Different p53 Genotypes Regulate PPARγ Post-Translational Modification in The Adipogenic Differentiation of Cancer Cells


 Background: Our previous studies confirmed that high concentrations of cobalt chloride (CoCl2) can induce the formation of polyploid giant cancer cells (PGCCs). PGCCs have the properties of cancer stem cell. In this study, we demonstrate that PGCCs can be induced to differentiate into adipose in vitro and in vivo. In addition, the molecular mechanism of adipogenic differentiation of PGCCs with daughter cells was investigated by detecting the expression of adipocyte differentiation related proteins in mutant and wild-type p53 cancer cell lines. Methods: HEY and MDA-MB-231 control cells and PGCCs with daughter cells were cultured with adipogenic differentiation medium and the cell cycle was detected by flow cytometry. The expression of adipocyte differentiation related proteins, and P300 histone acetyltransferase activity were compared before and after adipogenic differentiation. Immunoprecipitation was used to analyze the post-translational modification of peroxisome proliferator-activator receptor-γ (PPARγ) and P53 in HEY and MDA-MB-231 control cells and PGCCs with daughter cells cultured with adipogenic differentiation medium. Animal xenograft models were used to study the adipogenic differentiation of PGCCs with daughter cells.Results: Dexamethasone, rosiglitazone, insulin, and 3-isobutyl-1-methylxanthine (IBMX) can force the trans-differentiation of PGCCs into post-mitotic and functional adipocytes. Activation of PPARγ is a critical step in the process of adipogenic differentiation. The expression levels of p-CREBser133, PPARγ, C/EBPα, C/EBPβ were higher in PGCCs with daughter cells after adipogenic differentiation compared with those without adipogenic differentiation in HEY and MDA-MB-231. P53 regulates the acetylation and phosphorylation of PPARγ and the expression of P300- Acetyl-PPARγ(Lys)- FABP4 and ERK- Phospho-PPARγ (Ser112)-FABP4 depended upon the genotype of p53 in HEY and MDA-MB-231 cells after cultured with adipogenic differentiation medium. The invasion and migration abilities of PGCCs with their daughter cells after adipogenic differentiation decreased compared with those cells without adipogenic differentiation.Conclusion: P300-P53-ERK-CREB-PPARγ-CEBPα/β-FABP4 pathway may participate in the adipogenic differentiation of HEY, MDA-MB-231 PGCCs with daughter cells, which associated with genotype of p53.

Introduction Adipogenic differentiation of human mesenchymal stem cells (MSCs) can be induced by 3-isobutyl-1methylxanthine (IBMX), dexamethasone, and insulin (this being the typical cocktail) [1]. Adipocytes can be derived not only from preadipocytes and pluripotent MSCs but also from cancer stem cells (CSCs).
One of the treatments for malignant tumors is induced differentiation therapy-the application of chemicals that can differentiate malignant cells into normal cells. In dedifferentiated tumor cells, differentiation into normal cells is commonly induced using the following differentiation agents: retinoic acid (RA) [2], histone inhibitors, methylation inhibitors [3], peroxisome proliferator-activated receptor (PPAR) agonists [4,5], or hydroxymethylglutaryl coenzyme A (CoA) reductase inhibitors [6]. Welldifferentiated liposarcoma (WDLPS) and dedifferentiated liposarcoma (DDLPS) cells can also be differentiated into adipocytes using dexamethasone, indomethacin, insulin, and IBMX. These compounds induce adipogenesis by upregulating the transcription and translation of genes involved in maintaining cancer cell stemness and adipogenic differentiation [7].
Surgical resection combined with radiotherapy and chemotherapy is the most common treatment for malignant solid tumors. However, for solid tumors with high malignancy, metastasis and recurrence can occur after radiotherapy and chemotherapy. We previously reported that radiation and chemotherapy drugs induced the formation of polyploid giant cancer cells (PGCCs) and that the daughter cells derived from PGCCs via asymmetric division had strong invasion and in ltration abilities [8,9]. PGCCs (which had the properties of CSCs) and their daughter cells were positive for multiple normal and CSC markers.
