CIDEC impairs mitochondrial fitness and blocks protective autophagy via the inhibition of cGMP/PKG pathway to exert tumor-suppressive effects in breast cancer.

CIDEC impairs mitochondrial fitness and blocks protective autophagy via the inhibition of cGMP/PKG pathway to exert tumor-suppressive effects in breast cancer - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Cell Oncol (Dordr) . 2026 Apr 20;49(3):78. doi: 10.1007/s13402-026-01211-8 Search in PMC Search in PubMed View in NLM Catalog Add to search CIDEC impairs mitochondrial fitness and blocks protective autophagy via the inhibition of cGMP/PKG pathway to exert tumor-suppressive effects in breast cancer Xuchu Jin Xuchu Jin 1 Department of Breast Surgery, Sichuan Provincial People’s Hospital, School of Medicine, University of Electronic Science and Technology of China, Chengdu, 610072 China 2 Department of Breast Surgery & Lab 1, Cancer Institute, First Affiliated Hospital, China Medical University, No. 155, North Nanjing Street, Shenyang, 110001 China Find articles by Xuchu Jin 1, 2, # , Yu Liu Yu Liu 3 Department of Ophthalmology, Sichuan Provincial People’s Hospital, School of Medicine, University of Electronic Science and Technology of China, Chengdu, 610072 China Find articles by Yu Liu 3, # , Xinyu Zheng Xinyu Zheng 2 Department of Breast Surgery & Lab 1, Cancer Institute, First Affiliated Hospital, China Medical University, No. 155, North Nanjing Street, Shenyang, 110001 China Find articles by Xinyu Zheng 2, ✉ , Yangke He Yangke He 4 Department of Oncology, Sichuan Provincial People’s Hospital, School of Medicine, University of Electronic Science and Technology of China, Chengdu, 610072 China Find articles by Yangke He 4, ✉ Author information Article notes Copyright and License information 1 Department of Breast Surgery, Sichuan Provincial People’s Hospital, School of Medicine, University of Electronic Science and Technology of China, Chengdu, 610072 China 2 Department of Breast Surgery & Lab 1, Cancer Institute, First Affiliated Hospital, China Medical University, No. 155, North Nanjing Street, Shenyang, 110001 China 3 Department of Ophthalmology, Sichuan Provincial People’s Hospital, School of Medicine, University of Electronic Science and Technology of China, Chengdu, 610072 China 4 Department of Oncology, Sichuan Provincial People’s Hospital, School of Medicine, University of Electronic Science and Technology of China, Chengdu, 610072 China ✉ Corresponding author. # Contributed equally. Received 2025 Nov 29; Accepted 2026 Apr 9; Collection date 2026 Jun. © The Author(s) 2026, modified publication 2026 Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ . PMC Copyright notice PMCID: PMC13096410  PMID: 42010152 Abstract Objective Breast cancer continues to be a major contributor to cancer-associated deaths among the female population globally. This study aims to investigate the functional role, underlying mechanisms, and clinical relevance of Cell death-inducing DFFA-like effector C (CIDEC) in breast cancer pathogenesis. Methods Integrated bioinformatics analysis of three gene expression datasets identified hub genes via protein-protein interaction network and multiple machine learning algorithms. The tumor-suppressive effects of CIDEC were evaluated in vitro using breast cancer cell lines by assessing viability, proliferation, migration, invasion, and apoptosis, and in vivo via a xenograft model. Mitochondrial function, autophagy, and the cyclic guanosine monophosphate (cGMP)/protein kinase G (PKG) signaling pathway were assessed using a pathway agonist and an autophagy inhibitor. Results A marked reduction in CIDEC expression was observed in breast cancer tissues and cellular models. CIDEC effectively curtailed tumor progression by impeding proliferation, migration, and invasive capacity, coupled with the induction of apoptotic cell death. Mechanistically, CIDEC impaired mitochondrial fitness, characterized by reduced adenosine triphosphate (ATP) production, dissipated mitochondrial membrane potential, and elevated reactive oxygen species. Concurrently, CIDEC blocked protective autophagy. These effects were mediated through the suppression of the cGMP/PKG pathway. Activating this pathway with 8-Br-cGMP reversed the tumor-suppressive phenotypes and mitochondrial dysfunction induced by CIDEC, whereas inhibiting autophagy attenuated this rescue. Conclusion Our findings demonstrate that CIDEC functions as a novel tumor suppressor in breast cancer by disrupting mitochondrial fitness and inhibiting protective autophagy via the cGMP/PKG pathway. The CIDEC-cGMP/PKG axis represents a promising therapeutic target for breast cancer intervention. Clinical trial number Not applicable. Supplementary Information The online version contains supplementary material available at 10.1007/s13402-026-01211-8. Keywords: CIDEC, Breast cancer, cGMP/PKG pathway, Mitophagy Introduction According to global cancer epidemiology, breast cancer poses a substantial public health challenge as the predominant malignancy and primary contributor to cancer mortality in the female population worldwide [ 1 , 2 ]. The disease’s high mortality is primarily attributable to tumor metastasis, recurrence, and the development of resistance to conventional therapies, including chemotherapy, endocrine therapy, and targeted agents [ 3 – 6 ]. Although significant strides have been made in early detection and treatment, these limitations persistently undermine clinical outcomes for a substantial number of patients, particularly those with advanced or triple-negative breast cancer [ 7 , 8 ]. Therefore, unraveling the intricate molecular underpinnings of breast cancer progression and therapeutic resistance is imperative for identifying novel prognostic biomarkers and developing more effective treatment strategies. The cell death-inducing DFFA-like effector (CIDE) family proteins, including CIDEA, CIDEB, and CIDEC (also known as FSP27), are endoplasmic reticulum- and lipid droplet-associated proteins that have emerged as crucial regulators of lipid homeostasis, with their dysregulation contributing to various metabolic disorders and cancer development [ 9 ]. CIDEC is predominantly characterized by its role in lipid storage and metabolism within adipocytes [ 10 ]. CIDEC has been well-characterized in metabolic disorders such as non-alcoholic fatty liver disease and type 2 diabetes [ 11 , 12 ]. Its function in cancer, however, remains less extensively explored. Supporting its potential relevance in oncology, analysis of breast tumors from Indian women reveals that CIDEC is significantly downregulated and potentially contributes to tumorigenesis through its involvement in dysregulated lipid metabolism pathways [ 13 ]. Nevertheless, direct experimental evidence validating its biological functions and delineating the precise mechanistic pathways through which it might inhibit tumor development in breast cancer remains scarce, creating a significant gap in our understanding of its pathophysiological relevance. Mitochondrial homeostasis, governed by the precise balance between mitochondrial biogenesis and mitophagy, represents a fundamental regulatory axis for maintaining cellular fitness [ 14 ]. These coordinated processes dynamically respond to cellular energy status and environmental cues to ensure optimal mitochondrial quality and function [ 15 , 16 ]. Notably, both mitochondrial fitness and autophagy play pivotal roles in cellular homeostasis, and their dysregulation is increasingly recognized as a hallmark of cancer pathogenesis [ 17 ]. In particular, the interplay between mitochondrial dysfunction—characterized by compromised energy production, membrane potential dissipation, and oxidative stress—and altered autophagic activity constitutes a critical node in tumor biology [ 18 , 19 ]. Mitophagy, the selective autophagic clearance of damaged or superfluous mitochondria, represents a crucial quality control mechanism for maintaining mitochondrial fitness [ 20 ]. Concurrently, autophagy—a lysosomal degradation pathway—plays a dual role in cancer: basal autophagy maintains cellular homeostasis, whereas pathological activation can promote tumor growth via metabolic adaptation [ 21 ]. Protective autophagy serves as a critical survival mechanism in breast cancer that promotes chemoresistance and targeted therapy resistance by maintaining cellular homeostasis under treatment stress [ 22 – 24 ]. However, the mechanistic insights into how CIDEC coordinates this mitochondrial-mitophagy axis remain largely elusive. Based on these premises, this research systematically examined CIDEC’s tumor-suppressive role in breast cancer. Bioinformatic analyses pinpointed the cyclic guanosine monophosphate (cGMP)/protein kinase G (PKG) pathway - a key regulator of cardiovascular homeostasis [ 25 ], as a key downstream pathway of CIDEC. We therefore sought to determine whether CIDEC exerts its tumor-suppressive function through impairing mitochondrial fitness and blocking protective autophagy via this specific pathway. Employing integrated methodologies spanning bioinformatic analysis and functional assays, we delineated the molecular mechanisms through which CIDEC orchestrates mitochondrial homeostasis and autophagic activity via the cGMP/PKG pathway. Our investigation sought to validate the CIDEC-cGMP/PKG-mitochondria-autophagy axis as a candidate for therapeutic targeting in breast cancer. Materials and methods Microarray data acquisition and differential expression analysis Three gene expression microarray datasets ( GSE29431 , GSE42568 , and GSE65194 ) were sourced from the Gene Expression Omnibus (GEO) public repository ( https://www.ncbi.nlm.nih.gov/geo/ ). The GEO2R interactive web tool ( www.ncbi.nlm.nih.gov/geo/geo2r ), which implemented the limma R package for statistical analysis, was employed to conduct differential expression analysis for each dataset. Differentially expressed genes (DEGs) were defined using stringent criteria: an adjusted p-value (adj.P.Val) < 0.01 based on the Benjamini & Hochberg false discovery rate (FDR) method, coupled with an absolute log 2 fold change (|logFC|) ≥ 3. The ggplot2 R package was employed for generating volcano plots to display differential expression results, while the pheatmap package was applied to visualize expression patterns of the top significant DEGs in heatmap format. To identify common DEGs across the three datasets, the lists of significant DEGs from GSE29431 , GSE42568 , and GSE65194 were intersected. A Venn diagram was generated to illustrate the overlaps using the online tool Draw Venn Diagram ( http://bioinformatics.psb.ugent.be/webtools/Venn/ ). Gene ontology (GO) and kyoto encyclopedia of genes and genomes (KEGG) enrichment analysis A comprehensive functional characterization of the common DEGs was conducted using the Database for Annotation, Visualization and Integrated Discovery (DAVID, https://david.ncifcrf.gov ). This encompassed GO enrichment analysis across all three categories (Biological Process [BP], Cellular Component [CC], Molecular Function [MF]) and KEGG pathway analysis. Enrichment terms with p-values below 0.05 were deemed statistically significant, and all results were subsequently processed and visualized using the R environment. For the GO analysis, the results from the three categories were combined and sorted by p-value in ascending order. The top 6 most significant GO terms from this sorted list were selected for visualization. Similarly, for the KEGG analysis, all significant pathways were sorted by p-value, and the top 8 most significant pathways were selected. Dot plots for both GO and KEGG results were generated using the ggplot2 package. Protein-protein interaction (PPI) network construction The Search Tool for the Retrieval of Interacting Genes (STRING, https://string-db.org ) was employed to construct a PPI network for the common DEGs. The list of common DEGs was submitted, and the required interactions were retrieved. The resulting network data from STRING, containing the interaction pairs, was then imported into the Cytoscape software for further visualization and analysis. The Molecular Complex Detection (MCODE) algorithm, an integrated plugin within the Cytoscape environment, was employed to detect densely interconnected clusters in the PPI network. The module with the highest MCODE score was selected as the most significant functional cluster. All network visualizations were directly generated and exported from the Cytoscape platform. Identification and validation of hub genes The most significant module from the PPI network, identified by the MCODE algorithm, was selected for further analysis. The genes within this highest-scoring module were defined as candidate hub genes. The diagnostic utility of the hub genes was assessed through receiver operating characteristic (ROC) curve analysis. For each gene, ROC curves were independently generated using raw expression data from the three datasets ( GSE29431 , GSE42568 , GSE65194 ) with the pROC R package. The area under the curve (AUC) was then computed to quantify their discriminatory capacity between tumor and normal samples. Additional validation of hub gene differential expression was conducted through the GEPIA2 web platform ( http://gepia2.cancer-pku.cn/ ). Box plots were generated by comparing expression levels in breast cancer samples from The Cancer Genome Atlas (TCGA) database with normal breast tissue data pooled from both TCGA and the Genotype-Tissue Expression (GTEx) project. Furthermore, to visualize the co-expression relationships among the hub genes, a correlation matrix heatmap was constructed. This was done using the expression data from the GSE42568 cohort. The pheatmap R package was employed for this purpose, which included hierarchical clustering of both genes and samples based on Pearson correlation coefficients. Machine learning validation of hub genes in individual datasets To further validate the robustness of the candidate hub genes, a comprehensive machine learning analysis was performed independently on each of the three gene expression datasets ( GSE29431 , GSE42568 , and GSE65194 ) using multiple feature selection algorithms. Three machine learning methodologies were utilized for feature selection. The Least Absolute Shrinkage and Selection Operator (LASSO) regression analysis was implemented using the glmnet R package with 10-fold cross-validation to select the optimal lambda value and identify non-redundant features. The Support Vector Machine-Recursive Feature Elimination (SVM-RFE) algorithm was executed using the e1071 R package, and the feature subset size that yielded the minimum cross-validation error was selected as optimal. Additionally, the Boruta feature selection algorithm was performed using the Boruta R package, which employs a random forest-based approach to identify all relevant features by comparing original attributes with shadow features. The final set of validated hub genes for each dataset was determined by taking the intersection of the genes selected by all three independent methods (LASSO, SVM-RFE, and Boruta). This stringent consensus approach ensured the identification of the most robust and methodologically consistent key genes in each individual cohort. Prognostic analysis of hub gene expression For prognostic analysis, the Kaplan-Meier Plotter online tool ( https://kmplot.com/analysis/ ) was utilized to assess the association between hub gene expression and patient survival outcomes. A multi-dimensional survival analysis was conducted based on Overall Survival (OS, n = 1880), Relapse-Free Survival (RFS, n = 4934), and Post-Progression Survival (PPS, n = 458). Utilizing the platform’s auto-selected optimal cutoff values, we categorized patients into high- and low-expression subgroups for each gene. Kaplan-Meier survival analysis was then employed to generate survival curves, while group differences were statistically evaluated with the log-rank test. Gene set enrichment analysis (GSEA) of hub genes Pathway profiling for each hub gene was performed through GSEA using MSigDB’s curated pathway sets (c2.cp.kegg.v7.4.symbols.gmt) as the reference database. For each hub gene in the training sets ( GSE29431 , GSE42568 , GSE65194 ), a pre-ranked list of all other genes was generated based on Spearman’s correlation coefficients calculated using the psych R package. These gene lists, ranked from the highest positive correlation to the highest negative correlation with the respective hub gene, were then subjected to GSEA using the clusterProfiler R package. The normalized enrichment score (NES) was used to indicate the direction and magnitude of the enrichment. For each hub gene, the top 10 most significantly enriched KEGG pathways, sorted by p-value, were selected for visualization. Immune infiltration analysis The relative abundances of immune cells in each sample from the GSE29431 , GSE42568 , and GSE65194 datasets were estimated using the CIBERSORT algorithm. The estimated immune cell fractions for all qualified samples across the three datasets were then visualized as stacked bar charts using the ggplot2 R package. Immune infiltration patterns relative to hub gene expression were assessed, where samples were categorized into expression-based subgroups for subsequent comparative analysis of immune cell composition via Wilcoxon rank-sum tests. The results of these analyses were presented in separate box plots for each hub gene, generated using the ggplot2 R package. Screening for downstream regulatory mechanisms To investigate the downstream molecular mechanisms of CIDEC in breast cancer, a systematic bioinformatic analysis was performed. First, genes implicated in breast cancer pathogenesis were retrieved from the Genecards database ( https://www.genecards.org/ ) using “breast cancer” as the search term. Simultaneously, the Coxpresdb database ( https://coxpresdb.jp ) was employed to identify genes exhibiting significant co-expression patterns with CIDEC. The intersection of these two gene sets was identified and visualized using a Venn diagram, revealing the core set of breast cancer-related genes consistently co-expressed with CIDEC. The resulting list of CIDEC-coexpressed genes relevant to breast cancer was subsequently subjected to pathway enrichment analysis using the DAVID bioinformatics resource. KEGG pathway analysis was performed with a significance threshold of p < 0.05. The most significantly enriched pathways were selected and visualized as a bubble plot using the ggplot2 R package. Cell culture and transfection Cellular experiments utilized the MCF-10 A normal mammary epithelial cell line along with MCF-7 and MDA-MB-231 breast cancer lines, obtained from iCell bioscience Biotechnology Co., Ltd (Shanghai, China). Culture was performed using Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin under standard culture conditions (37 °C, 5% CO₂). To generate stable CIDEC-overexpressing cells, MCF-7 and MDA-MB-231 cells were transfected. The full-length human CIDEC cDNA sequence was cloned into a lentiviral vector. The primer sequences used for cloning were as follows: forward: 5’- CGCGGATCCATGAGAAACATGGAGT − 3’; reverse: 5’- CCGCTCGAGTGCAGTATCTTCAGAC − 3’. Lentiviral particles were produced and used to infect the target cells according to standard protocols. Lipofectamine 3000™ reagent (ThermoFisher, Waltham, Massachusetts, USA) was applied for transient transfections per manufacturer’s guidelines. Parallel transfections with empty vectors were conducted as negative controls (oe-NC). For subsequent functional assays, the experimental groups were established as follows. The Control group consisted of untreated cells, while the oe-NC group comprised cells transduced with an empty vector negative control. The oe-CIDEC group contained cells stably overexpressing CIDEC. Pharmacological interventions included the 8-Br-cGMP group, where untreated cells were treated with 100 µM 8-Br-cGMP for 24 h [ 26 ], and the oe-CIDEC + 8-Br-cGMP group, where CIDEC-overexpressing cells were co-treated with 100 µM 8-Br-cGMP for 24 h. To investigate autophagic involvement, the oe-CIDEC + 8-Br-cGMP + 3-MA group involved CIDEC-overexpressing cells co-treated with 100 µM 8-Br-cGMP and 5 mM 3-Methyladenine (3-MA) for 24 h [ 27 ]. For autophagic flux analysis, additional groups were included: the oe-NC + CQ group, where oe-NC-transfected cells were treated with 25 µM chloroquine (CQ) for 24 h, and the oe-CIDEC + CQ group, where CIDEC-overexpressing cells were treated with 25 µM CQ for 24 h, based on established protocols [ 28 ]. Cell counting kit-8 (CCK-8) assay Cellular viability was assessed with the CCK-8 assay kit (Beyotime Biotechnology, Beijing, China) per manufacturer’s specifications. Briefly, cells were seeded in 96-well plates at a density of 1 × 10⁴ cells/well and permitted to adhere for 24 h. Post-treatment, 10 µL of CCK-8 solution was introduced to each well containing 90 µL of fresh medium. After 2 h of incubation at 37 °C, absorbance at 450 nm was quantified using a microplate reader. 