FNDC1 is closely related to poor prognosis and immune cell infiltration in gastric cancer Authors: Meixia Zhang, Qingzi Fu, Bo Jiang, Zhenzhen Yang, Qiaofeng Chen, Peng Chen, Shiyu Zhang Published: 2026 ABSTRACT Fibronectin type III domain containing 1 (FNDC1) in gastric cancer (GC) progression and immune cell infiltration remains unclear. This study aimed to investigate the impact of FNDC1 on GC progression and to elucidate its association with immune cell infiltration in GC. FNDC1 expression profiles in GC and paired normal tissues were analyzed using datasets from the TCGA, GTEx, and GEO databases, with validation conducted through quantitative real-time PCR (qPCR) and immunohistochemistry (IHC). Differentially expressed genes (DEGs) associated with FNDC1 were identified through gene set enrichment analysis (GSEA) and functional enrichment analysis. We evaluated the role of FNDC1 in GC immune cell infiltration. In addition, Cox regression and Kaplan-Meier analyses were conducted on FNDC1 for its prognostic significance and correlation with clinical variables. We evaluated the role of FNDC1 in gastric cancer cell invasion and migration via Transwell and wound-healing assays following FNDC1 knockdown and overexpression. Finally, survival probabilities in GC patients were predicted by nomogram construction. FNDC1 expression in GC tissues significantly increased relative to that in matched normal samples (P < 0.001), as confirmed by qPCR and IHC (both P < 0.05). There were 359 DEGs associated with ECM Receptor Interaction, Focal Adhesion, and the Degradation of the Extracellular Matrix signaling pathway. FNDC1 expression was positively related to macrophage and NK cells infiltration (macrophage, r = 0.564, P < 0.001; NK cells, r = 0.566, P < 0.001; as validated by IHC and further confirmed by in vitro co-culture assays). Additionally, FNDC1 expression was strongly related to T stage (P < 0.001), pathological stage (P < 0.001), histological grade (P < 0.01), histological type (P < 0.001), anatomic neoplasm subdivision (P < 0.001), and unfavorable overall survival (OS, P < 0.05). In vitro cell experiments indicate that FNDC1 expression is closely associated with the invasion and migration of gastric cancer cells. Our established nomogram effectively predicted the 1-, 3-, and 5-year OS probabilities in patients with GC (C-index [95% CI] = 0.714 [0.689–0.738]). FNDC1 is strongly associated with a dismal prognostic outcome and immune cell infiltration in patients with GC. Introduction Gastric cancer (GC), a highly aggressive malignancy of the gastrointestinal tract, ranks fifth in cancer-related mortality and is the fourth leading cause of cancer-associated deaths globally1. Although the early diagnosis rate of GC exceeds 50% in certain countries, such as South Korea and Japan2, it remains markedly low in most parts of the world, contributing to poor prognostic outcomes3. Therefore, the identification of biomarkers closely associated with early GC detection and prognosis is critically needed. The extracellular matrix protein Fibronectin type III domain containing 1 (FNDC1) contributes to the biological mechanisms underlying blood vessel formation4,5. Accumulating evidence has demonstrated that elevated FNDC1 expression is significantly correlated with tumor progression and unfavorable clinical outcomes in colorectal cancer6 and lung adenocarcinoma7. Elevated FNDC1 expression has been observed across multiple cancer types, where it potentially facilitates lung adenocarcinoma (LUAD) metastasis through JAK2-STAT3 signaling pathway activation8. However, the role of FNDC1 in GC has not yet been elucidated. The emergence of immunotherapy has markedly improved clinical outcomes for many patients with advanced stage malignancies9. Compared with conventional chemotherapy, anti-her2 therapy and anti-programmed cell death 1 (PD-1) antibodies have shown remarkable therapeutic effects in untreated GC patients10. New anti-HER2 therapeutics, such as disitamab vedotin (RC48) and T-DXd, have made great achievements in the treatment of GC11. Antitumor immunotherapy efficacy should be predicated by taking effector cell infiltration in the tumor microenvironment (TME) as a prerequisite12. Anti-ctla-4 (T lymphocyte-associated antigen 4) monoclonal antibodies exert their effects on tumors through restoring T-cell growth and extending T-cell stimulation13. T cells are among the tumor-infiltrating lymphocytes (TILs), while additional prevalent TILs include NK cells and macrophages. Consequently, comprehending both the TME and TILs in GC might establish a basis for enhancing the anti-immunotherapy efficacy for GC. Nevertheless, it remains unreported whether FNDC1 can impact TIL levels in GC. The present work examined the effect of FNDC1 on GC via the Cancer Genome Atlas (TCGA), Gene Expression Omnibus (GEO) and Genotype-Tissue Expression (GTEx) databases through bioinformatics analyses, which cover differential gene (DEG) identification, functional enrichment, immune cell infiltration, clinical correlation, survival analyses and gene set enrichment analysis (GSEA). Additionally, overall survival (OS) in patients with GC was predicted through the construction of a nomogram. Finally, the aforementioned bioinformatics analyses were validated using clinical data and tissue samples from patients with histologically confirmed GC. Materials and methods Data acquisition and processing RNA-seq data were collected from 414 TCGA-derived GC and 36 adjacent non-carcinoma tissues (https://portal.gdc.cancer.gov/, November 24, 2024). Additionally, RNA-seq data from 174 non-carcinoma gastric tissues were acquired from the GTEx project through the UCSC Xena platform (http://xena.ucsc.edu/, November 24, 2024). Duplicate samples or those with insufficient information were excluded from subsequent analyses. R software (version 4.2.1) was used to perform log2-transformation of the RNA-seq data. Table 1 presents the clinical features of the GC patients. Table 1The clinical characteristic of Stomach Adenocarcinoma.CharacteristicLevelsOveralln375Gender, n (%)Female134 (35.7%)Male241 (64.3%)Age, n (%)<=65164 (44.2%)> 65207 (55.8%)T stage, n (%)T119 (5.2%)T280 (21.8%)T3168 (45.8%)T4100 (27.2%)N stage, n (%)N0111 (31.1%)N197 (27.2%)N275 (21%)N374 (20.7%)M stage, n (%)M0330 (93%)M125 (7%)Histological type, n (%)Diffuse type63 (16.8%)Mucinous type19 (5.1%)Not otherwise specified207 (55.3%)Papillary type5 (1.3%)Signet ring type11 (2.9%)Tubular type69 (18.4%)Pathologic stage, n (%)Stage I53 (15.1%)Stage II111 (31.5%)Stage III150 (42.6%)Stage IV38 (10.8%)Histologic grade, n (%)G110 (2.7%)G2137 (37.4%)G3219 (59.8%)Residual tumor, n (%)R0298 (90.6%)R115 (4.6%)R216 (4.9%)Primary therapy outcome, n (%)PD65 (20.5%)SD17 (5.4%)PR4 (1.3%)CR231 (72.9%)H pylori infection, n (%)No145 (89%)Yes18 (11%)Barretts esophagus, n (%)No193 (92.8%)Yes15 (7.2%)Anatomic neoplasm subdivision, n (%)Antrum/Distal138 (38.2%)Cardia/Proximal48 (13.3%)Fundus/Body130 (36%)Gastroesophageal Junction41 (11.4%)Other4 (1.1%)Age, median (IQR)67 (58, 73) Table 1The clinical characteristic of Stomach Adenocarcinoma.CharacteristicLevelsOveralln375Gender, n (%)Female134 (35.7%)Male241 (64.3%)Age, n (%)<=65164 (44.2%)> 65207 (55.8%)T stage, n (%)T119 (5.2%)T280 (21.8%)T3168 (45.8%)T4100 (27.2%)N stage, n (%)N0111 (31.1%)N197 (27.2%)N275 (21%)N374 (20.7%)M stage, n (%)M0330 (93%)M125 (7%)Histological type, n (%)Diffuse type63 (16.8%)Mucinous type19 (5.1%)Not otherwise specified207 (55.3%)Papillary type5 (1.3%)Signet ring type11 (2.9%)Tubular type69 (18.4%)Pathologic stage, n (%)Stage I53 (15.1%)Stage II111 (31.5%)Stage III150 (42.6%)Stage IV38 (10.8%)Histologic grade, n (%)G110 (2.7%)G2137 (37.4%)G3219 (59.8%)Residual tumor, n (%)R0298 (90.6%)R115 (4.6%)R216 (4.9%)Primary therapy outcome, n (%)PD65 (20.5%)SD17 (5.4%)PR4 (1.3%)CR231 (72.9%)H pylori infection, n (%)No145 (89%)Yes18 (11%)Barretts esophagus, n (%)No193 (92.8%)Yes15 (7.2%)Anatomic neoplasm subdivision, n (%)Antrum/Distal138 (38.2%)Cardia/Proximal48 (13.3%)Fundus/Body130 (36%)Gastroesophageal Junction41 (11.4%)Other4 (1.1%)Age, median (IQR)67 (58, 73) The clinical characteristic of Stomach Adenocarcinoma. FNDC1 differential expression in pan-cancer and GC tissues Statistical analyses were conducted using appropriate methods tailored to the distributional characteristics of the data, employing the stats and car packages in R. Data visualization was performed using the ggplot2 package. ROC curve analyses were carried out with the pROC package, and the resulting outputs were visualized using ggplot2.To further validate the differential expression of FNDC1 in GC compared with adjacent non-carcinoma tissues, two gene expression profile datasets from the GEO—GSE13911 and GSE54129—were analyzed using the GEO2R online tool (https://www.ncbi.nlm.nih.gov/geo/, accessed November 27, 2024). Subsequently, ROC curve analyses were performed to evaluate the prognostic significance of FNDC1 in GC. Quantitative real-time PCR (q-PCR) of FNDC1 expression in GC and non-carcinoma samples The Ethics Committee of the First Affiliated Hospital of Nanchang University (CDYFY-IACUC-202312QR022) approved our study. All human participants provided informed consent, and the study conformed to the Declaration of Helsinki. Between March and June 2024, GC and non-carcinoma biopsy samples were continuously harvested from 20 GC patients at the Endoscopy Center of the First Affiliated Hospital of Nanchang University (Nanchang, Jiangxi). These obtained tissues were soaked in RNA protective solution before immediate storage in a − 80 °C freezer. No patients received previous antitumor treatments, such as chemotherapy or radiotherapy. Moreover, patients who developed concomitant cancers were excluded from the analysis. RNA extraction and q-PCR First, TRIzol reagent (R0016, Beyotime) was used to extract total tissue RNA, which was subsequently used to prepare first-strand cDNA using the PrimeScript™ RT Reagent Kit with gDNA Eraser (RR047A, Takara). Thereafter, we adopted the 2–∆∆Ct method to compute FNDC1 gene expression, with GAPDH being used as an endogenous reference. Primer sequences: FNDC1-F: 5’- CGGCCTCAGTTGACCA − 3’.FNDC1-R: 5’- AGACTTTTCAGTGACTTCCC − 3’.GAPDH-F: 5’-GTCAGCCGCATCTTCTTT-3’.GAPDH-R: 5’-CGCCCAATACGACCAAAT-3’. FNDC1-F: 5’- CGGCCTCAGTTGACCA − 3’. FNDC1-F: 5’- CGGCCTCAGTTGACCA − 3’. FNDC1-R: 5’- AGACTTTTCAGTGACTTCCC − 3’. FNDC1-R: 5’- AGACTTTTCAGTGACTTCCC − 3’. GAPDH-F: 5’-GTCAGCCGCATCTTCTTT-3’. GAPDH-F: 5’-GTCAGCCGCATCTTCTTT-3’. GAPDH-R: 5’-CGCCCAATACGACCAAAT-3’. GAPDH-R: 5’-CGCCCAATACGACCAAAT-3’. IHC This study retrospectively enrolled 80 patients who underwent surgery for GC at the First Affiliated Hospital of Nanchang University between June 2022 and December 2022. All cases were pathologically confirmed as primary gastric adenocarcinoma, with no prior history of other malignancies or previous anticancer treatments. For the IHC analysis, paired GC and matched non-carcinoma samples (≥ 5 cm away from the tumor edge) were processed. Formalin-fixed, paraffin-embedded blocks were prepared to construct tissue microarrays. Section (4 μm) were subjected to xylene deparaffinization and gradient alcohol rehydration. Following antigen retrieval and blocking, a primary antibody against FNDC1 (PA5-56602; Thermo Fisher Scientific; 1:500) was added and incubated with the slides at 4 °C for 12 h. Subsequently, an HRP-conjugated secondary polymer (KIT-5009, MXB Biotechnologies, 1:10000) was applied and incubated for 40 min at 26 °C. Chromogenic detection was performed with diaminobenzidine (DAB) substrate (P0202, Beyotime). Three randomly selected fields per sample were imaged using a light microscope (Nikon Eclipse Ci-L). Staining intensity was quantitatively analyzed with ImageJ software (version 1.53t). Graphical presentation and statistical analyses were conducted using GraphPad Prism version 8.0. Identification of DEGs in the high- and low-FNDC1 expression groups In line with FNDC1 expression, the top 50% of samples were assigned to the high-expression group, whereas the remaining samples were assigned to the low-expression group. The DESeq2 package in R was used to identify DEGs for which P.adj < 0.05 and the logFC absolute value was greater than 1.5. The results of the identified differential genes are presented in volcano maps. Functional enrichment of DEGs When the IDs of the DEGs were converted via the “Hs.eg.db” package of R software, the clusterProfiler package was utilized for subsequent functional enrichment with thresholds of a q value < 0.2 and p.adj < 0.05. GSEA visualization and statistical analysis were completed with the clusterProfiler package, utilizing the C2 collection in MSigDB. Gene set permutations were conducted 10,000 times, with the significance thresholds adjusted to P < 0.05 and a false discovery rate (FDR) < 0.25. Immune cell infiltration Immune cell infiltration within tumor tissues was assessed through single-sample GSEA (ssGSEA) with the GSVA package14, which utilizes published gene signatures for quantification15. Statistical analysis was carried out to investigate the relationship between FNDC1 expression and immune infiltration levels via Spearman’s correlation, and the Wilcoxon signed-rank test was used to analyze differential immune infiltration patterns in the FNDC1 expression groups. For immunohistochemical evaluation, FNDC1 expression levels were quantified by calculating the average optical density (AOD) from tissue samples of 80 GC patients using ImageJ software. Patients were stratified into high and low FNDC1 expression cohorts based on the median AOD value. From each cohort, 10 cases were randomly selected for additional immunohistochemical staining to assess macrophage populations, following the protocol outlined in “IHC” section. NK cells and macrophages were respectively identified with anti-CD56 (MAB-0743) and anti-CD68 (MAB-0041) antibodies (both from MXB biotechnologies). Quantitative analysis of marker-positive areas was conducted via ImageJ, and GraphPad Prism 8.0 was used to visualize and