PGCCs were more tumorigenic than regular-differentiated cells in nude mice, possessed epithelialmesenchymal transition (EMT) phenotype changes, and could be induced into multiple benign lineages such as adipocytes, bone, and cartilage [10]. Lung cancer NCI-H446 cells can be induced to differentiate into neurons, adipocytes, and bone cells in vitro [11]; cancer cells with homologous recombination defects, such as ovarian and breast cancer cells with BRCA1/2 mutations, can be induced to differentiate using poly ADP-ribose polymerase (PARP) inhibitors [12]; and thyroid cancer cells expressing the PPARγ fusion protein (PPFP) can be induced to differentiate into adipocytes using pioglitazone [13].
In general, stemness is higher in cancer cells with high malignancy, which induces differentiation more easily [14]. In this study, we used CoCl 2 -treated HEY and MDA-MB-231 PGCCs with their daughter cells as a cell model with high metastatic ability. We con rmed that PGCCs are necessary for cancer dissemination but can be directly targeted and inhibited by a trans-differentiation approach, such as forced adipogenesis. We also highlight the crucial transcription factors that induce adipogenesis in HEY and MDA-MB-231 cells, including the well-studied PPARγ and CCAAT/enhancer-binding proteins (C/EBPs) that have recently been shown to play an important role in adipocyte differentiation [15]. Recent studies reported that the p53 gene deletion type of broblast-derived cancer cells can be induced to differentiate into adipocytes [16]. Previously, we showed that HEY cells express the wild-type p53 (p53 +/+) and MDA-MB-231 cells express the mutated p53 (p53 +/-) [17].
After induced differentiation, different p53 genotypes have different levels of FABP4 expression. The wildtype p53 shows higher levels of FABP4 expression and a lower degree of differentiation than that the mutant type. FABP4 negatively regulates the adipogenic differentiation of tumor cells. In PGCCs, adipogenic differentiation is negatively regulated by the P300 ubiquitination of the wild-type p53 and positively regulated by the P300 acetylation of the mutant p53. To understand the complex regulatory mechanisms in malignant tumors that can be transdifferentiated into mature adipocytes, we focused on different p53 genotypes that regulate the post-translational modi cation of PPARγ in the adipogenic differentiation of cancer cells. P300 histone acetyltransferase activity assay P300 histone acetyltransferase (HAT) activity assay was performed using the P300 HAT spectrophotometry quantitative detection kit (GenMed Scienti c Inc., USA) to test the differences in P300 HAT activity before and after adipogenic differentiation in vitro. Histone H3 peptide was used as the substrate for detection. HEY and MDA-MB-231 cells were treated with sodium butyrate for 48 h. Nuclear extracts were harvested as described above and subjected to the assays. Brie y, each nuclear extract was incubated with 100 µM acetyl-CoA and 1× HAT assay buffer on a histone H3-precoated enzyme-linked immunosorbent assay plate for 30 min. P300 transfers the acetyl group of acetyl-CoA to the histone peptide, releasing sulfhydryl-CoA, which then reacts with Ellman's reagent to produce a change in the absorbance peak. After several washes with PBS, acetylated histones were detected using the HAT assay kit according to the manufacturer's protocol. ExactiveTM Plus (Thermo) coupled online to an ultra-performance liquid chromatography system for the acquisition of MS/MS data. The peptides were identi ed and quanti ed using Proteome Discoverer 1.3. The peptide con dence was set at a high value, and the peptide ion score was set at >20.