5-Ethynyl-2’-deoxyuridine (EdU) assay Proliferation rates were further determined through EdU incorporation analysis. In brief, cells were plated in 96-well plates and allowed to attach overnight. Following treatment exposure, 10 µM EdU was administered for a 2-hour incubation at 37 °C. Cell fixation was performed with 4% paraformaldehyde (15 min) followed by permeabilization with 0.5% Triton X-100 (20 min, room temperature). Fluorescent detection of incorporated EdU was conducted according to the manufacturer’s protocol, with 4’,6-diamidino-2-phenylindole (DAPI) serving as nuclear counterstain. Image acquisition was carried out using a Nikon microscope (Tokyo, Japan). For quantification, the percentage of EdU-positive cells was calculated by dividing the number of EdU-positive cells by the total number of DAPI-stained nuclei in each field, with at least five randomly selected fields analyzed per condition. Wound healing assay Cell migration was assessed using a wound healing approach. After cells reached near confluence in 6-well plates, a straight wound was generated by scraping the monolayer with a sterile pipette tip. The plates were washed to remove displaced cells and incubated with serum-free medium for 24 h to minimize the contribution of cell proliferation to wound closure. Wound areas were imaged immediately (0 h) and after 24 h using an inverted microscope. Migration was quantified by calculating the percentage of wound closure based on the reduction in wound area over time using ImageJ software. Cell invasion assay Invasion assays were conducted using Matrigel-coated Transwell inserts. Cells (1 × 10⁵) in serum-free medium were added to the upper chambers, with 15% FBS medium as a chemoattractant below. After 24 h, non-invading cells were removed. Invaded cells were fixed, stained with 0.1% crystal violet, and subsequently quantified microscopically using ImageJ software. Assessment of apoptosis by flow cytometry Apoptosis detection was performed by flow cytometry employing Annexin V-fluorescein isothiocyanate (FITC)/propidium iodide (PI) staining (Beyotime Biotechnology). Cells were harvested by gentle trypsinization, washed, and resuspended in binding buffer before staining with Annexin V-FITC and PI for 15 min in the dark. Flow cytometric analysis was conducted on a Cytoflex system (Cytoflex; Beckman Coulter, Brea, CA, USA), and results were processed with FlowJo software. Measurement of mitochondrial membrane potential (MMP) Mitochondrial membrane potential was assessed using the 5,5’,6,6’-tetrachloro-1,1’,3,3’-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) Mitochondrial Membrane Potential Assay Kit (Servicebio, Wuhan, China; G1515-100T) according to the manufacturer’s instructions. Briefly, JC-1 staining working solution was prepared by mixing 2 µL of JC-1 (500×) with 900 µL of JC-1 dilution buffer, followed by addition of 100 µL of JC-1 staining buffer (10×) and thorough vortexing. Post-treatment MCF-7 cells were washed once with 1× JC-1 staining buffer, then incubated with 1 mL of culture medium (containing serum and phenol red) mixed with 1 mL of JC-1 working solution for 15–30 min at 37 °C in a CO₂ incubator protected from light. After incubation, the supernatant was removed, and cells were washed twice with 1× JC-1 staining buffer. Finally, 2 mL of 1× JC-1 staining buffer or phenol red-free culture medium was added, and images were acquired immediately using a fluorescence microscope. Detection parameters were set as follows: JC-1 monomers (green fluorescence) were detected at Ex = 514 nm, Em = 529 nm; JC-1 aggregates (red fluorescence) were detected at Ex = 585 nm, Em = 590 nm. Membrane polarization states were distinguished by fluorescence emission: red fluorescence indicates JC-1 aggregates in polarized mitochondria with high membrane potential, while green fluorescence indicates JC-1 monomers in depolarized mitochondria with low membrane potential. Quantitative analysis was performed by calculating the ratio of red to green fluorescence intensity using ImageJ software, with at least five randomly selected fields analyzed per condition. Measurement of intracellular reactive oxygen species (ROS) levels Intracellular ROS levels were assessed using the fluorescent probe 2’,7’-dichlorodihydrofluorescein diacetate (DCFH-DA). After experimental treatments, cells were loaded with 25 µM DCFH-DA and incubated at 37 °C for 30 min under light-protected conditions. Following phosphate-buffered saline washes to eliminate excess probe, fluorescence images were immediately acquired by fluorescence microscopy. Quantitative analysis of fluorescence intensity, representing ROS levels, was performed with ImageJ software, with at least five randomly selected fields analyzed per condition. Transmission electron microscopy (TEM) Autophagic structures were analyzed using TEM. Cells underwent primary fixation with glutaraldehyde in phosphate buffer (2 h, 4 °C) and secondary fixation with 1% osmium tetroxide (1 h). Following sequential ethanol dehydration, samples were embedded in epoxy resin. Ultrathin sections were obtained using an ultramicrotome and subsequently double-stained with uranyl acetate and lead citrate for TEM observation. For each sample, at least five randomly selected fields were examined, and representative images were captured. Xenograft tumor model All experimental procedures involving animals were carried out in compliance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals and received approval from our institution’s Institutional Animal Care and Use Committee. Female BALB/c nude mice (6–8 weeks old, 18–20 g) sourced from SPF Biotechnology Co., Ltd (Beijing, China) were housed under specific pathogen-free conditions throughout the study. Mice were randomly allocated into two experimental groups ( n ≥ 6 per group): the oe-NC group received subcutaneous injections of empty vector-transfected MCF-7 cells, while the oe-CIDEC group was injected with CIDEC-overexpressing MCF-7 cells. All injections contained 4 × 10⁶ MCF-7 cells suspended in a 1:1 mixture of serum-free DMEM and Matrigel with a total volume of 100 µL. Tumor dimensions were measured weekly using digital calipers, and volume was calculated as (length × width²)/2. After 28 days, mice were euthanized by overdose inhalation of isoflurane (5% in oxygen at a 2 L/min flow rate for 5 min) to ensure humane endpoints. Tumors were then excised for weight measurement and subsequent histological analysis. Histological processing, immunohistochemistry, and apoptosis detection Tumor specimens were processed through conventional histological techniques, including fixation with 4% paraformaldehyde, serial ethanol dehydration, and paraffin embedding. Subsequently, 4 μm sections were prepared and subjected to hematoxylin and eosin (H&E) staining according to established protocols for morphological evaluation. For immunohistochemical analysis, after antigen retrieval and blocking, sections were incubated with anti-Ki67 primary antibody (1:400 dilution; Abcam, Cambridge, UK, ab15580) overnight at 4 °C, followed by incubation with horseradish peroxidase (HRP)-conjugated goat anti-rabbit secondary antibody (1:200; Abcam, ab205718) for 1 h at room temperature. The immunoreactivity was visualized using 3,3’-diaminobenzidine (DAB) substrate followed by hematoxylin counterstaining. Ki67 expression was assessed in at least five randomly selected fields per section. Apoptotic cells were identified through the Terminal Deoxynucleotidyl Transferase dUTP Nick End Labeling (TUNEL) assay following the supplier’s recommended protocol. TUNEL-positive cells were counted in at least five randomly selected fields per section, with three sections per tumor analyzed using ImageJ software. Enzyme-linked immunosorbent assay (ELISA) Cellular adenosine triphosphate (ATP) content and tumor tissue cyclic guanosine monophosphate (cGMP) levels were measured using commercial ELISA kits (Esebio, Shanghai, China) according to the manufacturer’s instructions. After incubation with the detection antibody and washing, color development was carried out, followed by reaction termination. Absorbance was measured at 450 nm using a microplate reader. Real-time quantitative polymerase chain reaction (RT-qPCR) Total RNA was isolated with TRIzol reagent (Thermo Fisher Scientific) following the manufacturer’s instructions. After assessing RNA quality spectrophotometrically, 1 µg of RNA was reverse-transcribed into cDNA using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific). Quantitative PCR was carried out on an ABI 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) with SYBR Green Master Mix, using the following thermal cycling parameters: 95 °C for 10 min, followed by 36 cycles of 95 °C for 10 s and 60 °C for 30 s. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) served as the endogenous reference gene, and relative gene expression was determined using the 2 −ΔΔCt method with triple technical replicates. Corresponding primer sequences were documented in Supplementary Table 1. Western blot analysis Total protein was extracted from cells using Radioimmunoprecipitation Assay (RIPA) lysis buffer. Protein concentration was determined using the Bicinchoninic Acid (BCA) Protein Assay Kit (Thermo Fisher Scientific). Equal amounts of protein were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene fluoride (PVDF) membranes. After blocking with 5% non-fat milk for 1 h at room temperature, the membranes were incubated overnight at 4 °C with the following primary antibodies: β-actin (1:1000, ab8226, Abcam), CIDEC (1:1000, ab198204, Abcam), soluble guanylate cyclase (sGC, 1:1000, ab154841, Abcam), protein kinase G1 (PRKG1, 1:1000, 21646-1-AP, Proteintech, Chicago, IL, USA), protein kinase G2 (PRKG2, 1:1000, 55138-1-AP, Proteintech), phosphorylated vasodilator-stimulated phosphoprotein at Ser239 (p-VASP [Ser239], 1:2000, 3114, Cell Signaling Technology, Danvers, MA, USA), vasodilator-stimulated phosphoprotein (VASP, 1:2000, 3132, Cell Signaling Technology), dynamin-related protein 1 (DRP1, 1:1000, ab184247, Abcam), mitofusin 1 (MFN1, 1:1000, ab191853, Abcam), mitofusin 2 (MFN2, 1:1000, ab124773, Abcam), optic atrophy 1 (OPA1, 1:1000, ab157457, Abcam), Beclin-1 (1:1000, ab207612, Abcam), microtubule-associated protein 1 light chain 3 beta (LC3B, 1:1000, ab192890, Abcam), and sequestosome 1 (p62, 1:1000, ab109012, Abcam). Following washes with Tris-Buffered Saline with Tween-20 (TBST), the membranes were incubated with appropriate HRP-conjugated secondary antibodies: goat anti-rabbit immunoglobulin G (IgG) (1:5000, ab6721, Abcam) for rabbit primary antibodies and goat anti-mouse IgG (1:5000, ab6789, Abcam) for mouse primary antibodies, for 1 h at room temperature. Protein bands were visualized using enhanced chemiluminescence substrate on a Tanon 5200 imaging system (Shanghai, China). β-actin served as the loading control. Statistical analysis Data analysis was conducted using GraphPad Prism 8.0 and SPSS 20.0 software packages. For all in vitro experiments, at least three independent biological replicates were performed ( n ≥ 3). For the in vivo xenograft study, each group contained a minimum of six mice ( n ≥ 6 per group). Continuous variables were expressed as mean ± standard deviation. Group comparisons were analyzed by Student’s t-test (two groups) or one-way ANOVA with Tukey’s post-hoc test (multiple groups). Statistical significance was defined as p < 0.05 for all analyses. Results Identification of DEGs and GO/KEGG enrichment analysis Differential gene expression analysis was performed on the gene expression datasets GSE29431 , GSE42568 , and GSE65194 , comparing normal breast tissue samples to breast cancer tumor samples. In the GSE29431 dataset (12 Normal vs. 54 Tumor), a total of 195 DEGs were identified, comprising 13 upregulated and 182 downregulated genes. Analysis of the GSE42568 dataset (17 Normal vs. 104 Tumor) yielded 416 DEGs, with 119 being upregulated and 297 downregulated. The GSE65194 dataset (11 Normal vs. 153 Tumor) produced the largest set, with 843 DEGs, of which 590 were upregulated, and 