Comprehensive pan-cancer analysis of FSCN1 as a marker for prognosis and immunity

Rui Xu; Wanxin Xu; Changqing Dong; Guangliang Qiang

doi:10.1007/s12672-026-04949-7

. 2026 Apr 7;17:763. doi: 10.1007/s12672-026-04949-7

Comprehensive pan-cancer analysis of FSCN1 as a marker for prognosis and immunity

Rui Xu ^1,^#, Wanxin Xu ^2,^#, Changqing Dong ³, Guangliang Qiang ^3,^✉

PMCID: PMC13199569 PMID: 41945244

Abstract

Background

Fascin actin-bundling protein 1 (FSCN1), a member of actin cytoskeletal protein family genes, plays critical roles in cell migration, motility, adhesion, and other cellular interactions. It has shown its ascending importance in tumors of multiple systems, including the nervous system, respiratory system, digestive system and urinary system. Nevertheless, no systematic pan-cancer analysis has been carried out to investigate its function in diagnosis, prognosis, and immunological prediction.

Methods

We used UCSC Xena, The Cancer Genome Atlas (TCGA), Genotype Tissue Expression Project (GTEx), Human Protein Atlas (HPA), cBioPortal, STRING, Cancer single-cell state Atlas (CancerSEA), Genomics of Drug Sensitivity in Cancer (GDSC) and Xiantao Academic analysis tool websites and databases for the extraction of pan-cancer data on FSCN1. We adopted bioinformatics methods to explore the potential role of FSCN1 in pan-cancer, including analysis of the correlation between FSCN1 and clinicopathologic features, prognosis, genetic alteration, immune, DNA methylation, single cell function and drug sensitivity. Furthermore, a protein–protein interaction network was drawn, and gene set enrichment analysis was conducted to explore the functions of FSCN1. Finally, we validated the effects of FSCN1 on cell proliferation and apoptosis through experiments.

Results

The findings of the comprehensive pan-cancer analysis revealed that FSCN1 was highly expressed in most cancers, except in acute myeloid leukemia, prostate adenocarcinoma and thyroid carcinoma, and it has a certain diagnostic and prognosis value in many cancers. Furthermore, FSCN1 expression was found to correlate with immune cell infiltration, immune checkpoint (ICP) genes expression, DNA methylation and drug sensitivity in a variety of cancers. In addition, we discovered that FSCN1 participated in the proliferation and DNA repair at the single-cell level. Besides, Gene function enrichment analyses revealed that genes that were co-expressed with FSCN1 were enriched within the actin filament regulation pathways. CCK-8 assay and flow cytometry apoptosis assay results indicated that FSCN1 knockdown inhibited proliferation and induced apoptosis in mesothelioma (MESO).

Conclusion

FSCN1 is a prospective marker for the diagnosis, prognosis, immunology, chemotherapy, and small molecular drugs targeted for pan-cancer treatment.

Keywords: FSCN1, Pan-cancer, Diagnosis, Prognosis, Immunization

Introduction

Fascin actin-bundling protein 1 (FSCN1), also called fascin or fascin-1, is a globular, filamentous, actin-binding protein, which belongs to the actin cytoskeletal protein family [1]. FSCN1 contributes to tissue function through two different actin-based structures—dynamic cortical cell protrusions and cytoplasmic microfilament bundles. Filopodia, spikes, lamellipodial ribs, oocyte microvilli, and the dendrites on dendritic cells are examples of cortical structures, which are involved in cell–matrix adhesion, cell interactions and cell migration. The cytoplasmic actin bundles, on the other hand, seem to have a role in cell architecture [2, 3].