Immunocytochemical (ICC) and immunohistochemical (IHC) staining
For ICC staining, cells grown on coverslips were washed with sterile PBS, then xed with cold methanol for 30 min. After treating with 0.3% endogenous peroxidase inhibitor (Zhongshan Inc., Beijing, China) for 15 min, the cells were incubated with 1.5% normal goat serum (Zhongshan Inc., Beijing, China) for 20 min to block non-speci c protein binding. After incubation with primary antibodies at 4 °C overnight, biotinylated goat anti-mouse/rabbit IgG (Zhongshan Inc., Beijing, China) and horseradish peroxidaselabeled streptomycin (Zhongshan Inc., Beijing, China) were added to the slides for 20 and 15 min, respectively. For IHC staining, para n-embedded tissue sections were depara nized in xylene and dehydrated in a series of gradient alcohol solutions. Antigen retrieval was carried out by adding citrate buffer solution (Origene, Wuxi, China) at 160 °C for 1-2 min. After the sections were incubated with primary antibodies and reacted with biotinylated goat anti-rabbit IgG antibody for 20 min, the signal was detected using the labeled streptavidin-biotin system in the presence of the chromogen 3,3diaminobenzidine or alkaline phosphatase. Nuclei were counterstained with hematoxylin.

Wound healing assay
Control cells and PGCCs with daughter cells before and after adipogenic differentiation were seeded into 12-well plates (1×10 5 per well, three replicate wells per group) and cultured until they reached 100% con uence. Sterile pipette tips were then used to uniformly scratch the monolayer cells vertically to form wound tracks. After rinsing with PBS to remove oating cells, the wells were cultured in serum-free medium. The migration ability was evaluated by photographing the wound area at 0, 12, and 24 h for HEY and at 0, 16, and 32 h for MDA-MB-231 cells at the same scratch position. ImageJ software was used to outline the migration area and calculate the wound-healing index according to the following formula: [(wound area at 0 h) -(wound area at indicated time)]/(wound area at 0 h). A high score indicated strong migration ability.

Cell migration and invasion assays
Control cells and PGCCs with daughter cells before and after adipogenic differentiation were washed three times with FBS-free medium and counted using an automated cell counter (Invitrogen, CA, USA). Cell migration and invasion were assessed using transwell migration and invasion assays (8 μm; Corning Plate clone formation assay Control cells and PGCCs with daughter cells before and after adipogenic differentiation were counted. Cell suspensions (2 mL/well) with 30, 60, and 120 cells were cultured in a 12-well plate, and the plates were incubated for at least 2 weeks at 37 °C. Incubation was stopped when a white cell clone was visible.
The cell clones were washed with PBS and xed with cold methanol for 30 min. After staining with 0.1% crystal violet for 30 min, the number of cell clone groups per well was counted under a microscope (the number of cells in a single clone should be more than 50), and the e ciency of colony formation was calculated using the following formula: formation e ciency = (number of clones/number of cells inoculated).
Animal Experiments

H&E staining
Tumor sections were xed in formalin for 24 h at room temperature and embedded in para n, then 4-µmthick sections were made. The tissue sections were subsequently depara nized in xylene for 12 h at 75°C and rehydrated using a descending ethanol series. Sections were stained with 0.2% hematoxylin (Baso, Zhuhai, Guangzhou, China) at room temperature for 1 min and with 0.5% eosin for 2 min. After staining, the sections were dehydrated and mounted on coverslips.

Statistical Analysis
Data are presented as the mean ± SEM. Statistical analysis and graphs were generated using SPSS 22 software (SPSS Inc., Chicago, USA) and GraphPad Prism software. Statistical analyses were performed as indicated in the gure legends. In vitro studies were biologically repeated at least three times, and statistical analysis was performed using ANOVA and unpaired t-test. The number of animals in each experiment is indicated in the gure legends. In vivo statistical analysis was performed using the Kruskal-Wallis test corrected for multiple comparisons. Statistical signi cance was set at P<0.05. Error bars represent ± S.D., and p-values represent comparisons with each control (***P < 0.001; **P < 0.005; * P < 0.05; NS, non-signi cant). ORO staining in a 6-well plate also showed that after adipogenic differentiation, intracellular lipid accumulation increased more with time in the PGCCs with daughter cells than in the controls ( Figure 1F). The statistical results are shown in Figure 1G.

