Abstract Background The Coiled-coil domain-containing (CCDC) family, due to its unique protein structural domain and broad involvement in diverse biological processes, has emerged as a focus in oncology research. Nevertheless, its clinical significance and function in bladder cancer (BLCA) remain poorly defined. Methods Machine learning algorithms were employed to identify pivotal CCDC genes in the cancer genome atlas (TCGA), and a prognostic model was subsequently constructed. Multi-omics data encompassing pan-cancer cohorts, single-cell sequencing, and spatial transcriptomics were integrated to characterize the expression patterns and prognostic significance of Coiled-coil domain-containing 137 (CCDC137), a previously uncharacterized CCDC family member in BLCA. Tissue microarray confirmed CCDC137 abnormal expression in bladder carcinoma specimens. The effect of CCDC137 knockdown on BLCA progression was evaluated through CCK8 assay, clonogenic formation, wound healing, Transwell, and subcutaneous xenograft models. RNA sequencing, quantitative RT-PCR, and western blot were utilized to delineate its regulatory network. Results A prognostic model incorporating 10 CCDC genes was successfully established in the TCGA-BLCA cohort. Then, we found that CCDC137 exhibited pan-cancer overexpression and usually correlation with poor clinical outcomes. Immunohistochemistry further substantiated its dysregulation in bladder carcinoma. Integrated multi-omics analyses suggested associations between CCDC137 expression and a tumor immunosuppressive microenvironment. CCDC137 knockdown significantly suppressed bladder cancer cell proliferation and migratory capacity in vitro. Correspondingly, subcutaneous xenograft tumor growth was inhibited in vivo. Moreover, decreased expression of stearoyl-CoA desaturase (SCD), a key lipid metabolic enzyme, accompanied CCDC137 depletion. These findings collectively suggest a cancer-promoting role for CCDC137 in bladder carcinoma. Conclusions This systematic investigation combining multi-omics bioinformatics analyses and experimental validation demonstrates the role of CCDC137 in bladder carcinoma progression, providing novel mechanistic insights into the pathogenesis of BLCA and offering a theoretical foundation for therapeutic targeting of CCDC137 in urothelial malignancies. Supplementary Information The online version contains supplementary material available at 10.1186/s12967-025-07033-w. Keywords: Coiled-coil domain-containing proteins 137, Bladder cancer, Prognosis, Stearoyl-CoA desaturase Background Bladder cancer (BLCA) is one of the most common urological malignancies worldwide. In 2022, the global incidence and mortality rates of bladder cancer ranked ninth and thirteenth among all cancers, respectively, with 613,791 new cases and 220,349 deaths reported [[46]1]. The efficacy of conventional treatments, including radical cystectomy, chemotherapy, and radiotherapy, is limited by the high incidence of tumor recurrence and metastasis, particularly in patients with muscle invasive bladder cancer (MIBC) [[47]2]. In recent years, immune checkpoint therapies, represented by PD-1/PD-L1 inhibitors, have offered hope for advanced patients. However, the clinical response rate remains only 20–25%, and there is a lack of effective biomarkers to identify patients who may benefit [[48]3]. These challenges stem from the high heterogeneity of bladder cancer. Therefore, there is an urgent need to elucidate the molecular mechanisms of bladder cancer and develop novel prognostic biomarkers and therapeutic targets, which are critical to addressing this challenging treatment landscape. The coiled-coil domain-containing (CCDC) domain is a protein structural motif found in a variety of natural proteins, characterized by one or more alpha-helical peptides wrapped around each other in a supercoiled manner [[49]4]. Most CCDC proteins function as structural proteins, regulatory proteins, motor proteins, DNA-binding proteins, membrane proteins, and enzymes, participating in various physiological processes such as chromosome segregation, DNA recognition and cleavage, protein scaffolding, and protein transport [[50]5]. Additionally, several members of the CCDC protein family have been implicated in the pathogenesis of non-neoplastic diseases, including hypertension, asthma, male