Previous studies have discovered that FSCN1 is abnormally expressed in a variety of human carcinomas. It is present in all grades of glioblastoma, where its function appears to be related to cell motility, invasion, and immune response [4–6]. Quantitative RT-PCR and western blot analysis demonstrated that FSCN1 expression was markedly up-regulated in laryngeal squamous cell carcinoma and tongue squamous cell carcinoma in comparison with adjacent normal mucosal tissues [7, 8]. Compared to normal para-carcinoma tissue, FSCN1 exhibits differential expression in non-small cell lung cancer(NSCLC) tissue and correlates with poor prognosis in NSCLC patients [9]. FSCN1 expression is significantly elevated in esophageal squamous cell carcinoma and gastric adenocarcinoma, with a substantial correlation to histological grade and prognosis [10–13]. In hepatocellular carcinoma, FSCN1-positive tumors exhibit greater size, less differentiation, and an increased propensity for hematogenous metastasis compared to FSCN1-negative tumors [14]. FSCN1 expression is also increased in renal cell carcinoma and bladder cancer [15, 16], and its elevated level correlates with the aggressiveness of patients [17].

Although the studies mentioned above have shown the role of FSCN1 in some cancers, the potential value of FSCN1 for other cancers, including mesothelioma, remains unclear. In the present study, we conducted a comprehensive analysis to examine the role of FSCN1 in human cancer prognosis and immunology. We also investigated the relationship between FSCN1 and immune infiltration, DNA methylation, immune checkpoint genes, genomic alterations, potentially regulatory pathways and single cell function. Ultimately, we investigated effective small molecule drugs through analysing the relationship between drug sensitivity and FSCN1 expression.

Methods

Data and software availability

All originate data downloaded from the Cancer Genome Atlas (TCGA) (https://cancergenome.nih.

gov/) database and and the Genotype-Tissue Expression (GTEx) database. TCGA consists of more than 20,000 primary cancer samples over 33 cancer types [18]. We used R 4.4.1 and GraphPad Prism 8.0.2 for statistical analysis and graphing.

Gene expression and subcellular localization analysis of FSCN1 in pan-cancer

We downloaded RNA-seq data from The Cancer Genome Atlas (TCGA) database and the Genotype-Tissue Expression (GTEx) database by UCSC XENA (https://xenabrowser.net/datapages/). It should be pointed out that the data from two different databases, TCGA and GTEx, were integrated by the UCSC team and batch effects were removed [19]. The log2 (TPM + 0.001) transformed normalized expression profiles. Xiantao Academic (https://www.xiantaozi.com/) online analysis tools was used to analyze the expression of FSCN1 in tumor and normal tissues. Statistical analysis was performed using the Wilcoxon rank-sum test, and significant outcomes were defined at p < 0.05 (ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01; ***, p < 0.001). We used the immunofluorescence staining images of three human cancer cell lines (U251MG, A-431, and U2OS) to display the subcellular localization of FSCN1 in cells from the Human Protein Atlas (HPA) database (https://www.proteinatlas.org/) [20]. The abbreviations of the 33 tumors are shown in Table 1.

Table 1.

Pan-cancers and the corresponding abbreviations

Cancer type	Abbreviation
Adrenocortical carcinoma	ACC
Bladder urothelial carcinoma	BLCA
Breast invasive carcinoma	BRCA
Cervical squamous cell carcinoma and endocervical adenocarcinoma	CESC
Cholangiocarcinoma	CHOL
Colon adenocarcinoma	COAD
Lymphoid neoplasm diffuse large B-cell lymphoma	DLBC
Esophageal carcinoma	ESCA
Glioblastoma multiforme	GBM
Head and neck squamous cell carcinoma	HNSC
Kidney chromophobe	KICH
Kidney renal clear cell carcinoma	KIRC
Kidney renal papillary cell carcinoma	KIRP
Acute myeloid leukemia	LAML
Brain lower grade glioma	LGG
Liver hepatocellular carcinoma	LIHC
Lung adenocarcinoma	LUAD
Lung squamous cell carcinoma	LUSC
Mesothelioma	MESO
Ovarian serous cystadenocarcinoma	OV
Pancreatic adenocarcinoma	PAAD
Pheochromocytoma and paraganglioma	PCPG
Prostate adenocarcinoma	PRAD
Rectum adenocarcinoma	READ
Sarcoma	SARC
Skin cutaneous melanoma	SKCM
Stomach adenocarcinoma	STAD
Testicular germ cell tumors	TGCT
Thyroid carcinoma	THCA
Thymoma	THYM
Uterine corpus endometrial carcinoma	UCEC
Uterine carcinosarcoma	UCS
Uveal melanoma	UVM

