Adipocyte‐Derived Palmitic Acid Promotes Breast Cancer Malignancy via ZDHHC15‐Mediated S‐Palmitoylation of PPARγ Authors: Geon Ho Moon, Min Young Lee, Jiwon Koh, Ki Yong Hong, Junji Fukuda, Sung Joon Kim, Yang‐Sook Chun Published: 2026 ABSTRACT ABSTRACTPeroxisome proliferator‐activated receptor gamma (PPARγ), a well‐characterized regulator of adipogenesis, also functions as a tumor suppressor in breast cancer. However, the PPARγ agonist rosiglitazone has shown limited clinical efficacy, highlighting the importance of understanding PPARγ regulation within the tumor microenvironment. Here, we demonstrate that adipocyte‐derived palmitic acid promotes PPARγ S‐palmitoylation via zinc finger DHHC‐domain containing protein 15 (ZDHHC15), as confirmed by acyl‐biotin exchange (ABE) assay. Immunoblot, RT‐qPCR, and immunoprecipitation further revealed that this modification leads to ubiquitin‐mediated proteasomal degradation of PPARγ. Functional studies using adipocyte‐conditioned media and an oxygen‐permeable 3D spheroid culture system showed that PPARγ palmitoylation promotes breast cancer malignancy and confers resistance to rosiglitazone. Notably, the palmitoylation inhibitor 2‐bromopalmitate (2‐BP) stabilizes PPARγ, suppressing breast cancer malignancy. Furthermore, 2‐BP synergistically enhances the therapeutic efficacy of rosiglitazone, suggesting that targeting PPARγ S‐palmitoylation represents a promising therapeutic strategy in breast cancer. Abbreviations 2‐bromopalmitate acyl‐biotin exchange conditioned media free fatty acid palmitic acid polydimethylsiloxane peroxisome proliferator‐activated receptor gamma post‐translational modification zinc finger DHHC‐domain containing Introduction Breast cancer is the most commonly diagnosed malignancy among women worldwide, and its global burden is projected to continue rising [1]. Despite advances in early detection and treatment, metastatic breast cancer remains a leading cause of cancer‐related death [2]. In particular, the molecular and cellular heterogeneity of breast cancer distinguishes it from many other malignancies [3]. This diversity complicates therapy and often leads to treatment resistance and disease recurrence [2, 4]. These challenges highlight the urgent need for more effective therapeutic strategies that can overcome the diverse subtypes of breast cancer [5]. PPARγ is a nuclear receptor that regulates lipid metabolism, cell differentiation, and inflammatory signaling, as well as cancer‐related processes [6]. In breast cancer, PPARγ has been shown to regulate tumor growth across multiple subtypes, making it a promising therapeutic target [7]. PPARγ exhibits tumor‐suppressive functions in breast cancer through multiple mechanisms, including PTEN upregulation and p53‐mediated apoptosis [8, 9]. Consistent with these functions, the PPARγ agonist rosiglitazone has shown anti‐tumor effects in preclinical models [8, 9, 10]. However, clinical studies have reported that rosiglitazone treatment fails to increase PPARγ levels and does not produce significant therapeutic benefit [11]. Therefore, it is essential to understand PPARγ regulation in a physiological context to effectively inhibit breast cancer progression and improve therapeutic outcomes. Cancer progression depends not only on genetic alterations within tumor cells but also on signals from the surrounding microenvironment [12, 13]. This is particularly important in breast cancer, where the mammary gland is surrounded by adipose tissue [14]. Epidemiological studies have consistently identified obesity as a major risk factor for breast cancer incidence, progression, and poor prognosis [15]. Obesity promotes the expansion of adipose tissue, leading to increased adipocyte abundance and a tumor‐promoting microenvironment [16, 17]. Under these conditions, adipocytes exhibit increased secretion of cytokines, adipokines, and free fatty acids (FFAs), which enhance cancer malignancy [18]. Among these, FFAs are directly taken up by breast cancer cells and utilized as an energy source to support tumor growth and aggressiveness [19]. Furthermore, intracellular FFAs act as signaling molecules that regulate key oncogenic pathways by altering membrane composition and modulating protein function [20, 21]. These findings establish the role of FFAs as critical mediators of breast cancer malignancy. Palmitic acid derived from adipocytes serves as the substrate for palmitoylation, a post‐translational modification that finely regulates protein function beyond transcriptional control [22]. Protein S‐palmitoylation involves the covalent attachment of palmitoyl‐CoA to cysteine residues on target proteins, modulating their subcellular localization, stability, and activity [23]. This process is catalyzed by a family of ZDHHCs, comprising 23 members in mammals with distinct substrate specificities [24]. Recent studies have shown that palmitoylation stabilizes oncogenic proteins such as Flotillin‐1 in breast cancer, thereby promoting tumor growth and lung metastasis [25]. In addition, the palmitoylation inhibitor induces lysosome‐mediated degradation of the immune checkpoint protein B7 H4, enhancing anti‐tumor immune responses [26]. These findings underscore the need to investigate the role of protein palmitoylation in breast cancer malignancy. In this study, we investigated the effects of adipocyte‐derived palmitic acid on PPARγ in breast cancer. Mechanistically, ZDHHC15‐mediated palmitoylation of PPARγ reduces its stability, thereby promoting breast cancer progression and conferring resistance to rosiglitazone. These results offer a molecular explanation for the limited clinical efficacy of rosiglitazone in breast cancer patients and contribute to our understanding of potential strategies for targeting PPARγ palmitoylation in breast cancer therapy. Material and Methods Cell Culture and Pharmacological Treatment Human adipose‐derived stem cells (hADSCs) were isolated from human adipose tissue as previously described [13]. This procedure was conducted in accordance with the guidelines of the Institutional Review Board (IRB) of Seoul National University Hospital (Approval No. H‐1602‐110‐742). 