In ltration and invasion abilities in control cells and PGCCs with daughter cells before and after adipogenic differentiation
Cell invasion assays were performed using Matrigel-coated transwell inserts. The invasive (Figures S1A-S1B) and migratory abilities (Figures S1C-S1D) in cells cultured in adipogenic differentiation media were lower than those in cells cultured in complete media. Statistical analysis con rmed this result. PGCCs with daughter cells in adipogenic differentiation media had the lowest number of invasive and migratory cells among all the cell groups ( Figure S1E). Migration ability was also measured by wound-healing assay in HEY and MDA-MB-231 control and PGCCs with daughter cells before and after adipogenic differentiation. Without adipogenic differentiation, the migration ability of PGCCs with their daughter cells was stronger than those of the controls. However, migration abilities decreased in both the control and PGCCs with daughter cells after adipogenic differentiation (Figures S1F-S1G). A plate cloning assay was used to detect the cell proliferative ability. The numbers of clones formed in HEY and MDA-MB-231 PGCCs with daughter cells after adipogenic differentiation were signi cantly lower than those of the corresponding cells without adipogenic differentiation (Figures S1H-S1I). Migration, invasion, and proliferation abilities decreased when cells were cultured in adipogenic differentiation medium.
Flow cytometric analysis of control and PGCCs with daughter cells before and after adipogenic differentiation HEY and MDA-MB-231 control and PGCCs with daughter cells were cultured in adipogenic differentiation media for 168 h and used for DNA content and cell cycle analysis. Cell proliferation and differentiation are regulated by the cell cycle, which is divided into a series of phases based on the chromosome content of the cell. We previously reported that the cell cycle was arrested at the G2/M phase in PGCCs with daughter cells [17].  (Figures 2A and 2C). The proportion of cells in the S and G0/G1 phases in HEY and MDA-MB-231 control cells were higher after adipogenic differentiation than before adipogenic differentiation ( Figures 2B and 2D), indicating that these cells exhibited limited growth potential after adipogenic differentiation. The statistical results are shown in Figure 2E.
Expression of adipogenic differentiation-related proteins WB was performed to compare the expression of adipogenic differentiation-related proteins in control cells and PGCCs with daughter cells before and after adipogenic differentiation. The total CREB expression was lowest in PGCCs with daughter cells after adipogenic differentiation. CREB is activated by phosphorylation at serine 133 after inducing differentiation. On the other hand, the expression levels of phospho-CREB(Ser133), PPARγ, C/EBPα, C/EBPβ, and FABP4 were highest in PGCCs with daughter cells after adipogenic differentiation ( Figures 2F and 2G). The results of qPCR showed that the expression levels of C/EBPα, PPARγ, and FABP4 mRNA were higher in HEY and MDA-MB-231 control and PGCCs with daughter cells after adipogenic differentiation compared with those in the corresponding cells before adipogenic differentiation, and the differences for C/EBPα (P=0.  Figure 2H) were statistically signi cant.

Different p53 genotype regulated FABP4 expression
We previously reported that HEY cells have the wild-type p53 (p53 +/+) and MDA-MB-231 cells have the mutated p53 (p53 +/-) [17]. Figure 3A-a shows that the expression level of FABP4 in HEY cells increased after adipogenic differentiation. There were no obvious changes in FABP4 expression in MDA-MB-231 cells before and after adipogenic differentiation ( Figure 3A-b). The statistical results are shown in Figures   3A-c and -d. ICC staining showed that nuclear expression of FABP4 was higher in all cell groups after adipogenic differentiation (Figures 3B-a, -c and 3C-a, -c) than in the corresponding cells without adipogenic differentiation (Figures 3B-b, -d and 3C-b, -d). qPCR results also con rmed that FABP4 mRNA expression increased in both HEY and MDA-MB-231 cells after adipogenic differentiation compared with those in the corresponding cells without adipogenic differentiation ( Figures 2H-c, -f) and phospho-ERK(Thr202/Tyr204) expression slightly decreased ( Figure 5B). ORO staining also showed that the degree of differentiation in HEY control cells was higher after P53 knockdown ( Figures 5E-a, -b).