Open in a new tab

Relationship between FSCN1 and clinicopathologic features

The Wilcoxon test was used to explore the correlation between FSCN1 expression and clinicopathological features, including T stage, N stage and pathologic stage. Correlation analyses were completed using the Xiantao Academic (https://www.xiantaozi.com/) online analysis tool.

Prognostic analysis of FSCN1 in pan-cancer

The survival data of pan-cancer, including overall survival (OS), disease-specific survival (DSS) and progression-free interval (PFI), was downloaded from TCGA database (https://portal.gdc.cancer.gov/) for evaluating the prognostic significance of FSCN1. Patients were categorized into high and low FSCN1 groups based on the median FSCN1 expression value with 50% cut-off high and 50% cut-off low. Kaplan–Meier survival analysis and Cox regression were performed to explore the correlation between FSCN1 and survival prognosis. Correlation analyses were completed using the Xiantao Academic (https://www.xiantaozi.com/) online analysis tool.

Immune-related analysis of FSCN1

The Xiantao Academic (https://www.xiantaozi.com/) online analysis tool was used to analyze infiltration ratio of 24 types of immune cells based on ssGSEA algorithm and estimate algorithm. Spearman correction was used to analyze the correlation between FSCN1 and the enrichment scores of 24 types of immune cells. Furthermore, 11 immune checkpoint genes (including BTLA, CTLA4, HAVCR2, LAG3, LILRB2, LILRB4, PDCD1, SIGLEC7, SIRPA, TIGIT and VSIR) [21] were extracted from TCGA datasets for immune checkpoint gene correlation analysis.

Genetic alteration analysis of FSCN1

The cBioPortal platform (http://www.cbioportal.org/) was utilized to analyze and visualize cancer genetic data, aiding in the interpretation of molecular profiles derived from histologic and cytologic studies [22]. We retrieved gene alteration data for the "TCGA Pan-Cancer Atlas Studies" from the UCSC Xena and International Cancer Genome Consortium (ICGC) (https://www.icgc-argo.org) portals for subsequent analysis. Within cBioPortal, the "Cancer Types Summary" module was specifically used to explore the mutation landscape of the FSCN1 gene, detailing its mutation type, copy number alteration (CNA), and mutation frequency data.

DNA methylation analysis of FSCN1

Aberrant DNA methylation is a common form of epigenetic modifications in tumors [23]. Shiny Methylation Analysis Resource Tool (SMART, http://www.bioinfo-zs.com/smartapp/) was applied to discuss the distribution of methylation probes in chromosomes [24]. In addition, we explore the correlation between FSCN1 expression and DNA methylation levels using Spearman correlation analysis. DNA methylation profiles and phenotype data of pan-cancer was downloaded from UCSC Xena and TCGA. This analysis was conducted and displayed by the R packages psych and ggplot2.

Protein–protein interaction network and genomes enrichment analysis of FSCN1

The protein–protein interaction (PPI) networks of FSCN1 were acquired through the STRING database (https://cn.string-db.org/) [25]. We input FSCN1 in the Protein by name, and basic settings were as follows: the minimum required interaction score (medium confidence (0.400)), and max number of interactors to show (custom value (100)). Thereafter, we utilized genes that interact with FSCN1 which were obtained from the STRING database to perform enrichment analysis. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses of the proteins were conducted using the R package clusterProfiler [26] and results were visualized in the Cytoscape software (version 3.10.1) and R package ggplot2.