infertility, and congenital heart disease [[51]6]. In recent years, due to their critical roles in cytoskeleton formation, signal transduction, and genomic stability, CCDC proteins have emerged as a focus of cancer research. For example, Liu R et al. reported that CCDC25 promotes breast cancer metastasis by regulating neutrophil extracellular traps [[52]7]. CCDC106 promotes cervical carcinogenesis via p53 binding and subsequent proteasomal degradation [[53]8]. CCDC154 interacts with MCM2 to regulate the progression of colorectal cancer [[54]9]. CCDC178 enhances resistance to anoikis and promotes metastasis in hepatocellular carcinoma by activating the ERK signaling pathway [[55]10]. However, the biological roles of the CCDC protein family and their clinical significance in bladder cancer remain poorly understood. Stearoyl-CoA desaturase (SCD), famous for its function as a key rate-limiting enzyme in lipid metabolism [[56]11], has been extensively documented for its tumor-promoting roles in various malignancies, including bladder cancer [[57]12]. Piao C et al. reported SCD overexpression in bladder cancer and its association with poor patient prognosis [[58]13]. SCD knockdown suppresses bladder cancer cell proliferation and induces apoptosis in vitro, while inhibiting tumor growth in vivo [[59]14]. Furthermore, studies by Li Y et al. demonstrate that SCD inhibitors effectively mediate apoptosis of bladder cancer stem cells [[60]15]. Despite the well-established pro-oncogenic functions of SCD in bladder cancer, its relationship with CCDC proteins remains unknown.In recent years, owing to the rapid development of high-throughput sequencing technologies and bioinformatics analysis methods, an increasing number of studies have begun to explore the molecular mechanisms of bladder cancer at multi-omics data. These researches have not only revealed the complex heterogeneity of bladder cancer but also provided critical clues for identifying novel prognostic biomarkers and therapeutic targets [[61]16]. In this study, CCDC protein family members were obtained from the GeneCards database and differentially expressed CCDC proteins in bladder cancer were identified by intersecting with differentially expressed genes (DEGs) from The Cancer Genome Atlas (TCGA)-BLCA dataset. Subsequently, CCDC proteins for constructing a prognostic model were selected using random survival forest (RSF) analysis and least absolute shrinkage and selection operator (LASSO) regression (1000 times). Furthermore, our research focused for the first time on CCDC137, a CCDC family member with aberrantly high expression but unknown function in bladder cancer. By integrating bulk RNA sequencing, single-cell sequencing, and spatial transcriptomics, the expression characteristics of CCDC137 and its relationship with the tumor immune microenvironment (TIME) were revealed. Additionally, in vitro and in vivo experiments were conducted to explore biological functions of CCDC137 in bladder cancer progression. RNA sequencing and experimental validation finally identified a regulatory relationship between CCDC137 and SCD. In summary, our study utilized multi-omics integration to uncover the clinical significance and molecular mechanisms of CCDC137, providing new insights into the diagnosis, treatment and prognosis for BLCA. Methods Raw data acquisition and processing Members of the CCDC protein family were obtained from the GeneCards database ([62]https://www.genecards.org/, Additional file [63]1). Tumor datasets from the TCGA database were retrieved using the TCGAbiolinks R package [[64]17]. In addition to normal tissue samples, tumor tissue samples only retained patients with gene expression data and survival information. A total of 721 normal tissue samples and 9624 tumor tissue samples were included (Additional file [65]1) and corresponding methylation probe data were subsequently downloaded. TCGA pan-cancer gene-level copy number variation (CNV) data were acquired from the UCSC Xena database ([66]https://xenabrowser.net/datapages/) (dataset: copy number (gene-level)—gene-level copy number (gistic2)). RNA sequencing data and clinical information from the IMvigor 210 cohort [[67]18], a metastatic urothelial carcinoma with anti-programmed death-ligand 1 (anti-PD-L1) immunotherapy cohort, were derived with the