analyze the data. THP-1 cells (1 × 10⁶) were stimulated with 320 nM PMA (MedChemExpress, 16561-29-8) for 24 h to induce differentiation into macrophage-like cells. The differentiated cells were then co-cultured with gastric cancer cell line MKN7 in a Transwell system (0.4 μm pore size, Corning, 3413). Specifically, 5 × 10⁵ MKN7 cells were seeded into the upper chamber, while the PMA-treated THP-1-derived macrophages were placed in the lower chamber of a 6-well plate. Following 48 h of co-culture, macrophages were harvested for subsequent qRT-PCR analyses. Clinical correlation analysis of FNDC1 in GC The relationships between FNDC1 expression levels and clinicopathological features were assessed using logistic regression analysis and Wilcoxon signed-rank tests. To assess how the FNDC1 level and additional clinicopathological features affect survival outcomes in GC patients, univariate and multivariate Cox regression analyses were conducted. Multivariate Cox regression was further utilized to identify factors independently predicting GC prognosis. Nomogram establishment and validation We developed a nomogram integrating all independent prognostic factors to estimate 1-, 3-, and 5-year survival probabilities for GC patients. The prognostic variables included were based on previously published studies16. Nomogram construction was performed using R software, specifically employing the rms and survival packages. Model parameters were determined through bootstrap resampling, with 200 iterations conducted for each sample group to enhance stability. The predictive performance of the nomogram was assessed using the concordance index (C-index), with values approaching 1 indicating higher predictive accuracy. Cell culture Human gastric cancer cell lines MKN7 (ZM0715) were obtained from Zhongqiao Xinzhou Cell Research. Cells were cultured at 37 °C under 5% CO₂ in MEM medium (8120294, Gibco, USA) supplemented with 10% fetal bovine serum (A511-001, Lonsera, Uruguay) and 1% penicillin/streptomycin (P1400, Solarbio, China). Lentiviral transduction For FNDC1 knockdown, lentiviral vectors were constructed using the pLKO.1-Puro backbone, containing either FNDC1-targeting shRNAs or a non-targeting control. Lentiviral particles were generated by co-transfecting HEK-293T cells with the lentiviral vector plasmid, psPAX2 (BR036, Fenghui), and PMD2.G (BR037, Fenghui) using Lipofectamine™ 3000 (L3000150, Thermo Fisher). Viral supernatants were collected at 48 and 72 h, filtered (0.22 μm), and used to infect MKN7 cells. Stable cell lines were selected with 2 µg/mL puromycin. Protein extraction and western blotting Cells were lysed on ice using RIPA buffer (P0013K, Beyotime, China) containing PMSF (P0100, Solarbio, China). Lysates were centrifuged (12,000 × g, 15 min, 4 °C), and supernatants were mixed with loading buffer (P1015, Solarbio, China), denatured (100 °C, 10 min), and stored at − 20 °C. Proteins were separated by SDS-PAGE, transferred to PVDF membranes (0.22 μm), and blocked with 5% non-fat milk. Membranes were incubated with primary antibody (57946, CST, UK; 1:1000) overnight at 4 °C, followed by incubation with fluorescent secondary antibody (5151P, CST, UK; 1:10000) for 1 h at room temperature. Signals were detected using a Li-COR Odyssey system. Transwell invasion assay Cell invasion was analyzed using Matrigel-coated (60 µL; #256234, BD Biosciences) Transwell inserts (8 μm; Corning). Gastric cancer cells (5 × 10⁴) in serum-free medium were seeded in the upper chamber, with complete medium containing 10% FBS in the lower chamber. After 48 h, non-invading cells were removed, and invaded cells were fixed with 4% paraformaldehyde, stained with 0.1% crystal violet, and counted microscopically. Wound healing assay Cells were seeded in 6-well plates and cultured to approximately 95% confluency. A straight wound was generated in the monolayer using a sterile pipette tip. Following two gentle washes with PBS to remove detached cells, fresh serum-free medium was added to suppress cell proliferation. Images of the wound area were captured at 0 and 24 h under a phase-contrast microscope. The relative migration distance was quantified by measuring the wound width at each time point using ImageJ software. Results FNDC1 expression significantly increases inside pan-cancer and GC FNDC1 is significantly upregulated in multiple malignancies, such as GC (Fig. 1A, P < 0.001). TCGA- and GTEx-derived data, comprising 414 GC tissues and 210 normal or adjacent non-tumor tissues, revealed that FNDC1 expression markedly increased in GC samples compared with that in their non-carcinoma counterparts (Fig. 1B, P < 0.001). Furthermore, in 27 paired GC and adjacent tissue samples from the TCGA, FNDC1 expression apparently increased inside tumors compared with that in their matched non-tumor counterparts (Fig. 1C, P < 0.001). In addition, ROC curve analysis (Fig. 1D) demonstrated that FNDC1 expression has a high predictive value for GC diagnosis (AUC = 0.945, 95% CI 0.928–0.962). Fig. 1FNDC1 expression patterns across cancers and associated DEGs. (A) Comparison of FNDC1 expression between cancer and non-carcinoma samples in the TCGA dataset. (B, C) Differential FNDC1 expression in gastric cancer (GC). (D) Diagnostic performance of FNDC1 in distinguishing GC from non-tumor tissues, as evaluated by ROC curve analysis. NS, P ≥ 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001. FNDC1 expression patterns across cancers and associated DEGs. (A) Comparison of FNDC1 expression between cancer and non-carcinoma samples in the TCGA dataset. (B, C) Differential FNDC1 expression in gastric cancer (GC). (D) Diagnostic performance of FNDC1 in distinguishing GC from non-tumor tissues, as evaluated by ROC curve analysis. NS, P ≥ 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001. IHC analysis of 80 GC patients confirmed that FNDC1 expression inside tumor samples apparently increased compared with that in non-carcinoma samples (Fig. 2A–E, P < 0.0001). Similarly, according to the real-time PCR results, the FNDC1 transcript levels in the GC samples apparently increased relative to those in the non-carcinoma samples (Fig. 2F, P < 0.0001). Fig. 2IHC and qRT-PCR analysis of FNDC1 expression. (A–D) Representative IHC staining for FNDC1 in gastric cancer (GC) and non-carcinoma samples at 200× and 400× magnification. (E) Statistical analysis of FNDC1 protein expression in 80 clinically confirmed GC cases. (F) Comparative qRT-PCR assay for FNDC1 mRNA levels inside GC and matched normal samples from 20 patients. ****, P < 0.0001. IHC and qRT-PCR analysis of FNDC1 expression. (A–D) Representative IHC staining for FNDC1 in gastric cancer (GC) and non-carcinoma samples at 200× and 400× magnification. (E) Statistical analysis of FNDC1 protein expression in 80 clinically confirmed GC cases. (F) Comparative qRT-PCR assay for FNDC1 mRNA levels inside GC and matched normal samples from 20 patients. ****, P < 0.0001. To analyze differential FNDC1 expression and its diagnostic predictive value, we analyzed