Foxc2 enhances proliferation and inhibits apoptosis through activating Akt/mTORC1 signaling pathway in mouse preadipocytes.

Forkhead box C2 (Foxc2) protein is a transcription factor in regulation of development, metabolism, and immunology. However, the regulatory mechanisms of Foxc2 on proliferation and apoptosis of preadipocytes are unclear. In this study, we found that high-fat-diet-induced obesity elevated the expression of Foxc2 and cyclin E after 6 weeks. Additionally, Foxc2 suppressed preadipocyte differentiation, increased cell counts and augmented G1-S transition of preadipocytes, along with the elevation of cyclin E expression and the reduction levels of p27 and p53. Furthermore, Foxc2 knockdown reduced early apoptotic cells with accompanying reduction of mitochondrial membrane potential and increased fragmentation of genomic DNA. We show that Foxc2 reduces the expression of Bax, caspase-9, and caspase-3 in both serum-starved and palmitic acid-induced cell apoptotic models, which confirms the anti-apoptotic role of Foxc2. Moreover, the protein kinase B (Akt)/mammalian target of rapamycin (mTOR)C1 signaling pathway and the ERK/mTORC1 signaling pathway were activated along with preadipocyte proliferation in response to Foxc2 overexpression, whereas apoptosis marker genes were downregulated during this process. Those effects were blocked by the interference of Foxc2 or signal pathways specific inhibitors. These data collectively reveal that Foxc2 enhances proliferation of preadipocytes and inhibits apoptosis of preadipocytes by activating the Akt/mTORC1 and ERK/mTORC1 signaling pathways.

Then the adipose tissue was minced into fi ne sections (1 mm 3 ) with scissors and incubated in 10 ml of digestion buffer containing DMEM/F-12 (Gibco), 100 mM HEPES (Sigma), 1.5% BSA (Sigma), and 2 mg/ml type I collagenase (Sigma) for 50 min at 37°C in a water bath. After the incubation, growth medium [DMEM/F-12 (50:50)], 10% fetal bovine serum (Sigma), 100 U/ml penicillin, and 100 U/ml streptomycin were added to the digestion fl ask. Flask contents were mixed and fi ltered through nylon screens with 250 m and 20 m mesh openings to remove undigested tissue and large cell aggregates. The fi ltered cells were centrifuged at 300 g for 7 min at room temperature to separate fl oating adipocytes from cell pellets . Isolated cell pellets were suspended in DMEM/F12 (Invitrogen). Finally, cells were seeded into 35 mm primary culture dishes at a density of 8 × 10 4 cells/dish and incubated at 37°C under a humidifi ed atmosphere of 5% CO 2 and 95% air until confl uence. The medium was changed every other day.

Transfection of adipocyte with Foxc2 shRNA plasmid and Foxc2 overexpressed plasmid
Foxc2 expression plasmid vector pcDNA-Foxc2 was kept in our lab. shRNA sequence against Foxc2 was contrived and synthesized by GenePharma Company (Shanghai, China) using pGPU6/Neo shRNA expression vector, named shRNA-Foxc2. Two micrograms of interference or expression plasmid's DNA were mixed with 2 l X-tremeGENE HP reagent (Roche, Switzerland) and Opti-MEM I media (Invitrogen) and then added into the culture dish . The interference effi ciency and overexpression effi ciency were determined using Western blot 48 h after the transfection.

Cell proliferation assay
Cell proliferation was measured using Cell Counting Kit-8 (CCK-8) (Vazyme, China) assay. The transfected cells were fi rst seeded in a 96-well plate at a density of 2 × 10 3 for 12 h; then 10 l CCK-8 solution was added into each well and incubated further for 1 h at 37°C. Absorbance was quantifi ed at 450 nm using Vector 5 (Bio-Tech Instruments, USA).

Cell cycle assay
Cultured cells were harvested using trypsin/EDTA and washed twice with PBS buffer. Aliquots of 2 × 10 6 cells were centrifuged at 1,500 g for 5 min, precipitants of cells were fi xed in 70% ethanol and stained with 500 l propidium iodide (PI) solution (100 g/ml RNase and 50 g/ml PI in 1× PBS buffer). Percentages of cells within cell cycle compartments (G0/G1, S, and G2/M) were determined by BD FACScan (BD Biosciences, Franklin Lakes, NJ) and data were analyzed using Cell Quest software (BD Biosciences).