The statistical results are shown in Figure 5E-e. When blocking the wild-type p53 (p53 +/+) in MDA-MB-231 PGCCs with daughter cells undergoing differentiation ( Figure 5C) and in MDA-MB-231 control cells ( Figure 5D) after adipogenic differentiation, PPARγ expression was downregulated with a decrease in phospho-CREB(Ser133) compared with the negative groups, while phospho-PPARγ(Ser112) and phospho-ERK(Thr202/Tyr204) expression were signi cantly upregulated. ORO staining showed that the degree of differentiation in MDA-MB-231 control (Figures 5F-a, -b) and PGCCs with daughter cells (Figures 5F-c, -d) were higher after P53 knockdown. The statistical results are shown in Figure 5F-e.

Acetylation of PPARγ in HEY and MDA-MB-231 before and after adipogenic differentiation
PPARγ is regulated by a series of post-translational modi cations [20]. The effect of acetylation on PPARγ is mainly mediated by cell senescence and regulation of adipocyte metabolism. Many studies have con rmed that the deacetylation of PPARγ by SIRT1 can downregulate the expression of PPARγ and inhibit adipogenic differentiation [21]. PPARγ acetylation promotes lipid synthesis. When the conserved lysine motif (K154/155) of PPARγ1 is acetylated, it can promote lipid synthesis in ErbB2-positive breast cancer cells [22]. IP combined with WB was used to detect the post-translational acetylation level of PPARγ in HEY and MDA-MB-231 control and PGCCs with daughter cells before and after adipogenic differentiation. Using PPARγ IP-level antibodies, cell pellets were collected before and after adipogenic differentiation for IP, as shown in Figures 6A and 6B. We successfully pulled down the PPARγ protein after adipogenic differentiation of HEY control and PGCCs with daughter cells, and we used WB to detect broad-spectrum acetylation at the lysine site. We found that after adipogenic differentiation, the PPARγ lysine site was acetylated, and the level of modi cation in HEY control and PGCCs with daughter cells was higher after adipogenic differentiation (Figures 6C and 6D). PPARγ protein underwent acetylation during adipogenic differentiation, which was also observed in MDA-MB-231 control and PGCCs with daughter cells (Figures 6E-6H). The acetylation level of PPARγ increased most signi cantly after adipogenic differentiation of MDA-MB-231 PGCCs with daughter cells (Figures 6F and 6H). P300 expression increased after HEY and MDA-MB-231 adipogenic differentiation ( Figure 6I), and the difference was statistically signi cant ( Figure 6J). In vitro HAT assay was used to detect P300 acetyltransferase activity and showed that HEY and MDA-MB-231 control and PGCCs with daughter cells had increased P300 activity after adipogenic differentiation ( Figure 6K).
P300-P53 regulate the acetylation of PPARγ in adipogenic differentiation of HEY and MDA-MB-231 cells P300 is an HAT that activates genes by attaching acetyl groups to histones. Mutual regulation between P300 and P53 has previously been reported [23]. When P300 acetylates P53, P53 becomes activated and positively regulates P300. At the same time, P300 can regulate the ubiquitin-like degradation of P53. The acetylation and ubiquitin-like modi cation of P53 occur at the same lysine site [24] and show a competitive relationship. In the present study, P53 and P300 were immunoprecipitated, and WB was used to analyze their interaction with PPARγ in the adipogenic differentiation of HEY and MDA-MB-231 cells.
The results showed an interaction among these three proteins (Figure 7).