RNA-seq data from the GSE13911 and GSE54129 datasets obtained from GEO. The results revealed that FNDC1 expression in GC samples was markedly greater than that in adjacent samples in both datasets (Fig. 3A,B, P < 0.001). Moreover, FNDC1 expression exhibited strong predictive power for GC diagnosis (Fig. 3C, GSE13911: AUC = 0.931, 95% CI 0.864–0.999; Fig. 3D, GSE54129: AUC = 0.893, 95% CI: 0.837–0.949). Fig. 3Differential FNDC1 expression in gastric cancer and its diagnostic value in GEO datasets. (A, B) Differential FNDC1 expression in GC and matched non-tumor samples in two independent GEO datasets (GSE13911 and GSE54129). (C, D) ROC curves evaluating the ability of FNDC1 to discriminate GC samples from non-carcinoma samples in the respective datasets. *** P < 0.001. Differential FNDC1 expression in gastric cancer and its diagnostic value in GEO datasets. (A, B) Differential FNDC1 expression in GC and matched non-tumor samples in two independent GEO datasets (GSE13911 and GSE54129). (C, D) ROC curves evaluating the ability of FNDC1 to discriminate GC samples from non-carcinoma samples in the respective datasets. *** P < 0.001. DEG detection and functional enrichment A total of 174 downregulated DEGs and 185 upregulated DEGs related to FNDC1 expression were identified (Fig. 4A). Fig. 4FNDC1-associated DEGs and functional enrichment. (A) Volcano plot showing FNDC1-related DEGs. The red dots represent genes whose expression was significantly upregulated, whereas the blue dots indicate those whose expression was downregulated (|log2FC| > 1.5, adjusted P < 0.05). (B) Top three significantly enriched functional terms in the GO annotation, namely, BP, MF and CC, as well as KEGG. (C–H) The top six associated pathways obtained from GSEA. NES, normalized enrichment score; p.adj, adjusted p-value; FDR, false discovery rate. FNDC1-associated DEGs and functional enrichment. (A) Volcano plot showing FNDC1-related DEGs. The red dots represent genes whose expression was significantly upregulated, whereas the blue dots indicate those whose expression was downregulated (|log2FC| > 1.5, adjusted P < 0.05). (B) Top three significantly enriched functional terms in the GO annotation, namely, BP, MF and CC, as well as KEGG. (C–H) The top six associated pathways obtained from GSEA. NES, normalized enrichment score; p.adj, adjusted p-value; FDR, false discovery rate. To further analyze DEG functions, enrichment analyses involving biological process (BP), cellular component (CC), and molecular function (MF) categories and KEGG pathways were conducted. As shown in Fig. 4B, the DEGs were predominantly associated with the following terms: BP (external encapsulating structure organization, skin development, and keratinization), CC (collagen-containing extracellular matrix, cornified envelope and collagen trimer), MF (extracellular matrix structural constituent, glycosaminoglycan binding and extracellular matrix structural constituent conferring compression resistance), and KEGG (Protein digestion and absorption, ECM-receptor interaction and Staphylococcus aureus infection). Only the top three enriched terms are displayed here; detailed enrichment results are provided in Table 2. Table 2Details of functional enrichment analysis results of DEGs.OntologyIDDescriptionGene ratioBg ratiop valuep. adjustBPGO:0045229External encapsulating structure organization37/333310/18,8003.34e−201.01e−16BPGO:0043588Skin development35/333296/18,8005e−197.53e−16BPGO:0031424Keratinization21/33385/18,8001.36e−181.09e−15BPGO:0030198Extracellular matrix organization35/333307/18,8001.63e−181.09e−15BPGO:0043062Extracellular structure organization35/333308/18,8001.81e−181.09e−15CCGO:0062023Collagen-containing extracellular matrix64/347429/19,5941.46e−404.01e−38CCGO:0001533Cornified envelope12/34745/19,5941.34e−111.84e−09CCGO:0005581Collagen trimer14/34786/19,5943.3e−103.03e−08CCGO:0098644Complex of collagen trimers8/34722/19,5942.3e−091.58e−07CCGO:0005796Golgi lumen14/347104/19,5944.34e−092.39e−07MFGO:0005201Extracellular matrix structural constituent41/324172/18,4108.89e−353.8e−32MFGO:0005539Glycosaminoglycan binding31/324234/18,4101.69e−183.61e−16MFGO:0005518Collagen binding15/32468/18,4106.66e−139.51e−11MFGO:0030021Extracellular matrix structural constituent conferring compression resistance10/32422/18,4101.33e−121.42e−10MFGO:0048018Receptor ligand activity35/324489/18,4101.92e-121.48e−10KEGGhsa04974Protein digestion and absorption16/149103/81643.88e-118.47e−09KEGGhsa04512ECM-receptor interaction11/14988/81644.96e-075.4e−05KEGGhsa05150Staphylococcus aureus infection10/14996/81648.93e-060.0006KEGGhsa04972Pancreatic secretion10/149102/81641.54e−050.0007KEGGhsa05146Amoebiasis10/149102/81641.54e−050.0007DEGs: Differential Expression Genes; BP: Biological Process; MF: Molecular Function; CC: Cellular Component; KEGG: Kyoto Encyclopedia of Genes and Genomes. Table 2Details of functional enrichment analysis results of DEGs.OntologyIDDescriptionGene ratioBg ratiop valuep. adjustBPGO:0045229External encapsulating structure organization37/333310/18,8003.34e−201.01e−16BPGO:0043588Skin development35/333296/18,8005e−197.53e−16BPGO:0031424Keratinization21/33385/18,8001.36e−181.09e−15BPGO:0030198Extracellular matrix organization35/333307/18,8001.63e−181.09e−15BPGO:0043062Extracellular structure organization35/333308/18,8001.81e−181.09e−15CCGO:0062023Collagen-containing extracellular matrix64/347429/19,5941.46e−404.01e−38CCGO:0001533Cornified envelope12/34745/19,5941.34e−111.84e−09CCGO:0005581Collagen trimer14/34786/19,5943.3e−103.03e−08CCGO:0098644Complex of collagen trimers8/34722/19,5942.3e−091.58e−07CCGO:0005796Golgi lumen14/347104/19,5944.34e−092.39e−07MFGO:0005201Extracellular matrix structural constituent41/324172/18,4108.89e−353.8e−32MFGO:0005539Glycosaminoglycan binding31/324234/18,4101.69e−183.61e−16MFGO:0005518Collagen binding15/32468/18,4106.66e−139.51e−11MFGO:0030021Extracellular matrix structural constituent conferring compression resistance10/32422/18,4101.33e−121.42e−10MFGO:0048018Receptor ligand activity35/324489/18,4101.92e-121.48e−10KEGGhsa04974Protein digestion and absorption16/149103/81643.88e-118.47e−09KEGGhsa04512ECM-receptor interaction11/14988/81644.96e-075.4e−05KEGGhsa05150Staphylococcus aureus infection10/14996/81648.93e-060.0006KEGGhsa04972Pancreatic secretion10/149102/81641.54e−050.0007KEGGhsa05146Amoebiasis10/149102/81641.54e−050.0007DEGs: Differential Expression Genes; BP: Biological Process; MF: Molecular Function; CC: Cellular Component; KEGG: Kyoto Encyclopedia of Genes and Genomes. Details of functional enrichment analysis results of DEGs. DEGs: Differential Expression Genes; BP: Biological Process; MF: Molecular Function; CC: Cellular Component; KEGG: Kyoto Encyclopedia of Genes and Genomes. GSEA revealed that FNDC1-associated DEGs were most significantly enriched in six pathways, ranked in descending order by NES: Ecm Receptor Interaction, Focal Adhesion, Degradation of the Extracellular Matrix, Extracellular Matrix Organization, NABA Core Matrisome and Avb3 Integrin