253 were downregulated. Hierarchical clustering of the DEGs from all three datasets was visualized, generating a combined volcano plot (Supplementary Fig. 1A). Furthermore, to illustrate the expression patterns of the most significantly altered genes, heatmaps were constructed displaying the top 10 upregulated and top 10 downregulated genes from each individual dataset (Supplementary Fig. 1B). The gene lists corresponding to these top 10 up- and down-regulated DEGs for each dataset are provided in Supplementary Tables 2–4. A Venn diagram was employed to determine the common DEGs across the three studies. This analysis revealed a core set of 93 consistently DEGs, with 86 genes being commonly downregulated and only 7 genes being commonly upregulated (Supplementary Fig. 1C). This overlapping set of 93 DEGs was subsequently subjected to functional enrichment analysis. GO enrichment analysis, categorized into BP, CC, and MF, was performed. The BP terms were significantly enriched in processes including intestinal absorption, lipid storage, cholesterol homeostasis, cellular response to insulin stimulus, response to bacterium, and glucose homeostasis. The significantly enriched CC terms primarily involved the lipid particle, collagen trimer, basolateral plasma membrane, cell surface, extracellular space, and plasma membrane. For the MF category, significant enrichment was observed in activities such as phenanthrene 9,10-monooxygenase activity, dihydrotestosterone 17-beta-dehydrogenase activity, fatty acid binding, peptide hormone binding, macromolecular complex binding, and protein homodimerization activity. The six most significantly enriched terms from each of these three GO categories (BP, CC, MF) were presented in an enrichment bubble plot (Supplementary Fig. 1D). KEGG pathway enrichment analysis further indicated that these common DEGs were predominantly involved in several key pathways. The most significantly enriched pathways included proximal tubule bicarbonate reclamation, the PPAR signaling pathway, tyrosine metabolism, the adipocytokine signaling pathway, regulation of lipolysis in adipocytes, pyruvate metabolism, the AMPK signaling pathway, and protein digestion and absorption. The eight most significantly enriched pathways are summarized in a corresponding bubble plot (Supplementary Fig. 1E). PPI network construction and identification of hub genes To elucidate the functional interactions among the common DEGs, a PPI network was constructed using the STRING database and subsequently visualized in Cytoscape (Supplementary Fig. 2A). The network was first analyzed using the MCODE algorithm to identify densely connected clusters. The module with the highest score (Module 1) was selected, and the genes within this module were considered key functional units. In a parallel approach, the top six hub genes were identified using the Degree method from the CytoHubba plugin. The intersection of genes derived from these two independent algorithms yielded a final set of six hub genes: Leptin (LEP), Cluster of Differentiation 36 (CD36), Adiponectin, C1Q and Collagen Domain Containing (ADIPOQ), Perilipin 1 (PLIN1), CIDEC, and Fatty Acid Binding Protein 4 (FABP4). The corresponding visualizations for the highest-scoring MCODE module and the Degree-based hub gene network were directly generated from Cytoscape (Supplementary Fig. 2B-C). Machine learning validation of hub genes in individual datasets To further validate the robustness of the six candidate hub genes, a comprehensive machine learning analysis was performed independently across the three original datasets ( GSE29431 , GSE42568 , and GSE65194 ) using multiple feature selection algorithms. In the GSE29431 dataset, three distinct methods were employed. LASSO regression analysis, conducted using the “glmnet” R package with 10-fold cross-validation, identified two key genes: LEP and CIDEC (Supplementary Fig. 3A). Simultaneously, the SVM-RFE algorithm, implemented via the “e1071” R package, also selected LEP and CIDEC as the most significant features based on the point of minimum error (Supplementary Fig. 3B). The Boruta algorithm, using the “Boruta” R package, confirmed five important genes: LEP, ADIPOQ, PLIN1, CIDEC, and FABP4 (Supplementary Fig. 3C). The intersection of the results from these three methods yielded a final set of two critical genes for this dataset: LEP and CIDEC (Supplementary Fig. 3D). Analysis of the GSE42568 dataset produced a similar yet distinct profile. LASSO regression selected five genes: LEP, ADIPOQ, PLIN1, CD36, and FABP4 (Supplementary Fig. 4A). The SVM-RFE algorithm identified four genes: LEP, PLIN1, CIDEC, and FABP4 (Supplementary Fig. 4B). The Boruta algorithm confirmed all six candidate genes as relevant features (Supplementary Fig. 4C). The intersection of these three lists pinpointed three core genes for the GSE42568 dataset: LEP, PLIN1, and FABP4 (Supplementary Fig. 4D). For the GSE65194 dataset, LASSO regression analysis selected ADIPOQ, CD36, FABP4, LEP, and PLIN1 (Supplementary Fig. 5A). The SVM-RFE algorithm selected all six candidate genes (Supplementary Fig. 5B), a result that was corroborated by the Boruta algorithm (Supplementary Fig. 5C). Consequently, the intersection of the three methods identified five key genes for this dataset: ADIPOQ, CD36, FABP4, LEP, and PLIN1 (Supplementary Fig. 5D). Finally, the union of the key genes validated across all three independent datasets was taken. This integrative step confirmed a consistent set of six reliable hub genes: LEP, ADIPOQ, CD36, PLIN1, CIDEC, and FABP4. Comprehensive analysis of the identified hub genes To further characterize the six hub genes (LEP, CD36, ADIPOQ, PLIN1, CIDEC, FABP4), a series of in-depth analyses was performed. First, a correlation matrix heatmap was generated using expression data from the GSE42568 dataset to visualize the interrelationships among these hub genes across samples (Supplementary Fig. 6A). Subsequently, to elucidate the functional pathways associated with these genes and to illustrate the strength of their associations, an enrichment chord diagram was constructed based on the GSE42568 expression profile (Supplementary Fig. 6B). To rigorously evaluate the diagnostic potential of these hub genes for distinguishing tumors from normal samples, ROC curve analysis was conducted. The analysis was performed independently on all three original datasets ( GSE29431 , GSE42568 , and GSE65194 ). Notably, all six hub genes consistently exhibited strong discriminatory power, with AUC values exceeding 0.8 in each dataset. The comprehensive ROC curves were presented in Supplementary Fig. 6C. GSEA of hub genes To investigate the biological pathways potentially influenced by the identified hub genes, GSEA was performed. For each hub gene (ADIPOQ, CD36, CIDEC, FABP4, LEP, PLIN1) within each training dataset, a pre-ranked list of all other genes was generated based on the Spearman’s correlation coefficient calculated using the R package psych. These gene lists, ordered by the strength of their correlation with the respective hub gene, were then subjected to GSEA using the R package clusterProfiler. The analysis was conducted independently across the three datasets ( GSE29431 , GSE42568 , GSE65194 ). The results for each hub gene in the GSE29431 , GSE42568 , and GSE65194 datasets were comprehensively displayed in Supplementary Figs. 7, 8, and 9, respectively. Each figure presents the top 10 most significantly enriched pathways for every hub gene, sorted by p-value. A consistent and notable enrichment was observed across multiple datasets for pathways related to the PPAR signaling pathway, DNA replication, and cell cycle, underscoring the pivotal role of these hub genes in regulating both metabolic processes and proliferative signaling in breast cancer. Analysis of immune infiltration in the tumor microenvironment To investigate differences in the immune microenvironment between tumor and normal samples, the relative abundances of 22 types of immune cells were estimated using the CIBERSORT algorithm in the GSE29431 , GSE42568 , and GSE65194 datasets. Stacked bar charts illustrated the immune cell composition for all samples in each cohort (Supplementary Figs. 10 A, 11 A, and 12 A). The analysis revealed that CD4 + T cells, T follicular helper cells, NK cells, and macrophages constituted the most predominant immune cell populations across the majority of samples. Subsequently, we explored the relationship between the expression of each individual hub gene and the levels of immune cell infiltration. Tumor samples within each dataset were stratified into high and low expression groups for each hub gene, and their immune infiltration profiles were compared. In the GSE29431 dataset, specific correlations were observed between hub genes and immune cell infiltration. ADIPOQ expression showed a significant positive correlation with the infiltration levels of T follicular helper cells and regulatory T cells (Tregs) (Supplementary Fig. 10B). CD36 expression was significantly associated with the abundance of CD8 + T cells and monocytes (Supplementary Fig. 10C). CIDEC expression demonstrated a significant correlation with T follicular helper cell infiltration (Supplementary Fig. 10D). FABP4 expression was significantly positively correlated with the infiltration of both T follicular helper cells and Tregs (Supplementary Fig. 10E). LEP expression showed a significant association with T follicular helper cell infiltration (Supplementary Fig. 10F). For PLIN1, while associations with several immune cell types were observed, none reached statistical significance in this dataset (Supplementary Fig. 10G). Analysis of the GSE42568 dataset revealed distinct association patterns. ADIPOQ expression was significantly correlated with the abundance of CD4 + memory resting T cells and activated dendritic cells (Supplementary Fig. 11B). CD36 showed significant associations with CD4 + memory resting T cells, Tregs, and mast cells (Supplementary Fig. 11C). For CIDEC, no significant correlations with specific immune cells were observed (Supplementary Fig. 11D). FABP4 was significantly linked to CD4 + memory resting T cells and Tregs (Supplementary Fig. 11E). For LEP, no significant correlations with specific immune cells were identified (Supplementary Fig. 11F). PLIN1 expression was significantly associated with activated dendritic cells (Supplementary Fig. 11G). In the GSE65194 dataset, a broader spectrum of significant correlations was identified. ADIPOQ was significantly associated with B cells, multiple T cell subsets, plasma cells, and M0 macrophages (Supplementary Fig. 12B). CD36 expression correlated significantly with Tregs and M2 macrophages (Supplementary Fig. 12C). CIDEC showed associations with B cells, T cells, CD4 + T cells, M0 macrophages, activated dendritic cells, and neutrophils (Supplementary Fig. 12D). FABP4 demonstrated the most extensive correlations, being significantly associated with B cells, multiple T cell subsets (CD8+, CD4+), M0 and M2 macrophages, mast