4T1 (ATCC), MDA‐MB‐231 (KCLB), and ZR‐75‐1 (KCLB) cells were cultured in RPMI1640 (Welgene), and hADSCs were maintained in DMEM/F‐12 (Thermo Fisher Scientific), both supplemented with 10% FBS (GenDEPOT) and 1% penicillin/streptomycin (P/S; HyClone). 3T3‐L1 cells (KCLB) were maintained in DMEM supplemented with 10% bovine serum (BS; Gibco) and 1% P/S. All cells were incubated at 37°C in 5% CO2. For pharmacological treatment, cells were treated as described in the figure legends, and detailed information on reagents is provided in Table S1. Media and treatments were replaced daily. Human Breast Tissues for Immunohistochemistry A tissue microarray (TMA) containing normal breast tissues and breast carcinoma tissues from 26 patients was obtained from Seoul National University Hospital (Seoul, South Korea) with consent under IRB approval (Approval No. H‐2207‐198‐1345). TMA sections were subjected to immunohistochemistry using an anti‐PPARγ primary antibody (16643–1‐AP; Proteintech) or anti‐ZDHHC15 primary antibody (21627‐1‐AP; Proteintech), followed by incubation with Anti‐Rabbit IgG (H + L), Biotinylated Secondary Antibody (Vector Labs) for 1 h at room temperature and processed using the VECTASTAIN ABC kit (Vector Labs) according to the manufacturer's instructions. Histochemical score (H‐score) was calculated using the formula: 1 × (percentage of weakly stained cells [1+]) + 2 × (percentage of moderately stained cells [2+]) + 3 × (percentage of strongly stained cells [3+]). Adipocyte Differentiation 3T3‐L1 pre‐adipocytes were induced to differentiate using DMI (1 μM dexamethasone, 500 μM 3‐isobutyl‐1‐methylxanthine, and 5 μg/mL insulin; Sigma) in DMEM containing 10% FBS and 1% P/S. The medium was then replaced with fresh DMI‐containing differentiation medium for an additional 48 h. Subsequently, cells were cultured in DMEM supplemented with 10% FBS, 1% P/S, and 1 μg/mL insulin for 48 h. Thereafter, medium was replaced with DMEM supplemented with 10% FBS and 1% P/S for 48 h. For hADSC differentiation, cells were treated with DMI and 200 μM indomethacin under the same basal culture conditions. On day 12, the medium was replaced with maintenance medium. Media were changed every 2 days. Conditioned Media (CM) CM were collected after 24 h incubation in serum‐free medium from 3T3‐L1 adipocytes differentiated for 8 days or human adipocytes differentiated for 14 days. The collected CM was centrifuged at 12,000 rpm for 20 min to remove debris and subsequently filtered through 0.45 μm Minisart filters (Sigma). FFA‐BSA Complex FFAs were dissolved in ethanol to prepare stock solutions. Culture medium containing 10% charcoal‐stripped FBS and 1% penicillin/streptomycin was supplemented with 2% fatty acid–free bovine serum albumin (FFA‐free BSA; Sigma), and the FFA stock solutions were added to the medium. The mixture was incubated at 55°C for 10 min to ensure complete complex formation. The resulting FFA‐BSA solution was filtered through 0.45 μm Minisart filters before being applied to cells for treatment. Analysis of Differentially Expressed Genes (DEGs) Gene expression profile data were obtained from the Gene Expression Omnibus (GEO) database of the NCBI and The Cancer Genome Atlas (TCGA). Probe intensities from microarrays were log2‐transformed and normalized. DEGs were identified using thresholds of | log2 (fold change) |> 1 and −log10 (p‐value) > 1.4. Functional annotation analysis was conducted with the Database for Annotation, Visualization and Integrated Discovery (DAVID) to identify Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. Significantly enriched pathways were defined as those with a Benjamini–Hochberg‐adjusted p‐value < 0.05. The survival curves were measured based on the Kaplan–Meier method and assessed with the log‐rank test in breast cancer patients. Transfection of siRNAs and Plasmid siRNAs were purchased from Integrated DNA Technologies (IDT, Coralville, IA, USA), and their sequences are listed in Table S2. Flag‐PPARγ plasmid (#8895) was obtained from Addgene. Transfection of siRNAs or the plasmid into breast cancer cells was performed using Lipofectamine RNAiMAX (Invitrogen) or Lipofectamine 2000 (Invitrogen), respectively. Transwell Invasion Assay Transwell invasion assays were performed using Transwell inserts (6.5 mm diameter, 8 μm pore size; Corning Inc.) coated with 0.5 mg/mL collagen (Sigma). After the indicated treatments, breast cancer cells were resuspended in 200 μL of serum‐free medium and seeded in the upper chamber, while the lower chamber was filled with 500 μL of medium matching the treatment condition. After 24 h of incubation, invaded cells were fixed with methanol and stained with 0.1% crystal violet (Sigma) in 2% methanol. Images were acquired using a microscope, and the number of invaded cells was quantified using ImageJ software. Oxygen‐Permeable 