IMvigor210CoreBiologies package in R to explore immunotherapy response. For RNA-seq expression matrices, all analyses except differential expression analysis were performed using log[2](TPM + 1) transformed data, while raw count matrices were used for differential expression analysis. Additionally, GEO datasets were retrieved to validate differential CCDC137 expression between normal bladder tissues and bladder cancer tissues. Three inclusion criteria were applied: (1) Availability of CCDC137 mRNA expression data; (2) ≥ 5 normal bladder tissue samples; (3) ≥ 15 bladder cancer tissue samples. Ultimately, [68]GSE133624 (29 normal vs 36 BLCA samples) [[69]19] and [70]GSE40355 (8 normal vs 16 BLCA samples) [[71]20] were included. For prognostic validation of CCDC137, GEO datasets meeting the following criteria were analyzed: (1) CCDC137 mRNA expression profiling; (2) Availability of overall survival (OS) data; (3) ≥ 70 bladder cancer tissue samples. [72]GSE176307 (BLCA samples = 79) [[73]21] and [74]GSE48075 (BLCA samples = 73) [[75]22] were ultimately used. Construction of CCDC scoring model Differentially expressed CCDC genes in TCGA-BLCA were identified using the DESeq2 R package, with significance thresholds defined as false discovery rate (FDR) < 0.05 and |log[2] fold change|> 0.585 [[76]23]. Subsequently, the randomForestSRC R package was employed to perform random survival forest analysis, where 20 prognostically relevant CCDC genes were selected from 55 DEG-CCDC using the var.select function. A prognostic model comprising 10 core CCDC genes was ultimately constructed through least absolute shrinkage and selection operator (LASSO) regression (1000 times) implemented via the glmnet R package [[77]24]. The CCDC score for each TCGA-BLCA patient was calculated using these core genes and their corresponding LASSO penalty coefficients (Additional file [78]1). graphic file with name d33e628.gif Log-rank tests for intergroup prognostic differences were performed using the survminer and survival R packages, with Kaplan–Meier curves generated accordingly. Time-dependent receiver operating characteristic (Time-ROC) analysis was conducted using the timeROC R package to evaluate the prognostic capacity of the CCDC score. Furthermore, the prognostic value of the CCDC score was evaluated in two independent validation cohorts: [79]GSE32894 [[80]25] and E-MTAB-4321 [[81]26] ([82]https://www.ebi.ac.uk/) (Additional file [83]1). Evaluation of immune infiltration and retrieval of immunotherapy-related scores To characterize the tumor immune microenvironment landscape in the TCGA-BLCA cohort, the Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data (ESTIMATE), Cell-type Identification by Estimating Relative Subsets of RNA Transcripts (CIBERSORT), and single sample gene set enrichment analysis (ssGSEA) algorithms were employed. These analyses were performed using the immunedeconv [[84]27] and GSVA [[85]28] R packages, based on previously published immune function and immune cell-related gene sets [[86]29, [87]30]. The gene sets of immune checkpoints and chemokines were collected from previously published literature (Additional file [88]1) [[89]31, [90]32]. The tumor immune dysfunction and exclusion (TIDE) score and Immunophenoscore (IPS), which reflect tumor immune evasion and immunogenicity, respectively, were utilized to predict immunotherapy sensitivity in the TCGA-BLCA dataset. TIDE scores for the TCGA-BLCA cohort were obtained from the TIDE website ([91]http://tide.dfci.harvard.edu/) [[92]33], while IPS values were calculated according to the method described by Charoentong et al. [[93]30]. Single-cell sequencing data and spatial transcriptomics data analysis A single-cell sequencing dataset comprising 3 adjacent normal tissues and 8 bladder cancer tissues was downloaded from the EMBL’s European Bioinformatics Institute (EMBL-EBI) ([94]https://www.ebi.ac.uk/) (BioProject: PRJNA662018 [[95]34]) and processed using the Seurat package (version 4.4.0 [[96]35]) for analysis and visualization. High-quality cells were retained based on the following criteria: 1000 < unique molecular identifiers (UMIs) < 50,000, 200 < Genes < 7000, mitochondrial gene percentage < 15%, and hemoglobin gene percentage < 1%. The expression matrix was normalized using the NormalizeData function, followed by the identification