TNF- Represses -Klotho Expression and Impairs FGF21 Action in ...

ENERGY

BALANCE-OBESITY

TNF-␣ Represses ␤-Klotho Expression and Impairs FGF21 Action in Adipose Cells: Involvement of JNK1 in the FGF21 Pathway Julieta Díaz-Delfín,* Elayne Hondares,* Roser Iglesias, Marta Giralt, Carme Caelles, and Francesc Villarroya Department of Biochemistry and Molecular Biology and Institut de Biomedicina (J.D.-D., E.H., R.I., M.G., F.V.), University of Barcelona, and CIBER Fisiopatología de la Obesidad y Nutrición, and Institute for Research in Biomedicine (C.C.), Barcelona Science Park, Barcelona 08028, Spain

Fibroblast growth factor 21 (FGF21) is a member of the FGF family that reduces glycemia and ameliorates insulin resistance. Adipose tissue is a main target of FGF21 action. Obesity is associated with a chronic proinflammatory state. Here, we analyzed the role of proinflammatory signals in the FGF21 pathway in adipocytes, evaluating the effects of TNF-␣ on ␤-Klotho and FGF receptor-1 expression and FGF21 action in adipocytes. We also determined the effects of rosiglitazone on ␤-Klotho and FGF receptor-1 expression in models of proinflammatory signal induction in vitro and in vivo (high-fat diet-induced obesity). Because c-Jun NH2-terminal kinase 1 (JNK1) serves as a sensing juncture for inflammatory status, we also evaluated the involvement of JNK1 in the FGF21 pathway. TNF-␣ repressed ␤-Klotho expression and impaired FGF21 action in adipocytes. Rosiglitazone prevented the reduction in ␤-Klotho expression elicited by TNF-␣. Moreover, ␤-Klotho levels were reduced in adipose tissue from high-fat diet-induced obese mice, whereas rosiglitazone restored ␤-Klotho to near-normal levels. ␤-Klotho expression was increased in white fat from JNK1⫺/⫺ mice. The absence of JNK1 increased the responsiveness of mouse embryonic fibroblastderived adipocytes and brown adipocytes to FGF21. In conclusion, we show that proinflammatory signaling impairs ␤-Klotho expression and FGF21 responsiveness in adipocytes. We also show that JNK1 activity is involved in modulating FGF21 effects in adipocytes. The impairment in the FGF21 response machinery in adipocytes and the reduction in FGF21 action in response to proinflammatory signals may play important roles in metabolic alterations in obesity and other diseases associated with enhanced inflammation. (Endocrinology 153: 4238 – 4245, 2012)

F

ibroblast growth factor-21 (FGF21) is a hormonal factor reported to have powerful antidiabetic effects (1). FGF21 is abundantly expressed in adipose tissues, liver, and pancreas (2, 3). Although the liver is a candidate for paracrine and autocrine actions of FGF21, the most dramatic effects of FGF21 have been found in adipose tissue (1, 4), where it promotes glucose uptake and oxidation (1). FGF21 acts through FGF receptors (FGFR) associated with the auxiliary protein ␤-Klotho. The interaction of FGFR with ␤-Klotho confers FGF21 sensitivity and the

ability to activate intracellular signaling pathways that ultimately lead to biological effects. White adipose tissue (WAT) and brown adipose tissue express high levels of the FGFR 1c and ␤-Klotho, consistent with these tissues being highly sensitive targets of FGF21 (3, 5). Pharmacological studies have shown that treatment with FGF21 has broad systemic metabolic actions in obese rodents and primates that include enhancing insulin sensitivity, decreasing triglyceride concentrations, and causing weight loss (1, 6). However, observations in both ro-

ISSN Print 0013-7227 ISSN Online 1945-7170 Printed in U.S.A. Copyright © 2012 by The Endocrine Society doi: 10.1210/en.2012-1193 Received February 21, 2012. Accepted June 6, 2012. First Published Online July 9, 2012

* J.D.-D. and E.H. contributed equally to this study.

4238

endo.endojournals.org

Abbreviations: Cytc, Cytochrome c; ER, endoplasmatic reticulum; FGF, fibroblast growth factor; FGFR, FGF receptor; GLUT, glucose transporter; HF, high-fat, high-carbohydrate diet; JNK, Jun NH2-terminal kinase; MCP-1, monocyte chemoattractant protein-1; MEF, mouse embryonic fibroblast; PGC1, PPAR-␥ coactivator-1␣; PPAR, peroxisome proliferator-activated receptor; TZD, thiazolidinedione; UCP, uncoupling protein; WAT, white adipose tissue; WT, wild type.