Apoptosis assessment
The TUNEL assay was performed using an in situ cell death detection kit (Roche, Switzerland). Mouse preadipocytes were grown to 60% confl uence on a 60 mm culture dish and the Foxc2 reconstructed plasmid was transfected into cells with X-treme GENE HP. After a 48 h transfection, the cell suspension was then added with 4% (v/v) paraformaldehyde in PBS buffer to fi x the cells for 30 min at room temperature. Cells were permeabilized with 0.1% Triton X-100 on ice for 2 min and then stained by TUNEL fl uorochrome. The areas of cells stained with TUNEL were measured using Image-Pro Plus analyzer (Palmerton, Inc.). Then images of cells were obtained using inverted fl uorescent microscopy .
Apoptosis was also determined by FITC-labeled annexin V/PI double staining and fl ow cytometry analysis. An annexin V-FITC apoptosis assay kit (Vazyme, Piscataway, NJ) was used according and delayed cell-cycle withdrawal and apoptosis, as well as inhibited myotube formation ( 15,16 ). However, there are few reports on the relationship between Foxc2 and apoptosis of cells. Some studies have found that overexpression of Foxc2 increased both cell viability and cell number, whereas interference of Foxc2 has opposite effects on MC3T3-E1 cells and primary caldaria cells of mice; Foxc2 interference induced an arrest of the G0/G1 phase of the cell cycle along with a reduction of phosphorylation of both protein kinase B (Akt) and mammalian target of rapamycin (mTOR) signaling pathways ( 17 ). The mTOR signaling cascades play an important role in both the survival and proliferation of tumor cells. Kim et al. ( 18 ) found that embelin suppressed the constitutive activation of the Akt/mTOR signaling cascade, which correlated with the induction of apoptosis in prostate cancer cells of humans. The Akt/mTOR signaling cascade has diversifi ed functions on cell proliferation; Akt is the key upstream molecule that links the ligation of growth factor receptors to the phosphorylation and activation state of mTOR ( 19,20 ). The ERK pathway is involved in various cell types by activating a series of protein kinases and, downstream, targets phosphorylation in cellular events such as cell proliferation, differentiation, survival, and motility ( 21,22 ).
This study aimed to investigate the regulatory mechanism of Foxc2 in the proliferation and apoptotic process on preadipocytes . To do so, we enhanced the Foxc2 gene expression in preadipocytes and found that the overexpression of Foxc2 enhanced cell proliferation and inhibited cell apoptosis, and then the Akt and ERK1/2 signaling pathways were activated, which in turn activated downstream mTORC1 pathway . The results imply that manipulating Foxc2 could offer a new theoretical means to prevent and treat obesity and type 2 diabetes.

Animal experiments
Two-week-old Kunming male mice were purchased from the Laboratory Animal Center of the Fourth Military Medical University. All mice were treated in accordance with the applicable guidelines and regulations approved by the Animal Ethics Committee of Northwest A&F University . They were allowed ad libitum access to water and standard chow laboratory diet for the fi rst 2 weeks to allow them to adjust to the new environment. Mice were subsequently randomly assigned to two groups: an HFD-fed group (87.5% chow diet + 10% lard + 2% cholesterol + 0.5% bile salt; Animal Center of the Fourth Military Medical University) or a chow diet-fed group (Animal Center of the Fourth Military Medical University) for the next 10 weeks. The animal room was maintained under controlled conditions of temperature at 25 ± 1°C, humidity at 55 ± 5%, and a 12 h light/12 h dark cycle. Body weight was recorded once a week.