of the top 2000 highly variable genes across cells using the FindVariableFeatures function. Batch effects were removed by integrating expression data from different samples using the ScaleData function. Principal component analysis (PCA) was performed on the integrated single-cell expression matrix, and the top 30 principal components were extracted for downstream analysis. Cell clustering was performed using the FindClusters function (resolution = 0.3), and cell populations were annotated based on known cell markers (Additional file [97]1). Uniform Manifold Approximation and Projection (UMAP) was utilized for dimensionality reduction and visualization. To characterize CCDC137-related tumor cell features, epithelial cells in bladder cancer tissues were classified into CCDC137 positive (CCDC137 + , CCDC137 expression > 0) and CCDC137 negative (CCDC137 − , CCDC137 expression = 0) cell subgroups based on CCDC137 expression [[98]36]. The functional states of the two epithelial cell subgroups were evaluated using Gene Set Variation Analysis (GSVA) [[99]28] based on the Hallmark gene sets. Transcription factor activity, CNV, and differentiation scores for the two epithelial cell subgroups were calculated using the decoupleR [[100]37], infercnv, and CytoTRACE [[101]38] R packages, respectively. To analyze immune cell profile alterations in patients with differential CCDC137 expression, scRNA-seq of bladder cancer tissues (n = 8) from PRJNA662018 were stratified based on the median expression value of CCDC137 [[102]39]. Processed spatial transcriptomics data from 4 bladder cancer samples were downloaded from the GEO database (accession number: [103]GSE171351 [[104]40]). Module scores for each spatial spot were calculated using the AddModuleScore function in the Seurat package (version 4.4.0) based on known cell markers (Additional file [105]1). Tissue microarray and immunohistochemistry A bladder cancer tissue microarray (TMA) was purchased from Shanghai Zhuoli Biotechnology Co., Ltd (Cat. #ZL-BlaU961, Shanghai, China), containing 36 adjacent normal tissue spots and 60 bladder cancer tissue spots (diameter: 1.5 mm). Following standard deparaffinization and antigen retrieval procedures, tissue sections were incubated with 3% hydrogen peroxide solution at room temperature in the dark for 25 min to block endogenous peroxidase activity. Subsequently, the TMA sections were blocked with 3% bovine serum albumin (BSA) at room temperature for 30 min. Primary antibody binding was performed by incubating sections with anti-CCDC137 (1:50 dilution; Cat. #46419-1, Signalway Antibody) at 4 °C overnight (approximately 9 h). After washing with PBS three times (5 min each), sections were incubated with the corresponding secondary antibody (1:200 dilution; Cat. #GB23303, Servicebio) at room temperature for 50 min. Following another three PBS washes (5 min each), sections were stained with DAB chromogen (Cat. #G1212, Servicebio), counterstained with hematoxylin (Cat. #G1004, Servicebio) for nuclear visualization, and finally dehydrated and mounted. The prepared TMA slides were scanned, and QuPath software (version 0.3.0) [[106]41] was used to calculate QuPath scores for evaluating immunohistochemical (IHC) staining intensity. Additionally, immunohistochemical (IHC) images of CCDC137 expression were retrieved from the Human Protein Atlas (HPA) website ([107]https://www.proteinatlas.org/), encompassing both normal bladder tissues and bladder carcinoma specimens. These images were analyzed to validate the result from TMA. Cell culture The human normal ureteral epithelial cell line SV-HUC-1 and bladder cancer cell Lines 5637, J82, RT4, RT112, and SW780 were purchased from Pricella Biotechnology (Wuhan, China). The bladder cancer cell lines T24 and UMUC3 were obtained from the Cell Bank of the Shanghai Chinese Academy of Sciences (Shanghai, China). SV-HUC-1 cells were cultured in F-12 K medium (Gibco, USA). T24 and SW780 cells were maintained in DMEM (Gibco, USA). 