Endocrinology, September 2012, 153(9):4238 – 4245


dents and humans indicate that obesity, and especially insulin resistance, are associated with a paradoxically abnormal elevation of serum FGF21 levels (7, 8). It has been proposed that obesity is in fact an FGF21-resistant state (9). However, studies indicating the capacity of even low doses of FGF21, when administered to obese animals to improve metabolic parameters, argues against this statement (10, 11). Obesity is associated with a chronic proinflammatory state in WAT that is relevant to the development of insulin resistance. In obesity, c-Jun NH2-terminal kinase (JNK) activity is abnormally elevated whereas its deficiency results in decreased adiposity and improvement of insulin sensitivity in murine models of obesity (12). JNK family kinases are activated by cytokines such as TNF-␣, as well as by endoplasmatic reticulum (ER) stress, all of which are elevated in obesity and/or type 2 diabetes (13). Previous reports have indicated that thiazolidinediones (TZD), drugs used clinically as insulin-sensitizing agents, inhibit the JNK1 pathway in cell culture and in vivo (14) and also modulate FGF21 actions in 3T3-L1 adipocytes (4, 15). These findings raise the possibility that JNK1 could be involved in modulating the actions of FGF21. Here, we show, for the first time, that TNF-␣, a pivotal agent in inflammation in adipose tissue, represses the expression of ␤-Klotho and impairs the biological actions of FGF21 in adipocytes. Moreover, we found that JNK1 activity is involved in modulating FGF21 effects in adipose cells.

Materials and Methods Materials Reagents used for adipocyte cell cultures were from Sigma (St. Louis, MO), with the exception of rosiglitazone (Alexis Biochemicals, Carlsbad, CA) and mouse FGF21 (Phoenix Pharmaceuticals, Inc., Mountain View, CA).

Cell culture, differentiation, and treatment Mouse 3T3-L1 adipocytes and SGBS human adipocytes were cultured and differentiated as described previously (14, 16). Mouse embryonic fibroblasts (MEF) from wild-type (WT) and JNK1-deficient (JNK1⫺/⫺) mice were isolated and differentiated into adipocytes as described previously (17). Primary cultures of brown adipocytes from WT and JNK1⫺/⫺ mice were established and maintained as reported elsewhere (18). Experiments were performed when 80 –90% of cells had differentiated, as determined on the basis of acquisition of an adipocyte morphology. Differentiated adipocytes were treated for 24 h with 100 nM FGF21, 10 ng/ml TNF-␣, 1 ␮M thapsigargin, or 10 ␮M rosiglitazone. MEF-derived adipocytes and brown adipocytes were exposed to 100 nM and 50 nM of mouse recombinant FGF21 protein, respectively, for 24 h.


4239

Quantitative real-time RT-PCR RNA was extracted (NucleoSpin; Macherey Nagel), and transcript levels were determined by quantitative RT-PCR using TaqMan Assay-on-demand probes and the ABI/Prism-7700 Detector System (Applied Biosystems, Foster City, CA). Assay-ondemand probes used were: ␤-Klotho (Mm00473122), FGFR1 (Mm00438930), glucose transporter (GLUT)1 (Mm00441480), uncoupling protein (UCP)1 (Mm00494069), cytochrome c (Cytc) (Mm01621044), peroxisome proliferator-activated receptor (PPAR)-␥ coactivator-1␣ (PGC1) (Mm00447183), cytosolic fattyacid-binding protein (Mm00445880), TNF-␣ (Mm00443258), monocyte chemoattractant protein-1 (Mm00441242), GLUT4 (Mm00436615), and 18S rRNA (Hs99999901). The mean value of sample duplicates for each transcript was normalized to that of the 18S rRNA gene using the comparative (2-⌬CT) method.

Determination of 3H-labeled 2-deoxyglucose uptake Before assaying for glucose uptake, adipocytes were incubated for 24 h with TNF-␣ and then incubated for another 24 h in the presence or absence of FGF21. Glucose uptake assays were performed as described previously (1).

Immunoblotting ␤-Klotho and ␤-actin were detected using the antibodies BAF2619 (R&D Systems, Minneapolis, MN) and A5441 (Sigma), respectively. Immunoblotting was performed using an enhanced chemiluminescence detection system (Amersham Pharmacia Biotech, Piscataway, NJ).

Animals and in vivo studies All experimental protocols using animals were approved by the Animal Care Research Committee of the University of Barcelona. For studies in HF diet obese mice and effects of rosiglitazone treatment, male C57BL/6J mice were divided randomly into three groups at 4 wk of age. One group was fed standard chow (F4031 BioServ), and the other two received a high-fat, highcarbohydrate diet (HF) (F3282, BioServ) for 13 wk. One subgroup of HF-fed mice was subjected to rosiglitazone treatment. Rosiglitazone (1 mg/kg) or vehicle (PBS) was administered by oral gavage once a day for 17 consecutive days. In mice on d 9 of treatment with rosiglitazone and in the nontreated mice groups glucose tolerance tests were performed. Mice were fasted for 6 h before oral administration of glucose (2 mg/g body weight), and blood was collected for glucose quantification from the tail vein. Plasma insulin and FGF21 levels were determined by ELISA; [Mercodia (Uppsala Sweden) and Phoenix Secretomics, Burlingame, CA], respectively. JNK1⫺/⫺ mice and WT littermates were on a C57BL/6J strain background. Blood was obtained by cardiac puncture, and WAT (epididymal fat pads) were excised and frozen immediately in liquid nitrogen.

Statistics The distribution of data was controlled for normality. Unpaired Student’s t tests were used to test the level of significance of the differences between means.

4240

Díaz-Delfín et al.