We detected the expression of P300 when P53 was knocked down. Transfection with siRNA to inhibit the expression of P53 ( Figure 8A) showed that P53 knockdown markedly attenuated P300 expression levels in both HEY and MDA-MB-231 PGCCs (Figures 8B-8E). In HEY cells undergoing adipogenic differentiation, blocking wild-type p53 (p53 +/+) eliminated P300 expression, whereas in HEY PGCCs with daughter cells, the acetylation of PPARγ was upregulated compared with the negative groups ( Figure 8B). In HEY control cells after adipogenic differentiation, the acetylation of PPARγ showed no signi cant change compared to the NC cells ( Figure 8C). HEY cells that express wild-type p53 (p53 +/+) have a negative regulatory relationship between P300 and PPARγ acetylation levels after adipogenic differentiation. MDA-MB-231 cells that express mutation-type p53 (p53 +/-) have a positive regulatory relationship between P300 and PPARγ acetylation levels after adipogenic differentiation. As shown in Figures 8D-8E, blocking mutation P53 expression in MDA-MB-231 control and PGCCs with daughter cells eliminated P300 expression, and in adipogenic differentiation, the acetylation of PPARγ decreased compared with that of the NC groups. The results and detailed statistics are shown in Fig 8F. A485 is a potent and selective catalytic inhibitor of p300/CBP [18]. The effect of A485 on the viability of HEY and MDA-MB-231 control and PGCCs with daughter cells was analyzed using a CCK8 assay ( Figure   9F). Based on these results, 24 h pre-treatment with 10 μM A485 was chosen for further studies. As shown in Figures 9A-9D, the acetylation of PPARγ and of P53 at Lys-382 increased in HEY and MDA-MB-231 cells after adipogenic differentiation. However, SUMO-P53 levels also increased with the increase in P300 after adipogenic differentiation ( Figures 9A-9B), suggesting that P300 may be a negative factor for wild-type p53 (p53 +/+) in HEY cells. IP and WB analyses showed a decrease in PPARγ acetylation levels, accompanied by an increase in P53 acetylation levels in HEY cells ( Figures 9A-9B and 9E-a, -b, -e, -f) before and after adipogenic differentiation with the P300 inhibitor A485. PPARγ acetylation was negatively modulated by wild-type P53 (+/+) in HEY cells ( Figures 9A-9B), forming a P300-P53 (+/+)-ace-PPARγ negative feedback loop, and blocking P300 caused wild-type P53 expression and an increase in ace-PPARγ expression ( Figures 9A-9B and 9E-a, -b, -e, -f). IP and WB further showed that A485 treatment led to a decrease in the acetylation levels of PPARγ and P53 in MDA-MB-231 cells (Figures 9C-9D and 9Ec, -d, -g, -h) before and after adipogenic differentiation. PPARγ acetylation was positively modulated by mutation type P53 (+/-) in MDA-MB-231 cells (Figures 9C-9D), forming a P300-P53 (+/-)-ace-PPARγ positive feedback loop, and blocking mutated P53 decreased the expression of ace-PPARγ ( Figures 9C-9D and 9E-c, -d).

Adipogenic differentiation of cancer cells in vivo
To translate the in vitro ndings in vivo, xenograft ovarian and breast cancer models were established by subcutaneous injection of HEY ( Figure 10) and MDA-MB-231 cells (Figure 11) into BALB/c nude mice. The mice were divided into four treatment groups: (i) control cells, (ii) control cells after adipogenic differentiation, (iii) PGCCs, and (iv) PGCCs after adipogenic differentiation. On day 30, all mice were sacri ced, and the xenograft tumors were removed and analyzed. The xenograft tumor model exhibited a change in tumor volume before and after differentiation ( Figures 10B and 11B). The average volumes of xenograft tumors were signi cantly smaller in the in vitro-differentiated groups than in the undifferentiated (UD) groups ( Figures 10A and 11A). Tumor growth curves show that tumors in the in vitro-differentiated group grew markedly slower than those in the UD groups ( Figures 10C and 11C). H&E staining was performed to visualize morphology, and histological analysis revealed dense cancer cells in the UD groups. Compared with the UD groups, the in vitro-differentiated groups had more lipid droplets distributed in their tissues and had cells that became round and loosely bound ( Figures 10E and 11E). Adipogenic differentiation was con rmed by FABP4 expression (Figures 10D and 11D) and IHC staining ( Figures 10H and 11H). Immunostaining was performed to detect the human-derived marker vimentin and con rm the human origin of these cells (Figures 10F and 11F). Notably, when analyzing the Ki-67 IHC staining ( Figures 10G and 11G), we observed that in vitro adipogenic differentiation exhibited a tumor in ltration zone with a benign, segmental differentiation morphology. The in vitro differentiation model revealed that cancer cells successfully differentiated into mature adipocyte lineages under induction of adipocyte differentiation.