3D Spheroid Culture Oxygen‐permeable 3D spheroid culture systems were designed by Professor Fukuda [27, 28]. Cells (5 × 105) were seeded in 4 mL of culture medium and incubated for 5 days, with the medium replaced daily. Spheroid diameter and circularity were measured using ImageJ software. Immunoblot Analysis Cell lysates were subjected to SDS‐PAGE and transferred to Immobilon‐P membranes (Millipore). Membranes were blocked for 1 h in Tris‐buffered saline containing 0.1% Tween‐20 (TBS‐T) supplemented with 1% skim milk (BD Difco) and 1% BSA. Subsequently, the membranes were incubated at 4°C overnight with specific primary antibodies (listed in Table S3). Following primary antibody incubation, the membranes were incubated with a HRP‐conjugated secondary antibody (Invitrogen) or VeriBlot for IP Detection Reagent (HRP) (Abcam) for 1 h at room temperature. Protein bands were detected using the ECL Plus kit (Thermo Fisher Scientific). Quantitative RT‐PCR Total RNA was isolated using TRIzol reagent (Invitrogen) following the manufacturer's instructions. cDNA was synthesized with the EasyScript cDNA Synthesis Kit (Applied Biological Materials). RT–qPCR was performed using EvaGreen qPCR master mix (Applied Biological Materials) on a StepOne Real‐time PCR system (Applied Biosystems, Foster City, CA, USA). For each sample, the cycle threshold (Ct) value of the target gene was normalized to 18S rRNA (ΔCt), and the ΔCt was subtracted from that of the control group (ΔΔCt). Relative mRNA levels were expressed as 2−ΔΔCt. Primer sequences are listed in Table S4. Immunoprecipitation and ABE Assay Cells were lysed in buffer containing 50 mM Tris‐Cl (pH 7.5), 100 mM NaCl, 0.1% NP‐40 (Georgiachem), 5 mM EDTA, 1 mM PMSF (Sigma), and 100 μM protease inhibitor cocktail (Sigma). For ABE assay, 50 mM N‐ethylmaleimide (Sigma) was additionally included in the lysis buffer. Protein concentrations were determined using the DC protein assay kit (Bio‐Rad). Lysates were incubated with anti‐PPARγ antibody (sc‐7273; Santa Cruz Biotechnology) at 4°C overnight, followed by incubation with Protein A/G Sepharose beads (GE Healthcare Life Sciences) at 4°C for 4 h. Endogenous PPARγ was immunoprecipitated from cell lysates using anti‐PPARγ antibody (sc‐7273; Santa Cruz Biotechnology) at 4°C overnight, followed by incubation with Protein A/G Sepharose beads (GE Healthcare Life Sciences) at 4°C for 4 h. Flag‐tagged PPARγ was immunoprecipitated using EZview Red anti‐FLAG M2 Affinity Gel (Sigma‐Aldrich) at 4°C for 4 h. For the ABE assay, immunoprecipitated endogenous PPARγ or Flag‐tagged PPARγ was divided into two groups and treated with either lysis buffer alone or lysis buffer containing 1 M hydroxylamine (Sigma) at room temperature for 1 h. The beads were then incubated with EZ‐Link BMCC‐Biotin (Thermo Fisher Scientific) at 4°C for 1 h. Bound proteins were eluted with 2× SDS loading buffer and subjected to immunoblotting. In the ABE assay, palmitoylated proteins were detected using streptavidin‐HRP (Bio‐Rad). Site‐Directed Mutagenesis Site‐directed mutagenesis was performed to generate cysteine‐to‐serine substitutions at positions Cys159, Cys176, and Cys313 of PPARγ. Fragments of Flag‐PPARγ plasmid (#8895) were amplified using primer pairs #1/#2 for C159S, #3/#4 for C176S, and #5/#6 for C313S (sequences listed in Table S5). PCR amplification was performed using KOD‐Plus‐Mutagenesis Kit (SMK‐101; TOYOBO) according to the manufacturer's instructions. The PCR products were treated with DpnI to selectively digest the methylated template plasmid and then transformed into DH5α competent cells (Real Biotech Corporation). Luciferase Reporter Assay To evaluate PPARγ activity, cells were transfected with the proliferator response element (PPRE)‐luciferase (PPRE X3‐TK‐luc, Addgene) and β‐galactosidase plasmid. Following treatment with the indicated drugs, luciferase activity was measured using a Lumat LB9507 luminometer (Berthold Technologies, Bad Wildbad, Germany). β‐galactosidase activity was assessed using a spectrophotometer to normalize transfection efficiencies. MTT Assay Cells were seeded in 96‐well plates and cultured in the indicated media for 72 h, with media replaced daily. Subsequently, cells were incubated with 0.5 mg/mL MTT (Sigma) for 2 h. Formazan was solubilized in 200 μL of DMSO, and absorbance was measured at 570 nm. IC50 values and the combination index (CI) were calculated using GraphPad Prism 8. Synergistic effects were analyzed according to the Chou–Talalay method. Statistical Analysis Data are presented as mean ± SD. Statistical significance was determined using two‐tailed unpaired Student's t‐test or one‐way ANOVA followed by Tukey's HSD post hoc test. p‐values < 0.05 were considered statistically significant. The number of biologically independent experiments is represented by n. Graphical representations were generated using GraphPad Prism 8. Results Downregulation of PPARγ by Adipocyte‐Derived Factors Is Associated With Poor Prognosis in Breast Cancer To investigate pathway alterations in breast cancer, we performed KEGG pathway analysis of normal breast tissue and breast cancer using gene expression datasets. PPAR signaling showed the most significant downregulation in breast cancer relative to normal tissues (Figure 1A and Table S6). Among PPAR pathway genes, PPARG mRNA level was reduced in metastatic tumors compared to primary tumors (Figure 1B). Clinically, low PPARG expression was associated with worse overall and disease‐free