cells, and neutrophils (Supplementary Fig. 12E). LEP was significantly correlated with plasma cells, CD4 + naive T cells, and Tregs (Supplementary Fig. 12F). PLIN1 showed significant associations with B cells, plasma cells, T follicular helper cells, and M0 and M2 macrophages (Supplementary Fig. 12G). Validation of hub gene expression and prognostic analysis To validate the expression patterns of the identified hub genes, we first analyzed their expression levels in breast cancer versus normal tissue using the BOX Plot tool from the GEPIA2 database. This analysis revealed that the expression levels of LEP, CD36, ADIPOQ, PLIN1, CIDEC, and FABP4 were all significantly downregulated in breast cancer tissues compared to normal controls (Supplementary Fig. 13A). To further corroborate these findings experimentally, we performed RT-qPCR to measure the expression of these genes in the human normal mammary epithelial cell line MCF-10 A and in two human breast cancer cell lines, MCF-7 and MDA-MB-231. Consistent with the bioinformatics analysis, the qPCR results confirmed that the mRNA expression levels of all six hub genes were significantly lower in the breast cancer cell lines (Supplementary Fig. 13B). Subsequently, a multi-dimensional prognostic analysis for the six hub genes was conducted using the Kaplan-Meier Plotter tool, assessing their association with OS ( n = 1880), RFS ( n = 4934), and PPS ( n = 458), which correspond to core prognostic needs at different clinical stages of breast cancer. The results are summarized in Supplementary Fig. 14. High expression of ADIPOQ (ACDC) and PLIN1 was significantly associated with increased survival rates in both OS and RFS analyses, suggesting their potential role as core regulatory factors influencing the entire disease spectrum from early relapse to late-stage progression and ultimate mortality, possibly through involvement in key pathways such as tumor cell survival and immune evasion. In contrast, CIDEC and FABP4 showed a significant association only with RFS, with no notable links to OS or PPS, indicating that their function might be primarily relevant to early-stage invasion and metastasis. The genes CD36 and LEP demonstrated limited prognostic value across the survival endpoints analyzed. CIDEC overexpression suppresses proliferation, migration, and invasion, while promoting apoptosis in breast cancer cells in vitro To elucidate the functional role of CIDEC in malignant phenotypes of breast cancer cells, CIDEC was overexpressed in MCF-7 and MDA-MB-231 cell lines via lentiviral transduction with an oe-CIDEC vector. Western blot analysis confirmed a significant increase in CIDEC protein levels in the oe-CIDEC group compared to the control and oe-NC group, demonstrating efficient overexpression (Fig. 1 A). The impact of CIDEC overexpression on cellular viability and proliferation was assessed using CCK-8 and EdU assays, respectively. Overexpression of CIDEC significantly reduced cell viability and suppressed proliferation in both MCF-7 and MDA-MB-231 cell lines (Fig. 1B-C). Furthermore, cell migration and invasion capabilities were evaluated by wound healing and Transwell assays. The results indicated that CIDEC overexpression markedly inhibited the migratory and invasive capacities of the breast cancer cells (Fig. 1 D-E). Flow cytometric analysis for apoptosis revealed that enforced expression of CIDEC significantly promoted cell apoptosis in both tested cell lines (Fig. 1 F). Fig. 1. Open in a new tab Cell Death-Inducing DFFA-Like Effector C (CIDEC) overexpression inhibits malignant phenotypes of breast cancer cells in vitro. ( A ) Western blot analysis of CIDEC protein expression levels in MCF-7 and MDA-MB-231 cells from the Control, oe-NC, and oe-CIDEC groups. ( B ) Cell viability measured by the Cell Counting Kit-8 (CCK-8) assay after CIDEC overexpression. ( C ) Cell proliferation assessed by 5-Ethynyl-2’-deoxyuridine (EdU) staining (scale bar: 50 μm). ( D ) Cell migration ability evaluated by wound healing assay (scale bar: 50 μm). ( E ) Cell invasion capacity determined by Transwell assay (scale bar: 50 μm). ( F ) Cell apoptosis rate detected by flow cytometry. Data are presented as mean ± standard deviation (SD) ( n ≥ 3 independent experiments). Statistical significance was determined by unpaired two-tailed Student’s t-test for comparisons between oe-NC and oe-CIDEC groups. * p < 0.05, ** p < 0.01, *** p < 0.001 vs. oe-NC group CIDEC inhibits protective autophagy in breast cancer cells To investigate the effect of CIDEC on autophagy, TEM was employed to examine autophagic structures. Compared to the oe-NC group, cells overexpressing CIDEC (oe-CIDEC group) exhibited a notable absence of clear double-membraned autophagosomes in both MCF-7 and MDA-MB-231 cell lines (Fig. 2 A, B). To further corroborate these findings, the protein levels of key autophagy markers were analyzed by Western blot. Overexpression of CIDEC in MCF-7 and MDA-MB-231 cells resulted in a marked downregulation of Beclin-1 and the lipidated form of LC3 (LC3B-II), along with a concomitant increase in p62 expression (Fig. 2 C, D). Fig. 2. Open in a new tab CIDEC overexpression inhibits autophagy in breast cancer cells. ( A ) Representative transmission electron microscopy (TEM) images of MCF-7 cells (scale bar: upper panel, 2 μm; lower panel, 500 nm). ( B ) Representative TEM images of MDA-MB-231 cells (scale bar: upper panel, 2 μm; lower panel, 500 nm). ( C ) Western blot analysis of autophagy-related proteins in MCF-7 cells. (D) Western blot analysis of autophagy-related proteins in MDA-MB-231 cells. ( E ) Autophagic flux analysis in MCF-7 cells. Cells transfected with oe-NC or oe-CIDEC were treated with or without 25 µM chloroquine (CQ) for 24 h, and the levels of microtubule-associated protein 1 light chain 3 beta (LC3B) and sequestosome 1 (p62) were assessed by Western blot. ( F ) Autophagic flux analysis in MDA-MB-231 cells under the same conditions as in ( E ). Data are presented as mean ± SD ( n ≥ 3 independent experiments). Statistical significance was determined by unpaired two-tailed Student’s t-test for comparisons between two groups (oe-NC vs. oe-CIDEC) or one-way ANOVA followed by Tukey’s post-hoc test for multiple group comparisons (CQ treatment experiments). * p < 0.05, ** p < 0.01 vs. oe-NC group; ## p < 0.01, ### p < 0.01 vs. oe-CIDEC group To further distinguish whether CIDEC inhibits autophagy initiation or impairs autophagic flux, we performed autophagic flux analysis using the lysosomal inhibitor CQ. MCF-7 and MDA-MB-231 cells transfected with oe-NC or oe-CIDEC were treated with or without 25 µM CQ for 24 h, and the levels of LC3B and p62 were assessed by Western blot. As shown in Fig. 2 E-F, under basal conditions (without CQ), CIDEC overexpression significantly increased p62 and decreased the LC3B-II/I ratio compared to the oe-NC group in both cell lines, consistent with reduced autophagic degradation. Treatment with CQ in control cells (oe-NC + CQ) led to the expected accumulation of both LC3B-II and p62, confirming functional basal autophagic flux. In CIDEC-overexpressing cells, CQ treatment (oe-CIDEC + CQ) further increased both p62 levels and the LC3B-II/I ratio compared to oe-CIDEC alone in both MCF-7 and MDA-MB-231 cells, indicating that autophagosome formation still occurs, but degradation is impaired. Importantly, the LC3B-II/I ratio in oe-CIDEC + CQ remained significantly lower than that in oe-NC + CQ, and p62 levels remained higher, demonstrating that CIDEC overexpression attenuates autophagic flux rather than merely reducing autophagosome formation. These results confirm that CIDEC suppresses autophagic degradation in breast cancer cells. CIDEC overexpression inhibits tumor growth in vivo To investigate the impact of CIDEC overexpression on tumor growth in vivo, a xenograft model was established by subcutaneously injecting MCF-7 breast cancer cells into nude mice. As shown in Fig. 3 A-B, compared to the oe-NC group, the oe-CIDEC group exhibited a significant reduction in both tumor volume and final tumor weight. Subsequent histological examination via HE staining revealed extensive structural damage within the tumor tissues from the oe-CIDEC group compared to the oe-NC controls (Fig. 3 C). Immunohistochemical staining for Ki67 was employed to assess tumor cell proliferation. The results demonstrated a significant decrease in Ki67-positive cells in the oe-CIDEC group (Fig. 3 D). Furthermore, TUNEL fluorescence staining indicated that CIDEC overexpression markedly enhanced cellular apoptosis within the tumor tissues (Fig. 3 E). Consistent with the in vitro findings, Western blot analysis of the tumor tissues confirmed increased CIDEC expression in the oe-CIDEC group, as well as decreased protein levels of Beclin-1 and LC3B-II, alongside an accumulation of p62 (Fig. 3 F). These results collectively demonstrate that CIDEC overexpression effectively suppresses tumor growth in vivo, which is associated with the inhibition of proliferation, induction of apoptosis, and suppression of protective autophagy. Fig. 3. Open in a new tab CIDEC overexpression suppresses tumor growth and modulates autophagy in MCF-7 xenograft tumors in vivo. ( A ) Representative images of resected tumors from each group. ( B ) Tumor volume growth curves (left) and final tumor weight (right) of the xenograft models. ( C ) Representative images of hematoxylin-eosin (HE) staining of tumor tissues. Scale bar: 100 μm; Magnification: 400×. ( D ) Representative images of Ki67 immunohistochemical staining. Scale bar: 100 μm; Magnification: 400×. ( E ) Representative images of Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining and the quantitative analysis of TUNEL-positive cells. Scale bar: 100 μm; Magnification: 400×. ( F ) Western blot analysis of CIDEC and autophagy-related proteins in tumor tissues. Data are presented as mean ± SD ( n ≥ 6 mice per group). Statistical significance was determined by unpaired two-tailed Student’s t-test. * p < 0.05, ** p < 0.01, *** p < 0.001 vs. oe-NC group CIDEC suppresses breast cancer cell proliferation, migration, and invasion via the cGMP/PKG pathway To elucidate the downstream molecular mechanism through which CIDEC exerts its functions in breast cancer, a bioinformatic analysis was performed. Initially, 18,389 genes implicated in breast cancer were retrieved from the Genecards database. Subsequently, 2,000 genes co-expressed with CIDEC were identified using the Coxpresdb database. The intersection of these datasets yielded 1,496 potential CIDEC co-expressed genes relevant to breast cancer (Fig. 4 A). KEGG pathway enrichment analysis of these 1,496 genes, performed using DAVID and visualized via a bubble chart, revealed significant enrichment in various metabolic pathways, the phosphatidylinositol 3-kinase/protein kinase B (PI3K/AKT) signaling pathway, peroxisome proliferator-activated receptor (PPAR) signaling pathway, and the cGMP/ PKG signaling pathway, among others (Fig. 4 B). Fig. 4. Open in a new tab CIDEC inhibits malignant phenotypes in breast cancer