Cell culture
Epididymal white adipose tissues were harvested, visible fi bers and blood vessels were removed, and the adipose tissue was washed three times with PBS buffer containing 200 U/ml penicillin (Sigma, St. Louis, MO) and 200 U/ml streptomycin (Sigma). during the 10 weeks of HFD feeding ( P < 0.05), while there was no difference in food intake ( Fig. 1A, B ). Foxc2 expression elevated during the last 4 weeks, accompanied by increased cyclin E mRNA expression ( P < 0.05) ( Fig.  1C, D ). To further analyze the function of Foxc2, we used forced gene expression detection in mouse preadipocytes . Compared with the control group, the expression of Foxc2 increased signifi cantly in the pcDNA-Foxc2 group ( P < 0.01), while it decreased in the shRNA-Foxc2 group ( P < 0.01) ( Fig. 1E ). We next analyzed the signifi cance of Foxc2 in preadipocyte differentiation. Figure 1F, G shows that forced expression of Foxc2 blocked preadipocyte differentiation along with decreasing the levels of PPAR ␥ and HSL . In addition, Foxc2 enhanced the rate of cell counts while serum-starved treatment did not alter the proliferation status of preadipocytes ( P < 0.05) ( Fig. 1H, I ). Taken together, these discoveries suggest that Foxc2 has impacts on preadipocyte proliferation.

Foxc2 enhanced preadipocyte proliferation and induced cell cycle transition
The CCK-8 cell proliferation assay demonstrated an enhanced rate of proliferation in the pcDNA-Foxc2 group compared with that in the control group and the shRNA-Foxc2 group ( Fig. 1H ). To determine whether Foxc2 regulated preadipocyte cell cycle, cells were subjected to cell-cycle analysis by fl uorescence-activated cell sorter. A lower percentage of pcDNA-Foxc2 cells underwent G1 arrest after Foxc2 forced expression compared with shRNA-Foxc2 cells ( Fig. 2A ). This fi nding was confi rmed by the decreased p27 level observed in pcDNA-Foxc2 cells ( Fig.  2B ). Furthermore, p53 functions to prevent initiation of DNA replication in the G1/S checkpoint and maintains G0/G1 arrest. Indeed, data from Western blots demonstrated that p53 ( Fig. 2B ) was downregulated in the forced expression Foxc2 group ( P < 0.05), while it was upregulated after Foxc2 stable knockdown in the normal growth medium . The slight decrease in G2 arrest that we observed in pcDNA-Foxc2 cells showed that a higher proportion of pcDNA-Foxc2 cells underwent a transition from G1-phase to S-phase ( Fig. 2A ). As expected, cyclin E expression increased markedly in the pcDNA-Foxc2 group ( P < 0.05) ( Fig. 2B ). The protein levels of p27, p53, and cyclin E showed the same results after being cultured in the serumstarved medium for 6 h ( Fig. 2C ), which confi rmed that starvation did not change the role of Foxc2 in regulating proliferation of preadipocytes.

Foxc2 inhibited preadipocyte apoptosis
Mitochondria play a vital role as gatekeepers of apoptosis. In addition, mitochondrial membrane potential is an important parameter of mitochondrial function and an indicator of apoptosis ( 23 ). JC-1 staining was used to detect the mitochondrial membrane potential after transfection with Foxc2. As shown in Fig. 3A , pcDNA-Foxc2 increased ( P < 0.01) red fl uorescence intensity. Flow cytometry analysis showed that the overexpression of Foxc2 increased ( P < 0.01) the ratio of the red/green light by 50%, while stable knockdown of Foxc2 decreased ( P < 0.01) to the manufacturer's protocol . Only fl uorescein-positive cells without PI staining were regarded as apoptotic cells, and the percentages were determined by fl ow cytometry (BD FACScan; BD Biosciences) and data were analyzed using Cell Quest software (BD Biosciences).

Mitochondrial membrane potential measurement
Fluorescent probe JC-1 (Beyotime, China) was used to estimate mitochondrial membrane potential ( ⌬ ⌿ m). Cells were incubated with 5 g/ml JC-1 at 37°C for 10 min, then washed twice with PBS and placed in fresh medium without serum. Images were scanned using a fl uorescence microscope (Nikon TE2000-U, Japan). Cells were also gently harvested with trypsin and transferred on ice for measurement using a fl ow cytometer. JC-1 was excited at 488 nm and the monomer signal (green) was recorded at 525 nm (JC-1 monomer) in a minimum of 10,000 cells per sample. Simultaneously, aggregate signal (red) was recorded at 590 nm (JC-1 aggregates). The ratio of red/green fl uorescent intensity was then calculated.