5637 and RT112 cells were grown in RPMI 1640 medium (Gibco, USA). J82 and RT4 cells were cultured in MEM (Gibco, USA) and McCoy’s 5a medium (Gibco, USA), respectively. All media were supplemented with 10% fetal bovine serum (FBS; Gibco, USA) and 1% penicillin/streptomycin solution (Gibco, USA), and cells were cultured at 37 °C in 5% CO[2]. All cell lines were authenticated by short tandem repeat (STR) profiling. Quantitative reverse transcription PCR Total RNA was extracted using the MolPure® Cell/Tissue Total RNA Kit (Cat. #19221ES50, Yeasen). Reverse transcription was performed according to the manufacturer’s protocol for Hifair® III 1st Strand cDNA Synthesis SuperMix for qPCR (Cat. #11141ES60, Yeasen). The qRT-PCR reaction system was subsequently prepared using Hieff UNICON® Universal Blue qPCR SYBR Green Master Mix (Cat. #11184ES08, Yeasen). qRT-PCR was carried out on the QuantStudio® 3 Real-Time PCR System (Thermo Fisher Scientific). The mRNA level of CCDC137 and SCD normalized to β-actin in cells was determined using the methods of 2^−ΔΔCt. All experiments were conducted in triplicate with independent biological replicates. The following primer sequences were utilized in this study: CCDC137-forward 5’-3’ CCAAGAACCAGGACGAACAG and CCDC137-reverse 5’-3’ CCCCTTCCTCTGTTTGAACTTG, SCD-forward 5’-3’ AGTTCTACACCTGGCTTTGG and SCD-reverse 5’-3’ ACGAGCCCATTCATAGACATC, β-actin-forward 5’-3’ GAGAAAATCTGGCACCACACC and β-actin- reverse 5’-3’ GGATAGCACAGCCTGGATAGCAA. Western blot Cells were lysed with RIPA buffer (Cat. #P0013B, Beyotime) containing 1% protease inhibitor cocktail (Cat. #P1010, Beyotime) to extract total proteins. Protein concentrations were normalized using Pierce BCA Protein Assay Kit (Cat. #23227, Thermo Fisher Scientific). Proteins were mixed with 5 × SDS-PAGE Protein Loading Buffer (Cat. #20315ES20, Yeasen) and denatured at 100 °C for 5 min. Then, protein samples were loaded onto 10% FuturePAGE pre-cast gel (Cat. #ET12010Gel, Boyi Biotech). Electrophoresis was performed at 110V for 85 min, followed by membrane transfer at 120V for 60 min. Membranes were blocked with 5% non-fat milk in TBST for 2 h at room temperature. After three TBST washes (10 min each), primary antibodies were applied and incubated at 4 °C overnight (approximately 9 h). Following three additional TBST washes (10 min each), horseradish peroxidase (HRP)-conjugated secondary antibodies were incubated at room temperature for 50 min. After final TBST washes (10 min each), protein signals were visualized using Omni-ECL™ Pico Light Chemiluminescence Kit (Cat. #SQ202L, EpiZyme). Densities of the western blot bands were quantified using ImageJ software and normalized against β-actin as the internal loading control. All experimental procedures were performed in triplicate with independent biological replicates. Antibodies were diluted in antibody dilution buffer (Cat. #PS119L, EpiZyme) as follows: anti-CCDC137 (1:1000; Cat. #27201-1-AP, Proteintech), anti-β-actin (1:10000; Cat. #GXP6564, Genxspan), anti-SCD (1:3000; Cat. #28678-1-AP, Proteintech), anti-AKT (1:5000; Cat. #10176-2-AP, Proteintech), anti-p-AKT (1:5000; Cat. #66444-1-Ig, Proteintech), HRP-Goat Anti-Mouse Recombinant Secondary Antibody (1:10,000; Cat. #RGAM001, Proteintech), HRP-Goat Anti-Rabbit Recombinant Secondary Antibody (1:10,000; Cat. #RGAR001, Proteintech). Establishment of stably transfected cell lines The shCCDC137 lentiviral particles and corresponding control lentiviral particles (shCtrl) were purchased from Genechem Co., Ltd (Shanghai, China). The shRNA sequences were as follows: shCCDC137-1: TCCGAGACACGGTGAAGTTTG; shCCDC137-2: TTTCCGGCTCCGGGAGATTAT; shCtrl: TTCTCCGAACGTGTCACGT. The SCD-overexpression lentiviral particles (OE-SCD) corresponding control lentiviral particles (OE-NC) were likewise acquired from Genechem Co., Ltd (Shanghai, China). T24 and 5637 bladder cancer cells were maintained in 6-well plates until reaching 20–30% confluence for lentiviral infection. Following 72-h incubation, puromycin selection (2.5 μg/ml, Cat. #ST551, Beyotime) was performed for 48 h to establish stably transfected cells. The transfection efficiency was subsequently verified by western blot in both bladder cancer cell lines. Determination of cell proliferative capacity Stably transfected T24 (2 × 10^3 cells/well) and 5637 (3 × 10^3 cells/well) cells were seeded into 96-well plates for CCK8 assay. CCK8 solution (10 μl/well, Cat. #BMU106-CN, Abbkine) was added at 0h, 24h, 48h, and 72h time points. The plates were then incubated in darkness for 2 h. Absorbance was measured at 450 nm using 0h values as baseline references for calculating proliferation rates. For colony formation