survival (Figure 1C). Immunohistochemistry across disease stages confirmed progressive PPARγ protein loss from normal tissue through carcinoma in situ to invasive carcinoma (Figure 1D). These findings indicate that loss of PPARγ expression is associated with breast cancer progression and poor clinical outcome, supporting the clinical relevance of PPARγ in breast cancer. Consistent with these findings, PPARG knockdown enhanced invasion in MDA‐MB‐231, ZR‐75‐1, and 4T1 cells (Figure S1A), while its overexpression suppressed invasion (Figure S1B). These cell lines were selected to represent distinct molecular subtypes of breast cancer and to enable cross‐species validation of the proposed mechanism. MDA‐MB‐231, ZR‐75‐1, and 4T1 represent human triple‐negative (TNBC), human luminal A, and mouse TNBC‐like breast cancer models, respectively (Table S7) [29, 30]. We next examined additional factors associated with PPARγ expression beyond differences in invasiveness. Patients were stratified into high‐ and low‐PPARγ H‐score groups within all carcinomas, carcinoma in situ, and invasive carcinoma (Figure S1C). In all cases, body mass index (BMI) was significantly higher in the low PPARγ group (Figure 1E). Notably, adipocyte‐CM reduced PPARγ expression and increased invasion (Figure 1F). To mimic the previously described in vivo tumor environment [13, 31], breast cancer cells were cultured in an oxygen‐permeable PDMS‐based 3D spheroid system. The oxygen‐permeable PDMS membrane at the base of each culture well allows continuous oxygen diffusion to the spheroid, more closely mimicking physiological oxygen tension than conventional static 3D culture [27, 28]. Cancer cell proliferation was assessed by measuring spheroid diameter over time, and migration was evaluated based on spheroid circularity (Figure 1G). Spheroids cultured in adipocyte‐CM exhibited significantly increased proliferation and migration compared to controls, as evidenced by increased spheroid diameter and decreased circularity at day 5 (Figure 1H and Figure S1D). These results demonstrate that adipocyte‐derived factors suppress PPARγ expression, contributing to breast cancer progression. Downregulation of PPARγ by adipocyte‐derived factors is associated with poor prognosis in breast cancer. (A) KEGG analysis of transcriptomic datasets from normal and breast cancer tissues (GSE15852, GSE36295, GSE42568, GSE50567, GDC TCGA, and TCGA). (B) Volcano plot of TCGA breast cancer data showing significantly downregulated (blue), significantly upregulated (red), and non‐significant (black) genes. (C) Survival analysis of patients stratified by PPARG expression (median cutoff). Overall survival (GSE1456) and disease‐free survival (GSE12093). (D) Immunohistochemistry staining for PPARγ in normal breast tissue (n = 26), breast carcinoma in situ (n = 16), invasive breast carcinoma (n = 26). Scale bar, 200 μm. PPARγ expression was quantified using H‐score. (E) Body mass index (BMI) in patients stratified by PPARγ H‐score. Patients were divided into high‐ and low‐PPARγ groups in all carcinomas (high, n = 18; low, n = 24), carcinoma in situ (high, n = 10; low, n = 6), and invasive carcinoma (high, n = 8; low, n = 18) by k‐means clustering. (F) Transwell invasion assay (left) and immunoblot analysis of PPARγ protein expression (right) in MDA‐MB‐231, ZR‐75‐1, and 4 T1 cells (n = 3). Cells were cultured in the absence or presence of 10% adipocyte‐CM for 24 h. Scale bar, 200 μm. (G) Schematic of PDMS‐based 3D culture chips. (H) Representative images of spheroids cultured in the absence or presence of 10% adipocyte‐CM on 3D culture chips for 5 days. (n = 3). Scale bar, 100 μm. Data are presented as mean ± SD. Statistical significance was determined using two‐tailed unpaired Student's t‐test. *p < 0.05, **p < 0.01, ***p < 0.001, and n.s. > 0.05. Palmitic Acid Mediates PPARγ Proteasomal Degradation via S‐Palmitoylation Based on evidence that adipocytes increase cancer malignancy through FFA release [18], we tested which FFAs affect PPARγ expression in breast cancer cells. Treatment with various FFAs revealed that only palmitic acid (PA) consistently reduced PPARγ protein expression across all cell lines (Figure 2A). However, PPARG mRNA levels remained unchanged (Figure S2A), and the PA‐induced reduction in PPARγ protein was restored by the proteasome inhibitor MG132 but not by the lysosome inhibitor hydroxychloroquine (HCQ) (Figure 2B). These findings indicate that PA promotes proteasomal degradation of PPARγ. Given that palmitic acid can regulate protein stability through S‐palmitoylation [23], we assessed whether PPARγ undergoes PA‐induced S‐palmitoylation in breast cancer cells. ABE assay confirmed that PA treatment increased PPARγ S‐palmitoylation, which was abolished by the palmitoylation inhibitor 2‐BP (Figure 2C). Next, we conducted in silico analyses using the GPS‐Palm algorithm to predict potential palmitoylation sites on PPARγ in both Homo sapiens and Mus musculus . The analysis identified C159, C176, and C313 as the most probable palmitoylation sites (Figure S2B). To determine the palmitoylation site experimentally, we generated PPARγ cysteine‐to‐serine mutants (C159S, C176S, and C313S) and assessed their S‐palmitoylation by ABE assay following PA treatment. Among the three mutants, only the C159S mutant markedly decreased PA‐induced PPARγ S‐palmitoylation, whereas the C176S and C313S mutants exhibited palmitoylation levels comparable to those of wild‐type PPARγ (Figure 2D), identifying