Pathway (Fig. 4C,H). FNDC1 expression is tightly linked to immune cell infiltration FNDC1 expression was positively correlated with NK cells and macrophages infiltration (Fig. 5A–C, NK cells, R = 0.566, P < 0.001; macrophages, R = 0.564, P < 0.001). Compared to low-FNDC1 samples, high-FNDC1 samples exhibited significantly elevated enrichment scores for NK cells (Fig. 5E, P < 0.001), while macrophages enrichment indices showed a non-significant increasing trend (Fig. 5D, P > 0.05). To further investigate these associations, we performed IHC staining of NK cells (CD56) and macrophages (CD68). Compared to adjacent non-tumor tissues, GC specimens demonstrated significantly increased proportions of both CD56-positive areas (Fig. 6A, C, and E, P < 0.0001) and CD68-positive regions (Fig. 6B, D and F, P < 0.0001) as quantified by digital image analysis. We employed a co-culture model to investigate the impact of FNDC1 expression in GC cells on tumor-associated macrophages (TAMs). qPCR analysis demonstrated that macrophage markers were significantly upregulated in the FNDC1 overexpression group, whereas they were markedly downregulated in the FNDC1 knockout group ((Fig. 6G, P < 0.001). Fig. 5Analysis of FNDC1-associated immune cell infiltration within GCs. (A) Lollipop plot illustrating the association of FNDC1 with immune cell infiltration levels inside GCs. The circle size represents the strength of the correlation (larger size = stronger correlation), whereas the color intensity indicates statistical significance (blue gradient: darker = lower P value). (B, C) Scatter plots demonstrating the relationships of FNDC1 with (B) NK cell and (C) macrophage infiltration levels. (D–E) Box plots comparing the enrichment scores of (D) macrophages and (E) NK cells in the FNDC1-high versus FNDC1-low-expression groups. ***P < 0.001. Analysis of FNDC1-associated immune cell infiltration within GCs. (A) Lollipop plot illustrating the association of FNDC1 with immune cell infiltration levels inside GCs. The circle size represents the strength of the correlation (larger size = stronger correlation), whereas the color intensity indicates statistical significance (blue gradient: darker = lower P value). (B, C) Scatter plots demonstrating the relationships of FNDC1 with (B) NK cell and (C) macrophage infiltration levels. (D–E) Box plots comparing the enrichment scores of (D) macrophages and (E) NK cells in the FNDC1-high versus FNDC1-low-expression groups. ***P < 0.001. Fig. 6IHC and quantitative analysis of specific markers for macrophages and NK cells. (A–D) Representative IHC of CD68 (macrophage-specific marker) and CD56 (NK cell-specific marker) in adjacent non-tumor tissues and gastric cancer tissues at 200× magnification. Brown regions indicate positive staining. (H) Quantitative comparison of CD68-positive areas between gastric cancer and matched adjacent tissues. (E, F) Quantitative comparison of CD56 + and CD68 + immunopositive areas in gastric cancer versus adjacent non-tumor tissues. (G) Expression levels of CD68 mRNA in different co-culture groups. ***, P < 0.001; ****P < 0.0001. IHC and quantitative analysis of specific markers for macrophages and NK cells. (A–D) Representative IHC of CD68 (macrophage-specific marker) and CD56 (NK cell-specific marker) in adjacent non-tumor tissues and gastric cancer tissues at 200× magnification. Brown regions indicate positive staining. (H) Quantitative comparison of CD68-positive areas between gastric cancer and matched adjacent tissues. (E, F) Quantitative comparison of CD56 + and CD68 + immunopositive areas in gastric cancer versus adjacent non-tumor tissues. (G) Expression levels of CD68 mRNA in different co-culture groups. ***, P < 0.001; ****P < 0.0001. FNDC1 promotes the migration and invasion of gastric cancer cells To explore the functional role of FNDC1 in gastric cancer progression, we generated FNDC1-knockout (KO) and FNDC1-overexpression (OE) models in MKN7 gastric cancer cell lines. Western blot analysis validated the successful manipulation of FNDC1 expression, showing a significant reduction in FNDC1 protein levels in KO cells and a substantial increase in OE cells relative to control cells (Fig. 7A,B). Functional assays demonstrated that FNDC1 knockdown markedly impaired the migratory and invasive capacities of both cell lines. In contrast, FNDC1 overexpression robustly enhanced these aggressive phenotypes (Fig. 7C–E). Fig. 7FNDC1 promotes the invasion and migration of gastric cancer cells. (A, B) Western blot analysis revealed that FNDC1 expression was significantly decreased in the knockout (KO) group compared to the control group, while its expression was markedly increased in the overexpression (OE) group. Statistical results were derived from three independent experiments. (C, D) The OE group exhibited enhanced invasive capacity in both gastric cancer cell lines compared to the control group, whereas the KO group showed the opposite effect. (C, E) Similarly, the OE group demonstrated significantly greater migratory ability in both cell lines relative to the control group, while the KO group displayed reduced migration. ****P < 0.0001; ***P < 0.001; **P < 0.01; NS, Not Significant. FNDC1 promotes the invasion and migration of gastric cancer cells. (A, B) Western blot analysis revealed that FNDC1 expression was significantly decreased in the knockout (KO) group compared to the control group, while its expression was markedly increased in the overexpression (OE) group. Statistical results were derived from three independent experiments. (C, D) The OE group exhibited enhanced invasive capacity in both gastric cancer cell lines compared to the control group, whereas the KO group showed the opposite effect. (C, E) Similarly, the OE group demonstrated significantly greater migratory ability in both cell lines relative to the control group, while the KO group displayed reduced migration. ****P < 0.0001; ***P < 0.001; **P < 0.01; NS, Not Significant. FNDC1 expression is strongly associated with GC clinicopathological variables We obtained and analyzed RNAseq and clinical data from 375 STAD cases derived from TCGA. Statistical analysis revealed that FNDC1 expression was significantly associated with pathological T stage (Fig. 8A, T1 vs. T2&T3&T4, P < 0.001), pathological stage (Fig. 8D: Stage I vs. Stages II, III, and IV, P < 0.001), histological grade (Fig. 8E: G1&G2 vs. G3, P < 0.01), race (Fig. 8F, Asian vs. non-Asian, P < 0.01), histological type (Fig. 8G: Tubular type vs. non-tubular type, P < 0.001), and anatomic neoplasm subdivision (Fig. 8H, Gastroesophageal junction (GEJ) versus non-GEJ regions, P < 0.001). However, no significant associations were observed between FNDC1 expression and pathological M stage, N stage, primary therapy outcome, sex, race, or age (Fig. 8B, C, and I–L). Fig. 8Relationship of FNDC1 expression with GC clinicopathological features from the TCGA database. (A–L) Correlations between FNDC1 expression levels and (A) Pathological T stage, (B) Pathological N stage, (C) Pathological