TNF-␣ Represses ␤-Klotho in Adipose Cells


Results TNF-␣ inhibits the expression of ␤-Klotho in 3T3L1 adipocytes Mouse 3T3-L1 adipocytes were treated with TNF-␣, which is well known to efficiently stimulate proinflammatory pathways in different cells systems. TNF-␣ caused very significant reduction of the ␤-Klotho mRNA expression in 3T3-L1 adipocytes (Fig. 1A). A time-course analysis revealed that maximal inhibition of ␤-Klotho expression occurred after 24 h of TNF-␣ treatment (data not shown). FGFR1 mRNA levels were unchanged. Similar reductions of ␤-Klotho mRNA levels by TNF-␣ were obtained in distinct models of adipogenic cells, including SGBS human adipocytes (0.2-fold), MEF-derived adipocytes (0.65-fold), and brown adipocytes (0.7-fold) (data not shown). Consistent with the observed changes in mRNA levels, an analysis of ␤-Klotho protein expression showed that TNF-␣ significantly decreased ␤-Klotho protein levels (Fig. 1B). To determine whether down-regulation of ␤-Klotho by TNF-␣ was associated with alterations in FGF21 action, we evaluated FGF21 effects in 3T3-L1 adipocytes in the absence or presence of TNF-␣. As shown in Fig. 2A, FGF21 significantly induced glucose uptake (2-fold) in 3T3-L1 adipocytes, as previously reported (1). TNF-␣ alone also caused an increase in basal glucose uptake, as has also been shown (19, 20). However, exposure of cells to TNF-␣ significantly reduced the effects of FGF21 on glucose uptake (1.4-fold). We also analyzed the action of FGF21 on GLUT1 gene expression. The results essentially paralleled those found for glucose uptake: FGF21 up-regulated GLUT1 (Fig. 2B), in agreement with previous reports in adipocytes (5), and TNF-␣ alone also significantly increased GLUT1 mRNA levels, also as previously re-

FIG. 1. TNF-␣ inhibits ␤-Klotho expression in 3T3-L1 adipocytes. 3T3L1 adipocytes were treated with 10 ng/ml TNF-␣ and/or 100 nM FGF21 for 24 h. A, Effects of TNF-␣ on ␤-Klotho and FGFR1 mRNA levels. B, ␤-Klotho protein levels, expressed as a ratio of the densitometric intensity of the immunoreactive signal of ␤-Klotho protein to that of ␤-actin (top). Representative immunoblot showing ␤-Klotho and ␤actin from 3T3-L1 adipocytes treated with TNF-␣, as indicated (bottom). Data are presented as means ⫾ SEM from three to four independent experiments and are expressed relative to values from control cells (**, P ⬍ 0.01). C, Control.

FIG. 2. TNF-␣ inhibits FGF21 action in 3T3-L1 adipocytes. 3T3-L1 adipocytes were treated as Fig, 1. A, Effects of TNF-␣ on FGF21induced glucose uptake, measured as 2-deoxy-D-[3H]glucose uptake. B, Effects of FGF21 and TNF-␣ on GLUT1 mRNA levels. Data are presented as means ⫾ SEM from three to four independent experiments and are expressed relative to values from control cells (*, P ⬍ 0.05; ***, P ⬍ 0.001 for FGF21 effect in presence or absence of TNF-␣; #, P ⬍ 0.05, for comparisons between control and TNF-␣treated cells). C, Control.

ported (21); however, the presence of TNF-␣ blunted the FGF21-induced increase in GLUT1 (Fig. 2B). When the expression of another glucose transporter, GLUT4, was analyzed, we found that, although TNF-␣ reduced GLUT4 gene expression (0.6-fold reduction) as previously reported (16, 22), FGF21 had no effect on GLUT4 gene expression alone and did not alter TNF-␣ effects (data not shown). These data indicate that the TNF-␣induced decrease in ␤-Klotho expression in mouse adipocytes is associated with an impairment of FGF21 action on glucose uptake and GLUT1 gene expression. Rosiglitazone prevents the down-regulation of ␤Klotho expression caused by proinflammatory signaling in vitro and in vivo studies If the proinflammatory, insulin-resistant status of adipose tissue in obesity is associated with abnormal expression of ␤-Klotho, it is reasonable to postulate that an agent that protects against inflammation and insulin resistance, such as rosiglitazone, might prevent ␤-Klotho repression. To test this hypothesis, we treated 3T3-L1 adipocytes with rosiglitazone and/or TNF-␣ for 24 h. As shown in Fig. 3A, rosiglitazone induced ␤-Klotho mRNA expression in TNF-␣-treated cells, restoring ␤-Klotho mRNA to levels not significantly different from those in control cells not treated with TNF-␣. The expression levels of the FGFR1 receptor were unaltered (Fig. 3B). Like TNF-␣, treatment with thapsigargin, an ER stress activator and another stimulator of proinflammatory signals in adipocytes (23), also reduced ␤-Klotho mRNA expression, an effect that was partially prevented by rosiglitazone (Fig. 3A). An analysis of ␤-Klotho protein levels under these experimental conditions led to findings that paralleled those for ␤-Klotho mRNA: a reduction in ␤-Klotho protein levels by TNF-␣ and thapsigargin and a restorative action of rosiglitazone on ␤-Klotho protein expression (Fig. 3C).



4241

(Fig. 4B) and reduced their hyperinsulinemia (Fig. 4C) without altering body weight, in agreement with previous results using similar experimental designs (24). TNF-␣ and MCP-1 expression were significantly induced in WAT from HF diet-induced obese mice, and rosiglitazone treatment was also effective in decreasing the expression of these inflammation-related genes in obese mice (Fig. 4D). ␤-Klotho and FGFR1 expression was reduced in WAT from obese mice (Fig. 4E), in agreement with a previous report (9). Rosiglitazone treatment of HF-fed obese mice restored the expression of ␤-Klotho and FGFR1 receptors to levFIG. 3. Rosiglitazone (Rosi) reverses the down-regulation of ␤-Klotho induced by TNF-␣ and els similar to those of control mice (Fig. thapsigargin (Thap) in 3T3-L1 adipocytes. Effects of Rosi on TNF-␣ and Thap-regulated ␤-Klotho (A) and FGFR1 (B) mRNA levels in 3T3-L1 adipocytes. Data are presented as means ⫾ SEM from three to 4E). Consistent with the observed four independent experiments and are expressed relative to values from control cells (*, P ⬍ 0.05; **, changes in mRNA levels, an analysis of P ⬍ 0.01). C, Representative immunoblot showing ␤-Klotho and ␤-actin from 3T3-L1 adipocytes ␤ -Klotho protein expression showed treated with Rosi, TNF-␣, or Thap, as indicated. Numbers indicate the fold change in densitometric data of the immunoblot signals with respect to control untreated cells. C, Control. that a HF diet significantly decreased ␤-Klotho protein levels, in agreement To test whether the observed in vitro effects of rosiglitazone translate to the obesity state in vivo, we analyzed with previous reports (11). Furthermore, ␤-Klotho prothe mRNA levels of ␤-Klotho and FGFR1 in adipose tissue tein levels were essentially normalized by rosiglitazone from mice that develop obesity (60% body-weight gain vs. treatment (Fig. 4G). Moreover, plasma levels of FGF21 controls) after fed a HF diet (Fig. 4A). Treatment with were increased by a HF diet and were normalized by rosiglitazone caused a significant improvement in glucose rosiglitazone (Fig. 4F), consistent with reports using simtolerance in these HF-fed obese and hyperglycemic mice ilar experimental designs (24). The plasma levels of FGF21

FIG. 4. Rosiglitazone reverses the down-regulation of ␤-Klotho expression in WAT from obese mice. Mice were fed with control (C) or high-fat, high-carbohydrate diet (HF), and animals on HF were treated with rosiglitazone (HF⫹R). A, Body weight (g) after rosiglitazone treatment. B, Glucose tolerance test on d 9 of treatment with rosiglitazone. C, Plasma insulin levels. D, Transcript levels of TNF-␣ and MCP-1 in WAT. E, Transcript levels of ␤-Klotho and FGFR1 in WAT. F, Plasma FGF21 levels. G, ␤-Klotho protein levels (top), and representative immunoblot showing ␤-Klotho and ␤-actin (bottom) in WAT, as indicated. Bars are presented as means ⫾ SEM (n ⫽ 5– 8). (*, P ⬍ 0.05; **, P ⬍ 0.01; ***, P ⬍ 0.001 for comparisons between HF and controls; #, P ⬍ 0.05; ##, P ⬍ 0.01 for comparisons between HF⫹R and HF).

4242



in response to HF and after rosiglitazone treatment paralleled changes in hepatic FGF21 mRNA expression (6.5fold induction by HF vs. controls; 1.6-fold change in HF plus rosiglitazone vs. control; P ⬍ 0.001, not significant). In adipose tissue, HF increased FGF21 mRNA expression (3-fold induction in HF vs. controls, P ⬍ 0.05) and, in contrast with liver, rosiglitazone further increased FGF21 mRNA expression (7.3-fold increase in HF plus rosiglitazone vs. control; P ⬍ 0.05). These observations are in agreement with recent data in similar experimental settings (24) and support current assumptions of a main role of FGF21 expression in liver as determinant of FGF21 plasma levels under most pathophysiological conditions. Loss of JNK1 induces ␤-Klotho expression and enhances FGF21 action in MEF-derived adipocytes and brown adipocytes Considering that obesity causes chronic JNK1 activation, and experiments in both cell-based systems and whole animals have demonstrated that JNK serves as a sensing juncture for cellular stress and inflammatory status (12, 13), we analyzed whether JNK1 might be involved in suppression of ␤-Klotho expression and responsiveness to FGF21 action. JNK1 is known to be expressed in adipose tissue and adipocytes in culture (14), an observation that we corroborated through immunoblot detection of JNK1 in adipose tissue from mice as well as in the adipogenic cell culture models used in this study (data not shown). We first investigated the expression of ␤-Klotho in WAT from JNK1⫺/⫺ mice. ␤-Klotho mRNA was significantly up-regulated in JNK1⫺/⫺ mice compared with WT mice, although the FGFR1 mRNA levels were similar in the presence or absence of JNK1 (Fig. 5A). Changes in

FIG. 5. ␤-Klotho expression in WAT from JNK1⫺/⫺ mice. A, Transcript levels of ␤-Klotho and FGFR1 and B) a representative immunoblot showing ␤-Klotho and ␤-actin in WAT from WT and JNK1⫺/⫺ mice, as indicated. Data are presented as means ⫾ SEM (*, P ⬍ 0.05, WT vs. JNK1⫺/⫺ mice; n ⫽ 5– 8).


␤-Klotho mRNA in WAT were mirrored by changes in ␤-Klotho protein levels, which were also increased in the absence of JNK1 (Fig. 5B). Next, two different adipocyte models derived from JNK1⫺/⫺ mice were used to test the role of JNK1 in the FGF21 pathway: MEF-derived adipocytes and brown adipocytes. Treatment of WT MEF-derived adipocytes with FGF21 induced a moderate, but significant, increase in glucose uptake and GLUT1 expression, and treatment with TNF-␣ impaired the action of FGF21 (Fig. 6, A and B) in association with a significant reduction in ␤-Klotho mRNA expression (Fig. 6C), as previously observed in 3T3-L1 adipocytes. In JNK1⫺/⫺ MEF-derived adipocytes, the extent of induction of glucose uptake and GLUT1 mRNA expression by FGF21 was significantly higher than in WT cells (Fig. 6, A and B). Treatment of JNK1⫺/⫺ MEFderived adipocytes with TNF-␣ did not impair FGF21 effects on glucose uptake. Although ␤-Klotho expression was significantly reduced in response to TNF-␣ in JNK1⫺/⫺ MEF-derived adipocytes, the levels of ␤-Klotho remained significantly higher than those in TNF-␣-treated WT cells and were not significantly different from those in control cells (Fig. 6C). Finally, we determined the effects of FGF21 on the expression of genes other than GLUT1. We observed that expression of UCP1, another gene target of FGF21 action, was more intensely induced by FGF21 in JNK1⫺/⫺ MEFderived adipocytes than in WT cells (Fig. 7A), whereas aP2 (cytosolic fatty-acid-binding protein), a gene the expression of which is insensitive to FGF21 in WT MEF-derived adipocytes, was equally unresponsive to FGF21 in JNK1⫺/⫺ MEF-derived adipocytes. In parallel, we also studied the effects of FGF21 in primary cultures of brown adipocytes. This cell model is highly sensitive to FGF21 effects: in addition to promoting glucose uptake, FGF21 markedly induces thermogenic gene expression in these cells (18). Basal levels of expression of thermogenic genes tended to be higher in JNK1⫺/⫺ brown adipocytes (3.8fold for UCP1, 1.9-fold for PGC1␣, 1-fold for Cytc; data not shown). Treatment of WT brown adipocytes with FGF21 markedly induced the expression of the thermogenic genes UCP1, PGC1␣, and Cytc, as well as the FGF21-responsive gene GLUT1 (Fig. 7B), in agreement with previous reports (18). The absence of JNK1 strongly increased the response to FGF21 (Fig. 7B). Thus, changes in JNK1 expression resulted in altered FGF21 responsiveness in mouse adipocytes, consistent with an inhibitory role of JNK1 in FGF21 signaling pathways.

Discussion The present data show for the first time that TNF-␣, a pivotal mediator of inflammation in adipose tissue, dra-



4243

stimuli in vitro and in vivo. This is consistent with reports that rosiglitazone enhances FGF21 action on adipocytes through the activation of PPAR-␥ (4), and TZD induce ␤-Klotho expression (15). The fact that PPAR-␥ is down-regulated by TNF-␣ in adipocytes (27, 28) makes it likely that this is a mechanism by which TNF-␣ represses ␤-Klotho expression; it may also contribute to an explanation of why rosiglitazone-mediated PPAR-␥ activation favors FGF21 effects. Further investigation is required to confirm this mechanism. It is worth mentioning that rosiglitazone-induced promotion of FGF21 effects (glucose uptake, induction of the UCP1 gene) does not lead to enhanced energy expenditure to an extent capable of reducing body weight. The paradoxical observation that rosigilitazone up-regulates thermogenesisrelated processes in adipose tissues without increasing whole-body energy FIG. 6. Effects of JNK1 ablation on FGF21 action in MEF-derived adipocytes. A, Effects of FGF21 on glucose uptake, measured as 2-deoxy-D-[3H]glucose uptake and B) transcript levels expenditure has been previously reof GLUT1 in MEF-derived adipocytes from WT and JNK1⫺/⫺ mice. C, Effects of TNF-␣ on ␤ported (29 –32). The present observaKlotho mRNA levels in MEF-derived adipocytes from WT and JNK1⫺/⫺ mice. Data are tion that rosiglitazone favors FGF21 presented as means ⫾ SEM from three to four independent experiments and are expressed relative to values from control cells (*, P ⬍ 0.05; **, P ⬍ 0.01 for comparisons due to action in adipose tissues should be treatment with FGF21 in presence or absence of TNF-␣; #, P ⬍ 0.05; ##, P ⬍ 0.01 for added to this paradox, which will recomparisons between WT vs. JNK1⫺/⫺ adipocytes). quire further research to be solved. Whereas the foregoing results all are matically reduces ␤-Klotho expression and impairs the consistent with the hypothesis that inflammation may biological effectiveness of FGF21 in adipocytes. blunt the actions of FGF21 via ␤-Klotho down-regulation, ␤-Klotho is a cofactor essential for FGF21 activity (5, we found evidence linking inflammation, JNK1, and the 25); thus, cells lacking ␤-Klotho are unable to respond to FGF21 pathway. First, in agreement with the decreased FGF21 (26). We demonstrated that down-regulation of levels of ␤-Klotho in obesity, we observed that ␤-Klotho ␤-Klotho induced by TNF-␣ is associated with a reduction levels in WAT were abnormally high in mice lacking in the capacity of FGF21 to promote glucose uptake in JNK1. In accord with this observation, JNK1 activity is adipocytes. A reduction in ␤-Klotho expression in adipose increased in dietary and genetic mouse models of obesity tissue occurs in obesity (9), and the present findings (12). Second, TNF-␣ and ER stimulation are known to strongly support the possibility that the enhanced inflam- induce JNK1 activity (12, 13). And finally, rosiglitazone mation in adipose tissue in obesity, exemplified by in- may act as a JNK1 inhibitor (14). All these observations fit creased TNF-␣ levels, influences the actions of FGF21 with a role for the TNF-␣-JNK1 axis in the negative conthrough strong down-regulation of ␤-Klotho. Addition- trol of FGF21 effects, which may involve the repression of ally, we identified ER stress as a potential link between ␤-Klotho expression. Consistent with this, we showed obesity and the down-regulation of ␤-Klotho, showing that adipocytes derived from JNK1-null mice are particthat ER stress in adipose tissue, which occurs in obesity, ularly sensitive to FGF21 action. Although the increase in may alter the actions of FGF21. In light of these findings, ␤-Klotho expression may be highly relevant to the inwe hypothesized that drugs combining antidiabetic and creased responsiveness to FGF21 in JNK1-deficient adiantiinflammatory effects, such as TZD, could improve pocytes, further studies are required to fully define the FGF21 effects. Thus, rosiglitazone reversed the down-reg- precise mechanisms by which JNK1 influences the responulation of ␤-Klotho expression induced by inflammatory siveness of adipocytes to FGF21.

4244




Acknowledgments We thank Dr. R.A. Flavell (Department of Immunobiology, Yale University School of Medicine, New Haven, CT) for providing the JNK1⫺/⫺ mice. Address all correspondence and requests for reprints to: Julieta Díaz-Delfín, Department of Biochemistry and Molecular Biology, University of Barcelona, Avda Diagonal 643, 08028Barcelona, Spain. E-mail: [email protected]. This work was supported by grants from Ministerio de Ciencia e Innovación (Grants SAF2011-23636 and SAF201021682), Spain. Disclosure Summary: The authors declare no conflict of interest. FIG. 7. Effects of JNK1 ablation on FGF21 induction of gene expression in adipose cells. A, Effects of FGF21 on transcript levels of GLUT1, UCP1, and Cytc in MEF-derived adipocytes from WT and JNK1⫺/⫺ mice. B, Effects of FGF21 on UCP1, PGC1␣, Cytc, and aP2 (cytosolic fatty-acid-binding protein) mRNA levels in brown adipocytes. Data are presented as means ⫾ SEM from three to four independent experiments and are expressed relative to values from control cells (*, P ⬍ 0.05; **, P ⬍ 0.01; ***, P ⬍ 0.001 compared with untreated cells for each genetic background; #, P ⬍ 0.05; ##, P ⬍ 0.01; ###, P ⬍ 0.001 for comparisons between WT vs. JNK1⫺/⫺ adipocytes).

Few drugs that are effective for the treatment of obesity and related metabolic abnormalities are available. Inhibition of JNK1 expression in peripheral tissues can improve adiposity and could provide clinical benefit for these disorders (33). Indeed, JNK1 is a potential target for treatment of metabolic syndrome (34). In addition, FGF21 is now considered a potential candidate in the treatment of a variety of metabolic diseases (35). However, the acute response to FGF21 appears to be attenuated in obese/diabetic animals, and long-term administration is required to sustain the antiobesity effect of FGF21 (36). In this context, it would be interesting to explore combination therapies with antiinflammatory agents that promote the actions of FGF21 in obesity. Overall, the present study reveals a novel functional interaction of the FGF21 system with proinflammatory signaling involving the repressive effects of TNF-␣ on the expression of ␤-Klotho, a pivotal actor in the cellular machinery that mediates the response to FGF21. The complex cross-talk between the FGF21 system and inflammatory pathways is also highlighted by recent reports that FGF21 is capable of inhibiting NF-␬B activity (37) and is induced by inflammatory stimuli and protects animals against the toxic effects of lipopolysaccharide and sepsis (38). Moreover, the present results highlight the involvement of JNK in the negative regulation of ␤-Klotho expression and FGF21 action in adipose tissue, thus suggesting the potential therapeutic value of combinations of FGF21 and JNK1 inhibitors in obesity and associated diseases.

References 1. Kharitonenkov A, Shiyanova TL, Koester A, Ford AM, Micanovic R, Galbreath EJ, Sandusky GE, Hammond LJ, Moyers JS, Owens RA, Gromada J, Brozinick JT, Hawkins ED, Wroblewski VJ, Li DS, Mehrbod F, Jaskunas SR, Shanafelt AB 2005 FGF-21 as a novel metabolic regulator. J Clin Invest 115:1627–1635 2. Muise ES, Azzolina B, Kuo DW, El-Sherbeini M, Tan Y, Yuan X, Mu J, Thompson JR, Berger JP, Wong KK 2008 Adipose fibroblast growth factor 21 is up-regulated by peroxisome proliferator-activated receptor ␥ and altered metabolic states. Mol Pharmacol 74: 403– 412 3. Fon Tacer K, Bookout AL, Ding X, Kurosu H, John GB, Wang L, Goetz R, Mohammadi M, Kuro-o M, Mangelsdorf DJ, Kliewer SA 2010 Research resource: comprehensive expression atlas of the fibroblast growth factor system in adult mouse. Mol Endocrinol 24: 2050 –2064 4. Moyers JS, Shiyanova TL, Mehrbod F, Dunbar JD, Noblitt TW, Otto KA, Reifel-Miller A, Kharitonenkov A 2007 Molecular determinants of FGF-21 activity-synergy and cross-talk with PPAR ␥ signaling. J Cell Physiol 210:1– 6 5. Ogawa Y, Kurosu H, Yamamoto M, Nandi A, Rosenblatt KP, Goetz R, Eliseenkova AV, Mohammadi M, Kuro-o M 2007 ␤Klotho is required for metabolic activity of fibroblast growth factor 21. Proc Natl Acad Sci USA 104:7432–7437. 6. Kliewer SA, Mangelsdorf DJ 2010 Fibroblast growth factor 21: from pharmacology to physiology. Am J Clin Nutr 91:254S–257S 7. Badman MK, Kennedy AR, Adams AC, Pissios P, Maratos-Flier E 2009 A very low carbohydrate ketogenic diet improves glucose tolerance in ob/ob mice independent of weight loss. Am J Physiol Endocrinol Metab 297:E1197–E1204 8. Zhang X, Yeung DC, Karpisek M, Stejskal D, Zhou ZG, Liu F, Wong RL, Chow WS, Tso AW, Lam KS, Xu A 2008 Serum FGF21 levels are increased in obesity and are independently associated with the metabolic syndrome in humans. Diabetes 57:1246 –1253 9. Fisher FM, Chui PC, Antonellis PJ, Bina HA, Kharitonenkov A, Flier JS, Maratos-Flier E 2010 Obesity is a fibroblast growth factor 21 (FGF21)-resistant state. Diabetes 59:2781–2789 10. Coskun T, Bina HA, Schneider MA, Dunbar JD, Hu CC, Chen Y, Moller DE, Kharitonenkov A 2008 Fibroblast growth factor 21 corrects obesity in mice. Endocrinology 149:6018 – 6027 11. Hale C, Chen MM, Stanislaus S, Chinookoswong N, Hager T, Wang M, Véniant MM, Xu J 2012 Lack of overt FGF21 resistance in two mouse models of obesity and insulin resistance. Endocrinology 153:69 – 80 12. Hirosumi J, Tuncman G, Chang L, Görgün CZ, Uysal KT, Maeda


13. 14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

K, Karin M, Hotamisligil GS 2002 A central role for JNK in obesity and insulin resistance. Nature 420:333–336 Hotamisligil GS 2010 Endoplasmic reticulum stress and the inflammatory basis of metabolic disease. Cell 140:900 –917 Díaz-Delfín J, Morales M, Caelles C 2007 Hypoglycemic action of thiazolidinediones/peroxisome proliferator-activated receptor ␥ by inhibition of the c-Jun NH2-terminal kinase pathway. Diabetes 56: 1865–1871 Wang H, Qiang L, Farmer SR 2008 Identification of a domain within peroxisome proliferator-activated receptor ␥ regulating expression of a group of genes containing fibroblast growth factor 21 that are selectively repressed by SIRT1 in adipocytes. Mol Cell Biol 28:188 –200 Díaz-Delfín J, Domingo P, Wabitsch M, Giralt M, Villarroya F 2012 HIV-1 Tat protein impairs adipogenesis and induces the expression and secretion of proinflammatory cytokines in human SGBS adipocytes. Antivir Ther 17:529 –540 Giralt A, Hondares E, Villena JA, Ribas F, Díaz-Delfín J, Giralt M, Iglesias R, Villarroya F 2011 Peroxisome proliferator-activated receptor-␥ coactivator-1␣ controls transcription of the Sirt3 gene, an essential component of the thermogenic brown adipocyte phenotype. J Biol Chem 286:16958 –16966 Hondares E, Rosell M, Gonzalez FJ, Giralt M, Iglesias R, Villarroya F 2010 Hepatic FGF21 expression is induced at birth via PPAR␣ in response to milk intake and contributes to thermogenic activation of neonatal brown fat. Cell Metab 11:206 –212 Wang CN, O’Brien L, Brindley DN 1998 Effects of cell-permeable ceramides and tumor necrosis factor-␣ on insulin signaling and glucose uptake in 3T3–L1 adipocytes. Diabetes 47:24 –31 Fernández-Veledo S, Vila-Bedmar R, Nieto-Vazquez I, Lorenzo M 2009 c-Jun N-terminal kinase 1/2 activation by tumor necrosis factor-␣ induces insulin resistance in human visceral but not subcutaneous adipocytes: reversal by liver X receptor agonists. J Clin Endocrinol Metab 94:3583–3593 Qi C, Pekala PH 1999 The influence of mRNA stability on glucose transporter (GLUT1) gene expression. Biochem Biophys Res Commun 263:265–269 Szalkowski D, White-Carrington S, Berger J, Zhang B 1995 Antidiabetic thiazolidinediones block the inhibitory effect of tumor necrosis factor-␣ on differentiation, insulin-stimulated glucose uptake, and gene expression in 3T3–L1 cells. Endocrinology 136:1474 – 1481 Xu L, Spinas GA, Niessen M 2010 ER stress in adipocytes inhibits insulin signaling, represses lipolysis, and alters the secretion of adipokines without inhibiting glucose transport. Horm Metab Res 42: 643– 651 Dutchak PA, Katafuchi T, Bookout AL, Choi JH, Yu RT, Mangelsdorf DJ, Kliewer SA 2012 Fibroblast growth factor-21 regulates PPAR␥ activity and the antidiabetic actions of thiazolidinediones. Cell 148:556 –567 Suzuki M, Uehara Y, Motomura-Matsuzaka K, Oki J, Koyama Y, Kimura M, Asada M, Komi-Kuramochi A, Oka S, Imamura T 2008 ␤Klotho is required for fibroblast growth factor (FGF) 21 signaling


26.

27.

28.

29.

30.

31.

32.

33.

34. 35. 36. 37.

38.

4245

through FGF receptor (FGFR) 1c and FGFR3c. Mol Endocrinol 22:1006 –1014 Kurosu H, Choi M, Ogawa Y, Dickson AS, Goetz R, Eliseenkova AV, Mohammadi M, Rosenblatt KP, Kliewer SA, Kuro-o M 2007 Tissue-specific expression of ␤Klotho and fibroblast growth factor (FGF) receptor isoforms determines metabolic activity of FGF19 and FGF21. J Biol Chem 282:26687–26695 Zhang B, Berger J, Hu E, Szalkowski D, White-Carrington S, Spiegelman BM, Moller DE 1996 Negative regulation of peroxisome proliferator-activated receptor-␥ gene expression contributes to the antiadipogenic effects of tumor necrosis factor-␣. Mol Endocrinol 10:1457–1466 Ruan H, Hacohen N, Golub TR, Van Parijs L, Lodish HF 2002 Tumor necrosis factor-␣ suppresses adipocyte-specific genes and activates expression of preadipocyte genes in 3T3–L1 adipocytes: nuclear factor-␬B activation by TNF-␣ is obligatory. Diabetes 51: 1319 –1336 Wilson-Fritch L, Nicoloro S, Chouinard M, Lazar MA, Chui PC, Leszyk J, Straubhaar J, Czech MP, Corvera S 2004 Mitochondrial remodeling in adipose tissue associated with obesity and treatment with rosiglitazone. J Clin Invest 114:1281–1289 Festuccia WT, Oztezcan S, Laplante M, Berthiaume M, Michel C, Dohgu S, Denis RG, Brito MN, Brito NA, Miller DS, Banks WA, Bartness TJ, Richard D, Deshaies Y 2008 Peroxisome proliferatoractivated receptor-␥-mediated positive energy balance in the rat is associated with reduced sympathetic drive to adipose tissues and thyroid status. Endocrinology 149:2121–2130 Yang KJ, Noh JR, Kim YH, Gang GT, Hwang JH, Yang SJ, Yeom YI, Lee CH 2010 Differential modulatory effects of rosiglitazone and pioglitazone on white adipose tissue in db/db mice. Life Sci 87:405– 410 Pardo R, Enguix N, Lasheras J, Feliu JE, Kralli A, Villena JA 2011 Rosiglitazone-induced mitochondrial biogenesis in white adipose tissue is independent of peroxisome proliferator-activated receptor ␥ coactivator-1␣. PLoS One 6:e26989 Yu XX, Murray SF, Watts L, Booten SL, Tokorcheck J, Monia BP, Bhanot S 2008 Reduction of JNK1 expression with antisense oligonucleotide improves adiposity in obese mice. Am J Physiol Endocrinol Metab 295:E436 –E445 Bennett BL, Satoh Y, Lewis AJ 2003 JNK: a new therapeutic target for diabetes. Curr Opin Pharmacol 3:420 – 425 Kharitonenkov A, Shanafelt AB 2008 Fibroblast growth factor-21 as a therapeutic agent for metabolic diseases. BioDrugs 22:37– 44 Kharitonenkov A, Larsen P 2011 FGF21 reloaded: challenges of a rapidly growing field. Trends Endocrinol Metab 22:81– 86 Lee MS, Choi SE, Ha ES, An SY, Kim TH, Han SJ, Kim HJ, Kim DJ, Kang Y, Lee KW 5 Mar 2012 Fibroblast growth factor-21 protects human skeletal muscle myotubes from palmitate-induced insulin resistance by inhibiting stress kinase and NF-␬B. Metabolism 10.1016/j.metabol 2012.01.012 Feingold KR, Grunfeld C, Heuer JG, Gupta A, Cramer M, Zhang T, Shigenaga JK, Patzek SM, Chan ZW, Moser A, Bina H, Kharitonenkov A 2012 FGF21 Is increased by inflammatory stimuli and protects leptin-deficient ob/ob mice from the toxicity of sepsis. Endocrinology 153:2689 –2700