Molecular docking

In this study, the HDOCK server was employed for protein–protein molecular docking [27–29]. HDOCK incorporates a hybrid docking strategy that combines template-based docking and template-free docking [29], enabling efficient prediction of complex conformations even when structural templates are limited or unavailable. During the docking process, the server generated 100 candidate binding conformations, which were ranked according to a comprehensive score. The conformation with the lowest docking score was selected as the final binding model, as a more negative docking score typically indicates a more stable and plausible binding mode. The protein structures of interest were obtained from the UniProt database (https://www.uniprot.org/) [30]. After docking, the three-dimensional (3D) model of the complex was visualized and analyzed using PyMOL (Version 3.0.3) [31], while two-dimensional (2D) interactions were identified and illustrated using LigPlot + [32]. PyMOL primarily identifies and displays hydrogen bonds and conformations in 3D space based on geometric algorithms, whereas LigPlot + generates 2D interaction diagrams using built-in geometric and statistical rules. These two tools complement each other in their representation methods. The docking scores in HDOCK are calculated based on the knowledge-based iterative scoring functions ITScorePP or ITScorePR [33, 34]. Specifically, a more negative docking score indicates a higher likelihood of complex formation. Generally, docking scores for protein–protein or protein–DNA/RNA complexes are approximately –200. Therefore, comparative analysis of lower scores facilitates the identification of the most reasonable binding mode.

Single-cell functional analysis of FSCN1

The Cancer single-cell state Atlas (CancerSEA, http://biocc.hrbmu.edu.cn/CancerSEA/) is an analytic tool for studying cancer cell functions at the single-cell level, containing 14 tumor-related cellular functions of 900 cancer cells from 25 cancers [35]. We obtained correlation data of FSCN1 across 14 functional states in distinct cancers and used the Xiantao Academic (https://www.xiantaozi.com/) online analysis tool to visualize it.

Drug sensitivity analysis

We investigated the relationship between gene expression and drug sensitivity using cancer cell lines data from the Genomics of Drug Sensitivity in Cancer (GDSC) database via the GSCA platform (http://bioinfo.life.hust.edu.cn/GSCA/#/drug) [36]. The correlation between FSCN1 mRNA expression levels and drug IC50 values was assessed using Pearson correlation analysis. Subsequently, leveraging the GDSC database, we predicted the IC50 values for clinically relevant and statistically significant pharmacotherapeutic agents in each lung adenocarcinoma(LUAD) sample from the TCGA database. Differences in IC50 between the low-FSCN1 and high-FSCN1 expression groups were evaluated with the Wilcoxon signed-rank test, and results were visualized in box plots generated by the R packages oncoPredict [37] and ggplot2.

Proliferation and apoptosis analysis of FSCN1

We further investigated the role of FSCN1 knockdown in MESO through vitro cell experiments.

Cell and cell line culture

Human healthy mesothelial cell line MET-5A and MESO cell line NCI-H2452 were obtained from Pricella Life Science&Technology Co.,Ltd, and cultured in M199 medium supplemented with 10% fetal bovine serum and 1% penicillin–streptomycin at 37℃, with saturated humidity at 5% CO₂.

Small interfering RNA (siRNA) transfection

Cells were transfected with three siRNAs targeting FSCN1 and with a non-targeting scrambled control si-NC. The sequences (5'-3') were as follows: si-FSCN1 #1: forward GCAAGUUUGUGACCUCCAA, reverse UUGGAGGUCACAAACUUGC. si-FSCN1 #2: forward UCGAGUUCUGCGACUAUAACA, reverse UGUUAUAGUCGCAGAACUCGA. si-FSCN1 #3: forward GCUGCUACUUUGACAUCGAGU, reverse ACUCGAUGUCAAAGUAGCAGC (Baiouni biotechnology). The cells were transfected with siRNAs at a final concentration at 50 nM at 37℃.