M stage, (D) Pathological stage, (E) Histological grade, (F) Race, (G) Histological type, (H) Anatomic neoplasm subdivision, (I) Residual tumor, (J) Primary therapy outcome, (K) Age, (L) Gender. Gastroesophageal junction, GEJ. **P < 0.01; ***P < 0.001. Relationship of FNDC1 expression with GC clinicopathological features from the TCGA database. (A–L) Correlations between FNDC1 expression levels and (A) Pathological T stage, (B) Pathological N stage, (C) Pathological M stage, (D) Pathological stage, (E) Histological grade, (F) Race, (G) Histological type, (H) Anatomic neoplasm subdivision, (I) Residual tumor, (J) Primary therapy outcome, (K) Age, (L) Gender. Gastroesophageal junction, GEJ. **P < 0.01; ***P < 0.001. We further expanded our investigation by collecting clinicopathological variables from an additional cohort of 80 GC patients and analyzing their relationship with FNDC1 expression levels. The analysis revealed statistically significant correlations between FNDC1 expression and pathological stage (Table 3, pathological stage I vs. II & III & IV, P = 0.007), T stage (T1 vs. T2 & T3 & T4, P = 0.024), histological grade (G1 & G2 vs. G3, P = 0.007), and histological type (tubular type vs. non-tubular type, P = 0.002). These clinical findings provide substantial validation and reinforce the accuracy of our initial data mining analysis. Table 3The correlation between FNDC1 expression and clinical variables in 80 gastric cancer patients.CharacteristicsHighLowP valuen4139Gender (M/F), n (%)0.529 F12 (15%)9 (11.2%) M29 (36.2%)30 (37.5%)Age (years), median (IQR)66 (59, 70)63 (57.5, 72)0.773Pathologic.stage, n (%) 0.007 I10 (12.5%)21 (26.2%) II & III & IV31 (38.8%)18 (22.5%)T.stage, n (%) 0.024 T15 (6.2%)13 (16.2%) T2 & T3 & T436 (45%)26 (32.5%)N.stage, n (%)0.697 N1&N2&N330 (37.5%)27 (33.8%) N011 (13.8%)12 (15%)Histological type, n (%) 0.002 Tubular type6 (7.5%)18 (22.5%) Non-tubular type35 (43.8%)21 (26.2%)Histological grade, n (%) 0.007 G1 & G213 (16.2%)24 (30%) G328 (35%)15 (18.8%)Residual tumor, n (%)0.928 R028 (35%)27 (33.8%) R1 & R213 (16.2%)12 (15%)Primary therapy outcome, n (%)0.247 CR&PR33 (41.2%)35 (43.8%) PD&SD8 (10%)4 (5%)Anatomic neoplasm subdivision, n (%)0.655 Gastroesophageal Junction3 (3.8%)5 (6.2%) Non-gastroesophageal junction38 (47.5%)34 (42.5%)Significant values are in bold. Table 3The correlation between FNDC1 expression and clinical variables in 80 gastric cancer patients.CharacteristicsHighLowP valuen4139Gender (M/F), n (%)0.529 F12 (15%)9 (11.2%) M29 (36.2%)30 (37.5%)Age (years), median (IQR)66 (59, 70)63 (57.5, 72)0.773Pathologic.stage, n (%) 0.007 I10 (12.5%)21 (26.2%) II & III & IV31 (38.8%)18 (22.5%)T.stage, n (%) 0.024 T15 (6.2%)13 (16.2%) T2 & T3 & T436 (45%)26 (32.5%)N.stage, n (%)0.697 N1&N2&N330 (37.5%)27 (33.8%) N011 (13.8%)12 (15%)Histological type, n (%) 0.002 Tubular type6 (7.5%)18 (22.5%) Non-tubular type35 (43.8%)21 (26.2%)Histological grade, n (%) 0.007 G1 & G213 (16.2%)24 (30%) G328 (35%)15 (18.8%)Residual tumor, n (%)0.928 R028 (35%)27 (33.8%) R1 & R213 (16.2%)12 (15%)Primary therapy outcome, n (%)0.247 CR&PR33 (41.2%)35 (43.8%) PD&SD8 (10%)4 (5%)Anatomic neoplasm subdivision, n (%)0.655 Gastroesophageal Junction3 (3.8%)5 (6.2%) Non-gastroesophageal junction38 (47.5%)34 (42.5%)Significant values are in bold. The correlation between FNDC1 expression and clinical variables in 80 gastric cancer patients. Significant values are in bold. High FNDC1 expression is related to unfavorable GC prognostic outcomes We further performed survival analysis using clinical data from patients with GC obtained from the TCGA database. FNDC1 expression was strongly associated with OS, with the high-FNDC1 expression group showing poorer OS (Fig. 9A, P = 0.013). However, no significant correlation between FNDC1 expression and disease-specific survival (DSS) or the progression-free interval (PFI) was detected (Fig. 9B and C, P > 0.05). Fig. 9(A–C) Kaplan-Meier survival curves comparing OS, DSS, and PFI between the FNDC1-high and FNDC1-low-expression groups. (D) The nomogram constructed to predict 1-, 3-, and 5-year OS in patients with GC. (E) Calibration plots for our constructed nomogram. (A–C) Kaplan-Meier survival curves comparing OS, DSS, and PFI between the FNDC1-high and FNDC1-low-expression groups. (D) The nomogram constructed to predict 1-, 3-, and 5-year OS in patients with GC. (E) Calibration plots for our constructed nomogram. According to multivariate Cox regression, the histological grade, age, FNDC1 expression level, and primary therapy outcome were found to independently predict OS (Table 4). In contrast, FNDC1 expression was not significantly related to either DSS or PFI (Tables 5 and 6). Table 4Univariate and multivariate regression analysis of Overall Survival (OS) related factors in patients with gastric cancer.CharacteristicsTotal(N)Univariate analysisMultivariate analysisHazard ratio (95% CI)P valueHazard ratio (95% CI)P valuePathologic T stage362 T118ReferenceReference T2&T4&T33448.829 (1.234–63.151) 0.030 13092840.3686 (0.000–Inf)0.995Pathologic N stage352 N0107ReferenceReference N1&N2&N32451.925 (1.264–2.931) 0.002 1.738 (0.886–3.407)0.108Pathologic M stage352 M0327ReferenceReference M1252.254 (1.295–3.924) 0.004 1.005 (0.412–2.448)0.992Pathologic stage347 Stage I50ReferenceReference Stage II&Stage III&Stage IV2972.247 (1.210–4.175) 0.010 0.978 (0.349–2.740)0.966Histologic grade361 G1&G2144ReferenceReference G32171.353 (0.957–1.914)0.0871.582 (1.003–2.494) 0.048 Race320 Asian73Reference Black or African American&White2471.473 (0.890–2.439)0.132Gender370 Female133Reference Male2371.267 (0.891–1.804)0.188Age367 <= 65163ReferenceReference > 652041.620 (1.154–2.276) 0.005 1.909 (1.227–2.971) 0.004 Histological type369 Tubular type69Reference Diffuse Type&Mucinous Type&Not Otherwise Specified&Papillary Type&Signet Ring Type3001.157 (0.755–1.771)0.503FNDC1370 Low185ReferenceReference High1851.523 (1.093–2.123) 0.013 1.836 (1.199–2.811) 0.005 Anatomic neoplasm subdivision362 Gastroesophageal Junction39Reference Antrum/Distal&Cardia/Proximal&Fundus/Body&Other&Stomach (NOS)3231.247 (0.689–2.256)0.466Primary therapy outcome313 CR&PR233ReferenceReference PD&SD804.103 (2.823–5.964) < 0.001 4.647 (2.961–7.292) < 0.001 Residual tumor325 R0294ReferenceReference R1&R2313.445 (2.160–5.494) < 0.001 1.473 (0.796–2.726)0.217Significant values are in bold. Table 4Univariate and multivariate regression analysis of Overall Survival (OS) related factors in patients with gastric cancer.CharacteristicsTotal(N)Univariate analysisMultivariate analysisHazard ratio (95% CI)P valueHazard ratio (95% CI)P valuePathologic T stage362 T118ReferenceReference T2&T4&T33448.829 (1.234–63.151) 0.030 13092840.3686 (0.000–Inf)0.995Pathologic N stage352 N0107ReferenceReference N1&N2&N32451.925 (1.264–2.931) 0.002 1.738 (0.886–3.407)0.108Pathologic M stage352 M0327ReferenceReference M1252.254 (1.295–3.924) 0.004 1.005 (0.412–2.448)0.992Pathologic stage347 Stage I50ReferenceReference Stage II&Stage III&Stage IV2972.247 (1.210–4.175) 