C159 as the major S‐palmitoylation site on PPARγ. To determine whether this modification is functionally required for PA‐induced degradation, we pharmacologically inhibited palmitoylation and assessed PPARγ protein levels. Inhibition of palmitoylation restored PA‐reduced PPARγ protein expression without affecting mRNA (Figure 2E). Consistently, PPARγ transcriptional activity, measured by PPAR response element (PPRE) luciferase assay, paralleled changes in PPARγ protein expression (Figure S2C). We next evaluated PPARγ protein stability using the protein synthesis inhibitor cycloheximide (CHX) to measure PPARγ turnover by immunoblotting. PA treatment reduced PPARγ stability, which was restored by 2‐BP (Figure 2F). PPARγ levels were further reduced when PA was combined with CHX compared with PA alone, consistent with our finding that PA promotes proteasomal degradation of PPARγ rather than suppressing its synthesis. Immunoprecipitation further confirmed that PA increased PPARγ ubiquitination, which was suppressed by 2‐BP (Figure 2G). Importantly, PA failed to increase PPARγ ubiquitination in the C159S mutant (Figure S2D), indicating that C159 is required for PA‐induced PPARγ ubiquitination. These results reveal that palmitic acid promotes PPARγ proteasomal degradation through S‐palmitoylation of C159, which drives subsequent ubiquitination. Palmitic acid mediates PPARγ proteasomal degradation via S‐palmitoylation. (A) Immunoblot analysis of PPARγ in MDA‐MB‐231, ZR‐75‐1, and 4T1 cells treated with vehicle or fatty acids (palmitic, stearic, palmitoleic, oleic, linoleic, and α‐linolenic acid; 100 or 300 μM) for 24 h. (B) Immunoblot analysis of PPARγ in MDA‐MB‐231, ZR‐75‐1, and 4T1 cells. Cells were treated with vehicle or PA (300 μM) for 24 h. MG132 (10 μM) or HCQ (20 μM) was added during the last 8 h of PA treatment. (C) ABE assay for PPARγ S‐palmitoylation. Cells were treated with vehicle or 2‐BP (25 μM) in the absence or presence of PA (300 μM) for 24 h. MG132 (10 μM) was added during the last 8 h of PA treatment. (D) ABE assay for Flag‐PPARγ S‐palmitoylation. Cells transfected with Flag‐PPARγ WT, C159S, C176S, or C313S were treated with vehicle or PA (300 μM) for 24 h. MG132 (10 μM) was added during the last 8 h of PA treatment. (E–G) Cells were treated with vehicle or 2‐BP (25 μM) in the absence or presence of PA (300 μM) for 24 h. (E) RT–qPCR (upper) and immunoblot (lower) analyses of PPARγ. (F) Following 24 h treatment, cells were incubated with CHX (100 μM) for 1 h and analyzed by immunoblotting with band quantification. (G) Immunoprecipitation analysis for PPARγ ubiquitination. MG132 (10 μM) was added during the last 8 h of PA treatment. Data are presented as mean ± SD. Unless otherwise indicated, n = 3 independent experiments. Statistical significance was determined by one‐way ANOVA followed by Tukey's HSD post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001, and n.s. > 0.05. PPARγ Palmitoylation Enhances Breast Cancer Aggressiveness via AKT Signaling Next, we assessed the functional relevance of PPARγ restoration by palmitoylation inhibition in breast cancer. PA treatment significantly increased invasiveness in breast cancer cells, whereas co‐treatment with 2‐BP markedly attenuated PA‐induced invasion. In PPARG knockdown cells, this inhibitory effect was abolished, indicating that 2‐BP suppresses invasion through PPARγ (Figure 3A and Figure S3A). Similarly, adipocyte‐CM‐induced invasion was suppressed by 2‐BP in a PPARγ‐dependent manner (Figure 3B and Figure S3B). In the 3D spheroid system, PA treatment increased both breast cancer cell proliferation and migration, which were reversed by 2‐BP (Figure 3C and Figure S3C,D). Adipocyte‐CM also enhanced proliferation and migration, and 2‐BP similarly suppressed these effects (Figure 3D and Figure S3E,F). These responses were not observed in PPARG knockdown cells (Figure 3C,D and Figure S3C–F). To elucidate the signaling mechanism underlying the tumor‐suppressive effects of 2‐BP, we examined AKT and ERK phosphorylation. 2‐BP reduced PA‐induced AKT phosphorylation, whereas ERK phosphorylation remained unchanged (Figure 3E). Based on genes previously reported to play key roles in cancer cell proliferation and EMT regulation via the AKT pathway [32, 33], we examined whether 2‐BP‐mediated reduction of p‐AKT affected the expression of these target genes. 2‐BP treatment suppressed PA‐induced upregulation of proliferation‐associated genes (CCND1 and CCNE1) and EMT markers (VIM, CDH2, and SNAI1). Notably, 2‐BP treatment restored PTEN mRNA levels (Figure 3F and Figure S3G), consistent with previous reports that PPARγ positively regulates PTEN to suppress AKT signaling [8]. Together, these findings suggest that adipocyte‐derived palmitic acid promotes breast cancer malignancy by inducing PPARγ degradation through S‐palmitoylation, thereby activating AKT signaling and downstream oncogenic transcriptional programs. PPARγ palmitoylation enhances breast cancer aggressiveness via AKT signaling. (A–F) Cells transfected with si‐control or si‐PPARG (50 nM) were treated with vehicle or 2‐BP (25 μM) for 24 h in the absence or presence of PA (A, C, E–F) or 10% adipocyte‐CM (B, D). (A, B) Transwell invasion assay in MDA‐MB‐231, ZR‐75‐1, and 4 T1 cells. Scale bar, 200 μm. (C, D) Representative spheroids images on 3D culture chips after 5 days. Scale bar, 100 μm. (E) Immunoblot analysis of PPARγ, p‐AKT, AKT, p‐ERK and ERK in MDA‐MB‐231, ZR‐75‐1, and 4 T1 cells. (F) RT–qPCR analysis