Cell Counting Kit-8 (CCK-8) assay and flow cytometry apoptosis assay

To examine the effect of FSCN1 on the proliferative capacity of MESO cells, we used CCK-8 (C0039, Beyotime) assay. Cells from each group were seeded into 96-well plate at a certain density. After 24 h of adhesion in a 37℃, 5% CO2 incubator, 100 µL of prepared CCK-8 solution was added to each well, and the plate was incubated further. After 1 h, the absorbance (OD value) at 450 nm was measured using a microplate reader.

Following transfection, cells were digested with 0.25% trypsin, collected, and resuspended in a 5 mL flow tube. After adding 5 µL of Annexin V-FITC and mixing thoroughly, the cells were incubated in the dark at room temperature for 5 min. Then, The proportion of apoptosis was detected by using DAPI (Beyotime) and Annexin V (elabscience).

Statistical analysis

Wilcoxon rank-sum test was applied to study FSCN1 expression and its correlation with clinical features. Spearman correlation analysis estimated the correlation of FSCN1 with DNA methylation levels. Pearson correlation analysis was performed to study the correlation between FSCN1 and drug IC50. Follow-up correlation analyses were performed using a variety of online analysis websites. All quantitative data were presented as mean ± standard deviation (SD) from at least three times experiments and independent sample t-tests were used for comparisons between groups. For comparisons involving three or more groups, one-way ANOVA followed by Dunnett’s multiple comparison test was applied, with P < 0.05 indicated statistical significance.

Results

Gene expression and subcellular localization analysis of FSCN1 in pan-cancer

We used the Xiantao Academic online analysis tool to analyze the expression levels of FSCN1 in pan-cancer and normal tissues. Based on TCGA data, the results showed that FSCN1 expression was up-regulated in 12 cancers, including BLCA, CHOL, COAD, ESCA, HNSC, KIRC, KIRP, LIHC, LUAD, LUSC, READ, and STAD, while it was down-regulated in KICH (Fig. 1A). For TCGA paired tumor samples and corresponding normal samples, FSCN1 expression was significantly up-regulated in 10 cancer types, including BLCA, CHOL, COAD, HNSC, KIRC, LIHC, LUAD, LUSC, READ, and STAD, while it was down-regulated in KICH (Fig. 1C). In the meantime, the data of tumor and normal samples from TCGA and GTEx database showed that FSCN1 expression was significantly up-regulated in 24 cancer types, including ACC, BLCA, BRCA, CESC, CHOL, COAD, DLBC, ESCA, GBM, HNSC, KIRC, KIRP, LGG, LIHC, LIAD, LUSC, OV, PAAD, READ, STAD, TGCT, THYM, UCEC, and UCS, while it was down-regulated in 3 cancer types, including LAML, PRAD and THCA (Fig. 1B).

Fig. 1 — Expression and subcellular localization of FSCN1. A FSCN1 expression in cancers in TCGA. B FSCN1 expression in cancers in TCGA + GTEx. C FSCN1 expression between tumor and paired normal samples in TCGA. D Immunofluorescence staining of the subcellular localization of FSCN1 in HPA database. (*p < 0.05; **p < 0.01; ***p < 0.001, ns: no statistical differences)

The subcellular localization of FSCN1 was obtained by immunofluorescence localization of nucleus and microtubules in U-251 MG, A-431 and U-2OS cells. The results showed that FSCN1 was found in the plasma membrane and cytosol in all 3 cells (Fig. 1D).

Relationship between FSCN1 and clinicopathologic features

Subsequently, we explored the association between FSCN1 expression and clinicopathologic features. For pathologic stage, FSCN1 had a higher expression in the higher pathologic stage for ACC, LUAD, THCA and TGCT, while a lower expression in the higher pathologic stage for ESCA (Fig. 2A). For the T stage, FSCN1 had a higher expression in the higher T stage for ACC, CESC, COAD, PRAD and TGCT, and a lower expression in the higher T stage for THCA (Fig. 2B). Besides, FSCN1 expression in ACC, KIRC, LUAD and PRAD was higher in patients with lymph node metastasis than without lymph node metastasis. However, the FSCN1 expression level in BLCA, BRCA, ESCA, MESO and THCA was lower in patients with lymph node metastasis than in patients without lymph node metastasis (Fig. 2C).