by suppressing the cyclic guanosine monophosphate/protein kinase G (cGMP/PKG) signaling pathway. ( A ) Venn diagram identifying potential CIDEC co-expressed genes in breast cancer via Genecards and Coxpresdb databases. ( B ) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment bubble chart of the co-expressed genes. ( C ) cGMP levels in tumor tissues measured by Enzyme-Linked Immunosorbent Assay (ELISA). ( D ) Western blot analysis of key cGMP/PKG pathway components, including soluble guanylate cyclase (sGC) and protein kinase G 1/2 (PRKG1/2), in tumor tissues. ( E ) Western blot analysis of CIDEC expression and PKG activity, assessed by phosphorylated Vasodilator-Stimulated Phosphoprotein at Serine 239 (p-VASP [Ser239]) and total VASP levels. The p-VASP/VASP ratio was quantified and normalized to the oe-NC group. ( F ) Cell viability assessed by CCK-8 assay in MCF-7 cells treated with the cGMP analog 8-Br-cGMP. ( G ) Representative images of EdU staining for proliferation assessment (scale bar: 50 μm). ( H ) Quantification of EdU-positive cells. ( I ) Cell migration evaluated by wound healing assay (scale bar: 50 μm). ( J ) Cell invasion determined by Transwell assay (scale bar: 50 μm). ( K ) Apoptosis rate measured by flow cytometry. Data are presented as mean ± SD ( n ≥ 3 independent experiments). Statistical significance was determined by unpaired two-tailed Student’s t-test for comparisons between two groups (oe-NC vs. oe-CIDEC) or one-way ANOVA followed by Tukey’s post-hoc test for multiple group comparisons (8-Br-cGMP treatment experiments). * P <0.05, ** P <0.01, *** P <0.001 vs. oe-NC group; # P <0.05, ## P <0.01, ### P <0.001 vs. oe-CIDEC group Although pathways such as PI3K-AKT and PPAR exhibited stronger enrichment signals in the KEGG analysis, these pathways have been extensively studied in breast cancer. In contrast, the cGMP/PKG pathway remains relatively underinvestigated in this context. Notably, emerging evidence has linked the cGMP/PKG pathway to the regulation of mitochondrial function [ 25 ], which is central to CIDEC-mediated tumor suppression. Based on its novelty and potential mechanistic relevance to mitochondrial homeostasis, we selected the cGMP/PKG pathway for further experimental validation. In vivo validation using tumor xenograft tissues showed that CIDEC overexpression significantly reduced cGMP levels compared to the oe-NC group, as determined by ELISA (Fig. 4 C). Furthermore, Western blot analysis demonstrated that CIDEC overexpression suppressed the protein levels of key cGMP/PKG pathway components, including sGC and PRKG1/2 (Fig. 4D). To more specifically assess the impact of CIDEC on cGMP/PKG pathway activity, we examined the phosphorylation status of VASP at Ser239, a well-established direct downstream substrate of PKG. Western blot analysis was performed on cells from four experimental groups (oe-NC, oe-CIDEC, 8-Br-cGMP, and oe-CIDEC + 8-Br-cGMP). As shown in Fig. 4 E, CIDEC overexpression was first confirmed, with markedly increased CIDEC protein levels in the oe-CIDEC and oe-CIDEC + 8-Br-cGMP groups compared to the oe-NC group. Notably, the p-VASP (Ser239)/VASP ratio was significantly reduced in CIDEC-overexpressing cells relative to the oe-NC group, indicating that CIDEC suppresses basal PKG activity. Treatment with the cGMP analog 8-Br-cGMP alone substantially increased p-VASP levels compared to the oe-NC group, confirming effective PKG activation. Importantly, in the oe-CIDEC + 8-Br-cGMP co-treatment group, the p-VASP/VASP ratio was significantly higher than that in the oe-CIDEC group, demonstrating that 8-Br-cGMP treatment partially restores PKG activity suppressed by CIDEC overexpression. These results provide direct evidence that CIDEC negatively regulates the cGMP/PKG signaling pathway at the level of PKG kinase activity, and that this suppression can be partially reversed by pharmacological activation of the pathway. To further verify the regulatory role of CIDEC in the cGMP/PKG pathway, a functional rescue experiment was conducted by treating CIDEC-overexpressing MCF-7 cells with 8-Br-cGMP. The application of 8-Br-cGMP significantly reversed the inhibitory effects of CIDEC overexpression on breast cancer cell viability (Fig. 4 F), proliferation (Fig. 4 G, H), migration (Fig. 4 I), and invasion (Fig. 4 J), while also attenuating the pro-apoptotic effect (Fig. 4 K). Notably, treatment with 8-Br-cGMP alone in control cells enhanced proliferative, migratory, and invasive phenotypes and reduced apoptosis compared to the oe-NC group. CIDEC regulates mitochondrial function in breast cancer cells via the cGMP/PKG pathway To investigate whether CIDEC modulates mitochondrial function through the cGMP/PKG pathway, we performed a series of assays in MCF-7 cells. Measurement of ATP production by ELISA revealed that CIDEC overexpression (oe-CIDEC group) significantly reduced cellular ATP levels compared to the oe-NC group. Conversely, treatment with the cGMP analog 8-Br-cGMP alone increased ATP content. Notably, the co-treatment with oe-CIDEC and 8-Br-cGMP partially rescued the ATP generation deficit induced by CIDEC overexpression (Fig. 5 A). The MMP was assessed using the JC-1 fluorescent probe. Cells in the oe-NC group exhibited high MMP, indicated by predominant red fluorescence. This signal was further enhanced by 8-Br-cGMP treatment alone. In contrast, CIDEC-overexpressing cells showed a pronounced shift to green fluorescence, signifying MMP dissipation. The MMP loss in the oe-CIDEC group was partially restored upon co-administration of 8-Br-cGMP (Fig. 5 B). Intracellular ROS levels, detected by DCFH-DA staining, were significantly elevated in the oe-CIDEC group. Treatment with 8-Br-cGMP alone significantly reduced ROS accumulation compared to the oe-NC group. Furthermore, the combination of 8-Br-cGMP with oe-CIDEC effectively attenuated the ROS increase induced by CIDEC overexpression compared to the oe-CIDEC group (Fig. 5 C). Western blot analysis of key proteins governing mitochondrial dynamics showed that CIDEC overexpression downregulated the expression of both fusion proteins (MFN1, MFN2, OPA1) and the fission protein DRP1. Treatment with 8-Br-cGMP alone exhibited an opposing effect, upregulating these proteins. Importantly, the combined intervention of oe-CIDEC and 8-Br-cGMP partially rescued CIDEC-mediated suppression of key mitochondrial dynamics regulators (Fig. 5 D). Fig. 5. Open in a new tab CIDEC impairs mitochondrial function through the cGMP/PKG signaling pathway. ( A ) Adenosine Triphosphate (ATP) levels in MCF-7 cells measured by ELISA. ( B ) Analysis of mitochondrial membrane potential (MMP) using 5,5’,6,6’-tetrachloro-1,1’,3,3’-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) staining (scale bar: 50 μm, magnification: 200×). Red fluorescence indicates high MMP (polarized mitochondria), and green fluorescence indicates low MMP (depolarized mitochondria). ( C ) Intracellular reactive oxygen species (ROS) levels detected by 2’,7’-dichlorodihydrofluorescein diacetate (DCFH-DA) staining and quantitative analysis (scale bar: 50 μm, magnification: 200×). ( D ) Western blot analysis of mitochondrial dynamics proteins, including the fission protein Dynamin-Related Protein 1 (DRP1) and the fusion proteins Mitofusin 1/2 (MFN1/2) and Optic Atrophy 1 (OPA1). Data are presented as mean ± SD ( n ≥ 3 independent experiments). Statistical significance was determined by one-way ANOVA followed by Tukey’s post-hoc test. * P <0.05, ** P <0.01, *** P <0.001 vs. oe-NC group; # P <0.05, ## P <0.01 vs. oe-CIDEC group Collectively, these results demonstrate that CIDEC overexpression induces profound mitochondrial dysfunction, characterized by reduced ATP production, loss of MMP, and elevated ROS, largely through the inhibition of the cGMP/PKG pathway. The ability of 8-Br-cGMP to counteract these effects confirms the pivotal role of this pathway in mediating CIDEC’s impact on mitochondrial homeostasis. Correlation between CIDEC-cGMP/PKG axis in regulating mitochondrial function and autophagy Given that autophagy is a crucial process for maintaining cellular homeostasis and its dysregulation is frequently associated with mitochondrial dysfunction, we next investigated whether the regulation of mitochondrial function by the CIDEC-cGMP/PKG axis is linked to its impact on autophagy. The autophagic inhibitor 3-MA was applied to MCF-7 cells. TEM observations revealed that compared to the oe-NC group, autophagosomes were significantly reduced in the oe-CIDEC group, while the 8-Br-cGMP group showed an increase in autophagic structures. Co-treatment with oe-CIDEC and 8-Br-cGMP partially restored autophagy compared to the oe-CIDEC-alone group. As expected, the addition of 3-MA (oe-CIDEC + 8-Br-cGMP + 3-MA group) abolished autophagic structure formation due to pharmacological inhibition (Fig. 6 A). Western blot analysis confirmed that CIDEC protein levels were significantly increased in the oe-CIDEC group compared to the oe-NC group, while no significant differences were observed in the oe-CIDEC + 8-Br-cGMP and oe-CIDEC + 8-Br-cGMP + 3-MA groups compared to the oe-CIDEC group (Fig. 6 B). Treatment with 8-Br-cGMP alone increased the protein levels of Beclin-1 and the lipidated form of LC3 (LC3B-II) and decreased p62 accumulation, although these changes were not statistically significant compared to the oe-NC group. Importantly, 8-Br-cGMP treatment significantly reversed the CIDEC overexpression-induced downregulation of Beclin-1 and LC3B-II and the upregulation of p62. Conversely, intervention with 3-MA (oe-CIDEC + 8-Br-cGMP + 3-MA group) significantly increased p62 accumulation and inhibited the conversion of LC3B-I to LC3B-II relative to the (oe-CIDEC + 8-Br-cGMP group). Fig. 6. Open in a new tab The CIDEC-cGMP/PKG axis regulates mitochondrial function and autophagy in a correlated manner. ( A ) Representative TEM images of MCF-7 cells under different treatments (scale bar: upper panel, 2 μm; lower panel, 500 nm). ( B ) Western blot analysis of CIDEC and autophagy-related proteins and corresponding quantitative analysis. Data are presented as mean ± SD ( n ≥ 3 independent experiments). Statistical significance was determined by one-way ANOVA followed by Tukey’s post-hoc test. ** P <0.01, *** P <0.001 vs. oe-NC group; # P <0.05 vs. oe-CIDEC group; && P <0.01, &&& P <0.001 vs. oe-CIDEC + 8-Br-cGMP group Pharmacological inhibition of autophagy attenuates the functional rescue mediated by cGMP/PKG pathway activation To functionally interrogate the contribution of autophagy to the observed phenotypes, we administered the autophagy inhibitor 3-MA to MCF-7 cells in which the cGMP/PKG pathway was concurrently activated. CCK-8 and EdU staining assays demonstrated that, compared to the oe-CIDEC + 8-Br-cGMP group, co-treatment with 3-MA significantly suppressed cell viability (Fig. 7 A) and proliferation (Fig. 7 B, C). In CIDEC-overexpressing cells, 3-MA treatment markedly inhibited the migratory (Fig. 7 D) and invasive (Fig. 7 E) capacities that were partially restored by 