of CCND1, CCNE1, PTEN, VIM, CDH2 and SNAI1 in MDA‐MB‐231. Data are presented as mean ± SD. Unless otherwise indicated, n = 3 independent experiments. Statistical significance was determined by one‐way ANOVA followed by Tukey's HSD post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001, and n.s. > 0.05. ZDHHC15‐Mediated PPARγ Palmitoylation Promotes Breast Cancer Progression To identify candidate palmitoyltransferases responsible for PPARγ S‐palmitoylation, we analyzed PPARγ protein expression following knockdown of individual ZDHHC enzymes (ZDHHC1–24) in the presence of PA (Figure S4A). In both MDA‐MB‐231 and ZR‐75‐1 cells, knockdown of ZDHHC14, ZDHHC15, and ZDHHC16 elevated PPARγ expression (Figure S4B). Notably, ZDHHC15 knockdown significantly reduced PPARγ palmitoylation (Figure 4A and Figure S4C). Immunoprecipitation assays demonstrated that ZDHHC15 physically interacts with PPARγ (Figure 4B). Given that PPARγ palmitoylation facilitates ubiquitination, we examined whether ZDHHC15 knockdown affects PPARγ ubiquitination. PA‐induced PPARγ ubiquitination was reduced by ZDHHC15 knockdown (Figure 4C). Consistently, the increase in PPARγ protein levels following ZDHHC15 knockdown was not accompanied by changes in PPARG mRNA expression (Figure 4D). Next, we investigated the clinical relevance of ZDHHC15 in breast cancer by analyzing its expression in in situ and invasive breast carcinoma. Immunohistochemistry revealed that ZDHHC15 expression was upregulated in invasive breast carcinoma (Figure 4E). Furthermore, high ZDHHC15 expression was associated with reduced overall and disease‐free survival (Figure 4F). Transwell invasion assay revealed ZDHHC15 knockdown decreased adipocyte‐CM‐induced breast cancer cell invasion, an effect that was abolished upon PPARG co‐knockdown (Figure 4G and Figure S5A). Consistently, ZDHHC15 knockdown also suppressed adipocyte‐CM‐induced proliferation and migration in the 3D spheroid system, and these effects were similarly lost upon PPARG co‐knockdown (Figure 4H and Figure S3B,C). These results suggest that ZDHHC15 promotes breast cancer malignancy through PPARγ palmitoylation in an adipocyte‐conditioned microenvironment. ZDHHC15‐mediated PPARγ palmitoylation promotes breast cancer progression. (A, C, D) Cells transfected with si‐control or si‐ZDHHC15 (50 nM) were treated with vehicle in the absence or presence of PA for 24 h. (A) ABE assay for PPARγ S‐palmitoylation. MG132 (10 μM) was added during the last 8 h of PA treatment. (B) Cells were treated with MG132 (10 μM) for 8 h. Immunoprecipitation with anti‐PPARγ followed by immunoblotting for ZDHHC15. (C) Immunoprecipitation analysis for PPARγ ubiquitination. MG132 (10 μM) was added during the last 8 h of PA treatment. (D) RT–qPCR analysis of PPARG (upper) and immunoblot analysis of PPARγ and ZDHHC15 (lower) after 24 h treatment. (E) Immunohistochemistry staining for ZDHHC15 in carcinoma in situ (n = 16), invasive carcinoma (n = 26). Scale bar, 200 μm. (F) Survival analysis of breast cancer patients stratified by ZDHHC15 expression (upper tertile cutoff; GSE16446). (G–H) Cells transfected with si‐control, si‐ZDHHC15 (50 nM), or si‐PPARG (50 nM) were treated with 10% adipocyte‐CM for 24 h. (G) Transwell invasion assay in MDA‐MB‐231, ZR‐75‐1, and 4T1 cells. Scale bar, 200 μm. (H) Representative spheroids images on 3D culture chips after 5 days. Scale bar, 100 μm. Data are presented as mean ± SD. Unless otherwise indicated, n = 3 independent experiments. Statistical significance was determined using two‐tailed unpaired Student's t‐test for (E) and one‐way ANOVA followed by Tukey's HSD post hoc test for (D, G, H). *p < 0.05, **p < 0.01, ***p < 0.001, and n.s. > 0.05. Combined Rosiglitazone and 2‐BP Treatment Synergistically Suppresses Adipocyte‐CM‐Induced Breast Cancer Malignancy Since rosiglitazone, a PPARγ agonist, has been reported to increase PPARγ expression and suppress malignant phenotypes in breast cancer cells [34], we examined whether it could reverse the adipocyte‐CM–induced effects. Although rosiglitazone increased PPARγ protein expression, reduced AKT phosphorylation, and decreased invasion in control conditions, it failed to restore PPARγ protein expression, suppress AKT phosphorylation, or inhibit invasion in adipocyte‐CM (Figure 5A). This rosiglitazone resistance was overcome by 2‐BP co‐treatment, which increased PPARγ expression, further reduced AKT phosphorylation, and decreased invasion more effectively than 2‐BP alone (Figure 5B). Consistently, rosiglitazone failed to suppress adipocyte‐CM‐induced proliferation and migration in the 3D spheroid system. In contrast, its combination with 2‐BP significantly decreased both parameters (Figure 5C,D and Figure S6A), suggesting a synergistic effect of these treatments in suppressing breast cancer progression. To further explore the synergism between palmitoylation inhibition and rosiglitazone, the IC50 values of 2‐BP and rosiglitazone were determined in breast cancer cell lines (Figure S6B). Combination index (CI) analysis revealed that no synergistic interaction was observed with 2‐BP at 5 μM, whereas 2‐BP at 25 μM produced CI values indicative of synergism with rosiglitazone (Figure 5E). Based on previous reports indicating that inhibition of protein palmitoylation by 2‐BP occurs at concentrations above 10 μM [35], these results demonstrate that effective palmitoylation inhibition is required to overcome adipocyte‐CM–induced resistance to rosiglitazone. Combined rosiglitazone