Fig. 2 — Relationship between FSCN1 and clinicopathologic features. A The relationship of FSCN1 expression to pathologic stage in ACC, ESCA, LUAD, THCA, TGCT. B The relationship of FSCN1 expression to T stage in ACC, CESC, COAD, PRAD, TGCT, THCA. C The relationship of FSCN1 expression to N stage in ACC, BLCA, BRCA, ESCA, KIRC, LUAD, MESO, PRAD, THCA. (*p < 0.05; **p < 0.01; ***p < 0.001)

Prognostic value of FSCN1

Aiming to investigate the association between FSCN1 expression level and prognosis, we performed a survival association analysis for each cancer in the TCGA database, concentrating on overall survival (OS), disease specific survival (DSS) and progress free interval (PFI). Cox proportional hazards model analysis illustrated that the expression level of FSCN1 was associated with OS in ACC (p < 0.001, HR > 1), BLCA (p = 0.023, HR > 1), CESC (p = 0.043, HR > 1), COAD (p = 0.032, HR > 1), HNSC (p = 0.001, HR > 1), LUAD (p < 0.001, HR > 1) and MESO (p < 0.001, HR > 1), indicating that FSCN1 was a significant risk factor for OS in these tumor types (Fig. 3A). Kaplan–Meier survival analysis also demonstrated that among patients with ACC, BLCA, CESC, COAD, HNSC, LUAD and MESO, low FSCN1 expression was associated with better OS (Fig. 3B).

Fig. 3 — The Cox regression and Kaplan–Meier curves for overall survival (OS). A Forest plot of OS associations in pan-cancer. B Kaplan -Meier analysis of the association between FSCN expression and OS in ACC, BLCA, CESC, COAD, HNSC, LUAD, MESO

Moreover, DSS data analysis presented in Fig. 4 reflected associations between high FSCN1 expression and poor prognosis in patients with ACC (p < 0.001, HR > 1), BLCA (p = 0.017, HR > 1), CESC (p = 0.02, HR > 1), COAD (p = 0.016, HR > 1), HNSC (p = 0.009, HR > 1), LUAD (p = 0.049, HR > 1) and MESO (p < 0.001, HR > 1).

Referring to associations between FSCN1 expression and PFI, high expression of FSCN1 was associated with poor PFI in ACC (p < 0.001, HR > 1), COAD (p = 0.007, HR > 1), LGG (p = 0.006, HR > 1), MESO (p = 0.016, HR > 1), PAAD (p = 0.002, HR > 1), PRAD (p < 0.001, HR > 1), TGCT (p = 0.033, HR > 1) and UVM (p = 0.041, HR > 1) (Fig. 5).

Fig. 5 — The Cox regression and Kaplan–Meier curves for progression-free interval (PFI). A Forest plot of PFI associations in pan-cancer. B Kaplan -Meier analysis of the association between FSCN expression and PFI in ACC, COAD, LGG, MESO, PAAD, PRAD, TGCT, UVM

Considering that differences in some clinicopathologic features between groups may affect the results, we included factors with statistical differences (p < 0.1) in the univariate regression analysis into the multivariate regression analysis, as results shown in Fig. 6. For OS and DSS, high FSCN1 expression was a poor prognostic factor in ACC and MESO. For PFI, high FSCN1 expression was a poor prognostic factor in several tumors, including ACC, COAD, MESO, PAAD and TGCT. These findings suggested that FSCN1 may be an independent prognostic marker for multiple tumor types. More detailed information about univariate and multivariate COX regression analysis were exhibited in Supplementary Table S1-S22.

Fig. 6 — The multivariate COX regression analysis for OS (A), DSS (B), PFI (C)