8-Br-cGMP. Furthermore, flow cytometric analysis revealed that the addition of 3-MA (oe-CIDEC + 8-Br-cGMP + 3-MA group) significantly increased the apoptosis rate compared to the oe-CIDEC + 8-Br-cGMP group (Fig. 7F). These results indicate that the functional rescue achieved by activating the cGMP/PKG pathway in CIDEC-overexpressing cells is, at least partially, dependent on the restoration of autophagic activity. Fig. 7. Open in a new tab Autophagy inhibition attenuates the rescue of malignant phenotypes by cGMP/PKG activation. ( A ) Cell viability measured by CCK-8 assay. ( B ) Cell proliferation assessed by EdU staining (scale bar: 50 μm). ( C ) Quantification of EdU-positive cells. ( D ) Cell migration evaluated by wound healing assay (scale bar: 50 μm). ( E ) Cell invasion determined by Transwell assay (scale bar: 50 μm). ( F ) Apoptosis rate detected by flow cytometry. Data are presented as mean ± SD ( n ≥ 3 independent experiments). Statistical significance was determined by one-way ANOVA followed by Tukey’s post-hoc test. * P <0.05, ** P <0.01, *** P <0.001 vs. oe-NC group; # P <0.05, ## P <0.01, ### P <0.001 vs. oe-CIDEC group; & P <0.05, &&& P <0.001 vs. oe-CIDEC + 8-Br-cGMP group Pharmacological inhibition of autophagy abrogates the improvement in mitochondrial function mediated by cGMP/PKG pathway activation Subsequently, we assessed mitochondrial function following 3-MA intervention. ELISA results indicated that 3-MA treatment (oe-CIDEC + 8-Br-cGMP + 3-MA group) led to a decrease in cellular ATP levels compared to the oe-CIDEC + 8-Br-cGMP group (Fig. 8 A). JC-1 staining revealed that the oe-CIDEC + 8-Br-cGMP + 3-MA co-treatment group exhibited a greater proportion of low-potential green fluorescence compared to the oe-CIDEC + 8-Br-cGMP group, indicating a reduction in MMP (Fig. 8 B). Measurement of ROS levels using DCFH-DA staining showed a significant increase in ROS generation in the 3-MA treatment group (oe-CIDEC + 8-Br-cGMP + 3-MA group) relative to the oe-CIDEC + 8-Br-cGMP group (Fig. 8 C). Western blot analysis demonstrated that, compared to the oe-CIDEC + 8-Br-cGMP group, the combined treatment with oe-CIDEC, 8-Br-cGMP, and 3-MA resulted in downregulation of MFN1/2, OPA1, and DRP1 expression (Fig. 8D). It is noteworthy that the decreases in MFN2, OPA1, and DRP1 levels showed significant differences relative to the oe-CIDEC + 8-Br-cGMP group, whereas the downregulation of MFN1 expression did not reach statistical significance. Fig. 8. Open in a new tab Pharmacological inhibition of autophagy abrogates the improvement in mitochondrial function mediated by cGMP/PKG pathway activation. ( A ) ATP levels in MCF-7 cells measured by ELISA. ( B ) Analysis of MMP using JC-1 staining (scale bar: 50 μm, magnification: 200×). ( C ) Intracellular ROS levels detected by DCFH-DA staining and quantitative analysis (scale bar: 50 μm, magnification: 200×). ( D ) Western blot analysis of mitochondrial dynamics proteins with corresponding quantification. Data are presented as mean ± SD ( n ≥ 3 independent experiments). Statistical significance was determined by one-way ANOVA followed by Tukey’s post-hoc test. * P <0.05, ** P <0.01, *** P <0.001 vs. oe-NC group; # P <0.05, ## P <0.01, ### P <0.001 vs. oe-CIDEC group; & P <0.05, && P <0.01 vs. oe-CIDEC + 8-Br-cGMP group Discussion In this study, through integrated bioinformatics analysis of multiple gene expression datasets and rigorous machine learning approaches, we identified CIDEC as a core downregulated hub gene in breast cancer. Subsequent functional investigations using both in vitro and in vivo models revealed that its overexpression exerted potent tumor-suppressive effects by comprehensively inhibiting malignant phenotypes - including proliferation, migration, and invasion - while simultaneously promoting apoptosis. More importantly, we delineated a novel mechanistic pathway wherein CIDEC disrupts mitochondrial fitness and blocks protective autophagy through specific suppression of the cGMP/PKG signaling pathway. Crucially, rescue experiments demonstrated that pharmacological activation of this pathway reversed these effects, while additional inhibition of autophagy abrogated this rescue. Our integrated bioinformatics approach identified six core hub genes (LEP, CD36, ADIPOQ, PLIN1, CIDEC, and FABP4) through PPI network analysis and multiple machine learning algorithms, all demonstrating consistent downregulation in breast cancer tissues. The identification and validation of these biomarkers are further established through distinct methodological approaches: CD36 confirmation involves integrated computational docking screening and experimental validation, demonstrating its critical role in mediating fatty acid uptake and conferring resistance to anti-Human Epidermal Growth Factor Receptor 2 therapies [ 29 ]. LEP and ADIPOQ validation employs a comprehensive meta-analysis incorporating fixed/random effects models and subgroup analyses, revealing their consistent associations with breast cancer risk across multiple clinical studies [ 30 ]. Through integrated bioinformatics analyses of TCGA and multi-database platforms (Gepia2, Ualcan, Kaplan-Meier plotter), PLIN1 is identified as a potential breast cancer biomarker based on its significant downregulation across molecular subtypes and demonstrated association with poor prognosis [ 31 ]. CIDEC emerges as a highly accurate diagnostic biomarker through machine learning analysis of triple-negative breast cancer transcriptomic datasets, achieving 97.1% diagnostic accuracy with subsequent experimental validation [ 32 ]. Notably, FABP4 characterization applies single-cell RNA-sequencing analysis, revealing its specific upregulation in obesity-associated macrophages and demonstrating a role in promoting tumor cell proliferation [ 33 ]. The convergence of evidence from these complementary methodological approaches strongly supports the potential of this gene signature as a source of novel biomarkers and therapeutic targets for breast cancer intervention. Our functional characterization demonstrated that CIDEC overexpression significantly inhibited proliferation, migration, and invasion while promoting apoptosis in breast cancer cells in vitro, findings that were corroborated by in vivo studies showing suppressed tumor growth in xenograft models. This tumor-suppressive function of CIDEC exhibits tissue-specific mechanistic diversity across cancer types. In breast cancer, CIDEC acts as a key downstream effector of the Homeobox A2-Peroxisome Proliferator-Activated Receptor Gamma signaling axis, where its downregulation impairs lipid droplet formation and promotes disease progression [ 34 ]. In non-small cell lung cancer, CIDEC suppresses tumor progression through regulation of lipid metabolism by inhibiting adipose triglyceride lipase-mediated lipolysis, thereby maintaining lipid droplet stability and cellular energy homeostasis [ 35 ]. Similarly, in lung adenocarcinoma, CIDEC functions as a tumor suppressor by inhibiting the Extracellular Signal-Regulated Kinase/p38 Mitogen-Activated Protein Kinase signaling pathway and inducing G0 phase cell cycle arrest [ 36 ]. The concordance between in vitro and in vivo findings across multiple cancer types reinforces the physiological relevance of CIDEC-mediated tumor suppression and highlights its potential as a therapeutic target in oncology. Our mechanistic investigations revealed that CIDEC induced mitochondrial dysfunction, characterized by diminished ATP production, dissipated membrane potential, and elevated ROS, while concurrently suppressing protective autophagy. Notably, multiple regulatory pathways converge on mitophagy to influence breast cancer progression. Under glucose starvation, Sentrin-Specific Protease 1 (SENP1)-mediated deSUMOylation facilitates Mesencephalic Astrocyte-Derived Neurotrophic Factor (MANF) translocation to mitochondria, where it interacts with Parkin RBR E3 Ubiquitin Protein Ligase (PRKN) to restore E3 ligase activity, thereby initiating protective mitophagy that sustains breast cancer cell survival [ 37 ]. Complementarily, Sorting and Assembly Machinery component 50 (SAMM50) modulates mitochondrial dynamics and mitophagy to regulate CIDEC expression, establishing a functional link between mitochondrial quality control and metabolic activation [ 38 ]. Conversely, Unc-51 like autophagy activating kinase 1 (ULK1) deficiency under hypoxic conditions disrupts mitophagy, leading to damaged mitochondrial accumulation and subsequent NLRP3 inflammasome activation through elevated mitochondrial ROS, ultimately promoting breast cancer bone metastasis [ 39 ]. Pharmacologically, Urolithin A induces Transcription Factor EB (TFEB) nuclear translocation in a partially mechanistic Target of Rapamycin (mTOR)-dependent manner, enhancing mitophagy-lysosomal clearance of damaged mitochondria in tumor-associated macrophages and suppressing breast cancer progression [ 40 ]. Additionally, Mucin 1 (MUC1) translocation to mitochondria promotes breast cancer malignancy through degradation of ATPase family AAA domain-containing 3 A (ATAD3A), preventing PTEN-induced kinase 1 (PINK1) cleavage and enhancing mitophagy [ 41 ]. These findings collectively underscore the therapeutic potential of targeting the intricate crosstalk between mitochondrial homeostasis and autophagic processes in breast cancer intervention. Our results established the cGMP/PKG pathway as the pivotal upstream mediator of CIDEC’s tumor-suppressive function, evidenced by the complete reversal of CIDEC’s effects with 8-Br-cGMP. Furthermore, the abolition of this rescue by autophagy inhibition revealed that the restoration of protective autophagy was critically dependent on cGMP/PKG signaling. This tumor-suppressive role of cGMP/PKG signaling presents a notable contrast to its context-dependent functions in breast cancer progression, where the pathway can also promote metastasis through GTP-mediated activation that enhances tumor stemness and metastatic potential via downstream MAPK signaling [ 42 ]. The therapeutic potential of modulating this pathway is further evidenced by studies showing that inhibition of cGMP phosphodiesterase (PDE) elevates intracellular cGMP levels, activates cGMP-PKG, and subsequently suppresses β-catenin/Tcf transcriptional activity to induce apoptosis and inhibit cancer cell proliferation [ 43 ]. Additionally, in cardiac protection models, sildenafil activates the cGMP-PKG signaling pathway to enhance PINK1/Parkin-mediated mitophagy, thereby maintaining mitochondrial integrity [ 44 ]. Collectively, these findings establish the cGMP/PKG pathway as a central coordinator that integrates mitochondrial function with autophagic processes to execute CIDEC’s tumor-suppressive program, while highlighting the need for context-specific therapeutic targeting of this multifaceted signaling pathway. It is important to acknowledge the limitations of this study. First, although our data demonstrate that CIDEC overexpression reduces cGMP levels and downregulates sGC and PRKG1/2 expression, and that