Statistical analysis
Statistical calculations were performed with SAS v8.0 (SAS Institute, Cary, NC). The effects of the treatments were determined using the one-way ANOVA procedure. Comparisons among the means of individual treatments were made by Fisher's least signifi cant difference (LSD) post hoc test once the ANOVA analysis showed a signifi cant effect of the treatments. Data are presented as mean ± SD, and P < 0.05 was considered to be signifi cant.

Foxc2 inhibited mouse preadipocyte differentiation and increased cell counts
To study the effects of HFD on Foxc2, we placed 2-weekold male mice on a HFD. Body weight increased signifi cantly the control group ( P < 0.05). This verifi ed the anti-apoptotic role of Foxc2. The expression of apoptosis marker genes, Bax, cleaved caspase-3, and cleaved caspase-9, were all decreased ( P < 0.01) in the pcDNA-Foxc2 group ( Fig. 3E ). Taken together, these fi ndings indicated that Foxc2 affects cell survival and attenuates cell apoptosis.

Foxc2 attenuated serum-starved and palmitate-induced preadipocyte apoptosis
Palmitic acid had been shown to induce apoptosis in adipocytes ( 8,(24)(25)(26). To verify the role of Foxc2 during the red/green ratio ( Fig. 3B ). Therefore, Foxc2 was capable of maintaining the mitochondrial membrane potential and preventing the cells from early apoptosis. Additional evidence for Foxc2 reduction of apoptosis was provided by annexin V/PI staining, demonstrating that forced expression of Foxc2 attenuated the rate of preadipocyte apoptosis ( Fig. 3D ). During the later stage of apoptosis, some kinds of DNase are activated and cut genome DNA in the nucleus into fragments that can be detected by TUNEL staining. As shown in Fig. 3C , Foxc2 knockdown resulted in a higher number of TUNEL-positive cells compared with Values are mean ± SD. * P < 0.05, ** P < 0.01 compared with the control. apoptosis ( Fig. 4B ). The reduction of serum-starved and palmitate-induced apoptosis in mouse preadipocytes further confi rmed that Foxc2 had a negative effect on cell apoptosis.

Foxc2 regulated proliferation and apoptosis of preadipocytes via the mTORC1 pathway
We next explored possible molecular mechanisms underlying Foxc2 regulation of proliferation and apoptosis this process, 0.2 mM palmitic acid was used to stimulate mouse preadipocytes for 24 h to trigger apoptosis. The treatment with palmitic acid increased the expression of Bcl-2 ( P < 0.05), but reduced expression of Bax ( P < 0.05), cleaved caspase-9 ( P < 0.05), and cleaved caspase-3 ( P < 0.05), as shown in Fig. 4A ; so palmitic acid induced cell apoptosis successfully. With the expression of Foxc2, cell apoptosis was attenuated signifi cantly ( Fig. 4A ). Similarly, expression of Foxc2 protected against starvation-induced of mTOR Ser2448 . To further examine the regulation of the upstream pathways of mTORC1, we co-treated preadipocytes with MK2206 and U0126. Forced expression of Foxc2 failed to affect the expression of raptor and p-mTOR Ser2448 ( Fig. 5E ). Because rapamycin treatment can inhibit mTORC1, we used rapamycin as the inhibitor of mTORC1 signaling to determine the role of mTORC1 in the regulation of Foxc2 on preadipocyte proliferation and apoptosis. Figure 5F, G indicates that Foxc2 markedly positively regulated raptor and p27 and negatively regulated caspase-9. Taken together, our data demonstrate that Foxc2 regulates preadipocyte proliferation and apoptosis via both the Akt/mTORC1 pathway and the ERK/mTORC1 pathway.