0.010 0.978 (0.349–2.740)0.966Histologic grade361 G1&G2144ReferenceReference G32171.353 (0.957–1.914)0.0871.582 (1.003–2.494) 0.048 Race320 Asian73Reference Black or African American&White2471.473 (0.890–2.439)0.132Gender370 Female133Reference Male2371.267 (0.891–1.804)0.188Age367 <= 65163ReferenceReference > 652041.620 (1.154–2.276) 0.005 1.909 (1.227–2.971) 0.004 Histological type369 Tubular type69Reference Diffuse Type&Mucinous Type&Not Otherwise Specified&Papillary Type&Signet Ring Type3001.157 (0.755–1.771)0.503FNDC1370 Low185ReferenceReference High1851.523 (1.093–2.123) 0.013 1.836 (1.199–2.811) 0.005 Anatomic neoplasm subdivision362 Gastroesophageal Junction39Reference Antrum/Distal&Cardia/Proximal&Fundus/Body&Other&Stomach (NOS)3231.247 (0.689–2.256)0.466Primary therapy outcome313 CR&PR233ReferenceReference PD&SD804.103 (2.823–5.964) < 0.001 4.647 (2.961–7.292) < 0.001 Residual tumor325 R0294ReferenceReference R1&R2313.445 (2.160–5.494) < 0.001 1.473 (0.796–2.726)0.217Significant values are in bold. Univariate and multivariate regression analysis of Overall Survival (OS) related factors in patients with gastric cancer. Significant values are in bold. Table 5Univariate and multivariate regression analysis of Disease Specific Survival (DSS) related factors in patients with gastric cancer.CharacteristicsTotal (N)Univariate analysisMultivariate analysisHazard ratio (95% CI)P valueHazard ratio (95% CI)P valuePathologic T stage364 T119ReferenceReference T2&T4&T33454.394 (1.086–17.776) 0.038 2.065 (0.415–10.274)0.376Pathologic N stage354 N0108ReferenceReference N1&N2&N32461.640 (1.075–2.501) 0.022 1.186 (0.653–2.154)0.576Pathologic M stage353 M0328ReferenceReference M1252.224 (1.194–4.144) 0.012 1.184 (0.534–2.628)0.678Pathologic stage349 Stage I51ReferenceReference Stage II&Stage III&Stage IV2982.547 (1.288–5.039) 0.007 0.885 (0.338–2.315)0.804Histologic grade363 G1&G2145ReferenceReference G32181.540 (1.057–2.245) 0.025 1.566 (1.008–2.431) 0.046 Race322 Asian74Reference Black or African American&White2480.992 (0.624–1.578)0.973Gender372 Female133ReferenceReference Male2391.638 (1.099–2.440) 0.015 1.312 (0.814–2.115)0.265Age369 <= 65164Reference > 652050.858 (0.603–1.221)0.395Histological type371 Tubular Type69Reference Diffuse Type&Mucinous Type&Not Otherwise Specified&Papillary Type&Signet Ring Type3021.036 (0.667–1.607)0.875FNDC1372 Low186Reference High1861.316 (0.923–1.875)0.129Anatomic neoplasm subdivision364 Gastroesophageal Junction40Reference Antrum/Distal&Cardia/Proximal&Fundus/Body&Other&Stomach (NOS)3241.055 (0.594–1.877)0.854Primary therapy outcome315 CR&PR233ReferenceReference PD&SD827.808 (5.332–11.434) < 0.001 7.933 (5.037–12.492) < 0.001 Residual tumor326 R0295ReferenceReference R1&R2313.469 (2.127–5.656) < 0.001 1.581 (0.900–2.776)0.111Significant values are in bold. Table 5Univariate and multivariate regression analysis of Disease Specific Survival (DSS) related factors in patients with gastric cancer.CharacteristicsTotal (N)Univariate analysisMultivariate analysisHazard ratio (95% CI)P valueHazard ratio (95% CI)P valuePathologic T stage364 T119ReferenceReference T2&T4&T33454.394 (1.086–17.776) 0.038 2.065 (0.415–10.274)0.376Pathologic N stage354 N0108ReferenceReference N1&N2&N32461.640 (1.075–2.501) 0.022 1.186 (0.653–2.154)0.576Pathologic M stage353 M0328ReferenceReference M1252.224 (1.194–4.144) 0.012 1.184 (0.534–2.628)0.678Pathologic stage349 Stage I51ReferenceReference Stage II&Stage III&Stage IV2982.547 (1.288–5.039) 0.007 0.885 (0.338–2.315)0.804Histologic grade363 G1&G2145ReferenceReference G32181.540 (1.057–2.245) 0.025 1.566 (1.008–2.431) 0.046 Race322 Asian74Reference Black or African American&White2480.992 (0.624–1.578)0.973Gender372 Female133ReferenceReference Male2391.638 (1.099–2.440) 0.015 1.312 (0.814–2.115)0.265Age369 <= 65164Reference > 652050.858 (0.603–1.221)0.395Histological type371 Tubular Type69Reference Diffuse Type&Mucinous Type&Not Otherwise Specified&Papillary Type&Signet Ring Type3021.036 (0.667–1.607)0.875FNDC1372 Low186Reference High1861.316 (0.923–1.875)0.129Anatomic neoplasm subdivision364 Gastroesophageal Junction40Reference Antrum/Distal&Cardia/Proximal&Fundus/Body&Other&Stomach (NOS)3241.055 (0.594–1.877)0.854Primary therapy outcome315 CR&PR233ReferenceReference PD&SD827.808 (5.332–11.434) < 0.001 7.933 (5.037–12.492) < 0.001 Residual tumor326 R0295ReferenceReference R1&R2313.469 (2.127–5.656) < 0.001 1.581 (0.900–2.776)0.111Significant values are in bold. Univariate and multivariate regression analysis of Disease Specific Survival (DSS) related factors in patients with gastric cancer. Significant values are in bold. Table 6Univariate and multivariate regression analysis of Progress Free Interval (PFI) related factors in patients with gastric cancer.CharacteristicsTotal (N)Univariate analysisMultivariate analysisHazard ratio (95% CI)P valueHazard ratio (95% CI)P valuePathologic T stage345 T117ReferenceReference T2&T4&T332826817877.3790 (0.000 - Inf)0.99412236901.5563 (0.000 - Inf)0.995Pathologic N stage334 N0104ReferenceReference N1&N2&N32301.807 (1.075–3.036) 0.025 1.118 (0.519–2.410)0.776Pathologic M stage333 M0311ReferenceReference M1222.438 (1.221–4.870) 0.012 0.866 (0.356–2.104)0.750Pathologic stage331 Stage I48ReferenceReference Stage II & Stage III & Stage IV2833.259 (1.317–8.065) 0.011 1.166 (0.347–3.912)0.804Histologic grade340 G1&G2135Reference G32051.338 (0.862–2.078)0.194Race305 Asian71Reference Black or African American&White2341.050 (0.593–1.857)0.867Gender349 Female125ReferenceReference Male2241.573 (0.985–2.514)0.0581.168 (0.657–2.076)0.596Age346 <= 65160Reference > 651861.211 (0.797–1.840)0.371Histological type348 Tubular Type67Reference Diffuse Type&Mucinous Type&Not Otherwise Specified&Papillary Type&Signet Ring Type2810.912 (0.553–1.503)0.717FNDC1349 Low176ReferenceReference High1731.563 (1.024–2.384) 0.038 1.577 (0.954–2.606)0.076Anatomic neoplasm subdivision345 Gastroesophageal Junction38Reference Antrum/Distal&Cardia/Proximal&Fundus/Body&Other&Stomach (NOS)3071.379 (0.637–2.988)0.415Primary therapy outcome310 CR&PR232ReferenceReference PD&SD787.990 (5.064–12.606) < 0.001 8.375 (4.774–14.690) < 0.001 Residual tumor314 R0287ReferenceReference R1&R2275.142 (3.014–8.771) < 0.001 2.441 (1.302–4.575) 0.005 Significant values are in bold. Table 6Univariate and multivariate regression analysis of Progress Free Interval (PFI) related factors in patients with gastric cancer.CharacteristicsTotal (N)Univariate analysisMultivariate analysisHazard ratio (95% CI)P valueHazard ratio (95% CI)P valuePathologic T stage345 T117ReferenceReference T2&T4&T332826817877.3790 (0.000 - Inf)0.99412236901.5563 (0.000 - Inf)0.995Pathologic N stage334 N0104ReferenceReference N1&N2&N32301.807 (1.075–3.036) 0.025 1.118 (0.519–2.410)0.776Pathologic M stage333 M0311ReferenceReference M1222.438 (1.221–4.870) 0.012 0.866 (0.356–2.104)0.750Pathologic stage331 Stage I48ReferenceReference Stage II & Stage III & Stage IV2833.259 (1.317–8.065) 0.011 1.166 (0.347–3.912)0.804Histologic grade340 G1&G2135Reference