pharmacological activation of this pathway reverses CIDEC-induced phenotypes, we have not yet established whether CIDEC directly regulates these components or whether the observed effects are mediated through indirect mechanisms. Therefore, while we propose the cGMP/PKG pathway as a key downstream effector of CIDEC, we do not claim exclusivity or directness of this regulation. Additional genetic or mechanistic evidence is required to further substantiate these conclusions. Second, regarding the relationship between mitochondrial membrane potential depolarization and ROS accumulation induced by CIDEC overexpression, our current experimental design does not definitively establish the temporal sequence of these events. The coordinated downregulation of mitochondrial dynamics regulators (MFN1/2, OPA1, DRP1) suggested that CIDEC might directly impair mitochondrial quality control as an upstream event, but the potential for reciprocal feedback between ROS and mitochondrial damage remains to be elucidated. Future studies employing genetic approaches and time-course analyses are warranted to address these mechanistic questions. In conclusion, our study unveils a tumor-suppressive axis in breast cancer centered on CIDEC-mediated regulation of the cGMP/PKG pathway and its downstream effects on mitochondrial function and autophagy. These findings significantly advance our understanding of how CIDEC can orchestrate broad anti-tumor responses through integrated regulation of multiple cellular processes. The identification of this pathway not only provides new insights into breast cancer biology but also opens promising avenues for therapeutic intervention targeting the CIDEC-cGMP/PKG-mitochondria-autophagy network. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 (3MB, docx) Supplementary Material 2 (20.1KB, docx) Supplementary Material 3 (6.5MB, pdf) Acknowledgements Not applicable. Author contributions Xuchu Jin: Conceptualization; Formal analysis; Methodology; Writing - original draft; Validation; Resources; Yu Liu and Xinyu Zheng: Formal analysis; Methodology; Validation; Editing; Yangke He: Data curation; Investigation; Software; Review & editing; All authors have read and approved the manuscript. Funding Not applicable. Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate The experiments conformed to the Guide for the Care and Use of Laboratory Animals. Animal study has been approved by the Animal Ethics Committee of Sichuan Provincial People’s Hospital. All methods are reported in accordance with ARRIVE guidelines. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Xuchu Jin and Yu Liu are co first authors. Contributor Information Xinyu Zheng, Email: [email protected]. Yangke He, Email: [email protected]. References 1. X. Xiong et al., Breast cancer: pathogenesis and treatments. Signal. Transduct. Target. Ther. 10 (1), 49 (2025) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. D. Trapani et al., Global challenges and policy solutions in breast cancer control. Cancer Treat. Rev. 104 , 102339 (2022) [ DOI ] [ PubMed ] [ Google Scholar ] 3. A.H. Farkas, A.B. Nattinger, Breast cancer screening and prevention. Ann. Intern. Med., 2023. 176 (11): p. Itc161-itc176. 4. S. Valente, E. Roesch, Breast cancer survivorship. J. Surg. Oncol. 130 (1), 8–15 (2024) [ DOI ] [ PubMed ] [ Google Scholar ] 5. E. Nolan, G.J. Lindeman, J.E. Visvader, Deciphering breast cancer: from biology to the clinic. Cell. 186 (8), 1708–1728 (2023) [ DOI ] [ PubMed ] [ Google Scholar ] 6. Y. Liang et al., Metastatic heterogeneity of breast cancer: Molecular mechanism and potential therapeutic targets. Semin Cancer Biol. 60 , 14–27 (2020) [ DOI ] [ PubMed ] [ Google Scholar ] 7. R.L.N. Godone et al., Clinical and molecular aspects of breast cancer: Targets and therapies. Biomed. Pharmacother. 106 , 14–34 (2018) [ DOI ] [ PubMed ] [ Google Scholar ] 8. D. Kashyap et al., Global increase in breast cancer incidence: risk factors and preventive measures. Biomed Res Int, 2022. 2022: p. 9605439 9. F.J. Chen et al., CIDE family proteins control lipid homeostasis and the development of metabolic diseases. Traffic. 21 (1), 94–105 (2020) [ DOI ] [ PubMed ] [ Google Scholar ] 10. L. Xu et al., CIDE proteins and their regulatory mechanisms in lipid droplet fusion and growth (FEBS Lett, 2024) 11. K. Murakami et al., Selective PPARα modulator pemafibrate and sodium-glucose cotransporter 2 inhibitor tofogliflozin combination treatment improved histopathology in experimental mice model of non-alcoholic steatohepatitis. Cells, 2022. 11 (4) 12. Q.S. Xia et al., Ban-xia-xie-xin-tang ameliorates hepatic steatosis by regulating Cidea and Cidec expression in HFD-fed mice. Phytomedicine. 105 , 154351 (2022) [ DOI ] [ PubMed ] [ Google Scholar ] 13. S. Malvia et al., Study of Gene Expression Profiles of Breast Cancers in Indian Women. Sci. Rep. 9 (1), 10018 (2019) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. M. Onishi et al., Molecular mechanisms and physiological functions of mitophagy. Embo j. 40 (3), e104705 (2021) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 15. L. Liu et al., Crosstalk between mitochondrial biogenesis and mitophagy to maintain mitochondrial homeostasis. J. Biomed. Sci. 30 (1), 86 (2023) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. A. Roca-Portoles, S.W.G. Tait, Mitochondrial quality control: from molecule to organelle. Cell. Mol. Life Sci. 78 (8), 3853–3866 (2021) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. R. Iorio, G. Celenza, S. Petricca, Mitophagy: molecular mechanisms, new concepts on parkin activation and the emerging role of AMPK/ULK1 axis. Cells, 2021. 11 (1) 18. H.Y. Lee et al., Mitochondrial metabolic signatures in hepatocellular carcinoma. Cells, 2021. 10 (8) 19. Y. Luo, J. Ma, W. Lu, The significance of mitochondrial dysfunction in cancer. Int. J. Mol. Sci., 2020. 21 (16) 20. Y. Lu et al., Cellular mitophagy: Mechanism, roles in diseases and small molecule pharmacological regulation. Theranostics. 13 (2), 736–766 (2023) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. D.R. Miller, A. Thorburn, Autophagy and organelle homeostasis in cancer. Dev. Cell. 56 (7), 906–918 (2021) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. T. Wang et al., SOCS3 regulates pathological retinal angiogenesis through modulating SPP1 expression in microglia and macrophages. Mol. Ther. 32 (5), 1425–1444 (2024) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. S. Chen et al., Protective autophagy promotes the resistance of HER2-positive breast cancer cells to lapatinib. Tumour Biol. 37 (2), 2321–2331 (2016) [ DOI ] [ PubMed ] [ Google Scholar ] 24. H. Qu et al., CLDN6 induces chemoresistance through protective autophagy in breast cancer. Int. J. Biol. Sci. 21 (12), 5444–5459 (2025) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. J. Inserte, D. Garcia-Dorado, The cGMP/PKG pathway as a common mediator of cardioprotection: translatability and mechanism. Br. J. Pharmacol. 172 (8), 1996–2009 (2015) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Y.M. Yoo et al., Nitric oxide prevents H2O2-induced apoptosis in SK-N-MC human neuroblastoma cells. Int. J. Biol. Sci. 14 (14), 1974–1984 (2018) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. J. Niu et al., The COPS3-FOXO3 positive feedback loop regulates autophagy to promote cisplatin resistance in osteosarcoma. Autophagy. 19 (6), 1693–1710 (2023) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Y. Zhou et al., CDK11 Mediates Autophagy to Promote Breast Cancer Cell Proliferation and Migration by Regulating BCL-2. Faseb j. 39 (16), e70997 (2025) [ DOI ] [ PubMed ] [ Google Scholar ] 29. L. Castagnoli et al., Identification of new selective CD36 inhibitors to potentiate HER2-targeted therapy in HER2-positive breast cancer. Sci. Rep. 15 (1), 28709 (2025) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. W.Z. Peng et al., Genetic alterations in LEP and ADIPOQ genes and risk for breast cancer: a meta-analysis. Front. Oncol. 13 , 1125189 (2023) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. U. Yousuf et al., Identification and analysis of dysregulated fatty acid metabolism genes in breast cancer subtypes. Med. Oncol. 39 (12), 256 (2022) [ DOI ] [ PubMed ] [ Google Scholar ] 32. N. Rapier-Sharman et al., Secondary Transcriptomic analysis of triple-negative breast cancer reveals reliable universal and subtype-specific mechanistic markers . Cancers (Basel), 2024. 16(19) 33. G. Zhao et al., Single-cell RNA-sequencing reveals a unique landscape of the tumor microenvironment in obesity-associated breast cancer. Oncogene, 2024 34. De F.D.E. Palma et al., Epigenetic regulation of HOXA2 expression affects tumor progression and predicts breast cancer patient survival (Cell Death Differ, 2025) 35. Y. Chen et al., Cell death-inducing DFF45-like effector C intervenes the progression of non-small cell lung cancer via modulating lipid metabolism. Int. J. Surg., 2025 36. F. Guo et al., Silencing of ARL14 Gene Induces Lung Adenocarcinoma Cells to a Dormant State. Front. Cell. Dev. Biol. 7 , 238 (2019) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Z. Xiong et al., MANF facilitates breast cancer cell survival under glucose-starvation conditions via prkn-mediated mitophagy regulation. Autophagy, 2024 38. S.J. Park et al., SAMM50 regulates thermogenesis of beige adipocytes differentiated from human adipose-derived stem cells by balancing mitochondrial dynamics. Int. J. Mol. Sci., 2022. 23 (12) 39. R. Deng et al., MAPK1/3 kinase-dependent ULK1 degradation attenuates mitophagy and promotes breast cancer bone metastasis. Autophagy. 17 (10), 3011–3029 (2021) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. B. Zheng et al., Urolithin a inhibits breast cancer progression via activating TFEB-mediated mitophagy in tumor macrophages. J. Adv. Res., 2024 41. Q. Li et al., The oncoprotein MUC1 facilitates breast cancer progression by promoting Pink1-dependent mitophagy via ATAD3A destabilization. Cell. Death Dis. 13 (10), 899 (2022) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. Y. Lv et al., Nucleotide de novo synthesis increases breast cancer stemness and metastasis via cGMP-PKG-MAPK signaling pathway. PLoS Biol. 18 (11), e3000872 (2020) [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 43. H.N. Tinsley et al., Novel non-cyclooxygenase inhibitory derivative of sulindac inhibits breast cancer cell growth in vitro and reduces mammary tumorigenesis in rats. Cancers (Basel), 2023. 15 (3) 44. X. Zhang et al., Chronic inhibition of cGMP-specific phosphodiesterase 5 attenuates myocardial hypertrophy by promoting mitophagy in cardiomyocytes. Biochem. Pharmacol., 2025: p. 117366 Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary Material 1 (3MB, docx) Supplementary Material 2 (20.1KB, docx) Supplementary Material 3 (6.5MB, pdf) Data Availability Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Articles from Cellular Oncology are provided here courtesy of Springer ACTIONS View on publisher site PDF (18.9 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top