and 2‐BP treatment synergistically suppresses adipocyte‐CM‐induced breast cancer malignancy. (A–D) Cells were treated with vehicle, rosiglitazone (50 μM), or 2‐BP (25 μM) in the absence or presence of 10% adipocyte‐CM for 24 h. (A, B) Transwell invasion assay (left) and immunoblot of PPARγ, p‐AKT, and AKT (right) in MDA‐MB‐231, ZR‐75‐1, and 4T1 cells. Scale bar, 200 μm. (C, D) Representative spheroids images on 3D culture chips after 5 days. Scale bar, 100 μm. (E) Combination index (CI) analysis. MDA‐MB‐231, ZR‐75‐1, and 4T1 cells were cultured in 10% adipocyte‐CM with combinations of rosiglitazone (10, 25, and 50 μM) and 2‐BP (5 and 25 μM) for 72 h. Data are presented as mean ± SD. Unless otherwise indicated, n = 3 independent experiments. Statistical significance was determined by one‐way ANOVA followed by Tukey's HSD post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001, and n.s. > 0.05. Discussion Post‐translational modifications (PTMs) constitute critical regulatory mechanisms that fine‐tune protein function [36]. PPARγ is regulated by multiple PTMs with context‐dependent functional consequences. For instance, acetylation suppresses PPARγ activity in M2 macrophages, whereas it enhances PPARγ activity in adipocytes [37, 38]. Similarly, sumoylation generally reduces PPARγ transcriptional activity in hepatoblastoma, yet it is essential for maintaining PPARγ activity in lung cancer [39, 40]. These context‐dependent effects extend to S‐palmitoylation. Although previous studies in colon cancer reported that it stabilizes PPARγ [41], we demonstrate that palmitic acid from adipocytes induces PPARγ degradation via S‐palmitoylation at C159 in breast cancer (Figure 2). This tissue‐specific divergence underscores the importance of understanding PPARγ regulation within its physiological context. In breast cancer, ZDHHC‐mediated palmitoylation has emerged as an important regulatory mechanism influencing tumor progression and therapeutic response. ZDHHC22‐mediated palmitoylation of mTOR has been reported to suppress breast cancer growth and endocrine therapy resistance [42], whereas ZDHHC20‐mediated palmitoylation of KAT2A promotes lung metastasis in breast cancer [43], highlighting the context‐dependent roles of ZDHHCs in breast cancer. Notably, ZDHHC15 has been implicated in tumor progression across multiple malignancies, including glioma, ovarian cancer, and breast cancer [44, 45]. Our data indicate that ZDHHC15 mediates PPARγ palmitoylation in an adipocyte‐driven context, thereby promoting breast cancer progression (Figure 4), extending the oncogenic role of ZDHHC15 to this adipocyte‐conditioned microenvironment. Adipocytes are specialized for energy storage as triglycerides and for mobilizing these lipid reserves as FFAs through lipolysis [46]. Cancer cells engage in metabolic cross‐talk with adipocytes, stimulating lipolysis and increasing FFA availability to support tumor growth and invasion [47]. Breast cancer, which develops in adipocyte‐rich environments, has been extensively studied in this context [14]. Recent studies have shown that serum FFAs in patients with invasive ductal carcinoma are higher than those in healthy individuals [48]. Among the increased FFAs, palmitic acid (C16:0), stearic acid (C18:0), and linoleic acid (C18:2) have been identified as potential biomarkers for breast cancer [49]. We demonstrate that breast cancer cells exhibit resistance to rosiglitazone under adipocyte‐CM conditions, which is overcome by inhibition of palmitoylation (Figure 5). This may be explained by continuous ZDHHC15‐mediated S‐palmitoylation and degradation of PPARγ, which limits rosiglitazone‐mediated PPARγ activation and suppression of AKT phosphorylation. Consistent with this, synergism between rosiglitazone and 2‐BP was observed only at concentrations sufficient to inhibit protein palmitoylation (Figure 5E). This suggests that elevated palmitic acid in breast cancer patients may contribute to the limited clinical efficacy of rosiglitazone through PPARγ S‐palmitoylation. We demonstrated that adipocyte‐derived palmitic acid downregulates PPARγ via S‐palmitoylation at C159, revealing a pathway linking the adipocyte microenvironment to breast cancer progression (Figure 6). Identifying the ubiquitination site on PPARγ that is targeted following palmitoylation will be an important direction for future study. Although this study is based on in vitro models, the 3D spheroid culture system enhances the physiological relevance of our observations. In conclusion, our findings suggest that targeting ZDHHC15‐mediated PPARγ S‐palmitoylation represents a potential therapeutic strategy for breast cancer. Adipocyte‐derived palmitic acid induces ZDHHC15‐mediated PPARγ S‐palmitoylation at Cys159, triggering its proteasomal degradation and promoting breast cancer malignancy with rosiglitazone resistance. Inhibition of PPARγ S‐palmitoylation by 2‐bromopalmitate suppresses breast cancer progression and synergistically enhances rosiglitazone efficacy. Author Contributions Geon Ho Moon: conceptualization, formal analysis, writing – original draft, investigation. Min Young Lee: data curation, validation, visualization. Jiwon Koh: project administration, resources. Ki Yong Hong: project administration, resources. Junji Fukuda: resources, methodology. Sung Joon Kim: writing – review and editing, resources. Yang‐Sook Chun: supervision, funding acquisition, writing – review and editing. Funding This research was supported by the National Research