DISCUSSION
Obesity is determined by the number and volume of adipocytes in adipose tissue, which not only depends on intracellular triglyceride synthesis and catabolic effi ciency, but also on the balance between cell proliferation and death, including necrosis and apoptosis (30)(31)(32)(33)(34). Cell proliferation is the result of mitosis that is regulated by the cell cycle and requires a variety of regulatory mechanisms to genes, either positively or negatively, to coordinate each other (35)(36)(37). Cyclin E is a nuclear protein that can accelerate the cell cycle and enhance the ability of cell of preadipocytes. We fi rst determined the ratio of phosphorylated IRS-1 T896 , which showed a higher level in the Foxc2 forced expression group, along with elevated expression of phosphorylation (p)-Akt ser473 and p-mTOR Ser2448 ( Fig. 1A ), suggesting that the IRS-1 signaling pathway was involved and activated the Akt and mTOR pathways . Figure 1B shows that the expression of cyclin E and p27 were both elevated in the Foxc2 overexpressed group ( P < 0.05). To determine whether these effects were affected by Akt signaling, we treated preadipocytes with Akt phosphorylation inhibitor MK2206, and measured phosphorylated Akt Ser473 and p-mTOR Ser2448 . MK2206 is an allosteric inhibitor of Akt, and it could bind to the Akt kinase domain, which is an allosteric site, resulting in a conformational change, and then block the ATP binding site. Moreover, MK2206 had been used as a selective allosteric inhibitor of Akt phosphorylation (27)(28)(29). We found that MK2206 signifi cantly blocked phosphorylation of Akt Ser473 and decreased the expression of the cyclin E and p27 ( P < 0.05) ( Fig. 1A, B ). Foxc2 overexpression reversed this effect ( Fig. 5A ), while MK2206 had a minor effect on raptor and p-mTOR Ser2448 ( Fig. 5A ). Thus, we determined the levels of p-ERK1/2 and p-mTOR Ser2448 . As shown in Fig. 5C, D , Foxc2 increased the levels of p-ERK1/2, p-mTOR Ser2448 , and raptor signifi cantly, along with a higher expression of cyclin E, while the addition of ERK phosphorylated inhibitor U0126 blocked the phosphorylated ERK and largely reduced the level of raptor and the phosphorylation Cell counts by fl ow cytometry after staining with JC-1. The shift of green and red fl uorescence intensities were quantitated by subtracting the mean of treated cultures from the mean of the control. The ratio of red/green fl uorescence intensity was then used to quantitate the potential of mitochondrial membrane (n = 3). C: Images of preadipocyte apoptosis under fl uorescence microscope after TUNEL staining. Fluorescence intensity was used to quantitate the TUNEL-positive cells. Scale bar: 100 m (n = 3). D: Annexin V/PI staining analysis of apoptosis. The apoptotic rates were determined by fl ow cytometry (n = 3). E: Protein expressions of Bax, Bcl-2, caspase-3, caspase-9, cleaved caspase-3, cleaved caspase-9, and ␤ -actin. The expression level of total ␤ -actin was used as the loading control (n = 3). Values are mean ± SD. * P < 0.05, ** P < 0.01 compared with the control. caspase-9. We applied treatments of serum starvation and palmitic acid to activate caspase-3 and cleaved caspase-3, and then the Foxc2 interference reversed the treatment effects. By TUNEL and annexin-V/PI staining, we showed that the rate of apoptosis was attenuated in Foxc2-forced expression and increased in shRNA-Foxc2 group . These results all indicate that Foxc2 could inhibit preadipocyte apoptosis.
The Akt/mTOR signaling pathway has important functions in cell proliferation and apoptosis, and activation of the Akt/mTOR signaling pathway can promote adipocyte proliferation and inhibit apoptosis ( 19,20,(44)(45)(46). BSTA could suppress expression of Foxc2 and stimulate adipocyte differentiation, whereas Foxc2 could also promote osteoblast and colorectal cancer proliferation by activation of the Akt signaling pathway. It seems that Akt signaling plays an intricate role in Foxc2 function . In this study, we found that the expression of Foxc2 enhanced the phosphorylation of Akt and the phosphorylation of the Akt substrate, Foxo3a .
Another mechanism that has been implicated in the mechanical activation of mTOR signaling involves signaling by ERK1/2. For example, previous studies have demonstrated that ERK can induce mTOR signaling via the phosphorylation of raptor ( 21,(47)(48)(49)(50). In support of these data, we demonstrated that the expression of Foxc2 enhanced the phosphorylation of the ERK1/2 signaling pathway. Additionally, the mTORC1 path way key protein raptor, phosphorylation of mTOR, and proliferation, whereas p27 is a negative regulator of cell division, which binds to cyclin and then inhibits the biological activity of complex cyclin-dependent kinase in murine fi broblasts ( 38,39 ). Thus, manipulating the regulatory genes could alter proliferation and apoptosis of adipocytes, offering alternative approaches to prevent obesity. In this study, we found that a HFD signifi cantly induced Foxc2 and cyclin E expression. Foxc2 also inhibited preadipocyte differentiation and increased the cell counts, which is consistent with previous research. Additionally, Foxc2 promoted the expression of positive regulator cyclin E, while it inhibited the expressions of negative regulators p27 and p53. We verifi ed this effect further by cell cycle assay or in a serum-starved model, and obtained consistent results. Our fi ndings suggest that Foxc2 plays a positive regulatory role in cell proliferation by promoting the process of cell cycle transmission in preadipocytes.
Apoptosis of pancreatic ␤ cells induced by fatty acids involves the transcription factor FoxO1 ( 40 ), and Foxc1 and Foxc2 play pivotal roles in the early processes of heart development and cardiovascular development ( 10,(41)(42)(43). However, the relationship between Foxc2 and apoptosis of cells remains ambiguous. In this study, we found that shRNA-Foxc2 signifi cantly decreased the mitochondrial membrane potential and increased nuclear genomic DNA fragmentation, whereas Foxc2 expression reversed those effects. Moreover, Foxc2 showed reversal functions on the regulation of activities of genes involved in mitochondrial apoptosis, such as Bcl-2, Bax, caspase-9, and cleaved on proliferation promotion and apoptosis inhibition . All these results confi rm that Foxc2 functions necessarily via the mTORC1 signaling pathway, which is regulated by the upstream ERK1/2 pathway and partly by the Akt pathway .