Discussion
In this study, we reported the adipogenic differentiation of cancer cells. PGCCs undergo EMT to obtain a mesenchymal cell phenotype and stem cell characteristics, which promotes the trans-differentiation of cancer cells. We explored the relevant mechanisms of PGCC adipogenic differentiation. The rst part of the experiment con rmed that HEY and MDA-MB-231 cells were successfully induced to differentiate in vitro. The second part of the experiment con rmed that the P53-ERK-CREB-PPARγ-CEBPα/β-FABP4 pathway regulates the adipogenic differentiation process of HEY and MDA-MB-231 control and PGCCs with daughter cells. After adipogenic differentiation, PPARγ is regulated by post-translational modi cations. The third part of the experiment con rmed that the adipocytes produced by PGCCs after adipogenic differentiation are mature adipocytes and will no longer undergo dedifferentiation.
PPARγ is a master regulator of adipogenic differentiation [25]. It induces the expression of C/EBPα during adipogenic differentiation, thereby dramatically stimulating the induction of adipocyte genes [26]. In our study, PPARγ expression gradually increased after inducing differentiation, and both the protein levels and the phosphorylation levels of PPARγ were correlated with the expression of P53-ERK-CREB during differentiation. The transition between cell proliferation and cell differentiation during adipogenic differentiation is a tightly regulated process in which both cell cycle regulators and differentiating factors interact, creating a cascade of events leading to the commitment of the cells to the adipocyte phenotype [27]. Growth-arrested preadipocytes undergo several rounds of the cell cycle before terminally differentiating into adipocytes, suggesting that cross-talk may exist between the cell cycle or the cell proliferation machinery and the factors controlling cell differentiation. PPARγ ligand production is tightly linked to clonal expansion during the initiation of adipogenic differentiation. Post-translational modi cation of chromatin histones is a key mechanism of transcriptional regulation [28]. Protein phosphorylation is a common post-translational modi cation that regulates a wide range of signaling pathways involved in differentiation, apoptosis, proliferation, gene regulation, and metabolism [29].
PPARγ is a short-lived protein [30], and it has been shown to be regulated by a series of post-translational modi cations [31]. Studies have found that PPARγ has a variety of post-translational modi cations, including phosphorylation, ubiquitination, SUMO-mediated modi cation, acetylation, and nitrosylation [32]. The phosphorylation of PPARγ is the most well studied. It involves mitogen-activated protein kinases (MAPKs), cyclin-dependent kinase 5 (CDK5), and AMP-activated protein kinase (AMPK). Different kinases phosphorylate modi ed proteins to produce different biological functions. Studies have shown that during the differentiation of 3T3-L1 adipocytes, CDK9-mediated PPARγ is highly phosphorylated, indicating that CDK9 can mediate PPARγ phosphorylation, enhance the transcriptional activity of PPARγ, and promote adipogenic differentiation and lipid synthesis [33]. Our study con rmed that adipogenic differentiation of PGCCs activates the ERK pathway, which phosphorylates PPARγ, reduces the expression of PPARγ mRNA and protein, and inhibits adipogenic differentiation. The phosphorylation level of PPARγ protein at Serine 112 gradually decreased at the early adipocyte differentiation stage, while the downstream FABP4 protein expression gradually increased, enabling tumor cells to acquire the adipocyte phenotype. Helenius et al. also con rmed that phosphorylation inhibits the ability of PPARγ to promote adipogenic differentiation [34]. Phosphorylation of PPARγ2 at the N-terminal serine residue 112 reduces its transcriptional activity, promotes ubiquitination on lysine 107, and further reduces its ability as a transcriptional activator [35]. After inducing adipogenic differentiation of PGCCs, the regulatory effect between phosphorylation and ubiquitination of PPARγ is awaiting further study.