G32051.338 (0.862–2.078)0.194Race305 Asian71Reference Black or African American&White2341.050 (0.593–1.857)0.867Gender349 Female125ReferenceReference Male2241.573 (0.985–2.514)0.0581.168 (0.657–2.076)0.596Age346 <= 65160Reference > 651861.211 (0.797–1.840)0.371Histological type348 Tubular Type67Reference Diffuse Type&Mucinous Type&Not Otherwise Specified&Papillary Type&Signet Ring Type2810.912 (0.553–1.503)0.717FNDC1349 Low176ReferenceReference High1731.563 (1.024–2.384) 0.038 1.577 (0.954–2.606)0.076Anatomic neoplasm subdivision345 Gastroesophageal Junction38Reference Antrum/Distal&Cardia/Proximal&Fundus/Body&Other&Stomach (NOS)3071.379 (0.637–2.988)0.415Primary therapy outcome310 CR&PR232ReferenceReference PD&SD787.990 (5.064–12.606) < 0.001 8.375 (4.774–14.690) < 0.001 Residual tumor314 R0287ReferenceReference R1&R2275.142 (3.014–8.771) < 0.001 2.441 (1.302–4.575) 0.005 Significant values are in bold. Univariate and multivariate regression analysis of Progress Free Interval (PFI) related factors in patients with gastric cancer. Significant values are in bold. Nomogram establishment and validation According to multivariate Cox regression, risk factors independently affecting GC OS (histological grade, age, FNDC1 expression level, and primary therapy outcome) were incorporated to construct a nomogram (Fig. 9D) for predicting 1-, 3-, and 5-year survival rates in patients with GC. Calibration plots (Fig. 9E) demonstrated that our nomogram performed satisfactorily in predicting the 1-, 3-, and 5-year survival probabilities (C-index [95% CI] = 0.714 [0.689–0.738]). Discussion According to bioinformatics analysis, FNDC1 expression markedly increased in GC samples. These findings were further validated by qPCR and IHC analyses. Moreover, as revealed by the analysis of ROC curves, FNDC1 is a potential diagnostic biomarker related to GC. A total of 174 downregulated and 185 upregulated DEGs associated with FNDC1 were identified. GO annotation revealed that these DEGs were primarily enriched in terms related to collagen-containing extracellular matrix, extracellular matrix structural constituent and extracellular matrix structural constituent conferring compression resistance. The KEGG enrichment analysis and GSEA results demonstrated that FNDC1-associated DEGs were predominantly enriched in the following pathways: ECM-receptor interaction, focal adhesion, degradation of the extracellular matrix, extracellular matrix organization, NABA core matrisome, and the αvβ3 integrin pathway. The tumor cells and tumor microenvironment (TME), which constitute the major components of solid tumors, comprise diverse cellular elements, including immune cells and extracellular matrix17. Extensive studies have demonstrated that ECM remodeling in tumor microenvironment is critically involved in gastric cancer initiation and progression18,19. Compelling evidence has established MMP-220, MMP-721, and COL12A121 as pivotal ECM regulators in gastric cancer pathogenesis, with demonstrated roles in driving tumor invasion and predicting adverse survival outcomes. Transmembrane integrins play critical roles in promoting tumorigenesis and drug resistance22,23. Matrix stiffening not only facilitates cellular differentiation, autophagy, and invasive behavior24,25, but also accelerates hepatocellular carcinoma cell migration by enhancing aerobic glycolysis26. Furthermore, immune infiltration analysis demonstrated a significant association between FNDC1 expression levels and tumor immune cell infiltration, particularly involving macrophages and NK cells. The TME comprises various components, including stromal cells and immune cells, which interact dynamically to influence cancer progression and metastasis27,28. The complex crosstalk between stromal and immune elements plays a pivotal role in shaping the TME and driving the development of GC28. Moreover, interactions between the extracellular matrix and its receptors significantly impact the structure and function of the TME29, influencing carcinogenesis30, and tumor development30. M2-like macrophages promote GC metastasis by secreting multiple potent proangiogenic cytokines31. Consequently, FNDC1 may have a significant effect on modulating the TME and extracellular matrix in GC, warranting further investigation. Previous bioinformatic analyses have demonstrated that upregulated FNDC1 expression is significantly associated with poor prognosis in GC32. Overexpression of FNDC1 may promote metastasis and invasion in lung adenocarcinoma through activation of the JAK2-STAT3 signaling pathway8. Furthermore, emerging evidence suggests that FNDC1 could enhance proliferation, invasion, and migration of breast cancer cells by modulating the Wnt/β-catenin pathway32. The study revealed significant associations between FNDC1 expression and GC features such as pathological T stage, pathological stage, histological grade, race, histological type, and anatomic neoplasm subdivision. High FNDC1 expression was strongly correlated with poor OS. Overexpression of FNDC1 promoted the invasive and migratory capabilities of gastric cancer cells, whereas knockdown of FNDC1 inhibited these malignant behaviors in vitro. These findings collectively suggest that FNDC1 may serve as a pivotal regulator in gastric tumorigenesis, progression, and differentiation. Studies have demonstrated that inhibition of FNDC1 suppresses gastric cancer progression by interfering with the formation of the Gβγ‑VEGFR2 complex33. However, the precise underlying mechanisms involved, particularly how FNDC1 influences the immune microenvironment of gastric cancer, still require further investigation. Multivariate Cox regression analysis revealed that histological grade, age, FNDC1 expression, and primary therapy outcome were prognostic factors that independently predict the OS of GC patients. From these factors, a prognostic nomogram was established to predict survival probability in patients with GC. The calibration curves demonstrated good predictive accuracy of the established nomogram, suggesting its potential clinical utility for guiding individualized treatment strategies in GC management. Conclusions The main findings of this study can be summarized as follows: First, FNDC1 expression is significantly elevated in GC tissues compared to non-carcinoma samples, suggesting its potential as a diagnostic biomarker for GC. Second, FNDC1 may contribute to tumor progression and prognosis by remodeling the TME and extracellular matrix, highlighting its promise as a therapeutic target. Third, histological grade, age, FNDC1 expression level, and primary therapy outcome appear to function as independent risk factors influencing overall survival in GC patients. Finally, the nomogram developed in this study demonstrated predictive accuracy for estimating patient survival probability. Nonetheless, despite the integration of data from multiple online databases and clinical specimens, the findings require further validation in larger, independent cohorts to confirm their reliability.