CONCLUSION
In conclusion, this study provides a new insight into the regulatory mechanisms of Foxc2 on proliferation and apoptosis of preadipocytes. We further confi rmed that the effects of Foxc2 on the proliferation and apoptosis were via the Akt and ERK1/2 pathways and their downstream mTORC1 signaling pathway ( Fig. 6 ). Our results would phosphorylation of S6K1 were activated by Foxc2 and inhibited by rapamycin . The phosphorylation level was attenuated by MK-2206, a highly specifi c inhibitor to the Akt signaling pathway, and attenuated by rapamycin, a highly specifi c inhibitor to the mTORC1 signaling pathway. And the same data were detected for adipocytes treated with the specifi c inhibitor of ERK1/2 signal pathway U0126. Furthermore, when MK-2206 or U0126 was added, we also detected the expression of those genes regulating cell proliferation, such as cyclin E, p53, and p27, and apoptotic key factors caspase-3 and caspase-9. We anticipated that co-treatment with MK-2206 and U0126 would inhibit the mTORC1 signaling pathway, and that they should alleviate the effects of Foxc2 Foxc2 regulated proliferation and apoptosis of preadipocytes via the mTORC1 pathway. Mouse preadipocytes were pretreated with pcDNA-Foxc2, shRNA-Foxc2, MK-2206, U0126, and rapamycin, respectively. A: Representative immunoblots and densitometric quantifi cation for p-IRS1, p-Akt ser473 , p-FoxO3a, and p-mTOR Ser2448 (n = 3) with or without MK-2206. B: Representative immunoblots and densitometric quantifi cation for cyclin E, p53, and caspase-3 (n = 3) with or without MK-2206. C: Representative immunoblots and densitometric quantifi cation for p-ERK1/2, raptor, and p-mTOR Ser2448 (n = 3) with or without U0126. D: Representative immunoblots and densitometric quantifi cation for cyclin E and caspase-9 (n = 3) with or without U0126. E: Representative immunoblots and densitometric quantifi cation for p-Akt Ser473 , p-ERK1/2, raptor, and p-mTOR Ser2448 (n = 3) with or without MK-2206 or U0126. F: Representative immunoblots and densitometric quantifi cation for raptor, p-mTOR Ser2448 , and p-S6K1 Thr389 (n = 3) with or without rapamycin. G: Representative immunoblots and densitometric quantifi cation for p27 and caspase-9 (n = 3) with or without rapamycin. The level of total ␤ -actin was used as the loading control. Values are mean ± SD. * P < 0.05, ** P < 0.01 compared with the control. Fig. 6. Summary of the effects of Foxc2 on proliferation and apoptosis of adipocytes. Forced expression of Foxc2 promoted preadipocyte proliferation and inhibited preadipocyte apoptosis via the activation of the Akt/mTORC1 and ERK/mTORC1 signaling pathways.