Foundation of Korea (RS‐2024‐00344706). Ethics Statement Approval of the research protocol by an Institutional Reviewer Board: Approval No. H‐1602‐110‐742, and H‐2207‐198‐1345. Consent All informed consent was obtained from the subject(s) and/or guardian(s). Conflicts of Interest The authors declare no conflicts of interest. Supporting information Figure S1: Downregulation of PPARγ by adipocyte‐derived factors is associated with poor prognosis in breast cancer. (A, B) Transwell invasion assay (left) and immunoblot analysis of PPARγ and Flag protein expression (right) in MDA‐MB 231, ZR‐75‐1, and 4 T1 cells (n = 3). Scale bar, 200 μm. (A) Cells were transfected with si‐control or si‐PPARG (50 nM). (B) Cells were transfected with control or Flag‐PPARγ. (C) Dot plot of PPARγ H‐scores with k‐means clustering–based stratification (R‐package v.4.4.2). Patients were divided into high and low PPARγ groups within each disease category: all carcinomas (high, n = 18; low, n = 24), carcinoma in situ (high, n = 10; low, n = 6), and invasive carcinoma (high, n = 8; low, n = 18). (D) Representative images of spheroids cultured on 3D culture chips in the absence or presence of 10% adipocyte CM at the indicated time points (Day 0, 2, and 5). Scale bar, 200 μm. Data are presented as mean ± SD. Statistical significance was determined using two‐tailed unpaired Student's t‐test. *p < 0.05, **p < 0.01, ***p < 0.001, and n.s. > 0.05. Figure S2: Palmitic acid mediates PPARγ proteasomal degradation via S‐palmitoylation at C159. (A) RT–qPCR analysis of PPARG mRNA levels in MDA‐MB‐231, ZR‐75‐1, and 4 T1 cells (n = 3). Cells were treated with vehicle or PA at concentrations of 100 μM or 300 μM for 24 h. (B) Predicted palmitoylation sites on PPARγ in Homo sapiens (upper) and Mus musculus (lower) using GPS‐Palm algorithm. (C) Luciferase assay of PPARγ activity using a PPRE reporter. (D) Immunoprecipitation analysis for Flag‐PPARγ ubiquitination. Cells transfected with Flag‐PPARγ WT or C159S were treated with vehicle or PA (300 μM) for 24 h. MG132 (10 μM) was added during the last 8 h of PA treatment. Data are presented as mean ± SD. Statistical significance was determined by one‐way ANOVA followed by Tukey's HSD post hoc test. n.s. > 0.05. Figure S3: PPARγ palmitoylation enhances breast cancer aggressiveness via AKT signaling. (A–G) Cells transfected with si‐control or si‐PPARG (50 nM) were treated with vehicle or 2‐BP (25 μM) for 24 h in the absence or presence of PA (A, C, D, and G) or 10% adipocyte‐CM (B, E, and F). (A, B) Transwell invasion assay (left) and immunoblot analysis of PPARγ protein expression (right) in MDA‐MB‐231, ZR‐75‐1, and 4T1 cells (n = 3). Scale bar, 200 μm. (C–F) Representative images of spheroids cultured on 3D culture chips at the indicated time points (Day 0, 2, and 5). Scale bar, 200 μm. (G) RT–qPCR analysis of CCND1, CCNE1, PTEN, VIM, CDH2 and SNAI1 mRNA levels (n = 3). Data are presented as mean ± SD. Statistical significance was determined by one‐way ANOVA followed by Tukey's HSD post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001, and n.s. > 0.05. Figure S4: Identification of candidate enzymes responsible for PPARγ S‐palmitoylation. (A, B) Cells were transfected with siRNAs targeting 23 ZDHHCs (50 nM). RT–qPCR analysis of ZDHHCs (A) (n = 3). Immunoblot analysis of PPARγ (B). (C) Cells transfected with si‐control, si‐ZDHHC14, si‐ZDHHC15, or si‐ZDHHC16 (50 nM) were treated with PA for 24 h. ABE assay for PPARγ S‐palmitoylation. MG132 (10 μM) was added during the last 8 h of PA treatment. Data are presented as mean ± SD. Statistical significance was determined using two‐tailed unpaired Student's t‐test. **p < 0.01 and ***p < 0.001. Figure S5: Effects of ZDHHC15 knockdown on adipocyte‐CM‐driven proliferation and migration are dependent on PPARγ. (A–C) Cells transfected with si‐control, si‐ZDHHC15 (50 nM), or si‐PPARG (50 nM) were treated with 10% adipocyte‐CM for 24 h. (A) Transwell invasion assay (left) and immunoblot analysis of PPARγ and ZDHHC15 protein expression (right) in MDA MB‐231, ZR‐75‐1, and 4 T1 cells (n = 3). Scale bar, 200 μm. (B–C) Representative images of spheroids cultured on 3D culture chips at the indicated time points (Day 0, 2, and 5). Scale bar, 200 μm. Data are presented as mean ± SD. Statistical significance was determined by one‐way ANOVA followed by Tukey's HSD post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001, and n.s. > 0.05. Figure S6: Effects of rosiglitazone and 2‐BP in breast cancer cells. (A) Representative images of spheroids cultured on 3D culture chips at the indicated time points (Day 0, 2, and 5). Scale bar, 200 μm. Cells were treated with vehicle, rosiglitazone (50 μM), or 2‐BP (25 μM) in the absence or presence of 10% adipocyte‐CM for 24 h. (B) IC50 determination of rosiglitazone and 2‐BP. MDA‐MB‐231, ZR‐75‐1, and 4 T1 cells were cultured in 10% adipocyte‐CM with vehicle, rosiglitazone (10, 25, and 50 μM), or 2‐BP (5 and 25 μM) for 72 h. Cell viability was assessed relative to vehicle control (n = 3). Data are presented as mean ± SD. Table S1: Reagents used in experiments. Table S2: Nucleotide sequences of siRNAs used in experiments. Table S3: List of antibodies used in experiments. Table S4: Nucleotide sequences of primers used in experiments. Table S5: Nucleotide sequences of site‐directed mutagenesis primers used in experiments. Table S6: KEGG pathway analysis of normal breast tissue and breast cancer. Table S7: Breast cancer cell lines used in experiments.