The tumor suppressor gene p53 is functionally involved in cell cycle control, apoptosis, and genomic stability and is mutated and inactivated in most human cancers. A hallmark of virtually all cancers is the dysregulated expression or functioning of pRB or P53 [16]. There are several p53 gene mutations in human tumors, including wild-type (p53+/+), knockdown (p53 +/-), and deletion (p53 -/-) [36]. Although p53 is one of the most well-described genes, its role in adipocytes is poorly understood. Regulation of PPARγ expression by P53 depends on the p53 genotype. Our experiments con rmed that the P53-ERK-CREB pathway regulates the activation and phosphorylation of PPARγ during the adipogenic differentiation of HEY and MDA-MB-231 control and PGCCs with daughter cells. The level of FABP4 expression in the p53 wild-type is higher than that in the p53 mutant type, but the degree of adipogenic differentiation in the p53 wild-type is lower than that in the p53 mutant type. FABP4 negatively regulates the adipogenic differentiation of HEY and MDA-MB-231 PGCCs with daughter cells. Knockdown of P53 protein expression in wild-type p53 HEY and mutant p53 MDA-MB-231 cells showed that, in HEY control cells, PPARγ phosphorylation is reduced and the degree of differentiation is higher, while in MDA-MB-231 PGCCs with daughter cells, PPARγ phosphorylation is increased and the degree of differentiation is lower.
Wild-type p53 negatively regulates the adipogenic differentiation of PGCCs with daughter cells, whereas mutant p53 promotes their adipogenic differentiation.
P300 is an HAT transcriptional coactivator that is critical for a multitude of cellular processes [37]. A485 is a potent, selective, and drug-like catalytic inhibitor of P300 and CBP. Lasko et al. con rmed that A485 competes with acetyl-CoA to inhibit the acetyltransferase activity of p300 [18]. Silencing P53 or p300 disrupted the P53-P300 complex and its subsequent binding to the P53 minimal promoter [38]. P300-P53 regulates the acetylation level of PPARγ, as con rmed by our experiments. In HEY control and PGCCs with daughter cells, P53 is regulated by P300 and undergoes ubiquitination/degradation; after adipogenic differentiation, the expression of P300 is upregulated to increase the acetylation level of PPARγ. In MDA-MB-2331 control and PGCCs with daughter cells, PPARγ is regulated by P300 and undergoes acetylation; after adipogenic differentiation, the level of P53 acetylation is signi cantly increased, and the level of PPARγ acetylation also increases. After adding the P300 inhibitor A485 in HEY cells, the level of P53 ubiquitination decreased, the level of P53 acetylation increased, the level of PPARγ acetylation increased, and the level of FABP4 expression decreased. In MDA-MB-231 cells, the level of P53 acetylation signi cantly decreased, the level of PPARγ acetylation decreased, and the level of FABP4 expression increased. We speculate that PPARγ acetylation is positively modulated by mutation type p53 (p53 +/-) in MDA-MB-231 cells, forming a P300-P53-ace-PPARγ positive feedback loop. P300 negatively regulates wild-type p53, and wild-type p53 negatively regulates PPARγ acetylation. However, the precise mechanism underlying P53 and P300 interaction requires further investigation, and the function of PPARγ acetylation in the adipogenic differentiation of PGCCs with daughter cells remains to be explored.

Conclusion
FABP4 attenuates PPARγ and adipogenesis in cancer cells. Our study con rms a causative link between FABP4 and PPARγ and may explain some differences between the adipocyte subtypes comprising these two kinds of p53 genotypes. For instance, HEY cells proliferate and differentiate into mature adipocytes less readily than do MDA-MB-231 cells. Thereafter, the success of inducing adipogenesis is likely due to the loss-of-function of wild-type p53 and the acquisition of mutated p53. FABP4 is induced by PPARγ, the master regulator of adipogenesis, yet these two regulators exert opposite effects on various metabolic parameters, such as insulin resistance and in ammation [39]. Previous studies have found that PPARγ activity is elevated in FABP4-null macrophages. FABP4 was reported to physically interact with PPARγ [40], and it is likely that such a physical interaction triggers the ubiquitination and subsequent proteasomal degradation of PPARγ