tissue but the correlation between adiposity and circulating concentrations of TNF
remains controversial (for review see 20). A variety of experimental and clinical studies suggest that TNF
may act as an important auto/paracrine regulator of fat cell function which limits adipose tissue expansion, but which also promotes insulin resistance that may, in turn causes metabolic disturbances. The chronic and graduated elevation of TNF
in obesity may also promote modifications in the secretory function of the fat cell. It has already been described that increased PAI-1, TGF-ß, and IL-6 productions are linked with high TNF
expression (21)
. Although clearly demonstrated to be directly regulated by TNF
, such adipokines are over-expressed in massively obese patients and are associated with a cluster of abnormalities including hyperinsulinemia insulin resistance or hypertension (22)
. The present study clearly demonstrated a tight positive correlation between apelin and TNF
expression in adipose tissue of massively obese as well as in lean subjects. These results are in concordance with the results obtained by Heinonen et al. (23)
showing that apelin plasma levels correlate positively with body mass index. Moreover, the time course of TNF
and apelin mRNA expression in isolated adipocytes from human cultured explants is likely to explain a paracrine effect of endogenously produced TNF
on apelin expression in the surrounding adipocytes. This hypothesis has been supported by the absence of up-regulation of apelin mRNA in IBMX-treated explants since this phosphodiesterase inhibitor has already been described to inhibit TNF
expression and secretion (16)
.
The up-regulation of both TNF
and apelin expression has also been shown in adipocytes isolated from different mouse models of obesity. Adipocytes from obese mice (or obese subjects) are bigger and could content more mRNA than smaller ones when results are expressed as mg of total mRNA per cell. It is therefore possible that the increased mRNA levels measured might be greater than reported. To further study, the effect of TNF
on apelin expression in adipocytes, an i.p. injection of TNF
was performed and analyses were performed 8 h after injection, time already used to show increased levels of other adipokines in adipose tissue (24)
. TNF
induced a significant increase of apelin mRNA in both isolated adipocytes and the stroma-vascular fraction. However the expression of apelin in other tissues such as kidney and heart was not modified. Moreover, the rise in apelin mRNA observed in 3T3F442A adipocytes led us to conclude to a direct effect of TNF
on apelin expression. Furthermore, the signaling pathways by which TNF
induced apelin expression were depicted. TNF
has been shown to activate several signaling molecules such as PI3-kinase, PKC, MAP kinase or JNK in adipocytes (25)
. By using specific inhibitors of these distinct TNF
signaling pathways, we found that TNF
(8 h) regulates apelin expression in adipocytes through different pathways involving the MAPK, JNK and PI3-kinase. We previously described that insulin induced apelin expression in adipocytes by using also the MAPK and PI3-kinase pathways (14)
. The expression of the forkhead transcription factor gene Foxc2 in 3T3-L1 adipocytes involved similar MAPK- and PI3K-dependent pathways (26)
. Common pathways, used by TNF
and insulin to induce gene expression in 3T3-L1 adipocytes have been already demonstrated (27)
. GFX 109203X, a broad-spectrum inhibitor of PKC isoforms, did not modified the TNF
-induced expression of apelin. Interestingly, this signaling pathway differs from TNF
-mediated PAI-1 expression in 3T3-L1 adipocytes where PKC pathway plays a central role (27)
. So far, very little is known about the signaling pathways mediating apelin expression in other cells type.
The present work clearly demonstrates for the first time that TNF
may act as a direct regulator of apelin expression in both human and mouse adipocytes. In addition, i.p. injection of TNF
in mice induced increased both apelin expression in adipose tissue and blood plasma levels of apelin. In different mouse models of obesity, both apelin expression in adipocytes and apelin plasma levels were increased. Moreover, in obese humans, plasma apelin levels were significantly higher than in lean subjects (14)
. These findings were very recently strengthened by a study showing that apelin plasma levels were dramatically increased in morbid obese patients (23)
. Thus, when considering the total adipose tissue mass in the body, the endocrine role of adipocyte-produced apelin might be of importance. These results presented herein suggest that the increased apelin expression in adipose tissue could contribute to apelin plasma levels. In addition, the data from Xu et al. (28)
showed that an increased macrophage and inflammation gene levels (such as TNF
) seems to occur selectively in adipose tissue before the rise in circulating-insulin levels. Such an observation indicates that inflammation is observed before the appearance of insulin resistance. Thus, a mutual regulation of apelin production in adipocytes by both insulin and TNF
could be hypothesized, leading to the maintenance of sustained apelin expression and secretion in obesity. Taking in account the physiological roles of apelin in the cardiovascular system, the over-production of apelin in the obese could be one of the last protections before the emergence of obesity-related disorder such as hypertension and cardiovascular dysfunctions. On the other hand, apelin up-regulation in obesity could contribute to endocrine or metabolic dysfunctions such as diabetic retinopathy since apelin has been proposed as a novel angiogenic factor in retinal endothelial cells (29)
. Thus, further studies to understand the contribution of apelin in associated regulations/dysregulations of obesity are of major interest.
ACKNOWLEDGMENTS
We gratefully acknowledge the animal facilities staff (Zootechnie, IFR31), the plastic surgery service of Rangueil Hospital, Bruno Séguy (INSERM U466, IFR31) for help in TNF
assay and Max Lafontan for fruitful discussions.
Received for publication November 14, 2005. Accepted for publication February 16, 2006.
REFERENCES
. Cytokine Growth Factor Rew. 14,447-455
initiated signal transduction in 3T3L1 adipocytes. J. Cell. Physiol. 179,58-66[CrossRef][Medline]
-mediated plasminogen activator inhibitor-1 expression in adipocytes. FASEB J. 19,1317-1319This article has been cited by other articles:
![]() |
M. Eyries, G. Siegfried, M. Ciumas, K. Montagne, M. Agrapart, F. Lebrin, and F. Soubrier Hypoxia-Induced Apelin Expression Regulates Endothelial Cell Proliferation and Regenerative Angiogenesis Circ. Res., August 15, 2008; 103(4): 432 - 440. [Abstract] [Full Text] [PDF] |
||||
![]() |
I. Castan-Laurell, M. Vitkova, D. Daviaud, C. Dray, M. Kovacikova, Z. Kovacova, J. Hejnova, V. Stich, and P. Valet Effect of hypocaloric diet-induced weight loss in obese women on plasma apelin and adipose tissue expression of apelin and APJ. Eur. J. Endocrinol., June 1, 2008; 158(6): 905 - 910. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Lambin, R. van Bree,, I. Vergote, and J. Verhaeghe Chronic Tumor Necrosis Factor-{alpha} Infusion in Gravid C57BL6/J Mice Accelerates Adipose Tissue Development in Female Offspring Reproductive Sciences, December 1, 2006; 13(8): 558 - 565. [Abstract] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |