Mifepristone Promotes Adiponectin Production and Improves Insulin Sensitivity in a Mouse Model of Diet-Induced-Obesity

The steroid receptor antagonist mifepristone is used as an anti-cancer agent, eliciting both cytostatic and cytotoxic effects on malignant cells. However, the metabolic effects of long-term treatment with mifepristone have remained unclear. The effects of mifepristone on insulin sensitivity and adiponectin secretion were evaluated both in in vivo and in vitro. First, we explored the effects of mifepristone, on metabolic functions in obese mice receiving a high-fat diet. When these mice were fed mifepristone, they exhibited a marked improvement in insulin sensitivity, attenuated hepatic injury, and decreased adipocyte size, compared with mice that received only the high-fat diet. Intriguingly, mifepristone-treated mice showed significantly elevated plasma adiponectin levels. Second, we tested the effects of mifepristone on differentiated 3T3-L1 adipocytes in vitro. When differentiated adipocytes were treated with mifepristone for 48 h, adiponectin was upregulated at both mRNA and protein levels. Collectively, these results reveal novel actions of mifepristone on metabolic functions, in vivo and in vitro, in which the drug exerts antidiabetic effects associated with an upregulation in adiponectin-secretion.


Introduction
In the last few decades, a significant amount of research has been dedicated to exploring the complex relationship between obesity and type 2 diabetes mellitus (T2DM) [1,2]. Although understanding this relationship is still a matter of intense study, clinical and basic science research efforts have led to the consensus that obesity can contribute to the development of several health complications, including T2DM, insulin resistance, dyslipidemia and hepatic steatosis. When occurring together, these medical disorders are termed as metabolic syndrome and this has been associated with an increased risk of developing cardiovascular disease and diabetes [3]. Indeed, the current perspective on the etiology of T2DM is that environmental factors, such as high calorie intake, in addition to the genetic abnormalities (i.e., leptin or leptin receptor abnormalities) [4], are considered a major risk factor in the development of obesity-induced T2DM.
Insulin resistance is a key predisposing factor in the pathogenesis of T2DM. Adiponectin is a peptide hormone secreted specifically from adipose tissue. Increasing evidence indicates that adiponectin plays a major role in attenuating insulin resistance. For example, plasma levels of adiponectin are suppressed in obese and diabetic mice compared with healthy littermates [5], as well as in obese humans compared with lean subjects [6]. Conversely, adiponectin supplementation decreases body adiposity and ameliorates insulin resistance that results from either genetic or environmental factors [7]. Pharmacological agents that promote secretion of adiponectin may be candidate therapeutic targets against metabolic disorders including T2DM.
In the early 80s, mifepristone was first introduced as a synthetic steroid with high affinity and antagonistic action at the glucocorticoid receptor (GR) [8]. Although approved as a GR antagonist, mifepristone has been clinically used as an agent for inducing abortion, both for its action as a progesterone receptor antagonist in the luteal phase of the cycle and for early abortion [9]. Recently, experimental studies have revealed that mifepristone treatment decreases blood glucose levels in genetically-modified diabetic mice, suggesting a role for the 11β-hydroxysteroid-dehydrogenase-1 signaling pathway in lowering blood glucose levels [10,11]. However, the mechanisms underlying mifepristone-induced improvement of insulin sensitivity remains less well defined.
Plasma adiponectin levels are also significantly lower in patients with breast cancer compared with age-and sexmatched healthy women, suggesting a possible association of decreased adiponectin levels with the growth and differentiation of breast epithelial cells and breast cancer cells [12]. These finding are compelling in view of the fact that mifepristone is also used as an anti-cancer agent [9], and suggest the existence of additional unidentified pleiotropic functions. In this study, we will provide evidence that mifepristone augments adiponectin secretion and improves multiple aspects of metabolic functions both in mice in vivo and in cultured adipocytes in vitro.

Reagents
Mifepristone and T0070907 were obtained from Cayman Chemical (Ellsworth Road, MI). GW9662 was from MERCK Chemical (Tokyo, Japan). Primer oligonucleotides for real-time quantitative RT-PCR assays (qRT-PCR) were synthesized by Operon Technologies (Tokyo, Japan). FBS was purchased from JRH bioscience (Lenexa, KS). Other reagents were obtained from Sigma (St. Louis, MO) unless otherwise noted.

Animals and Ethics Statement
All studies were approved by the Kagawa University Institutional Animal Care and Use Committee (IACUC). Sixweek-old C57BL/6NCr Slc male mice were purchased from SLC (Shizuoka, Japan). These mice were handled in compliance with the guidelines for conducting Animal Experiments at Kagawa University. All mice were housed one mouse per cage with temperature-and light-control (25°C and 12 h light/12 h dark cycle, respectively). Animals were provided with ad libitum access to a regular diet (RD) (Oriental MF diet, Oriental Yeast. Co., Tokyo, Japan) and water for a minimum of 1-week acclimation period before beginning the study. Mice were then divided into five groups and were maintained for 22 weeks on either an RD diet or high-fat diets (HFDs). Mice receiving HFDs were subdivided on the basis of orally administered mifepristone (no mifepristone, 0.1 mg/kg, 1.0 mg/kg, and 30 mg/kg). The maximum concentration of mifepristone in this study was chosen on the basis of an earlier report [13]. The HFD consisted of 35% of calories from fat (powdered beef tallow 15%/100 g and high oleic acid safflower oil 20%/100 g; HFD32, CLEA, Tokyo, Japan) We measured consumption of food, as well as body weight, every 7 days. Twenty-two weeks after treatment, an insulin tolerance test (ITT) was performed. Following a 7-day recovery period, blood samples were collected, all mice were euthanized, and organ weight was measured.
Liver and white adipose tissues (WAT) were isolated, immediately frozen in liquid nitrogen, and were stored at -80°C until RNA extraction. WAT was taken from the perirenal and epididymal regions. Tissue samples were homogenized using a Beads Cell Disrupter (Micro Smash MS-100, Tomys Seiko Co., Tokyo, Japan) and subjected to column-based RNA isolation procedure as below.

ITT
An ITT was performed between 1400 h and 1700 h in mice deprived of food for 6 h before the test. After intraperitoneal injection of human insulin (Humulin, Eli Lilly Japan K.K., Kobe, Japan) (0.75 U/kg body weight), blood samples were collected from the tail vein in series, and blood glucose levels were determined with the use of a Glutest sensor (Sanwakagaku-Arkray, Nagoya, Japan).

Hematoxylin-eosin staining
Each section from liver and perirenal adipose tissue was fixed in 4% (w/v) paraformaldehyde and embedded in paraffin. 4-µm sections were cut and processed for histopathological examination using hematoxylin and eosin stain. Micrographs were taken using a fluorescent microscope BioRevo BZ9000 (Keyence Japan) and analyzed by using the Keyence software, Dynamic Cell Count.

Blood sample analysis
Non-fasting serum levels of total serum and the highmolecular-weight (HMW) isoform of adiponectin were measured by enzyme immunoassay (ALPCO Diagnostics, Salem, NH). Serum aspartate aminotransferase (AST) levels were measured by using the Wako Transaminase CII Test (Wako Pure Chemical, Tokyo, Japan).

Cell culture and treatment with mifepristone
Mouse 3T3-L1 fibroblasts were obtained from JCRB Cell Bank (Osaka, Japan) and were cultured and induced to differentiate as we previously reported [14]. A stock solution of mifepristone was prepared in ethanol at a concentration of 10 mM. Cells were treated either with mifepristone (0.1, 1, or 10 µM) or with vehicle (ethanol).

Immunoblot analyses (culture medium)
The amount of adiponectin secreted into the culture medium by the 3T3-L1 cells was quantified by Western blot analysis. Cells were cultured in serum-free medium for 3 days. Medium was collected and treated with a same volume of 2x Laemmli sample buffer. The sample loading was based on volume normalization without further sample manipulation such as concentration or precipitation [15]. As controls, we utilized loading Prestained Molecular Weight Markers (cat.7720) (Cell Signaling Technology CST, Beverly, MA).
Each sample was separated by gel electrophoresis and transferred to nitrocellulose membranes (Amersham). Mouse adiponectin antibody (MAB3608) (Millipore, Billerica, MA) was used as the primary antibody. Immunolabeling was visualized using a species-specific horseradish peroxidase (HRP)conjugated IgG secondary antibody and an enhanced chemiluminescence reagent (SuperSignal West Pico Chemiluminescent Substrate, Pierce, Rockford, IL) using a LAS1000-Plus chemiluminescence image analyzer (Fujifilm Co, Tokyo, Japan).

[ 3 H]-2-deoxy-D-glucose uptake assay in differentiated 3T3-L1 adipocytes
Twelve days after induction of differentiation, mifepristone was added for an additional 3 days. Glucose uptake assay was essentially performed as previously described [16].

Statistical analysis
Results are expressed as mean ± standard error of mean. The statistical significance of the differences between mean values was determined with an ANOVA followed by Scheffe's F test for multiple groups (Figures 1, 2A, 2B, 3A, 3C, 4A, 4C, 5B, 6, 7, and 8C). ANOVA repeated-measures test was used for the adiponectin secretion analysis ( Figure 5A). In Figure 8B, unpaired Student's t-test was performed to analyze the statistical significance of the difference between the two groups. The threshold of significance was set at a p value of < 0.05.

Effects of mifepristone on HFD-induced insulin resistance in vivo
In order to investigate whether mifepristone influences peripheral insulin-sensitivity and adiponectin secretion, we first explored its effects in HFD-induced obese mice. To this end, we measured the body weight of 7-week-old C57BL/6NCr Slc mice just before the start of the feeding program and randomly assigned eight mice each into to one of the following five groups: RD, HFD alone, or HFD supplemented with 0.1, 1 or 30 mg/kg body weight/day of crude mifepristone. The increase in body weight observed in the HFD groups receiving mifepristone was comparable with that of the HFD alone group, indicating that drug treatment did not affect weight gain (Table  1). Consistent with this, energy intake remained unchanged in mice fed the HFD alone or the HFD treated with concentrations of crude mifepristone (Table 1). However, mifepristone treatment had an effect on fasting blood glucose levels in HFDinduced obese mice. Although fasting blood glucose levels were significantly higher in the HFD-induced obese mice than in the RD-fed mice (Table 1 and Table S1), the HFD mifepristone-treated groups exhibited significantly decreased glucose levels compared with the HFD-alone group (Table 1). Indeed, the fasting blood glucose level in the three HFD groups receiving concentrations of crude mifepristone was not significantly different from the levels observed in the RD mice (Table S1). An ITT revealed that the reduction in blood glucose levels after intraperitoneal administration of insulin was enhanced in all three mifepristone treated groups relative to the HFD alone group (Figure 1, left panel), and this was further supported by a significantly decreased glucose AUC (0-120) ( Figure 1, right panel).
We hypothesized that mifepristone augments responses to insulin in HFD mice by promoting adiponectin secretion. Moreover, because the HMW isoform of adiponectin exerts the most potent effects on blood glucose levels in mice [19], we compared the HMW and the total serum adiponectin levels measured by ELISA. As expected, the HFD-fed group had significantly decrease levels of circulating adiponectin compared with the age-matched RD group (6.7 and 21.9 µg/mL, respectively; Figure 2A). Significantly, treatment with mifepristone induced a recovery in total serum adiponectin levels in a dose-dependent manner ( Figure 2A). We also noted that HFD induced a decrease in the total-HMW adiponectin ratio. This decrease was attenuated by the 30 mg/kg bw/day dose of mifepristone ( Figure 2B). Indeed, similar to the changes induced in circulating adiponectin, treatment with mifepristone blunted the HFD-induced decrease in adiponectin mRNA production in the perirenal WAT ( Figure 3A). In addition, mifepristone upregulated mRNA expression of other molecules known to improve insulin sensitivity, such as aP2, FAS, leptin and PPARγ, in a dose-dependent manner without any effect on the housekeeping gene 18S rRNA in the perirenal WAT ( Figure  S1).
It has been reported that small-sized adipocytes secrete more insulin-sensitizing adipokines than hypertrophic adipocytes [20]. Therefore, we performed histological studies on the perirenal WAT, to evaluate the impact of mifepristone treatment on adipose tissue morphology. As shown in Figure  3B-a-e, fat pads from the 30-mg/kg bw/day mifepristone group contained a relatively uniform population of small adipocytes compared with the mixed distribution of large and small adipocyte sizes in WAT from HFD mice ( Figure 3B-b). Next, we calculated the average cross-sectional area of adipocytes in the perirenal WAT of mice across experimental groups. The mean adipocyte size of the mice fed the HFD supplemented with the three concentrations of crude mifepristone was significantly smaller than that of the mice fed only HFD ( Figure  3C, p < 0.01).
Obesity leads to liver injury characterized by steatohepatitis and ballooning degeneration [21], which can be attenuated by administering exogenous adiponectin [22]. We therefore assessed the degree of liver injury in HFD mice treated with or without mifepristone. Liver weight was markedly lower in mice treated with mifepristone at a dose of 30 mg/kg bw/day, despite being unchanged at lower doses ( Figure 4A and Figure S2). Histological examination indicated that mice fed the HFD alone exhibited marked ballooning degeneration of hepatocytes, characterized by the presence of numerous vacuoles within the hepatocyte cytoplasm ( Figure 4B-b). Interestingly, mifepristone decreased the number of cells undergoing ballooning degeneration in a dose-dependent manner ( Figure 4B-c-e). A comparable reduction of serum AST levels was also observed in mifepristone treated mice ( Figure 4C). Overall, these findings demonstrate that mifepristone enhances adiponectin secretion from mature adipocytes, which is associated with improvement in insulin sensitivity, amelioration of liver injury, and attenuation of adipocyte hypertrophy.

The effects of mifepristone in differentiated 3T3-L1 adipocytes in vitro
To gain insight into the mechanisms underlying the abovementioned effects of mifepristone in vivo, we explored the effects of mifepristone on adiponectin secretion from differentiated 3T3-L1 adipocytes cultured in serum-free medium. Mifepristone (1-µM) increased adiponectin secretion by 3T3-L1 adipocytes compared with cells treated with medium alone ( Figure 5A). Similarly, adiponectin mRNA expression was also increased by 1.3-or 1.4-fold after 3 days of treatment with 0.1-µM or 1-µM mifepristone, respectively ( Figure 5B, upper panel). The effect was also time-dependent; 2 days in vitro, or longer, were required to observe a significant increase in adiponectin mRNA levels ( Figure S3). The relative LDH release, i.e., the ratio of LDH released over total LDH, was <0.2% of the total for each incubation (Table S2). Subsequently, we assessed 2-deoxy-D-glucose transport activity in 3T3-L1 adipocytes preincubated with either 1-or 10-µM mifepristone, or vehicle ( Figure 6). Insulin caused a 1.9-fold increase in glucose uptake in untreated 3T3-L1 adipocytes, and a significantly greater increase in mifepristone-pretreated cells. Figure 6 summarizes how mifepristone promoted both basal and insulin-stimulated 2-deoxy-D-glucose uptake in these cells. Collectively these experiments show that mifepristone  Mice were fed a HFD and treated with mifepristone as described above. Perirenal adipose tissues were collected when the animals were sacrificed. Panel A shows the results of qRT-PCR assays in perirenal adipose tissue taken from mice with HFD-induced obesity treated with or without mifepristone. Total RNA was isolated and subjected to qRT-PCR using primers and probes directed to adiponectin. The expression level of each transcript was normalized to 18S. The results shown represent the pooled data. The fold increase in expression levels of the genes of interest are indicated for each mifepristone concentration, relative to the signals obtained in the HFD alone group. Each data point represents the mean ± S.E.M. derived from eight independent experiments. ** p < 0.01 versus HFD fed mice that did not receive mifepristone. Panel B shows the results of the histological analysis in perirenal adipose tissues stained with hematoxylin and eosin. Each subpanel demonstrates a representative image obtained from eight animals in each treatment condition. Panel C shows the results of adipose tissue depot quantification in these tissue samples. The adipocyte cross sectional area in each tissue depot was measured as described in the methods; mean values obtained from each mifepristone treatment group are summarized in the graph. ** p < 0.01 versus HFD fed mice that did not receive mifepristone.  . Protective effects of mifepristone on hepatic injury in HFD mice. At 29 weeks of age, mice were euthanized and liver weights were measured (Panel A). Panel B shows liver tissues from mice fed with a HFD and treated with mifepristone. Livers obtained from HFD mice that received various concentrations of mifepristone (or vehicle) as described in Figure 1 were stained with hematoxylin and eosin. Each subpanel demonstrates a representative image obtained from eight animals in each treatment condition. Panel C shows their serum AST levels at the time of sacrifice. * p < 0.05, ** p < 0.01 versus HFD fed mice that did not receive mifepristone. we sought to explore the functional association between adiponectin and the PPARγ pathway. We performed pharmacological experiments in which 3T3-L1 adipocytes were treated with inhibitors of PPARγ, such as T0070907 and GW9662 (up to 10 µM from Day 0 to Day 3 in adipocytes) prior to treating them with mifepristone. As expected, the PPARγ inhibitors markedly attenuated mifepristone-induced adiponectin secretion (Figure 7). Neither T0070907 nor GW9662 altered cell numbers, total protein recovery, or LDH release into culture medium of 3T3-L1 adipocytes after mifepristone stimulation (Table S2, Figure S4). Collectively, these results demonstrate that pharmacological inhibition of the PPARγ pathway attenuates mifepristone-induced adiponectin secretion from 3T3-L1 adipocytes.
As a complementary experiment to the pharmacological studies, we performed a genetic knockdown of PPARγ in differentiated adipocytes. We performed a transient transfection of siRNA against different target sequences of mouse PPARγ mRNA by electroporation. We first attempted to determine the effects of siRNA on endogenously expressed PPARγ protein 2 days after siRNA transfection using the nuclear/cytosol translocation kit. As shown in Figures 8A and  8B, transfection of PPARγ siRNA led to a significant 64 ± 5% reduction in PPARγ protein expression but not Lamine A/C expression, confirming the target specificity of the PPARγ siRNA. Transfection with siRNA specific to PPARγ in 3T3-L1 adipocytes cultured in medium alone did not produce a change in adiponectin release at 3 days in vitro, in comparison with 3T3-L1 adipocytes grown in medium alone, and transfected with control siRNA. This indicates that this PPARγ siRNA effectively attenuates mifepristone-induced adiponectin release by differentiated 3T3-L1 adipocytes ( Figure 8C). Together, these data indicate that transfection with siRNA targeting mouse PPARγ leads to specific knockdown of PPARγ protein expression as well as marked attenuation of adiponectin release induced by mifepristone stimulation. Off target effects are unlikely because similar effects were obtained in cells transfected with two other distinct siRNA oligonucleotides.

Discussion
We have provided experimental results demonstrating that mifepristone, a pharmacological agent previously known as a steroid analogue, promotes secretion of adiponectin both in an in vivo model of HFD-induced obesity and in an in vitro model of 3T3-L1 cells. Importantly, these effects of mifepristone are associated with improvement of several major metabolic Figure 7. Effects of mifepristone on adiponectin secretion levels in differentiated 3T3-L1 adipocytes. Shown are the results of immunoblot analyses in differentiated 3T3-L1 adipocytes. Cells were treated with mifepristone or vehicle at the indicated concentrations for 3 days. 3T3-L1 adipocytes were serum starved and treated with 1-µM mifepristone and/or 10-µM PPARγ inhibitors (T0070907 or GW9662). Equal volumes of culture medium were collected and analyzed by Western blot analysis using an antibody specific to adiponectin. Lower half panel shows the results of densitometric analysis, plotting the fold increase in the degree of adiponectin secretion levels after mifepristone addition. Bands corresponding to adiponectin were quantified by densitometry and were normalized to the signals obtained in the absence of mifepristone. Pooled data are summarized in the graphs in each assay set. * p < 0.05 versus for mifepristone compared with vehicle; † p < 0.05 versus for mifepristone treated cells.  parameters, including insulin tolerance, hepatic injury as well as adipose tissue hypertrophy in vivo and glucose uptake in vitro.
Our results demonstrate that increased adiponectin secretion from adipocytes is one of the predominant mechanisms underlying the effects mediated by mifepristone in vivo. Previous basic studies using genetically modified mice (ob/ob and db/db) have reported that short term intraperitoneal injection of mifepristone improved blood sugar levels in the early stages of obesity [10,23]. Subsequently, two clinical studies in healthy male subjects demonstrated that mifepristone significantly attenuated not only the side effect of weight gain caused by the antipsychotic medications olanzapine and risperidone [24,25], but also attenuated the increase in fasting plasma insulin and triglycerides levels caused by the risperidone [25]. Because obesity-induced T2DM is usually caused by an energy imbalance over a prolonged period of time, we first examined the effects of mifepristone in a more relevant long-term HFD-induced obesity mouse model. Mifepristone treatment did not influence energy intake, body weight gain, or body fat percentage (Table 1 and Figure S2). However, mifepristone produced significant protection against insulin resistance in a dose dependent manner (Table 1, Figures 1, 3A, 3B, 3C, 4A, 4B, and 4C). Simultaneous increases in gene expression and serum levels of adiponectin, a well-established antidiabetic adipokine [26] suggests that these mifepristone-mediated effects are in part attributable to an enhancement in adiponectin secretion from adipocytes (Table 1, Figures 2A, 2B, 3A, 3B, 3C, and Table  S1). Of note, the effects of mifepristone in the ITT are statistically more robust at the 30-mg/kg bw/day treatment group than at the 1 or 0.1 mg/kg (γ). This may be partly explained by the fact that the drug elevated HMW-to-total adiponectin levels at the 30 mg/kg bw/day treatment, only ( Figure 2B). It will be more important to rigorously determine the necessary amounts of mifepristone to improve metabolic parameters and/or adiponectin secretion using multiple experimental models. Glucose tolerance tests (GTT) in mifepristone-treated animals may provide further insights as well. It is now well established that insulin sensitizing drugs such as thiazolidinedione produce their effect through body-wide repartitioning of lipid storage and by altering signaling molecules such as adiponectin [27]. Notably, we have observed a significant reduction in hepatocyte ballooning and hepatic steatosis by mifepristone treatment (Figure 4B), suggesting enhanced lipid mobilization in this organ. Moreover, adiponectin has been shown to ameliorate insulin resistance through insulin receptor substrates (IRSs) in the liver [28]. Although we did not directly examine the effects of mifepristone on IRS in the current study, either at the level of protein expression or tyrosine phosphorylation, enhanced secretion of serum adiponectin from adipose tissues suggests that mifepristone exerts its effects through their remote activity [27]. However, further studies are required to understand the role of mifepristone in the liver exposed to a HFD.
The mechanisms underlying mifepristone-mediated improvement of insulin sensitivity remain to be fully elucidated. We have observed an enhanced expression of PPARγ, aP2, FAS, and leptin, which are genes known to promote lipid storage and lipogenesis [29] in adipose tissues of mifepristonetreated mice. Similarly, basal adiponectin secretion from differentiated 3T3-L1 adipocytes was significantly increased by the addition of mifepristone to the culture medium, which is in agreement with an enhancement of adiponectin mRNA production ( Figure 5B, upper panel). Note that the expression levels of IRS-1 mRNA remain unchanged in these cells following mifepristone stimulation ( Figure 5B, lower panel). Although IRS-1 signaling is regulated at multiple levels including tyrosine phosphorylation [30], we speculate that mifepristone modulates metabolic functions without altering the expression levels of IRS-1. Indeed, identification of the molecular mechanisms underlying the effects of mifepristone on metabolic functions remains to be resolved. We tested several other pharmacological agents chemically or pharmacologically related to mifepristone: GR antagonists, pregnenolone and DHEA; synthetic progestin causing abortion, levonorgestrel; MR antagonist, spironolactone [8,[31][32][33][34]. None of these drugs swere capable of promoting adiponectin secretion (Hashimoto et al. unpublished observation). We therefore propose that the ability to enhance adiponectin secretion is a function that is specific to mifepristone, rather than representing a class effect. Our data indicate that pharmacological (T0070907 and GW9662) as well as genetic (siRNA) inhibition of PPARγ is able to attenuate mifepristoneinduced adiponectin secretion by differentiated 3T3-L1 adipocytes in vitro, suggesting that mifepristone exerts its effects through the PPARγ pathway.
Moreover, a recent study has reported that mifepristone can bind the PPARγ ligand-binding pocket [35]. Interestingly, we have observed increased PPARγ expression along with its downstream target aP2 gene in both mice and cell culture studies ( Figure S1 and Hashimoto et al. unpublished observation). Although there is evidence that adiponectin can modulate cellular functions in both PPARγ-dependent andindependent fashions [36], we speculated that PPARγ might be involved in upregulating adiponectin, a molecule that when suppressed is linked to insulin resistance [37,38]. It is of note that an earlier study reported that the PPARγ agonist, pioglitazone improves HFD-induced insulin resistance and steatohepatitis in rodents, whereas body weight is still increased [39]..Consistent with this, our data indicate that mifepristone improves insulin sensitivity without affecting total body weight. Thus, mifepristone and classical PPARγ agonists may modulate insulin sensitivity and whole body weight in distinct manners. Unraveling the mechanisms underlying the effects of mifepristone on these and other adipocyte signaling pathways would clearly benefit our understanding of the pathophysiology of metabolic disorders and provide novel therapeutic pharmacological targets.
Furthermore, mifepristone augmented 2-deoxy-D-glucose uptake both in the basal and insulin-stimulated state of 3T3-L1 adipocytes ( Figure 6). A previous report has shown that the potency of these insulin-sensitizing effects is comparable with what is observed in adiponectin-treated cultured adipocytes [40]. It is important to mention that treatment with mifepristone does not lead to any observable variations in LDH release, cellular protein level or cell number (Table S2, Figures S5, and  S6). Collectively, these results support our hypothesis that both mRNA expression and secretion of adiponectin are induced by mifepristone in adipocytes.
Our in vivo and in vitro results show that treatment with mifepristone leads to an increase in the proportion of small adipocytes, which likely reflects enhanced adipocyte proliferation and/or suppressed adipocyte hypertrophy. A similar effect has been reported for thiazolidinedione, a PPARγ agonist, and its impact on adipocyte function is considered to be an additional mechanism by which this agent improves insulin sensitivity [41]. As a previous report has shown that thiazolidinedione enhances preadipocyte proliferation [42], the number of adipocytes in perirenal adipose tissue may be elevated by mifepristone. However, in our in vitro studies using differentiated 3T3-L1 adipocytes, treatment with mifepristone does not lead to any observable variations in the cell number detected by DAPI stained nuclei ( Figure S5). Moreover, mifepristone did not enhance or suppress preadipocyte proliferation of 3T3-L1 cells during adipocyte differentiation (Hashimoto et al. unpublished observation). We speculate that these seemingly contradictory findings stem from many differences in assay conditions between our in vivo and in vitro experiments; for example, the in vivo assays were performed over a much longer period than the in vitro assays (30 weeks versus 3 days). Thus, our data highlight the differential activity of mifepristone, which exerts anti-proliferative actions in cultured breast cancer cells [43,44], and effects on the metabolic actions in adipose tissues and adipocytes. It is therefore plausible that the effect of mifepristone on cell proliferation is highly cell-specific, remarkably reminiscent of the effect of thiazolidinedione [45].
It is of note that mifepristone is able to promote secretion of adiponectin from fully differentiated adipocytes ( Figure 5A and Figure S6). In these experiments we used serum-free medium because FBS contains endogenous circulating adiponectin [46], which makes it difficult to discriminate 3T3-L1 cell-derived molecules ( Figure S6). Note that serum-free medium and transfection of siRNAs did not alter LDH release, cellular protein levels, or the morphology of these cells (during 3-day) ( Figures S5, S7, and S8 as well as Table S2). Mifepristone, in contrast, modestly increased the secretion ability of adiponectin into the serum-free medium ( Figure 5A and Figure S6). Although, we focused on PPARγ in our experiments with siRNA, more detailed quantitative characterizations of both types of PPARγ, in terms of adiponectin-producing capacity as well as functional regulation of adipose tissues and differentiated adipocytes, remain to be elucidated. However, in both cases, mifepristone is capable of augmenting adiponectin secretion by cultured 3T3-L1 cells, as revealed by immunoblot analysis. When combined with our adiponectin data in vivo (Figures 2A and 3A) and qRT-PCR data in vitro ( Figure 5B), these data point to mifepristone as a novel agent that promotes adiponectin production from adipose tissues.
In conclusion, we have demonstrated that mifepristone induces expression of adiponectin in adipose tissue from mice with HFD-induced obesity as well as in differentiated mouse 3T3-L1 adipocytes, and that these are associated with improvements in several key metabolic parameters. We propose that a deeper understanding of the relationship between mifepristone and adiponectin signaling in adipose tissue may lead to the identification of novel therapeutic targets in obesity-induced T2DM along with other potentially related pathophysiological conditions, including breast cancer and metabolic diseases. Figure S1. Effects of mifepristone on perirenal adipose tissues in HFD mice. Mice had been fed with HFD and mifepristone as described above. Perirenal adipose tissues was collected when the animals were sacrificed. Panel shows the results of qRT-PCR assays in perirenal adipose tissue of HFD-induced obesity mice treated with or without mifepristone. Total RNA was isolated and subjected to qRT-PCR using primers and probes directed to aP2, FAS, leptin, and PPARγ. Expression level of each transcript was normalized to that of 18S. Shown are the results derived from pooled data, plotting the fold increase of the expression levels of indicated gene at the mifepristone-concentration indicated, relative to the signals obtained in the absence of mifepristone (HFD alone). Each data represents the mean ± S.E.M. derived from 8 independent experiments. * p < 0.05, ** p < 0.01 versus HFD fed mice that did not receive mifepristone. (PPT) Figure S2. Effects of mifepristone on percent of total body weight (Liver and Perirenal adipose tissue) in HFD mice. Tissue weights (liver and perirenal adipose tissues) in HFD induced obese mice are expressed as percentage of body weight. Each data represents the mean ± S.E.M. derived from 8 independent experiments. * p < 0.05 versus HFD fed mice that did not receive mifepristone.  Figure S8. Effects of mifepristone on total cellular protein level in matured adipocytes transfected with each siRNA. Shown are the results of total cellular protein. Cells were transfected with siRNA for control or PPARγ treated with/ without mifepristone. Total cellular proteins were extracted using a buffer containing n-octyl b-glucopyranoside. Protein determinations were made with Bicinchoninate Protein Assay Kit (Nacalai, Kyoto, Japan) with bovine serum albumin as a standard. Shown are the results derived from pooled data, relative to the values obtained in the absence of mifepristone (day 3). Each data represents the mean ± S.E.M. derived from 4 independent experiments. (PPT) Table S1. Parameters of regular diet mice. Seven-week-old C57BL/6NCr Slc mice received regular diet (RD) and were orally treated with either mifepristone (0.1, 1 or 30 mg/kg bw/ day) or vehicle (RD alone) for twenty-two consecutive weeks (n = 8 in each group). At the age of week 28, animals were subjected to fasting blood glucose test. * p < 0.05, ** p < 0.01 versus RD fed mice that did not receive mifepristone. (PPT) Table S2. LDH activity assay. LDH activity of culture medium was measured using a CytoTox-ONE Homogenous Membrane Integrity Assay kit (Promega), using DMEM and cell lysates extracted with a lysis buffer containing Triton-X (0.1% v/v) as negative and positive controls, respectively. (PPT)